Ventilation Model and Analysis Report Rev 04, ICN 00 ANL-EBS-MD-000030 October 2004 1. PURPOSE 1.1 BACKGROUND Yucca Mountain, approximately 100 miles northwest of Las Vegas, Nevada, has been selected as the site for the nation’s geologic repository for high-level nuclear waste. The Yucca Mountain Project (YMP) is currently developing the design for the underground facilities. The design includes a network of parallel drifts that will hold the waste (emplacement drifts), branching from a main drift. There are two distinct phases considered in the repository operation. The first phase, or preclosure phase, which includes emplacement of the waste, is a period when heat generated from the decay of radionuclides contained in waste packages is actively removed from the repository by ventilating the emplacement drifts. In the second phase, or postclosure phase, forced ventilation of the drifts is stopped and the repository is closed. A prerequisite for designing the YMP repository is the ability to both understand and control the heat generated from the decay of the radionuclides. The decay heat affects the performance of both the waste packages and the emplacement drift. During the preclosure period, heat transfer from the waste packages occurs through mixed convection (a combination of forced and natural convection), conduction through the waste package supports, and thermal radiation to the invert and drift walls. In the postclosure phase, heat is transferred from the waste package by natural convection (as opposed to mixed convection before closure), conduction, and thermal radiation. The purpose of the ventilation model, described in this report, is to simulate the heat transfer processes in and around a waste emplacement drift and predict the heat removal by ventilation during the preclosure period. The heat removal by ventilation is temporally and spatially dependent, and is expressed as the fraction or percentage of the heat produced by radionuclide decay that is carried away by the ventilation air. The heat removal by ventilation is also referred to as the ventilation efficiency. 1.2 SCOPE This document describes the ventilation model. Technical Work Plan for: Near-Field Environment and Transport In-Drift Heat and Mass Transfer Model and Analysis Reports Integration (BSC 2004 [DIRS 170950]) describes work performed for this revision. Sections 4.2 and 8.3 of this report discuss acceptance criteria that were not assigned by the technical work plan (TWP). Otherwise, the work presented in this report is consistent with, and contains no deviations from, the governing TWP. The objectives of this model report are to: 1. Develop and validate a conceptual model for preclosure ventilation of an emplacement drift (Sections 6.3 and 7). 2. Implement the ventilation conceptual model using numerical and analytical methods, and use the License Application design basis inputs and parameters to predict the preclosure ventilation efficiency (Section 6.4). Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 1-2 October 2004 3. Verify the results of the numerical and analytical implementations of the ventilation conceptual model through comparative analyses (Section 6.6). 4. Develop an alternative conceptual model for preclosure ventilation which includes the impacts of water and water vapor mass transfer on the heat transfer (Section 6.7). 5. Implement the alternative conceptual model using analytical calculations to assess the impacts of moisture on the ventilation efficiency (Sections 6.8 and 6.9). 6. Demonstrate the applicability of the use of the ventilation efficiency as an abstraction method for downstream postclosure models to account for the preclosure heat removal (Section 6.10). 7. Demonstrate the sensitivity of the ventilation efficiency to discretization and uncertainties in key input parameters associated with the host rock and engineered components including thermal conductivity, matrix and lithophysal porosity, specific heat, emissivity, and convection heat transfer coefficient (Sections 6.6.1 and 6.11). This report conforms to the prescribed outline of AP-SIII.10Q, Models, as described in Table 1-1. Table 1-1. Outline of the Ventilation Model Documentation Section Content 1. PURPOSE Purpose and introduction to the model report. 2. QUALITY ASSURANCE Identifies the applicability of the YMP Quality Assurance program. 3. USE OF SOFTWARE Lists controlled and exempt software used in the development, implementation, and validation of the model. 4. INPUTS Lists data, parameters, and other inputs used in the development, implementation, and validation of the model. Also lists the appropriate criteria, codes, and standards. 5. ASSUMPTIONS Lists assumptions and the rationale for their use in the development, implementation, and validation of the model. 6. MODEL DISCUSSION Describes the conceptual model, the mathematical implementations of the conceptual model and the results, the alternative conceptual model, the mathematical implementation of the alternative conceptual model and the results, the appropriate use of the model output (i.e., ventilation efficiency), and the sensitivity of the ventilation efficiency to uncertainties in key model inputs and parameters. 7. VALIDATION Presents the analyses that validate the conceptual model, which includes corroboration of the engineered barrier system (EBS) Ventilation Test Phase I results with modeling results. 8. CONCLUSIONS Summarizes the modeling activities and describes the appropriate use of the model output (i.e., ventilation efficiency) in downstream models. 9. INPUTS AND REFERENCES Lists input and output data tracking numbers (DTNs) and cited references. APPENDICES Document supporting analyses. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 1-3 October 2004 1.3 LIMITATIONS Applicability of the ventilation model documented in this report is limited to: • Ventilation air flow rates between 10 and 30 m3/s. • Configurations in which the waste packages are spaced in the drift such that, during the preclosure period, the average heat generation per unit length in each small group of waste packages is approximately the same as the average over the entire drift. • Conditions in which conduction heat transfer from the waste package to the invert or drift wall is small compared to the heat transfer to the invert and drift wall via thermal radiation. • Repository average waste package heat loads (or waste streams) that produce sub-boiling temperature conditions in the host rock during the ventilation period. • Single drift analyses where the repository edges do not significantly affect the near field host rock thermal conduction. • Simultaneous emplacement of the waste packages at the start of the preclosure period which is conservative with respect to the total heat load applied to the system. 1.4 DOWNSTREAM USE OF THE RESULTS The main output of the ventilation model is the ventilation efficiency. Downstream models that do not explicitly model the preclosure period rely on the ventilation efficiency as a means of initializing their postclosure analyses. Such models include those presented in Multiscale Thermohydrologic Model, Drift Degradation Analysis, and Drift-Scale Coupled Processes (DST and THC Seepage) Models. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 1-4 October 2004 INTENTIONALLY LEFT BLANK Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 2-1 October 2004 2. QUALITY ASSURANCE This document was prepared in accordance with Technical Work Plan for: Near-Field Environment and Transport In-Drift Heat and Mass Transfer Model and Analysis Reports Integration (BSC 2004 [DIRS 170950]), which directs the work identified in work package ARTM02. As described in the technical work plan (TWP), the Quality Assurance program applies to the development of this document (BSC 2004 [DIRS 170950], Section 8). The methods used to control the electronic management of data as required by AP-SV.1Q, Control of the Electronic Management of Information, were accomplished in accordance with the TWP (BSC 2004 [DIRS 170950], Section 8). There was no variance from the methods for controlling the electronic management of data. As directed in the TWP, this document was prepared in accordance with AP-SIII.10Q, Models, LP-SI.11Q-BSC, Software Management, AP-3.15Q, Managing Technical Product Inputs, and reviewed in accordance with AP-2.14Q, Document Review. This report supports the investigation of the performance of the engineered barrier system. In accordance with the Q-list (BSC 2004 [DIRS 168361], Table A-2, p. A-5), the engineered barrier system is designated as “important to waste isolation,” and the Safety Category (SC) is “SC.” Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 2-2 October 2004 INTENTIONALLY LEFT BLANK Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 3-1 October 2004 3. USE OF SOFTWARE Table 3-1 lists the software used to perform the analyses as well as the software tracking numbers (where appropriate), CPU(s), operating systems, and physical location where the software was installed. All software listed in Table 3-1 was obtained from Software Configuration Management, was appropriate for the applications used, and was used within the range of validation in accordance with LP-SI.11Q-BSC, Software Management. Use of software has been documented in accordance with LP-SI.11Q-BSC. Table 3-1. Software Code Software Tracking Number CPU Physical Location Operating System ANSYS v5.6.2 10145-5.6.2-00 10145-5.6.2-01 SGI Octane Sun Microsystems UltraSPARC Las Vegas, NV IRIX 6.5 Solaris 2.6 and Solaris 2.7 rme6 v1.2 10617-1.2-00 Sun Microsystems Blade 100 Livermore, CA Solaris 8 YMESH v1.54 10172-1.54-00 Sun Microsystems Blade 100 Livermore, CA Solaris 8 Mathcad 2001i Professional Exempt Dell Pentium Workstation Las Vegas, NV Windows 2000 Microsoft Excel 97 Exempt Various YMP M&O Computers Las Vegas, NV Windows 95, Windows 2000 3.1 ANSYS v5.6.2 ANSYS v5.6.2 (BSC 2001 [DIRS 164464]) is a commercially available computer program and is classified as qualified software (per LP-SI.11Q-BSC). ANSYS v5.6.2 is used to implement the ventilation conceptual model. ANSYS v5.6.2 is a general purpose finite element analysis code, and is used in many disciplines of engineering that deal with topics including structural, geotechnical, mechanical, thermal, and fluids. ANSYS was selected for its capability of modeling heat transfer processes and predicting the ventilation efficiency in the thermal model for the License Application design. The use of this software was consistent with the intended use and was within the validation range defined by the test cases for this model (Test 01, Test 02, Test 05, Test 06 and Test 11) identified in the Software Validation Test Report ANSYS Version 5.6.2 Software (CRWMS M&O 2001 [DIRS 155138], pp. 8, 9, 11, 12, 16, 17) and Validation Test Report for ANSYS Version 5.6.2 Software (Doraswamy 2001 [DIRS 171331], pp. 4, 5, 7, 8, 13). There are no known limitations on outputs. 3.2 rme6 v1.2 rme6 v1.2 (LLNL 2003 [DIRS 163892]) is a developed computer program and is classified as qualified software (per LP-SI.11Q-BSC). rme6 v1.2 was selected for its unique capability of converting the numerical grid from the three-dimensional site scale unsaturated zone (UZ) flow and transport model to a format that is readable by YMESH v1.54. The use of this software was consistent with the intended use and was within the validation range defined by the test cases 1, Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 3-2 October 2004 2, 3, and 4 identified in Software Management Report: rme6 Version 1.2 (DOE 2003 [DIRS 171333], pp. 28, 33, 39, 45). Limitations on the software are also specified in Software Management Report: rme6 Version 1.2 (DOE 2003 [DIRS 171333], Section 2.6). There are no known limitations on outputs. Work performed with the rme6 v1.2 software prior to qualification was reperformed with baselined software; the outputs were identical (see Section 6.5.1). 3.3 YMESH v1.54 YMESH v1.54 (LLNL 2003 [DIRS 163894]) is a developed computer program and is classified as qualified software (per LP-SI.11Q-BSC). YMESH v1.54 was selected for its unique capability of generating the thickness of the geologic layers for a stratigraphic column given by some easting and northing coordinates. The use of this software was consistent with the intended use and was within the validation range defined by the test cases 1 through 7 identified in the Software Management Report: ymesh Version 1.54 (DOE 2003 [DIRS 171332], pp. 14 to 29). Limitations on the software are also specified in Software Management Report: ymesh Version 1.54 (DOE 2003 [DIRS 171332], Section 2.6). There are no known limitations on outputs. Work performed with the YMESH v1.54 software prior to qualification was reperformed with baselined software; the outputs were identical (see Section 6.5.1). 3.4 MATHCAD 2001i PROFESSIONAL Mathcad 2001i Professional is a commercially available software package. The Mathcad software provides a technical computing environment using standard mathematical notation for equations and operations. The use of the Mathcad software in this report is exempt from LP-SI.11Q-BSC per AP-SIII.10Q. The formulas, inputs and outputs of those formulas, and additional information required for an independent technically qualified person to verify the results of these Mathcad analyses are provided in Section 6 and Appendices III and XIII. 3.5 MICROSOFT EXCEL 97 Microsoft Excel 97 is a commercially available spreadsheet software package. Excel 97 is used in conjunction with the ANSYS v5.6.2 software to predict the ventilation efficiency, and as a stand alone implementation of the ventilation conceptual model to predict ventilation efficiency. Each of these applications uses only standard or built in functions. It is also used to make plots of data and perform other computations using standard functions. The use of Excel 97 in this report is exempt from LP-SI.11Q-BSC, per AP-SIII.10Q. The formulas, inputs and outputs of those formulas, and additional information required for an independent technically qualified person to verify the results of these Excel analyses are provided in Section 6, Appendices I, II, IV, V, VI, VII, VIII, X, XI, XII, XIV, XV, and the DTNs listed in Table 8-1. The user must select ‘Analysis ToolPak’ from the Tools/Add-Ins menu and must disable macros if prompted. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 4-1 October 2004 4. INPUTS 4.1 DIRECT INPUT The following data were used as direct inputs to develop the models and analyses described in Section 6. Qualification and justification for use of the direct inputs which are obtained from outside sources, as listed in Tables 4-8, 4-14, and 4-17 through 4-21, are provided in Appendix XVIII. 4.1.1 Thermophysical Properties of the Invert Tables 4-1 and 4-2 list measured thermophysical properties of 4-10 crushed tuff. The 4-10 crushed tuff is crushed welded tuff that passes through a size 4 sieve (4.75 mm mesh) and is retained on a size 10 sieve (2.00 mm mesh). The justification for the use of the material properties of 4-10 crushed tuff for the invert ballast material is described in Section 5.5. That is, though the 4-10 crushed tuff properties are not exact, they are adequate for use in the ventilation calculations. Tables 4-1 and 4-2 are for the 4-10 crushed tuff, and are used as inputs to the models and analyses described in Section 6 (see Section 5.5). Table 4-1. Specific Heat, Thermal Conductivity, and Thermal Diffusivity of 4-10 Crushed Tuff Sample Type Sample Number Specific Heat (J/cm3·°C) Thermal Conductivity (W/m·°C) Thermal Diffusivity (mm2/s) Temperature (°C) 4-10 crushed tuff TK-CT-01 0.82 0.17 0.21 16.2 4-10 crushed tuff TK-CT-02 0.84 0.14 0.16 15.8 4-10 crushed tuff TK-CT-03 0.98 0.17 0.17 16.1 4-10 crushed tuff TK-CT-04 0.98 0.17 0.17 16.4 4-10 crushed tuff TK-CT-05 0.99 0.17 0.17 17.1 4-10 crushed tuff TK-CT-06 0.92 0.16 0.18 17.5 4-10 crushed tuff TK-CT-07 0.96 0.17 0.17 17.6 4-10 crushed tuff TK-CT-07a 0.86 0.15 0.18 18.9 4-10 crushed tuff TK-CT-08 0.88 0.16 0.18 18 4-10 crushed tuff TK-CT-09 1.06 0.17 0.16 18.1 4-10 crushed tuff TK-CT-10 0.94 0.17 0.18 18.5 DTN: GS000483351030.003 [DIRS 152932]. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 4-2 October 2004 Table 4-2. Bulk Density of 4-10 Crushed Tuff Sample Type Sample Number Bulk Density (g/cm3) Sample Type Sample Number Bulk Density (g/cm3) 4-10 Crushed Tuff UNCBD41A 1.3 4-10 Crushed Tuff UNCBD53B 1.3 4-10 Crushed Tuff UNCBD41B 1.2 4-10 Crushed Tuff UNCBD54A 1.3 4-10 Crushed Tuff UNCBD42A 1.3 4-10 Crushed Tuff UNCBD54B 1.3 4-10 Crushed Tuff UNCBD42B 1.3 4-10 Crushed Tuff UNCBD55A 1.3 4-10 Crushed Tuff UNCBD43A 1.3 4-10 Crushed Tuff UNCBD55B 1.2 4-10 Crushed Tuff UNCBD43B 1.2 4-10 Crushed Tuff UNCBD56A 1.3 4-10 Crushed Tuff UNCBD44A 1.3 4-10 Crushed Tuff UNCBD56B 1.3 4-10 Crushed Tuff UNCBD44B 1.2 4-10 Crushed Tuff UNCBD57A 1.3 4-10 Crushed Tuff UNCBD45A 1.3 4-10 Crushed Tuff UNCBD57B 1.2 4-10 Crushed Tuff UNCBD45B 1.2 4-10 Crushed Tuff UNCBD58A 1.3 4-10 Crushed Tuff UNCBD46A 1.2 4-10 Crushed Tuff UNCBD58B 1.3 4-10 Crushed Tuff UNCBD46B 1.2 4-10 Crushed Tuff UNCBD59A 1.2 4-10 Crushed Tuff UNCBD47A 1.3 4-10 Crushed Tuff UNCBD59B 1.2 4-10 Crushed Tuff UNCBD47B 1.3 4-10 Crushed Tuff UNCBD60A 1.2 4-10 Crushed Tuff UNCBD48A 1.3 4-10 Crushed Tuff UNCBD60B 1.3 4-10 Crushed Tuff UNCBD48B 1.3 4-10 Crushed Tuff UNCBD61A 1.3 4-10 Crushed Tuff UNCBD49A 1.3 4-10 Crushed Tuff UNCBD61B 1.3 4-10 Crushed Tuff UNCBD49B 1.2 4-10 Crushed Tuff UNCBD62A 1.3 4-10 Crushed Tuff UNCBD50A 1.2 4-10 Crushed Tuff UNCBD62B 1.2 4-10 Crushed Tuff UNCBD50B 1.2 4-10 Crushed Tuff UNCBD63A 1.2 4-10 Crushed Tuff UNCBD51A 1.3 4-10 Crushed Tuff UNCBD63B 1.2 4-10 Crushed Tuff UNCBD51B 1.3 4-10 Crushed Tuff UNCBD64A 1.3 4-10 Crushed Tuff UNCBD52A 1.3 4-10 Crushed Tuff UNCBD64B 1.3 4-10 Crushed Tuff UNCBD52B 1.3 4-10 Crushed Tuff UNCBD65A 1.3 4-10 Crushed Tuff UNCBD53A 1.3 4-10 Crushed Tuff UNCBD65B 1.3 DTN: GS020183351030.001 [DIRS 163107], ROWS 321-370. 4.1.2 Relative Humidity Relative humidity in the ventilated Enhanced Characterization of the Repository Block (ECRB) Cross-Drift was measured during November 1998. The relative humidity ranged from 10 to 41 percent. Appendix XIII requires a single relative humidity to represent average conditions in an open drift. That appendix uses 30% RH, a central value of the measurements taken in the ventilated ECRB, rounded off to one significant digit. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 4-3 October 2004 Table 4-3. Measured Relative Humidity in the Ventilated Zone of the ECRB Cross-Drift Location Range of Relative Humidity Measurements Ventilated ECRB Fluctuated between 10% and 41% DTN: LB990901233124.006 [DIRS 135137]. 4.1.3 Laboratory Measured Saturation for Tptpll from Borehole Cores Table 4-4 lists laboratory-measured values of saturation from borehole core data. Column headings designate area boreholes that pass through the Tptpll (lower lithophysal unit). These data are used in Appendix XIII, and the average of these measurements is referred to in Section 6.9.1. Table 4-4. Laboratory Measured Saturation for Tptpll from Borehole Cores USW SD-7 a USW SD-9 b USW NRG-6 c USW NRG-7/7A d USW UZ-7A e Depth (ft) Sat. Depth (ft) Sat. Depth (ft) Sat. Depth (m) Sat. Depth (m) Sat. 809.2 0.862 847.2 0.852 816.6 0.24 269.1 0.8 184.2 0.606 819 0.904 849.6 0.775 817.9 0.71 271.9 0.84 185.3 0.702 824.7 0.911 853.4 0.843 820.8 0.8 272.8 0.71 186.3 0.669 835.4 0.874 859 0.974 823 0.87 274.1 0.61 188.6 0.636 836.8 0.698 865.3 0.907 826.1 0.63 276.7 0.67 189 0.715 842.5 0.891 879.6 1.02 829.2 0.78 280 0.57 190.7 0.733 847.6 0.862 888.8 0.774 831.7 0.98 285.7 0.71 191 0.635 848.4 0.775 897 0.898 835.4 0.54 287.5 0.64 197.1 0.73 856.9 0.794 899.5 0.854 838.6 0.71 288.3 0.61 198.3 0.76 857.7 0.845 905.8 0.886 841.7 0.39 290.2 0.62 198.9 0.84 862.3 0.863 921.9 0.794 844.8 0.89 291.1 0.63 203.6 0.705 864.9 0.778 924.2 0.717 851.9 0.75 292.1 0.56 205.1 0.818 867.4 0.942 936.1 0.728 854.9 0.83 293.9 0.66 205.4 0.839 872 0.72 938.9 0.812 857.8 0.12 295.7 0.6 206.6 0.779 874.4 0.772 944.6 0.787 861 0.08 296.4 0.57 207 0.803 875.5 0.835 948 0.796 862.7 0.28 297.2 0.56 208.1 0.85 878.8 0.844 954 0.865 865.8 0.64 298.2 0.7 210 0.846 884.2 0.821 958.1 0.776 867.7 0.7 300.3 0.69 210.7 0.844 885 0.879 962.6 0.791 871.5 0.83 301.1 0.78 211.7 0.876 887.6 0.888 968.7 0.837 873.8 0.64 304.8 0.6 212.8 0.749 891 0.843 971.9 0.716 877.6 0.58 306.7 0.97 213.2 0.776 894 0.864 975.5 0.91 879.7 0.77 313 0.46 213.7 0.744 897.3 0.904 981 0.793 886 0.86 314.1 0.55 214.8 0.784 899.5 0.924 984.7 0.742 890.7 0.66 314.9 0.5 215.7 0.678 Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 4-4 October 2004 Table 4-4. Laboratory Measured Saturation for Tptpll from Borehole Cores (Continued) USW SD-7 a USW SD-9 b USW NRG-6 c USW NRG-7/7A d USW UZ-7A e Depth (ft) Sat. Depth (ft) Sat. Depth (ft) Sat. Depth (m) Sat. Depth (m) Sat. 904.9 0.793 986.6 0.744 892.8 0.31 316.9 0.4 216.7 0.751 910.7 0.855 995.7 0.809 898.6 0.72 317.4 0.65 217.5 0.852 914.7 0.854 1003 0.672 901.6 0.75 318.5 0.56 218.2 0.719 916.2 0.902 1007.3 0.781 904.8 0.85 319.4 0.34 219.2 0.706 919.1 0.818 1012.3 0.731 910.7 0.69 322.1 0.72 220.2 0.756 920.4 0.831 1017.2 0.886 912.8 0.8 323.1 0.74 220.9 0.678 924.1 0.847 1023.8 0.89 917.1 0.78 324.9 0.57 221.7 0.814 928.4 0.903 1028.9 0.81 920.4 0.71 326.7 0.72 222.9 0.723 929.7 0.871 1033.1 0.903 928.8 0.8 328.5 0.66 224.4 0.779 932.8 0.798 1035.1 0.922 932 0.72 331.3 0.69 225.4 0.499 936.7 0.781 1038.8 0.95 936 0.67 332.2 0.73 227.7 0.664 940.7 0.903 1041 0.895 942.7 0.81 334.2 0.58 228.7 0.769 941.5 0.879 1044.2 0.874 949.3 0.49 334.9 0.53 230.1 0.762 946.4 0.825 1047.2 0.932 952.5 0.65 336.9 0.63 231.2 0.784 951.2 0.906 1050.2 0.871 955.4 0.75 337.8 0.68 234.2 0.579 954.5 0.847 1053.6 0.985 959 0.71 338.4 0.51 — — 957 0.778 1055.8 0.909 962 0.77 340 0.52 — — 961.4 0.762 1064.8 0.958 968.2 0.69 342.4 0.38 — — 962.5 0.956 1068.1 0.798 970.8 0.65 344 0.79 — — 966.9 0.839 1070.4 0.821 975.1 0.64 346 0.59 — — 968.9 0.881 1076.7 0.92 977 0.82 348 0.53 — — 971.4 0.905 1080.1 0.837 978.9 0.72 348.8 0.57 — — 974.5 0.85 1086.4 0.918 985.1 0.77 353.2 0.48 — — 978.1 0.918 1091.1 0.863 989 0.73 354.3 0.39 — — 981 0.846 1095.4 0.84 991.6 0.75 355 0.52 — — 983.8 0.831 1098.4 0.712 995.6 0.41 357 0.39 — — 986.2 0.965 1101.3 0.757 1004.1 0.71 357.9 0.5 — — 990.2 0.918 1104.1 0.596 1010.2 0.62 358.9 0.38 — — 993.1 0.995 1106.4 0.761 1015.7 0.84 359.6 0.66 — — 994.3 0.985 1110.3 0.729 1018.5 0.88 360.5 0.42 — — 999 0.878 1113.5 0.706 1024.1 0.5 361.5 0.53 — — 1005 0.901 1116 0.749 1033.8 0.41 362.6 0.53 — — 1008.2 0.909 1119.2 0.755 1036 0.62 363.2 0.35 — — 1013.3 0.955 1125.1 0.806 1040.1 0.84 366 0.76 — — 1017.6 0.952 1128.6 0.877 1042.7 0.87 366.9 0.56 — — — — 1133.6 0.799 1049 0.66 367.8 0.7 — — — — 1139.6 0.84 1054.8 0.37 368.9 0.75 — — — — 1142 0.903 1058.3 0.87 370.6 0.68 — — — — 1146.1 0.863 1060.9 0.83 373.2 0.56 — — — — 1149 0.865 1063.5 0.81 — — — — — — 1152.7 0.855 1067 0.52 — — — — Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 4-5 October 2004 Table 4-4. Laboratory Measured Saturation for Tptpll from Borehole Cores (Continued) USW SD-7 a USW SD-9 b USW NRG-6 c USW NRG-7/7A d USW UZ-7A e Depth (ft) Sat. Depth (ft) Sat. Depth (ft) Sat. Depth (m) Sat. Depth (m) Sat. — — 1158.5 0.69 1069.8 0.82 — — — — — — 1161.1 0.864 1076.1 0.68 — — — — — — 1163.8 0.828 1079.1 0.72 — — — — — — 1166.6 0.862 1081.9 0.75 — — — — — — 1170.5 0.813 1084.2 0.77 — — — — — — 1172.8 0.88 1087.1 0.8 — — — — — — 1179 0.868 1090.3 0.86 — — — — — — — — 1096.6 0.8 — — — — a DTN: GS951108312231.009 [DIRS 108984], S96037_007. b DTN: GS950408312231.004 [DIRS 108986], S96021_007. c DTN: GS000508312231.006 [DIRS 153237]. d DTN: GS951108312231.010 [DIRS 108983]. e DTN: GS951108312231.011 [DIRS 108992]. 4.1.4 Water Potential Measurements Taken at the ECRB Station 15+00 Table 4-5 lists measurements of water potential taken at the ECRB Station 15+00 on 7/29/00. This data set was chosen because it was the first set of measurements that appear to be typical of later measurements. These data are used in Appendix XIII to demonstrate that the latent heat of vaporization can be neglected in the calculation of ventilation efficiency. Table 4-5. Water Potential Measurements Taken at the ECRB Station 15+00 Station Distance from Borehole (m) Water Potential (m) ST-1500-0.62 0.62 -259 ST-1500-1.12 1.12 -91 ST-1500-1.69 1.69 -10 ST-1500-2.12 2.12 -24 ST-1500-2.62 2.62 -37 ST-1500-3.12 3.12 -5 ST-1500-3.62 3.62 4 ST-1500-4.12 4.12 -12 ST-1500-4.62 4.62 -14 ST-1500-5.12 5.12 -8 ST-1500-5.62 5.62 -10 DTN: LB0110ECRBH2OP.001 [DIRS 156883], C7-1500.xls, worksheet “wp-2000-plot”, row 4. NOTE: Water potential in the source DTN and in this table follows the convention that a positive pressure results in a positive water potential while a negative capillary pressure results in a negative water potential. Except as presented in this table, this report uses the opposite convention: positive water potential corresponds to negative capillary pressure. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 4-6 October 2004 4.1.5 Thermophysical Properties of the Stratigraphic Layers Tables 4-6 through 4-11 list the thermophysical properties of the repository and non-repository stratigraphic units. Except for emissivity, these properties are obtained from qualified data found in the Technical Data Management System. The emissivity values for rocks (see Table 4-8) are from Fundamentals of Heat and Mass Transfer (Incropera and DeWitt 1996 [DIRS 108184], Table A.11 for rocks). Their range of 0.88 to 0.95 is adapted from sources for hemispherical emissivity of rock at 300K. The range is corroborated by handbook values (Knudsen et al. 1984 [DIRS 170057], Table 10-17, pp. 10-51 to 10-52) for normal emissivity of rough silica and rough fused quartz, ranging from 0.8 to 0.93. Therefore, the data are qualified for use as emissivity of the repository stratigraphaphic units and the invert material (see Section 4.1.15) in the calculation of ventilation efficiency. Parameter distributions are only included for the repository stratigraphic units. These parameters are used as inputs to the models and analyses described in Section 6. Table 4-6. Thermophysical Properties of the Repository Stratigraphic Units Dry Bulk Thermal Conductivity (W/m·K) Wet Bulk Thermal Conductivity (W/m·K) Dry Bulk Density (g/cc) Matrix Porosity Lithophysal Porosity Unit (UZ Model Layer) Mean Std. Dev. Mean Std. Dev. Mean Std. Dev. Mean Std. Dev. Mean Std. Dev. Tptpul (tsw33) 1.1829 0.2440 1.7749 0.2474 1.8344 0.1496 0.1667 0.0412 0.1228 0.0613 Tptpmn (tsw34) 1.4189 0.2654 2.0741 0.2517 2.1483 0.0932 0.1287 0.0323 0.0254 0.0225 Tptpll (tsw35) 1.2784 0.2511 1.8895 0.2484 1.9793 0.1381 0.1486 0.0340 0.0883 0.0540 Tptpln (tsw36) 1.4900 0.2844 2.1303 0.2676 2.2114 0.0857 0.1058 0.0264 0.0302 0.0253 Dry Matrix Thermal Conductivity (W/m·K) Wet Matrix Thermal Conductivity (W/m·K) Solid Thermal Conductivity (W/m·K) Solid Connectivity Unit (UZ Model Layer) Mean Std. Dev. Mean Std. Dev. Mean Std. Dev. Mean Std. Dev. Tptpul (tsw33) 1.3453 0.2639 2.0201 0.2484 2.6011 0.3493 0.8517 0.1158 Tptpmn (tsw34) 1.4553 0.2690 2.1276 0.2519 2.6033 0.3518 0.8476 0.1094 Tptpll (tsw35) 1.3998 0.2640 2.0707 0.2455 2.6030 0.3413 0.8531 0.1130 Tptpln (tsw36) 1.5356 0.2908 2.1958 0.2764 2.6017 0.3505 0.8492 0.1151 DTN: SN0404T0503102.011 [DIRS 169129], file ReadMe.Doc, Tables 7-10 and 7-11. NOTE: Nomenclature correlation between stratigraphic units and UZ model layer is based on BSC 2004 [DIRS 169855], Table 6-11. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 4-7 October 2004 Table 4-7. Specific Heat of the Repository Stratigraphic Units Average Rock Grain Specific Heat (J/g·K) Unit UZ Model Layer Mean Std. Dev. Tptpul tsw33 0.93 0.12 Tptpmn tsw34 0.93 0.14 Tptpll tsw35 0.93 0.13 Tptpln tsw36 0.93 0.10 DTN: SN0307T0510902.003 [DIRS 164196], file rock_grain_heat_capacity.xls, worksheet “Cp grain 25-325”, rows 8- 11, columns y and z. NOTES: T = 25 to 325°C. Nomenclature correlation between stratigraphic units and UZ model layer is based on BSC 2004 [DIRS 169855], Table 6-11. Table 4-8. Emissivity of Rocks Used as Inputs for the Repository Stratigraphic Units Emissivity Unit UZ Model Layer Minimum Maximum Tptpul tsw33 0.88 0.95 Tptpmn tsw34 0.88 0.95 Tptpll tsw35 0.88 0.95 Tptpln tsw36 0.88 0.95 Source: Incropera and DeWitt 1996 [DIRS 108184], Table A.11 for Rocks. NOTE: Nomenclature correlation between stratigraphic units and UZ model layer is based on BSC 2004 [DIRS 169855], Table 6-11. Table 4-9. Matrix Permeability and Van Genuchten Parameters of the Repository Stratigraphic Units Unit UZ Model Layer Permeability (m2) Residual Saturation a (1/Pa) Van Genuchten’s m Tptpul tsw33 (tswM3) 6.57e-18 0.12 6.17e-6 0.283 Tptpmn tsw34 (tswM4) 1.77e-19 0.19 8.45e-6 0.317 Tptpll tsw35 (tswM5) 4.48e-18 0.12 1.08e-5 0.216 Tptpln tsw36 (tswM6) 2.00e-19 0.20 8.32e-6 0.442 DTN: LB0208UZDSCPMI.002 [DIRS 161243], drift-scale calibrated properties for mean infiltration2.xls. NOTE: Nomenclature correlation between stratigraphic units and UZ model layer is based on BSC 2004 [DIRS 169855], Table 6-11. (In the source spreadsheet, one character of the UZ model is changed to “M” for matrix properties or “F” for fracture properties.) Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 4-8 October 2004 Table 4-10. Thermophysical Properties of the Non-Repository Stratigraphic Units Dry Matrix Thermal Conductivity (W/m·K) Wet Matrix Thermal Conductivity (W/m·K) Dry Bulk Density (kg/m3) Matrix Porosity Unit UZ Model Layer Mean Std. Dev. Mean Std. Dev. Mean Std. Dev. Mean Std. Dev. Tpcr tcw11 1.30 0.23 1.81 0.20 2190 177 0.12 0.05 Tpcp 1.30 0.23 1.81 0.20 2190 177 0.12 0.05 TpcLD tcw12 1.30 0.23 1.81 0.20 2190 177 0.12 0.05 Tpcpv3 0.69 0.23 0.80 0.25 2310 89 0.04 0.04 Tpcpv2 tcw13 0.49 0.16 1.06 0.15 1460 337 0.39 0.13 Tpcpv1 ptn21 0.49 0.16 1.06 0.15 1460 337 0.39 0.13 Tpbt4 0.49 0.16 1.06 0.15 1460 337 0.39 0.13 ptn22 ptn23 Yucca 0.49 0.16 1.06 0.15 1460 337 0.39 0.13 Tpbt3_dc ptn24 0.49 0.16 1.06 0.15 1460 337 0.39 0.13 Pah ptn25 0.49 0.16 1.06 0.15 1460 337 0.39 0.13 Tpbt2 0.49 0.16 1.06 0.15 1460 337 0.39 0.13 Tptrv3 0.49 0.16 1.06 0.15 1460 337 0.39 0.13 Tptrv2 ptn26 0.49 0.16 1.06 0.15 1460 337 0.39 0.13 Tptrv1 0.69 0.23 0.80 0.25 2310 89 0.04 0.04 tsw31 Tptrn tsw32 1.30 0.23 1.81 0.20 2190 177 0.12 0.05 Tptrl 1.30 0.23 1.81 0.20 2190 177 0.12 0.05 Tptf tsw33 1.30 0.23 1.81 0.20 2190 177 0.12 0.05 Tptpv3 tsw38 0.69 0.23 0.80 0.25 2310 89 0.04 0.04 Tptpv2 tsw39 0.49 0.16 1.06 0.15 1460 337 0.39 0.13 Tptpv1 0.49 0.16 1.06 0.15 1460 337 0.39 0.13 Tpbt1 ch1 0.49 0.16 1.06 0.15 1460 337 0.39 0.13 ch2 ch3 ch4 Calico ch5 0.60 0.11 1.26 0.14 1670 157 0.33 0.05 Calicobt ch6 0.60 0.11 1.26 0.14 1670 157 0.33 0.05 Prowuv pp4 0.57 0.10 1.13 0.12 1790 117 0.30 0.04 Prowuc pp3 0.57 0.10 1.13 0.12 1790 117 0.30 0.04 Prowmd 1.06 0.18 1.63 0.17 2070 139 0.21 0.06 Prowlc pp2 0.57 0.10 1.13 0.12 1790 117 0.30 0.04 Prowlv 0.57 0.10 1.13 0.12 1790 117 0.30 0.04 Prowbt 0.57 0.10 1.13 0.12 1790 117 0.30 0.04 Bullfroguv pp1 0.66 0.13 1.19 0.14 1880 167 0.23 0.06 Bullfroguc 0.66 0.13 1.19 0.14 1880 167 0.23 0.06 Bullfrogmd 1.30 0.24 1.81 0.20 2260 138 0.12 0.05 Bullfroglc bf3 0.66 0.13 1.19 0.14 1880 167 0.23 0.06 Bullfroglv 0.66 0.13 1.19 0.14 1880 167 0.23 0.06 Bullfrogbt 0.66 0.13 1.19 0.14 1880 167 0.23 0.06 Tramuv bf2 0.54 0.11 1.10 0.12 1760 195 0.33 0.06 Tramuc tr3 0.54 0.11 1.10 0.12 1760 195 0.33 0.06 Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 4-9 October 2004 Table 4-10. Thermophysical Properties of the Non-Repository Stratigraphic Units (Continued) Dry Matrix Thermal Conductivity (W/m·K) Wet Matrix Thermal Conductivity (W/m·K) Dry Bulk Density (kg/m3) Matrix Porosity Unit UZ Model Layer Mean Std. Dev. Mean Std. Dev. Mean Std. Dev. Mean Std. Dev. Trammd 1.06 0.18 1.63 0.17 2140 78 0.21 0.06 Tramlc tr3 0.54 0.11 1.10 0.12 1760 195 0.33 0.06 Tramlv 0.54 0.11 1.10 0.12 1760 195 0.33 0.06 Trambt tr2 0.54 0.11 1.10 0.12 1760 195 0.33 0.06 DTN: SN0303T0503102.008 [DIRS 162401]. NOTE: Nomenclature correlation between stratigraphic units and UZ model layer is based on BSC 2004 [DIRS 169855], Table 6-11. Table 4-11. Specific Heat of the Non-Repository Stratigraphic Units Average Rock Grain Specific Heat (J/g·K) Unit UZ Model Layer Mean Std. Dev. Tpc tcw11, tcw12 0.93 0.11 Tpcpv23 tcw13 0.95 0.11 pTn ptn21 to ptn26 0.96 0.23 Tptrv1 tsw31 0.95 0.10 Tptrnf tsw32 0.93 0.13 Tptpv3 tsw38 0.98 0.24 Tptpv2 tsw39 0.98 0.19 Tptpv1-Tpbt1 ch1 1.08 0.42 Tac4 ch2 1.07 0.42 Tac3 ch3 1.07 0.38 Tac2 ch4 1.07 0.36 Tac1 ch5 1.07 0.35 Tacbt ch6 1.02 0.24 Tcpuv pp4 1.04 0.28 Tcpuc-Tcplc pp3, pp2 0.93 0.13 Tcplv-Tcbuv pp1 1.10 0.19 Tcbuc-Tcblc bf3 0.93 0.12 Tcblv-Tctuv Bf2 1.05 0.22 Tctuc-Tctlc tr3 0.94 0.12 Tctlv-Tctbt tr2 0.94 0.12 DTN: SN0307T0510902.003 [DIRS 164196]. NOTES: T = 25 to 325°C. Nomenclature correlation between stratigraphic units and UZ model layer is based on BSC 2004 [DIRS 169855], Table 6-11. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 4-10 October 2004 4.1.6 Ground Surface and Water Table Elevations and Temperatures UZ Flow Models and Submodels (BSC 2004 [DIRS 169861], pp. 6-36 to 6-37) develops a linear correlation of measured mean surface temperature with elevation. The reference surface temperature is 18.23°C at an elevation of 1,231.0 m, averaged using measured data from borehole NRG-6. Based on measurements at NRG-7a, the calculated mean lapse rate is 0.009°C/m. The water table elevation and temperature are contained in Table 4-12. The location identified as R5C10 (Northing 170730, Easting 234913) (see Section 5.1), used in this report, is nearest to the grid column BTb76 (Northing 170840, Easting 234950) in DTN: LB0303THERMSIM.001 [DIRS 165167]. Table 4-12. Information Used to Calculate the Ground Surface and Water Table Temperatures Grid/Mesh Column ID Easting (m) Northing (m) Water Table Elevation a (m) Water Table Temperature (°C) BTb76 170840 234950 730 28.27 DTN: LB0303THERMSIM.001 [DIRS 165167] (file bot_temp_thermal_grid.dat). a DTN: MO0106RIB00038.001, Water Table Altitude, midpoint of small-gradient area. 4.1.7 Waste Package Heat Decay Table 4-13 shows the repository average lineal heat load as a function of time since waste emplacement. This design information is used as input to the models and analyses described in Section 6. Table 4-13. Waste Package Heat Decay Time Since Emplacement (years) Lineal Heat Load (kW/m) Time Since Emplacement (years) Lineal Heat Load (kW/m) 0.000001 1.45E+00 26 8.525E-01 1 1.399E+00 27 8.382E-01 2 1.357E+00 28 8.245E-01 3 1.321E+00 29 8.114E-01 4 1.289E+00 30 7.992E-01 5 1.259E+00 31 7.858E-01 6 1.232E+00 32 7.730E-01 7 1.206E+00 33 7.610E-01 8 1.181E+00 34 7.493E-01 9 1.157E+00 35 7.381E-01 10 1.135E+00 36 7.262E-01 11 1.110E+00 37 7.150E-01 12 1.088E+00 38 7.042E-01 13 1.068E+00 39 6.938E-01 14 1.049E+00 40 6.838E-01 15 1.033E+00 41 6.733E-01 16 1.012E+00 42 6.632E-01 17 9.934E-01 43 6.535E-01 Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 4-11 October 2004 Table 4-13. Waste Package Heat Decay (Continued) Time Since Emplacement (years) Lineal Heat Load (kW/m) Time Since Emplacement (years) Lineal Heat Load (kW/m) 18 9.759E-01 44 6.441E-01 19 9.595E-01 45 6.351E-01 20 9.443E-01 46 6.258E-01 21 9.267E-01 47 6.169E-01 22 9.103E-01 48 6.083E-01 23 8.950E-01 49 6.000E-01 24 8.805E-01 50 5.920E-01 25 8.666E-01 Source: BSC 2004 [DIRS 167754], Table 12. 4.1.8 Kuehn and Goldstein Parameters for Natural Convection Table 4-14 lists constants for large Rayleigh numbers in the Kuehn and Goldstein correlations for natural convection. These constants are used in the mixed convection correlation to calculate convection heat transfer coefficients. This information is used as input to the models and analyses described in Sections 6 and 7. Table 4-14. Constants for Large Rayleigh Numbers in the Kuehn and Goldstein Correlations for Natural Convection Term Value ci 0.5 i C 0.12 co 1 o C 0.12 m 15 Source: Kuehn and Goldstein 1978 [DIRS 130084], Eq. 1a and 1b. 4.1.9 Thermophysical Properties of the Waste Package Section 5.2 provides rationale for using a 21-PWR as a representative waste package. Table 4-15 shows the thermophysical properties and dimensions of a 21-PWR waste package, its inner stainless steel shell, and its outer Alloy 22 shell. This design information was used as input to a multilayer model of the waste package in the ANSYS calculations described in Section 6. Subsequent to completion of the analyses, the design-basis dimensions for a typical 21-PWR were superseded due to the evolution of waste package design; the design-basis thickness of the inner shell became 50.8 mm instead of 50 mm, and the nominal diameter became 1718.3 mm instead of 1644 mm (BSC 2004 [DIRS 169472], Table 1). The increase of the inner shell thickness is less than 2%, which is expected to have insignificant effect on the results because the magnitude of change is far smaller than the range of variations in different waste package Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 4-12 October 2004 diameters. The more significant change is in the nominal diameter, an increase of about 5%. This would result in a 5% increase in the surface area available for heat transfer. However, because nominal diameters of waste packages vary from 1375.4 mm to 2126.0 mm (BSC 2004 [DIRS 169472], Table 1), the inputs in Table 4-15 are suitable for intended use because they are within the range of waste package dimensions of interest and, therefore, justified for use as dimensions of a representative waste package. Table 4-15. Thermophysical Properties of the Waste Package Reference Temperature (°C) a Thermal Conductivity (W/m·K) a Specific Heat (J/kg·K) a Density (kg/m3) a Emissivity a Thickness 21-PWR (mm) Nominal Diameter 21-PWR (mm) Waste Package Internal Cylinder (21- PWR) N/A 1.5 378 3495 N/A N/A 21.11 13.33 482.93 37.78 13.67 488.19 65.56 14.19 499.38 93.33 14.54 500.68 121.11 15.06 511.31 148.89 15.58 521.64 176.67 15.92 522.43 Waste Package Inner Shell (316NG) 204.44 16.44 528.75 7980 N/A 50 b 48, 52 10.1 414 100 11.1 423 Waste Package Outer Shell (Alloy 22) 200 13.4 444 8690 0.87 20 c 1644b a BSC 2001 [DIRS 156276], pp. 13 and 14. b BSC 2003 [DIRS 165406], Table 1 (superseded data, justified in text). c BSC 2004 [DIRS 169472], Table 1. 4.1.10 In-Drift Geometry and Ventilation Parameters Table 4-16 lists various in-drift geometric and preclosure ventilation parameters. This design information is used as input to the models and analyses described in Section 6. The design-bases configurations for the height from the invert top to the center of 21-PWR waste package and the invert height were superseded because of the changes in waste package design and invert structure design. The design-basis height from the invert top to the center of 21-PWR waste package and invert height have been changed to 1051 mm (BSC 2004 [DIRS 168489]) and 864 mm (2’ 10”) (BSC 2004 [DIRS 169503]) from 1018 mm (BSC 2003 [DIRS 164069]) and 806 mm (BSC 2003 [DIRS 164101]), respectively. Since the heights from the invert top to the center of different types of waste packages vary from 887 mm to 1286 mm (BSC 2004 [DIRS 168489]), the inputs in Table 4-16 are within the range of properties of interest. The use of the invert height of 806 mm is justified through comparison of the analytical model to the ANSYS model (Section 6.6.2). The analytical model does not explicitly account for thermal conduction through the invert, whereas such thermal conduction is included in the ANSYS Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 4-13 October 2004 model. A comparison shows that the results are not sensitive to the presence of the invert because of the agreement between the ANSYS and analytical models. Therefore, the inputs in Table 4-16 are justified for their intended use in this report. Table 4-16. Emplacement Drift Geometries, Ventilation Flow Rate, Ventilation Duration Parameter Value Source Emplacement Drift Diameter (m) 5.5 BSC 2004 [DIRS 168489] Emplacement Drift Spacing (m) 81 BSC 2004 [DIRS 168489] Nominal Ventilation Airflow Rate Preclosure (m3/s) 15 BSC 2004 [DIRS 168489] Ventilation Duration After Final Emplacement (years) 50 BSC 2004 [DIRS 168489] Height from Invert Top to Center of 21-PWR (mm) 1018 BSC 2003 [DIRS 164069]a Invert Height (mm) 806 BSC 2003 [DIRS 164101]a a Superseded data, justified in text. 4.1.11 Thermophysical Properties of Air Table 4-17 lists the thermophysical properties of air at one atmosphere, which corresponds to pressure at sea level. This information is used as input to the models and analyses described in Section 6. However, the emplacement drifts will be located above sea level, where the pressure is about 0.88 atmosphere (Appendix XIX). The major effect of this 0.12 atmosphere difference in pressure is that the air density will be about 12 percent lower than shown in Table 4-17, resulting in a 12 percent reduction in the mass of air flowing through the drift for a given volumetric flow rate, based on the ideal gas law. The effect of this can be calculated with a 12 percent reduction in air density. The effect of reducing the total air pressure as indicated results in a reduction of the ventilation efficiencies by approximately 1% as reported in Section 6.11. Table 4-17. Thermophysical Properties of Air Reference Temperature (K) Density (kg/m3) Specific Heat (kJ/kg·K) Viscosity 107 (N·s/m2) Kinematic Viscosity 106 (m2/s) Thermal Conductivity 103 (W/m·K) Thermal Diffusivity 106 (m2/s) Prandtl Number 250 1.3947 1.006 159.6 11.44 22.3 15.9 0.720 300 1.1614 1.007 184.6 15.89 26.3 22.5 0.707 350 0.995 1.009 208.2 20.92 30.0 29.9 0.700 400 0.8711 1.014 230.1 26.41 33.8 38.3 0.690 Source: Incropera and DeWitt 1996 [DIRS 108184], Table A.4. This change in total pressure from one atmosphere to 0.88 atmosphere does not affect the other pertinent physical properties, which are specific heat, viscosity, and thermal conductivity. The other properties in Table 4-17, the kinematic viscosity, thermal diffusivity, and Prandtl number, are derived quantities and need no further discussion. Because the specific heat has units of energy per unit mass per degree K, the volumetric heat capacity depends on the gas density, which can be predicted by the ideal gas law. The heat capacity of an ideal gas is not dependent upon pressure (Reid et al. 1977 [DIRS 130310], Section Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 4-14 October 2004 7-1), and air behaves as an ideal gas around one atmosphere total pressure because its compressibility factor (usually denoted as Z) is close to unity, which defines an ideal gas (Reid et al. 1977 [DIRS 130310], Section 3-2). The compressibility at conditions of interest can be determined from reduced properties. Using a critical pressure of 37.2 atmospheres and a critical temperature of –140.7°C for air (Perry et al. 1984 [DIRS 125806], p. 3-111, Table 3-161), the reduced pressure at one atmosphere is P/Pc = 1/37.2 ˜ 0.027, and the reduced temperature at 100°C is T/Tc = 373/132.4 ˜ 2.8. According to a generalized compressibility chart for these reduced properties (Reid et al. 1977 [DIRS 130310], Figure 3-1), Z ˜ 1 for pressures and temperatures relevant to the ventilation caclulations, and thus air behaves as an ideal gas. The other two properties are not significantly affected by a small pressure change. The viscosity of air is essentially constant with pressure changes around one atmosphere (Reid et al. 1977 [DIRS 130310], Figure 9-8). The thermal conductivity of gases at low pressure (up to 10 atmospheres) increases about 1 percent per atmosphere (Reid et al. 1977 [DIRS 130310], Section 10-5) so that a change of 0.12 atmosphere results in a thermal conductivity relative change of about 0.12 percent, which is small enough to be ignored. Therefore, the gas-phase density is the only physical property in Table 4-17 that changes significantly as the pressure changes from one atmosphere to 0.88 atmosphere, and the gas-phase density can be calculated from the ideal gas law. 4.1.12 Thermophysical Properties of Water Table 4-18 lists the thermophysical properties of pure water. This information is used in Section 6.9 and Appendix XIII to demonstrate that the contribution of latent heat of vaporization may be neglected in calculating the ventilation efficiency. The composition of the water is not relevant. Table 4-18. Thermophysical Properties of Water Reference Temperature (K) Specific Volume 103 (m3/kg) Heat of Vaporization (kJ/kg) Specific Heat (kJ/kg·K) Viscosity 106 (N·s/m2) Thermal Conductivity 103 (W/m·K) 273.15 1.000 2502 4.217 1750 569 300 1.003 2438 4.179 855 613 350 1.027 2317 4.195 365 668 Source: Incropera and DeWitt 1996 [DIRS 108184], Table A.6. 4.1.13 Kays and Leung Parameters for Forced Convection Table 4-19 lists Kays and Leung parameters used in the mixed convection correlation to calculate forced convection heat transfer coefficients. This information is used as input to the models and analyses described in Sections 6 and 7. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 4-15 October 2004 Table 4-19. Kays and Leung Parameters for Forced Convection Annulus Radius Ratio (r*) Reynolds Number (Re) Nusselt Number – Inner Surface Condition, Inner Surface Heated Alone (Nuii) Non-Dimensional Temperature – Inner Surface (.i) Nusselt Number - Outer Surface Condition, Outer Surface Heated Alone (Nuoo) Non- Dimensional Temperature – Outer Surface (.o) 1.00E+04 38.6 0.412 29.4 0.063 3.00E+04 79.8 0.338 64.3 0.055 1.00E+05 196 0.286 165 0.049 3.00E+05 473 0.26 397 0.044 0.2 1.00E+06 1270 0.235 1070 0.04 1.00E+04 30.9 0.3 28.3 0.137 3.00E+04 66 0.258 62 0.119 1.00E+05 166 0.225 158 0.107 3.00E+05 400 0.206 380 0.097 Fluid with Prandtl Number = 0.700 0.5 1.00E+06 1080 0.185 1040 0.09 Source: Kays and Leung 1963 [DIRS 160763], Table 1. 4.1.14 Physical Constants Table 4-20 lists physical constants used as inputs to the model and analyses of Sections 6 and 7. Table 4-20. Physical Constants Property Value Source Stefan-Boltzmann (W/m2·K4) 5.670 × 10-8 Incropera and DeWitt 1996 [DIRS 108184], Back cover Gravity (m/s2) 9.8 Incropera and DeWitt 1996 [DIRS 108184], Back cover Ideal Gas Law Constant (kJ/kmol·K) 8.315 Incropera and DeWitt 1996 [DIRS 108184], Back cover Prandtl Number Exponent (Dittus-Boelter Correlation) 0.4 Incropera and DeWitt 1996 [DIRS 108184], Section 8.5 Molecular Weight of Water (g/mol) 18 Weast 1977 [DIRS 106266], p. B117 Molecular Weight of Dry Air (g/mol) 29 Weast 1977 [DIRS 106266], p. F13 – F15 4.1.15 Emissivity of the Invert Material Table 4-21 lists the emissivity of rocks used for the invert material. Justification for use of the data for the invert material is provided in Section 4.1.5. This information is used as input to the models and analyses described in Section 6 and the validation exercises described in Section 7. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 4-16 October 2004 Table 4-21. Emissivity of Rocks Used for the Invert Minimum Maximum 0.88 0.95 Source: Incropera and DeWitt 1996 [DIRS 108184], Table A.11 for Rocks at 300K. 4.1.16 Direct Inputs from Outside Sources Table 4-22 lists the direct inputs that are obtained from outside sources. Appendix XVIII contains the demonstrations for each of these sources that it is suitable for its use in this report. Table 4-22. Direct Inputs Obtained from Outside Sources Information Used Reference Identification Equations for annular radiant heat transfer Bird et al., 1960 [DIRS 103524] Linearization of radiant heat transfer, conduction equations in cylindrical and slab systems Carslaw and Jaeger 1959 [DIRS 100968] Van Genuchten and retention relations, steady-state unsaturated flow equation Fetter 1993 [DIRS 102009] Emissitivity of rock and concrete, thermophysical properties of air and water, Dittus-Boelter heat transfer correlation, definitions, radiation equation for annulus, treatment of air as non-radiant absorber, conditions for boundary layer, constants Incropera and DeWitt 1996 [DIRS 108184] Nusselt number definition for forced convection Kays and Leung 1963 [DIRS 160763] Natural convection heat transfer in annulus, correlation from experiment Kuehn and Goldstein 1978 [DIRS 130084] Effective Reynolds number in mixed flow Morgan 1975 [DIRS 160791] Superposition principle Nagle and Saff 1994 [DIRS 100922] Physical properties of air Reid et al. 1977 [DIRS 130310] Standard atmosphere White 1986 [DIRS 111015] 4.2 CRITERIA This section addresses the applicable acceptance criteria from Yucca Mountain Review Plan, Final Report (NRC 2003 [DIRS 163274]), the required documentation of level of accuracy (BSC 2004 [DIRS 170950], Section 3.3), and the completion criteria (BSC 2004 [DIRS 170950], Section 3.4). Each of these criteria is detailed below, in separate sections. 4.2.1 Yucca Mountain Review Plan Criteria The TWP does not state specific acceptance criteria for this report. However, this report provides results that feed indirectly into the model abstraction for quantity and chemistry of water contacting engineered barriers and waste forms. Yucca Mountain Review Plan, Final Report (NRC 2003 [DIRS 163274]) lists the following acceptance criteria for that model abstraction (NRC 2003 [DIRS 163274], Section 2.2.1.3.3.3) that are applicable to this report: Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 4-17 October 2004 • Acceptance Criterion 1 – System Description and Model Integration Are Adequate. (1) Total system performance assessment adequately incorporates important design features, physical phenomena, and couplings, and uses consistent and appropriate assumptions throughout the quantity and chemistry of water contacting engineered barriers and waste forms abstraction process; (3) Important design features, such as waste package design and material selection, backfill, drip shield, ground support, thermal loading strategy, and degradation processes, are adequate to determine the initial and boundary conditions for calculations of the quantity and chemistry of water contacting engineered barriers and waste forms; (6) The expected ranges of environmental conditions within the waste package emplacement drifts, inside of breached waste packages, and contacting the waste forms and their evolution with time are identified. These ranges may be developed to include: (i) the effects of the drip shield and backfill on the quantity and chemistry of water (e.g., the potential for condensate formation and dripping from the underside of the shield); (ii) conditions that promote corrosion of engineered barriers and degradation of waste forms; (iii) irregular wet and dry cycles; (iv) gamma-radiolysis; and (v) size and distribution of penetrations of engineered barriers; • Acceptance Criterion 2 – Data Are Sufficient for Model Justification. (2) Sufficient data were collected on the characteristics of the natural system and engineered materials to establish initial and boundary conditions for conceptual models of thermal-hydrologic-mechanical-chemical coupled processes, that affect seepage and flow and the engineered barrier chemical environment. • Acceptance Criterion 3 – Data Uncertainty Is Characterized and Propagated Through the Model Abstraction. (1) Models use parameter values, assumed ranges, probability distributions, and bounding assumptions that are technically defensible, reasonably account for uncertainties and variabilities, and do not result in an under-representation of the risk estimate; (2) Parameter values, assumed ranges, probability distributions, and bounding assumptions used in the total system performance assessment calculations of Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 4-18 October 2004 quantity and chemistry of water contacting engineered barriers and waste forms are technically defensible and reasonable, based on data from the Yucca Mountain region (e.g., results from large block and drift-scale heater and niche tests), and a combination of techniques that may include laboratory experiments, field measurements, natural analog research, and process-level modeling studies; (3) Input values used in the total system performance assessment calculations of quantity and chemistry of water contacting engineered barriers (e.g., drip shield and waste package) are consistent with the initial and boundary conditions and the assumptions of the conceptual models and design concepts for the Yucca Mountain site. Correlations between input values are appropriately established in the U.S. Department of Energy total system performance assessment. Parameters used to define initial conditions, boundary conditions, and computational domain in sensitivity analyses involving coupled thermal-hydrologic-mechanical-chemical effects on seepage and flow, the waste package chemical environment, and the chemical environment for radionuclide release, are consistent with available data. Reasonable or conservative ranges of parameters or functional relations are established; 4.2.2 Required Documentation of Level of Accuracy The TWP requires this report to state the level of accuracy, precision, and representativeness for the results of the analyses, and how these were determined (BSC 2004 [DIRS 170950], Section 3.3). 4.2.3 Completion Criteria The TWP (BSC 2004 [DIRS 170950], Section 3.4) requires that the work that is done be consistent with the activities performed as part of Technical Work Plan: Regulatory Integration Evaluation of Analysis and Model Reports Supporting the TSPA-LA (BSC 2004 [DIRS 169653]) and that it fulfill a portion of the Phase 2 work identified in that plan. That is, the work should address the prioritized list of actions selected in Phase 1 for disposition in Phase 2 (BSC 2004 [DIRS 169653], Section 1.3). Another completion criterion in the TWP (BSC 2004 [DIRS 170950], Section 3.4) is that the work satisfy the requirements of AP-16.1, Condition Reporting and Resolution, to enable closure of Condition Reports (CRs) CR-2049 and CR-1841D. CR-2049 pertains to providing discussion of criteria to establish that the adequacy of the scientific basis for the model is consistent with the intended use of the model. CR-1841D (Level D) pertains to transparency in the documentation of the conceptual model processes, the validation method, and the criteria used to determine validation. 4.3 CODES, STANDARDS AND REGULATIONS 4.3.1 Codes This report was prepared to comply with 10 CFR Part 63, the U.S. Nuclear Regulatory Commission rule on high-level radioactive waste. Subparts of this rule that are applicable to data Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 4-19 October 2004 include Subpart E, Section 114 (Requirements for Performance Assessment). The subpart applicable to models is also outlined in Subpart E, Section 114. The subparts applicable to features, events, and processes (FEPs) are 10 CFR 63.114(d), (e), and (f). Other codes and standards used in this report are ANSI/NCSL Z540-2-1997 [DIRS 157394], American National Standard for Calibration — U.S. Guide to the Expression of Uncertainty in Measurement, and ASME PTC 19.1-1998 [DIRS 153195], Test Uncertainty, Instruments and Apparatus. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 4-20 October 2004 INTENTIONALLY LEFT BLANK Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 5-1 October 2004 5. ASSUMPTIONS 5.1 REPRESENTATIVE LOCATION WITHIN THE REPOSITORY FOOTPRINT Northing 234913 and Easting 170730 was chosen as the location within the repository footprint to perform the ventilation analyses because it is representative of rock properties, in situ temperature, and stratigraphy information. This assumption does not require confirmation. The rationale for choosing this location is that the repository lies within the tsw35 geologic unit in this area. In addition, this location is representative because it does not lie on an edge or corner of the repository footprint, and it experiences average infiltration rates. This is used in Section 6.5.1. 5.2 THERMAL PROPERTIES OF A 21-PWR WASTE PACKAGE AS REPRESENTATIVE The thermal properties of a 21-PWR waste package were used as representative properties for all waste packages emplaced in the repository. This assumption does not require confirmation. The rationale for using these thermal properties is that the 21-PWR accounts for the majority of the repository inventory. This is used throughout Section 6. 5.3 INITIAL WATER SATURATION OF EACH OF THE STRATIGRAPHIC LAYERS The initial water saturation of the stratigraphic layers is assumed to be approximately 90.5% (a value of 90.54% is used in Appendix II). This assumption does not require confirmation. The rationale for this assumption is that measurements and hydrologic models demonstrate the range of saturation to be between 35 and 99.5%. Sections 6.9.1 and 6.9.2 investigate the effect of a range of matrix saturations on ventilation and Section 6.11 shows that the ventilation efficiency is not sensitive to the choice of saturation. This is used in Section 6.5.2 and Appendix II to account for saturation in obtaining effective thermophysical properties of the stratigraphic units. The sensitivity of the ventilation efficiency with respect to saturation is documented in Section 6.11. 5.4 LITHOPHYSAL PORES ARE AIR-FILLED The lithophysal pores are assumed to be 100% air-filled. This assumption does not require confirmation. The rationale for this assumption is that, based on water retention theory, large voids do not retain liquid water. This is used in Section 6.5.2 and Appendix II to account for airfilled lithophysal porosity in obtaining effective thermophysical properties of the stratigraphic units. 5.5 INVERT BALLAST MATERIAL Repository Design Project, Repository/PA IED Emplacement Drift Committed Materials (2) (BSC 2003 [DIRS 164101]) describes the invert ballast material as crushed tuff. The nominal particle diameter of the crushed tuff is not specified. Therefore, the thermophysical properties of a 4-10 crushed tuff (for which these properties have been measured) are used. This assumption Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 5-2 October 2004 does not require confirmation as the model is not sensitive to this parameter (Section 6.6.2). The rationale for this assumption is that difference in particle sizes has little effect on the bulk thermophysical properties of the material. This is used in Section 6.5.3. 5.6 MIXED CONVECTION CORRELATION The mixed convection correlation incorporates forced and natural convection correlations from experimental data. The correlations are for idealized configurations that are not the same as the EBS configuration. With one exception, the development of the correlation (documented in Appendix IX) recognizes these idealizations and considers their effects in an uncertainty analysis. The one exception applies to natural convection when the outer convective surface is hotter than the air. During the preclosure period the ventilating air removes heat. Because the drift wall is heated by thermal radiation from the waste package, the drift wall (outer convective surface) will be hotter than the air. The development of the mixed convection correlation assumes that the Kuehn-Goldstein correlation remains valid when the drift wall is hotter than the ventilation air. This assumption does not require confirmation. This is used in Appendix IX. The mixed convection correlation is used throughout Sections 6 and 7 to calculate convection heat transfer coefficients. 5.7 TEMPERATURE OF THE VENTILATION AIR AT THE INLET The average temperature of the ventilation air at the inlet to the drift is assumed to be equal to the ambient temperature of the host rock. The reason for choosing equality of the inlet air temperature and initial rock temperature is to avoid removing or putting energy into the rock due to a difference in these temperatures. The concept behind this choice is illustrated by considering an inlet air temperature hotter than the rock which will result (initially anyway) in sensible heat from the air being transferred to the rock. This will increase the efficiency because the efficiency is calculated as a temperature difference at the inlet and outlet multiplied by the volumetric heat capacity of the air. Thus, if energy from the hot air is transferred (initially) to the rock from the air, this does not have relevance to the energy removed from the waste package power sources. This difference of the inlet air and initial rock temperatures complicates the calculation of the efficiency because of the transfer of sensible heat from the air to the rock when the objective is to determine how much energy from the waste package power sources is transferred to the air, and to the rock. Thus, the choice of equal inlet air and initial rock temperatures is justified when the ventilation efficiency for the heat removal from waste packages is the calculation objective. For the Northing and Easting coordinates chosen for analysis, the average temperatures in the ambient state vary from 17°C at the surface to 28°C at the water table (Section 6.5.5). The calculated ambient temperature of the host rock is 22.8°C (Section 6.5.6), which this assumption assigns to the average temperature of the air entering the drift. The study of sensitivity to uncertainty (Section 6.11) assigns 5°C as the standard deviation in inlet air temperature. Because this assignment captures uncertainty introduced by the assumption, the assumption does not require confirmation. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 5-3 October 2004 5.8 EMPLACEMENT SCHEDULE Heat transfer to the rock during emplacement is assumed to be negligible. This assumption does not require confirmation. The rationale for neglecting this transient is that the time scale for placing waste in a drift is small compared to the preclosure period and that the drift will be ventilated during emplacement. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 5-4 October 2004 INTENTIONALLY LEFT BLANK Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-1 October 2004 6. MODEL DISCUSSION Section 6 provides a description of the conceptual models and the scientific, engineering, and mathematical concepts/principles on which the mathematical models are based. Section 6.1 establishes the appropriateness of the model for the purpose of predicting ventilation efficiency, within the limitations stated in Section 1. Direct inputs are listed in Section 4.1. No other corroborating/supporting data, models, or product output were used to develop the model. A conceptual model and an alternative conceptual model for the preclosure heat transfer in and around a waste emplacement drift are developed, implemented, and documented in this section. Table 6-1 outlines the organization of this section. The primary output of the ventilation model is the ventilation efficiency, defined as the fraction of source heat removed by the ventilating air. The ventilation efficiency is expressed as both an instantaneous efficiency (time and distance from the drift inlet dependent), and an integrated efficiency (instantaneous efficiencies integrated over time and drift length). Table 6-1. Outline of the Ventilation Model Documentation Section Content 6.1 Modeling and analysis objectives. 6.2 Lists and describes FEPs assigned to the ventilation model and a summary of their disposition. 6.3 Develops the conceptual model for preclosure heat transfer in and around a ventilated emplacement drift including the basic mathematical equations. The conceptual model includes thermal radiation, convection, and conduction heat transfer. 6.4 Describes the numerical implementations of the conceptual model using the ANSYS/Excel methodology and an analytical approach. 6.5 Lists additional inputs developed from the inputs of Section 4. 6.6 Presents and discusses the results of the numerical and analytical implementations of the conceptual model described in Section 6.4. 6.7 Develops the alternative conceptual model for preclosure heat transfer in and around a ventilated emplacement drift which includes the effects of moisture in the host rock. 6.8 Describes the implementations of the alternative conceptual model using analytical approaches. 6.9 Presents and discusses the results of the analytical approaches which implement the alternative conceptual model. 6.10 Discusses the applicability of the downstream use of the output of the ventilation model (i.e., ventilation efficiency) as a means of representing the preclosure heat transfer to initialize postclosure analyses. 6.11 Discusses the uncertainties associated with the ventilation modeling approaches and the design inputs and parameters, and quantifies the sensitivity of the model output (i.e., ventilation efficiency). Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-2 October 2004 6.1 MODELING OBJECTIVES The thermal energy removed by ventilation is determined by analyzing thermal radiation, thermal convection, and thermal conduction which occur simultaneously in the drift and the surrounding rock mass. The ventilation efficiency, expressed as the percentage of the total thermal energy removed by convection, is the primary output of the ventilation modeling. The ventilation efficiency is used as input in downstream models that do not explicitly simulate the preclosure period. Examples of these models include the multiscale thermohydrologic, UZ coupled process, and drift degradation models. The ventilation modeling and analysis objectives are to: 1. Develop a conceptual model for preclosure ventilation of an emplacement drift (Section 6.3). 2. Implement the ventilation conceptual model using developed software and methods, and the License Application design basis inputs and parameters to predict the preclosure ventilation efficiency (Section 6.4). 3. Verify the results of the numerical application of the ventilation conceptual model through comparative analyses (Section 6.6). 4. Develop an alternative conceptual model for preclosure ventilation which includes the impacts of water and water vapor mass transfer on the heat transfer (Section 6.7). 5. Implement the alternative conceptual model using analytical calculations to assess the impacts of moisture on the ventilation efficiency (Sections 6.8 and 6.9). 6. Demonstrate the applicability of the use of the ventilation efficiency as an abstraction method for downstream postclosure models to account for the preclosure heat removal (Section 6.10). 7. Demonstrate the sensitivity of the ventilation efficiency to discretization and uncertainties in key input parameters associated with the host rock and engineered components including thermal conductivity, matrix and lithophysal porosity, specific heat, emissivity, and convection heat transfer coefficient (Sections 6.6.1 and 6.11). 6.2 FEATURES, EVENTS, AND PROCESSES Table 6-2 provides a listing of FEPs addressed in this document, in accordance with the TWP (BSC 2004 [DIRS 170950]) and DTN: MO0407SEPFEPLA.000 [DIRS 170760]. The table provides specific references to sections within this document. Table 6-2. Included FEPs Addressed in This Document FEP Number FEP Name Section Where Addressed 1.1.02.02.0A Preclosure ventilation Section 6.6 2.1.08.03.0A Repository dryout due to waste heat Sections 6.6 and 6.7 Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-3 October 2004 6.3 CONCEPTUAL MODEL FOR IN-DRIFT VENTILATION Thermal energy released from the waste packages is transferred to the in-drift and host rock surroundings. During the preclosure period, some heat will be removed from the waste packages and emplacement drift by the ventilation system. The heat transfer processes are time and axial position (i.e., the distance down the length of the drift from the airflow entrance point) dependent. A description of the conceptual model follows. No corroborating/supporting data, models, or product output was used to develop the model. 6.3.1 Heat Transfer Processes The heat transfer processes for ventilation of an emplacement drift are shown in Figure 6-1. Figure 6-1 also includes other heat and mass transfer processes that will be outlined later in Section 6.7 where the alternative conceptual model for ventilation is presented. Figure 6-1. Conceptual Model for Heat and Mass Transfer Within and Around an Emplacement Drift Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-4 October 2004 The heat transfer processes depicted in Figure 6-1 include: Process 1. Thermal radiation heat transfer from the surface of the waste package to the drift wall. The rate at which the heat is transferred is calculated using the Stefan-Boltzmann Law for gray surface radiation exchange, at any time during the preclosure period, using the waste package surface and drift wall temperatures. This calculation also requires knowledge of the geometry and emissivities of the waste package and drift wall surfaces. Process 2. Convective heat transfer from the surface of the waste package to the airflow due to the temperature differences between the surface and the moving air. The heat flow rate can be calculated using Newton's Law of Cooling at any time during the preclosure period, using the bulk temperature of the airflow and the temperature of the waste package surface. This calculation also requires knowledge of the convective heat transfer coefficients that implicitly describe the effects of the airflow, the drift geometry, the thermal conductivity of air, and the surface properties on the heat transfer rates. The correlations used (Apendix IX) have simple thermal conduction as a limiting case. Process 3. Convective heat transfer from the drift wall surface directly to the airflow due to the temperature differences between the wall surface and the moving air, similar to process 2. The sum of the two convective heat transfer rates determines the rate of energy addition to the moving air, and can be used to calculate the axial rate of air temperature increase. This calculation also requires knowledge of the convective heat transfer coefficients that implicitly describe the effects of the airflow, the drift geometry, and surface properties on the heat transfer rates. Axially along the drift, the convective heat transfer (processes 2 and 3) is combined with the air mass flow rate and its specific heat to calculate the axial change of air temperature. Process 4. Conductive heat transfer within the rock mass due to changes in drift wall temperature. The heat flow rate into the rock can be determined using Fourier’s Law of Conduction, at any time during the preclosure period, using the temperature gradient in the rock mass. This calculation requires knowledge of the thermal conductivity, saturation, density, and heat capacity of the rock (which vary spatially). The heat transfer for the processes described above can be related by considering the overall conservation of thermal energy except during the early transient response when the waste package temperature is rapidly changing. The following summarizes the coupled components of Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-5 October 2004 the thermal energy conservation during quasi-steady-state conditions when energy storage is relatively constant: • The sum of the radiative heat transfer rate from the waste package to the drift wall (process 1 from above), and the convective heat transfer rate from the waste package into the airflow (process 2), must equal the total rate of heat released from the waste package. • The sum of the convective heat transfer rates from the waste package and drift wall into the airflow (processes 2 and 3), and the conductive heat transfer rate into the rock (process 4), must equal the total rate of heat released from the waste package. • The sum of the convective heat transfer rates from the drift wall into the airflow (process 3), and the conductive heat transfer rate into the rock (process 4), must equal the rate of radiant heat released from the waste. Five additional processes have not been explicitly included in the conceptual model. The first includes the mass transport of water and water vapor and the coupled latent and sensible heat transfer associated with the phase change and movement of water. However, these latent heat effects and near-field host rock mass transport processes can be approximated using boiling point temperature dependent values for the thermal conductivity and specific heat of the rock. This can account for vaporization of pore water and dryout in an approximate sense, but cannot accurately track changes of saturation and evolution of properties. In most cases, the temperatures needed to change these properties are not reached during the preclosure period. These processes are presented in greater detail in Section 6.7. The second process excluded from the conceptual model is the axial transport of heat and mass within the rock domain. This process has negligible influence on the ventilation efficiency during the 50-year preclosure period due to the small thermal diffusivity of rock (~1×10-7 m2/s) and the large (hundreds of meters) scale of the repository footprint. The axial heat transport process, especially about the end of the drift, is captured in the multiscale thermohydrologic model. The third process not included in the ventilation conceptual model is the frictional heating of the air and engineered components due to the moving air. This process is negligible when compared to the waste package heat source due to the low air flow velocities. The fourth process not included in the ventilation conceptual model is episodic flow of liquid water into the drift air (due to heterogeneities in the host rock and episodic infiltration). The total heat added to the airflow by vaporizing such seeps is small compared to the heat from radionuclide decay. It should be noted here that the alternative conceptual model does account for vaporization of liquid water within the host rock and movement of the vapor into the drift, but that process adds only the sensible heat due to the temperature difference between the entering water vapor and the airflow. Finally, the fifth process not included in the ventilation conceptual model is the participation of the drift gas in the radiation process. Water vapor is an effective absorber of infrared radiation; Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-6 October 2004 however, the effect of its absorption and re-radiation of thermal energy is negligible due to low in-drift relative humidity (absolute humidity much less than one atmosphere) during the ventilation period. 6.3.2 Heat Transfer Equations for the Ventilation Model The following three equations represent energy balances for processes 1, 2, and 3 as outlined above in Section 6.3.1 and Figure 6-1: ( ) ( ) [ ] bulk air s s s 4 w 4 s rad s s T T h d T T h d pL Q - - + - · = (Eq. 6-1) ( ) ( ) [ ] bulk air w w w 4 w 4 s rad s w T T h d T T h d pL Q - - - - · = (Eq. 6-2) ( ) ( ) ( ) ( ) [ ] bulk air w w w bulk air s s s in out air p T T h d T T h d pL T T C m - - - + - · = - & (Eq. 6-3) where Qs = heat generated by the waste package (W) Qw = heat transferred into the rock by convection and radiation (W) Ts = waste package surface temperature (K) Tw = drift wall temperature (K) Tair-bulk = (Tair-in + Tair-out)/2 (K) Tin = ventilation air temperature at the drift segment inlet (K) Tout = ventilation air temperature at the drift segment outlet (K) hs = waste package surface convection heat transfer coefficient (W/m2·K) hw = drift wall convection heat transfer coefficient (W/m2·K) hrad = radiation heat transfer coefficient (W/m2·K4) L = drift segment length (m) ds = waste package diameter (m) dw = drift diameter (m) m& = ventilation mass flow rate (kg/s) Cp = specific heat of air (J/kg·K) In the above definitions, temperatures are averages over applicable surfaces. The heat transfer coefficients are effective coefficients for circumferentially integrated heat transfer using averaged temperatures. 6.3.3 Mixed Convection Heat Transfer Coefficient Correlation Energy from the waste package is transferred to the ventilating air by a combination of forced and natural convection, or mixed convection. Morgan developed a general approach for calculating the average heat transferred from horizontal cylinders in mixed convection for various flow regimes and various flow directions (Gebhart et al. 1988 [DIRS 152234], Section 10.4.1). While this approach can be used for the YMP geometry, the specific correlations cannot (they are for external flow). The approach is simplistic: Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-7 October 2004 • Calculate an effective Reynolds number for mixed convection. • Use the mixed convection Reynolds number to calculate a mixed convection Nusselt number. • Use the mixed convection Nusselt number to calculate a mixed convection heat transfer coefficient. The drift wall and waste package surfaces are considered independently, thus coefficients for each wall are derived. Calculating the Reynolds number for forced convection is completed using the definition for Reynolds number for flow in a circular tube (Incropera and DeWitt 1996 [DIRS 108184], p. 421, Equation 8.1). Calculating the Reynolds number for natural convection is not as straightforward. It involves first using literature-provided correlations to calculate a Nusselt number for pure natural convection, and then using this value with the chosen forced convection correlation to determine a Reynolds number that would result in the same heat transfer. The Kuehn-Goldstein (1978 [DIRS 130084]) correlation is generally accepted as the best available model for natural convection, and was chosen for the mixed convection model. The correlation defines Nusselt numbers for the inner and outer cylinders as a function of the Rayleigh number and constants derived from experimental data. The Kays-Leung (1963) model for forced convection in a circular annulus was chosen as the forced convection correlation. In this model the Nusselt number is defined as a function of the heat fluxes and temperatures of the surfaces, and influence coefficients. The influence coefficients are semi-empirical in nature and were determined in conjunction with experimental data. The radii of the cylinders, Reynolds number, and the fluid’s Prandtl number influence the values. The two Reynolds numbers are then combined to give a “mixed” Reynolds number (using Morgan’s approach) as the square root of the sum of the squares of the Reynolds number for forced convection and an equivalent Reynolds number for natural convection. Once a “mixed” Reynolds number is calculated, it can be used in conjunction with the chosen forced convection model (Kays and Leung 1963 [DIRS 160763]) to determine the heat transfer coefficients from the inner (waste package) and outer (drift wall) surfaces. Appendix IX of this report provides a detailed review of the mixed convection correlation, including the development of the method, a review of the sensitivity of the method to each of its parameters, the estimated uncertainty in the heat transfer coefficients predicted by the method, and a comparison of the method results to experimental data from the ventilation tests. Based on these analyses, the mixed convection correlation is valid for the flow conditions attributed to the design parameters presented in Section 4, including a ventilation air flow rate between 10 and 30 m3/s. 6.3.4 Radiation Heat Transfer Coefficient The radiation heat transfer coefficient is calculated from an analytical solution for concentric cylinders (Incropera and DeWitt 1996 [DIRS 108184], p. 739, Table 13-3): Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-8 October 2004 w s w w s rad d d A h · . .. . . .. . - + = e e 1 e 1 s s (Eq. 6-4) where s = Stefan-Boltzmann constant (W/m2·K4) es= surface emissivity of source ew= surface emissivity of drift wall Because the geometry is eccentric rather than concentric, the local radiation heat transfer coefficients are not constant, being larger at the bottom of the con figuration and smaller at the top. However, the value calculated from Equation 6-4 is a reasonable approximation to the effective coefficient for circumferentially integrated energy transfer using averaged temperatures. 6.3.5 Ventilation Efficiency The instantaneous ventilation efficiency is both a function of time and distance from the entrance and is defined by: ( ) ( ) ( ) t Q x t Q x t s air , , = . (Eq. 6-5) where .(t,x) = instantaneous ventilation efficiency (dimensionless) Qair = heat convected to the air from the waste package and drift wall surfaces (W/m) Qs = heat generated by the waste package (W/m) t = time since ventilation began x = distance from the drift entrance (m) The integrated ventilation efficiency is defined by: ( ) ( ) . . . · · .. . .. . · = b s b a air dt t Q x dt dx x t Q 0 0 0 integrated , . (Eq. 6-6) where .integrated = integrated ventilation efficiency (dimensionless) a = limit of integration in terms of the total drift length b = limit of integration in terms of the total ventilation duration Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-9 October 2004 6.4 NUMERICAL APPLICATION OF THE CONCEPTUAL MODEL Two numerical applications and one analytical application of the conceptual model for in-drift ventilation heat transfer are performed. The two numerical applications use the ANSYS software code, and the analytical uses a spreadsheet. The results of each application are compared later in Section 6.6. The first ANSYS based application, named ANSYS-LA-Coarse, divides the drift into segments of 1, 10, 100, 200, 300, 400, 500, 600, 700, and 800 meters. The second ANSYS based application, named ANSYS-LA-Fine, divides the drift into 24 equal segments of 25 meters, for a total of 600 meters. The spreadsheet application, named Analytical- LA-Coarse, is similar to the ANSYS-Coarse model, and was developed to benchmark the analytical approach against ANSYS. 6.4.1 ANSYS Methodology The ANSYS methodology implemented to calculate the various dependent variables in the ventilation model is based on the following energy balances shown in Equations 6-1 through 6-3: • The waste package is the power source in the drift and transfers heat (actually power, i.e., energy per unit time) to the flowing air by forced convection and to the drift wall by radiant heat transfer. The energy balance based on these two heat transfer mechanisms is written in Equation 6-1. In the methodology, the energy is removed uniformly from the surface of the waste package. • The drift wall, as a cylindrical surface, receives energy by radiant heat transfer from the waste package, transfers energy to the flowing air by forced convection, and transfers energy into the rock by conduction. The energy balance based on these three heat transfer mechanisms is written in Equation 6-2. In the methodology, the energy from radiation and convection is transferred uniformly into the drift wall. • The flowing air stream receives energy from the two convection surfaces (i.e., the waste package surface and the drift wall, and the resulting temperature change is written in the energy balance in Equation 6-3). The energy balance that describes the temperature of the drift wall, and in the rock, is written as a two-dimensional transient heat conduction equation for a cylinder in a medium bounded vertically by the location of the mountain surface, the water table, and two vertical insulated boundaries located (usually) equidistant horizontally (to the left and right). There is no heat transfer in the rock along the axis of the drift. Thus, at this point in this methodology description there are three explicit energy balance equations and one implicit (the transient energy balance). Implementation of the ANSYS methodology proceeds by dividing the total drift length into a number of equal lengths, or segments. Within each segment the energy balances for the wastepackage surface, drift wall surface, and rock mass, are solved with the restriction that the inlet air temperature is held (fixed) constant at its inlet value for the duration of a time step (a form of explicit time-stepping). For the first segment that receives (fresh) air this temperature is usually fixed for the entire ventilation duration. Information supplied to ANSYS includes the heat transfer coefficients for the waste package and drift wall surfaces, the dimensions of the waste Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-10 October 2004 package and drift wall, the waste package power as a function of time, and the inlet air temperatures in the form of a lookup table, and the thermophysical properties of the surrounding rock layers. The transient solution is then calculated for each time step up to some specified ventilation duration. Then, in order to calculate the exiting air temperature from the segment, the energy that was transferred to the fixed air temperature for each time step is used to calculate this exiting air temperature based on the total flow (in the time step) and heat capacity of air. This exiting air temperature for this segment for this time step then becomes the inlet air temperature to the next segment. This air exit-temperature calculation is performed external to ANSYS in a spreadsheet. The waste-package power is specified as the average linear power density, for example, kilowatts per meter. This specified linear power density is applied to the entire segment as if the waste package were continuous over the entire segment. This representation is appropriate for the close end-to-end spacing of the waste packages (i.e., 0.1 meter). By fixing the air temperature at the inlet value for the duration of the time step, an assertion is made that the (air) temperature within a segment is everywhere the same (i.e., the air is wellmixed). The concept of a well mixed segment, sometimes referred to as a volume element, is invoked in the engineering design of plug-flow, or “pipe” reactors, and thus this concept has been extensively used in other applications. It can be shown that a series of well-mixed volume elements approximates a plug flow reactor with the restriction that the total volume of the well-mixed volume elements equals that of the plug flow reactor (Levenspiel 1972 [DIRS 156839], p. 137). This concept of a well-mixed volume element means that the air temperature is not a function of location in a segment, even though it is intuitive that the air temperature increases as a function of increasing position within the segment. However, when invoking the concept of a well-mixed volume element, there is no difference in the temperature at the beginning of the segment relative to that at the end of the segment. The question that then arises is: How many segments must be specified in order to obtain results that are considered to be descriptive of the tubular flow situation? The number of series segments is determined by comparing results when the number of segments is increased (through a range) and it is observed that the results do not change; this is sometimes called a “discretization” study. 6.4.1.1 Radiation Heat Transfer Model Use of the Stefan-Boltzmann Law to calculate the radiative heat transfer between the surface of an eccentrically located waste package and the drift wall requires the following: 1. An assumption that the in-drift air does not participate in the radiation heat transfer by absorbing significant amounts of energy that would have been otherwise transferred to the drift wall. 2. Appropriate values for the emissivities of the waste package and drift wall surfaces. 3. An assumption that the use of equation 6-4 is appropriate for the YMP geometry. Emissivity values and their corresponding sources are presented in Section 4 of this report. Emmisivity values were taken from standard engineering sources that are generally accepted in the engineering community. Further confidence is established in the emissivity values through Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-11 October 2004 the fact that the standard deviations of the emissivities of the waste package and drift walls have a negligible effect on the ventilation efficiency based on the sensitivity study presented in Section 6.11 (Table 6-10). The use of the Stephan-Boltzmann law for thermal radiation heat transfer between the surfaces of the waste package and the drift wall, and also the use in this report of a radiation heat transfer coefficient, are corroborated extensively in the engineering literature. The manner in which radiation heat transfer is described in Section 6.4.2.1 (Eq. 6-11) and described in detail in Section 6.4.2.3, is corroborated by the following engineering references. Kern (1950 [DIRS 130111], p.77) describes a fictitious film coefficient to represent the rate at which radiation transfers heat from one surface to another. This film coefficient is of the form of a heat transfer rate that is proportional to a temperature difference, not a difference of the fourth power of the temperatures. Perry et al. (1984 [DIRS 125806], p. 10-13) also describe this technical approach for radiant heat transfer in the form of a “radiation film coefficient.” This description clearly shows that the radiation heat transfer coefficient is a slowly varying function of the temperature difference of the two surfaces, and thus allows the linearization and use as illustrated in Section 6.4.2.1. Also, McAdams (1954 [DIRS 161435], p.78) describes yet another form of this technical approach for the expression of radiation in the form of a first power difference relation for use in combining radiation heat transfer with convection. Thus, the method of determining radiation heat transfer as described in Section 6.4.2.1, and specifically derived in Section 6.4.2.3 is validated through the substantive corroboration with open engineering literature. Additional confidence is achieved by demonstrating that the use of this equation conforms to generally accepted physical principals (i.e. conservation of energy) through its use in the energy balances described in Section 6.3.2 for the ventilation heat transfer model, and the subsequent validation of that model. For enclosures such as an emplacement drift, a medium such as air that separates the radiating surfaces is said to be nonparticipating if it neither absorbs nor scatters the thermal radiation, and it emits no radiation itself. Incropera and DeWitt (1996 [DIRS 108184], Section 13.5) state that: The foregoing conditions and the related equations [summarized in Section 6.3.4 of this report] may often be used to obtain reliable first estimates and, in most cases, highly accurate results for radiation heat transfer in an enclosure…. For nonpolar gases, such as O2 or N2, such neglect [of participating gaseous radiation] is justified, since the gases do not emit radiation and are essentially transparent to the incident thermal radiation. However, the same may not be said for polar molecules, such as CO2, H2O (vapor), NH3, and hydrocarbon gases, which emit and absorb over a wide temperature range. The design of the preclosure ventilation system draws air from the outside environment to the intake shafts and then to the emplacement drifts. The initial composition of the ventilation airstream will resemble that of the outside air, or approximately 78% N2 and 22% O2 with some small fraction of water vapor. The composition of the ventilation airstream may change as it proceeds through the emplacement drift and acquires additional water vapor and CO2 from the host rock. The analysis in Section 6.9.1, Moisture Effects on the In-Drift Ventilation Air Stream, shows that the ventilating air would have a relative humidity of 7.3% and a temperature of 42°C at the end of 50 years. Under these conditions, the amount of water vapor in the air would be less than 1% by mass. The range of relative humidities observed in the Exploratory Studies Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-12 October 2004 Facility was 10 to 40% with occasional increases over 60% (Section 4.1.2). The maximum temperatures recorded in this zone of the ECRB were near 30°C. Because warmer air can hold more water vapor than cooler air, a conservative estimate of the effects of water vapor on the radiation calculations would be to assume 60% relative humidity and 30°C. These conditions represent a conservative estimate because the water vapor content in the air would be a theoretical maximum under the measured conditions. Air at 30°C and 60% relative humidity has a water vapor content of less than 2% by mass, which supports the assumption that the effects of water vapor on radiation can be neglected. 6.4.1.2 Convection Heat Transfer Model The use of convective heat transfer coefficients as used in this report to describe the transfer of energy between a flowing fluid and a surface is seen extensively in the engineering literature. The manner in which convective heat transfer is described and used in section 6.4.2.1 (Equations 6-8, 6-9, and 6-11) is corroborated by the following engineering references. Kern (1950 [DIRS 130111], p.3) presents a mathematical form of convective heat transfer which imitates the form of the conduction equation; that is the energy that is transferred is proportional to the product of the area through which the heat is transferring and the temperature difference between the participating media. The proportionality constant is called the heat transfer coefficient and is usually denoted by the letter “h.” When the fluid is flowing, as is the case during ventilation, the heat transfer process is called “forced convection.” McAdams (1954 [DIRS 161435], p. 187) also discusses the concept of convective heat transfer coefficients, and further distinguishes local and overall heat transfer coefficients. These coefficients pertain to one surface (local) or many surfaces such as the inside and outside of a pipe (overall). The heat transfer coefficients as used for the waste package surface and drift wall surface in section 6.4.2.1 are local heat transfer coefficients. 6.4.1.3 Host Rock Conduction Heat Transfer Model Conduction heat transfer dominates other heat transfer mechanisms (i.e., convection in fractures and lithophysae, and latent heat) in the host rock (Sass et al. 1988 [DIRS 100644], p. 35). This is supported by conclusions of data and modeling of the Drift Scale Test (Birkholzer and Tsang 2000 [DIRS 154608], p. 1439). For the level of confidence required for the ventilation model, the assertion that conduction dominates the heat transfer in the host rock is consistent with the validity of the conceptual model. 6.4.2 Analytical Approach The ventilation calculation technique described in this section is based on the same heat-transfer physics used in the previous ANSYS methodology description in Section 6.4.1. The only change relative to the ANSYS methodology here is in the calculation techniques used to solve the heat transfer equations. This technique is based on two technical approaches to problem solving: the use of a steady-state approximation, and the principle of superposition to calculate the temperature response of the drift wall due to an arbitrary heat flux. By implementing these two techniques, it is not necessary to perform a stand-alone spreadsheet calculation for the air temperature from segment to segment, and there is no requirement to solve the energy equation for the drift wall (rock mass) for every segment. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-13 October 2004 The use of the steady-state approximation, sometimes referred to as a quasi-steady-state approximation, allows the energy balance equations to be written with no time derivatives, only algebraic equations which can then be solved by any number of methods. The solution method used here is to algebraically solve the resulting equations, where there are four equations and four unknowns. The energy balance equations derived as a result of using the steady-state approximation apply for the duration of a time step. The progress of the calculation through time is exactly like that of integrating a function using Euler’s method for numerical integration, summing a “stair-step” approximation. Each step represents a steady state for a particular time interval. Application of the superposition technique is based on the repeated use of a single temperature response of the drift wall due to a short-duration constant flux. This short-duration constant flux is referred to as a “pulse.” By repeatedly applying a series of short-duration scaled constant fluxes to the drift wall, the resulting temperature due to an arbitrary flux can be calculated. Thus the arbitrary flux is approximated like “stair steps.” Part of the ventilation calculation then involves calculating the temperature response of the drift wall due to a single short-duration constant flux, and demonstrating that the short duration, which is the time step, is sufficient to allow the drift-wall temperature to be calculated by superposition for the time-varying fluxes of interest. This temperature response must be calculated independently of the ventilation calculation itself, but is calculated only once for a given set of thermophysical rock properties. The use of a time-series of a constant flux (pulses) to calculate the temperature of the drift wall as a function of time is presented in Section 6.4.2.2. 6.4.2.1 Derivation of the Energy Balance Equations that Describe an Algebraic Solution for Ventilation Calculations This section describes the derivation of the energy balance equations of the analytical ventilation heat-transfer process. This derivation uses a common engineering concept described earlier, well-mixed volume elements. In a well mixed volume element the variables of interest, such as temperature, are everywhere the same. The concept of a well-mixed volume element appears under different names in the engineering literature such as backmix reactor, or continuous-stirred-tankreactor (Levenspiel 1972 [DIRS 156839], p. 139). Using this concept, the drift is divided into a number of well-mixed volume elements that are in series, and the output of air from one is the input to the next (as is done in the ANSYS methodology). The energy balance equations for the algebraic ventilation calculation derivation that follows uses a linearized radiant heat transfer coefficient, as discussed previously (Perry et al. 1984 [DIRS 125806], p. 10-13). The use of a linearized radiant heat transfer coefficient introduces a trial-and-error calculation itself, but has been found to converge very quickly using a successive approximation solution. This linearization does away with the nonlinear nature of radiant heat transfer. The details on the use of the linearized radiant heat transfer coefficient are presented in Section 6.4.2.3. The objective of this derivation is to obtain algebraic expressions for the four dependent variables of interest, these are the air temperature, T (no subscript), drift-wall temperature, Tw, the power-source (waste-package) surface temperature, Ts, and the total energy per unit time conducted into the drift wall (into the rock), Qwall (which when divided by the drift-wall area yields an energy flux). Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-14 October 2004 Consider a well-mixed volume element of a tunnel, or tube, with a heated source inside, air moving through this tunnel, at steady state. A volume element is defined by the “air” volume in a specified length of tunnel. The net energy per time transported by air advection through the element is written as: air p in Q C m T T = - & ) ( (Eq. 6-7) where m& (“m dot”) is the air mass flow rate (kg/s), Cp is the heat capacity of air at constant pressure, Tin is the inlet air temperature (K), and Qair is the net energy/time transported by the air. The air in the volume element is considered well mixed (i.e., a continuous stirred tank reactor or backmix reactor) thus T is the same everywhere in the volume element. The energy per time transferred from the heated source or waste package to the air by convection is written as: )( T T A h q s s s sa - = (Eq. 6-8) where hs is the appropriate heat transfer coefficient (W/m2·K), As is the area of the source in the well-mixed volume element (m2), and Ts is the temperature of the source (surface). The energy per time transferred from the tunnel wall to the air by convection (qwa) is written as: ) ( T T A h q w w w wa - = (Eq. 6-9) where hw is the appropriate heat transfer coefficient (W/m2·K), Aw is the tunnel-wall area in the well-mixed volume element (m2), and Tw is the tunnel-wall temperature (K). Note that the sign convention is positive toward the air and negative away from the air. Energy per time transferred to the air occurs only from the wall and source, so replacing Qair in Equation 6-7 with qsa and qwa from Equations 6-8 and 6-9: )( ) ( ) ( T T A h T T A h Q C m T T w w w s s s air p in - + - = = - & (Eq. 6-10) Now consider the heated source. All of the energy per time is lost instantaneously; the change, if any, in the heat content of the source is negligible compared to the heat lost. The energy per time balance for the source is written as: 0 ) ( ) ( = - - + - s w s s rs s s s p T T A h T T A h (Eq. 6-11) where ps is the source power in the well-mixed volume element (W), the second term on the right is the energy per time transferred to the wall by radiant heat transfer, and hrs is a linearized radiant heat transfer coefficient discussed in Section 6.4.2.3. The energy per time balance at the wall is written on a coordinate frame where energy per time transferred to the wall (surface) is positive, thus: Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-15 October 2004 0 ) ( ) ( = - - - - wall w w w w s s rs Q T T A h T T A h (Eq. 6-12) where the first term is the radiant energy per time transferred to the wall for Ts > Tw, and is thus positive; the second term is the convective energy per time transferred from the wall to the air for Tw > T, and is thus negative; and Qwall is the energy per time transferred by conduction into (from) the wall (medium) in the well-mixed volume element, and is thus negative. Qwall itself can be either negative or positive. Consider now the approximations that can be made for short time intervals in the well-mixed volume element. Fix the wall flux, Qwall/Aw, and the source energy per time, ps, for a yet to be determined time interval (time step). In order to progress with respect to time, approximate the power source and wall flux as a series of constant fluxes (not the same). Section 6.4.2.2 describes the details of how a series of constant fluxes (also known as finite-width pulses) can be used to predict the drift-wall temperature. With Qwall and ps fixed for a short time interval, Equations 6-11 and 6-12 can be used to eliminate Ts and Tw from Equation 6-10. Rearrange Equation 6-11 as: T h A p T h A T h h A s s s w rs s s rs s s + = - + ) ( (Eq. 6-13) And rearrange Equation 6-12 as: T h A Q T h A h A T h A w w wall w w w rs s s rs s - = + - ) ( (Eq. 6-14) Rewrite Equations 6-13 and 6-14 in matrix notation as: ( ) ( ) . .. . . .. . . .. . . .. . . .. . . .. . T h A - Q T h A + p = T T h A + h A - h A h A - h + h A w w wall s s s w s w w rs s rs s rs s rs s s (Eq. 6-15) Write the determinant of the 2×2 matrix as: 2 rs 2 s w w rs s rs s s h A h A h A h h A D + + + - = ) )( ( (Eq. 6-16) Expand D to obtain: 2 rs 2 s w rs w s 2 rs 2 s w s w s rs s 2 s h A h h A A h A h h A A h h A D + - - - - = (Eq. 6-17) Cancel the squared terms to obtain: ) ( w rs w s w s w s rs s 2 s h h A A h h A A h h A D + + - = (Eq. 6-18) And finally obtain by factoring out AsAwhw from the second and third terms: Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-16 October 2004 [ ] ) ( rs s w w s rs s 2 s h h h A A h h A D + + - = (Eq. 6-19) Now solve for Ts using Cramer’s rule. Do this by replacing column 1 in the 2×2 matrix with the right side of matrix Equation 6-15, the forcing vector, and obtain (multiply by D to obtain DTs): . . . . . . . . + - - - + = ) ( w w rs s w w wall rs s s s s s h A h A T h A Q h A T h A p DT (Eq. 6-20) Expand the 2×2 determinant to obtain: ) ( ) )( ( T h A Q h A h A h A T h A p DT w w wall rs s w w rs s s s s s - + + + - = (Eq. 6-21) Collect the coefficient of T, and a constant: ) ( ) ) ( ( w w rs s s wall rs s w w rs s w w rs s s s s h A h A p Q h A T h A h A h A h A h A DT + - + - + - = (Eq. 6-22) Rearrange the coefficient of T to obtain: ) ( )) ( ( w w rs s s wall rs s rs s w w s rs s 2 s s h A h A p Q h A T h h h A A h h A DT + - + + + - = (Eq. 6-23) Note that the coefficient of T above is D as given by Equation 6-19, so dividing by D to obtain Ts yields: ) ( ) ( rs s w w s rs s 2 s w w rs s s rs s wall s h h h A A h h A h A h A p h A Q T T + + + + - + = (Eq. 6-24) Solve for Tw in the same manner from Equation 6-15 by replacing column 2 with the forcing vector to obtain: . . . . . . . . - + + = T h A Q h A T h A p h h A DT w w wall rs s s s s rs s s w ) ( (Eq. 6-25) Expand the determinant to obtain: rs s s s s w w wall rs s s w h A T h A p T h A Q h h A DT ) ( ) )( ( + - - + = (Eq. 6-26) Collect coefficient of T, and a constant: Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-17 October 2004 s rs s rs s s wall rs s w w s rs s 2 s w p h A h h A Q T h h h A A h h A DT - + + + + - = ) ( )) ( ( (Eq. 6-27) So that the result for Tw after dividing by D, Equation 6-19, becomes: ) ( ) ( rs s w w s rs s 2 s s rs s rs s s wall w h h h A A h h A p h A h h A Q T T + + + + - + = (Eq. 6-28) Write Ts from Equation 6-24, and Tw from Equation 6-28, as: s s B T T + = (Eq. 6-29) w w B T T + = (Eq. 6-30) And the coefficients Bw and Bs are defined (use the = sign) from Equations 6-24 and 6-28 as: ) ( ) ( rs s w w s rs s 2s s rs s rs s s wall w h h h A A h h A p h A h h A Q B + + + + - = (Eq. 6-31) ) ( ) ( rs s w w s rs s 2s w w rs s s rs s wall s h h h A A h h A h A h A p h A Q B + + + + - = (Eq. 6-32) Use Equation 6-29 for Ts and Equation 6-30 for Tw to rewrite the air energy balance as Equation 6-10 for T as: ) ( ) ( ) ( T B T A h T B T A h C m T T w w w s s s p in - + + - + = - & (Eq. 6-33) The expression for T becomes: p w w w s s s in C m B A h B A h T T & + + = (Eq. 6-34) Consider a simplification of Equation 6-34, by rewriting this equation as: w w w s s s p in B A h B A h C m T T + = - & ) ( (Eq. 6-35) and substitute Bw and Bs from Equation 6-31 and Equation 6-32 to obtain: Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-18 October 2004 ) ( ) ) ( ( ) ( )) ( ( ) ( rs s w w s rs s 2s s rs s rs s s wall w w rs s w w s rs s 2s w w rs s s rs s wall s s p in h h h A A h h A p h A h h A Q A h h h h A A h h A h A h A p h A Q A h C m T T + + + + - + + + + + - = - & (Eq. 6-36) Regroup Qwall and ps and work on the numerator(s) above to obtain: = + + - + + - rs w w s s rs s w w s wall w w rs s s s s rs s 2 s wall h h A A p h h h A A Q h A h A A h p h h A Q ) ( ) ( = + + + + + - ) ( )) ( ( 2 rs w w s w s w s rs s s s rs s w w s rs s 2s wall h h A A h h A A h h A p h h h A A h h A Q )) ( ( )) ( ( rs s w w s rs s 2s s rs s w w s rs s 2s wall h h h A A h h A p h h h A A h h A Q + + + + + - (Eq. 6-37) The coefficients of Qwall and ps cancel with the denominator(s) in Equation 6-36, so the net result is: s wall p in p Q C m T T + - = - & ) ( (Eq. 6-38) This result can be obtained by writing an energy balance on just the air in the well-mixed volume element. To see this, consider the control envelope to be the air in the volume element, so T - Tin = .T, and multiplication by m& Cp yields the net rate of energy transported through the volume element carried by the air. Since ps is the energy per time added by the source, and +Qwall is the energy per time transferred by conduction into the wall (see the text following Equation 6-12 for the sign convection), -Qwall + ps is the net energy per time removed from the volume element by the air (moving through). This rather simple energy balance is recovered from the preceding equations. At this point an equation is required that relates Tw and Qwall, and this is obtained from use of the superposition principle as described in Section 6.4.2.2. This equation is (and is also Equation 6-48): w b w wall N T P A Q Pt = + . . 1 2 (Eq. 6-39) The summation symbol denotes the pulse contributions to the temperature Tw from all previous wall fluxes, N denotes the number of time steps, and for the situation where the time step is one year, N denotes the total time. The summation runs from 2 to N, not 1 to N, because the current wall flux is not (yet) known (it is Qwall/Aw). The current wall flux is multiplied by Pt1 which is the pulse temperature response at an age of 1 year due to the application of a constant flux of a known strength (for example, 1.0 W/m2). In other words, in the stand-alone term above, the contribution to the wall temperature is being calculated at the end of 1 year due to the flux Qwall/Aw being applied for 1 year. But all the other wall fluxes are known and do not change, they are “history,” and their contribution to Tw diminishes with respect to time because with each time step they get “older.” The contribution of all of the older fluxes to the temperature Tw are taken into account in the summation. The factor Pb is a conversion factor that takes into account any units conversion necessary between the wall flux, Qwall/Aw, and the pulse flux basis. For example, suppose that the Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-19 October 2004 applied constant flux is 1.0 W/m2 and the units of Qwall/Aw are W/m2, then the conversion factor is unity. However, if English units for Qwall/Aw are used, such as Btu/(hr·ft2), then Pb would be 0.3171 (see List of Conversions), which is the conversion of 1.0 W/m2 to Btu/(hr ft2). Now write Equation 6-38 as: in p s wall T C m p Q T + + - = & (Eq. 6-40) And substitute this expression for T in Equation 6-30 to obtain: w in p s wall w B T C m p Q T + + + - = & (Eq. 6-41) There are now two equations for Tw, Equations 6-39 and 6-41. Equate Tw from each of these equations to obtain one equation with one unknown, and that unknown is Qwall. Proceeding: ) ( ) ( 1 2 rs s w w s rs s 2s s rs s rs s s wall in p s wall b w wall N h h h A A h h A p h A h h A Q T C m p Q P A Q Pt + + + + - + + + - = + . . & (Eq. 6-42) The denominator on the right is –D in Equation 6-19, so condense notation one more time keeping –D: ) ( ) ( 1 2 D p h A h h A Q T C m p Q P A Q Pt s rs s rs s s wall in p s wall b w wall N - + + - + + + - = + . . & (Eq. 6-43) Now solve for Qwall. ) ( ) ( ) ( 1 2 1 D p h A T C m p D h h A C m P A Pt Q s rs s in p s rs s s p b w wall N - + + + . - = . . . . . . . . - + + + . & & (Eq. 6-44) Or: ) ( ) ( 1 ) ( 1 2 D h h A C m P A Pt D p h A T C m p Q rs s s p b w s rs s in p s wall N - + + + - + + . + - = . & & (Eq. 6-45) The net result is an equation for Qwall in terms of the knowns of the calculation. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-20 October 2004 The implementation of the calculation proceeds from Qwall above. The calculated value of Qwall is then used to calculate Bw from Equation 6-31, and Bs from Equation 6-32. Then Ts follows immediately from Equation 6-29, and Tw follows from Equation 6-30. T, which is the temperature of the air, follows immediately from Equation 6-38. Thus, the four variables of interest, T, Tw, Ts, and Qwall, are determined. The use of a linearized radiant heat transfer coefficient introduces a trial-and-error calculation which is implemented as follows. An initial guess of the radiant heat transfer coefficient, hrs, is used to start the calculation. A reasonable value can be obtained by examining the information given in the engineering literature (Perry et al. 1984 [DIRS 125806], p. 10-13). Using the initial guess, the calculation proceeds as described above and values for Ts and Tw are obtained. These just-calculated values are now used to calculate hrs, as described in Section 6.4.2.3, and the entire calculation repeated. This “successive approximation” is repeated until the temperatures Ts and Tw change very little from one trial to the next, say 0.1 degrees. The radiant heat transfer coefficient does not vary excessively for the parameters of the problem, again seen by examining the engineering literature. In other words, hrs varies by about a factor of 2 over the range of parameters of interest, and as a result the convergence is easily obtained. The calculation progresses with respect to time by solving for the variables of interest at a time step using the equations noted above, and then stepping to the next time interval. The summation in Equation 6-39 then increases by 1 (which is N), and the calculation repeated out to the specified ventilation duration. 6.4.2.2 Description of the Use of a Constant-Flux Temperature Response to Calculate the Temperature Due to an Arbitrary Flux Consider an arbitrary energy flux applied to a solid. The temperature response of this solid can be calculated by summing the temperature responses from individual constant fluxes applied over short time intervals in such a manner that the constant fluxes approximate the arbitrary flux. The temperature response being referred to here is the temperature at the surface of the drift wall. The temperature response can be calculated in this way due to the use of the superposition principle for the heat conduction equation (Nagle and Saff 1994 [DIRS 100922], p. 166). In order to illustrate this calculation technique and establish an indexing system, the following description is presented. Suppose that the temperature response S due to a single unit flux pulse f is tabulated at every n·.t for n = 1,2,3,...., refer to Figure 6-2, specifically to the upper plot of the temperature S versus time. In the first time interval, .t, the unit flux pulse is “on,” and from here on refer to the unit flux pulse as the “pulse.” After .t the pulse is “off,” and the boundary condition where the flux was applied is flux = 0. The temperature response S decreases with respect to time because the energy delivered to the solid is being conducted, or diffused, into the solid (rock mass), and as such the temperature decreases. Now suppose an arbitrary flux is available in functional or tabulated form. Refer to the middle plot in Figure 6-2 of an arbitrary flux f as a function of time. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-21 October 2004 NOTE: Illustration shows how to calculate a temperature t due to an arbitrary flux f using the repeated application of the temperature response s due to a unit flux pulse applied initially between time = 0 and t1. This calculation methodology is based on the superposition principle and thus adds the temperature contributions from each scaled flux, s·f, to obtain the temperature t at the indicated time. Figure 6-2. Diagram of the Pulse Response by the Superposition Method Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-22 October 2004 In order to calculate the temperature T illustrated in Figure 6-2 due to the arbitrary flux applied up to time = 7·.t (the 7 is arbitrary, for illustration only), the temperature contribution from each of the applied single pulses within each .t is scaled by the flux at the time the flux was applied, and the temperature contributions summed. In order to illustrate this, consider the contribution to the temperature T due to the pulse applied in the first .t between t0 and t1. The temperature response will “age,” or “decay,” to the value indicated at S7. But the S-versus-time plot is based on a unit flux (or whatever flux one chooses). Therefore S7 must be scaled by the value of the arbitrary flux applied in the first .t, so the contribution to the temperature T at t7 due to this flux is S7·f1, and this is illustrated in the plot of T versus time with a “line” connecting S7 and f1. This “line” means multiply these two values. Instead of using f1 as indicated, a midpoint or average value of the flux in this time interval can be used. Likewise, consider the contribution to the temperature T due to the flux applied between t6 and t7. This temperature response is S1 because it is only one .t from its origin in time. This value of S1 is scaled by the flux used between the times indicated. Thus the contribution to the temperature T at t7 due to this flux is S1·f7, and this is added to the sum of contributions, and also illustrated in the plot of T versus time with a “line” connecting S1 and t7. In general, suppose the time index of interest is N, and the time is t = N·.t, then TN is written as: .= + - = N 1 n 1 n N n N f S T (Eq. 6-46) To illustrate the indexing in this summation, consider N = 7, let n = 1, then N – n + 1 = 7, and the product is S1·f7. Now let n = 7, then N – n + 1 = 1, and the product is S7·f1. A table of how the indices run for N = 7 can be found in Table 6-3. Table 6-3. Example of the Indexing for the Pulse Response Method n N - n +1 Sn·fN-n+1 1 7 S1·f7 2 6 S2·f6 3 5 S3·f5 4 4 S4·f4 5 3 S5·f3 6 2 S6·f2 7 1 S7·f1 The temperature TN is then the sum of SnfN-n+1 in the last column. This computation scheme is intended to calculate the temperature on the time “nodes” as indicated. Now suppose that the flux f in the last time interval, .t between t6 and t7, is an unknown. All the other fluxes are “history” because they have already occurred, and hence are known. Thus the above summation can be written as a sum of what occurred (known), and what is going to occur (unknown) in the current time step as: Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-23 October 2004 N 1 N 2 n 1 n N n N f S f S T + = .= + - (Eq. 6-47) In the illustration using N = 7 the indices of the last term above are S1f7 which illustrates, referring to Figure 6-2, that the flux in the indicated time interval (the last one) is being scaled by S1. This form of the summation equation for the temperature is used in Section 6.4.2.1 as Equation 6-39, and rewritten with the following notation: . . = + N w b w wall T P A Q Pt 2 1 (Eq. 6-48) In this form of the summation equation, Pt1 corresponds to S1, Qwall/Aw corresponds to fN, and Pb is a scale factor (inserted for future convenience). Qwall in Section 6.4.2.1 is the total energy per time delivered to the total drift wall in the segment; thus, dividing by the total drift wall area in the segment, Aw, yields the indicated flux Qwall/Aw. 6.4.2.3 Linearized Radiant Heat Transfer Coefficient The linearization of radiant energy transfer is discussed in numerous texts (Carslaw and Jaeger 1959 [DIRS 100968], p. 21, and Perry et al. 1984 [DIRS 125806], p. 10-13, use the terminology “radiation film coefficient”; Kern 1950 [DIRS 130111], p. 77, describes a fictitious film coefficient to represent the rate at which radiation passes from one surface of a radiator). In order to derive a linearized radiant heat transfer coefficient for a heated tunnel, consider the transport of heat by radiation in an annulus as given by Bird et al. (1960 [DIRS 103524], p. 453, problem 14.G2): length time energy ] [ 1 - e 1 A 1 + e A 1 ) T - T ( = Q 2 2 1 1 42 4 1 12 · = .. . .. . . .. . . .. . s (Eq. 6-49) Q12 (W/m) is the net radiant energy interchange between surface 1 and 2, T1 (K) and T2 (K) are the respective surface absolute temperatures, e1 and e2 are the respective emissivities, s is the Stephan Boltzmann constant, and A1 is the surface area of the inner cylinder per unit length (m2/m) (see Bird et al. 1960 [DIRS 103524], p. 448, Example 14.5-2, for a similar problem where “unit length” is used), and the [=] symbol means “has units of.” Therefore, change the subscripts from 1 . s (the source which is the inner cylinder), and from 2 . w (the wall which is the outer cylinder). So the energy per time (heat) transferred becomes: length time energy ] [ 1 - e 1 A 1 + e A 1 ) T - T ( = Q w w s s 4 w 4 s sw · = .. . .. . . .. . . .. . s (Eq. 6-50) The energy (heat) transferred for a length .x (i.e., the length of the well-mixed volume element) is: Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-24 October 2004 time energy ] [ 1 - e 1 A 1 + e A 1 x ) T - T ( = x Q w w s s 4 w 4s sw = .. . .. . . .. . . .. . . . s (Eq. 6-51) At this point it is necessary to recognize that the areas here, As and Aw, as written above in Equation 6-51 are not the same areas that appear in Equation 6-10. The areas in Equation 6-51 are more appropriately “specific” areas (i.e., area per unit length). Those areas in Equation 6-10 are areas in the well-mixed volume element. Therefore, change the notation in Equation 6-51 to denote “specific” areas; to do this, define the specific area for As as Aus where the subscript “us” denotes per unit length. Likewise for Aw use Auw. Equation 6-51 now appears as: time energy ] [ 1 - e 1 A 1 + e A 1 x ) T - T ( = x Q w uw s us 4 w 4s sw = .. . .. . . .. . . .. . . . s (Eq. 6-52) Note that Qsw.x is the total energy per time transferred from the source in the well-mixed volume element. Now define (use the = symbol for “define”) a linearized radiant transfer coefficient based on the power source area As in the well-mixed volume element (see Equation 6-11) as: x Q ) T - T ( x) D ( h = ) T - T ( A h sw w s s rs w s s rs . = . p (Eq. 6-53) So using Equation 6-52 for QSW: .. . .. . .. . .. . = 1 - e 1 A 1 + e A 1 ) T - T ( ) T - T ( D h w uw s us 4 w 4sw s s rs s p (Eq. 6-54) So that by definition: e temperatur · area · time energy ] [ 1 - e 1 A 1 + e A 1 D ) T - T )/( T - T ( h w uw s us s w s 4 w 4s rs = .. . .. . . .. . . .. . = p s (Eq. 6-55) Carrying out the indicated division yields: .. . .. . .. . .. . = 1 - e 1 A 1 + e A 1 D ) T + T T + T T + T ( h w uw s us s 3 w 2 w s w 2s 3s rs p s (Eq. 6-56) And hence when using such a linearization a trial-and-error calculation is introduced because Ts and Tw must be specified. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-25 October 2004 Therefore, the linearized radiant-heat transfer coefficient defined in Equation 6-55, hrs, is the coefficient that multiplies the source area in the well-mixed volume element, but note that the areas appearing in Equation 6-56 are specific areas. 6.4.2.4 Thermal Pulse Calculation In order to implement the ventilation calculation using the analytical approach described in Section 6.4.2, it is necessary to have a temperature response of the drift wall due to the application of a pulse of energy put into the drift wall. This temperature pulse response was introduced in Equation 6-39 in the derivation of the analytical equations, and its use further described in Section 6.4.2.2. The sections that follow here describe how to calculate this temperature response analytically using results from the open literature. This analytical temperature pulse response is based on using two analytical temperature solutions; these are the temperature in the infinite region bounded internally by a cylinder for a constant heat flux (Carslaw and Jaeger 1959 [DIRS 100968], p. 338), and the temperature in the semi-infinite solid for a constant heat flux (Carslaw and Jaeger 1959 [DIRS 100968], p. 75). The first analytical solution, that for the region bounded internally by a cylinder, is used to describe the drift-wall temperature for the early times of the pulse response, and the second analytical solution, that for the semi-infinite solid, is used for the drift-wall temperature for the later, or long-term, times of the pulse response. The reason that the temperature response from the semi-infinite solid can be used for later times is that a pulse of energy entering the drift wall spreads out to the adiabatic boundary at midpillar at later times, and then transports vertically within the rock. A pulse response for each of these time frames is obtained from these constant-flux solutions by shifting the analytical result by one year (for a one-year pulse) and subtracting from the unshifted solution. This shift-and-subtract operation to yield the pulse is based on the superposition principle as described by Nagle and Saff (1994 [DIRS 100922], p. 166). The entire temperature pulse response is then generated by taking the maximum of these two pulses out to the maximum time of interest. The discussions presented in Sections 6.4.2.4.1 and 6.4.2.4.2 below pertain to how to compute the drift wall temperature in an infinite medium. This is then used to calculate the pulse response. 6.4.2.4.1 The Infinite Region Bounded Internally by a Cylinder The temperature in the infinite region bounded internally by a cylinder is (Carslaw and Jaeger 1959 [DIRS 100968], p. 338, Equation 17): du (ua)] Y + (ua) J [ u (ua) J (ur) Y - (ua) Y (ur) J ) e - (1 K Q 2 - = v 2 1 2 1 2 1 o 1 o t u - 0 2 . p . 8 (Eq. 6-57) where v is the temperature, Q is a constant flux, a is the cylinder radius, Jo, J1, Yo, and Y1 are Bessel functions as used by Carslaw and Jaeger (1959 [DIRS 100968]), and the other symbols are previously described. This equation can be put into a dimensionless form that is convenient because it is then necessary to perform the calculation indicated only once for any value of drift Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-26 October 2004 radius (m), a, thermal conductivity (W/m·K), K, and thermal diffusivity (m2/s), .. To put the above equation in dimensionless form, proceed by defining the dimensionless variable . as: ua = . (Eq. 6-58) which differentiating with respect to the integration variable u yields: du a = d . (Eq. 6-59) Substituting the above two results into Equation 6-57 yields: a d )] ( Y + ) ( J [ a ) ( J a r Y - ) ( Y a r J ) e - (1 K Q 2 - = v 2 1 2 1 2 2 1 o 1 o t ) /a ( - 0 2 . . . . . . . . p . . .. . .. . .. . .. . . 8 (Eq. 6-60) Now define a dimensionless time t as: a t 2 . t= (Eq. 6-61) And evaluate the temperature at the cylinder surface (i.e., drift wall) by setting r = a and obtain a dimensionless temperature written as: . . . . . . . . p . t . d )] ( Y + ) ( J [ ) ( J ) ( Y - ) ( Y ) ( J ) e - (1 2 - = a Q a) = v(r K 2 1 2 1 2 1 o 1 o - 0 2 . 8 = (Eq. 6-62) This equation is used to generate the temperatures of interest at specific times as follows. Suppose that the dimensionless temperatures have been generated as a function of dimensionless time, t, for .t = 1, 2, … up to some maximum tmax. Now suppose that the temperature is required at times of every year, .t = 1; use Equation 6-61 to write: 1,2,3... = n for t), (n a = 2 n . .. . .. . . t (Eq. 6-63) To further illustrate, suppose that . = 26 m2/year and a = 2.75 meters (for a 5.5-meter diameter drift), so that the above becomes: 1,2,3... = n for t), 3.44(n n . ˜ t (Eq. 6-64) To generate the temperature v at the desired times of one-year increments, the a priori calculated dimensionless temperature and dimensionless time at discrete values of t can be interpolated accordingly from those at tn in Equation 6-63. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-27 October 2004 6.4.2.4.2 The Semi-Infinite Slab The temperature in the semi-infinite slab is (Carslaw and Jaeger 1959 [DIRS 100968], p. 75, Equation 7): . . . . . . . . . .. . . .. . · · - · .. . .. . · · · = · · - t x erfc x e t K F t v t x . p . . 2 2 2 ) ( 4 2 1 0 2 (Eq. 6-65) where v is the temperature, F0 is a constant flux (equal to one-half the linear power at the drift wall, applied over the area determined by the drift spacing), erfc(z) is the complementary error function, and the other symbols are as previously described. The temperature at the face, or x = 0, is: p . t K F t v · · · = 0 2 ) ( (Eq. 6-66) To generate the temperature response due to a one-year pulse, again shift the solution by one year and subtract from the unshifted solution: . .. . . .. . - · - · · · = p . p . ) 1 ( 2 ) ( 0 t t K F t v (Eq. 6-67) for t = 1. For geometries modeled here, F0 is 1 W/m applied at the drift wall. 6.5 DEVELOPED INPUTS, BOUNDARY CONDITIONS, AND MESHES This section summarizes the inputs developed from Section 4.1.1, which are used in the ANSYS and analytical models. 6.5.1 Thickness of Each of the Stratigraphic Layers The rme6 v1.2 and YMESH v1.54 software routines (Sections 3.2 and 3.3) were used to generate product output in the form of stratigraphic layer thickness at a specified northing and easting coordinate pair (DTN: MO0306MWDSLTLC.000, P2WR5C10.col). This product output was used in the ANSYS based ventilation models. The stratigraphic layer thicknesses are presented in Table 6-4. The computed surface elevation and water table elevation are 1363.4m and 774.4m, respectively (DTN: MO0306MWDSLTLC.000, P2WR5C10.col). The computed center of the emplacement drift is located at 310.5m from the surface (DTN: MO0306MWDSLTLC.000, P2WR5C10.col). The data used in ANSYS ventilation models was based on the preliminary product outputs generated by the unqualified rme6 v1.2 and YMESH v1.54 software routines (DTN: MO0303MWDSLTLC.000). Qualified versions of rme6 v1.2 and YMESH v1.54 were Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-28 October 2004 later used to generated the another set of stratigraphic layer thickness at the same northing and easting coordinate (DTN: MO0306MWDSLTLC.000). The two sets of data were compared. There was no difference between the product outputs from the preliminary and the qualified versions of rme6 v1.2 and YMESH v1.54. Therefore, DTN: MO0306MWDSLTLC.000 for the qualified data set is listed as the source in Table 6-4. Table 6-4. Thickness of the Stratigraphic Layers rme6 v.12 and YMESH v1.54 Northing 234912.719 Easting 170730.297 Stratigraphic Unit Thickness (m) tcw12 20.2 tcw13 4.0 ptn21 7.2 ptn22 5.6 ptn23 2.0 ptn24 12.5 ptn25 36.5 ptn26 11.3 tsw31 2.0 tsw32 45.6 tsw33 85.3 tsw34 33.0 tsw35 104.7 tsw36 25.8 tsw37 12.9 tsw38 21.9 tsw9z 6.6 ch1z 15.0 ch2z 20.3 ch3z 20.3 ch4z 20.3 ch5z 20.3 ch6z 17.6 pp4 19.7 pp3 14.3 pp2 4.1 Output DTN: MO0306MWDSLTLC.000, P2WR5C10.col. 6.5.2 Effective Thermophysical Properties of the Stratigraphic Layers Table 6-5 lists the effective thermophysical properties of the stratigraphic units which take into account the effects of 90.5% water saturation of the matrix porosity (Section 5.3) and 100% air saturation of the lithophysal porosity (Section 5.4) on the thermal conductivity, density, and Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-29 October 2004 specific heat. These properties were obtained using Table 4-6, Table 4-7, Table 4-10, and Table 4-11. The calculation of these properties is documented in Appendices I and II. These properties were used in the ANSYS-based models. Table 6-5. Effective Thermophysical Properties of the Stratigraphic Units Used in the ANSYS Models Unit Effective Thermal Conductivity (W/m·K) Effective Specific Heat (J/kg·K) Effective Density (kg/m3) tcw12 1.76 930 2673 tcw13 0.90 950 2721 ptn21 1.01 960 2973 ptn22 1.01 960 2973 ptn23 1.01 960 2973 ptn24 1.01 960 2973 ptn25 1.01 960 2973 ptn26 1.01 960 2973 tsw31 1.27 940 2561 tsw32 1.76 930 2673 tsw33 1.74 930 2578 tsw34 2.01 930 2665 tsw35 1.83 930 2563 tsw36 2.07 930 2635 tsw37 2.07 930 2635 tsw38 0.79 980 2449 tsw9z 1.01 980 2942 ch1z 1.01 1080 2805 ch2z 1.20 1070 2844 ch3z 1.20 1070 2844 ch4z 1.20 1070 2844 ch5z 1.20 1070 2844 ch6z 1.20 1020 2902 pp4 1.08 1040 2878 pp3 1.08 930 3007 pp2 1.33 930 2962 Source: Appendices I and II of this report. 6.5.3 Average Thermophysical Properties of the Invert and Impact Evaluation for Error in Input Heat Capacity Table 6-6 lists the average thermophysical properties of the invert ballast material taken from Tables 4-1 and 4-2. A specific heat value of 1177 J/kg·K instead of the value of 735 J/kg·K was used in the ANSYS model for calculating the ventilation efficiency. Because the mass of the invert per unit length is 3,095 kg/m (BSC 2003 [DIRS 164101], Table 3), the error in specific heat of 442 J/kg·K Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-30 October 2004 produced an error in heat capacity per unit length of 1.37 × 106 J/m·K. The average temperature rise in the invert was always less than 50 K, so that the extra heat required to raise the temperature was less than 7 × 107 J/m. At the initial heat decay of 1.4 kJ/m·s, the time required to produce that amount of heat is less than 5 × 104 s, which is less than one day. Therefore, the impact of the discrepancy in specific heat had no significant impact on the ANSYS results. As will be seen in Section 6.6, this is verified by comparison of the ANSYS results with the results of the analytical model. Table 6-6. Average Thermophysical Properties of the Invert Specific Heat Thermal Conductivity Bulk Density (J/cm3·°C)a (J/kg·K)b (W/m·°C)c (W/m·K)d Thermal Diffusivity (mm2/s)e (g/cm3)f (kg/m3)g 0.93 735 (1177) 0.16 0.16 0.18 1.266 1266 a Average of Table 4-1 for Specific Heat. b Convert a from J/cm3·°C to J/kg·K using the Bulk Density f; value in parentheses was used in ANSYS model. c Average of Table 4-1 for Thermal Conductivity. d Convert c from °C to K. e Average of Table 4-1 for Thermal Diffusivity. f Average of Table 4-2 for Bulk Density. g Convert f from g/cm3 to kg/m3. 6.5.4 In-Drift Cross-Sectional Area Available for Flow The in-drift cross-sectional area available for air flow is calculated in Appendix XVI and is 19.5 m2. This calculation takes into account the cross-sectional area of the drift, minus the cross-sectional area of the waste package and the cross-sectional area of the invert. 6.5.5 Temperature and Flux Boundary Conditions at the Ground Surface, Water Table, and Mid-Pillar The temperature at the ground surface is calculated from the following equation (BSC 2004 [DIRS 169861], Equation 6.3-1): ( ) ref s s ref s s Z Z T T - - - - = . (Eq. 6-68) where Ts = ground surface temperature (°C) Ts-ref = surface temperature at the reference elevation Zref (°C) . = mean lapse rate Zs = ground surface elevation (m) Zs-ref = surface elevation for which the temperature Ts-ref is known (m) Ts is calculated using the information provided in Section 4.1.6, and Section 6.5.1: Ts-ref = 18.23°C . = 0.009°C/m Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-31 October 2004 Zs-ref = 1231.0 m Zs = 1363.4 m Ts = 17.04°C = 17.0°C The water table temperature is calculated by linear interpolation using the following equation: ( )( ) ref w ref w s ref w s ref w w w T Z Z T T Z Z T - - - - + - - - = (Eq. 6-69) where Tw = water table surface temperature (°C) Tw-ref = water table surface temperature at the reference elevation Zw-ref (°C) Zw = water table surface elevation (m) Zw-ref = water table surface elevation for which the temperature Tw-ref is known (m) Tw is calculated using the information provided in Table 4-12, and Section 6.5.1: Tw-ref = 28.27°C Zw-ref = 730.0 m Zw = 774.4 m Tw = 27.48°C = 27.5°C The calculations in this report used preliminary boundary values of 17°C at the surface and 28°C at the water table. These are the round-off values of the qualified values. In each case, the difference between the preliminary (round-off) values and the qualified values is not large enough to have a significant effect on the calculated efficiency or on the conclusions of this report. The flux boundary condition at the mid-pillar is adiabatic. 6.5.6 Temperature of the Ventilation Air at the Drift Inlet The temperature of the ventilation air at the drift inlet is assumed to be equal to the temperature of the host rock at the repository horizon prior to preclosure (Section 5.7). The temperature of the host rock at the repository horizon prior to preclosure was calculated by ANSYS using the boundary conditions described in Section 6.5.5 and the thermophysical properties of the rock layers described in Section 6.5.2. The temperature was calculated to be 22.8°C (DTN: MO0306MWDASLCV.001, air_temp_co.input). 6.6 RESULTS OF THE NUMERICAL APPLICATION OF THE CONCEPTUAL MODEL The results for the ANSYS-LA-Coarse, ANSYS-LA-Fine, and Analytical-LA-Coarse models are presented in terms of temporally and spatially varying temperatures. In addition, ventilation efficiencies are presented for the ANSYS-LA-Coarse and Analytical-LA-Coarse models. A comparison between the ANSYS-LA-Coarse and ANSYS-LA-Fine models quantifies the impact Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-32 October 2004 of the axial discretization along the drift length and serves as a model verification exercise. A comparison between the ANSYS-LA-Coarse and Analytical-LA-Coarse models benchmarks the analytical approach in preparation for further use in the implementation of the alternative conceptual model and the uncertainty/sensitivity analysis (Sections 6.9.2 and 6.11). 6.6.1 The Effects of Axial Discretization The general trends of waste package, drift wall, and drift air temperatures as functions of time and drift length for the ANSYS-based models are shown in Figures 6-3 and 6-4 for a 600 meter drift length case. The waste package and drift wall temperatures are perimeter-averaged results, while the in-drift air temperatures are bulk averaged. The temperatures for the waste package, drift wall, and drift air for the ANSYS-LA-Coarse and ANSYS-LA-Fine models are within 0.4°C for all distances from the drift entrance and times since emplacement. The following general observations with respect to waste package, drift wall, and drift air temperatures for the ANSYS-based models can be made: • With respect to time, temperatures peak at one year into the ventilation period and afterwards decline in an exponential fashion similar to the waste package heat energy input decay curve. • With respect to location along the length of the drift, temperatures increase linearly, with the maximum temperatures occurring at the end of the drift. • The ANSYS methodology is insensitive to the number and length of sub-divisions in the axial direction. 6.6.2 Temperature and Ventilation Efficiency Comparisons for the ANSYS-LA-Coarse and Analytical-LA-Coarse Models The same general trends of temperature variation with time and distance from the drift entrance, as noted in Section 6.6.1 for the ANSYS numerical models, are observed for the Analytical-LACoarse. Figure 6-5 shows the ANSYS-LA-Coarse and Analytical-LA-Coarse temperatures as a function of time for locations 100 m, 600 m, and 800 m from the drift entrance. Figure 6-6 shows temperatures as function of axial distance from the drift entrance for ventilation durations of 5 and 50 years. The temperatures for the waste package, drift wall, and drift air for the two models are within 5°C for all distances from the drift entrance and times since emplacement. The instantaneous ventilation efficiency is calculated using Equation 6-5. Figure 6-7 shows the ANSYS-LA-Coarse and Analytical-LA-Coarse instantaneous ventilation efficiencies as a function of time for locations 100 m, 600 m, and 800 m from the drift entrance. Figure 6-8 shows ventilation efficiencies as function of axial distance from the drift entrance for ventilation durations of 5 and 50 years. The ventilation efficiencies for the two models are within 4% for all distances from the drift entrance and times since emplacement. The overall or integrated ventilation efficiency is calculated using Equation 6-6. Table 6-7 shows the integrated efficiency over 600 and 800 meters of drift length, and 50 years of ventilation for the two models. Use of the ventilation efficiency is discussed in Section 6.10. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-33 October 2004 It should be noted that the ANSYS-based model simulates an eccentrically located waste package and an invert, while the analytical model simulates a concentrically located waste package and no invert. Based on the reasonable comparisons of temperature and efficiency, the ventilation model is not sensitive to the eccentricity of the waste package, nor the presence of the invert (including its thermophysical properties). (a) 100 meters (b) 600 meters Waste Package Drift Wall Drift Air Waste Package Drift Wall Drift Air 0 10 20 30 40 50 60 70 80 90 100 110 0 5 10 15 20 25 30 35 40 45 50 Time (yr) Temperature (°C) ANSYS-LA-Coarse ANSYS-LA-Fine 0 10 20 30 40 50 60 70 80 90 100 110 0 5 10 15 20 25 30 35 40 45 50 Time (yr) Temperature (°C) ANSYS-LA-Coarse ANSYS-LA-Fine Output DTNs: MO0306MWDASLCV.001; MO0306MWDALAFV.000. Figure 6-3. Waste Package, Drift Wall, and Drift Air Temperatures as Function of Time for (a) 100 Meters and (b) 600 Meters from the Drift Entrance for the ANSYS-LA-Coarse and ANSYS-LA-Fine Models Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-34 October 2004 (a) 5 years (b) 50 years Waste Package Drift Wall Drift Air Waste Package Drift Wall Drift Air 0 10 20 30 40 50 60 70 80 90 100 110 0 100 200 300 400 500 600 Drift Length (m) Temperature (°C) ANSYS-LA-Coarse ANSYS-LA-Fine 0 10 20 30 40 50 60 70 80 90 100 110 0 100 200 300 400 500 600 Drift Length (m) Temperature (°C) ANSYS-LA-Coarse ANSYS-LA-Fine Output DTNs: MO0306MWDASLCV.001; MO0306MWDALAFV.000. Figure 6-4. Waste Package, Drift Wall, and Drift Air Temperatures as Function of Drift Length for (a) 5 Years and (b) 50 Years from the Time of Waste Emplacement for the ANSYS-LA-Coarse and ANSYS-LA-Fine Models Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-35 October 2004 (a) 100 meters (b) 600 meters Waste Package Drift Wall Drift Air Waste Package Drift Wall Drift Air 0 10 20 30 40 50 60 70 80 90 100 110 0 5 10 15 20 25 30 35 40 45 50 Time (yr) Temperature (°C) ANSYS-LA-Coarse Analytical-LA-Coarse 0 10 20 30 40 50 60 70 80 90 100 110 0 5 10 15 20 25 30 35 40 45 50 Time (yr) Temperature (°C) ANSYS-LA-Coarse Analytical-LA-Coarse 0 10 20 30 40 50 60 70 80 90 100 110 0 5 10 15 20 25 30 35 40 45 50 Time (yr) Temperature (°C) ANSYS-LA-Coarse Analytical-LA-Coarse (c) 800 meters Waste Package Drift Wall Drift Air Output DTNs: MO0306MWDASLCV.001, MO0307MWDAC8MV.000. Figure 6-5. Waste Package, Drift Wall, and Drift Air Temperatures as Function of Time for (a) 100 Meters, (b) 600 Meters, and (c) 800 Meters from the Drift Entrance for the ANSYS-LA-Coarse and Analytical-LA-Coarse Models Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-36 October 2004 (a) 5 years (b) 50 years Waste Package Drift Wall Drift Air Waste Package Drift Wall Drift Air 0 10 20 30 40 50 60 70 80 90 100 110 0 100 200 300 400 500 600 700 800 Drift Length (m) Temperature (°C) ANSYS-LA-Coarse Analytical-LA-Coarse 0 10 20 30 40 50 60 70 80 90 100 110 0 100 200 300 400 500 600 700 800 Drift Length (m) Temperature (°C) ANSYS-LA-Coarse Analytical-LA-Coarse Output DTNs: MO0306MWDASLCV.001; MO0307MWDAC8MV.000. Figure 6-6. Waste Package, Drift Wall, and Drift Air Temperatures as Function of Drift Length for (a) 5 Years and (b) 50 Years from the Time of Waste Emplacement for the ANSYS-LA-Coarse and Analytical-LA-Coarse Models Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-37 October 2004 (a) 100 meters (b) 600 meters 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 0 5 10 15 20 25 30 35 40 45 50 Time (yr) Ventilation Efficiency ANSYS-LA-Coarse Analytical-LA-Coarse 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 0 5 10 15 20 25 30 35 40 45 50 Time (yr) Ventilation Efficiency ANSYS-LA-Coarse Analytical-LA-Coarse (c) 800 meters 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 0 5 10 15 20 25 30 35 40 45 50 Time (yr) Ventilation Efficiency ANSYS-LA-Coarse Analytical-LA-Coarse Output DTNs: MO0306MWDASLCV.001; MO0307MWDAC8MV.000. Figure 6-7. Ventilation Efficiency as Function of Time for (a) 100 Meters, (b) 600 Meters, and (c) 800 Meters from the Drift Entrance for the ANSYS-LA-Coarse and Analytical-LA-Coarse Models Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-38 October 2004 (a) 5 years (b) 50 years 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 100 200 300 400 500 600 700 800 Drift Length (m) Ventilation Efficiency ANSYS-LA-Coarse Analytical-LA-Coarse 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 100 200 300 400 500 600 700 800 Drift Length (m) Ventilation Efficiency ANSYS-LA-Coarse Analytical-LA-Coarse Output DTNs: MO0306MWDASLCV.001; MO0307MWDAC8MV.000. Figure 6-8. Ventilation Efficiency as Function of Drift Length for (a) 5 Years and (b) 50 Years from the Time of Waste Emplacement for the ANSYS-LA-Coarse and Analytical-LA-Coarse Models Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-39 October 2004 Table 6-7. Integrated Ventilation Efficiency for a 600-meter and 800-meter Drift and 50 Years of Ventilation Model Integrated Ventilation Efficiency (Eq. 6-6) 600-meter Drift, 50 years of Ventilation ANSYS-LA-Coarse a 88.3% Analytical-LA-Coarse b 88.0% 800-meter Drift, 50 years of Ventilation ANSYS-LA-Coarse a 85.8% Analytical-LA-Coarse b 86.0% a DTN: MO0406MWDLACVD.001. b DTN: MO0406MWDAC8VD.001. 6.7 ALTERNATIVE CONCEPTUAL MODEL FOR IN-DRIFT VENTILATION The alternative conceptual model for in-drift ventilation includes the addition of water and water vapor mass transport in the host rock, across the drift wall, and into the ventilation airstream. Water and water vapor mass transport is directly coupled to the heat transfer processes described in the conceptual model for in-drift ventilation. The impacts of the mass transport, in terms of latent heat transfer, temperature, heat removal rates, and near-field host rock dryout are evaluated using analytical approaches. 6.7.1 Alternative Conceptual Model Heat and Mass Transfer Processes The coupled heat and mass transfer processes for the alternative conceptual model for in-drift ventilation are the same as those for the conceptual model described in Section 6.3.1 and Figure 6-1 with the addition of two other processes: Process 5. Water phase change (evaporation and condensation) occurs within the host rock as the temperature and vapor pressure change which causes the host rock saturation to change, thus altering the thermal conductivity of the rock. Process 6. Water (liquid and vapor phases) mass transfer occurs within the host rock and the in-drift air. Water vapor may move within the host rock via diffusion to cooler regions where it condenses. It also may enter the in-drift airflow at the drift wall, causing a change in relative humidity, and can potentially condense in cooler regions of the ventilation system, such as the exhaust main drift and exhaust shafts. The heat transfer rates for processes 5 and 6 can be related to processes 1 through 4 by again considering the overall conservation of thermal energy except during the early transient response when the waste package temperature is rapidly changing. The following is an addition to the thermal energy conservation described in Section 6.3.1: Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-40 October 2004 • The sum of the convective heat transfer rates from the waste package and the drift wall into the airflow (processes 2 and 3), and the heat of the water vapor transported back across the drift wall (process 6) equal the total heat added to the ventilation air. Additionally, the mass transfer rates for processes 1 through 6 can be related by considering the overall conservation of mass during the ventilation period. The following summarizes the coupled components of the mass balance: • The sum of the mass of the ventilation air into the drift and the water vapor that moves across the drift wall from the surrounding host rock equals the mass of the air exiting the drift. Vapor diffusion or enhanced vapor diffusion has the potential of locally increasing the heat flux rate for saturations that are intermediate to full matrix-fracture saturation, and in the dry condition in which water vapor is absent. Vapor diffusion is defined as the movement of water vapor under Fick’s Law. Enhanced vapor diffusion is the movement of water vapor to areas where water vapor is retained, condensed and then evaporated. 6.8 IMPLEMENTATION OF THE ALTERNATIVE CONCEPTUAL MODEL To assess the impact on the ventilation efficiency of the alternative conceptual model processes, the following analyses were performed: • An analytical calculation which bounds the latent heat contribution to the in-drift ventilation air stream (Appendix XIII). • Ventilation analyses using the analytical spreadsheet calculation (named Analytical-LAWet- vs-Dry-kth, Appendix VIII) for host rock at different levels of saturation (and therefore different values of thermal conductivity). • An analysis of experimental data for vapor diffusion and enhanced vapor diffusion. 6.9 RESULTS OF THE APPLICATION OF THE ALTERNATIVE CONCEPTUAL MODEL The results of the analytical calculation to bound the latent heat contribution to the in-drift ventilation air stream, the Analytical-LA-Wet-vs-Dry-kth model, and the analysis of experimental data to quantify the effects of vapor diffusion and enhanced vapor diffusion follow. 6.9.1 Moisture Effects on the In-Drift Ventilation Air Stream An analytical calculation was performed which bounds the latent heat contribution to the in-drift ventilation air stream over a 600 m drift length and 50-year ventilation period in terms of the matrix hydrologic properties. Analytical equations for steady-state unsaturated flow in porous media to a specified moisture potential boundary condition at the drift wall were developed with the help of Jury et al. (1991 [DIRS 102010], pp. 51, 60, 113, 151, Section 3.4) and Fetter (1993 [DIRS 102009], pp. 172, 181, 182). Using 30% relative humidity in the drift, the moisture potential at the drift wall was calculated to be 1.985 × 106 cm (see Appendix XIII, p. XIII-5). Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-41 October 2004 The moisture potential in the surrounding host rock at some distance from the drift wall is calculated using two different sets of measured data: the first being the mean of the measurements from borehole core of matrix saturation in the Tptpll (tsw35) geologic unit (Table 4-4) and the second being measurements of water potential taken from the ECRB Cross-Drift in the Tptpll (tsw35) geologic unit (Table 4-5). The average saturation from the borehole core measurements (Table 4-4) is 74%, which translates to a water potential of 2908 cm (see Appendix XIII, p. XIII-6). Using these potentials, a radius of influence of 6 m, and the hydrologic properties of Tptpll (tsw35) from Table 4-9, the steady-state liquid flux toward the drift (which evaporates) is calculated to be 0.061 mm/year (see Appendix XIII, p. XIII-6). The measured water potential at 5.62 m from the drift wall is 1000 cm (Table 4-5). Using this value, the potential calculated at the drift wall based on relative humidity conditions, and the hydrologic properties of Tptpll (tsw35) from Table 4-9, the steady-state liquid flux toward the drift (which evaporates) is calculated to be 0.278 mm/year (see Appendix XIII, p. XIII-7). If all the moisture which fluxes to the drift wall over the entire length of the emplacement drift is evaporated at some constant temperature, the total latent heat contribution to the in-drift air over the 50-year preclosure period can be calculated. The latent heat contribution is then divided by the total heat output by the waste packages over the same 50-year period and 600 meter long drift. The results are presented in Table 6-8. Table 6-8. Latent Heat Contribution Expressed as a Percentage of the Total Waste Package Heat Over 50 Years and 600 Meters of Drift Model Latent Heat Contribution Analytical model with a moisture flux = 0.061 mm/yeara 0.01% Analytical model with a moisture flux = 0.278 mm/yearb 0.04% a Appendix XIII, p. XIII-6, flux based on the mean saturation from Table 4-4. b Appendix XIII, p. XIII-7, flux based on measured water potential from a borehole in the ECRB Cross-Drift (Table 4-5). The analytical calculation indicates that: • The contribution of heat by vaporization of moisture is rate limited by the hydrogeologic properties of the host rock. • The contribution of heat by vaporization of moisture is a small percentage of the total heat input. As corroboration, consider a comparative calculation which bounds the latent heat contribution. We take a present day percolation flux and apply it at the drift wall. For northing 234913 and easting 170730, the closest UZ grid mesh column ID is g_9 (see Section 4.1.6). The percolation flux (present day climate, upper case) reported for UZ grid mesh column ID g_9 at the base of the ptn unit is 15.70959 mm/year (DTN: LB0302PTNTSW9I.001 [DIRS 162277], file preq_uz_ptn.q). If this percolation flux is flow focused through matrix and fracture network over the width of two drift diameters, and arrives at the drift wall where it is evaporated, the latent Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-42 October 2004 heat contribution can be calculated. Using the thermophysical properties of water at 350 K (Table 4-18), the total latent heat contribution over the 50-year ventilation period and 600 meters of drift is calculated as follows: J 10 1.170 1kJ 1000J kg 2317kJ m 973.7kg 1 50yr 1 600m 5.5m 2 1mm 0.001m yr 15.70959mm 13 3 · = .. . .. . · . .. . . .. . · .. . .. . · .. . .. . · .. . .. . · · · .. . .. . · . .. . . .. . The total waste package heat input over the 50-year ventilation period and 600 meter long drift is 8.605·1014 J (DTN: MO0307MWDAC8MV.000, worksheet “Ventilation Efficiency”). The latent heat contribution expressed as a percentage of the total waste package heat input is: 1.4% J 10 8.605 J 10 1.170 14 13 = · · This calculation supports the conclusion reached earlier, that the contribution of heat by vaporization of moisture is a small percentage of the total heat input. Therefore, neglecting latent heat in the calculation of ventilation efficiencies does not introduce significant error. The reduction in relative humidity to the in-drift ventilation air for a 600 meter long drift is calculated using the methodology outlined in Attachment XXVII of ANSYS Calculations in Support of Natural Ventilation Parametric Study for SR (BSC 2001 [DIRS 155246]) and Mine Ventilation and Air Conditions (Hartman 1982 [DIRS 128009], pages 596-597). Using the percolation flux of 15.7 mm/yr, the conversion factor from page xix, and the thermophysical properties of water at 350 K (Table 4-18), the mass flux of water which arrives at the drift wall is: s grainswater kg grain m kg m m s yr mm m yr mm 371 . 49 10 479891 . 6 1 7 . 973 1 600 5 . 5 2 31556926 1 1 001 . 0 70959 . 15 5 3 = . .. . . .. . × · .. . .. . · .. . .. . · · · .. . .. . · .. . .. . · . .. . . .. . - The mass flux of the ventilation air at 350K (Table 4-17) is: s lbair kg lb m kg s m 904 . 32 1 2046 . 2 995 . 0 15 3 3 = . .. . . .. . · .. . .. . · . .. . . .. . The distribution of mass flux of water to the mass flux of ventilation air (percolation component) is: air water air water lb grains s lb s grains 500 . 1 904 . 32 371 . 49 = . . . . . . . . The relative humidity of the inlet air is taken to be 20.31%, which has a moisture content of 28.210 grainswater/lbair (BSC 2002 [DIRS 161233], Table 5 for P1 Early Emplacement Drift, Intake Main). The new moisture content of the ventilation air is then the sum of the percolation and relative humidity components: Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-43 October 2004 air water air water air water lb grains lb grains lb grains 710 . 29 500 . 1 210 . 28 = + Converting 29.710grainswater/lbair to lbwater/lbair: air water lb lb air water kg lb grain kg lb grains 00424 . 0 1 2046 . 2 1 10 479891 . 6 1 710 . 29 5 = . .. . . .. . · . .. . . .. . × · . .. . . .. . - The average barometric pressure is 26.3322 inHg (BSC 2002 [DIRS 161233], Table 5 for P1 Early Emplacement Drift, Intake Main). The partial pressure can be calculated by (Hartman 1982 [DIRS 128009], Eq. 21-5 rearranging to solve for pv): inHg inHg air water air water lb lb lb lb 178 . 0 00424 . 0 622 . 0 3322 . 26 00424 . 0 = + · The air temperature at the outlet of the 600 meter long drift after 50 years of ventilation is approximately 42°C from the results of the Analytical-LA-Coarse ventilation model (Table 8-6). The saturated vapor pressure at 42°C (107.6°F) is (Hartman 1982 [DIRS 128009], Eq. 21-1): ( ) inHg e F F 427 . 2 18079 . 0 14 . 395 6 . 107 64 . 552 6 . 107 27 . 17 = + ° - ° · The relative humidity at the outlet of the 600 meter long drift is (Hartman 1982 [DIRS 128009], Eq. 21-4): % 3 . 7 % 100 427 . 2 178 . 0 = · inHg inHg Neglecting the contribution from percolation, the relative humidity at the outlet is 6.9%. Therefore, while the ventilation air stream picks up moisture through evaporation of the near field host rock pore water at a rate of approximately 15.7 mm/yr, the relative humidity over 600 meters of drift and 50 years of ventilation decreases from approximately 20% to 7%, primarily due to the increase in air temperature. 6.9.2 Ventilation Analysis for Host Rock at Varying Degrees of Saturation An analytical spreadsheet ventilation analysis (Output DTN: MO0306MWDRTCCV.000, worksheet “Wet vs. Dry” of Analytical-LA-Wet-vs-Dry-kth.xls) assessed the impact of varying degrees of host rock saturation on the waste package, drift wall, and in-drift air temperatures and the ventilation efficiency for a 600 meter long drift. These analyses used the thermophysical properties of the tws35 unit for the repository horizon and matrix water saturation ranging from 0% to 100%. Figure 6-9 shows the impact of a “wet” versus “dry” thermal conductivity on the temperatures of the waste package, drift wall, and in-drift air. The temperatures for the two cases are within 4°C for all distances from the drift entrance and times since emplacement, with the “dry” case being consistently hotter. Figure 6-10 plots the integrated ventilation efficiency as a function of matrix saturation and host rock thermal conductivity. The integrated ventilation Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-44 October 2004 efficiency changes from approximately 87.7% to 90.7% when the matrix saturation goes from wet to dry. 0 10 20 30 40 50 60 70 80 90 100 110 0 5 10 15 20 25 30 35 40 45 50 Time (yr) Temperature (C) Waste Package (wet) Drift Wall (wet) Drift Air (wet) Waste Package (dry) Drift Wall (dry) Drift Air (dry) (a) 100 meters 0 10 20 30 40 50 60 70 80 90 100 110 0 5 10 15 20 25 30 35 40 45 50 Time (yr) Temperature (C) Waste Package (wet) Drift Wall (wet) Drift Air (wet) Waste Package (dry) Drift Wall (dry) Drift Air (dry) (b) 600 meters Output DTN: MO0306MWDRTCCV.000, plots generated based on data from worksheet “Wet vs. Dry“ of Analytical- LA-Coarse-Wet-vs-Dry-kth.xls. Figure 6-9. Waste Package, Drift Wall, and Drift Air Temperatures as Function of Time for (a) 100 Meters and (b) 600 Meters from the Drift Entrance for the Analytical-LA-Wet-vs-Dry-kth Ventilation Model (Attachment VIII) Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-45 October 2004 87.0% 87.5% 88.0% 88.5% 89.0% 89.5% 90.0% 90.5% 91.0% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Matrix Saturation Ventilation Efficiency 87.0% 87.5% 88.0% 88.5% 89.0% 89.5% 90.0% 90.5% 91.0% 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0 Bulk Thermal Conductivity (W/mK) Ventilation Efficiency Output DTN: MO0306MWDRTCCV.000, plots generated based on data in worksheet “Wet vs. Dry” of Analytical-LAWet- vs-Dry-kth.xls. Figure 6-10. Ventilation Efficiency as Function of Matrix Saturation and Bulk Thermal Conductivity Calculated Using the Analytical-LA-Wet-vs-Dry-kth Ventilation Model (Attachment VIII) Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-46 October 2004 6.9.3 Evaluation of Vapor Diffusion and Enhanced Vapor Diffusion on the Host Rock Thermal Conductivity and Thus Ventilation Efficiency The following discussion relates the issue of enhanced vapor diffusion in the surrounding walls of the emplacement drift. Moyne et al. 1990 [DIRS 153164] present information on the effect of enhanced vapor diffusion on increasing the effective thermal conductivity of the rock mass. If enhanced vapor diffusion occurs over intermediate ranges of saturation, the effective thermal conductivity could be two to three times higher than the thermal conductivity under saturated conditions in the matrix (Moyne et al. 1990 [DIRS 153164], Figures 2 through 4). In this case the ventilation efficiency could be lower based upon the higher effective thermal conductivity than would be the case when the saturated thermal conductivity applies. In addition to the conduction of heat under a temperature gradient, it is possible to have vapor phase diffusion (Jury et al. 1991 [DIRS 102010], p. 211) within the rock mass. In addition, enhanced vapor phase diffusion may occur (Jury et al. 1991 [DIRS 102010], p. 212). These related phenomena are not included in the ventilation analysis presented above nor in the ANSYS calculations. Vapor diffusion and/or enhanced vapor diffusion (due to evaporation and condensation in the pores) tend to increase the aggregate thermal conductivity over that of stagnant fluid components. The following discussion presents information regarding vapor diffusion and enhanced vapor diffusion by Jury et al. (1991 [DIRS 102010], pp. 211 to 213). Experimental results obtained by Moyne et al. (1990 [DIRS 153164]) are then discussed to illustrate vapor and enhanced vapor diffusion effects. Experimental results obtained from laboratory measurements on core samples and the results from the large scale Drift Scale Test are then presented. 6.9.3.1 Vapor Diffusion by Jury et al. Jury et al. (1991 [DIRS 102010]) provide a general discussion of vapor diffusion. Laboratory tests have shown that when temperature gradients were placed across soil samples, the measured vapor fluxes were 10 times larger than that predicted by Fick’s Law. It was found that two mechanisms could increase the potential for vapor diffusion. The first mechanism is that water vapor that is fluxing towards moisture that is retained in the pore space may condense on one side and evaporate on the other side (Jury et al. 1991 [DIRS 102010], p. 212, Figure 6.7). The second enhancement mechanism relates to the thermal gradients across the liquid phase contained within the pore space. The thermal conductivity of the solid phase is several times larger than the thermal conductivity of water. Since the thermal gradients within the pore space are more likely to be influenced by liquid water, the effective thermal gradients for a uniform heat flux would be higher. Theoretical considerations suggest that the thermal gradients might be a factor of two to three higher. Moyne et al. 1990 ([DIRS 153164] Figures 2 through 4) show that for materials that have a higher interconnected porosity, that the effective thermal conductivity could be higher and provides a theoretical analysis of the enhanced vapor diffusion. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-47 October 2004 6.9.3.2 Thermal Conductivity – Saturation Relationship for Welded Tuff The following presents a discussion of the relationship of apparent thermal conductivity to saturation. The experimental data are for samples from the middle nonlithophysal (Tptmn) unit, whish is adequate to represent response of the host rock units. Sandia National Laboratories conducted a laboratory investigation of thermal conductivity as a function of saturation state, using welded and nonwelded tuff specimens associated with the Drift Scale Test. Rock core samples were recovered from the repository site to determine the relationship between thermal conductivity and saturation state for both welded and nonwelded tuffs. Welded tuff from the Tptpmn unit was taken from Alcove 5 of the Exploratory Studies Facility. All thermal conductivity tests were conducted at 30°C and at intermediate moisture conditions. The results of the laboratory investigations (DTN: SNL22100196001.006 [DIRS 158213]) below a saturation of 90 percent are presented in Figure 6-11 with a trend line. Thermal Conductivity of the Tptpll as a Function of Saturation 1.4 1.6 1.8 2 2.2 2.4 0 10 20 30 40 50 60 70 80 90 100 Saturation (%) Thermal Conductivity (W/(m*K)) DTN: SNL22100196001.006 [DIRS 158213]. Figure 6-11. Rock Matrix Thermal Conductivity as a Function of Saturation, with Trend Line The measured results show a similar trend with results for a highly compacted clay from Moyne et al. (1990 [DIRS 153164]), which do not exhibit enhanced vapor diffusion effects. It should be emphasized that at a low temperature of 30°C, the vapor pressure of water is not very high, and the general theory presented by Moyne et al. (1990 [DIRS 153164]) would not predict a very large increase in aggregate thermal conductivity. Wildenschild and Roberts (1999 [DIRS 131055]) performed an investigation of thermally driven water vapor diffusion, for tuff from the middle nonlithophysal zone of the Topopah Spring Tuff (Tptpmn), associated with the Large Block Test. Thermal conductivity was measured for a single sample of welded tuff indirectly as a function of total pore pressure, temperature and water content. Enhancement of vapor diffusion in welded tuff was not observed at any of the combinations of saturation, temperature and imposed pressures. At a temperature of approximately 50°C, the aggregate rock matrix thermal conductivity increased modestly with the Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-48 October 2004 degree of saturation from about 1.1 W/m·K at a saturation of 0.1 to about 1.3 W/m·K at a saturation of 0.78 (Wildenschild and Roberts 1999 [DIRS 131055], Figure 5). From a saturation of 0.1 to a saturation of 0.35, the thermal conductivity was approximately constant. The results showed a stronger dependence on temperature than on saturation. The guarded heat flow results are corroborative with the experimental results for a single sample of welded tuff presented by Wildenschild and Roberts (1999 [DIRS 131055]). Wildenschild and Roberts (1999 [DIRS 131055]) show that the aggregate thermal conductivity increased by 18 percent as the saturation increased from zero to 78 percent. This increase compares well with increase of about 16 percent in Figure 6-11 over this same saturation range. In conclusion, experimental studies on welded tuff show that thermal conductivity varies with liquid saturation in a straightforward, monotonic manner. Targeted experiments do not show evidence of enhanced vapor diffusion that increases the aggregate thermal conductivity and affects ventilation efficiency. The results apply over the range of temperature of the host rock during the preclosure period. 6.10 APPLICABILITY OF THE VENTILATION EFFICIENCY AS AN ABSTRACTION METHOD The ventilation efficiency can be expressed as a single value by integrating over both the duration of the preclosure period and the length of the drift (Equation 6-6). It may also be applied as a function of time and drift length (Equation 6-5). Downstream models that model the ventilation period may implement ventilation efficiency either way. The first way is to introduce the heat flux adjusted by the ventilation efficiency directly to the drift wall. Typically, a downstream model that uses the ventilation efficiency in this manner does not model the in-drift components. In this case, the only heat transfer mechanism being simulated is the conduction from the drift wall out to the host rock. Because the solution of the heat conduction equation is linear in nature with constant temperature heat sinks at the upper and lower boundaries of the domain, a unique solution for the temperature of the drift wall exists. Therefore, this method will result in both the same heat flux at the drift wall and the same drift wall temperature history as that predicted by the ventilation model from which the ventilation efficiency was derived. The second way the ventilation efficiency may be used involves downstream models that include the in-drift components in their domains, but cannot model boundary layers and therefore cannot include convective heat transfer. These models typically reduce the waste package heat generation rate by the ventilation efficiency and apply this new heat flux directly to the waste package rather than the drift wall. This type of application relies on both radiation and conduction heat transfer to deliver the right amount of heat to the drift wall, and replicate the drift wall temperature history as predicted by the upstream ventilation model. This approach is less straightforward than the first and requires further discussion as to its feasibility. 6.10.1 Theoretical Use of the Ventilation Efficiency at the Waste Package Consider the case where the preclosure waste package heat output reduced by the ventilation efficiency (calculated by an upstream ventilation model) is used as a substitute for the preclosure Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-49 October 2004 convection to represent the preclosure heat removal by ventilation. An energy balance for the ventilation model is: rad s conv s Q Q Q + = - (Eq. 6-70) where ( ) T T A Q s s conv - · = - (Eq. 6-71) ( ) 4w 4 s rad T T C Q - · = (Eq. 6-72) The fraction of heat removed by the ventilation (i.e., by convection) is: s w conv s conv Q Q Q - - + = . (Eq. 6-73) where ( ) T T B Q w w conv - · = - (Eq. 6-74) The constants A, B, and C are defined as: s s h d A · = (Eq. 6-75) w w h d B · = (Eq. 6-76) rad s h d C · = (Eq. 6-77) Substituting Equations 6-71 and 6-72 into Equation 6-70 yields: ( ) ( ) 4w 4 s s s T T C T T A Q - + - · = (Eq. 6-78) Substituting Equations 6-71, 6-74, and 6-78 into Equation 6-73 yields: ( ) ( ) [ ] ( ) ( ) [ ].. . .. . - · + - · - · + - · = 4w 4 s s w s T T C T T A T T B T T A . (Eq. 6-79) For the downstream model, the waste package heat output is multiplied by the ventilation efficiency to account for the heat removed during the preclosure ventilation period. Equation 6-80 represents the fraction of heat delivered to the drift wall: ( ) . - · = 1 Q Q s w ' (Eq. 6-80) Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-50 October 2004 Substituting Equations 6-78 and 6-79 into Equation 6-80 yields: ( ) ( ) [ ] ( ) ( ) [ ] ( ) ( ) [ ].. . .. . - · + - · - · + - · - · - · + - · = 4w 4 s s w s 4w 4 s s w T T C T T A T T B T T A 1 T T C T T A Q' (Eq. 6-81) An energy balance for the downstream model considered in this case (i.e., where the ventilation efficiency is used as a substitute for the heat transfer via convection) is: ' ' rad w Q Q = (Eq. 6-82) where ( ) 4 w 4 s rad T T C Q ' ' ' - · = (Eq. 6-83) where 's T = waste package temperature of the downstream model (K) 'w T = drift wall temperature of the downstream model (K) Substituting Equations 6-81 and 6-83 into Equation 6-82 and simplifying yields: ( ) ( ) ( ) T T C B T T T T w 4w 4 s 4 w 4 s - · - - = - ' ' (Eq. 6-84) If s s T T = ' and w w T T = ' are to be true, then the term ( ) T T C B w - · must be zero. For this to be true, either the coefficient B must be zero, and/or the terms Tw and T must be equal. The implication for either of these conditions is that there is no convective heat transfer between the drift wall and the drift air, which of course is not true. Therefore, a downstream application in which the ventilation efficiency is used as a substitute for the convective heat transfer to simulate the preclosure heat removal by ventilation cannot accurately represent both the preclosure waste package and drift wall temperatures as calculated by the ventilation model. Because Tw is controlled by heat flux through the wall and far-field boundary conditions, the preclosure waste package temperatures will bear most of the inaccuracy. 6.10.2 Numerical Example Using the Ventilation Efficiency as an Abstraction Method Two numerical examples that apply the theoretical use of the ventilation efficiency as described in Section 6.10.1 are presented below in Sections 6.10.2.3 and 6.10.2.4. Beforehand, total energy balances are presented using the results of the ANSYS-LA-Coarse model and Equation 6-5 for calculating the instantaneous heat removal efficiency as a function of time and drift length, and Equation 6-6 for calculating an integrated heat removal efficiency. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-51 October 2004 6.10.2.1 Using Equation 6-5 to Calculate the Total Energy Delivered to the Host Rock Using the results of the ANSYS-LA-Coarse model and Equation 6-5 to calculate the heat removal efficiency as a function of both time and drift length, the total energy delivered to the host rock over the 50-year preclosure period and a 600 meter long drift becomes: ( ) ( ) ( ) dx dt x , t t Q Energy m yr s total rock · · - · = . . - 600 0 50 0 1 . (Eq. 6-85) where Energyrock-total = total energy to the host rock (J) Qs(t) = waste package lineal heat decay as a function of time (W/m) .(t,x) = instantaneous ventilation heat removal efficiency at some time, t, and some distance from the drift entrance, x, (dimensionless) Using the heat decay from Table 4-13, and the heat removal efficiencies calculated as a function of time and drift length for the ANSYS-LA-Coarse model (DTN: MO0306MWDASLCV.001), the total energy delivered to the host rock over 50 years and 600 meters is 1.02 × 1014 J (DTN: MO0306MWDASLCV.001, worksheet “Efficiency data” of ANSYS-LA-Coarse.xls, row 49, column O). 6.10.2.2 Using Equation 6-6 to Calculate the Total Energy to the System Finally, using the results of the ANSYS-LA-Coarse model and Equation 6-6 to calculate an integrated ventilation heat removal efficiency, the total energy to the system over the 50-year preclosure period and 600 meter long drift becomes: ( ) ( ) dt t Q m Energy yr integrated s total · - · · = . 50 0 1 600 . (Eq. 6-86) where .integrated = integrated ventilation heat removal efficiency given by Eq. 6-6 (dimensionless) Using the heat decay from Table 4-13 and the integrated ventilation efficiency of 88.3% reported in Table 6-7 for the ANSYS-LA-Coarse model, the total energy to the system over 50 years and 600 meters is 1.02 × 1014 J (DTN: MO0306MWDASLCV.001, worksheet “Efficiency data” of ANSYS-LA-Coarse.xls, row 100, column M). This result balances with the energy calculated in Section 6.10.2.1. 6.11 SENSITIVITY OF THE VENTILATION EFFICIENCY TO UNCERTAINTIES IN KEY INPUTS AND DESIGN PARAMETERS The sensitivity of the ventilation efficiency to uncertainties in key inputs and design parameters was investigated using the Delta Method, also referred to as the “generation of system moments” or “statistical error propagation.” The Delta Method involves calculating the mean system Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-52 October 2004 performance, in this case the integrated ventilation efficiency, and its standard deviation using the means and variances of the component variables which make up the system. Equations 6-87 and 6-88 describe the Delta Method mathematically (Hahn and Shapiro 1967 [DIRS 146529], pp. 228 to 231). Equation 6-89 describes the standard deviation based on the variance (Hahn and Shapiro 1967 [DIRS 146529], pp. 228 to 231). ( ) ( ) ( ) ( ) [ ] ( ) .= . . + = n i i i n x Var x h x E ,..., x E , x E h z E 1 2 2 2 1 2 1 (Eq. 6-87) ( ) ( ) .= . .. . . .. . . . = n i i i x Var x h z Var 1 2 (Eq. 6-88) ( ) ( ) [ ]2 i i x x Var s = (Eq. 6-89) where E(z) = mean system performance E(x1, 2, …, n) = mean of the xth component variable h = function that describes the system performance based on the components variable (set of equations from Section 6.4.2 describing the ventilation model) Var(x1, 2, …n) = variance of the xth component variable Var(z) = variance of the system performance s(xi) = standard deviation of the xth component variable When the system performance is a linear function of the component variables, the second and higher order partial derivatives are zero. In other words, the second term of Equation 6-87 goes to zero and the mean system performance can be calculated using only the means of the component variables. In terms of the ventilation model, E(z) represents the mean integrated ventilation efficiency where h[E(x1), E(x2),…, E(xn)] represents the equations of Section 6.4.2 used to perform the algebraic ventilation calculation, and E(x1, 2, …, n) represents the mean values of the inputs and design parameters. Var(x1, 2, …, n) then represents the variances of the inputs and design parameters. h[E(x1), E(x2),…, E(xn)] is evaluated using the analytical method. The variance or standard deviation of the integrated efficiency is calculated using Equation 6-88 which propagates the uncertainties in select inputs and design parameters (expressed by variances or standard deviations). Table 6-9 shows the key inputs and design parameters selected for the Delta Method, along with their respective standard deviations. Where available, standard deviations were assigned from DTNs. Where unknown, standard deviations using normal distributions were determined based on engineering judgment. The source for each standard deviation is documented in Table 6-9. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-53 October 2004 Table 6-9. Inputs and Design Parameters, and Their Respective Standard Deviations, Selected for the Delta Method to Assess the Sensitivity of the Integrated Ventilation Efficiency Input/Design Parameter Central Value Source Standard Deviation Source for Standard Deviation Dry Bulk Thermal Conductivity (W/m·K) 1.2784 Table 4-6 for Tptpll (tsw35) 0.2511 Table 4-6 for Tptpll (tsw35) Wet Bulk Thermal Conductivity (W/m·K) 1.8895 Table 4-6 for Tptpll (tsw35) 0.2484 Table 4-6 for Tptpll (tsw35) Grain Density (kg/m3) 2593 Attachment II for tsw35 (column K) 138 Table 4-6 for Tptpll (tsw35) dry bulk density Solids Specific Heat (J/kg·K) 930 Table 4-7 for Tptpll (tsw35) 130 Table 4-7 for Tptpll (tsw35) Matrix Porosity 14.86% Table 4-6 for Tptpll (tsw35) 3.4% Table 4-6 for Tptpll (tsw35) Matrix Saturation 90.5% Section 5.3 9.5% Saturation cannot exceed 100% Lithophysal Porosity 8.83% Table 4-6 for Tptpll (tsw35) 5.4% Table 4-6 for Tptpll (tsw35) Drift Diameter (m) 5.5 Table 4-16 0.5 Constrained by construction tolerance Waste Package Diameter (m) 1.644 Table 4-15 0.5 Cover range of waste packages from 24-BWR to DHLW Inlet Air Temperature (°C) 22.82 Output DTN: MO0306MWDASLCV.001, air_temp_co.input 5 Constrained by average temperature at the surface and at the water table Air Flow Rate (m3/s) 15 Table 4-16 2 Controlled process variable Drift Wall Emissivity 0.9 Table 4-8 for Tptpll 0.1 Emissivity cannot exceed 1.0 Waste Package Emissivity 0.87 Table 4-15 0.13 Emissivity cannot exceed 1.0 Inner Convection Heat Transfer Coefficient (W/m2·K) 4.23 Output DTN: MO0307MWDAC8MV.000 (Analytical) (average of hs for all time steps, columns F, H, J, L, N, P, R T, in worksheet “CSTR Analysis” of Analytical-LA-Coarse- 800m.xls; converted from 0.74 Btu/h·ft2·°F) 0.63 15% of the Mean (typical combined random standard uncertainty from Tables IX-24 and IX-25 of Appendix IX) Outer Convection Heat Transfer Coefficient (W/m2·K) 3.86 Output DTN: MO0307MWDAC8MV.000 (Analytical) (average of hw for all time steps, columns G, I, K, M, O, Q, S, U, in worksheet “CSTR Analysis” of Analytical-LA-Coarse- 800m.xls; converted from 0.68 Btu/h·ft2·°F) 0.58 15% of the Mean (typical combined random standard uncertainty from Tables IX-24 and IX-25 of Appendix IX) Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-54 October 2004 Using the central values for the inputs and design parameters listed in Table 6-9, the integrated ventilation efficiency is 88% for a 600-meter-long drift, and 86% for an 800-meter-long drift. By employing the Delta Method to propagate the standard deviations of the inputs and design parameters listed in Table 6-9 through the analysis, the standard deviation (normally distributed) of the integrated ventilation efficiency about the central value of 88% is 3% for the 600-meterlong drift, and 3% about the central value of 86% for the 800-meter-long drift. Expressed in terms of the normal distribution, the integrated ventilation efficiency for the 600-meter-long drift will be between approximately 85% and 91%, 68% of the time; between 83% and 93%, 96% of the time; and between 80% and 96%, 99% of the time (Hahn and Shapiro 1967 [DIRS 146529]). The documentation of this analysis is in Appendix VII. Table 6-10 summarizes the first step of the Delta Method, which is to calculate the system performance (ventilation efficiency) using the central values of the system components (input/design parameters) from Equation 6-87. Then, independently and one at a time, each system component value is replaced by its central value plus/minus a standard deviation, and a new system performance is calculated using Equation 6-87. The standard deviation of the ventilation efficiency is calculated using the 5th and 7th columns of Table 6-10 and Equations 6-88 and 6-89. In addition to the uncertainty values in Table 6-10, the change in ventilation efficiency due to a reduction in total pressure from 1 atmosphere to 0.88 atmosphere (due to the repository elevation) as discussed in Section 4.1.11 results in the ventilation efficiency for a 600-meter drift changing from 87.98% to 87.00%, or a change from 88% to 87%, a 1% reduction. This calculation is performed using the Excel analytical model presented in DTN: MO0307MWDAC8MV.000. Likewise for an 800-meter drift the ventilation efficiency changes from 85.93% to 84.72%, or a change from 86% to 85%, again a 1% reduction. This 1% change is considered to be insignificant compared to the standard deviation of 3% discussed above. The sensitivity results with respect to the other key input and design variables are not expected to change if the central values are based on a pressure of 0.88 atmosphere. In addition, the influence of each of the standard deviations of the inputs and design parameters on the integrated ventilation efficiency was determined. Their individual influence on the standard deviation of the integrated efficiency was also determined. These influences are plotted against each other in Figure 6-12. All values of influence for the respective axes of the plot were normalized by dividing by the largest corresponding value. The purpose of Figure 6-12 is to give a qualitative assessment of which inputs and design parameters are most significant (and those that are not significant) in the ventilation model. The significance is determined by the variable’s influence, relative to other variables, on both the integrated ventilation efficiency and its standard deviation. Figure 6-12 shows that the most significant variables in the ventilation model are the inlet air temperature, the air flow rate, the host rock wet bulk thermal conductivity (as a function of matrix saturation and specific heat), and the convection heat transfer coefficients. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-55 October 2004 Table 6-10. Using the Delta Method to Determine the Sensitivity of the Ventilation Efficiency Due to Uncertainties in Key Inputs and Design Parameters for a 600-Meter-Long Drift Input/Design Parameter Central Value Efficiency (Eq. 6-88) Central Value + Standard Deviation Efficiency (Eq. 6-88, replacing the mean of the xth component variable with the value in the previous column) Central Value - Standard Deviation Efficiency (Eq. 6- 88, replacing the xth component variable with the value in the previous column) Dry Bulk Thermal Conductivity (W/m·K) 1.2784 1.5295 87.88% 1.0273 88.06% Wet Bulk Thermal Conductivity (W/m·K) 1.8895 2.1379 87.12% 1.6411 88.87% Grain Density (kg/m3) 2593 2731 87.88% 2455 88.06% Solids Specific Heat (J/kg·K) 930 1060 87.74% 800 88.22% Matrix Porosity 14.86% 18.26 87.93% 11.46 88.01% Matrix Saturation 90.5% 100 87.70% 81.084 88.24% Lithophysal Porosity 8.83% 14.23 88.10% 3.43 87.85% Drift Diameter (m) 5.5 6 87.86% 5 88.07% Waste Package Diameter (m) 1.644 2.144 88.11% 1.144 87.80% Inlet Air Temperature (°C) 22.82 27.82 85.82% 17.82 90.12% Air Flow Rate (m3/s) 15 17 88.81% 13 86.89% Drift Wall Emissivity 0.9 1 87.94% 0.8 88.00% Waste Package Emissivity 0.87 1 87.85% 0.74 88.12% Inner Convection Heat Transfer Coefficient (W/m2·K) 4.23 4.86 88.17% 3.60 87.74% Outer Convection Heat Transfer Coefficient (W/m2·K) 3.87 88.0% 4.45 88.24% 3.29 87.63% DTN: MO0406MWDLACVD.001. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 6-56 October 2004 Output DTN: MO0406MWDLACVD.001. Figure 6-12. Qualitative Plot Showing the Influence of Ventilation Model Inputs and Design Parameters on the Integrated Ventilation Efficiency and Its Standard Deviation Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 7-1 October 2004 7. VALIDATION AP-SIII.10Q requires that total system performance assessment model components be validated for their intended purpose and stated limitations, and to the level of confidence required by a component’s relative importance to the performance of the repository. Section 1 of this report provides the intended use of the ventilation model and the model limitations. The governing technical work plan (BSC 2004 [DIRS 170950], Section 2.3.2) identifies Level I as the appropriate level of validation for the ventilation model. The appropriateness of Level I is based on recognition that the model results are not extrapolated over large distances or time frames, and that ventilation efficiency (model output) represents the preclosure response to forced ventilation subject to engineering verification and controls. Variation in the output of the ventilation model is estimated to have only a small effect (less than 0.1 mrem/year) on the estimated mean annual dose for the repository system (BSC 2004 [DIRS 170950], Section 2.3.1). 7.1 CONFIDENCE BUILDING DURING MODEL DEVELOPMENT TO ESTABLISH SCIENTIFIC BASIS AND ACCURACY FOR INTENDED USE In accordance with AP-2.27Q, Planning for Science Activities, Level I validation includes a discussion of model development. In particular, this report documents decisions implemented during model development that build confidence and verify that a reasonable, credible technical approach using scientific and engineering principles was taken. The development of the model is documented in accordance with the requirements of Section 5.3.2(b) of AP-SIII.10Q and Attachment 3 of AP-2.27Q. The development of the ventilation model has been conducted according to these criteria, as follows: 1. Selection of input parameters and/or input data, and a discussion of how the selection process builds confidence in the model. [AP-SIII.10Q 5.3.2(b) (1) and AP-2.27Q Attachment 3 Level I (a)] The parameters of the mixed convection correlation (Sections 4.1.8 and 4.1.13), standard properties of air and water (Sections 4.1.11 and 4.1.12), and physical constants (Section 4.1.14) are from standard sources that are qualified and justified in Appendix XVIII. Inputs relevant to the design of the EBS, including ventilation, are almost entirely from current information exchange drawings (IEDs), except for minor changes to the design subsequent to completion of the calculations reported here (Sections 4.1.7, 4.1.9, and 4.1.10). The input properties of the rock and the EBS materials, as well as the initial conditions in the rock and ventilating air (Sections 4.1.1 through 4.1.6, 4.1.9, and 4.1.15) are from DTNs and controlled engineering calculations that are specific to the site, except for one outside source for emissivity (Tables 4-8 and 4-21) that is qualified and justified in Appendix XVIII. The method of selecting input parameters and data builds confidence in the model. 2. Description of calibration activities, and/or initial boundary condition runs, and/or run convergences, simulation conditions set up to span the range of intended use and avoid Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 7-2 October 2004 inconsistent outputs, and a discussion of how the activity or activities build confidence in the model. Inclusion of a discussion of impacts of any non-convergence runs. [AP-SIII.10Q 5.3.2(b)(2) and AP-2.27Q Attachment 3 Level I (e)]. The only calibration activities affecting the model parameters (Tables 4.1.8 and 4.1.13) were performed independently of the model development by Kuehn and Goldstein (1978 [DIRS 130084]) and Kays and Leung (1963 [DIRS 160763]). The reliability of these sources, including their citation in a handbook and textbook, is presented in Appendix XVIII. Initial and boundary conditions (Tables 4.1.2 and 4.1.6) for the calculations of ventilation efficiency are from DTNs that are specific to the site. All calculations converged. Simulations were performed for drifts of 600 m and 800 m, spanning the expected lengths of emplacement drifts. Simulations also varied many input parameters over their range of uncertainty (Section 6.11). The independence of the calibration activities, the specificity of the initial and boundary conditions, the convergence of the calculations, and the range of simulations combine to build additional confidence in the model. 3. Discussion of the impacts of uncertainties to the model results including how the model results represent the range of possible outcomes consistent with important uncertainties. [AP-SIII.10Q 5.3.2(b)(3) and AP-2.27Q Attachment 3 Level 1 (d) and (f)]. Simulations varied many input parameters over their range of uncertainty (Section 6.11). These results are combined, using the Delta Method, to calculate that the uncertainty in the ventilation efficiency is about 3% efficiency. This quantification of the uncertainty provides additional confidence in the results. 4. Formulation of defensible assumptions and simplifications. [AP-2.27Q Attachment 3 Level I (b)]. The limitations stated in Section 1 are equivalent to assumptions and simplifications that are applicable to the current EBS design and are therefore defensible. In particular: • The ventilation air flow rate is between 10 and 30 m3/s. • The waste packages are spaced in the drift such that, during the preclosure period, the average heat generation per unit length in each small group of waste packages is approximately the same as the average over the entire drift. • Conduction from the waste package is small compared to thermal radiation. • Average heat loads produce sub-boiling conditions in the host rock. • Repository edges do not significantly affect the near field host rock thermal conduction. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 7-3 October 2004 • Simultaneous emplacement of the waste packages, which is conservative with respect to total heat load. The following further assumptions are presented and defended in Section 5: • The chosen location for the drift is representative. • The thermal properties of a 21-PWR waste package are representative. • An initial water saturation of approximately 90.5% is representative. • The lithophysal porosity is 100% air-filled. • The thermophysical properties of 4-10 crushed tuff are representative of the invert. • The convection correlations for idealized configurations are adequate for the non-ideal configuration of the EBS, and the effects of the differences are captured in the uncertainty analysis. • The average temperature of the ventilation air at the inlet to the drift is equal to the ambient temperature of the host rock. The fact that these assumptions and simplification are defensible provides additional confidence in the results of the model calculations. 5. Consistency with physical principles, such as conservation of mass, energy, and momentum. [AP-2.27Q Attachment 3 Level I (c)] Consistency with physical principles is demonstrated by the conceptual and mathematical formulations for the mass and energy balance equations in Section 6.4 and the selection and use of the ANSYS code based on those physical principles. 7.2 CONFIDENCE BUILDING AFTER MODEL DEVELOPMENT TO SUPPORT THE SCIENTIFIC BASIS OF THE MODEL Level I validation must include at least one post-development method described in Paragraph 5.3.2 of AP-SIII.10Q. The governing TWP (BSC 2004 [DIRS 170950], Section 2.3.3) specifies one such activity, comparison of relevant model predictions with the results from the one-quarter scale ventilation tests (BSC 2003 [DIRS 160724]). This activity is described in Section 7.2.1. The validation criteria specified by the TWP are listed in Table 7-1. The following discussion, which describes the importance of the ventilation efficiency as a parameter and the criteria for acceptance of the comparison with test data, was provided by the TWP (BSC 2004 [DIRS 170950], Section 2.3.3). The purpose of the ventilation model is to support the engineering feasibility of the recommended primary value of the ventilation efficiency. It is important not to overestimate the magnitude of the ventilation efficiency, and it is not important if the ventilation efficiency is underestimated, as long as the model meets the requirements for Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 7-4 October 2004 confidence building during development. Therefore, the validation criterion for comparison with test data is that the model results do not overestimate (i.e., calculated efficiency is less than or equal to) the measured results from testing, expressed in terms of representative or ensemble values and accounting qualitatively for the effects of scale. This section shows not only that this validation criterion is met, but also that a previously proposed criterion is met: matching the model results to the test temperature data within 5°C (BSC 2003 [DIRS 2003 165601], Table 5). Past modeling of thermal performance of the repository has used values of 70 percent ventilation efficiency without generating concern for the effectiveness of ventilation, signifying that this efficiency is feasible. Therefore, a second validation criterion in the TWP is that the estimate of expected ventilation efficiency for the repository be more than 70 percent (BSC 2004 [DIRS 170950], Section 2.3.4). The second criterion is met by the results of Section 6, which provide average ventilation efficiencies of 86% and 88%, depending on the length of the drift. Table 7-1. Validation Criteria Activity Parameter Criterion Simulation of ventilation tests Ventilation efficiency Predicted efficiency does not exceed measured efficiency Simulation of repository ventilation Ventilation efficiency Predicted efficiency exceeds 70% Source: BSC 2004 [DIRS 170950], Section 2.4.3. The corroborating/supporting data used to complete model validation activities (and as direct input) are contained in Tables 4-1 through 4-5 (also used in Section 6 for the model development and application), 7-3 through 7-6, and IX-6 through IX-10. 7.2.1 One-Quarter Scale Ventilation Tests Phases 1 and 2 of the one-quarter scale ventilation tests were performed at the North Las Vegas Atlas Facility during 2001 and 2002. A detailed description of the ventilation tests is provided in the Phase 1 report (BSC 2003 [DIRS 160724], Section 2). The ventilation test train was constructed by connecting segments of concrete pipes. Twenty-five simulated waste packages were fabricated from steel pipe. A steel structure designed to simulate the current waste package support structure and emplacement pallet was used to support the simulated waste packages. Crushed tuff from Yucca Mountain was used as the invert ballast material. Electric heaters within the waste packages simulated the decay heat (BSC 2003 [DIRS 160724], Section 2.2.2). The test configuration is described in the test report (BSC 2003 [DIRS 160724], Sections 2.2.2.1 and 2.2.2.2). The test setup was nominally ¼ scale of a repository drift segment. The tests were conducted in two phases. The primary difference between Phases 1 and 2 is that the ventilation air in Phase 2 was conditioned to better control its inlet temperature and relative humidity. The Phase 1 test brought in ambient air from outside the test train that exhibited diurnal temperature changes of around 4°C. The same ventilation air flow rates and linear heat loads were used for both phases. Considering these aspects, that the ANSYS methodology for simulating ventilation does not account for the relative humidity of the in-drift air, and that the Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 7-5 October 2004 results of Section 6.9 show that the moisture has no significant impact on the ventilation, the Phase 1 test data are sufficient to provide the level of validation required for the ANSYS model. Therefore, the use of the ventilation test data for post-test ANSYS modeling and validation for this revision of the ventilation model report is confined to the Phase 1 cases. Table 7-2 lists the Phase 1 ventilation tests and cases for which ANSYS post-test modeling was performed. Table 7-2. Ventilation Phase 1 Test Matrix Case No. Nominal Flow (m3/s) Nominal Power (kW/m) 1 1 0.36 2 2 0.36 3 0.5 0.36 4 1 0.18 5 0.5 0.18 Source: BSC 2003 [DIRS 160724], Table 3-1. 7.2.2 Post-Test ANSYS Model Figure 7-1 shows the saddle-like temperature trends for the waste packages of Case 4 of Ventilation Test Phase 1. The same trend is observed in all the other cases. The temperature peaks that occur around Station 3 are due to heat losses at the inlet and outlet of the test train. However, the ANSYS methodology outlined in Section 6.4.1 is not capable of modeling the profile of axial temperature exhibited by the test data. An underlying assumption of the ANSYS methodology is that temperatures of the in-drift components, drift wall, and ventilation air are always increasing as the calculation proceeds down the length of the drift. This limitation forced the development of a two-dimensional ANSYS-based ventilation model. In other words, only a two-dimensional cross-section at Station 3 was modeled using ANSYS, rather than the ANSYS/Excel methodology described in Section 6.4.1 for a pseudo-three-dimensional analysis from Station 1 to Station 5. One consequence is that there is more model uncertainty in the validation calculations than in the methodology described in Section 6.4.1. 7.2.2.1 Mesh Figure 7-2 shows a detailed drawing through a cross-section of the test train. It also includes the relative locations of the instrumentation. Figure 7-3 shows the discretization of the test domain or the computational mesh used for the ANSYS post-test modeling. The pallet that supports the simulated waste package is not continuous in the test configuration. Rather, it supports only the ends of the package. The contribution of heat transfer via conduction from the package through the pallet and into the invert is considered to be negligible in comparison to the amount of heat transferred by radiation. For this reason, the pallet was not modeled. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 7-6 October 2004 25 30 35 40 45 50 55 60 65 70 0 10 20 30 40 50 Temperature (°C) right top bottom left Axial Distance (m) Station 3 Station 5 Station 1 Source: BSC 2003 [DIRS 160724], Figure 5-7. Figure 7-1. Ventilation Phase 1, Case 4 Waste Package Temperatures versus Axial Distance Down the Test Train for Data Recorded 10/15/00 Source: BSC 2003 [DIRS 160724], Figure 2-11. NOTE: All dimensions are in meters. Figure 7-2. Cross-Sectional View of the Ventilation Test Train Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 7-7 October 2004 Insulation Concrete Pipe Waste Package Invert Output DTN: MO0209MWDANS30.017, file vti.db of vti.tar.Z. NOTE: The plot was generated using the post-processor of ANSYS software and the model database file vti.db. Figure 7-3. Mesh of ANSYS Model 7.2.2.2 Thermophysical Properties for Model Validation Thermophysical properties of the invert, simulated waste package, concrete pipe, and insulation are listed in Sections 7.2.2.2.1 through 7.2.2.2.4. These properties are used in the model validation. 7.2.2.2.1 Thermophysical Properties of the Invert Table 7-3 lists the average thermophysical properties for the fine crushed tuff. The justification for the use of the material properties of fine crushed tuff for the invert ballast material is described in Section 5.5. Although only the volume specific heat is relevant to the heat transfer calculation, the ANSYS software does not accept that input. Instead, it requires the density and the mass specific heat. The particle density (2530 kg/m3) was input to the program in lieu of the bulk density. Because the particle density was also used to correct the volume specific heat to a mass specific heat of 363.24 (J/kg · K), the effective volume specific heat in the calculation was the same as the value shown in Table 7-3. Therefore, the choice of bulking factor had no effect on the calculation. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 7-8 October 2004 Table 7-3. Average Thermophysical Properties of the Invert Specific Heata Thermal Conductivityb Particle Densityc (J/cm3·°C) (W/m·°C) (g/cm3) 0.919 0.14 2.53 a DTN: GS000483351030.003 [DIRS 152932], average of specific heat values for fine crushed tuff, samples TK-FT-01 to TK-FT-10, rows 33-43. b DTN: GS000483351030.003 [DIRS 152932], average of thermal conductivity values for fine crushed tuff, samples TK-FT-01 to TK-FT-10, rows 33-43. c DTN: GS000383351030.002 [DIRS 148445], Table S00193_002, particle density for fine (crushed) Topopah Spring tuff, rows 16-21. 7.2.2.2.2 Thermophysical Properties of the Simulated Waste Package Table 7-4 lists thermophysical properties of the simulated waste package used in the ventilation tests performed at the North Las Vegas Atlas Facility. This information is used as input to the model validation exercises. Table 7-4. Thermophysical Properties of the Simulated Waste Package Property Value Source Density (kg/m3) 7840 Stroe 2001 [DIRS 155633], p. 3 for steel pipe (averaged over 20 to 50°C) Thermal Conductivity (W/m·K) 38.37 Stroe 2001 [DIRS 155633], p. 3 for steel pipe (averaged over 20 to 50°C) Specific Heat (J/kg·K) 410.98 Stroe 2001 [DIRS 155633], p. 3 for steel pipe (averaged over 20 to 50°C) Emissivity 0.8 Holman 1997 [DIRS 101978], Table A-10 for Sheet Steel Outside Diameter (in.) 16 CRWMS M&O 2000 [DIRS 153503] 7.2.2.2.3 Thermophysical Properties of the Concrete Pipe Table 7-5 lists thermophysical properties of the concrete pipe used in the ventilation tests performed at the North Las Vegas Atlas Facility. This information is used as input to the model validation exercises. Table 7-5. Thermophysical Properties of the Concrete Pipe Property Value Source Density (kg/m3) 2280 Stroe 2001 [DIRS 155633], p. 3 for Concrete Pipe (averaged over 20 to 50°C) Thermal Conductivity (W/m·K) 2.75 Stroe 2001 [DIRS 155633], p. 3 for Concrete Pipe (averaged over 20 to 50°C) Specific Heat (J/kg·K) 1016.16 Stroe 2001 [DIRS 155633], p. 3 for Concrete Pipe (averaged over 20 to 50°C) Emissivity 0.93 Incropera and DeWitt 1996 [DIRS 108184], Table A.11 for Concrete Inner Diameter (in.) 54 CRWMS M&O 2000 [DIRS 153503] Outside Diameter (in.) 65 CRWMS M&O 2000 [DIRS 153503] (Wall Thickness [5.5”] + Inner Diameter) Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 7-9 October 2004 7.2.2.2.4 Thermophysical Properties of the Insulation Table 7-6 lists thermophysical properties of the insulation used in the ventilation tests performed at the North Las Vegas Atlas Facility. This information is used as input to the model validation exercises. Table 7-6. Thermophysical Properties of the Insulation Property Value Source Density (kg/m3) 12 CertainTeed 1996 [DIRS 153512] for Type 75 Standard Fiber Glass Duct Wrap Thermal Conductivity (W/m·K) 0.04 CertainTeed 1996 [DIRS 153512] for Type 75 Standard Fiber Glass Duct Wrap Specific Heat (J/kg·K) 700 Holman 1997 [DIRS 101978], Table A-3 for Insulation Thickness (m) 0.0508 CRWMS M&O 2000 [DIRS 153503] (Converted from 2 in.) 7.2.2.3 Boundary Conditions The recorded temperatures on the outer insulation at Station 3 of the test train were the basis for the outer boundary conditions for each ANSYS post-test model (DTN: SN0208F3409100.007 [DIRS 161729], worksheets “RTDs” of vent_test_C1.xls [case 1], vent_test_C2.xls [case 2], vent_test_C3.xls [case 3], vent_test_C4.xls [case 4], vent_test_C5.xls [case 5]). Each test case had a different set of recorded temperatures over its life span. To aid in the implementation of each outer boundary condition, the several hundred time-stamped results from the test were replaced by a data fit. The data fit consisted of linear interpolations between twenty-four times, chosen to capture the fluctuations in that test’s boundary conditions. Figure 7-4 is an example of a working plot that was used to choose the data fit for an outer insulation temperature history. The scatter seen in the temperature data for cases 4 and 5 (Tables 7-7d and 7-7e) was caused by the HVAC system at the test facility, and is representative of test conditions (BSC 2003 [DIRS 160724], Section 3.3.3.6). The measured outer insulation temperature histories contained in the output DTN: MO0410MWDANS30.018 (worksheets “Measured Air and Insu Temp” of vti-aa.xls [case 4], vti-ba.xls [case 5], vti-ca.xls [case 1], vti-da.xls [case 2], vti-ea.xls [case 3]) were imported from DTN: SN0208F3409100.007 ([DIRS 161729], worksheets “RTDs” of vent_test_C1.xls [case 1], vent_test_C2.xls [case 2], vent_test_C3.xls [case 3], vent_test_C4.xls [case 4], vent_test_C5.xls [case 5]) for the data fitting purposes. The “fitted data” were all from the measured values, and selected using visual observation to best represent the bounds and variations of the measured outer insulation temperature histories. These selected (fitted) data from measurements, listed in Tables 7-7a through 7-7e, were then used as inputs for the outer boundary conditions in the ANSYS post-test modeling. The temperatures at the side of outer insulation, as listed in Column E of Tables 7-7a to 7-7e and shown in Figure 7-4, are the average values of those measured on the left and right sides at Station 3. The ANSYS methodology requires that an inlet ventilation air stream temperature be specified. Therefore, the temperatures of the ventilation air stream recorded at Station 3 were used as input to the ANSYS post-test models (DTN: SN0208F3409100.007 [DIRS 161729], worksheets “RTDs” of vent_test_C1.xls, vent_test_C2.xls, vent_test_C3.xls, vent_test_C4.xls, vent_test_C5.xls). Each test case had a different set of air stream temperature histories. Again, a Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 7-10 October 2004 data fit for each case was performed on the recorded temperature data to simplify its implementation into the models. Figure 7-5 is an example of a working plot that was used to choose the data fit for a ventilation air stream temperature history. The measured ventilation airstream temperature histories contained in the output DTN: MO0410MWDANS30.018 (worksheets “Measured Air and Insu Temp” of vti-aa.xls [case 4], vti-ba.xls [case 5], vti-ca.xls [case 1], vti-da.xls [case 2], vti-ea.xls [case 3]) were imported from DTN: SN0208F3409100.007 ([DIRS 161729], worksheets “RTDs” of vent_test_C1.xls [case 1], vent_test_C2.xls [case 2], vent_test_C3.xls [case 3], vent_test_C4.xls [case 4], vent_test_C5.xls [case 5]) for the data fitting purposes. Similarly, the fitted data were all from the measured values, and selected based on visual observation to best represent the bounds and variations of the measured air stream temperature histories. These selected (fitted) data from measurements were then used as inputs for ventilation air stream temperatures in the ANSYS post-test modeling. The temperatures shown in Figure 7-5 are the average values of those measured near the left and right sides of concrete pipe (inside) at Station 3. The simulated waste packages were hollow rolled steel tubes, with heater rods suspended concentrically inside. Due to the nature of the experimental set-up, natural convection cells developed within the placid annulus of the waste packages. This caused a non-uniform heat flux, and hence temperature distribution, around the circumference of the waste package. No temperature measurements were recorded inside the waste package (i.e., the annulus air or the rod-heater). Rather than model the complexity of the natural convection inside of the waste package, the ANSYS model supplied a heat flux at the waste package wall. The heat flux was partitioned around the waste package circumference using the recorded steady-state temperature distributions as a basis. Table 7-8 summarizes the distributions for the test cases listed in Table 7-2. The validity of this partitioning methodology is confirmed by the consistency of the calculated distributions from case to case. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 7-11 October 2004 Table 7-7a. Ventilation Test Phase 1, Case 1, Outer Insulation and Air Temperatures Measured at Station 3 Measured Outer Insulation Temperature at Station 3 (°C) Measured Air Temperature at Station 3 (°C) V3-RTD-03 V3-RTD-04 V3-RTD-05 V3-RTD-06 Average of RTDs –04 and 06 V3-RTD-01 V3-RTD-02 Average of RTDs-01 and 02 Date / Time A B C D E ([B+D]/2) F G H ([F+G]/2) 11/3/2000 14:03 29.14 28.78 28.42 28.08 28.43 27.34 27.36 27.35 11/3/2000 14:18 29.15 28.78 28.79 28.33 28.55 27.68 27.54 27.61 11/3/2000 14:49 29.17 28.32 28.05 27.72 28.02 28.40 28.36 28.38 11/3/2000 15:18 28.11 27.80 27.51 27.44 27.62 29.37 29.03 29.20 11/3/2000 15:48 27.39 27.30 27.03 26.97 27.13 29.46 29.23 29.34 11/3/2000 16:18 27.81 28.41 28.56 27.66 28.03 30.03 29.97 30.00 11/3/2000 16:48 27.75 28.43 28.14 27.75 28.09 30.28 30.14 30.21 11/3/2000 17:18 27.37 28.05 27.78 27.46 27.76 30.42 30.14 30.28 11/3/2000 17:49 27.30 27.75 27.54 27.28 27.52 30.40 30.16 30.28 11/3/2000 18:18 27.08 27.76 27.40 27.19 27.47 30.23 30.19 30.21 11/3/2000 18:49 27.07 27.51 27.24 27.26 27.39 30.50 29.91 30.21 11/3/2000 19:33 26.81 27.39 27.19 26.97 27.18 30.15 29.98 30.07 11/4/2000 7:03 27.16 25.68 24.44 25.23 25.45 29.04 28.77 28.90 11/4/2000 15:48 29.65 29.32 28.21 29.22 29.27 31.76 31.75 31.76 11/5/2000 6:33 28.22 26.60 24.73 25.61 26.11 29.22 29.11 29.16 11/5/2000 11:48 30.99 29.83 29.22 30.04 29.93 32.08 32.10 32.09 11/6/2000 8:03 27.23 26.45 24.80 25.61 26.03 28.93 28.83 28.88 11/6/2000 13:48 31.87 30.04 29.30 29.98 30.01 32.57 32.28 32.43 11/7/2000 6:48 26.41 25.14 23.74 24.27 24.70 27.97 28.03 28.00 11/7/2000 13:33 29.07 28.14 27.60 28.13 28.13 31.18 31.34 31.26 11/8/2000 6:48 27.05 24.84 22.95 23.80 24.32 27.65 27.47 27.56 11/8/2000 13:03 30.93 28.45 28.03 28.39 28.42 31.49 31.18 31.34 11/9/2000 6:18 27.09 25.76 23.95 24.78 25.27 28.08 28.09 28.09 11/9/2000 16:03 29.09 28.32 27.56 28.31 28.31 31.74 31.66 31.70 DTN: SN0208F3409100.007 [DIRS 161729], selected measurements from worksheet “RTDs” of vent_test_C1.xls. Averages from DTN: MO0410MWDANS30.018, vti-ca.xls, worksheet “Measured Air and Insu Temp.” Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 7-12 October 2004 Table 7-7b. Ventilation Test Phase 1, Case 2, Outer Insulation and Air Temperatures Measured at Station 3 Measured Outer Insulation Temperature at Station 3 (°C) Measured Air Temperature at Station 3 (°C) V3-RTD-03 V3-RTD-04 V3-RTD-05 V3-RTD-06 Average of RTDs –04 and 06 V3-RTD-01 V3-RTD-02 Average of RTDs –01 and 02 Date / Time A B C D E ([B+D]/2) F G H ([F+G]/2) 11/20/2000 16:03 26.14 24.60 22.42 24.11 24.35 25.91 25.43 25.67 11/20/2000 16:18 26.98 24.40 22.55 24.39 24.40 26.29 25.99 26.14 11/20/2000 16:48 25.01 23.90 22.60 24.06 23.98 25.82 25.49 25.66 11/20/2000 17:19 24.08 23.77 22.58 22.71 23.24 25.08 24.80 24.94 11/20/2000 17:49 25.89 23.51 22.40 23.20 23.35 26.13 25.52 25.83 11/20/2000 18:18 25.97 24.20 22.63 23.36 23.78 25.98 25.70 25.84 11/20/2000 18:49 25.65 23.71 22.43 23.13 23.42 25.98 25.72 25.85 11/20/2000 19:19 25.81 23.31 22.27 23.27 23.29 26.13 25.74 25.93 11/20/2000 19:49 25.27 23.25 22.71 23.03 23.14 25.87 25.52 25.70 11/20/2000 20:19 25.07 23.65 22.52 22.87 23.26 25.76 25.33 25.54 11/21/2000 6:03 24.91 23.59 21.39 22.69 22.98 24.80 24.42 24.61 11/21/2000 14:03 29.57 27.78 24.80 28.23 22.85 29.74 29.30 29.52 11/22/2000 6:33 23.09 22.24 21.21 21.58 24.10 23.55 23.09 23.32 11/22/2000 13:49 29.19 27.41 26.00 27.57 25.58 29.17 29.02 29.10 11/23/2000 5:03 24.77 23.92 23.29 23.68 24.80 25.12 24.78 24.95 11/23/2000 14:48 30.62 29.66 27.00 29.87 27.49 31.38 30.73 31.05 11/24/2000 6:19 25.22 24.12 23.16 23.72 27.10 25.44 24.82 25.13 11/24/2000 14:18 30.47 29.53 27.35 29.69 27.91 30.98 30.58 30.78 11/25/2000 6:48 25.42 24.01 23.51 23.50 25.30 24.85 24.37 24.61 11/25/2000 14:03 30.32 29.22 27.79 29.43 28.15 30.61 30.40 30.51 11/26/2000 6:33 24.33 23.49 22.78 22.66 26.59 24.45 24.09 24.27 11/26/2000 13:33 29.78 28.85 27.48 29.01 28.72 30.67 30.32 30.49 11/27/2000 6:48 24.61 23.95 23.40 23.20 25.64 25.03 24.59 24.81 11/27/2000 11:33 28.96 27.22 26.22 27.17 25.80 28.81 28.19 28.50 DTN: SN0208F3409100.007 [DIRS 161729], selected measurements from worksheet “RTDs” of vent_test_C2.xls. Averages from DTN: MO0410MWDANS30.018, vti-da.xls, worksheet “Measured Air and Insu Temp.” Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 7-13 October 2004 Table 7-7c. Ventilation Test Phase 1, Case 3, Outer Insulation and Air Temperatures Measured at Station 3 Measured Outer Insulation Temperature at Station 3 (°C) Measured Air Temperature at Station 3 (°C) V3-RTD-03 V3-RTD-04 V3-RTD-05 V3-RTD-06 Average of RTDs –04 and 06 V3-RTD-01 V3-RTD-02 Average of RTDs –01 and 02 Date / Time A B C D E ([B+D]/2) F G H ([F+G]/2) 12/1/2000 15:03 20.39 20.82 20.61 20.41 20.61 21.10 20.84 20.97 12/1/2000 15:19 20.38 20.22 20.69 20.30 20.26 21.44 21.42 21.43 12/1/2000 15:48 20.00 19.98 20.46 20.31 20.15 22.77 23.04 22.91 12/1/2000 16:18 19.86 20.05 20.06 19.92 19.98 23.98 23.99 23.99 12/1/2000 16:48 19.51 19.10 19.54 19.30 19.20 24.89 24.81 24.85 12/1/2000 17:18 18.76 18.96 19.15 18.98 18.97 25.27 25.11 25.19 12/1/2000 17:49 20.49 21.30 21.45 20.60 20.95 25.46 25.26 25.36 12/1/2000 18:19 18.79 18.90 19.25 18.96 18.93 26.02 25.95 25.99 12/1/2000 18:48 21.13 21.81 22.32 21.38 21.59 26.18 26.07 26.13 12/1/2000 19:19 18.52 18.49 18.58 18.49 18.49 26.29 26.35 26.32 12/2/2000 7:03 22.42 22.90 21.78 22.05 22.47 27.75 27.45 27.60 12/2/2000 15:48 23.60 23.18 22.31 23.27 23.23 30.83 30.61 30.72 12/3/2000 7:18 19.62 18.37 17.99 18.81 18.59 28.36 28.39 28.37 12/3/2000 14:33 25.63 24.85 23.92 24.96 24.90 32.31 31.98 32.14 12/4/2000 7:19 21.23 20.53 20.08 21.08 20.80 28.74 28.78 28.76 12/4/2000 14:49 26.25 25.38 24.53 25.98 25.68 32.54 32.44 32.49 12/5/2000 7:18 22.20 21.96 21.48 22.18 22.07 29.14 29.32 29.23 12/5/2000 15:18 25.17 23.81 23.56 24.31 24.06 32.55 32.42 32.48 12/6/2000 6:34 23.07 22.58 20.64 21.83 22.21 28.97 29.10 29.04 12/6/2000 14:33 25.27 24.12 23.40 24.37 24.25 31.98 32.03 32.01 12/7/2000 7:18 24.60 24.00 23.32 24.28 24.14 29.95 30.04 30.00 12/7/2000 14:33 25.23 24.01 23.47 24.74 24.38 32.65 32.96 32.81 12/8/2000 7:03 22.08 21.63 20.88 21.62 21.62 29.39 29.55 29.47 12/8/2000 11:48 26.08 25.09 24.60 25.56 25.32 32.08 32.08 32.08 DTN: SN0208F3409100.007 [DIRS 161729], selected measurements from worksheet “RTDs” of vent_test_C3.xls. Averages from DTN: MO0410MWDANS30.018, vti-ea.xls, worksheet “Measured Air and Insu Temp.” Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 7-14 October 2004 Table 7-7d. Ventilation Test Phase 1, Case 4, Outer Insulation and Air Temperatures Measured at Station 3 Measured Outer Insulation Temperature at Station 3 (°C) Measured Air Temperature at Station 3 (°C) V3-RTD-03 V3-RTD-04 V3-RTD-05 V3-RTD-06 Average of RTDs –04 and 06 V3-RTD-10 V3-RTD-02 Average of RTDs –10 and 02 Date / Time A B C D E ([B+D]/2) F G H ([F+G]/2) 10/9/2000 8:48 35.04 28.39 27.52 29.27 28.83 26.80 26.45 26.62 10/9/2000 9:19 26.45 26.78 26.39 26.33 26.55 27.32 27.08 27.20 10/9/2000 9:49 28.60 27.80 27.30 28.04 27.92 27.66 27.39 27.53 10/9/2000 10:18 27.72 26.88 26.91 27.54 27.21 28.25 27.84 28.04 10/9/2000 10:48 27.40 27.18 26.91 27.35 27.27 28.51 28.28 28.40 10/9/2000 11:19 28.14 27.33 27.28 27.52 27.42 29.12 28.50 28.81 10/9/2000 11:49 28.60 27.79 27.54 27.84 27.81 29.26 28.83 29.05 10/9/2000 12:18 28.34 27.88 27.57 28.33 28.10 29.48 29.54 29.51 10/9/2000 12:48 28.13 27.85 27.42 27.83 27.84 29.66 29.46 29.56 10/9/2000 13:33 28.50 27.85 27.64 28.36 28.11 29.82 29.48 29.65 10/10/2000 5:48 35.53 28.73 25.71 28.18 28.45 26.42 25.97 26.20 10/10/2000 14:19 27.09 26.92 26.69 26.97 26.95 29.04 28.63 28.83 10/11/2000 5:03 37.48 25.34 25.62 28.81 27.07 25.38 25.19 25.28 10/11/2000 12:48 27.93 27.23 26.83 27.40 27.32 28.61 28.36 28.49 10/12/2000 3:48 30.78 26.06 24.92 26.70 26.38 25.70 25.36 25.53 10/12/2000 15:33 34.28 28.56 27.93 29.94 29.25 28.41 28.33 28.37 10/13/2000 5:33 33.54 26.48 24.99 27.95 27.21 25.42 25.01 25.21 10/13/2000 13:33 36.86 28.41 27.91 30.40 29.40 29.13 28.85 28.99 10/14/2000 5:18 25.68 25.42 24.32 24.49 24.96 26.00 25.62 25.81 10/14/2000 14:33 29.05 28.84 27.80 28.86 28.85 29.43 29.13 29.28 10/15/2000 5:33 25.58 25.60 24.57 24.65 25.13 26.07 25.44 25.75 10/15/2000 12:33 28.39 26.87 27.26 27.34 27.10 29.44 29.21 29.33 10/16/2000 5:03 32.50 27.41 26.23 28.70 28.05 26.19 25.60 25.89 10/16/2000 8:03 29.04 27.63 27.26 27.66 27.65 28.26 27.68 27.97 DTN: SN0208F3409100.007 [DIRS 161729], selected measurements from worksheet “RTDs” of vent_test_C4.xls. Averages from DTN: MO0410MWDANS30.018, vti-aa.xls, worksheet “Measured Air and Insu Temp.” Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 7-15 October 2004 Table 7-7e. Ventilation Test Phase 1, Case 5, Outer Insulation and Air Temperatures Measured at Station 3 Measured Outer Insulation Temperature at Station 3 (°C) Measured Air Temperature at Station 3 (°C) V3-RTD-03 V3-RTD-04 V3-RTD-05 V3-RTD-06 Average of RTDs –04 and 06 V3-RTD-10 V3-RTD-02 Average of RTDs –10 and 02 Date / Time A B C D E ([B+D]/2) F G H ([F+G]/2) 10/20/2000 8:33 28.89 27.33 27.20 27.44 27.39 26.54 26.25 26.40 10/20/2000 9:03 26.66 25.64 26.08 25.66 25.65 27.10 26.91 27.00 10/20/2000 9:33 28.10 27.41 27.04 27.64 27.52 28.13 27.85 27.99 10/20/2000 10:03 27.63 25.82 26.27 26.05 25.94 28.49 28.39 28.44 10/20/2000 11:03 27.43 26.63 26.50 26.05 26.34 29.28 29.24 29.26 10/20/2000 13:03 28.19 26.94 26.93 27.27 27.10 30.38 29.79 30.08 10/20/2000 20:33 27.77 27.81 27.15 28.25 28.03 30.63 30.64 30.63 10/20/2000 23:18 26.30 26.75 25.92 26.71 26.73 30.30 29.99 30.14 10/21/2000 5:33 26.29 26.58 25.29 26.52 26.55 29.63 29.59 29.61 10/21/2000 10:18 29.61 28.41 27.69 28.80 28.60 31.12 31.17 31.15 10/21/2000 15:48 29.08 28.95 28.12 28.81 28.88 31.49 31.49 31.49 10/22/2000 4:48 28.28 26.95 25.15 26.34 26.65 29.88 29.83 29.85 10/22/2000 12:33 27.61 27.27 26.08 26.85 27.06 30.30 30.01 30.15 10/22/2000 23:18 31.39 27.82 25.40 28.27 28.04 29.17 29.06 29.11 10/23/2000 3:33 32.31 27.53 26.10 28.69 28.11 28.97 29.08 29.03 10/23/2000 15:18 36.94 30.51 26.30 29.27 29.89 29.08 29.24 29.16 10/24/2000 5:18 26.47 25.21 23.93 24.97 25.09 28.59 28.49 28.54 10/24/2000 13:18 35.20 29.25 28.52 30.44 29.85 30.85 30.99 30.92 10/25/2000 6:19 30.79 28.14 25.58 28.70 28.42 28.98 29.04 29.01 10/25/2000 14:18 28.28 27.23 26.99 27.40 27.32 30.98 31.28 31.13 10/26/2000 6:48 35.53 29.25 27.59 30.26 29.76 29.40 29.61 29.51 10/26/2000 12:18 28.52 28.02 27.86 28.09 28.06 30.72 30.85 30.79 10/26/2000 18:48 28.99 28.63 27.97 28.40 28.52 31.44 31.54 31.49 10/27/2000 7:48 29.06 27.91 25.50 26.59 27.25 30.57 30.51 30.54 DTN: SN0208F3409100.007 [DIRS 161729], selected measurements from worksheet “RTDs” of vent_test_C5.xls. Averages from DTN: MO0410MWDANS30.018, vti-ba.xls, worksheet “Measured Air and Insu Temp.” Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 7-16 October 2004 Temperatures on Insulation Surface (I; #3; HL=0.36kW/m; FR=1.0m3/s; a) 15 20 25 30 35 0 24 48 72 96 120 144 168 Time (Hours) Temperatures (Degrees C) Top (Used) Side (Used) Bottom (Used) Top (Measured) Side (Measured) Bottom (Measured) C DTN: MO0410MWDANS30.018, worksheet “Measured Air and Insu Temp” of file vti-ca.xls. NOTE: Data used as input are connected with smooth curves for clarity, but ANSYS uses linear interpolation. Figure 7-4. Example (Case 1) of Working Plot for Fitting Outer Insulation Boundary Temperatures for the ANSYS Post-Test Ventilation Model Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 7-17 October 2004 Air Temperatures near Side of Waste Package (I; #3; HL=0.36kW/m; FR=1.0m3/s; a) 25 27 29 31 33 35 0 24 48 72 96 120 144 168 Time (Hours) Temperatures (Degrees C) Used Measured DTN: MO0410MWDANS30.018, worksheet “Measured Air and Insu Temp” of file vti-ca.xls NOTE: Data used as input are connected with smooth curves for clarity, but ANSYS uses linear interpolation. Figure 7-5. Example (Case 1) of Working Plot for Fitting Temperatures at Station 3 for Use as Inlet Air for the ANSYS Post-Test Ventilation Model Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 7-18 October 2004 Table 7-8. Distribution of Total Power to the Top, Sides, and Bottom Quarters of the Waste Package Based on Temperature Measurements Case No. WP Top Quarter (%) WP Side Quarters (%) WP Bottom Quarter (%) 1 32% 24% 20% 2 32% 24% 20% 3 31% 24% 21% 4 32% 24% 20% 5 32% 24% 20% DTN: MO0410MWDANS30.018; worksheets “Heat Removal”, rows 10 to 13 of column B of vti-aa.xls (case 4), vti-ba.xls (case 5), vti-ca.xls (case 1), vti-da.xls (case 2), vti-ea.xls (case 3). 7.2.2.4 Correlating the Model Results to the Test Data Using Heat Transfer Coefficients Having determined appropriate distributions of power around the circumference of the waste package, ANSYS models were run iteratively using different values for the heat transfer coefficients until the model results matched the test data. 7.2.2.5 Results Table 7-9 shows the heat transfer coefficient values which resulted in close agreement to the measured temperature data. The temperature results from the ANSYS models are compared to the recorded test data in Figures 7-6 through 7-10. Table 7-10 compares the fitted average heat transfer coefficient for each test case from Table 7-9 to heat transfer coefficients calculated using the Mixed Convection Correlation and the Dittus-Boelter correlation for fully developed turbulent flow inside a smooth circular tube (Incropera and DeWitt 1996 [DIRS 108184], Section 8.5). The Dittus-Boelter formula gives the asymptotic Nusselt number for fully developed turbulent flow (Re > 10,000) in circular tubes. This formula was used for the Nusselt number at the outer wall in earlier ventilation model calculations with fully developed turbulent flow. Incropera and DeWitt (1985 [DIRS 114109], p. 400) consider the Dittus-Boelter equation to be a first approximation, in which the inner and outer convection coefficients are assumed to be equal. Appendix XVII presents the calculation of Dittus-Boelter heat transfer coefficients for the ventilation test Phase 1 cases 1 through 5. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 7-19 October 2004 Case 1 - Top of Exterior Concrete 20 25 30 35 40 45 0 24 48 72 96 120 144 168 Time (Hours) Temperatures (°C) Sta. 3 (Calculated) Sta. 3 (Measured) Case 1 - Side of Waste Package 20 30 40 50 60 70 80 0 24 48 72 96 120 144 168 Time (Hours) Temperatures (°C) Sta. 3 (Calculated) Sta. 3 (Measured) Case 1 - Bottom of Waste Package 20 30 40 50 60 70 80 0 24 48 72 96 120 144 168 Time (Hours) Temperatures (°C) Sta. 3 (Calculated) Sta. 3 (Measured) Case 1 - Side of Interior Concrete 20 25 30 35 40 45 0 24 48 72 96 120 144 168 Time (Hours) Temperatures (°C) Sta. 3 (Calculated) Sta. 3 (Measured) Case 1 - Bottom of Interior Concrete 20 25 30 35 40 45 0 24 48 72 96 120 144 168 Time (Hours) Temperatures (°C) Sta. 3 (Calculated) Sta. 3 (Measured) Case 1 - Side of Exterior Concrete 20 25 30 35 40 45 0 24 48 72 96 120 144 168 Time (Hours) Temperatures (°C) Sta. 3 (Calculated) Sta. 3 (Measured) Case 1 - Bottom of Exterior Concrete 20 25 30 35 40 45 0 24 48 72 96 120 144 168 Time (Hours) Temperatures (°C) Sta. 3 (Calculated) Sta. 3 (Measured) Case 1 - Top of Waste Package 20 30 40 50 60 70 80 0 24 48 72 96 120 144 168 Time (Hours) Temperatures (°C) Sta. 3 (Calculated) Sta. 3 (Measured) Case 1 - Top of Interior Concrete 20 25 30 35 40 45 0 24 48 72 96 120 144 168 Time (Hours) Temperatures (°C) Sta. 3 (Calculated) Sta. 3 (Measured) DTN: MO0410MWDANS30.018MO0209MWDANS30.017, data in worksheet “raw data” of vti-c-data.xls and worksheet “ANSYS Results” of vti-ca.xls; plots in worksheet “ANSYS Results” of vti-ca.xls and modified for clarity. Figure 7-6. ANSYS Post-Test Ventilation Model versus Measured Results for Ventilation Test Phase 1, Case 1 Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 7-20 October 2004 Case 2 - Top of Waste Package 20 30 40 50 60 70 0 24 48 72 96 120 144 168 Time (Hours) Temperatures (°C) Sta. 3 (Calculated) Sta. 3 (Measured) Case 2 - Top of Interior Concrete 20 25 30 35 0 24 48 72 96 120 144 168 Time (Hours) Temperatures (°C) Sta. 3 (Calculated) Sta. 3 (Measured) Case 2 - Top of Exterior Concrete 20 25 30 35 0 24 48 72 96 120 144 168 Time (Hours) Temperatures (°C) Sta. 3 (Calculated) Sta. 3 (Measured) Case 2 - Side of Waste Package 20 30 40 50 60 70 0 24 48 72 96 120 144 168 Time (Hours) Temperatures (°C) Sta. 3 (Calculated) Sta. 3 (Measured) Case 2 - Bottom of Waste Package 20 30 40 50 60 70 0 24 48 72 96 120 144 168 Time (Hours) Temperatures (°C) Sta. 3 (Calculated) Sta. 3 (Measured) Case 2 - Side of Interior Concrete 20 25 30 35 0 24 48 72 96 120 144 168 Time (Hours) Temperatures (°C) Sta. 3 (Calculated) Sta. 3 (Measured) Case 2 - Bottom of Interior Concrete 20 25 30 35 0 24 48 72 96 120 144 168 Time (Hours) Temperatures (°C) Sta. 3 (Calculated) Sta. 3 (Measured) Case 2 - Side of Exterior Concrete 20 25 30 35 0 24 48 72 96 120 144 168 Time (Hours) Temperatures (°C) Sta. 3 (Calculated) Sta. 3 (Measured) Case 2 - Bottom of Exterior Concrete 20 25 30 35 0 24 48 72 96 120 144 168 Time (Hours) Temperatures (°C) Sta. 3 (Calculated) Sta. 3 (Measured) DTN: MO0410MWDANS30.018MO0209MWDANS30.017, data in worksheet “raw data” of vti-d-data.xls and worksheet “ANSYS Results” of vti-da.xls; plots in worksheet “ANSYS Results” of vti-da.xls and modified for clarity. Figure 7-7. ANSYS Post-Test Ventilation Model versus Measured Results for Ventilation Test Phase 1, Case 2 Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 7-21 October 2004 Case 3 - Top of Waste Package 20 30 40 50 60 70 80 0 24 48 72 96 120 144 168 Time (Hours) Temperatures (°C) Sta. 3 (Calculated) Sta. 3 (Measured) Case 3 - Top of Interior Concrete 20 25 30 35 40 0 24 48 72 96 120 144 168 Time (Hours) Temperatures (°C) Sta. 3 (Calculated) Sta. 3 (Measured) Case 3 - Top of Exterior Concrete 20 25 30 35 40 0 24 48 72 96 120 144 168 Time (Hours) Temperatures (°C) Sta. 3 (Calculated) Sta. 3 (Measured) Case 3 - Side of Waste Package 20 30 40 50 60 70 80 0 24 48 72 96 120 144 168 Time (Hours) Temperatures (°C) Sta. 3 (Calculated) Sta. 3 (Measured) Case 3 - Bottom of Waste Package 20 30 40 50 60 70 80 0 24 48 72 96 120 144 168 Time (Hours) Temperatures (°C) Sta. 3 (Calculated) Sta. 3 (Measured) Case 3 - Side of Interior Concrete 20 25 30 35 40 0 24 48 72 96 120 144 168 Time (Hours) Temperatures (°C) Sta. 3 (Calculated) Sta. 3 (Measured) Case 3 - Bottom of Interior Concrete 20 25 30 35 40 0 24 48 72 96 120 144 168 Time (Hours) Temperatures (°C) Sta. 3 (Calculated) Sta. 3 (Measured) Case 3 - Side of Exterior Concrete 20 25 30 35 40 0 24 48 72 96 120 144 168 Time (Hours) Temperatures (°C) Sta. 3 (Calculated) Sta. 3 (Measured) Case 3 - Bottom of Exterior Concrete 20 25 30 35 40 0 24 48 72 96 120 144 168 Time (Hours) Temperatures (°C) Sta. 3 (Calculated) Sta. 3 (Measured) DTN: MO0410MWDANS30.018, data in worksheet “raw data” of vti-e-data.xls and worksheet “ANSYS Results” of vti-ea.xls; plots in worksheet “ANSYS Results” of vti-ea.xls and modified for clarity. Figure 7-8. ANSYS Post-Test Ventilation Model versus Measured Results for Ventilation Test Phase 1, Case 3 Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 7-22 October 2004 Case 4 - Top of Waste Package 20 30 40 50 60 0 24 48 72 96 120 144 168 Time (Hours) Temperatures (°C) Sta. 3 (Calculated) Sta. 3 (Measured) Case 4 - Top of Interior Concrete 20 25 30 35 40 45 0 24 48 72 96 120 144 168 Time (Hours) Temperatures (°C) Sta. 3 (Calculated) Sta. 3 (Measured) Case 4 - Top of Exterior Concrete 20 25 30 35 40 45 0 24 48 72 96 120 144 168 Time (Hours) Temperatures (°C) Sta. 3 (Calculated) Sta. 3 (Measured) Case 4 - Side of Waste Package 20 30 40 50 60 0 24 48 72 96 120 144 168 Time (Hours) Temperatures (°C) Sta. 3 (Calculated) Sta. 3 (Measured) Case 4 - Bottom of Waste Package 20 30 40 50 60 0 24 48 72 96 120 144 168 Time (Hours) Temperatures (°C) Sta. 3 (Calculated) Sta. 3 (Measured) Case 4 - Side of Interior Concrete 20 25 30 35 40 45 0 24 48 72 96 120 144 168 Time (Hours) Temperatures (°C) Sta. 3 (Calculated) Sta. 3 (Measured) Case 4 - Bottom of Interior Concrete 20 25 30 35 40 45 0 24 48 72 96 120 144 168 Time (Hours) Temperatures (°C) Sta. 3 (Calculated) Sta. 3 (Measured) Case 4 - Side of Exterior Concrete 20 25 30 35 40 45 0 24 48 72 96 120 144 168 Time (Hours) Temperatures (°C) Sta. 3 (Calculated) Sta. 3 (Measured) Case 4 - Bottom of Exterior Concrete 20 25 30 35 40 45 0 24 48 72 96 120 144 168 Time (Hours) Temperatures (°C) Sta. 3 (Calculated) Sta. 3 (Measured) DTN: MO0410MWDANS30.018, data in worksheet “raw data” of vti-a-data.xls and worksheet “ANSYS Results” of vti-aa.xls; plots in the worksheet “ANSYS Results” of vti-aa.xls and modified for clarity. Figure 7-9. ANSYS Post-Test Ventilation Model versus Measured Results for Ventilation Test Phase 1, Case 4 Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 7-23 October 2004 Case 5 - Top of Waste Package 20 30 40 50 60 0 24 48 72 96 120 144 168 Time (Hours) Temperatures (°C) Sta. 3 (Calculated) Sta. 3 (Measured) Case 5 - Top of Interior Concrete 20 25 30 35 40 45 0 24 48 72 96 120 144 168 Time (Hours) Temperatures (°C) Sta. 3 (Calculated) Sta. 3 (Measured) Case 5 - Top of Exterior Concrete 20 25 30 35 40 45 0 24 48 72 96 120 144 168 Time (Hours) Temperatures (°C) Sta. 3 (Calculated) Sta. 3 (Measured) Case 5 - Side of Waste Package 20 30 40 50 60 0 24 48 72 96 120 144 168 Time (Hours) Temperatures (°C) Sta. 3 (Calculated) Sta. 3 (Measured) Case 5 - Bottom of Waste Package 20 30 40 50 60 0 24 48 72 96 120 144 168 Time (Hours) Temperatures (°C) Sta. 3 (Calculated) Sta. 3 (Measured) Case 5 - Side of Interior Concrete 20 25 30 35 40 45 0 24 48 72 96 120 144 168 Time (Hours) Temperatures (°C) Sta. 3 (Calculated) Sta. 3 (Measured) Case 5 - Bottom of Interior Concrete 20 25 30 35 40 45 0 24 48 72 96 120 144 168 Time (Hours) Temperatures (°C) Sta. 3 (Calculated) Sta. 3 (Measured) Case 5 - Side of Exterior Concrete 20 25 30 35 40 45 0 24 48 72 96 120 144 168 Time (Hours) Temperatures (°C) Sta. 3 (Calculated) Sta. 3 (Measured) Case 5 - Bottom of Exterior Concrete 20 25 30 35 40 45 0 24 48 72 96 120 144 168 Time (Hours) Temperatures (°C) Sta. 3 (Calculated) Sta. 3 (Measured) DTN: MO0410MWDANS30.018, data in worksheet “raw data” of vti-b-data.xls and worksheet “ANSYS Results” of vti-ba.xls; plots in the worksheet “ANSYS Results” of vti-ba.xls and modified for clarity. Figure 7-10. ANSYS Post-Test Ventilation Model versus Measured Results for Ventilation Test Phase 1, Case 5 Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 7-24 October 2004 Table 7-9. Developed Heat Transfer Coefficients from the ANSYS Post-Test Modeling of Phase 1 of the Ventilation Test Heat Transfer Coefficient (W/m2·K) Case No. WP Top Quarter WP Side Quarters WP Bottom Quarter Upper Concrete Lower Concrete Invert a 1 2.0 8.0 8.0 7.0 11.0 6.0 2 3.5 9.5 10.5 13.0 15.0 2.0 3 1.0 7.5 8.5 9.5 16 5.0 4 0.5 7.5 7.0 5.0 9.0 9.0 5 0.0 7.0 7.0 5.0 15.0 9.0 DTN: MO0410MWDANS30.018, worksheets “Heat Removal” of vti-aa.xls (case 4), vti-ba.xls (case 5), vti-ca.xls (case 1), vti-da.xls (case 2), vtiea. xls (case 3). a Values from cells (row 12, column F), identified as “dw h coef.”, in worksheets “Heat Removal” of vti-aa.xls (case 4), vti-ba.xls (case 5), vti-ca.xls (case 1), vtida. xls (case 2), vti-ea.xls (case 3). Table 7-10. Comparison of Heat Transfer Coefficients Using Data-Fitting to the Mixed Convection and Dittus-Boelter Correlations Heat Transfer Coefficient (W/m2·K) Case No. Flow Rate (m3/s) ANSYS WP a Mixed Convection Correlation – Inner Surface (i.e., waste package surface) ANSYS Concrete and Invert a Mixed Convection Correlation – Outer Surface (i.e., inner concrete surface and invert surface) Dittus- Boelter b 1 1 6.5 5.9 c 8.8 6.6 h 2.9 2 2 8.3 7.5 d 11.3 9.2 i 5.0 3 0.5 6.1 5.7 e 11.6 7.3 j 1.7 4 1 5.6 5.3 f 8.0 6.5 k 2.9 5 0.5 5.3 4.7 g 11.0 7.1 l 1.7 a Average of values from Table 7-9, with side quarter or lower concrete counted twice. b Value from Table XVII-1. c Value from Table IX-31, Test 3. d Value from Table IX-31, Test 4. e Value from Table IX-31, Test 5. f Value from Table IX-31, Test 1. g Value from Table IX-31, Test 2. h Value from Table IX-34, Test 3. i Value from Table IX-34, Test 4. j Value from Table IX-34, Test 5. k Value from Table IX-34, Test 1. l Value from Table IX-34, Test 2. Iterating the heat transfer coefficients input to ANSYS until its results matched the test data resulted in heat transfer coefficients very close to those predicted by the mixed-convection correlation. However, the average of the heat transfer coefficients ranges from approximately two to five times larger than heat transfer coefficients calculated using the Dittus-Boelter correlation. Two reasons would tend to account for the differences. First, the Dittus-Boelter equation is a forced convection correlation. Analyses of the ventilation test data indicate a mixed (i.e., natural and forced) convection regime inside the concrete pipe annulus. Second, the Dittus-Boelter correlation for calculating a forced convection heat transfer coefficient was Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 7-25 October 2004 developed for hollow tube geometries. The correlation can be extended to a cylinder within a tube (i.e., waste package inside a drift) by using the hydraulic diameter instead of the geometric diameter. However, a cylinder within a tube, eccentrically located, is a different geometry which would tend to invalidate the Dittus-Boelter correlation within the range of air flow velocities being considered for preclosure. Add an invert, and the geometry of the problem lies even farther beyond the range of the Dittus-Boelter correlation. The values presented in Table 7-10 for the heat transfer coefficients used in the ANSYS post-test calculations argue that both natural and forced convection are important heat removal mechanisms for the experimental set-up of the Ventilation Test. Although scaling the quarter scale test results to a full scale drift is beyond the scope of this report, it stands to reason that a convection coefficient correlation which considers both natural and forced convection is more appropriate for use than the Dittus-Boelter equation (for the current drift design, heat load range, and ventilation flow rate). Table 7-11 summarizes the ventilation heat removal ratios for the five cases as modeled by ANSYS. The uncertainty in the predicted efficiency resulting from uncertainties in input data is estimated to be 3 percent on the basis of the analysis in Section 6.11. For four of the five cases (Cases 1, 3, 4, and 5), the predicted efficiency is below the uncertainty bounds, even without considering the additional model uncertainty described in Section 7.2.2. Therefore, the validation criterion is satisfied. Table 7-11. Heat Removal Ratios for the ANSYS Post-Test Ventilation Models Case No. Ventilation Efficiencya Ventilation Efficiency from Measurementsb 1 78 ± 3% 86.4 ± 0.6% 2 93 ± 3% 79.7 ± 0.2% 3 80 ± 3% 81 ± 2% 4 87 ± 3% 83.8 ± 0.6% 5 83± 3% 79 ± 2% a DTN: MO0410MWDANS30.018, identified as “percent of heat removal” in worksheets “Heat Removal” of vti-aa.xls (case 4), vti-ba.xls (case 5), vti-ca.xls (case 1), vti-da.xls (case 2), vti-ea.xls (case 3). b BSC 2003 [DIRS 160724], Table 5-17. 7.2.2.6 Additional Criterion Met for the Convection Heat Transfer Model The ANSYS numerical model matched the Phase I Ventilation Test results within the criterion of ±5°C using a reasonable range of heat transfer coefficients (Table 7-10). The range of heat transfer coefficients required to match the test results indicates a mixed convection regime inside the test train. The Dittus-Boelter correlation for calculating forced convection heat transfer coefficients is therefore a conservative approach. The impact of using such a correlation is a lower or more conservative rate of heat removal by ventilation because the Dittus Boelter equation is only valid for forced convection. A lower rate of heat removal translates to a lower efficiency and thus higher temperatures within the drift, which represents a conservative estimate of ventilation efficiency. A more realistic correlation is one that accounts for both natural and forced convection to remove heat from the drift, such as the mixed convection correlation used in these analyses. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 7-26 October 2004 7.3 VALIDATION SUMMARY The ventilation model has been validated by applying acceptance criteria based on an evaluation of the model’s relative importance to the performance of the repository system. All validation requirements defined in the applicable TWP (BSC 2004 [DIRS 170950], Section 2.3) have been fulfilled. Requirements for confidence building during model development have also been satisfied. The model development activities and postdevelopment validation activities described herein establish the scientific bases for the ventilation model. Based on this, the model is considered to be sufficiently accurate and adequate for the intended purpose with the stated limitations and to the level of confidence required by the model’s relative importance to the performance of the repository system. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 8-1 October 2004 8. CONCLUSIONS This report develops, validates, and implements a conceptual model for heat transfer in and around a ventilated emplacement drift. This conceptual model includes thermal radiation between the waste package and the drift wall, convection from the waste package and drift wall surfaces into the flowing air, and conduction in the surrounding host rock. These heat transfer processes are coupled and vary both temporally and spatially, so numerical and analytical methods are used to implement the mathematical equations which describe the conceptual model. These numerical and analytical methods predict the transient response of the system, at the drift scale, in terms of spatially varying temperatures and ventilation efficiencies. The ventilation efficiency describes the effectiveness of the ventilation process in removing radionuclide decay heat from the drift environment. An alternative conceptual model is also developed which evaluates the influence of water and water vapor mass transport on the ventilation efficiency. These effects are described using analytical methods which bound the contribution of latent heat to the system, quantify the effects of varying degrees of host rock saturation (and hence host rock thermal conductivity) on the ventilation efficiency, and evaluate the effects of vapor and enhanced vapor diffusion on the host rock thermal conductivity. 8.1 SUMMARY OF RESULTS As described by the conceptual model and its mathematical implementations, ventilation is found to be an effective way to remove heat produced by the decay of radionuclides in the in-drift environment, and to mitigate the peak waste package and drift wall temperatures that would otherwise occur. Given the License Application design parameters and inputs listed in Section 4 (including a ventilation flow rate of 15 m3/s, an inlet air temperature of 22.8°C, and a preclosure period of 50 years), the integrated ventilation efficiency is 88% for a 600 meter long drift and 86% for an 800 meter long drift. Temperatures of the waste package, drift wall, and drift air do not exceed 105, 85, or 80°C respectively (Figure 6-5c). The most influential parameters on the effectiveness of the ventilation to remove heat from the drift are the temperature of the inlet ventilation air and the ventilation flow rate (Figure 6-12). The effects of water and water vapor mass transport under sub-boiling conditions, described by the alternative conceptual model, on the ventilation efficiency and the waste package, drift wall, and drift air temperatures are minor (Section 6.9). The latent heat contribution associated with the evaporation of host rock matrix water near the drift wall is limited by the hydrologic properties of the rock, and is determined to be less than 1% of the total waste package energy provided to the in-drift and host rock environment. The change in temperatures associated with varying the host rock matrix saturation (and hence the bulk thermal conductivity) from completely dry to completely wet is found to be less than 5°C at any given time and distance from the drift entrance. The integrated ventilation efficiency for a 600 meter long drift ranges from about 90% for completely dry conditions, to about 87% for completely wet conditions. Finally, there was no evidence of enhancement to the host rock thermal conductivity due to vapor and enhanced vapor diffusion. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 8-2 October 2004 8.2 MODEL OUTPUTS The DTNs produced by this report are given in Table 8-1. Table 8-1. DTNs Produced by the Ventilation Model and Analysis Report DTN Description MO0303MWDSLTLC.000 Stratigraphic Layer Thickness (Section 6.5.1) MO0306MWDASLCV.001 Input/Output and Analysis of the ANSYS-LA-Coarse Ventilation Model (Section 6.6.1 and 6.6.2) MO0306MWDALAFV.000 Input/Output and Analysis of the ANSYS-LA-Fine Ventilation Model (Section 6.6.2) MO0307MWDAC8MV.000 Analytical-LA-Coarse Ventilation Model (Section 6.6.1) MO0306MWDRTCCV.000 Analytical-LA-Coarse-Wet-vs-Dry-kth Ventilation Model (Section 6.9.2) MO0406MWDLACVD.001 Analytical-LA-Coarse-Delta-Method Ventilation Model (Section 6.11) MO0406MWDAC8VD.001 Analytical-LA-Coarse-Delta-Method-800m Ventilation Model (Section 6.11) MO0410MWDANS30.018 Input/Output and Analysis of ANSYS Post Test Modeling of the Ventilation Test Phase I for Model Validation (Section 7.1.3) MO0306MWDVTPH2.000 Ventilation Test Phase II Data Analysis (Appendix XI) MO0306MWDMXCNV.000 Analyses to Support the Mixed Convection Correlation (Appendix X) 8.2.1 Summary of Model Outputs Given the design parameters and inputs listed in Section 4, the primary outputs of the ventilation model are: • Ventilation efficiencies as a function of time and location from the inlet of the drift, whose trend is to increase with time, but decrease with distance from the drift inlet (DTN: MO0307MWDAC8MV.000, from the Analytical-LA-Coarse results. Note that the results of the ANSYS-LA-Coarse model may also be used). • Integrated ventilation efficiencies and standard deviations for 600 and 800 meter long drifts (Table 8-2, integrated efficiencies from Analytical-LA-Coarse results and standard deviations, from the Analytical-LA-Coarse-Delta-Method results. Note that the integrated efficiencies are slightly different in the Delta-Method spreadsheet because that spreadsheet uses an average heat transfer coefficient, independent of time and position). • Waste package, drift wall, and in-drift air preclosure temperatures as function of time and location from the inlet of the drift, whose trend is to decrease with time, but increase with distance from the drift inlet (DTN: MO0307MWDAC8MV.000, from the Analytical-LA-Coarse results. Note that the results of the ANSYS-LA-Coarse model may also be used). Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 8-3 October 2004 Table 8-2. Integrated Ventilation Efficiency Over 50 Years of Preclosure, and 600 and 800 Meters of Drift Length of Drift (meters) Length of Ventilation (years) Mean Efficiency Uncertainty 600 50 88%a 3%b 800 50 86%a 3%c a DTN: MO0307MWDAC8MV.000, worksheet “Ventilation Efficiency” of Analytical-LA-Coarse- 800m.xls, rounded. b DTN: MO0406MWDLACVD.001, worksheet “Delta Method” of Analytical-LA-Coarse-Delta- Method.xls, rounded. c DTN: MO0406MWDAC8VD.001, worksheet “Delta Method” of Analytical-LA-Coarse-Delta-Method- 800m.xls, rounded. 8.2.2 Recommendations for Downstream Use of the Model Outputs Use of the ventilation model outputs, specifically the ventilation efficiency, is recommended for downstream thermal models that do not explicitly model in-drift behavior, such as waste package temperature, during the preclosure period. Such models include the multiscale thermohydrologic model, the drift degradation model, and the UZ coupled processes models. Either the ANSYSLA- Coarse or the Analytical-LA-Coarse results may be used. The ventilation efficiency can be used to reduce the thermal energy produced by the waste package during the preclosure period as a means of initializing postclosure conditions in the host rock. If the intent of the initialization of the downstream postclosure thermal model is only to account for the correct amount of heat energy supplied to the host rock during the preclosure period, then the use of the ventilation efficiency in this manner (either as a function of time and distance from the drift inlet, or the integrated efficiency) is appropriate. However, if the intent is also to predict drift wall temperature at the start of postclosure, then the use of the ventilation efficiency as described above is inadequate. 8.2.3 Output Uncertainty Uncertainty in the model output, specifically the integrated ventilation efficiency, is characterized using the mean and standard deviation listed in Table 8-3. Key input data and parameter uncertainties are also characterized using the means and standard deviations identified in Table 6-9. Input data and parameter uncertainties were propagated through the ventilation analysis using the Delta Method (see Section 6.11). The most influential design inputs and parameters on the uncertainty of the model output are the inlet air temperature, the air flow rate, the host-rock wet bulk thermal conductivity and specific heat (as a function of matrix saturation), and the convection heat transfer coefficients. The uncertainties associated with the model and methods of analyses are characterized by comparing the results of the implementations of the conceptual model to the results of the implementations of the alternative conceptual model, and by comparing the results of the actual methods themselves to each other (i.e., comparing the ANSYS results to the analytical results). Two examples demonstrate these points. First, the uncertainty associated with including, or not including, water and water vapor mass transport in the ventilation analysis is shown to be minimal when comparing the results of both models (Section 6.9.1). Second, there is some uncertainty associated with linearizing the radiation heat transfer in the analytical method. When Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 8-4 October 2004 comparing the temperature results of the analytical method to the results of the ANSYS method which explicitly treats the fourth order radiation heat transfer equation, there is little difference and hence no impact (Section 6.6.2). These same arguments may be made for other sources of model uncertainty such as: substituting the thermal pulse methodology in the analytical method for the transient conduction heat transfer analysis performed by ANSYS; the insensitivity of the length of well mixed volume elements in which the coupled heat transfer occurs, as addressed in the discretization study; or using finite element numerical iteration techniques in the ANSYS method compared to exact solutions obtained by the analytical approach. The uncertainty in the output of the ventilation model is propagated through downstream models that use this output by taking the standard deviation of the integrated ventilation efficiency about the mean, and comparing the sensitivity of the downstream results to the results obtained by the mean. 8.3 YUCCA MOUNTAIN REVIEW PLAN CRITERIA ASSESSMENT This model and analysis report provides preclosure information on ventilation efficiency that other models and analyses use to establish initial conditions for postclosure thermohydrologic calculations. The postclosure calculations feed into the model abstraction of the quantity and chemistry of water contacting engineered barriers and waste forms. This section summarizes the contents of this report as they apply to NRC criteria for a detailed review of that abstraction. These are the relevant criteria from Yucca Mountain Review Plan, Final Report (NRC 2003 [DIRS 163274], Section 2.2.1.3.3.3), which are based on meeting the requirements of 10 CFR 63.114(a)-(c) and (e)-(g) [DIRS 156605]. • Acceptance Criterion 1 – System Description and Model Integration Are Adequate. (1) Total system performance assessment adequately incorporates important design features, physical phenomena, and couplings, and uses consistent and appropriate assumptions throughout the quantity and chemistry of water contacting engineered barriers and waste forms abstraction process. For the ventilation calculations, inputs relevant to the design of the EBS, including ventilation, are almost entirely from current IEDs, except for minor changes to the design subsequent to completion of the calculations (Sections 4.1.7, 4.1.9, and 4.1.10). The input properties of the rock and the EBS materials, as well as the initial conditions in the rock and ventilating air (Sections 4.1.1 through 4.1.6, 4.1.9, and 4.1.15) are from DTNs and controlled engineering calculations that are specific to the site, except for one outside source for emissivity (Tables 4-8 and 4-21) that is qualified and justified in Appendix XVIII. Section 6.3 describes the physical phenomena and couplings incorporated in the ventilation model. Heat transfer in the EBS couples radiative energy transfer with convective heat transfer. Convective heat transfer is modeled as a coupling of forced convection and natural convection. Coupling of additional moisture-related processes is omitted as not important on the basis of analyses in Section 6.9. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 8-5 October 2004 The limitations stated in Section 1 are applicable to the current EBS design and are therefore appropriate. The assumptions are shown in Section 5 to be appropriate for the preclosure period and the EBS and ventilation design. The ventilation efficiency is used by downstream thermohydrologic models that recalculate the details of the preclosure period before proceeding to the postclosure period. Because the processes in the postclosure period differ from those during preclosure and must be projected over a much longer time frame, the assumptions in the downstream models may differ from those in the ventilation model. Therefore, the ventilation efficiency incorporates important design features, physical phenomena and couplings, and uses appropriate assumptions. (3) Important design features, such as waste package design and material selection, backfill, drip shield, ground support, thermal loading strategy, and degradation processes, are adequate to determine the initial and boundary conditions for calculations of the quantity and chemistry of water contacting engineered barriers and waste forms. The inputs to the ventilation analysis include details of the EBS features, including waste package dimensions and material properties (Section 4.1.9), waste package heat decay (Section 4.1.7), drift dimensions, and ventilation design parameters (Section 4.1.10). During the preclosure period, there is no drip shield. Therefore, the ventilation analysis incorporates the important design features that are adequate to determine the ventilation efficiency, which modifies the heat-generation boundary conditions for coupled process models. (6) The expected ranges of environmental conditions within the waste package emplacement drifts, inside of breached waste packages, and contacting the waste forms and their evolution with time are identified. These ranges may be developed to include: (i) the effects of the drip shield and backfill on the quantity and chemistry of water (e.g., the potential for condensate formation and dripping from the underside of the shield); (ii) conditions that promote corrosion of engineered barriers and degradation of waste forms; (iii) irregular wet and dry cycles; (iv) gamma-radiolysis; and (v) size and distribution of penetrations of engineered barriers. The ventilation model and analysis predicts heat transfer only during the preclosure period. The calculated ventilation efficiency can be applied in coupled process models to identify the expected range of saturation in the host rock at closure and the expected range of environmental conditions in the host rock, within the emplacement drifts, inside of breached waste packages, and contacting the waste forms and their evolution with time. • Acceptance Criterion 2 – Data Are Sufficient for Model Justification. (2) Sufficient data were collected on the characteristics of the natural system and engineered materials to establish initial and boundary conditions for conceptual models of thermal-hydrologic-mechanical-chemical coupled processes, that affect seepage and flow and the engineered barrier chemical environment. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 8-6 October 2004 The ventilation model does not couple thermal processes with hydrologic, mechanical, or chemical processes. However, the ventilation analysis provides information that affects boundary conditions for conceptual models of thermal-hydrologic-mechanical-chemical coupled processes. Specifically, the ventilation analysis provides time- and position-dependent ventilation efficiencies that modify the thermal output of the waste package, which may be treated as a boundary. Section 6.11 shows that the most important parameters for determining ventilation efficiency are inlet air temperature, air flow rate, and the wet bulk thermal conductivity of the rock. The inlet air temperature is assumed to be equal to the ambient host rock temperature, which is calculated (Section 6.5.6) primarily from the boundary conditions at the surface and the water table (Section 6.5.5), with consideration of the thermophysical properties of the rock layers (Section 6.5.2). The boundary temperatures are supported by ample data, as described in Section 4.1.6. The air flow rate is a design parameter, as described in Section 4.1.10. The thermal conductivity is developed from qualified project data, as described in Section 4.1.5. The collected data are sufficient that the standard deviations in wet bulk thermal conductivity are less than 15 percent of the values. Therefore, sufficient data were collected to establish initial and boundary conditions for the ventilation model. In turn, the calculated ventilation efficiency is sufficient to establish the effect of ventilation on boundary conditions for conceptual models of thermalhydrologic- mechanical-chemical coupled processes. • Acceptance Criterion 3 – Data Uncertainty Is Characterized and Propagated Through the Model Abstraction. (1) Models use parameter values, assumed ranges, probability distributions, and bounding assumptions that are technically defensible, reasonably account for uncertainties and variabilities, and do not result in an under-representation of the risk estimate. Section 6.11 demonstrates that the ventilation model uncertainty analysis uses parameter values, assumed ranges, probability distributions, and bounding assumptions that are technically defensible and reasonably account for uncertainties and variabilities. Section 7.2.2.5 shows that the model does not result in an overestimate of ventilation efficiency, which would tend to under-represent the risk estimate by under-representing the perturbations of pre-emplacement conditions due to repository heating. (2) Parameter values, assumed ranges, probability distributions, and bounding assumptions used in the total system performance assessment calculations of quantity and chemistry of water contacting engineered barriers and waste forms are technically defensible and reasonable, based on data from the Yucca Mountain region (e.g., results from large block and drift-scale heater and niche tests), and a combination of techniques that may include laboratory experiments, field measurements, natural analog research, and process-level modeling studies. Sections 4.1, 6.6.1, and 6.11 provide assurances that the parameter values, assumed ranges, probability distributions and bounding assumptions used in the ventilation calculations are Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 8-7 October 2004 technically defensible and reasonable, based on data from field measurements in the Yucca Mountain region and on process-level modeling studies. (3) Input values used in the total system performance assessment calculations of quantity and chemistry of water contacting engineered barriers (e.g., drip shield and waste package) are consistent with the initial and boundary conditions and the assumptions of the conceptual models and design concepts for the Yucca Mountain site. Correlations between input values are appropriately established in the U.S. Department of Energy total system performance assessment. Parameters used to define initial conditions, boundary conditions, and computational domain in sensitivity analyses involving coupled thermal-hydrologic-mechanicalchemical effects on seepage and flow, the waste package chemical environment, and the chemical environment for radionuclide release, are consistent with available data. Reasonable or conservative ranges of parameters or functional relations are established. Sections 4.1 and 5 show that the input values used in the ventilation calculations are consistent with the initial and boundary conditions and the assumptions of the conceptual models and design concepts for the Yucca Mountain site. Section 6.5.6 explains how correlation between ambient temperatures and inlet air temperature was appropriately established. Sections 6.6.1 and 6.11 demonstrate that parameters used to define initial conditions, boundary conditions, and computational domain in sensitivity analyses are consistent with available data and that reasonable or conservative ranges of parameters were established. 8.4 REQUIRED DOCUMENTATION OF LEVEL OF ACCURACY Table 8-3 includes the uncertainty in ventilation efficiency. Section 6.11 explains how the uncertainty was determined, thereby meeting the requirement for documentation for level of accuracy (BSC 2004 [DIRS 170950], Section 3.3). 8.5 COMPLETION CRITERIA This evaluation is consistent with the activities performed as part of Technical Work Plan: Regulatory Integration Evaluation of Analysis and Model Reports Supporting the TSPA-LA (BSC 2004 [DIRS 169653]) and fulfills a portion of the Phase 2 work identified in that plan. That is, the work addresses the prioritized list of actions selected in Phase 1 for disposition in Phase 2 (BSC 2004 [DIRS 169653], Section 1.3). Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 8-8 October 2004 INTENTIONALLY LEFT BLANK Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 9-1 October 2004 9. INPUTS AND REFERENCES 9.1 DOCUMENTS CITED Alenitsyn, A.; Butikov, E.I.; and Kondratyev, A.S. 1997. Concise Handbook of Mathematics and Physics. Boca Raton, Florida: CRC Press. TIC: 233486. 171443 Azizi, S.; Moyne, C.; and Degiovanni, A. 1988. “Approche Experimentale et Theorique de la Conductivite Thermique des Milieux Poreux Humides—I. Experimentation.” International Journal of Heat and Mass Transfer, 31, (11), 2305-2317. New York, New York: Pergamon Press. TIC: 243729. 154108 Bird, R.B.; Stewart, W.E.; and Lightfoot, E.N. 1960. Transport Phenomena. New York, New York: John Wiley & Sons. TIC: 208957. 103524 Birkholzer, J.T. and Tsang, Y.W. 2000. “Modeling the Thermal-Hydrologic Processes in a Large-Scale Underground Heater Test in Partially Saturated Fractured Tuff.” Water Resources Research, 36, (6), 1431-1447. Washington, D.C.: American Geophysical Union. TIC: 248278. 154608 BSC (Bechtel SAIC Company) 2001. ANSYS Calculations in Support of Natural Ventilation Parametric Study for SR. CAL-SVS-HV-000003 REV 00 ICN 01. Las Vegas, Nevada: Bechtel SAIC Company. ACC: MOL.20010613.0250. 155246 BSC 2001. Repository Multiple Waste Package Thermal Calculation. CAL-WIS-TH-000010 REV 00. Las Vegas, Nevada: Bechtel SAIC Company. ACC: MOL.20010814.0330. 156276 BSC 2002. Properties of Air Entering Emplacement Drifts Calculation. 800-P0C-VU00-00200-000-00A. Las Vegas, Nevada: Bechtel SAIC Company. ACC: MOL.20030106.0302. 161233 BSC 2003. Design and Engineering, D&E/PA/C IED Typical Waste Package Components Assembly 1 of 9. 800-IED-WIS0-00201-000-00C. Las Vegas, Nevada: Bechtel SAIC Company. ACC: ENG.20030917.0002. 165406 BSC 2003. Repository Design Project, Repository/PA IED Emplacement Drift Committed Materials (2). 800-IED-WIS0-00302-000-00A. Las Vegas, Nevada: Bechtel SAIC Company. ACC: ENG.20030627.0004. 164101 BSC 2003. Repository Design Project, Repository/PA IED Emplacement Drift Configuration 1 of 2. 800-IED-EBS0-00201-000-00A. Las Vegas, Nevada: Bechtel SAIC Company. ACC: ENG.20030630.0002. 164069 Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 9-2 October 2004 BSC 2003. Technical Work Plan for: Engineered Barrier System Department Modeling and Testing FY03 Work Activities. TWP-MGR-MD-000015 REV 04 ICN 02. Las Vegas, Nevada: Bechtel SAIC Company. ACC: DOC.20031001.0008. 165601 BSC 2003. Testing to Provide Data for Ventilation System Design: Phase 1. TDR-EBS-MD-000021 REV 00. Las Vegas, Nevada: Bechtel SAIC Company. ACC: DOC.20030711.0001. 160724 BSC 2004. D&E / PA/C IED Emplacement Drift Configuration and Environment. 800-IED-MGR0-00201-000-00B. Las Vegas, Nevada: Bechtel SAIC Company. ACC: ENG.20040326.0001. 168489 BSC 2004. D&E / PA/C IED Subsurface Facilities. 800-IED-WIS0-00101-000- 00A. Las Vegas, Nevada: Bechtel SAIC Company. ACC: ENG.20040309.0026. 164519 BSC 2004. D&E / PA/C IED Typical Waste Package Components Assembly. 800-IED-WIS0-00203-000-00B. Las Vegas, Nevada: Bechtel SAIC Company. ACC: ENG.20040202.0011. 167754 BSC 2004. D&E/PA/C IED Typical Waste Package Components Assembly. 800-IED-WIS0-00202-000-00C. Las Vegas, Nevada: Bechtel SAIC Company. ACC: ENG.20040517.0008. 169472 BSC 2004. Development of Numerical Grids for UZ Flow and Transport Modeling. ANL-NBS-HS-000015 REV 02. Las Vegas, Nevada: Bechtel SAIC Company. ACC: DOC.20040901.0001. 169855 BSC 2004. Multiscale Thermohydrologic Model. ANL-EBS-MD-000049 REV 02. Las Vegas, Nevada: Bechtel SAIC Company. ACC: DOC.20041014.0008. 169565 BSC 2004. Q-List. 000-30R-MGR0-00500-000-000 REV 00. Las Vegas, Nevada: Bechtel SAIC Company. ACC: ENG.20040721.0007. 168361 BSC 2004. Repository Subsurface Emplacement Drifts Steel Invert Structure Plan & Elevation. 800-SS0-SSE0-00101-000-00B. Las Vegas, Nevada: Bechtel SAIC Company. ACC: ENG.20040520.0004. 169503 BSC 2004. Technical Work Plan for: Near-Field Environment and Transport In-Drift Heat and Mass Transfer Model and Analysis Reports Integration. TWP-MGR-PA-000018 REV 01. Las Vegas, Nevada: Bechtel SAIC Company. ACC: DOC.20040729.0006. 170950 BSC 2004. Technical Work Plan for: Regulatory Integration Evaluation of Analysis and Model Reports Supporting the TSPA-LA. TWP-MGR-PA-000014 REV 00 ICN 01. Las Vegas, Nevada: Bechtel SAIC Company. ACC: DOC.20040603.0001. 169653 Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 9-3 October 2004 BSC 2004. UZ Flow Models and Submodels. MDL-NBS-HS-000006, Rev. 02. Las Vegas, Nevada: Bechtel SAIC Company. 169861 Carslaw, H.S. and Jaeger, J.C. 1959. Conduction of Heat in Solids. 2nd Edition. Oxford, Great Britain: Oxford University Press. TIC: 206085. 100968 CertainTeed. 1996. Submittal Sheet, Standard Fiber Glass Duct Wrap. Valley Forge, Pennsylvania: CertainTeed Corporation. TIC: 249257. 153512 Cho, Y.I.; Ganic, E.N.; Hartnett, J.P.; and Rohsenow, W.M. 1998. “Basic Concepts of Heat Transfer.” Chapter 1 of Handbook of Heat Transfer. 3rd Edition. Rohsenow, W.M.; Hartnett, J.P.; and Cho, Y.I., eds. New York, New York: McGraw-Hill. TIC: 253612. 160802 Conte, S.D. and de Boor, C. 1972. Elementary Numerical Analysis, An Algorithmic Approach. 2nd Edition. New York, New York: McGraw-Hill. TIC: 224146. 159800 CRWMS (Civilian Radioactive Waste Management System) M&O (Management and Operating Contractor) 2000. Conceptual Arrangement Simulated Emplacement Ventilation Test. Las Vegas, Nevada: CRWMS M&O. ACC: MOL.20001219.0107. 153503 CRWMS M&O 2001. Software Validation Test Report ANSYS Version 5.6.2 Software. SAN: LV-2000-169. SDN: 10145-VTR-5.6.2-00. Las Vegas, Nevada: CRWMS M&O. ACC: MOL.20010323.0064. 155138 DOE (U.S. Department of Energy) 2003. Software Management Report: ymesh Version 1.54. Software Document Number: 10172-SMR-1.54-00. Las Vegas, Nevada: U.S. Department of Energy, Office of Repository Development. ACC: MOL.20030529.0026. 171332 DOE 2003. Software Management Report: rme6 Version 1.2. Software Document Number: 10617-SMR-1.2-00. Las Vegas, Nevada: U.S. Department of Energy, Office of Repository Development. ACC: MOL.20030523.0042. 171333 Doraswamy, N. 2001. Validation Test Report for ANSYS Version 5.6.2 Software. Document Number: 10145-VTR-5.6.2-01. Las Vegas, Nevada: Bechtel SAIC Company. ACC: MOL.20020219.0076. 171331 Ebadian, M.A. and Dong, Z.F. 1998. “Forced Convection, Internal Flow in Ducts.” Chapter 5 of Handbook of Heat Transfer. 3rd Edition. Rohsenow, W.M.; Hartnett, J.P.; and Cho, Y.I., eds. New York, New York: McGraw-Hill. TIC: 253612. 160728 Fetter, C.W. 1993. Contaminant Hydrogeology. Upper Saddle River, New Jersey: Prentice Hall. TIC: 240691. 102009 Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 9-4 October 2004 Gebhart, B.; Jaluria, Y.; Mahajan, R.L.; and Sammakia, B. 1988. Buoyancy-Induced Flows and Transport. Textbook Edition. New York, New York: Hemisphere Publishing. TIC: 102802. 152234 Haar, L.; Gallagher, J.S.; and Kell, G.S. 1984. NBS/NRC Steam Tables: Thermodynamic and Transport Properties and Computer Programs for Vapor and Liquid States of Water in SI Units. New York, New York: Hemisphere Publishing Corporation. TIC: 241793. 105175 Hahn, G.J. and Shapiro, S.S. 1967. Statistical Models in Engineering. New York, New York: John Wiley & Sons. TIC: 247729. 146529 Hartman, H.L. 1982. Mine Ventilation and Air Conditioning. Mutmansky, J.M. and Wang, Y.J., eds. 2nd Edition. New York, New York: John Wiley & Sons. TIC: 210152. 128009 Hillel, D. 1998. Environmental Soil Physics. San Diego, California: Academic Press. TIC: 254422. 165404 Holman, J.P. 1997. Heat Transfer. 8th Edition. New York, New York: McGraw-Hill. TIC: 239954. 101978 Incropera, F.P. and DeWitt, D.P. 1985. Fundamentals of Heat and Mass Transfer. 2nd Edition. New York, New York: John Wiley & Sons. TIC: 208420. 114109 Incropera, F.P. and DeWitt, D.P. 1996. Fundamentals of Heat and Mass Transfer. 4th Edition. New York, New York: John Wiley & Sons. TIC: 243950. 108184 Jury, W.A.; Gardner, W.R.; and Gardner, W.H. 1991. Soil Physics. 5th Edition. New York, New York: John Wiley & Sons. TIC: 241000. 102010 Kays, W.M. and Leung, E.Y. 1963. “Heat Transfer in Annular Passages— Hydrodynamically Developed Turbulent Flow with Arbitrarily Prescribed Heat Flux.” International Journal of Heat and Mass Transfer, 6, (7), 537-557. New York, New York: Pergamon. TIC: 253626. 160763 Kays, W.M. and Perkins, H.C. 1973. “Forced Convection, Internal Flow in Ducts.” Section 7 of Handbook of Heat Transfer. Rohsenow, W.M. and Hartnett, J.P., eds. New York, New York: McGraw-Hill. TIC: 253611. 160782 Kern, D.Q. 1950. Process Heat Transfer. New York, New York: McGraw-Hill. TIC: 248066. 130111 Knudsen, J.G.; Bell, K.J.; Holt, A.D.; Hottel, H.C.; Sarofim, A.F.; Standiford, F.C.; Stuhlbarg, D.; and Uhl, V.W. 1984. “Heat Transmission.” Section 10 of Perry's Chemical Engineers' Handbook. 6th Edition. Perry, R.H.; Green, D.W.; and Maloney, J.O., eds. New York, New York: McGraw-Hill. TIC: 246473. 170057 Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 9-5 October 2004 Kuehn, T.H. and Goldstein, R.J. 1976. “Correlating Equations for Natural Convection Heat Transfer Between Horizontal Circular Cylinders.” International Journal of Heat and Mass Transfer, 19, (10), 1127-1134. New York, New York: Pergamon Press. TIC: 238411. 100675 Kuehn, T.H. and Goldstein, R.J. 1978. “An Experimental Study of Natural Convection Heat Transfer in Concentric and Eccentric Horizontal Cylindrical Annuli.” Journal of Heat Transfer, 100, (4), 635-640. New York, New York: American Society of Mechanical Engineers. TIC: 244433. 130084 Levenspiel, O. 1972. Chemical Reaction Engineering. 2nd Edition. New York, New York: John Wiley & Sons. TIC: 224877. 156839 McAdams, W.H. 1954. Heat Transmission. 3rd Edition. New York, New York: McGraw-Hill. TIC: 242359. 161435 Morgan, V.T. 1975. “The Overall Convective Heat Transfer from Smooth Circular Cylinders.” Advances in Heat Transfer. Volume 11. Irvine, T.F., Jr. and Hartnett, J.P., eds. Pages 199-264. New York, New York: Academic Press. TIC: 254306. 160791 Moyne, C.; Batsale, J-C.; and Degiovanni, A. 1988. “Approche Experimentale et Theorique de la Conductivite Thermique des Milieux Poreux Humides—II. Theorie.” International Journal of Heat and Mass Transfer, 31, (11), 2319-2329. New York, New York: Pergamon Press. TIC: 249402. 154107 Moyne, C.; Batsale, J.C.; Degiovanni, A.; and Maillet, D. 1990. “Thermal Conductivity of Wet Porous Media: Theoretical Analysis and Experimental Measurements.” Thermal Conductivity 21, Proceedings of the Twenty-First International Thermal Conductivity Conference, October 15-18, 1989, Lexington, Kentucky. Cremers, C.J. and Fine, H.A., eds. Pages 109-120. New York, New York: Plenum Press. TIC: 249322. 153164 Nagle, R.K. and Saff, E.B. 1994. Fundamentals of Differential Equations and Boundary Value Problems. Reading, Massachusetts: Addison-Wesley Publishing. TIC: 238891. 100922 NRC (U.S. Nuclear Regulatory Commission) 2003. Yucca Mountain Review Plan, Final Report. NUREG-1804, Rev. 2. Washington, D.C.: U.S. Nuclear Regulatory Commission, Office of Nuclear Material Safety and Safeguards. TIC: 254568. 163274 Perry, R.H.; Green, D.W.; and Maloney, J.O., eds. 1984. Perry's Chemical Engineers' Handbook. 6th Edition. New York, New York: McGraw-Hill. TIC: 246473. 125806 Raithby, G.D. and Hollands, K.G.T. 1998. “Natural Convection.” Chapter 4 of Handbook of Heat Transfer. 3rd Edition. Rohsenow, W.M.; Hartnett, J.P.; and Cho, Y.I., eds. New York, New York: McGraw-Hill. TIC: 253612. 160764 Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 9-6 October 2004 Reid, R.C.; Prausnitz, J.M.; and Sherwood, T.K. 1977. The Properties of Liquids and Gases. New York, New York: McGraw-Hill Book Company. TIC: 240958. 130310 Reynolds, W.C.; Lundberg, R.E.; and McCuen, P.A. 1963. “Heat Transfer in Annular Passages. General Formulation of the Problem for Arbitrarily Prescribed Wall Temperatures or Heat Fluxes.” International Journal of Heat and Mass Transfer, 6, (6), 483-493. New York, New York: Pergamon Press. TIC: 253625. 160770 Rohsenow, W.M.; Hartnett, J.P.; and Cho, Y.I. 1998. Handbook of Heat Transfer. 3rd Edition. New York, New York: McGraw-Hill. TIC: 253612. 169241 Sass, J.H.; Lachenbruch, A.H.; Dudley, W.W., Jr.; Priest, S.S.; and Munroe, R.J. 1988. Temperature, Thermal Conductivity, and Heat Flow Near Yucca Mountain, Nevada: Some Tectonic and Hydrologic Implications. Open-File Report 87-649. Denver, Colorado: U.S. Geological Survey. TIC: 203195. 100644 Stroe, D.E. 2001. “PO#: A18763CM0A, Transmittal of Test Results.” Letter from D.E. Stroe (Anter Laboratories) to M. Knudsen (CRWMS M&O), January 31, 2001, PR20939-51554c, with attachment. ACC: MOL.20010220.0057. 155633 Sutherland, W.A. and Kays, W.M. 1964. “Heat Transfer in an Annulus with Variable Circumferential Heat Flux.” International Journal of Heat and Mass Transfer, 7, (11), 1187-1194. New York, New York: Pergamon. TIC: 253693. 160789 Weast, R.C., ed. 1977. CRC Handbook of Chemistry and Physics. 58th Edition. Cleveland, Ohio: CRC Press. TIC: 242376. 106266 Weinberger, H.F. 1965. A First Course in Partial Differential Equations. New York, New York: John Wiley & Sons. TIC: 254310. 163216 White, F.M. 1986. Fluid Mechanics. 2nd Edition. New York, New York: McGraw-Hill. TIC: 243415. 111015 Wildenschild, D. and Roberts, J.J. 1999. Experimental Tests of Enhancement of Vapor Diffusion in Topopah Spring Tuff. UCRL-JC-134850. Livermore, California: Lawrence Livermore National Laboratory. TIC: 246923. 131055 Yovanovich, M.M. 1998. “Conduction and Thermal Contact Resistances (Conductances).” Chapter 3 of Handbook of Heat Transfer. 3rd Edition. Rohsenow, W.M.; Hartnett, J.P.; and Cho, Y.I., eds. New York, New York: McGraw-Hill. TIC: 253612. 171591 Zwillinger, D., ed. 1996. CRC Standard Mathematical Tables and Formulae. 30th Edition. Boca Raton, Florida: CRC Press. TIC: 233960. 152179 Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 9-7 October 2004 9.2 CODES, STANDARDS, REGULATIONS, AND PROCEDURES 10 CFR 63. Energy: Disposal of High-Level Radioactive Wastes in a Geologic Repository at Yucca Mountain, Nevada. Readily available 156605 ANSI/NCSL Z540-2-1997. American National Standard for Calibration — U.S. Guide to the Expression of Uncertainty in Measurement. Boulder, Colorado: NCSL International. TIC: 251472. 157394 AP-2.14Q, Rev. 3, ICN 0. Document Review. Washington, D.C.: U.S. Department of Energy, Office of Civilian Radioactive Waste Management. ACC: DOC.20030827.0018. AP-2.27Q, Rev. 1, ICN 4. Planning for Science Activities. Washington, D.C.: U.S. Department of Energy, Office of Civilian Radioactive Waste Management. ACC: DOC.20040610.0006. AP-3.15Q, Rev. 4, ICN 5. Managing Technical Product Inputs. Washington, D.C.: U.S. Department of Energy, Office of Civilian Radioactive Waste Management. ACC: DOC.20040812.0004. AP-SIII.10Q, Rev. 2, ICN 7. Models. Washington, D.C.: U.S. Department of Energy, Office of Civilian Radioactive Waste Management. ACC: DOC.20040920.0002. AP-SV.1Q, Rev. 1, ICN 1. Control of the Electronic Management of Information. Washington, D.C.: U.S. Department of Energy, Office of Civilian Radioactive Waste Management. ACC: DOC.20040308.0001. ASME PTC 19.1-1998. Test Uncertainty, Instruments and Apparatus. New York, New York: American Society of Mechanical Engineers. TIC: 249327. 153195 LP-SI.11Q-BSC Rev 0, ICN 1. Software Management. Washington, D.C.: U.S. Department of Energy, Office of Civilian Radioactive Waste Management. ACC: DOC.20041005.0008. 9.3 SOURCE DATA, LISTED BY DATA TRACKING NUMBER GS000383351030.002. Angle of Repose, Particle Density, and Uncompacted Bulk Density Data for Analyses Performed on Potential Candidate Backfill Materials. Submittal date: 03/24/00. 148445 GS000483351030.003. Thermal Properties Measured 12/01/99 to 12/02/99 Using the Thermolink Soil Multimeter and Thermal Properties Sensor on Selected Potential Candidate Backfill Materials Used in the Engineered Barrier System. Submittal date: 11/09/2000. 152932 Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 9-8 October 2004 GS000508312231.006. Physical Properties and Water Content from Borehole USW NRG-6, 3/19/94 to 3/27/95. Submittal date: 05/23/2000. 153237 GS020183351030.001. Uncompacted Bulk Density for Analyses Performed 02/02/00 to 05/23/00 on Potential Backfill Materials Used in the Engineered Barrier System. Submittal date: 01/22/2002. 163107 GS950408312231.004. Physical Properties and Water Potentials of Core from Borehole USW SD-9. Submittal date: 03/01/1995. 108986 GS951108312231.009. Physical Properties, Water Content, and Water Potential for Borehole USW SD-7. Submittal date: 09/26/1995. 108984 GS951108312231.010. Physical Properties and Water Content for Borehole USW NRG-7/7A. Submittal date: 09/26/1995. 108983 GS951108312231.011. Physical Properties, Water Content, and Water Potential for Borehole USW UZ-7A. Submittal date: 09/26/1995. 108992 LB0110ECRBH2OP.001. Water Potential Data from Three Locations in the ECRB. Submittal date: 11/12/2001. 156883 LB0208UZDSCPMI.002. Drift-Scale Calibrated Property Sets: Mean Infiltration Data Summary. Submittal date: 08/26/2002. 161243 LB0302PTNTSW9I.001. PTN/TSW Interface Percolation Flux Maps for 9 Infiltration Scenarios. Submittal date: 02/28/2003. 162277 LB0303THERMSIM.001. UZ Thermal Modeling: Simulations. Submittal date: 03/28/2003. 165167 LB990901233124.006. Moisture Data from the ECRB Cross Drift for AMR U0015, “In Situ Testing of Field Processes”. Submittal date: 11/01/1999. 135137 MO0106RIB00038.001. Water-Level Data and the Potentiometric Surface. Submittal date: 06/22/2001. 155631 MO0406SEPTVDST.000. Temperature and Volume Water Content for Drift Scale Test (DST) Heating Phase for Boreholes 79 and 80. Submittal date: 06/29/2004. 170616 MO0407SEPFEPLA.000. LA FEP List. Submittal date: 07/20/2004. 170760 SN0208F3409100.007. Preclosure Ventilation Test: 1/4 Scale, Including Data from Cases 1 through 6 (with Results from 10/05/2000 through 12/22/2000), Final Data Revised August 2002. Submittal date: 08/27/2002. 161729 Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 9-9 October 2004 SN0208F3409100.009. Preclosure Ventilation Test: 1/4 Scale, Phase 2, Including Data from Tests 1 through 16 (with Results from 4/25/01 to 10/01/2001), Final Data Revised August 2002. Submittal date: 08/27/2002. 163079 SN0303T0503102.008. Revised Thermal Conductivity of the Non-Repository Layers of Yucca Mountain. Submittal date: 03/19/2003. 162401 SN0307T0510902.003. Updated Heat Capacity of Yucca Mountain Stratigraphic Units. Submittal date: 07/15/2003. 164196 SN0404T0503102.011. Thermal Conductivity of the Potential Repository Horizon Rev 3. Submittal date: 04/27/2004. 169129 SNL22100196001.006. Laboratory Measurements of Thermal Conductivity as a Function of Saturation State for Welded and Nonwelded Tuff Specimens. Submittal date: 06/08/1998. 158213 9.4 SOFTWARE CODES BSC 2001. Software Code: ANSYS. V5.6.2. Sun, Solaris 2.6 and Solaris 2.7. 10145-5.6.2-01. 164464 CRWMS M&O 2001. Software Code: ANSYS. V5.6.2. IRIX 6.5. 10145-5.6.2-00. 154671 LLNL (Lawrence Livermore National Laboratory) 2003. Software Code: rme6. v1.2. SUN, SOLARIS 8. 10617-1.2-00. 163892 LLNL 2003. Software Code: YMESH. v1.54. SUN, SOLARIS 8. 10172-1.54-00. 163894 Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 9-10 October 2004 INTENTIONALLY LEFT BLANK Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 October 2004 APPENDIX I USING THE GEOLOGIC FRAMEWORK MODEL AND MINERALOGIC HYDROSTRATIGRAPHIC UNITS TO ASSIGN THERMOPHYSICAL PROPERTIES TO THE UZ UNITS Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 October 2004 Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 I-1 October 2004 This appendix documents the calculation of thermophysical properties of the UZ model layers based on the thermophysical properties of the lithostratigraphic units. The inputs (thermophysical properties of the lithostratigraphic units) are presented in Tables I-1 and I-2. The outputs (thermophysical properties of the UZ model layers) are presented in Table I-3. The formulae used in the calculation are listed in Table I-4. Table I-5 provides the nomenclature correlation between lithostratigraphic units and UZ model layers. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 I-2 October 2004 Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 I-3 October 2004 Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 I-4 October 2004 Table I-5. Nomenclature Correlation Between Lithostratigraphic Unit and UZ Model Layer Lithostratigraphic Unit UZ Model Layer Tpcr tcw11 Tpcp TpcLD tcw12 Tpcpv3 Tpcpv2 tcw13 Tpcpv1 ptn21 Tpbt4 ptn22 ptn23 Tpy (Yucca) Tpbt3 ptn24 Tpp (Pah) ptn25 Tpbt2 Tptrv3 Tptrv2 ptn26 Tptrv1 tsw31 Tptrn tsw32 Tptrl, Tptf Tptpul, RHHtop tsw33 Tptpmn tsw34 Tptpll tsw35 tsw36 Tptpln tsw37 Tptpv3 tsw38 Tptpv2 tsw39 Tptpv1 Tpbt1 ch1 ch2 ch3 ch4 Tac (Calico) ch5 Tacbt (Calicobt) ch6 Tcpuv (Prowuv) pp4 Tcpuc (Prowuc) pp3 Tcpm (Prowmd) Tcplc (Prowlc) pp2 Tcplv (Prowlv) Tcpbt (Prowbt) Tcbuv (Bullfroguv) pp1 Tcbuc (Bullfroguc) Tcbm (Bullfrogmd) Tcblc (Bullfroglc) bf3 Tcblv (Bullfroglv) Tcbbt (Bullfrogbt) Tctuv (Tramuv) bf2 Tctuc (Tramuc) Tctm (Trammd) Tctlc (Tramlc) tr3 Tctlv (Tramlv) Tctbt (Trambt) tr2 Source: BSC 2004 [DIRS 169855], Table 6-11. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 October 2004 APPENDIX II CALCULATING EFFECTIVE THERMOPHYSICAL PROPERTIES FOR THE ANSYS-BASED MODELS Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 October 2004 Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 II-1 October 2004 This appendix documents the calculation of effective thermophysical properties used in the ANSYS model discussed in Section 6 of this report. For geologic media composed of air, water, and rock, the heat capacity per unit volume of the composite material is the sum of the heat capacities of the constituents weighted by volume fractions. Jury et al. (1991 [DIRS 102010], p. 179) express this capacity as: .= · + · + · = N j sj sj vw w a a soil C C C C 1 . . . (Eq. II-1) where .a = volume fraction of the air .w = volume fraction of the water .sj = volume fraction of jth component of the solids Ca = volumetric heat capacity of the air Cvw = volumetric heat capacity of the water Csj = volumetric heat capacity of the jth component of the solids More specifically for the geologic units at Yucca Mountain, Equation II-1 can be written: s s vw w a al a am rock C C C C C · + · + · + · = . . . . (Eq. II-2) where .am = volume fraction of the air in the matrix .al = volume fraction of the air in the lithophysae .w = volume fraction of the water in the matrix and .s = volume fraction of the solids The various volume fractions can be written as: al am wm s al al V V V V V + + + = . (Eq. II-3) al am wm s am am V V V V V + + + = . (Eq. II-4) al am wm s w w V V V V V + + + = . (Eq. II-5) al am wm s s s V V V V V + + + = . (Eq. II-6) Substituting these equations into Equation II-2 and using the identity that the product of the density and the specific heat of a material is the volumetric heat capacity (Incropera and DeWitt 1996 [DIRS 108184], Section 2.2.2) results in the following: Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 II-2 October 2004 p g s vw w a al a am rock al am wm s C V C V C V C V C ) V V V V ( · · + · + · + · = · + + + . (Eq. II-7) where Val = volume of the air in the lithophysae Vam = volume of the air in the matrix Vwm = volume of the water in the matrix Vs = volume of the solids (which is set to 1) Cp = specific heat of the solids .g = grain density of the solids Now consider the definitions for the matrix porosity and the lithophysal porosity. The matrix porosity is defined as the ratio of the void volume of the matrix to the total matrix volume (of solids): m s m m V V V + = f (Eq. II-8) where Vm = Vam + Vwm. Solving for the matrix void volume in terms of the matrix porosity: s m m m V V · - = f f 1 (Eq. II-9) The lithophysal porosity is defined as the ratio of the volume of the lithophysae to the total volume: al m s al l V V V V + + = f (Eq. II-10) Solving for the volume of lithophysae: ) V V V ( V al m s l al + + · = f (Eq. II-11) Substituting Equation II-9 into Equation II-11 yields: s m m l l al V ) ( V · - + · - = f f f f 1 1 1 (Eq. II-12) The matrix saturation (S) is used to estimate the volume occupied by water: s m m w V S V · - · = f f 1 (Eq. II-13) Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 II-3 October 2004 Substituting Equation II-12 into Equation II-7, and neglecting the heat capacity of the air (Ca< 10,000) in circular tubes (r* = 0). For air (Pr = 0.70), and the tube hotter than the fluid, the correlation is (Incropera and DeWitt 1985 [DIRS 114109], p. 394, Eq. 8.58): 8 . 0 020 . 0 Re Nu = (Eq. IX-12) Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-10 October 2004 This formula was used for the Nusselt number at the outer wall in earlier ventilation model calculations with fully developed turbulent flow. Incropera and DeWitt (1985 [DIRS 114109], p. 400) consider the Dittus-Boelter equation to be a first approximation, in which the inner and outer convection coefficients are assumed to be equal. For concentric circular cylinders with a uniform heat rate on each cylinder, Kays and Leung (1963 [DIRS 160763], p. 539, Eqs. 15 and 16) derived the following expressions: ( ) ) ( ) ( 1 ) ( " " i o * i ii i q q x . x Nu x Nu - = (Eq. IX-13) ( ) ) ( ) ( 1 ) ( " " o i * o oo o q q x . x Nu x Nu - = (Eq. IX-14) where ts coefficien influence ) ( and ) ( heated is alone it hen cylinder w outer the of number Nusselt the ) ( heated is alone it hen cylinder w inner the of number Nusselt the ) ( * * = = = x x x Nu x Nu o i oo ii . . Using empirical velocity and eddy distribution profiles, Kays and Leung (1963 [DIRS 160763]) evaluated the parameters by obtaining asymptotic solutions of the energy differential equations, using fluid properties evaluated at ( ) x Tf and empirical equations for turbulent region diffusivity. Kays and Leung (1963 [DIRS 160763], Table 1) tabulated the asymptotic values of the parameters, * i . , * o . , ii Nu , and oo Nu as functions of r*, the Reynolds number, and the Prandtl number. Table IX-2 contains the values for Pr=0.7 and r* of 0.2 and 0.5, which are used in Section IX.1.6. The methodology is limited to values of r* between 0.2 and 0.5, which includes the design value for the EBS. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-11 October 2004 Table IX-2. Parameters for Forced Convection Correlation for Fully Developed Flow and Pr = 0.7 r*: 0.2 Re: 10000 30000 100000 300000 1000000 Nuii: 38.6 79.8 196 473 1270 .I: 0.412 0.338 0.286 0.260 0.235 Nuoo: 29.4 64.3 165 397 1070 .o: 0.063 0.055 0.049 0.044 0.040 r*: 0.5 Re: 10000 30000 100000 300000 1000000 Nuii: 30.9 66 166 400 1080 .I: 0.3 0.258 0.225 0.206 0.185 Nuoo: 28.3 62 158 380 1040 .o: 0.137 0.119 0.107 0.097 0.090 Source: Kays and Leung 1963 [DIRS 160763], Table 1 However, Equations IX-13 and IX-14 do not provide an explicit form for the calculation of heat transfer coefficients for models that have known boundary temperatures rather than known fluxes. In convective processes involving heat transfer from a boundary surface exposed to a relatively low-velocity fluid stream, it is convenient to introduce a local convective heat transfer coefficient, ) , ( . x h (W/m2·K), defined implicitly by Newton’s law of cooling, which is: ( ) [ ] x T x T x h x q f - = ' ' ) , ( ) , ( ) , ( . . . (Eq. IX-15) where ( ) . , x q ' ' is the convective heat flux (W/m2) (positive into the fluid) and ) , ( . x T is the surface temperature (K) (Incropera and DeWitt 1985 [DIRS 114109], p. 8, Eq. 1.3). If the temperature difference is zero, then ( ) . , x q ' ' is zero, and ) , ( . x h is not defined. Because the methodology does not permit heat transfer coefficients that vary around the circumference, a nominal value, the “effective circumferential” convective heat transfer coefficient ) (x h (W/m2·K), is defined such that: ( ) [ ] x T x T x h x q f - = ) ( ) ( ) ( " (Eq. IX-16) where ( ) x q" is the circumferential average convective heat flux (W/m2) and ) (x T is the circumferential average surface temperature (K). If the cylinder has a uniform temperature around its circumference, then ) (x h is the circumferential average of ) , ( . x h , but if the temperature varies around the circumference, ) (x h may differ from the average of ) , ( . x h . From Equations IX-16, IX-2, and IX-3, the ratio of heat fluxes is: ( ) ( ) ( ) x x Nu x Nu x q x q o i o i t ) ( ) ( " " = (Eq. IX-17) Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-12 October 2004 where ( ) ( ) ( ) x T x T x T x T x f o f i - - = ) ( ) ( t (Eq. IX-18) Substituting for the ratio in Equations IX-13 and IX-14, using the asymptotic values of the parameters, yields the following two solutions for the asymptotic Nusselt numbers: ( ) ( ) ) ( ) ( 1 x Nu x Nu x . Nu x Nu i o * i ii i t - = (Eq. IX-19) ( ) ( ) x x Nu x Nu . Nu x Nu o i * o oo o t ) ( ) ( 1- = (Eq. IX-20) The forced-convection correlation is valid only for fully developed flows. This permits replacing the asymptotic limits with the local values. Solving the above simultaneous equations yields the following formulas for explicit calculation of the effective circumferential Nusselt numbers from the temperatures: ( ) ( ) ( ) ( ) ( ) ( ) ( ) Re . Re . x Re . Re Nu Re Nu x Nu * i * o * i oo ii Fi - + = 1 / t (Eq. IX-21) ( ) ( ) ( ) ( ) ( ) ( ) ( ) Re . Re . x Re . Re Nu Re Nu x Nu * i * o * o ii oo Fo - + = 1 t (Eq. IX-22) In comparing experimental data with their correlations, Kays and Leung (1963 [DIRS 160763], pp. 544-545, Figures 6 through 8) did not correct the experimental data for Re effects of natural convection. Because the flow was vertically upward, so that buoyancy effects were longitudinal rather than transverse, the effects of natural convection are minimized. The Kays-Leung correlations are for idealized configurations that differ from the EBS in that the cylinders are concentric. The effects of this idealization are considered in the uncertainty analysis presented later in this appendix. IX.1.6 Mixed Convection Review of the literature shows little research has been completed for mixed convection conditions, and no information was found for configurations similar to the YMP drifts. For internal flows, Incropera and DeWitt (1985 [DIRS 114109], p. 445) and Raithby and Hollands (1998 [DIRS 160764], pp. 4.78 to 4.79) limit consideration of mixed convection to laminar flows within heated cylinders. Earlier ventilation model calculations neglected natural convection, using only a model for forced convection. The method used for mixed convection in the Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-13 October 2004 methodology documented here is based on the method of Morgan (1975 [DIRS 160791], p. 244, Eq. 21). The literature search found no published correlations of experimental data for the flow pattern of Figure IX-1d. In order to use published correlations, the following statements need to be true: • Mixed-convection in the EBS configuration is approximately the same as mixed-convection in an idealized configuration in which a hotter cylinder is inside a cooler cylinder. • The effective circumferential Nusselt number at each surface is related to the Reynolds number and dimensionless temperature difference across the boundary layer, but is independent of the conditions at the other surface. This relation is given by the correlation of Kays and Leung (1963 [DIRS 160763], p. 539, Eqs. 15 and 16) for natural convection. • Natural convection in the EBS may be predicted by the correlation of Kuehn and Goldstein (1978 [DIRS 130084], p. 639, Eq. 1). In particular, the Kuehn-Goldstein correlation must remain valid when the outer surface is hotter than the air. • The effective Reynolds number for natural convection at a surface depends only on the circumferential Nusselt number at that surface, as predicted by the Kuehn-Goldstein correlation. In particular, the effective Reynolds number is approximately the same as the Reynolds number for the particular forced flow at that surface that would give the same effective Nusselt number. • As proposed by Morgan (1975 [DIRS 160791]) for configurations in which the direction for natural convection is 90º from the direction for forced convection, the effective Reynolds number for mixed convection is the square root of the sum of the square of the Reynolds number for forced convection and the square of the effective Reynolds number for natural convection. The validity of these statements is demonstrated in comparison of the methodology to experimental data. IX.1.6.1 Methodology Morgan (1975 [DIRS 160791]) proposed a method for calculating the Nusselt number when both natural and forced modes of convection are present. He considered an equivalent Reynolds number for natural convection, ReN, such that the Nusselt number for natural convection would be equal to the Nusselt number for a forced convection that had a Reynolds number of ReN. In other words, for the forced-convection flow pattern of Figure IX-1b, Equations IX-21 and IX-22 provide relationships among the Reynolds number and the two effective circumferential Nusselt numbers. The Morgan approach applies these relationships to the natural-convection flow pattern of Figure IX-1a to obtain effective Reynolds numbers for natural convection. The Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-14 October 2004 equations are similar, with the Nusselt numbers and Reynolds number replaced by ( ) x NuNi , ( ) x NuNi , Ni Re , and No Re . For the natural-convection flow pattern of Figure IX-1a, the two surfaces need not have the same effective Reynolds number. By conservation of mass, the mass flow rates must be related, but the channel widths are not known. For example, the flow speed may be higher when the air is rising past the inner cylinder, because the motion is in the direction of buoyancy. At the outer cylinder the flow may be slower and occupy a wider channel. Therefore, in applying Equations IX-21 and IX-22 to natural convection (or mixed convection) each equation uses the Reynolds number appropriate to the surface. For steady pure natural convection, with or without radiation, energy conservation requires that the ratio of the convective fluxes at the two surfaces be related to the inverse of their circumferences. This additional relationship might have permitted simultaneous solution of Equations IX-21 and IX-22 in the case of natural convection. However, the appearance of an additional variable, the second Reynolds number, precludes solving the equations simultaneously. Therefore, each surface must be treated separately. Once the effective Reynolds number for natural convection at a surface is available, the Morgan procedure defines the effective Reynolds number for the mixed flow to be ReM, such that (Morgan 1975 [DIRS 160791], p. 244, Eq. 21): ( ) ( ) ( ) ( )( ) f cos 2 2 2 2 Re Re Re Re Re N N M + + = (Eq. IX-23) where Re is the Reynolds number for forced convection and f is the angle between the direction of gravity and the direction of forced flow. The total heat transfer is found by using ReM in place of Re in the forced convection correlation. Section IX.1.6.2 uses Equation IX-23, in the special case that o 90 = f , ( 0 cos = f ), for prediction of mixed convection in the EBS during ventilation. Morgan (1975 [DIRS 160791]) applied the mixed-convection technique to predict the effective Reynolds number for mixed convection in external horizontal flow that is transverse to a horizontal cylinder. He compared the predicted ratio of effective Nusselt number to forced-flow Nusselt number to the experimental values from two data sets (Morgan 1975 [DIRS 160791], p. 249, Figure 10). Section IX.3.5 uses this comparison as a sound basis for estimating the additional uncertainty that arises when Equation IX-23 is used. The mixed-convection methodology incorporates correlations of experimental data. The correlations are for idealized configurations that are not the same as the EBS configuration. With one exception, methodology development recognizes that the idealizations are not true and considers their effects in the uncertainty analysis (Section IX.3). The one exception applies to natural convection when the outer surface is hotter than the air. During forced ventilation, the ventilating air removes heat. Because the outer cylinder is heated by thermal radiation from the inner cylinder, the outer surface will be hotter than the air. As discussed in Section 5, the development of the mixed-convection methodology assumes that the Kuehn-Goldstein correlation remains valid when the outer surface is hotter than the air. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-15 October 2004 IX.1.6.2 Mathematical Methodology The methodology documented here combines the natural and forced convection formulas into mixed convection formulas. The following formulas give the effective circumferential Nusselt number for mixed convection at each surface: ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) x Re . x Re . x x Re . x Re Nu x Re Nu x Nu Mo * i Mo * o Mo * i Mo oo Mo ii Mi - + = 1 / t (Eq. IX-24) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) x Re . x Re . x x Re . x Re Nu x Re Nu x Nu Mi * i Mi * o Mi * o Mi ii Mi oo Mo - + = 1 t (Eq. IX-25) where the parameters are interpolated linearly between Reynolds numbers and values of r* in Table IX-2, ( ) ( ) [ ]2 2 x Re Re x Re Ni Mi + = (Eq. IX-26) ( ) ( ) [ ]2 2 x Re Re x Re No Mo + = (Eq. IX-27) The Morgan procedure entails finding the equivalent natural-convection Reynolds numbers, Ni Re and No Re , such that Equations IX-24 and IX-25 are satisfied with the subscript M replaced by the subscript N. To simplify the implicit equations to be solved, the methodology documented here makes the following approximations: ( ) ( ) ( ) x Re x Nu Ni Ni ii Nu = (Eq. IX-28) ( ) ( ) ( ) x Re x Nu No No oo Nu = (Eq. IX-29) with ( ) x NuNi and ( ) x NuNo given by Equations IX-10, IX-11 and Table IX-1 and with linear interpolations with respect to Reynolds number and r* in Table IX-2. The uncertainty analysis of Section IX.3.5 includes the effects of these approximations. Equations IX-10 and IX-11 are appropriate to the flow patterns of Figures IX-1a and IX-1c. The methodology documented here applies those correlations to the general case, including the flow pattern of Figure IX-1d. An application of the methodology may be represented as an algorithm, with a preparation phase to establish the dimensionless groups that are input to the methodology, a Nusselt number prediction phase that accords with the methodology, and a phase for interpretation of the calculated Nusselt numbers. The preparation phase consists of the following steps: Step P1. (Geometry) Calculate Di/Do, which is r*. Also calculate the hydraulic diameter, Do-Di. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-16 October 2004 Step P2. (Reynolds Number) Choose the axial position of interest, x. Determine the mass flow rate, um(x), and the mean fluid temperature, ( ) x Tf . Calculate Re, using Equation IX-1. Step P3. (Rayleigh Numbers) Estimate the local value of g. Determine the circumferential average temperature on each surface, ( ) x Ti and ( ) x To . Calculate ( ) ( ) x T x T f i - and ( ) ( ) x T x T f o - . Calculate ( ) x t using Equation IX-18. Calculate ) (x Rai and ) (x Rao , using Equations IX-5 and IX-6. The following steps apply the methodology to predict the mixed-convection Nusselt numbers: Step N1. (Forced-Convection Parameters) Using linear interpolation for r* in Table IX-2, establish tables for ii Nu , oo Nu , * i . , and * o . as functions of Re. Step N2. (Natural Convection) Using Equations IX-10 and IX-11 with Table IX-1, calculate ( ) x NuNi and ( ) x NuNo . Using the table created in Step N1, and using linear dependence on Re between table values, find ( ) x ReNi to satisfy Equation IX-28 and ( ) x ReNo to satisfy Equation IX-29. Step N3. (Inner-Surface Nusselt Number) Using Equation IX-26, calculate ( ) x ReMi . Using linear interpolation in the table created in Step N1, look up the values of ( ) ( ) x Re Nu Mi ii , ( ) ( ) x Re Nu Mi oo , ( ) ( ) x Re . Mi * i , and ( ) ( ) x Re . Mi * o . Using Equation IX-24, calculate ( ) x NuMi . Step N4. (Outer-Surface Nusselt Number) Using Equation IX-27, calculate ( ) x ReMo . Using linear interpolation in the table created in Step N1, look up the values of ( ) ( ) x Re Nu Mo ii , ( ) ( ) x Re Nu Mo oo , ( ) ( ) x Re . Mo * i , and ( ) ( ) x Re . Mo * o . Using Equation IX-25, calculate ( ) x NuMo . The development of the methodology supports the following interpretation of the mixed-convection Nusselt numbers: Step I1. Using Equations IX-2 and IX-3, calculate ( ) x hi and ( ) x ho . Step I1. Using Equation IX-16, calculate the two circumferential average heat fluxes and apply them uniformly over each surface. Appendix X contains an Excel spreadsheet that may be used for this algorithm. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-17 October 2004 IX.1.6.3 Methodology Limitations This section summarizes the limitations of the ventilation methodology discussed in the above sections of the appendix. The impacts of these limitations are addressed in Section IX.3. Sentences describing the limitations appear in italics. Although forced ventilation is proposed during the preclosure period, the anticipated flow rate is low enough that both natural and forced convection play a significant role in the transfer of energy. This combination increases the complexity of predicting heat transfer. Review of the literature showed that little research has been completed for mixed convection conditions, with no information found for configurations similar to the YMP drifts. The available correlations are based on measurements of stationary processes. Therefore, the mixed-convection methodology documented here applies only when the temperatures at the surfaces are varying so slowly with time that the convective processes are nearly stationary. The ventilating fluid must be air, and its velocity and other properties at every location must be varying slowly enough that processes are nearly stationary. Also, the methodology documented here uses a forced-convection correlation that is valid only for fully developed flows. Therefore, the methodology is limited to situations in which the flow is fully developed over most of the length of the drift. The EBS drift configuration is similar to an air-filled horizontal cylinder (the drift) with an interior cylindrical solid (the train of waste packages), as shown in Figure IX-2. The methodology documented here is limited to configurations in which the waste packages are spaced in the drift such that the heat generation per unit length will be nearly uniform throughout the drift. The cylinders are neither concentric nor of equal length (coextensive). The diameters of the inner and outer cylinders are Di and Do, respectively. Because the core of the methodology uses only dimensionless parameters, any consistent system of units is acceptable. The applications discussed in this appendix use SI units, so the diameters are in meters. To improve readability, this appendix indicates the SI units for each variable. e Di Do Figure IX-2. Configuration Treated by Methodology of Mixed-Convection Heat Transfer Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-18 October 2004 The parameter r* is the ratio of the diameters, Di/Do. The methodology documented here has the limitation that 0.5 = r* = 0.2. The distance between the central axes of the cylinders is e. The dimensionless eccentricity, e*, is ) /( 2 i o D D - e , positive upward. Another methodology limitation is that 0 = e* > -2/3. That is the range covered by the experimental data for natural convection. The driver for forced convection is the mean axial fluid velocity, um (m/s). Its surrogate is the dimensionless Reynolds number, Re, a measure of the ratio of inertial to viscous forces. A limitation of the mixed convection methodology is that the Reynolds number be at least 15,000 (turbulent). This is the range in which both the natural convection correlation and the forced convection correlation are valid. The coordinate pair (x, .o) specifies positions on the inner surface of the outer cylinder, with x being the longitudinal coordinate along the cylinder in the direction of flow and .o being the angle from the vertically upward direction (zenith) relative to the axis of the outer cylinder. Similarly, (x, .i) specifies positions on the outer surface of the inner cylinder, with .i being the angle from the zenith relative to the axis of the inner cylinder. If a statement applies to either surface, the subscript on . is omitted. A thermal boundary layer must develop whenever the surface temperature differs from the fluid free-stream temperature (Incropera and DeWitt 1985 [DIRS 114109], p. 251). At each longitudinal position along the annulus, the central region of the fluid has a mean temperature ( ) x Tf (K). The current methodology is limited to air, with a Prandtl number of 0.7 and all other properties evaluated at ( ) x T f . In convective processes involving heat transfer from a boundary surface exposed to a relatively low-velocity fluid stream, it is convenient to introduce a local convective heat transfer coefficient, ) , ( . x h (W/m2·K), defined implicitly by Newton’s law of cooling, Equation IX-15. The mixed-convection methodology documented here addresses the heat transfer coefficients used to predict convection in the EBS and in scaled tests of EBS designs. Therefore, another limitation of the methodology is that the inner cylinder be hotter than the outer cylinder. The methodology does not predict local heat transfer coefficients. Rather, it leads to an effective circumferential convective heat transfer coefficient ) (x h (W/m2·K), defined by Equation IX-16. IX.2 SENSITIVITY STUDY Table IX-3 presents the results of a simple sensitivity study for the algorithm, Steps N1 through N3. The table shows how the values of the two Nusselt numbers change as each of the five inputs are varied. Each sensitivity for each input is the ratio of the change in Nusselt number to the change in the dimensionless input. Appendix X contains the Excel spreadsheet that produced this sensitivity study. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-19 October 2004 Table IX-3. Sensitivity Study Ra i Ra o t Re r* Nu Mi Nu Mo Nu Mi Nu Mo 1.E+08 7.E+08 6 45,000 0.3 165 202 low 1.E+08 7.E+08 6 15,000 0.3 145 177 high 1.E+08 7.E+08 6 150,000 0.3 285 348 low 1.E+08 7.E+08 3 45,000 0.3 171 169 high 1.E+08 7.E+08 15 45,000 0.3 161 301 low 5.E+07 7.E+08 6 45,000 0.3 144 202 high 2.E+08 7.E+08 6 45,000 0.3 195 202 low 1.E+08 2.E+08 6 45,000 0.3 165 168 high 1.E+08 2.E+09 6 45,000 0.3 165 243 low 1.E+08 7.E+08 6 45,000 0.2 253 203 high 1.E+08 7.E+08 6 45,000 0.5 103 189 1E-03 1E-03 -9E-01 1E+01 3E-07 4E-08 -5E+02 -5E+01 Sensitivity Re t Ra i Ra o r* Base Values Case IX.3 UNCERTAINTY ANALYSIS Both the forced and the natural convection correlations used in developing the mixed convection methodology are empirical or semi-empirical in nature. Thus, there is some inherent uncertainty associated with each correlation separately. Combining these equations into a mixed convection equation further increases the uncertainty. This section describes a comprehensive analysis of the overall uncertainty in the mixed convection equations. This analysis discusses the uncertainty in the predictions without reference to any particular application. Therefore, it does not consider uncertainties in the dimensionless groups that are inputs to the methodology. Those uncertainties must be addressed by making use of the sensitivity study of Section IX.2. Section IX.4 provides examples of quantifying uncertainty from all sources using data from the EBS Ventilation Test Series. IX.3.1 Definitions There is no standard for the expression of uncertainty in predictions made with algorithms. However, algorithms are used to predict measurements. The treatment of uncertainty in this appendix is based on ANSI/NCSL Z540-2-1997, American National Standard for Calibration — U.S. Guide to the Expression of Uncertainty in Measurement [DIRS 157394]. The following are adapted from definitions that appear in the standard: 1. The measurand is the particular quantity subject to measurement and therefore to prediction. Its definition may require specification of the conditions under which the quantity is measured. The standard avoids the phrase “true value of the measurand” because the word “true” is viewed as redundant. The “true value of the measurand” is simply the value of the measurand (ANSI/NCSL Z540-2-1997 [DIRS 157394], p. 41). 2. The measurement error is the result of the measurement minus the value of the measurand (ANSI/NCSL Z540-2-1997 [DIRS 157394], p. 34). As used in this — — Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-20 October 2004 appendix, the prediction error is the result of the prediction minus the value of the measurand. 3. A random component of prediction error is an effect that, for multiple predictions with varying inputs, produces a mean error that is small relative to the standard deviation of the error from that effect. An example of a random component is the residual error after a formula has been adjusted to correlate with data. 4. A systematic component of prediction error is an effect that is not a random component. 5. If the systematic component of prediction error includes a systematic effect that is quantifiable, one may add a correction to the prediction to compensate for that effect. However, the necessary correction may not be practical in the intended application of the prediction. 6. The uncertainty of the result of a prediction is an estimate of the likelihood of nearness to the best value that is consistent with presently available knowledge (adapted from ANSI/NCSL Z540-2-1997 [DIRS 157394], p. 41). Components of uncertainty include estimates of random error, uncertainties in corrections, and estimates of uncorrected or unrecognized systematic effects. 7. Standard uncertainty u (x), of a predicted value x is the uncertainty of the result of a prediction expressed as a standard deviation. It does not correspond to a high level of confidence. 8. A Type A evaluation of uncertainty is an evaluation by statistical analysis of a series of observations. A Type B evaluation of uncertainty is an evaluation by any other method. A Type B evaluation is founded on an a priori distribution of the possible values (ANSI/NCSL Z540-2-1997 [DIRS 157394], p. 3). 9. If the result of a prediction is a function of the values of a number of other quantities, the standard uncertainty in the prediction is the combined standard uncertainty. 10. For contributions to uncertainty that are independent, the law of propagation of uncertainty (ANSI/NCSL Z540-2-1997 [DIRS 157394], p. 19) determines the combined standard uncertainty. For y = f (xi . . . , xn), the combined standard uncertainty uc (y) is given by: ( ) i i N i c x u x f y u 2 2 1 2 ) ( . .. . . .. . =. = d d (Eq. IX-30) 11. In some applications, it may be necessary to have a measure of uncertainty that encompasses a large fraction of the values that one could reasonably attribute to the measurand. If necessary, the user may multiply the standard uncertainty by a coverage factor to obtain an expanded uncertainty. In general, the coverage factor will be in the range 2 to 3 (ANSI/NCSL Z540-2-1997 [DIRS 157394], p. 24). This Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-21 October 2004 appendix uses a coverage factor of 2 to approximate a 95% confidence interval (ASME PTC 19.1-1998 [DIRS 153195], p. 95). 12. The relative combined standard uncertainty in a predicted positive value y is uc (y)/y (ANSI/NCSL Z540-2-1997 [DIRS 157394], p. 25, Sect. 7.2.1). 13. For non-zero values of the xi, Equation IX-30 may be rewritten for propagation of relative uncertainty: ( ) ( ) 2 2 1 2 .. . .. . . .. . . .. . . . = .. . .. . .= i i i i N i c x x u x f y x y y u (Eq. IX-31) IX.3.2 Interpolation Errors The uncertainty analysis considers errors arising from interpolation in Table IX-2. Consider one variable at a time. Let x be the variable and y be a parameter defined by f(x). The error in linear interpolation for y is (Conte and de Boor 1972 [DIRS 159800], pp. 211-212, Example 4.5): ) ( 2 ) )( ( 2 1 . f x x x x ' - - where x is the input variable, (x1, x2) is the interpolation interval, and ( ) . f ' is a value of the second derivative of y with respect to x at some point . in the interval (x1, x2). The relative error is: M x x x x ' - - 2 ) )( ( 2 1 The maximum value of 2 ) )( ( 2 1 x x x x - - in the interval (x1, x2) is 8 ) ( 2 1 2 x x - . Therefore, the maximum relative error in the interval is 8 ) ( 2 1 2 M x x ' - . Suppose the interpolation interval is not at the edge of the table, so that for x0 < x1 < x2 < x3 we have the values y0, y1, y2, and y3. By the mean value theorem for derivatives (Conte and de Boor 1972 [DIRS 159800], p. 23, Theorem 1.6), there is xa in (x0, x1) where ( ) 0 1 0 1 x x y y x f a - - = ' (Eq. IX-32) Similarly, there is xb in (x2, x3) where ( ) 2 3 2 3 x x y y x f b - - = ' (Eq. IX-33) Applying the mean value theorem for derivatives one more time, there is xm in (xa, xb) where Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-22 October 2004 ( ) ( ) ( ) a b a b m x x x f x f x f - ' - ' = ' (Eq. IX-34) Because the smallest possible value of a b x x - is 1 2 x x - , ( ) 1 2 0 1 0 1 2 3 2 3 2 2 d d x x x x y y x x y y x x y m - - - - - - = (Eq. IX-35) so that the right hand side is a high estimate of the second derivative somewhere in (x0, x3). We use it as though it were a high estimate of the magnitude of ( ) . f ' . Therefore, a high estimate for the magnitude of M ’ is: ( ) 0 1 0 1 2 3 2 3 2 1 1 2 ) , Min( 1 x x y y x x y y y y x x - - - - - - where Min (y1, y2) is the minimum value of y between x1 and x2. We now address interpolation with respect to r*. Table IX-4 provides the parameter table for r* values of 0.1, 0.2, 0.5, and 0.8 (Kays and Leung 1963 [DIRS 160763], pp. 552 to 554) and the calculation of M ’ . All calculations in this appendix were performed to many significant digits, with the results being rounded for presentation in tables. Consider, for example, the parameter Nuii as a function of r*, with Re fixed at 10,000. The values of x0, x1, x2, and x3 are 0.1, 0.2, 0.5, and 0.8. The values of y0, y1, y2, and y3 are 48.5, 38.6, 30.9, and 28.5. At some unknown location xa between 0.1 and 0.2, the derivative of the function is (38.6 - 48.5)/(0.2 - 0.1), which is -99. Similarly, there is a location xb between 0.5 and 0.8 where the derivative of the function is (28.5 - 30.9)/(0.8 - 0.5), or -8.0. Therefore, there is some xm between xa and xb, where the second derivative is [-8.0 - (-99]/(xb - xa), or 91/(xb - xa). We do not know the value of either xa or xb, but we know that one is not larger than 0.2 and the other is no smaller than 0.5, so that their difference must be at least 0.3. Therefore, we know that there is some point between 0.1 and 0.8 where the second derivative is less than 303 (about 91 divided by 0.3). If the function is sufficiently smooth, xm will be between x1 and x2. To get a high estimate of M ’ , we divide by the smallest value of y between x1 and x2, which is 30.9. Therefore, the high estimate of M ’ is 9.8 (303 divided by 30.9). In fact, this is the largest value for any function in Table IX-4, so for interpolation with respect to r*, we take the upper bound on M ’ to be 10. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-23 October 2004 Table IX-4. High Estimate of Second Derivative with Respect to r* Re 10,000 r* Nuii .i Nuoo .o 0.1 48.5 0.512 29.8 0.032 0.2 38.6 0.412 29.4 0.063 0.5 30.9 0.300 28.3 0.137 0.8 28.5 0.224 28.0 0.192 dy /dx (x a ) -99.0 -1.0 -4.0 0.3 dy /dx (x b ) -8.0 -0.3 -1.0 0.2 d2y /dx 2 (x m ) 303.3 2.5 10.0 0.4 M' 9.8 8.296 0.4 6.702 Re 30,000 r* Nuii .i Nuoo .o 0.1 98.0 0.407 66.0 0.028 0.2 79.8 0.338 64.3 0.055 0.5 66.0 0.258 62.0 0.119 0.8 62.3 0.212 61.0 0.166 dy /dx (x a ) -182.0 -0.7 -17.0 0.3 dy /dx (x b ) -12.3 -0.2 -3.3 0.2 d2y /dx 2 (x m ) 565.6 1.8 45.6 0.4 M' 8.6 6.934 0.7 6.869 Re 100,000 r* Nuii .i Nuoo .o 0.1 235.0 0.338 167.0 0.024 0.2 196.0 0.286 165.0 0.049 0.5 166.0 0.225 158.0 0.107 0.8 157.0 0.186 156.0 0.150 dy /dx (x a ) -390.0 -0.5 -20.0 0.3 dy /dx (x b ) -30.0 -0.1 -6.7 0.1 d2y /dx 2 (x m ) 1200.0 1.3 44.4 0.4 M' 7.2 5.778 0.3 7.256 Now consider interpolation with respect to Reynolds numbers. For each interval (Ren, Ren+1) in the table, Ren+1 is approximately 3Ren. Therefore, the maximum relative error, 8 ) ( 2 1 M Re Re n n ' - + , is ( ) M Ren ' 2 5 . 0 . For the EBS Ventilation Test Series, the inner and outer diameters were 40.64 cm and 1.37 m (BSC 2003 [DIRS 160724], Sections 2.2.2.1 and 2.2.2.2), so that r* = 0.297. Table IX-5 contains the parameter table after interpolation to r* = 0.297. In order to have four values of Re, we take the values for 300,000 from Kays and Leung (1963 [DIRS 160763], pp. 552 to 554). Table IX-5 shows the derivation of M ’ and the values of ( ) M Re ' 2 1 5 . 0 . Again consider the calculation for the parameter Nuii, this time as a function of Re, with r* fixed at 0.297. Now the Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-24 October 2004 values of x0, x1, x2, and x3 are 10,000, 30,000, 100,000, and 300,000, and the values of y0, y1, y2, and y3 are 36.1, 75.4, 186.3, and 449.5. At some unknown Reynolds number xa between 10,000 and 30,000, the derivative of the function is (75.4 - 36.1)/(30000 - 10000), which is 1.96 × 10-3. Similarly, there is a Reynold number xb between 100,000 and 300,000 where the derivative of the function is (449.5 - 186.3)/(300000 - 100000), or 1.32 × 10-3. Therefore, there is some xm between xa and xb, where the second derivative is (1.32 × 10-3 - 1.96×10-3)/(xb - xa), or -6.4×10-4/(xb - xa). We do not know the value of either xa or xb, but we know that one is not larger than 30,000 and the other is no smaller than 100,000, so that their difference must be at least 70,000. Therefore, we know that there is some Reynolds number between 10,000 and 300,000 where the magnitude of the second derivative is less than 9.23 × 10-9 (about 6.4×10-4 divided by 70,000). If the function is sufficiently smooth, xm will be between x1 and x2. To get a high estimate of M ’ , we divide by the smallest value of y between x1 and x2, which is 75.4. Therefore, the high estimate of M ’ is 1.22 × 10-10 (about 9.2 × 10-9 divided by 75.4) and the maximum relative error for Reynolds numbers between 30,000 and 100,000 is 5.5%, which is 0.5 × (30,000)2 × 1.22 × 10-10. The results in Table IX-5 indicate that the relative error should be no more than about 7%. Table IX-5. High Estimate of Second Derivative with Respect to Re r* 0.297 Re Nuii .i Nuoo .o 10,000 36.1 0.376 29.0 0.087 30,000 75.4 0.312 63.6 0.076 100,000 186.3 0.266 162.7 0.068 300,000 449.5 0.243 391.5 0.061 dy /dx (x a ) 1.96E-03 -3.18E-06 1.73E-03 -5.61E-07 dy /dx (x b ) 1.32E-03 -1.19E-07 1.14E-03 -3.31E-08 d2y /dx 2 (x m ) 9.23E-09 4.38E-11 8.31E-09 7.54E-12 M' 1.22E-10 1.64E-10 1.31E-10 1.11E-10 0.5M' (30,000)^2 5.5% 7.4% 5.9% 5.0% dx /dy (y a ) 5.10E+02 5.79E+02 dx /dy (y b ) 7.60E+02 8.74E+02 d2x /dy 2 (y m ) 2.26E+00 2.97E+00 M' 7.52E-05 9.90E-05 M' (y 2-y 1)2/8 11.6% 12.2% At the bottom of Table IX-5 is an error analysis for the reverse interpolation for ( ) x ReNi or ( ) x ReNo , starting from Nuii or Nuoo. The roles of the variables are reversed. For example, the first derivative of Re with respect to Nuii, at some value of Nuii between 36.1 and 75.4, is 509 [=(30,000 - 10,000)/(75.4 - 36.1)]. Similarly, the value of the derivative is 760 somewhere between 186 and 450. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-25 October 2004 Therefore, the value of the second derivative is less than (760 - 509)/(186.3 - 75.4), which is about 2.26, at some value of Nuii between 36.1 and 450. Dividing by the smallest value of Re in the intermediate interval, 30,000, yields 7.5 × 10-5 as a high estimate for M ’ . Therefore, the relative error should be no more than 11.6%, which is the result of M ’ × (186.3 - 75.4)2/8. Table IX-5 shows that a similar analysis for reverse interpolation from Nuoo gives 12.2% as a high estimate of the relative error. IX.3.3 Uncertainty in Nusselt Numbers from the Natural Convection Methodology This section evaluates the following sources of uncertainty in effective circumferential Nusselt numbers calculated from the natural convection correlation: 1. The extent to which measured Nusselt numbers for concentric, coextensive, isothermal cylinders deviate from the effective circumferential Nusselt numbers predicted by the correlation. 2. The uncertainty arising from applying a correlation for a diameter ratio of 0.38 to configurations with other diameter ratios in the range 0.2 to 0.5. 3. Uncertainty in the effective circumferential Nusselt numbers arising from temperature variation along the lengths of the cylinders. This section also discusses the following sources of uncertainty, which must be evaluated for each particular application: 1. The uncertainty in effective circumferential Nusselt numbers for coextensive, isothermal cylinders arising from eccentric location of the inner cylinder. 2. The uncertainty in effective circumferential Nusselt numbers for isothermal cylinders arising from unequal cylinder lengths. 3. Uncertainty in the effective circumferential Nusselt numbers arising from temperature variation along the circumferences of the cylinders. This uncertainty analysis does not estimate the error from applying the natural convection correlation to situations in which the outer surface is hotter than the air (Figure IX-1d). IX.3.3.1 Deviations in Measured N Nu for Concentric, Coextensive, and Isothermal Cylinders This section estimates the uncertainty inherent in the Kuehn-Goldstein correlation, even when applied to idealized configurations. One definition of a Nusselt number for overall convection between the two cylinders is: Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-26 October 2004 ( ) o i i i conv T T k q D Nu - ' ' = (Eq. IX-36) As suggested by Kuehn and Goldstein (1978 [DIRS 130084], p. 639), the three Nusselt numbers are related by: ( ) ( ) ( ) ( ) ( ) 1 * * 1 1 1 1 1 1 - - . . . . . . . . - + - = . .. . . .. . ' + ' = x Nu r x Nu r u N u N x Nu o i o i conv (Eq. IX-37) Kuehn and Goldstein (1978 [DIRS 130084]) correlated results of 40 tests for which Pr was 0.7, r* was 0.38, and the Rayleigh number ranged widely. After correcting test data for end losses and radiation, they determined that their correlation fit conv Nu for Ra > 5,000 (33 tests) with a standard deviation of 1.7% (Kuehn and Goldstein 1978 [DIRS 130084], p. 639). However, a fit to conv Nu does not require a fit to each of the values of Ni Nu and No Nu . For example, Ni Nu could be too large and No Nu too small. These could combine to produce the correct value of conv Nu , but the predicted value of f T would be too large. Kuehn and Goldstein (1978 [DIRS 130084], p. 636) define an average dimensionless fluid temperature, b F , by ( ) ( ) o i o f b T T T T F - - = (Eq. IX-38) They report that the dimensionless average fluid temperature near the center of the gap obtained from the correlation agrees “fairly well” with the experimental results. They give only one example, for which the experimental result is 0.25 compared to 0.28 given by the correlation (Kuehn and Goldstein 1978 [DIRS 130084], p. 639). In steady natural convection, the total heat flux at the two cylinders must be equal and opposite. That is: 0 0 D q D q i i ' - = ' (Eq. IX-39) The average dimensionless fluid temperature, b F , is related to the ratio 0 Nu Nuconv , or alternatively to the ratio i conv Nu Nu . A derivation of the relationship between b F and 0 Nu Nuconv starts with Equation IX-28, first substitutes for the temperature differences from Equations IX-16 and IX-36, and then uses Equations IX-39 and IX-3 to simplify. The resulting expression can be converted to use i conv Nu Nu by applying Equation IX-37. The result is: Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-27 October 2004 ( ) ( ) i conv o conv o conv i o conv i i o o b Nu Nu r Nu Nu r l Nu D Nu D D Nu k q D h q F .. . .. . - - = - = - = ' ' - = 1 1 1 " * * 0 (Eq. IX-40) Therefore, if conv Nu is relatively accurate, an error of +12% in b f corresponds to a value of Ni Nu that is about 12% too high and a value of No Nu that is about 12% too low. This appendix uses an a priori normal distribution for a Type B evaluation of the uncertainty inherent in the Kuehn-Goldstein correlation. This analysis assigns a relative standard uncertainty of 12% to predictions of Ni Nu and No Nu for concentric, coextensive, isothermal cylinders with r* = 0.38. The smaller error for conv Nu indicates that the errors in Ni Nu and No Nu tend to be equal and opposite. With only one data point available, this uncertainty analysis treats the error as random rather than systematic. IX.3.3.2 Uncertainty in N Nu from Diameter Ratio For Ra > 108, Pr = 0.7, and r* = 0.33 or 0.5, the agreement between the Kuehn-Goldstein correlation for natural convection was within 5% of the experimental data (Kuehn and Goldstein 1976 [DIRS 100675], Figure 2). Kuehn and Goldstein (1978 [DIRS 130084], p. 639, Eq. 1) presented a modified correlation for the same data which provided an even better fit. Therefore, this appendix neglects any additional uncertainty for r* between 0.2 and 0.5. IX.3.3.3 Uncertainty in N Nu from Longitudinal Temperature Variation In natural convection, there should be no longitudinal gradient. To the extent that there are longitudinal gradients in a natural convection test, they are considered to be the result of end effects, and the test results are corrected for these effects. Therefore, there is no uncertainty associated with longitudinal temperature gradients in natural convection. In mixed convection, there is a longitudinal gradient that is expected from forced convection. In this appendix, any effects on the natural convection Nusselt number from a longitudinal temperature gradient are included in the uncertainties inherent in combining the two correlations into a mixed convection model. IX.3.3.4 Uncertainty in N Nu from Eccentricity Kuehn and Goldstein (1978 [DIRS 130084], p. 637) reported the effects of eccentricity on heat transfer coefficients. The overall heat transfer coefficients tend to increase by 10 percent as the inner cylinder is moved downward from the concentric position to an e* of -2/3. The methodology is limited to values between 0 and –2/3 (Section IX.1.6.3). This appendix uses that information for a Type B evaluation of uncertainty by assuming that the error is linear with the eccentricity. That is, the use of the concentric correlation systematically underestimates the heat transfer coefficients. The fractional error is about -0.15 .e*., so that it would be zero if the cylinders were concentric and is -0.1 when the value of e* is -2/3. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-28 October 2004 IX.3.3.5 Uncertainty in N Nu from Unequal Cylinder Lengths In a particular configuration, the outer cylinder may be longer than the inner cylinder. The inner cylinder may be a series of waste packages with gaps between them, reducing the heated length of the inner cylinder. The additional area of the outer cylinder may permit more convective heat transfer from the air to the outer cylinder. Therefore, the air temperature may be closer to the temperature of the outer cylinder than it would be if the inner cylinder extended the entire length of the outer cylinder. The contribution to uncertainty from the length difference may be neglected if the following conditions hold: 1. The greater length of the outer cylinder does not cause a qualitative change in the flow from natural convection other than mild divergence and convergence along the axis. 2. The change in the Rayleigh numbers appearing in the correlation, caused by the change in air temperature, properly accounts for most of the changes in the circumferential average Nusselt numbers. 3. The remaining effects of the longer outer cylinder are not significant compared to the other contributions to uncertainty. IX.3.3.6 Uncertainty in N Nu from Circumferential Temperature Variation It may be that heat transfer by thermal conductivity within one or both cylinders is not sufficient to maintain a cylinder at nearly uniform temperature. In such a case, one must consider how accurately the natural convection correlation predicts an effective circumferential Nusselt number. Kuehn and Goldstein (1978 [DIRS 130084], p. 637) observed that moving a heated inner cylinder below its concentric position results in more uniform local coefficients on the outer cylinder. However, this uncertainty analysis uses results for concentric cylinders. Because the inner cylinder is hotter than the outer cylinder and the flow develops as shown in Figure IX-1a, natural convection cools the bottom of the inner cylinder more effectively than the top and transfers heat to the top of the outer cylinder more effectively than to the bottom. Therefore, both cylinders are hotter at the top than at the bottom. For pure natural convection, in which Tf is between the temperatures of the cylinders, the magnitude of the temperature difference between the inner cylinder and the fluid is smallest at the bottom. For the outer cylinder, on the other hand, the difference is smallest at the top. Kuehn and Goldstein (1978 [DIRS 130084]) obtained temperature distributions and local heat transfer coefficients using time-averaged interferograms. For four Rayleigh numbers, they plotted local equivalent conductivities (which are proportional to the local heat transfer coefficients) for isothermal cylinders as a function of angular position numbers (Kuehn and Goldstein 1978 [DIRS 130084], Figure 8). Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-29 October 2004 First, consider the inner cylinder. Because the inner cylinder is hotter than the outer cylinder and the flow develops as shown in Figure IX-1a, natural convection cools the bottom of the inner cylinder more effectively than the top. Therefore, if conduction within the cylinder is not sufficient to maintain a uniform temperature, the inner cylinder is hotter at the top than at the bottom. The magnitude of the temperature difference between the inner cylinder and the fluid is smallest at the bottom. For four Rayleigh numbers, Kuehn and Goldstein (1978 [DIRS 130084], Figure 8) plotted local heat transfer coefficients for concentric isothermal cylinders as a function of angular position. Their plot shows that hi is smallest at the top of the cylinder, may increase by a factor of five or more at the sides, and stays within 50% of that value along the bottom half of the cylinder. To estimate the effect of deviations from temperature uniformity around the inner cylinder, we use the approximation that the heat transfer coefficients are not affected. We let hT be the heat transfer coefficient around the top quarter of the cylinder and assign 5hT as the heat transfer coefficient around the rest of the circumference. For an isothermal inner cylinder, the effective circumferential heat transfer coefficient is the same as the average, which is 4ht. For varying temperatures at the top, left, bottom, and right, each representing the average over one-quarter of the circumference, we have: 4 ] [ 5 ] [ 5 ] [ 5 ] [ ] [ f b t f r t f l t f t t f i i T T h T T h T T h T T h T T h - + - + - + - = - (Eq. IX-41) Adding and subtracting ht [Tt – Tf] on the right and letting i T be the average of the four temperatures, we obtain: ] [ ] [ 5 ] [ f t t f i t f i i T T h T T h T T h - - - = - (Eq. IX-42) ] [ ] [ ] [ ] [ 4 ] [ ] [ 5 f i f t t f i f i t t f i f t t t i T T T T h T T T T h h T T T T h h h - - - - - + = - - - = (Eq. IX-43) ] [ ] [ 4 f i i t t t i T T T T h h h - - - = (Eq. IX-44) The error from using the average, t t t t t i h h h h h h 4 4 5 5 5 ~ = + + + = (Eq. IX-45) is ] [ ] [ ~ 25 . 0 ] [ ] [ ~ f i i t i f i i t i i i T T T T h T T T T h h h - - + = - - = - (Eq. IX-46) Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-30 October 2004 so that the relative error is about ] [ ] [ 25 . 0 f i i t T T T T - - + . This is also the relative error in the inner-cylinder Nusselt number. Now, consider the outer cylinder. As shown in Figure IX-1a, natural convection transfers heat to the top of the outer cylinder more effectively than to the bottom. Therefore, if conduction within the cylinder is not sufficient to maintain a uniform temperature, the outer cylinder is hotter at the top than at the bottom. For pure natural convection, in which Tf is between the temperatures of the cylinders, the magnitude of the temperature difference between the inner cylinder and the fluid is smallest at the top. The Kuehn and Goldstein chart (1978 [DIRS 130084], Figure 8) shows that ho is largest at the top of the outer cylinder, drops by a factor of three or more at the sides, and drops to zero along the bottom of the cylinder. To estimate the effect of deviations from temperature uniformity around the outer cylinder, we again use the approximation that the heat transfer coefficients are not affected. We let ht be the heat transfer coefficient around the top quarter of the cylinder and assign ht/3 as the heat transfer coefficient at the sides. As in the application to the EBS tests, we exclude the bottom quarter from the analysis. For an isothermal outer cylinder, the effective circumferential heat transfer coefficient is the same as the average of the three coefficients, which is 5ht/9. For varying temperatures at the top, left, and right, each representing the average over one-quarter of the circumference, and using the relative heat transfer coefficients from the previous paragraph, we have 3 ] [ 3 1 ] [ 3 1 ] [ ] [ R f t l f t t f t o f o T T h T T h T T h T T h - + - + - = - (Eq. IX-47) In this case, we subtract and add [ ] t f t T T h - 9 2 on the right and let o T be the average of these temperatures to obtain ] [ 9 2 ] [ 3 1 ] [ t f t o f t o f o T T h T T h T T h - + - = - (Eq. IX-48) ] [ ] [ 9 2 ] [ ] [ 9 2 9 5 ] [ ] [ 9 2 3 1 o f t f t o f o f t t o f t f t t o T T T T h T T T T h h T T T T h h h - - + - - - = - - + = (Eq. IX-49) ] [ ] [ 9 2 9 5 o f o t t t o T T T T h h h - - - = (Eq. IX-50) The error from using the average, Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-31 October 2004 t t t t o h h h h h 9 5 3 3 1 3 1 ~ = + + = (Eq. IX-51) is ] [ ] [ ~ 5 2 ] [ ] [ 9 2 ~ o f o t o o f o t t o o T T T T h T T T T h h h - - + = - - + = - (Eq. IX-52) so that the relative error is about ] [ ] [ 4 . 0 o f o t T T T T - - + . This is also the relative error in the outer-cylinder Nusselt number. IX.3.4 Uncertainty in Nusselt Numbers from the Forced Convection Correlation This uncertainty analysis evaluates the following sources of uncertainty in effective circumferential Nusselt numbers calculated from the forced convection correlation: 1. Uncertainty in the Nusselt numbers arising from flux variation along the lengths of the cylinders. 2. Uncertainty in the effective circumferential Nusselt numbers arising from flux variation along the circumferences of the cylinders. This section also discusses the following sources of uncertainty, which must be evaluated for each particular application: 1. The extent to which measured Nusselt numbers for fully developed flow in concentric, coextensive, uniform-flux cylinders deviate from the Nusselt numbers predicted by the correlation. 2. Uncertainty from linear interpolation of the Kays-Leung parameters to the diameter ratio of 0.3. 3. Uncertainty from linear interpolation of the Kays-Leung parameters to the appropriate Reynolds number. 4. The uncertainty in effective circumferential Nusselt numbers for fully developed flow in uniform-flux cylinders arising from unequal cylinder lengths. 5. The uncertainty in effective circumferential Nusselt numbers for fully developed flow in coextensive, uniform-flux cylinders arising from eccentric location of the inner cylinder. 6. Uncertainty arising from deviations from fully developed flow. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-32 October 2004 IX.3.4.1 Uncertainty in Measured F Nu from Flux Variation along the Lengths of the Cylinders The methodology documented here is limited to configurations in which the waste packages are spaced in the drift such that the heat generation will be roughly constant per unit length of drift. Therefore, the surface flux should vary sufficiently slowly in the axial direction that the flow at each location is approximately the same as if that flux were uniform over the length of the cylinder. Consequently, this analysis neglects that source of error. IX.3.4.2 Uncertainty in Measured F Nu from Flux Variation along the Circumferences of the Cylinders In pure forced convection with uniform boundary conditions, there is no dependence on the angle .. Sutherland and Kays (1964 [DIRS 160789], p. 1189) considered fully developed flow in a concentric annulus with heat flux varying circumferentially, but not axially. They represented the heat fluxes at each surface as a Fourier series of the form: ( ) ( ) ( ) [ ] n. b n. a q n n n cos sin 0 + = ' .8 = . (Eq. IX-53) Neglecting thermal conduction in the walls, they derived (Sutherland and Kays 1964 [DIRS 160789], p. 1189, Eqs. 3a and 3b): ( ) ( ) ( ) [ ] ( ) ( ) [ ] n. b n. a R n. b n. a R T . T o o io i o i i ii i o n n n n k D D n n n n k D D f i cos sin cos sin 0 0 + + + = - . . 8 = - 8 = - (Eq. IX-54) ( ) ( ) ( ) [ ] ( ) ( ) [ ] n. b n. a R n. b n. a R T . T o o oo i o i i oi i o n n n n k D D n n n n k D D f cos sin cos sin 0 0 0 + + + = - . . 8 = - 8 = - (Eq. IX-55) where the Rn are the eigenfunctions when only one wall is heated. The index n indicates the harmonic of the Fourier expansion, its first subscript is the affected wall, and its second subscript is the heated wall. Integration of those equations over . (over 2p) yields: i i b q 0 2 p = ' (Eq. IX-56) 0 0 2 b qo p = ' (Eq. IX-57) Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-33 October 2004 ( ) o i i o f i q R q R b R b R D D T T k io ii o io i ii ' + ' = + = - - 0 0 0 0 0 0 2 2 p p (Eq. IX-58) ( ) o i i o f q R q R b R b R D D T T k oo oi o oo i oi ' + ' = + = - - 0 0 0 0 0 0 0 2 2 p p (Eq. IX-59) Therefore, the relationships between the mean heat fluxes and the mean temperatures are independent of any axial variation. Consequently, this appendix neglects the uncertainty arising from flux variation around the circumference. IX.3.4.3 Deviations in Measured F Nu for Concentric, Coextensive, Uniform-Flux Cylinders The concentric tubes were mounted vertically with airflow from the bottom upward (Reynolds et al. 1963 [DIRS 160770], p. 489). The reported experimental data reflect correction for radiative heat transfer (Kays and Leung 1963 [DIRS 160763], p. 540). Correction for the effects of natural convection were not necessary (as explained previously). They reported measurement uncertainties of about 3% in ii Nu and oo Nu after correction for radiative heat transfer (Kays and Leung 1963 [DIRS 160763], p. 541). They presented the asymptotic Nusselt numbers, both analytical and experimental, for various values of r*, including 0.255, 0.376, and 0.5 (Kays and Leung 1963 [DIRS 160763], pp. 544-545, Figures 6 to 8). The measurements were consistently within 3% of the correlation, except that Nuii tended to deviate from the experimental data at Reynolds numbers below 20,000. At Re = 15,000 and r* = 0.255, for instance, the correlation predicts a value for Nuii that is about 10% high (the two labels for “Present analysis” in their Figure 8 having been transposed inadvertently). The mixed-convection methodology is limited to Reynolds numbers greater than 15,000 (Section IX.1.6.3). Because the contributions of natural convection to the experimental results can be neglected (Section IX.1.5), this appendix uses an a priori normal distribution for a Type B evaluation of the uncertainty inherent in the Kays-Leung correlation. For concentric, coextensive, uniform-flux cylinders, this appendix assigns a relative uncertainty of 3% as the random component and an additional systematic error in ( ) x NuFi that decreases linearly from 10% to zero as Re increases from 15,000 to 20,000. IX.3.4.4 Uncertainty in F Nu from Linear Interpolation in Diameter Ratio The error in linear interpolation for y is (Conte and de Boor 1972 [DIRS 159800], pp. 211-212, Example 4.5). 2/ ) )( ( 1 0 M x x x x - - where x is the input variable, (x0, x1) is the interpolation interval, and M is a value of the second derivative of y with respect to x somewhere in (x0, x1). The analysis in Section IX.3.2 suggests Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-34 October 2004 that the absolute value of M is no more than 10 times the value of the parameters. For 0.5 = r* = 0.2, and letting the worst case value of M correspond to the 95% confidence limit, one may take: )5 . 0 )( 2 . 0 ( 5 * * r r - - as the upper 95% confidence limit in the relative error caused by interpolation in r*, so that the standard relative uncertainty would be one-half of that value. IX.3.4.5 Uncertainty in F Nu from Linear Interpolation in Reynolds Number The maximum error in an interpolation interval is (Conte and de Boor 1972 [DIRS 159800], pp. 211-212, Example 4.5): 8/ ) ( 2 0 1 M x x - where x is the input variable, (x0, x1) is the interpolation interval, and M is a value of the second derivative of y with respect to x somewhere in (x0, x1). For the Kays-Leung tables, in which 0 1 3Re Re = , the maximum error is ( ) M Re 2 0 5 . 0 . Section IX.3.2 provides an example demonstrating the evaluation of M from the table developed in Step P2. One may take ( ) M Re 2 0 5 . 0 as the upper 95% confidence limit in the error caused by interpolation in Re. IX.3.4.6 Uncertainty in F Nu from Eccentric Location of the Inner Cylinder Here we estimate the error caused by using a correlation developed for concentric cylinders to predict forced-convection Nusselt numbers for eccentric configurations. Our analysis is based on a review of experimental results for the turbulent flow of air in an eccentric annulus with fully developed constant heat rate (Kays and Perkins 1973 [DIRS 160782], pp. 7-109 to 7-110, Figures 89 and 90). Although the Nusselt number is uniform around the cylinder for the concentric configuration, eccentricity introduces circumferential variation in the Nusselt numbers. The cited charts provide the ratio of the local Nusselt number to the concentric value, as a function of positive eccentricity, for two opposite locations on the cylinder and two values of r*. For pure forced convection, there is no difference between positive and negative eccentricity. The locations where the cylinders are most separated (labeled A in the figures) correspond to the tops of the cylinders in a configuration with negative eccentricity. First, we consider the inner cylinder. We take the effect on Nuii as an estimate of the effect on ( ) x NuFi . We consider only the effect at the bottom, where the local heat transfer coefficient may be greater by about a factor of 5 (from natural convection; see Kuehn and Goldstein 1978 [DIRS 130084], Figure 8). For the two values of r*, with heating from the inner surface and the outer surface insulated, piecewise linear fits (by inspection) to the data in the region of interest (Kays and Perkins 1973 [DIRS 160782], pp. 7-109, Figure 89, “B”) result in the following approximations: Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-35 October 2004 ( ) ( ) , 15 . 0 1 Nu * ii e Re x NuFi + = r* = 0.255, 0 = e* = -0.67 (Eq. IX-60) ( ) ( ) , 1 Nuii = Re x NuFi r* = 0.5, 0 = e* = -0.27 (Eq. IX-61) ( ) ( ) ( ), 27 . 0 5 . 0 1 Nu * ii + + = e Re x NuFi r* = 0.5, -0.27 = e* = -0.67 (Eq. IX-62) The following general form for the relative error, ( ) ( ) Re x NuFi ii Nu 1- , covers the range 0.5 = r* = 0.2 and matches the above equations at r* = 0.255 and r* = 0.5: ( ) ( ) , 7 . 0 15 . 0 Nu 1 * * ii e r Re x NuFi - - = - 0.3 = r* = 0.2, 0 = e* = -0.67 (Eq. IX-63) ( ) ( ) , 0 Nu 1 ii = - Re x NuFi 0.5 = r* = 0.3, 0 = e* = 0.405-1.35r* (Eq. IX-64) ( ) ( ) ( ), 405 . 0 35 . 1 7 . 0 15 . 0 Nu 1 * * * ii - + - - = - r e r Re x NuFi 0.5 = r* = 0.3, 0.405-1.35r* = e* = -0.67 (Eq. IX-65) For the outer cylinder, we take the effect on Nuoo as representative of the effect on ( ) x NuFo . We consider only the effect at the top, because the local heat transfer coefficient may drop to zero at the bottom (from natural convection; see Section IX.3.3). For the two values of r*, with heating from the outer surface and the inner surface insulated, the ratio is approximately (Kays and Perkins 1973 [DIRS 160782], p. 7-110, Figure 90, “A”): ( ) ( ) , 1 Nuoo = Re x NuFo r* = 0.255, 0 = e* = -0.4 (Eq. IX-66) ( ) ( ) ( ), 4 . 0 15 . 0 1 Nu * oo + + = e Re x NuFo r* = 0.255, -0.4 = e* = -0.67 (Eq. IX-67) ( ) ( ) , 1 Nuoo = Re x NuFo r* = 0.5, 0 = e* = -0.53 (Eq. IX-68) Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-36 October 2004 ( ) ( ) ( ), 53 . 0 35 . 0 1 Nu * oo + + = e Re x NuFo r* = 0.5, -0.53 = e* = -0.67 (Eq. IX-69) The following is the linear form for the relative error, ( ) ( ) Re x NuFo oo Nu 1- , 0.5 = r* = 0.2, that matches the above equations at r* = 0.255 and r* = 0.5: ( ) ( ) , 0 Nu 1 oo = - Re x NuFo 0.5 = r* = 0.2, 0 = e* = 49 97 . 12 26 * + - r (Eq. IX-70) ( ) ( ) , 49 97 . 12 26 225 . 1 07125 . 0 Nu 1 * * * oo . .. . . .. . + + - - = - r e r Re x NuFo 0.5 = r* = 0.2, 49 97 . 12 26 * + - r = e* = -0.67 (Eq. IX-71) That is, the error in ( ) x NuFi is positive, with a formula that depends on the value of r*. If 2 . 0 3 . 0 * = = r and 67 . 0 e 0 * - = = , Equation IX-63 shows that there is a systematic relative error in ( ) x NuFi of about ( ) 7 . 0 15 . 0 e * * - - r . For 3. 0 5 . 0 * = = r , the error in ( ) x NuFi is not significant if * * 35 . 1 405 . 0 e 0 r - = = ; otherwise, there is a systematic relative error in ( ) x NuFi of about (Equation IX-65) ( ) 7 . 0 15 . 0 35 . 1 405 . 0 e * * * - - + - + r r . The error in ( ) x NuFo is negligible if 49 97 . 12 26 e 0 * * + - = = r . Otherwise, Equation IX-71 shows that there is a systematic relative error in ( ) x NuFo of about . .. . . .. . + - - - + 49 97 . 12 26 e 225 . 1 07125 . 0 * * * r r . IX.3.4.7 Uncertainty in Measured F Nu from Unequal Cylinder Lengths As in natural convection, any additional area in the outer cylinder may permit more convective heat transfer from the air to the outer cylinder. Also, where the inner cylinder is not present, the orifice area increases from p (Do – Di)2/4 to p Do 2/4, by a factor of 1/(1-r*)2. Because the mass flow rate must be the same and density does not change significantly, air velocity must drop by a factor of (1-r*)2. If the additional length of the cylinder is sufficiently small, the contribution to uncertainty from the length difference may be neglected. Alternatively, if r* is sufficiently small, the effect of the greater length of the outer cylinder may be accounted for by applying the predicted Nusselt number to the additional area. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-37 October 2004 IX.3.4.8 Uncertainty in Measured F Nu from Deviations from Fully Developed Flow Kays and Leung also considered thermally developing annular flow. They presented non-dimensional fluid temperatures, including parameters labeled .ii and .oo, for thermally developing annular flow with r*=0.255 (Kays and Leung 1963 [DIRS 160763], p. 542, Figure 2). The parameters Nuii and Nuoo are approximately the inverses of .ii and .oo, respectively. The Nusselt numbers start out at about twice their asymptotic value but decay to within 10% of their asymptote in a distance of about ten hydraulic diameters. For x = 11 (Do-Di), this uncertainty analysis assigns systematic errors in predictions of ( ) x NuFi and ( ) x NuFo , based on a linear fit to the errors at x = 0 and x = 10 (Do-Di), that amount to ( ) % 100 11 1 .. . .. . - - + i o D D x IX.3.5 Uncertainty in Nusselt Numbers from the Mixed Convection Methodology The uncertainty in the mixed-convection methodology is affected by the uncertainties in natural convection and forced convection in accordance with the Law of Propagation of Uncertainty (ANSI/NCSL Z540-2-1997 [DIRS 157394], p. 19). Because the preliminary steps are not part of the methodology, this section considers only the uncertainties in Steps N1 through N4. In addition to the uncertainty contributed by the underlying convection correlations, this uncertainty analysis considers two sources of uncertainty in the mixed convection methodology. One is the error from using an approximation to the forced convection correlation (Equations IX-28 and IX-29) to find the equivalent Reynolds number for natural convection. The second source of uncertainty is the variation of measured mixed convection results from the Morgan approximation. As noted at the beginning of Section IX.3, uncertainties in the input dimensionless groups must be evaluated by using the sensitivity analysis of Section IX.2. In addition, the uncertainty inherent in the methodology depends on the input parameters. Therefore, the prediction uncertainty is not quantified in this section. Section IX.4 provides examples of the evaluation of uncertainty in specific applications. Step N1 uses interpolation in r* to create a table of forced-convection parameters that are functions of Re only. As discussed in Section IX.3.4, the standard relative uncertainty is )5 . 0 )( 2 . 0 ( 5 . 2 * * r r - - Step N2 produces the equivalent Reynolds numbers for natural convection, ( ) x ReNi and ( ) x ReNo . The uncertainty in each of these equivalent Reynolds numbers is a combined relative uncertainty, composed of the following contributions: 1. Relative uncertainty in the appropriate Nusselt number for natural convection, calculated in accordance with Section IX.3.3. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-38 October 2004 2. Relative uncertainty in the values for ii Nu and oo Nu produced by Step N1, calculated as described above. 3. Relative uncertainty in the reverse linear interpolation to get ( ) x ReNi and ( ) x ReNo , calculated in accordance with the discussion of forward interpolation in Section IX.3.4. 4. Relative uncertainty introduced by the approximations represented by Equations IX-28 and IX-29. The errors from using Equation IX-28 for ( ) x ReNi and Equation IX-29 for ( ) x ReNo , instead of Equations IX-24 and IX-25, depend on the value of t(x) and must therefore be evaluated separately for each application. In Step N3, the equivalent Reynolds number for inner surface natural convection combines with the Reynolds number for forced convection to produce an equivalent Reynolds numbers for mixed convection. The uncertainty in ( ) x ReNi propagates through Equation IX-23 (specialized to the inner surface by adding the subscript i). Taking the partial derivative of that equation with respect to ( ) x ReNi and multiplying by ( ) ( ) x Re x Re Mi Ni / yields: ( ) ( ) ( ) ( ) ( ) ( ) 2 . .. . . .. . = . . x Re x Re x Re x Re x Re x Re Mi Ni Ni Mi Mi Ni (Eq. IX-72) This factor, applied to the relative uncertainty in ( ) x ReNi , produces its contribution to the combined relative uncertainty in ( ) x Re Mi (see Equation IX-31). Next, Step N3 produces the Nusselt number for the inner surface. There are three contributors to the uncertainty in the Nusselt number: 1. The uncertainty in ( ) x Re Mi , propagated according to the Law of Propagation of Uncertainty and making use of the sensitivity study (Section IX.2) 2. The uncertainty in the forced convection methodology when the input Reynolds number is known, calculated in accordance with Section IX.3.4 3. The uncertainty in mixed-convection Nusselt numbers inherent in the Morgan approximation. Morgan (1975 [DIRS 160791], p. 249, Figure 10) compared the experimental values from two data sets to the predicted ratio of effective Nusselt number to forced-flow Nusselt number. The experimental value for the ratio was consistently within 15% of the theoretical value. Taking 15% as the 95% confidence limit of an a priori normal distribution for a Type B evaluation of the uncertainty, this appendix assigns a standard uncertainty of 7.5% as the relative error inherent in the Morgan approximation for mixed convection. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-39 October 2004 Step N4 is the same as Step N3, except that it applies to the outer surface. The uncertainty considerations are the same as those for Step N3. IX.4 COMPARISON OF METHODOLOGY RESULTS TO TEST DATA This section evaluates the methodology under YMP specific conditions by corroboration of methodology results with data acquired from the EBS Ventilation Test series. Uncertainties in both the measurements and the predictions are considered. The calculated uncertainties in the previous section determine the accuracy of the predictions for the EBS forced ventilation test configuration, taking into consideration that the cylinders in the EBS model were of different lengths, were not held to either constant temperature or constant flux conditions, and were not concentric. However, the invert and waste package support systems make the EBS test geometry more complex than that for which the uncertainty was evaluated. To determine how appropriate the methodology is for the EBS configuration, it was applied to the EBS ventilation tests. A description of how this was done and the overall results are given below. IX.4.1 Test Data of Ventilation Test Phase 1 and 2 The Ventilation Test Phase 1 report, Testing to Provide Data for Ventilation System Design: Phase 1 (BSC 2003 [DIRS 160724], Sections 2.2.4 and 3) presents 24-hour averages of measurements taken at the rate of four per hour. For the Phase 1 tests, the time period chosen was the last full day of data in cases where quasi steady-state conditions were achieved, or the last 24 hours of data collected in cases where recorded temperatures were still increasing with time. For the Phase 2 tests, the averaging period was chosen as the last 24-hour period of the test where the design test conditions were maintained. Appendix XI describes the averaging process, starting from the raw data in DTN: SN0208F3409100.009 [DIRS 163079]. A brief description of the phase 1 and phase 2 ventilation tests is presented in Section 7.1.2 of this report. A volume flow rate for each test within each phase was calculated using measured differential pressure, relative humidity, barometric pressure, and air temperatures at both the inlet (designated Station A) and the outlet (designated Station D) (BSC 2003 [DIRS 160724], Section 5.2). The 24-hour average flow rates for each of the forced ventilation tests were within 10% of the nominally desired values, as shown in Tables IX-6 and IX-7. No flow rate measurements were reported at Station D for Tests 15 or 16 of Phase 2. Tables IX-6 and IX-7 also show the 24-hour average line load for each test, which is the total power input divided by the total heated length of the test train, 33.9 m (BSC 2003 [DIRS 160724], Section 2.2.2.2). The standard uncertainty in the 24-hour-average total load was 5.8 W (BSC 2003 [DIRS 160724], Section 3.3.2.4), which is equivalent to a standard uncertainty of 0.2 W/m (5.8W÷33.9m˜0.2W/m) in the average line load, much less than 1% (0.2W/m÷179W/m˜0.1%) of the measured average. The test reports also tabulate average temperatures for 24-hour periods. Tables IX-8 through IX-10 present the calculated average temperatures at a point midway along the heated portion of the test train (Station 3). Values in the tables are reported by quadrant (top, right, bottom, and left) for sensors located on the external surface of the waste package, the internal and external Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-40 October 2004 surfaces of the concrete pipe, the external surface of the insulation, and within the annulus between the waste package and concrete pipe (ventilation air). Table IX-6. Averaged Flow Rates and Line Loads and Their Standard Uncertainties for EBS Ventilation Tests, Phase 1 Test No. Nominal Flow Rate (m3/s) Station A, Flow Rate (m3/s) Station D, Flow Rate (m3/s) Flow Rate Uncertainty (m3/sec) Nominal Line Load (W/m) Avg. Line Load (W/m) 1 1 0.997 1.001 0.014 180 182 2 0.5 0.501 0.495 0.03 180 179 3 1 0.998 1.016 0.014 360 359 4 2 1.990 1.990 0.008 360 362 5 0.5 0.519 0.525 0.03 360 362 6 3 3.048 3.052 0.02 360 364 NC1 — — — — 120 120 NC2 — — — — 240 242 Source: BSC 2003 [DIRS 160724], Tables 3-16 and 5-6. NC1=Natural Convection Test 1; NC2=Natural Convection Test 2. Table IX-7. Averaged Flow Rates and Line Loads and Their Standard Uncertainties for EBS Ventilation Tests, Phase 2 Test No. Nominal Flow Rate (m3/s) Station A, Flow Rate (m3/s) Station D, Flow Rate (m3/s) Flow Rate Uncertainty (m3/sec) Nominal Line Load (W/m) Avg. Line Load (W/m) 1 1 1.021 0.972 0.014 220 218 2 1 1.037 0.986 0.014 220 218 3 1 1.058 1.012 0.014 220 216 4 1 1.054 1.003 0.014 220 215 5 1 1.024 0.989 0.014 360 360 6 1 1.041 1.005 0.014 360 359 7 1 1.053 1.011 0.014 360 357 8 1 1.055 1.013 0.014 360 358 9 0.5 0.516 0.506 0.03 220 215 10 0.5 0.552 0.544 0.03 220 215 11 0.5 0.554 0.530 0.03 220 216 12 0.5 0.547 0.553 0.03 360 360 13 0.5 0.553 0.550 0.03 360 360 14 0.5 0.553 0.537 0.03 360 361 15 1 0.993 N/A 0.014 360 360 16 1 0.991 N/A 0.014 360 364 Source: Appendix XI of this document. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-41 October 2004 Table IX-8. Averaged Temperature Values (C) at Station 3 for EBS Ventilation Tests, Phase 1 Ventilating Air (°C) WP Surface (°C) Concrete Pipe Wall (°C) Concrete/ Insulation Interface (°C) Outside Insulation Surface (°C) top — 47.4 30.5 30.7 30.3 right 27.8 42.0 29.8 30.0 27.5 bottom — 39.1 28.4 28.2 26.7 Test 1 left 27.4 40.8 30.2 30.1 27.9 top — 51.6 33.6 33.1 28.8 right 31.4 45.9 32.8 32.7 28.4 bottom — 42.9 30.5 30.1 27.1 Test 2 left 31.4 44.7 33.4 32.8 27.6 top — 63.4 33.8 33.5 27.8 right 29.3 54.6 32.7 33.0 26.5 bottom — 50.3 30.6 30.0 25.1 Test 3 left 29.2 53.3 33.5 33.0 25.9 top — 57.4 29.9 30.0 27.0 right 27.2 49.4 30.0 30.2 26.0 bottom — 45.3 28.8 28.3 25.2 Test 4 left 26.8 48.0 29.9 29.9 25.7 top 65.6 34.6 33.7 24.5 right 31.0 56.7 33.5 33.3 23.6 bottom — 52.0 29.2 28.4 22.6 Test 5 left 31.0 55.0 34.6 33.6 23.6 top — 48.0 23.4 23.5 21.2 right 21.7 40.6 23.5 23.7 20.6 bottom — 36.8 23.3 23.0 20.0 Test 6 left 21.7 39.8 23.7 23.7 20.1 top — 60.9 46.3 43.8 28.2 right 48.7 56.9 45.7 43.6 28.3 bottom — 55.0 38.0 36.7 25.5 Test NC1 left 48.4 56.3 45.5 43.3 27.4 top — 93.2 69.4 64.4 33.0 right 73.4 87.4 68.6 64.2 32.8 bottom — 84.9 56.1 51.6 29.0 Test NC2 left 72.9 86.3 68.3 63.8 31.8 Source: BSC 2003 [DIRS 160724], Tables 5-7 through 5-14. WP = waste package; NC1=Natural Convection Test 1; NC2=Natural Convection Test 2. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-42 October 2004 Table IX-9. Averaged Temperature Values (C) at Station 3 for EBS Ventilation Tests, Phase 2, Tests 1 through 8 Ventilating Air (°C) WP Surface (°C) Concrete Pipe Wall (°C) Concrete/ Insulation Interface (°C) Outside Insulation Surface (°C) top — 50.5 30.8 30.8 28.5 right 27.5 44.3 30.2 30.6 28.5 bottom — 41.1 29.4 30.7 28.0 Test 1 left 27.3 43.5 30.6 30.5 28.3 top — 59.3 38.8 37.9 30.0 right 36.9 52.8 38.5 37.8 29.2 bottom — 49.9 35.3 35.8 28.4 Test 2 left 36.7 51.9 38.6 37.6 28.4 top — 68.4 48.4 47.1 36.6 right 47.0 62.2 48.2 47.2 36.6 bottom — 59.6 44.2 44.7 35.1 Test 3 left 46.7 61.3 48.2 46.9 35.8 top — 67.9 48.0 46.6 35.9 right 46.7 61.7 47.8 46.8 36.3 bottom — 59.1 43.8 44.2 34.4 Test 4 left 46.5 60.9 47.9 46.6 36.0 top — 64.1 34.4 34.4 31.1 right 29.8 55.2 33.4 33.8 30.1 bottom — 50.8 31.8 33.0 28.7 Test 5 left 29.7 54.0 34.1 34.0 30.4 top — 73.6 43.8 43.4 37.2 right 39.6 64.4 42.9 43.0 36.3 bottom — 60.4 39.8 40.9 34.4 Test 6 left 39.4 63.1 43.6 43.1 36.5 top — 81.4 51.8 50.3 37.8 right 49.1 72.7 51.3 50.4 37.9 bottom — 69.0 46.4 47.0 35.4 Test 7 left 48.7 71.4 51.8 50.4 38.2 top — 81.4 51.6 50.1 37.2 right 49.0 72.7 51.2 50.2 37.3 bottom — 69.0 46.4 46.9 34.8 Test 8 left 48.7 71.4 51.6 50.1 37.1 Source: Appendix XI of this document. WP = waste package. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-43 October 2004 Table IX-10. Averaged Temperature Values (C) at Station 3 for EBS Ventilation Tests, Phase 2, Tests 9 through 16 Ventilating Air (°C) WP Surface (°C) Concrete Pipe Wall (°C) Concrete/ Insulation Interface (°C) Outside Insulation Surface (°C) top — 55.2 35.3 35.3 33.0 right 31.6 48.5 34.5 34.8 32.3 bottom — 45.1 33.1 34.8 31.0 Test 9 left 31.6 47.6 35.0 35.0 32.7 top — 63.5 43.0 42.4 36.2 right 40.4 56.5 42.5 42.2 36.0 bottom — 53.5 39.5 40.9 34.2 Test 10 left 40.3 55.7 42.9 42.2 36.2 top — 71.7 51.0 49.7 39.6 right 49.7 65.1 50.6 49.7 39.4 bottom — 62.3 46.1 47.1 36.9 Test 11 left 49.6 64.3 50.9 49.7 39.9 top — 71.2 41.0 40.9 35.6 right 35.9 61.5 40.0 40.4 35.2 bottom — 56.9 38.0 39.6 33.8 Test 12 left 35.9 60.1 40.8 40.5 35.1 top — 78.0 47.3 46.3 34.8 right 44.0 68.4 46.9 46.2 35.0 bottom — 64.4 42.3 43.1 33.1 Test 13 left 43.9 67.2 47.3 46.1 34.5 top — 86.1 55.6 53.7 38.4 right 53.4 77.0 55.2 53.8 38.2 bottom — 73.2 48.3 48.9 35.1 Test 14 left 53.3 75.7 55.7 53.9 38.8 top — 68.6 38.4 37.8 29.4 right 34.6 59.7 37.7 37.7 30.4 bottom — 55.4 35.0 36.4 28.8 Test 15 left 34.4 58.4 38.3 37.5 28.7 top — 68.6 37.9 37.2 27.9 right 34.3 59.5 37.2 37.0 28.2 bottom — 55.2 34.8 36.1 27.4 Test 16 left 34.1 58.2 37.8 36.9 27.1 Source: Appendix XI of this document. WP = waste package. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-44 October 2004 IX.4.2 Prediction of Nusselt numbers This section describes the prediction of Nusselt numbers for the EBS Ventilation Tests, in accordance with the algorithm of Section IX.1.6. Step P1. (Geometry) For Di = 0.4064 m and Do = 1.37 m (Tables 7-4 and 7-5), r* = 0.297. The predictions ignore the effect of the invert, so that they use a hydraulic diameter of 0.96 m. The effect of the invert would be to reduce the hydraulic diameter to about 0.93 m. Step P2. (Reynolds Number) To minimize the influence of end effects, the axial position of interest is the most centrally located measurement station (Station 3). The cross-sectional area is 1.34 m2. For each test, the mean fluid temperature at Station 3, Tf, is the average of the two reported measurements (Tables IX-8 through IX-10). Table 4-17 contains the properties of air at temperatures relevant to the EBS Ventilation Test Series. For each test, Table IX-11 or IX-12 shows the value of Tf, the kinematic viscosity, ., of air at Tf, linearly interpolated in Table 4-17, the average of the two reported flow rates (Tables IX-6 and IX-7), the mean axial flow velocity calculated by dividing the flow rate by the annulus cross-sectional area, and the value of Re calculated using Equation IX-1. Step P3. (Rayleigh Numbers) The value for g is 9.8 m/s2 (Table 4-20). For each test (at Station 3), Table IX-13 or IX-14 gives the circumferential average temperature on each surface, based on the measurements in Tables IX-8 through Table IX-10. On the inner surface (the waste package), i T is the average of the four reported measurements. However, because the bottom of the outer surface is covered by the invert, the circumferential average temperature on the outer surface, o T , is the average of only the top and side measurements. Tables IX-13 and IX-14 also show the amount that each average differs from its associated Tf , as well as the Rayleigh numbers calculated from Equations IX-5 and IX-6 and the relevant air properties. Step N1. (Forced-Convection Parameters) Table IX-15 contains the forced-convection parameters as a function of Re, for Pr=0.7 and r*=0.297, linearly interpolated from Table IX-2. Step N2. (Natural Convection) For each test (at Station 3), Table IX-16 or IX-17 gives the effective circumferential Nusselt numbers for natural convection, on the inner and outer surfaces. The tables also show the equivalent Reynolds numbers, calculated in accordance with Equations IX-28 and IX-29. Step N3. (Inner-Surface Nusselt Number) Tables IX-18 and IX-19 report the mixed-convection Reynolds number at the inner surface for each test, calculated in accordance with Equation IX-26. These tables also report the forcedconvection parameters associated with each such Reynolds number, from interpolation in Table IX-11. The last column contains the effective Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-45 October 2004 circumferential Nusselt number convection at the inner surface, from the mixedconvection methodology, Equation IX-24. Step N4. (Outer-Surface Nusselt Number) Tables IX-20 and IX-21 report the mixed-convection Reynolds number at the outer surface for each test, calculated in accordance with Equation IX-27. These tables also report the forced-convection parameters associated with each such Reynolds number, from interpolation in Table IX-11. The last column contains the effective circumferential Nusselt number for mixed convection at the outer surface, from the mixed-convection methodology, Equation IX-25. Table IX-11. Reynolds Numbers for EBS Ventilation Test Series, Phase 1 Location Tf (ºC) Tf (K) . (m2/s) V (m3/s) um (m/s) Re (Thousands) Test 1, Station 3 27.60 300.75 1.597E-05 0.999 0.74 44.8 Test 2, Station 3 31.40 304.55 1.635E-05 0.498 0.37 21.8 Test 3, Station 3 29.25 302.40 1.613E-05 1.007 0.75 44.7 Test 4, Station 3 27.00 300.15 1.591E-05 1.990 1.48 89.7 Test 5, Station 3 31.00 304.15 1.631E-05 0.522 0.39 22.9 Test 6, Station 3 21.70 294.85 1.543E-05 3.050 2.27 141.7 Output DTN: MO0306MWDMXCNV.000, worksheet “Dimensionless Inputs” of Phase 1 Supporting Calculations for Mixed Convection.xls. Table IX-12. Reynolds Numbers for EBS Ventilation Test Series, Phase 2 Location Tf (ºC) Tf (K) . (m2/s) V (m3/s) um (m/s) Re (Thousands) Test 1, Station 3 27.40 300.55 1.595E-05 0.996 0.74 44.8 Test 2, Station 3 36.80 309.95 1.689E-05 1.012 0.75 42.9 Test 3, Station 3 46.85 320.00 1.790E-05 1.035 0.77 41.4 Test 4, Station 3 46.60 319.75 1.788E-05 1.028 0.76 41.2 Test 5, Station 3 29.75 302.90 1.618E-05 1.006 0.75 44.6 Test 6, Station 3 39.50 312.65 1.716E-05 1.023 0.76 42.7 Test 7, Station 3 48.90 322.05 1.811E-05 1.032 0.77 40.8 Test 8, Station 3 48.85 322.00 1.810E-05 1.034 0.77 40.9 Test 9, Station 3 31.60 304.75 1.637E-05 0.511 0.38 22.4 Test 10, Station 3 40.35 313.50 1.725E-05 0.548 0.41 22.8 Test 11, Station 3 49.65 322.80 1.818E-05 0.542 0.40 21.4 Test 12, Station 3 36.15 309.30 1.683E-05 0.550 0.41 23.4 Test 13, Station 3 43.95 317.10 1.761E-05 0.552 0.41 22.5 Test 14, Station 3 53.35 326.50 1.856E-05 0.545 0.41 21.1 Test 15, Station 3 34.50 307.65 1.666E-05 0.993 0.74 42.7 Test 16, Station 3 34.20 307.35 1.663E-05 0.991 0.74 42.7 Output DTN: MO0306MWDMXCNV.000, worksheet “Dimensionless Inputs” of Phase 2 Supporting Calculations for Mixed Convection.xls. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-46 October 2004 Table IX-13. Rayleigh Numbers for EBS Ventilation Test Series, Phase 1 Location Avg Ti (ºC) Avg To (ºC) (Ti-Tf) (ºC) (To-Tf) (ºC) t a (m2/s) Rai (Millions) Rao (Millions) Test 1, Station 3 42.33 30.17 14.73 2.567 5.74 2.26E-05 89 596 Test 2, Station 3 46.28 33.27 14.88 1.867 7.97 2.32E-05 85 408 Test 3, Station 3 55.40 33.33 26.15 4.083 6.40 2.29E-05 154 923 Test 4, Station 3 50.03 29.93 23.03 2.933 7.85 2.25E-05 141 687 Test 5, Station 3 57.33 34.23 26.33 3.233 8.14 2.31E-05 151 711 Test 6, Station 3 41.30 23.53 19.60 1.833 10.69 2.18E-05 130 465 Output DTN: MO0306MWDMXCNV.000, worksheet “Dimensionless Inputs” of Phase 1 Supporting Calculations for Mixed Convection.xls. Table IX-14. Rayleigh Numbers for EBS Ventilation Test Series, Phase 2 Location Avg Ti (ºC) Avg To (ºC) (Ti-Tf) (ºC) (To-Tf) (ºC) t a (m2/s) Rai (Millions) Rao (Millions) Test 1, Station 3 44.85 30.53 17.45 3.133 5.57 2.26E-05 106 730 Test 2, Station 3 53.48 38.63 16.68 1.833 9.10 2.40E-05 87 368 Test 3, Station 3 62.88 48.27 16.03 1.417 11.31 2.55E-05 72 245 Test 4, Station 3 62.40 47.90 15.80 1.300 12.15 2.54E-05 72 225 Test 5, Station 3 56.03 33.97 26.28 4.217 6.23 2.29E-05 154 945 Test 6, Station 3 65.38 43.43 25.88 3.933 6.58 2.44E-05 130 758 Test 7, Station 3 73.63 51.63 24.73 2.733 9.05 2.58E-05 108 458 Test 8, Station 3 73.63 51.47 24.78 2.617 9.47 2.58E-05 109 439 Test 9, Station 3 49.10 34.93 17.50 3.333 5.25 2.32E-05 99 726 Test 10, Station 3 57.30 42.80 16.95 2.450 6.92 2.45E-05 84 466 Test 11, Station 3 65.85 50.83 16.20 1.183 13.69 2.59E-05 70 196 Test 12, Station 3 62.43 40.60 26.28 4.450 5.90 2.39E-05 139 902 Test 13, Station 3 69.50 47.17 25.55 3.217 7.94 2.50E-05 120 580 Test 14, Station 3 78.00 55.50 24.65 2.150 11.47 2.64E-05 101 338 Test 15, Station 3 60.53 38.13 26.03 3.633 7.16 2.36E-05 141 756 Test 16, Station 3 60.38 37.63 26.18 3.433 7.62 2.36E-05 143 718 Output DTN: MO0306MWDMXCNV.000, worksheet “Dimensionless Inputs” of Phase 2 Supporting Calculations for Mixed Convection.xls. Table IX-15. Parameters for Annular Forced Convection at Pr = 0.7 and r* = 0.297 Re: 10000 30000 100000 300000 1000000 Nuii: 36.1 75.4 186.3 449.5 1208.8 .i: 0.376 0.312 0.266 0.243 0.219 Nuoo: 29.0 63.6 162.7 391.5 1060.3 .o: 0.087 0.076 0.068 0.061 0.056 Output DTN: MO0306MWDMXCNV.000, worksheet “Predicted Nusselt Numbers” of Phase 1 Supporting Calculations for Mixed Convection.xls. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-47 October 2004 Table IX-16. Natural Convection Nusselt Numbers for EBS Ventilation Test Series, Phase 1 Location NuNi NuNo ReNi (Thousands) ReNo (Thousands) Test 1, Station 3 131 109 65 62 Test 2, Station 3 129 99 64 55 Test 3, Station 3 156 122 81 71 Test 4, Station 3 152 113 78 65 Test 5, Station 3 155 114 80 66 Test 6, Station 3 148 103 76 58 Output DTN: MO0306MWDMXCNV.000, worksheet “Predicted Nusselt Numbers” of Phase 1 Supporting Calculations for Mixed Convection.xls. Table IX-17. Natural Convection Nusselt Numbers for EBS Ventilation Test Series, Phase 2 Location NuNi NuNo ReNi (Thousands) ReNo (Thousands) Test 1, Station 3 139 115 70 66 Test 2, Station 3 130 97 65 53 Test 3, Station 3 123 87 60 47 Test 4, Station 3 123 85 60 45 Test 5, Station 3 156 123 81 72 Test 6, Station 3 148 116 76 67 Test 7, Station 3 139 102 70 57 Test 8, Station 3 140 101 71 57 Test 9, Station 3 136 115 68 66 Test 10, Station 3 129 103 64 58 Test 11, Station 3 122 83 59 43 Test 12, Station 3 151 121 78 71 Test 13, Station 3 144 108 73 62 Test 14, Station 3 137 95 69 52 Test 15, Station 3 152 116 78 67 Test 16, Station 3 152 114 79 66 Output DTN: MO0306MWDMXCNV.000, worksheet “Predicted Nusselt Numbers” of Phase 2 Supporting Calculations for Mixed Convection.xls. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-48 October 2004 Table IX-18. Predicted Inner-Surface Nusselt Numbers for EBS Ventilation Test Series, Phase 1 Location ReMi (Thousands) Nuii(ReMi) Nuoo(ReMi) .i(ReMi) .o(ReMi) NuMi Test 1, Station 3 79 153 133 0.280 0.070 163 Test 2, Station 3 68 135 117 0.288 0.071 142 Test 3, Station 3 92 174 152 0.271 0.069 184 Test 4, Station 3 119 211 184 0.264 0.067 221 Test 5, Station 3 83 160 139 0.277 0.070 168 Test 6, Station 3 161 266 232 0.259 0.066 276 Output DTN: MO0306MWDMXCNV.000, worksheet “Predicted Nusselt Numbers” of Phase 1 Supporting Calculations for Mixed Convection.xls. Table IX-19. Predicted Inner-Surface Nusselt Numbers for EBS Ventilation Test Series, Phase 2 Location ReMi (thousands) Nuii(ReMi) Nuoo(ReMi) .i(ReMi) .o(ReMi) NuMi Test 1, Station 3 83 159 139 0.278 0.070 170 Test 2, Station 3 78 151 131 0.281 0.070 158 Test 3, Station 3 73 143 124 0.284 0.071 150 Test 4, Station 3 73 143 124 0.284 0.071 149 Test 5, Station 3 92 174 152 0.271 0.069 184 Test 6, Station 3 87 166 144 0.275 0.069 175 Test 7, Station 3 81 157 136 0.279 0.070 164 Test 8, Station 3 82 157 137 0.278 0.070 164 Test 9, Station 3 72 141 123 0.285 0.071 151 Test 10, Station 3 68 135 117 0.288 0.071 143 Test 11, Station 3 63 128 110 0.291 0.072 133 Test 12, Station 3 81 156 136 0.279 0.070 166 Test 13, Station 3 77 149 130 0.282 0.070 157 Test 14, Station 3 72 142 123 0.285 0.071 148 Test 15, Station 3 89 169 147 0.273 0.069 178 Test 16, Station 3 89 170 148 0.273 0.069 178 Output DTN: MO0306MWDMXCNV.000, worksheet “Predicted Nusselt Numbers” of Phase 2 Supporting Calculations for Mixed Convection.xls. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-49 October 2004 Table IX-20. Predicted Outer-Surface Nusselt Numbers for EBS Ventilation Test Series, Phase 1 Location ReMo (thousands) Nuii(ReMo) Nuoo(ReMo) .i(ReMo) .o(ReMo) NuMo Test 1, Station 3 77 149 130 0.282 0.070 194 Test 2, Station 3 59 122 105 0.293 0.072 179 Test 3, Station 3 84 161 140 0.277 0.069 216 Test 4, Station 3 111 201 175 0.265 0.067 286 Test 5, Station 3 70 138 120 0.286 0.071 204 Test 6, Station 3 153 256 223 0.260 0.066 411 Output DTN: MO0306MWDMXCNV.000, worksheet “Predicted Nusselt Numbers” of Phase 1 Supporting Calculations for Mixed Convection.xls. Table IX-21. Predicted Outer-Surface Nusselt Numbers for EBS Ventilation Test Series, Phase 2 Location ReMo (thousands) Nuii(ReMo) Nuoo(ReMo) .i(ReMo) .o(ReMo) NuMo Test 1, Station 3 80 155 134 0.279 0.070 198 Test 2, Station 3 69 136 118 0.287 0.071 211 Test 3, Station 3 62 127 110 0.291 0.072 217 Test 4, Station 3 61 125 108 0.292 0.072 222 Test 5, Station 3 84 162 141 0.277 0.069 215 Test 6, Station 3 79 154 134 0.280 0.070 209 Test 7, Station 3 70 139 121 0.286 0.071 215 Test 8, Station 3 70 138 120 0.286 0.071 218 Test 9, Station 3 70 138 120 0.286 0.071 175 Test 10, Station 3 62 126 109 0.291 0.072 175 Test 11, Station 3 48 104 90 0.300 0.074 199 Test 12, Station 3 74 146 127 0.283 0.071 191 Test 13, Station 3 66 132 114 0.289 0.072 193 Test 14, Station 3 56 117 101 0.295 0.073 202 Test 15, Station 3 79 154 134 0.280 0.070 215 Test 16, Station 3 79 152 132 0.280 0.070 218 Output DTN: MO0306MWDMXCNV.000, worksheet “Predicted Nusselt Numbers” of Phase 2 Supporting Calculations for Mixed Convection.xls. IX.4.3 Uncertainty in Predicted Nusselt Numbers This section evaluates the sources of uncertainty that have a quantitative dependence on the configuration or environment. These are the sources for which Section IX.3 does not provide a numerical uncertainty. The sources of uncertainty, both those evaluated here and those evaluated in Section IX.3, become inputs to the combined uncertainty. IX.4.3.1 Uncertainty in the Predicted Nusselt Numbers for Natural Convection For natural convection, Section IX.3.3 evaluates the deviation of measured Nusselt numbers from the Kuehn-Goldstein correlation of those measurements. The Type B evaluation gives a Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-50 October 2004 random relative standard uncertainty of 12%. It also finds that the contribution from the effects of diameter ratio are negligible for the value in the EBS Ventilation Test Series, r* = 0.3. This section evaluates the following sources of uncertainty for the particular configuration and conditions of the EBS Ventilation Tests, based on the discussions in Section IX.3.3: 1. The uncertainty in predicted effective circumferential Nusselt numbers for coextensive, isothermal cylinders arising from eccentric location of the inner cylinder 2. The uncertainty in predicted effective circumferential Nusselt numbers for isothermal cylinders arising from unequal cylinder lengths 3. Uncertainty in the effective circumferential Nusselt numbers arising from temperature variation along the circumferences of the cylinders. As described in Section IX.3.3, the use of the concentric correlation for natural convection systematically underestimates the natural convection heat transfer by * 15 . 0 e - . For the EBS Ventilation Test Series configuration, with an e* of –0.42, this source of uncertainty causes a systematic error of about –6%. The use of data from Station 3 minimizes the effects of the extra length of wall beyond the ends of the waste package train. As suggested in Section IX.3.3, this analysis neglects the error caused by those extensions and by the gaps between the waste packages, because the following conditions hold: 1. The greater length of the outer cylinder does not cause a qualitative change in the flow from natural convection other than mild divergence and convergence along the axis. 2. The changes in the Rayleigh numbers appearing in the correlation, caused by the change in air temperature, account for most of the changes in the circumferential average Nusselt numbers. 3. The remaining effects of the longer outer cylinder are not significant compared to the other contributions to uncertainty. The following is an evaluation of the effects of the circumferential temperature variations in the EBS Ventilation Tests. From Section IX.3.3, the relative error in the inner-cylinder Nusselt number is about ] [ ] [ 25 . 0 f i i t T T T T - - + and the relative error in the outer-cylinder Nusselt number is about ] [ ] [ 4 . 0 o f o t T T T T - - + . Table IX-22 provides an evaluation of the errors for each test in the EBS Ventilation Test Series. The average is a Type A evaluation of the effects of the circumferential temperature variations. The predictions for ( ) x NuNi have a systematic error of +8% with a random standard uncertainty that is 0.5% of ( ) x NuNi . Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-51 October 2004 At the outer surface, the values are negative, because unlike the situation for pure natural convection, the ventilation tests have f o T T > . The negative error is reasonable because the magnitude of the temperature difference is largest at the top. The predictions have a systematic error of –3% in ( ) x NuNo with a random standard uncertainty of 2%. Table IX-22. Errors in ( ) x NuN for EBS Ventilation Test Series Location NUNI Error NUNO Error Phase 1 Test 1, Station 3 8.6% -5.2% Test 2, Station 3 8.9% -7.1% Test 3, Station 3 7.6% -4.6% Test 4, Station 3 8.0% 0.5% Test 5, Station 3 7.9% -4.5% Test 6, Station 3 8.5% 2.9% Phase 2 Test 1, Station 3 8.1% -3.4% Test 2, Station 3 8.7% -3.6% Test 3, Station 3 8.6% -3.8% Test 4, Station 3 8.7% -3.1% Test 5, Station 3 7.4% -3.4% Test 6, Station 3 7.9% -3.7% Test 7, Station 3 7.9% -2.4% Test 8, Station 3 7.8% -2.0% Test 9, Station 3 8.7% -4.4% Test 10, Station 3 9.1% -3.3% Test 11, Station 3 9.0% -5.6% Test 12, Station 3 8.3% -3.6% Test 13, Station 3 8.3% 3.3% Test 14, Station 3 8.2% -1.9% Test 15, Station 3 7.8% -2.9% Test 16, Station 3 7.9% -3.1% mean 8.3% -3.0% std dev 0.5% 2.46% Output DTN: MO0306MWDMXCNV.000, file: “Phase 1 Supporting Calculations for Mixed Convection.xls,” spreadsheet: “Circum.T Vartn,” col. F; and file: “Phase 2 Supporting Calculations for Mixed Convection.xls,” spreadsheet: “Circum.T Vartn,” col. F. In all of the EBS ventilation tests, the outer surface was hotter than the air (Tables IX-13 and IX-14). Therefore, the flow patterns were more like Figure IX-1d than Figure IX-1c. There may be an unknown error from applying the Kuehn-Goldstein correlation to the flow pattern of Figure IX-1d. Table IX-23 presents the contributions to N Nu uncertainty from other causes and their combined standard uncertainty. Systematic effects are shown as corrections, which have the opposite signs Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-52 October 2004 from the errors. The 95% confidence interval for ( ) x NuNi is from –26% to +22%. At the outer surface, the systematic effects are in the same direction, so that the 95% confidences limit is from –15% to +33%. Of the effects considered in Table IX-23, the dominant source of uncertainty is the deviation of measured Nusselt numbers reported by Kuehn and Goldstein (1976 [DIRS 100675]) from their correlation of those measurements. Table IX-23. Uncertainty Budget, Predicted N Nu Relative Standard Uncertainties from Random Effects Corrections for Systematic Effects Source of Uncertainty Type A Evaluation Type B Evaluation Type A Evaluation Type B Evaluation Correlation for concentric, coextensive, isothermal cylinders — 12% — — Eccentricity — — — +6% Circumferential temperature variation Ni Nu : 0.6% No Nu : 2% — — Ni Nu : -8% No Nu : +3% NOTES: Ni Nu : Correction for systematic effects: -2% Combined standard uncertainty from random effects: 12% 95% confidence interval: -26% to +22% No Nu : Correction for systematic effects: +9% Combined standard uncertainty from random effects: 12% 95% confidence interval: -15% to +33% IX.4.3.2 Uncertainty in the Predicted Nusselt Numbers for Forced Convection For forced convection, Section IX.3.4 finds that the contribution from the effects of flux variations along the lengths and the circumferences of the cylinder are negligible. This section evaluates the following sources of uncertainty for the particular configuration and conditions of the EBS Ventilation Tests, based on the discussions in Section IX.3.4: 1. The extent to which measured circumferential average Nusselt numbers for fully developed flow in concentric, coextensive, uniform-flux cylinders deviate from the circumferential average Nusselt numbers predicted by the correlation. 2. Uncertainty from linear interpolation of the Kays-Leung parameters to the diameter ratio of 0.3. 3. Uncertainty from linear interpolation of the Kays-Leung parameters to the appropriate Reynolds number. 4. The uncertainty in effective circumferential Nusselt numbers for fully developed flow in uniform-flux cylinders arising from unequal cylinder lengths. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-53 October 2004 5. The uncertainty in effective circumferential Nusselt numbers for fully developed flow in coextensive, uniform-flux cylinders arising from eccentric location of the inner cylinder. 6. Uncertainty arising from deviations from fully developed flow. For the uncertainty inherent in the Kays-Leung correlation, the Type B evaluation in Section IX.3.4 assigns a relative uncertainty of 3% as the random component. Because all of the ventilation tests had Re greater than 20,000, the systematic component is negligible. For 0.2 < r* < 0.5, one may take ) 5 . 0 )( 2 . 0 ( 5 * * r r - - as the upper 95% confidence limit in the relative error caused by interpolation of r* (Section IX.3.4). For interpolation to r*=0.3, the 95% confidence limit is 10%. This appendix uses an a priori normal distribution for a Type B evaluation of the uncertainty, with a standard uncertainty of 5%, for the relative error caused by interpolation in r*. The maximum error in an interpolation in Re is ( ) M Re 2 0 5 . 0 , where 0 Re is the value at the lower end of the interval and M is a second derivative (Section IX.3.2). The analysis in Section IX.3.2 indicates that for r* = 0.3, the relative error should be no more than 7%. Taking 7% as the 95% confidence limit of an a priori normal distribution for a Type B evaluation of the uncertainty, this appendix assigns a standard uncertainty of 3.5% as the relative error caused by interpolation in Re. Section IX.3.4 derived expressions for the relative systematic errors in the Nusselt numbers arising from the eccentricity of the configuration. For the configuration of the EBS Ventilation Test Series, r* = 0.3 and e* = -0.4. Applying these values to the expressions, the relative systematic errors for the inner and outer Nusselt numbers are: ( ) ( ) % 6 . 8 Nu 1 ii + = - Re x NuFi (Eq. IX-73) and ( ) ( ) 0 Nu 1 oo = - Re x NuFo (Eq. IX-74) As suggested in Section IX.3.4, this analysis neglects the error caused by the difference in total lengths of the cylinders, because the necessary conditions hold. That is, the additional length of the cylinder is sufficiently small. Section IX.3.4 provides an estimate of the systematic error arising from applying Nusselt numbers predicted for fully developed flow to regions of thermally developing flow. According to that estimate, the systematic error becomes negligible at a distance of 10 (Di - Do) into the flow. For the EBS Ventilation Test configuration, that distance is 10 m. Because Station 3 is about 20 m from the inlet (BSC 2003 [DIRS 160724], Section 2), this analysis neglects that source of uncertainty. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-54 October 2004 Table IX-24 presents the contributions to F Nu uncertainty from various causes and the combined standard uncertainty. Systematic effects are shown as corrections and therefore have opposite signs. There are three major random effects that are approximately equal in significance, causing the 95% confidence limit for Fo Nu to range from –14% to +14%. Fi Nu also has a systematic effect from eccentricity, so that its 95% confidence limit extends from -23% to +5%. Table IX-24. Uncertainty Budget, Predicted F Nu Relative Uncertainties from Random Effects Corrections for Systematic Effects Source of Uncertainty Type A Evaluation Type B Evaluation Type A Evaluation Type B Evaluation Correlation for fullydeveloped flow in concentric, coextensive, uniform-flux cylinders — 3% — — Linear interpolation in r* — 5% — — Linear interpolation in Re — 3.5% — — Eccentricity — — — Fi Nu : -9% NOTES: Correction for systematic effects, Fi Nu only: -9% Combined standard uncertainty from random effects: 7% Fi Nu 95% confidence interval: -23% to +5% Fo Nu 95% confidence interval: -14% to +14% IX.4.3.3 Uncertainty in the Predicted Nusselt Numbers for Mixed Convection This section evaluates the uncertainties in the predicted Nusselt numbers by propagating uncertainty through the calculations of Step N1 through Step N3 that were reported in Section IX.4.2. The uncertainty analysis reflects the discussion in Section IX.3.5. Uncertainties in measured temperatures and flow rates are neglected. Step N1 uses interpolation in r* to create Table IX-15, in which the forced-convection parameters are functions of Re only. As discussed above, the standard uncertainty in each interpolated parameter, such as ii Nu or oo Nu , is 5%. Step N2 begins with the calculation of the two natural convection Nusselt numbers, ( ) x NuNi and ( ) x NuNo , reported in Tables IX-16 and IX-17. As reported in Table IX-23, each is missing a correction for systematic effects and has a combined standard uncertainty of 12% from random effects. These combine with the 5% standard uncertainties in ii Nu and oo Nu to produce the total uncertainty in the Nusselt numbers that ( ) x ReNi and ( ) x ReNo are supposed to represent. That is, the total uncertainty before the reverse interpolation consists of a combined standard uncertainty of 13% from random effects, as well as a systematic error. This does not include the error arising from having the flow pattern of Figure IX-1d instead of the pattern of Figure IX-1c. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-55 October 2004 The factor for propagating relative uncertainty is ( ) ii ii Nu Re Re Nu . . or ( ) oo oo Nu Re Re Nu . . (see Equation IX-31). For the interval between Re of 30,000 and Re of 100,000 in Table IX-15, for example, the last two weighting factors are both about 1.3. Taking 1.3 as a representative propagation factor, the uncertainty in NuN contributes 17% to the random standard uncertainty in ReN from random effects. The corrections of -2% and +9% in the ( ) x NuNi and ( ) x NuNo become corrections of -3% and +12% in ( ) x ReNi and ( ) x ReNo , respectively. Another source of uncertainty in ( ) x ReNi and ( ) x ReNo is the interpolation in Reynolds number. The analysis in Section IX.3.2 shows that the error may range up to 12%. That result is the basis for a Type B evaluation of the uncertainty and an assignment of 6% as the standard relative uncertainty from the reverse interpolation. Together with the 17% of the previous paragraph, this yields 18% as the combined standard uncertainty from random effects (applying Equation IX-30). A final source of uncertainty in ( ) x ReNi and ( ) x ReNo is the use of the approximations represented by Equations IX-28 and IX-29. For purposes of this appendix, Tables IX-25 and IX-26 present the values of ( ) x ReNi and ( ) x ReNo that would have been obtained without the approximation. These “correct” values are the result of applying the bisection method (Conte and de Boor 1972 [DIRS 159800], p. 28, Algorithm 2.1) until the interval in Re was less than 50.0. Tables IX-25 and IX-26 also contain the percentage corrections that are implied for the values of ( ) x ReNi and ( ) x ReNo appearing in Tables IX-16 and IX-17, the random uncertainties, and the 95% confidence limits. For the confidence limits, the percentage of random uncertainty was applied after the correction. For Test 1 of Phase 1, for example, the correction is –15%, the random standard uncertainty is 14%, and the lower confidence limit of –39% is the value of the expression (100% - 28%) (100% - 15%) - 100%. The confidence limits do not include the error from the approximation that the effects of the Figure IX-1d flow pattern are negligible. Of the evaluated sources of uncertainty in ( ) x ReNi the dominant source is the uncertainty in ( ) x NuNi , which stems from the deviation of measured Nusselt numbers reported by Kuehn and Goldstein (1976 [DIRS 100675]) from their correlation of those measurements. The dominant contribution to the evaluated uncertainty in ( ) x ReNo is the approximation represented by Equation IX-29. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-56 October 2004 Table IX-25. Combined Uncertainty in ReNi and ReNo for EBS Ventilation Test Series, Phase 1 95% Confidence Interval Location ReN from Eq. IX-28 or IX-29 (Simplified) (Thousands) ReN from Eq. IX-24 or IX-25 (Not Simplified) (Thousands) To Correct for Simplified Equation Correction for Systematic Effect in NuN Combined Correction for Systematic Effects Standard Uncertainty from Random Effects Lower Upper ReNi Test 1, Station 3 65 60 -8% -3% -11% 18% -43% 21% Test 2, Station 3 64 60 -7% -3% -10% 18% -42% 22% Test 3, Station 3 81 76 -7% -3% -10% 18% -42% 23% Test 4, Station 3 78 74 -6% -3% -9% 18% -42% 24% Test 5, Station 3 80 76 -6% -3% -9% 18% -42% 24% Test 6, Station 3 76 72 -5% -3% -8% 18% -41% 25% ReNo Test 1, Station 3 62 35 -44% 12% -32% 18% -56% -7% Test 2, Station 3 55 25 -54% 12% -42% 18% -63% -21% Test 3, Station 3 71 39 -45% 12% -33% 18% -57% -9% Test 4, Station 3 65 31 -52% 12% -40% 18% -62% -19% Test 5, Station 3 66 31 -54% 12% -42% 18% -63% -21% Test 6, Station 3 58 22 -62% 12% -50% 18% -68% -32% Output DTN: MO0306MWDMXCNV.000, worksheet “Uncertainties” of Phase 1 Supporting Calculations for Mixed Convection.xls. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-57 October 2004 Table IX-26. Combined Uncertainty in ReNi and ReNo for EBS Ventilation Test Series, Phase 2 95% Confidence Interval Location ReN from Eq. IX-28 or IX-29 (Simplified) (Thousands) ReN from Eq. IX-24 or IX-25 (Not Simplified) (Thousands) To Correct for Simplified Equation Correction for Systematic Effect in NuN Combined Correction for Systematic Effects Standard Uncertainty from Random Effects Lower Upper ReNi Test 1, Station 3 70 65 -8% -3% -11% 18% -43% 21% Test 2, Station 3 65 61 -7% -3% -10% 18% -42% 23% Test 3, Station 3 60 57 -6% -3% -9% 18% -42% 24% Test 4, Station 3 60 57 -5% -3% -8% 18% -41% 26% Test 5, Station 3 81 75 -8% -3% -11% 18% -43% 21% Test 6, Station 3 76 71 -7% -3% -10% 18% -42% 23% Test 7, Station 3 70 66 -6% -3% -9% 18% -42% 23% Test 8, Station 3 71 67 -6% -3% -9% 18% -42% 24% Test 9, Station 3 68 63 -8% -3% -11% 18% -43% 21% Test 10, Station 3 64 60 -7% -3% -10% 18% -42% 23% Test 11, Station 3 59 57 -5% -3% -8% 18% -41% 25% Test 12, Station 3 78 72 -7% -3% -10% 18% -43% 22% Test 13, Station 3 73 69 -6% -3% -9% 18% -42% 24% Test 14, Station 3 69 65 -5% -3% -8% 18% -41% 25% Test 15, Station 3 78 74 -6% -3% -9% 18% -42% 24% Test 16, Station 3 79 74 -6% -3% -9% 18% -42% 23% ReNo Test 1, Station 3 66 39 -42% 12% -30% 18% -55% -5% Test 2, Station 3 53 23 -58% 12% -46% 18% -65% -26% Test 3, Station 3 47 16 -66% 12% -54% 18% -70% -37% Test 4, Station 3 45 15 -68% 12% -56% 18% -72% -40% Test 5, Station 3 72 41 -44% 12% -32% 18% -56% -7% Test 6, Station 3 67 36 -47% 12% -35% 18% -58% -12% Test 7, Station 3 57 24 -58% 12% -46% 18% -65% -26% Test 8, Station 3 57 23 -59% 12% -47% 18% -66% -28% Test 9, Station 3 66 40 -40% 12% -28% 18% -54% -2% Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-58 October 2004 Table IX-26. Combined Uncertainty in ReNi and ReNo for EBS Ventilation Test Series, Phase 2 (Continued) Location ReN from Eq. IX-28 or IX-29 (Simplified) (Thousands) ReN from Eq. IX-24 or IX-25 (Not Simplified) (Thousands) To Correct for Simplified Equation Correction for Systematic Effect in NuN Combined Correction for Systematic Effects Standard Uncertainty from Random Effects 95% Confidence Interval ReNo Test 10, Station 3 58 29 -50% 12% -38% 18% -60% -15% Test 11, Station 3 43 12 -72% 12% -60% 18% -74% -45% Test 12, Station 3 71 40 -43% 12% -31% 18% -56% -7% Test 13, Station 3 62 29 -54% 12% -42% 18% -63% -21% Test 14, Station 3 52 18 -65% 12% -53% 18% -70% -36% Test 15, Station 3 67 34 -49% 12% -37% 18% -60% -15% Test 16, Station 3 66 32 -51% 12% -39% 18% -61% -18% Output DTN: MO0306MWDMXCNV.000, worksheet “Uncertainties” of Phase 2 Supporting Calculations for Mixed Convection.xls. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-59 October 2004 Steps N3 and N4 begin with calculations of ( ) x ReMi and ( ) x ReMo , for which the only uncertainty is from the uncertainty in ( ) x ReNi and ( ) x ReNo . As explained in Section IX.3.5, the relative uncertainty in N Re propagates into M Re with a factor of [ ]2 M N Re Re . Tables IX-27 and IX-28 show that propagation for the EBS Ventilation Test Series, still omitting the error from the qualitatively different flow pattern. Finally, steps N3 and N4 calculate ( ) x NuMi and ( ) x NuMo . Tables IX-27 and IX-28 show the factor ( ) M M M M Re Nu Nu Re . . by which the relative uncertainty in M Re propagates into M Nu . For purposes of this appendix, the derivatives were estimated by taking a small increment in M Re and evaluating M Nu again. Table IX-27. Combined Uncertainty in ReMi and ReMo for EBS Ventilation Test Series, Phase 1 Location (ReN/ReM)2 Combined Correction for Systematic Effects Standard Uncertainty from Random Effects (ReM/NuM) Times Partial of NuM with Respect to ReM ReMi Test 1, Station 3 0.68 -8% 12% 0.80 Test 2, Station 3 0.90 -9% 16% 0.78 Test 3, Station 3 0.77 -7% 14% 0.80 Test 4, Station 3 0.43 -4% 8% 0.74 Test 5, Station 3 0.92 -8% 17% 0.81 Test 6, Station 3 0.22 -2% 4% 0.79 ReMo Test 1, Station 3 0.66 -21% 12% 0.78 Test 2, Station 3 0.86 -36% 16% 0.74 Test 3, Station 3 0.72 -24% 13% 0.78 Test 4, Station 3 0.34 -14% 6% 0.70 Test 5, Station 3 0.89 -37% 16% 0.76 Test 6, Station 3 0.14 -7% 3% 0.74 Output DTN: MO0306MWDMXCNV.000, worksheet “Uncertainties” of Phase 1 Supporting Calculations for Mixed Convection.xls, rows 34 to 37, columns B, D, E, and J. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-60 October 2004 Table IX-28. Combined Uncertainty in ReMi and ReMo for EBS Ventilation Test Series, Phase 2 Location (ReN/ReM)2 Combined Correction for Systematic Effects Standard Uncertainty from Random Effects (ReM/NuM) Times Partial of NuM with Respect to ReM ReMi Test 1, Station 3 0.71 -8% 13% 0.81 Test 2, Station 3 0.69 -7% 12% 0.80 Test 3, Station 3 0.68 -6% 12% 0.80 Test 4, Station 3 0.68 -5% 12% 0.80 Test 5, Station 3 0.77 -8% 14% 0.80 Test 6, Station 3 0.76 -8% 14% 0.82 Test 7, Station 3 0.75 -7% 13% 0.81 Test 8, Station 3 0.75 -6% 13% 0.81 Test 9, Station 3 0.90 -10% 16% 0.79 Test 10, Station 3 0.89 -9% 16% 0.78 Test 11, Station 3 0.89 -7% 16% 0.77 Test 12, Station 3 0.92 -9% 17% 0.81 Test 13, Station 3 0.91 -8% 16% 0.80 Test 14, Station 3 0.91 -8% 16% 0.79 Test 15, Station 3 0.77 -7% 14% 0.82 Test 16, Station 3 0.77 -7% 14% 0.82 ReMo Test 1, Station 3 0.69 -21% 12% 0.79 Test 2, Station 3 0.61 -28% 11% 0.75 Test 3, Station 3 0.56 -30% 10% 0.74 Test 4, Station 3 0.55 -31% 10% 0.73 Test 5, Station 3 0.72 -23% 13% 0.79 Test 6, Station 3 0.71 -25% 13% 0.78 Test 7, Station 3 0.66 -30% 12% 0.76 Test 8, Station 3 0.66 -31% 12% 0.75 Test 9, Station 3 0.90 -25% 16% 0.77 Test 10, Station 3 0.86 -33% 16% 0.75 Test 11, Station 3 0.80 -48% 14% 0.70 Test 12, Station 3 0.90 -28% 16% 0.78 Test 13, Station 3 0.88 -37% 16% 0.75 Test 14, Station 3 0.86 -46% 15% 0.73 Test 15, Station 3 0.71 -26% 13% 0.78 Test 16, Station 3 0.70 -28% 13% 0.77 Output DTN: MO0306MWDMXCNV.000, worksheet “Uncertainties” of Phase 2 Supporting Calculations for Mixed Convection.xls, rows 54 to 87, columns B, D, E, F, and J. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-61 October 2004 The other contributors to the uncertainty in each mixed-convection Nusselt number are: 1. The uncertainty in the forced convection correlation when the input Reynolds number is known, which is a correction of -9% for systematic effects, at the inner surface only, and a standard uncertainty of 7% from random effects at both surfaces (Table IX-24). 2. The 7.5% standard uncertainty in mixed-convection Nusselt numbers inherent in the Morgan approximation (Section IX.3.5). Tables IX-29 and IX-30 present the combined uncertainties in ( ) x NuMi and ( ) x NuMo , with the various contributors to those uncertainties. As before, this does not include the effect of the qualitatively different flow pattern. Of the evaluated sources of uncertainty in each mixed-convection Nusselt number, the dominant source is the uncertainty in the effective Reynolds number. For the inner surface Nusselt number, the root source is the deviation of measured Nusselt numbers reported by Kuehn and Goldstein (1976 [DIRS 100675]) from their correlation of those measurements. At the outer surface, the root source is the approximation represented by Equation IX-29. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-62 October 2004 Table IX-29. Uncertainty in Predicted M Nu for EBS Ventilation Test Series, Phase 1 From Uncertainty in ReM From NuF Uncertainty From Morgan Approx. Combined Uncertainty 95% Confidence Interval Location Correction for Systematic Effects Random Standard Uncert. Correction for Systematic Effects Random Standard Uncert. Random Standard Uncert. Correction for Systematic Effects Random Standard Uncert. Lower Upper NuMi Test 1, Station 3 -6% 10% -9% 7% 7.5% -15% 14% -39% 9% Test 2, Station 3 -7% 13% -9% 7% 7.5% -16% 16% -43% 11% Test 3, Station 3 -6% 11% -9% 7% 7.5% -15% 15% -41% 11% Test 4, Station 3 -3% 6% -9% 7% 7.5% -12% 12% -33% 9% Test 5, Station 3 -7% 14% -9% 7% 7.5% -16% 17% -44% 13% Test 6, Station 3 -1% 3% -9% 7% 7.5% -10% 11% -30% 9% NuMo Test 1, Station 3 -16% 9% 0% 7% 7.5% -16% 14% -39% 7% Test 2, Station 3 -27% 12% 0% 7% 7.5% -27% 15% -50% -5% Test 3, Station 3 -19% 10% 0% 7% 7.5% -19% 14% -42% 5% Test 4, Station 3 -10% 4% 0% 7% 7.5% -10% 11% -30% 10% Test 5, Station 3 -28% 12% 0% 7% 7.5% -28% 16% -51% -5% Test 6, Station 3 -5% 2% 0% 7% 7.5% -5% 10% -25% 14% Output DTN: MO0306MWDMXCNV.000, worksheet “Uncertainties” of Phase 1 Supporting Calculations for Mixed Convection.xls, rows 54 to 67, columns B through K. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-63 October 2004 Table IX-30. Uncertainty in Predicted M Nu for EBS Ventilation Test Series, Phase 2 From Uncertainty in ReM From NuF Uncertainty From Morgan Approx. Combined Uncertainty 95% Confidence Interval Location Correction for Systematic Effects Random Standard Uncert. Correction for Systematic Effects Random Standard Uncert. Random Standard Uncert. Correction for Systematic Effects Random Standard Uncert. Lower Upper NuMi Test 1, Station 3 -6% 10% -9% 7% 7.5% -15% 15% -40% 10% Test 2, Station 3 -5% 10% -9% 7% 7.5% -14% 14% -39% 10% Test 3, Station 3 -5% 10% -9% 7% 7.5% -14% 14% -38% 11% Test 4, Station 3 -4% 10% -9% 7% 7.5% -13% 14% -38% 11% Test 5, Station 3 -7% 11% -9% 7% 7.5% -16% 15% -41% 10% Test 6, Station 3 -6% 11% -9% 7% 7.5% -15% 15% -41% 11% Test 7, Station 3 -6% 11% -9% 7% 7.5% -15% 15% -40% 11% Test 8, Station 3 -5% 11% -9% 7% 7.5% -14% 15% -40% 11% Test 9, Station 3 -8% 13% -9% 7% 7.5% -17% 16% -44% 10% Test 10, Station 3 -7% 13% -9% 7% 7.5% -16% 16% -43% 12% Test 11, Station 3 -5% 12% -9% 7% 7.5% -14% 16% -42% 13% Test 12, Station 3 -8% 13% -9% 7% 7.5% -17% 17% -45% 11% Test 13, Station 3 -7% 13% -9% 7% 7.5% -16% 17% -44% 13% Test 14, Station 3 -6% 13% -9% 7% 7.5% -15% 17% -43% 13% Test 15, Station 3 -6% 11% -9% 7% 7.5% -15% 15% -41% 11% Test 16, Station 3 -6% 11% -9% 7% 7.5% -15% 15% -41% 11% NuMo Test 1, Station 3 -16% 10% 0% 7% 7.5% -16% 14% -40% 8% Test 2, Station 3 -21% 8% 0% 7% 7.5% -21% 13% -42% 0% Test 3, Station 3 -22% 7% 0% 7% 7.5% -22% 13% -42% -2% Test 4, Station 3 -23% 7% 0% 7% 7.5% -23% 13% -42% -3% Test 5, Station 3 -18% 10% 0% 7% 7.5% -18% 14% -42% 6% Test 6, Station 3 -19% 10% 0% 7% 7.5% -19% 14% -42% 4% Test 7, Station 3 -23% 9% 0% 7% 7.5% -23% 14% -44% -2% Test 8, Station 3 -23% 9% 0% 7% 7.5% -23% 14% -44% -3% Test 9, Station 3 -20% 13% 0% 7% 7.5% -20% 16% -46% 6% Test 10, Station 3 -25% 12% 0% 7% 7.5% -25% 16% -48% -1% Test 11, Station 3 -34% 10% 0% 7% 7.5% -34% 14% -53% -15% Test 12, Station 3 -22% 13% 0% 7% 7.5% -22% 16% -47% 3% Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-64 October 2004 Table IX-30. Uncertainty in Predicted M Nu for EBS Ventilation Test Series, Phase 2 (Continued) From Uncertainty in ReM From NuF Uncertainty From Morgan Approx. Combined Uncertainty 95% Confidence Interval Location Correction for Systematic Effects Random Standard Uncert. Correction for Systematic Effects Random Standard Uncert. Random Standard Uncert. Correction for Systematic Effects Random Standard Uncert. Lower Upper NuMo Test 13, Station 3 -28% 12% 0% 7% 7.5% -28% 16% -51% -5% Test 14, Station 3 -33% 11% 0% 7% 7.5% -33% 15% -53% -13% Test 15, Station 3 -21% 10% 0% 7% 7.5% -21% 14% -43% 2% Test 16, Station 3 -21% 10% 0% 7% 7.5% -21% 14% -44% 1% Output DTN: MO0306MWDMXCNV.000, worksheet “Uncertainties” of Phase 2 Supporting Calculations for Mixed Convection.xls, rows 94 to 127, columns B through K. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-65 October 2004 IX.4.4 Measurement of Nusselt Numbers This appendix follows the practice in the open literature on convective heat transfer, such as Kuehn and Goldstein (1978 [DIRS 130084]). “Measured” circumferential average Nusselt numbers are based on measured heat input and measured temperatures, with corrections for non-convective mechanisms, such as radiative heat transfer and conductive heat transfer. At the Rayleigh numbers and Reynolds numbers in the EBS Ventilation Test Series, conduction to the air is not a significant mechanism. Therefore, a measured value for circumferential average convective heat flux from the inner surface at a central location (Station 3) is: ( ) ( ) x q L D q x q i rad i i in i , ' ' - = ' ' p (Eq. IX-75) where in q is the 24-hour average power generated in the waste packages, i L is the combined length of the waste packages, and ( ) x q i rad , ' ' is the circumferential average radiative flux from the waste packages at location x. For each ventilation test, Table IX-6 or IX-7 gives the value of the average line load, i in L q . For transparent air between concentric cylinders (Incropera and DeWitt 1985 [DIRS 114109], p. 647, Eq. 13.25): ( ) ( ) * 4 4 , 1 1 ) ( s r x T x T x q o o i o i i rad e e e - + .. . .. . - = ' ' (Eq. IX-76) where s is the Stefan-Boltzmann constant (Table 4-28), i e and o e are the measured emissivity of the waste package steel and concrete pipe (Tables 7-4 and 7-5), and each ( ) x T 4 is the 24-hour and circumferential average of the fourth power of the absolute temperature (K). This appendix approximates the averages of the fourth powers from the 24-hour averages, ( ) x T , of absolute temperatures (K) at the top (t), left (l), bottom (b), and right (r) positions on the surfaces as follows: { } [ ] { }4 4 4 4 / ) ( ) ( ) ( ) ( ) ( ) ( x T x T x T x T x T x T ri li bi ti i i + + + = . (Eq. IX-77) { } [ ] { }4 4 4 3 / ) ( ) ( ) ( ) ( ) ( x T x T x T x T x T ro lo to o o + + = . (Eq. IX-78) Because the bottom of the concrete pipe is covered by the invert, its bottom temperature is not included for radiation. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-66 October 2004 By factoring, Equation IX-78 may be put in the form: ( ) ( ) ( ) ( ) ( ).. . .. . - . . . . . . .. . .. . + .. . .. . .. . .. . + .. . .. . .. . .. . + .. . .. . - + . x T x T x T x T x T x T x T x T r x q o i o o i o i i o o i i rad ) ( ) ( ) ( ) ( 1 1 s 3 2 2 3 * ' ' , e e e ( ) ( ) ( ) ( ) ( ) . . . . . . .. . .. . - .. . .. . + + .. . .. . - .. . .. . + - + = 3 3 * ) ( ) ( ) ( 1 1 s 5 . 0 x T x T x T x T x T x T x T x T r o i o i o i o i o o i e e e (Eq. IX-79) Ignoring the term containing the cube of the temperature difference, the approximation becomes: ( ) ( ) ( ).. . .. . - .. . .. . + - + . x T x T x T x T r x q o i o i o o i i rad ) ( ) ( 1 1 s 5 . 0 3 * ' ' , e e e (Eq. IX-80) For each ventilation test, Table IX-31 or IX-32 shows ( ) x q i rad " , calculated in accordance with Equation IX-80, the measured ( ) x qi " , the value of ( ) x hi from Equation IX-16, and the value of ( ) x Nui from Equation IX-2, using k = 0.0263 W/mK. Table IX-31. Inner-Surface Nusselt Number Measurements, EBS Ventilation Test Series, Phase 1 Location qin/pDiL (W/m2) q”rad,© (W/m2) q”© (W/m2) Effective hi (W/m2K) Effective Nui Test 1, Station 3 143 64 78 5.32 195 Test 2, Station 3 140 71 69 4.65 170 Test 3, Station 3 281 126 155 5.94 218 Test 4, Station 3 284 110 174 7.54 276 Test 5, Station 3 284 134 150 5.70 209 Test 6, Station 3 285 90 195 9.94 364 Output DTN: MO0306MWDMXCNV.000, worksheet “Effective Nui” of Phase 1 Supporting Calculations for Mixed Convection.xls. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-67 October 2004 Table IX-32. Inner-Surface Nusselt Number Measurements, EBS Ventilation Test Series, Phase 2 Location qin/pDiL (W/m2) q”rad,© (W/m2) q”© (W/m2) Effective hi (W/m2K) Effective Nui Test 1, Station 3 171 77 94 5.39 198 Test 2, Station 3 171 86 85 5.08 186 Test 3, Station 3 169 92 77 4.79 175 Test 4, Station 3 168 91 77 4.87 178 Test 5, Station 3 282 127 155 5.91 217 Test 6, Station 3 281 137 144 5.55 204 Test 7, Station 3 280 148 131 5.31 194 Test 8, Station 3 280 149 131 5.29 194 Test 9, Station 3 168 79 89 5.11 187 Test 10, Station 3 168 87 81 4.79 175 Test 11, Station 3 169 98 72 4.42 162 Test 12, Station 3 282 133 149 5.66 208 Test 13, Station 3 282 145 137 5.36 196 Test 14, Station 3 283 157 125 5.08 186 Test 15, Station 3 282 134 148 5.69 209 Test 16, Station 3 285 136 150 5.71 209 Output DTN: MO0306MWDMXCNV.000, worksheet “Effective Nui” of Phase 2 Supporting Calculations for Mixed Convection.xls. At the outer surface, heat arrives by radiation and leaves by conduction into the wall and by convection into the air. Therefore, the measured value for circumferential average convective flux is: ( ) ( ) ( ) ( ).. . .. . - - ' ' = ' ' x T x T h x q r x q o cond i rad o a , * (Eq. IX-81) where the radiative flux at the waste package has been multiplied by r* to reflect the larger circumference at the outer wall, cond h is the overall conductive heat transfer coefficient for the combined thickness of concrete and insulation, and ) ( a x T is an average ambient temperature external to the insulation, defined by [ ] 3 / ) ( ) ( ) ( ) ( a a a a x T x T x T x T r l t + + = (Eq. IX-82) The two natural convection tests conducted at the end of Phase 1 establish a value for cond h . Because there is no heat removed by ventilating air in the natural convection tests, conduction through the wall must be equal to the heat input. That is, neglecting end effects, ( ) ( ) [ ] x T x T h L D q r o cond i i in a * - = p (Eq. IX-83) Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-68 October 2004 where the flux from the heat source has been multiplied by r* to reflect the larger circumference at the outer wall. Table IX-33 shows that the calculation of cond h from data in Table IX-8 for the two natural convection tests gives an average value of 1.99 W/m2K, with a standard deviation of less than 1%. For each ventilation test, Table IX-34 or IX-35 shows the measured ( ) x qo ' ' , the value of ( ) x ho from Equation IX-16, and the value of ( ) x Nuo from Equation IX-3, with k = 0.0263 W/mK. For the direct measurements taken during the EBS Ventilation Test Series, the uncertainties are small. Averaging 96 measurements to get a 24-hour average reduces further the effects of random errors. For example, the uncertainty in the average heat input is only 2.6 W out of a total input of 4 kW or more (BSC 2003 [DIRS 160724], Tables 3-15 and 3-16). The major uncertainty in the measured Nusselt numbers is in the approximation for radiative heat transfer. One source of uncertainty is the absence of measured temperatures below the center of the concrete pipe. Other sources are the approximations that underlie the radiation formula, including: • Concentric cylinders • Isothermal surfaces • Transparent air • No end effects The effort documented here did not include a literature search for data regarding deviations from these approximations. This appendix does not provide numerical uncertainties for the measured Nusselt numbers. Table IX-33. Determination of Conductive Heat Transfer Coefficient from Natural Convection Tests Conducted During EBS Ventilation Test Series Avg. To Avg. (To-Ta) hcond Location qin/L (W/m) Source at Wall (W/m2) Avg. Ta (ºC) (ºC) (ºC) (W/m2ºC) Test NC1, Station 3 120 27.9 27.97 45.83 17.87 1.561 Test NC2, Station 3 242 56.2 32.53 68.77 36.23 1.552 average 1.556 std. dev. 0.006 NC1=Natural Convection Test 1; NC2=Natural Convection Test 2. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-69 October 2004 Table IX-34. Outer-Surface Nusselt Number Measurements, EBS Ventilation Test Series, Phase 1 Location r*q”rad,© (W/m2) Avg Ta (ºC) qcond (W/m2) q”o (W/m2) Effective ho (W/m2K) Effective Nuo Test 1, Station 3 19.0 28.57 2.5 16.6 6.45 236 Test 2, Station 3 21.1 28.27 7.8 13.3 7.12 261 Test 3, Station 3 37.4 26.73 10.3 27.1 6.63 243 Test 4, Station 3 32.6 26.23 5.8 26.9 9.16 336 Test 5, Station 3 39.6 23.90 16.1 23.5 7.28 267 Test 6, Station 3 26.8 20.63 4.5 22.3 12.16 445 Output DTN: MO0306MWDMXCNV.000, worksheet “Effective Nui” of Phase 1 Supporting Calculations for Mixed Convection.xls. Table IX-35. Outer-Surface Nusselt Number Measurements, EBS Ventilation Test Series, Phase 2 Location r*q”rad,© (W/m2) Avg Ta (°C) qcond (W/m2) q”o (W/m2) Effective ho (W/m2K) Effective Nuo Test 1, Station 3 22.7 28.43 3.3 19.5 6.21 228 Test 2, Station 3 25.5 29.20 14.7 10.8 5.92 217 Test 3, Station 3 27.4 36.33 18.6 8.9 6.26 229 Test 4, Station 3 27.1 36.07 18.4 8.7 6.71 246 Test 5, Station 3 37.6 30.53 5.3 32.2 7.64 280 Test 6, Station 3 40.8 36.67 10.5 30.2 7.69 282 Test 7, Station 3 44.0 37.97 21.3 22.8 8.33 305 Test 8, Station 3 44.3 37.20 22.2 22.1 8.46 310 Test 9, Station 3 23.5 32.67 3.5 19.9 5.98 219 Test 10, Station 3 25.9 36.13 10.4 15.5 6.33 232 Test 11, Station 3 28.9 39.63 17.4 11.5 9.72 356 Test 12, Station 3 39.5 35.30 8.2 31.2 7.02 257 Test 13, Station 3 43.0 34.77 19.3 23.7 7.37 270 Test 14, Station 3 46.7 38.47 26.5 20.2 9.40 345 Test 15, Station 3 39.7 29.50 13.4 26.3 7.23 265 Test 16, Station 3 40.2 27.73 15.4 24.8 7.22 265 Output DTN: MO0306MWDMXCNV.000, worksheet “Effective Nui” of Phase 2 Supporting Calculations for Mixed Convection.xls. IX.4.5 Corroboration of Predicted Results With Test Data Figures IX-3 through IX-6 compare “measured” and predicted Nusselt numbers at Station 3 for the Phase 1 and Phase 2 tests in the EBS Ventilation Test Series. When the prediction agrees with the measurement, the point lies on the diagonal line. These plots show the 95% confidence limits for the predictions. In some cases, because of systematic errors that are not corrected in the methodology, the predicted value is outside of the confidence limits. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-70 October 2004 The “measured” values of i Nu agree with the predicted values to within the uncertainty (Figures IX-3 and IX-4). However, the “measured” values for o Nu are consistently higher than the predicted values (Figures IX-5 and IX-6). The predicted values of o Nu (Figures IX-5 and IX-6) could be brought within their own 95% confidence limits by solving Equation IX-25 implicitly, thereby eliminating the systematic error caused by the approximation of Equation IX-29. For each case, one could evaluate the right hand side of Equation IX-25 for two or three values of Re, then interpolate in the resulting small table. However, this would not improve the agreement with the measured values. A striking feature of Figure IX-6, in particular, is that the measured values of o Nu span a factor of three, while the predicted values are relatively constant. Most of the variation in measured o Nu occurred in tests at the lowest flow rate, 0.5 m3/s. Table IX-36 is a summary of the measured values of o Nu for all tests that had controlled inlet air conditions and a nominal flow rate of 0.5 m3/s. There appears to be a strong dependence on air inlet temperature that was not seen at higher flow rates. At the lower flow rates, natural convection has a greater influence. The natural convection correlation is based on the flow pattern of Figure IX-1c, but the actual circulation near the outer surface is in the opposite direction, as shown in Figure IX-1d. Because the temperature of the outer surface is near the temperature of the air (Table IX-16 and IX-17), details of the flow pattern may be sensitive to the inlet temperature. Table IX-36 also contains the predicted Nusselt numbers and the values that would result from the Dittus-Boelter formula. The Dittus-Boelter predictions are the result of applying Equation IX-12, using Reynolds numbers from Table IX-12. Although the predictions are low, the Dittus-Boelter values are even lower, by about a factor of three. Concentrating attention on the Nusselt numbers tends to exaggerate the significance of the errors with respect to overall energy transfer in the EBS Ventilation Tests. To provide another perspective, an energy balance can be represented by expressing the various components of energy transfer as percentages of the total input energy. A certain percentage was convected from the inner wall to the air, a percentage was convected from the outer wall to the air, and a percentage was conducted through the outer wall. Using measured data, these percentages must sum to 100%. Figures IX-7 and IX-8 show an energy balance using the methodology for convection to the air and measured conduction losses. All of these plots are based on a vertical section at the center of the configuration (Station 3) and contain no adjustment for longitudinal effects other than the airflow. The figures show the sum, ( ) cond q q D q D o o i i + ' + ' p p , as a percentage of the average line load given in Table IX-6 or IX-7. In forming the sum, qcond is from Table IX-34 or IX-35. The convective heat flux at each surface is: ( ) i o f D D T T Nu k q - - = ' (Eq. IX-84) Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-71 October 2004 The Nusselt numbers are from Tables IX-18 through IX-21, each temperature is from Table IX-13 or IX-14, and k is 0.0263 W/m K. The percentages cluster around 85%. Of the energy convected, 75 to 85% was directly from the waste package, with the remaining 15 to 25% being convected from the drift wall. From this perspective, the effects of errors in o Nu are limited because f o T T - is small. Considered in terms of the effects of the errors in a ventilation model that conserves total energy, the surface temperatures might have to rise enough to remove an additional 10% of the energy. For the ventilation tests, for example, a ventilation model that used the mixed-convection methodology might predict a waste package temperature that was too high by about 2°C and a wall temperature that was too high by about 0.3ºC. In summary, the results of the EBS Ventilation Tests support the mixed convection methodology for prediction of the Nusselt number at the waste package, which is the dominant source for heat transfer to the air. They also support the use of the methodology, rather than a forced convection formula, at the drift wall. The determination of accuracy and precision followed conventional scientific standards, and used sensitivity analyses and bounding techniques, as appropriate. This appendix accounts for uncertainties and variabilities in parameter values and provides for the technical basis for parameter ranges, probability distributions, or bounding values that may be used in predictions. Also, this appendix considers alternative conceptual models of processes that are consistent with available data and current scientific understanding and evaluates the effects that alternative conceptual models have on the predictions. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-72 October 2004 0 100 200 300 0 100 200 300 Measured Nui Predicted Nui Output DTN: MO0306MWDMXCNV.000, modified from worksheet “Nusselt no. plots sta. 3” of Phase 1 Supporting Calculations for Mixed Convection.xls. Figure IX-3. Comparison of “Measured” and Predicted Inner Surface Nusselt Numbers for Phase 1 Ventilation Tests Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-73 October 2004 0 100 200 300 0 100 200 300 Measured Nui Predicted Nui Output DTN: MO0306MWDMXCNV.000, modified from worksheet “Nusselt no. plots sta. 3” of Phase 2 Supporting Calculations for Mixed Convection.xls. Figure IX-4. Comparison of “Measured” and Predicted Inner Surface Nusselt Numbers for Phase 2 Ventilation Tests Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-74 October 2004 0 200 400 600 0 200 400 600 Measured Nuo Predicted Nuo Output DTN: MO0306MWDMXCNV.000, modified from worksheet “Nusselt no. plots sta. 3” of Phase 1 Supporting Calculations for Mixed Convection.xls. Figure IX-5. Comparison of “Measured” and Predicted Outer Surface Nusselt Numbers for Phase 1 Ventilation Tests Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-75 October 2004 0 200 400 600 0 200 400 600 Measured Nuo Predicted Nuo Output DTN: MO0306MWDMXCNV.000, modified from worksheet “Nusselt no. plots sta. 3” of Phase 2 Supporting Calculations for Mixed Convection.xls. Figure IX-6. Comparison of “Measured” and Predicted Outer Surface Nusselt Numbers for Phase 2 Ventilation Tests Table IX-36. Outer-Surface Nusselt Number Measurements for Flow Rate of 0.5 m3/s Phase Test Nominal Line Load (W/m) Inlet Air Temperature (C) Measured Nuo Predicted Nuo Dittus-Boelter Nuo 2 9 220 25 219 175 60 2 10 220 35 232 175 61 2 11 220 45 356 199 58 2 12 360 25 257 191 63 2 13 360 35 270 193 61 2 14 360 45 345 202 58 Source: Tables IX-7, IX-35, and IX-21, and Equation IX-12, using Re from Table IX-12. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 IX-76 October 2004 0% 20% 40% 60% 80% 100% 120% 0 1 2 3 Flow Rate (m3/s) Percent of Total Output DTN: MO0306MWDMXCNV.000, modified from worksheet “Energy balance” of Phase 1 Supporting Calculations for Mixed Convection.xls. Figure IX-7. Energy Balance Using Convection Prediction for Phase 1 Tests (Heat Convected to Air, Augmented by Heat Conducted Through Concrete, as a Percentage of Input Energy) 0% 20% 40% 60% 80% 100% 120% 0 1 2 3 Nominal Flow Rate (m3/s) Percent of Total Output DTN: MO0306MWDMXCNV.000, modified from worksheet “Energy balance” of Phase 2 Supporting Calculations for Mixed Convection.xls. Figure IX-8. Energy Balance Using Convection Prediction for Phase 2 tests (Heat Convected to Air, Augmented by Heat Conducted Through Concrete, as a Percentage of Input Energy) Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 October 2004 APPENDIX X VERIFICATION CALCUATIONS IN SUPPORT OF THE MIXED CONVECTION CORRELATION METHODOLOGY (APPENDIX IX) Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 October 2004 Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 X-1 October 2004 This appendix documents the spreadsheets calculations used in Appendix IX. The electronic copies of these Microsoft Excel spreadsheets are contained in Mixed Convection.zip (DTN: MO0306MWDMXCNV.000). Table X-1 summarizes the contents of the spreadsheets. Further documentation of the cell formulas and referencing is found within the electronic copy of the file. Table X-1. Contents of Spreadsheet used in the Mixed Convection Methodology of Appendix IX File Name (.xls) Contents Phase 1 Supporting Calculations for Mixed Convection Mixed Convection model applied to the EBS Ventilation Test Series, Phase I; Evaluation of uncertainty for EBS Ventilation Tests Series, Phase I; Determination of measured Nusselt numbers for EBS Ventilation Test Series, Phase I; Calculated Energy Balance for Phase I. Phase 2 Supporting Calculations for Mixed Convection Mixed Convection model applied to the EBS Ventilation Test Series, Phase II; Evaluation of uncertainty for EBS Ventilation Tests Series, Phase II; Determination of measured Nusselt numbers for EBS Ventilation Test Series, Phase II; Calculated Energy Balance for Phase II. h-cond from NC tests Evaluation of effective heat transfer coefficient for conduction Mixed Convection Sensitivity Sensitivity of Mixed Convection model to input parameters Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 X-2 October 2004 INTENTIONALLY LEFT BLANK Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 October 2004 APPENDIX XI ANALYSIS OF THE VENTILATION TEST PHASE 2 DATA IN SUPPORT OF THE CALCULATIONS PERFORMED IN APPENDIX IX Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 October 2004 Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 XI-1 October 2004 Results from the Phase 2 Ventilation Tests were used to support validation of the mixed convection correlation in Appendix IX. The raw data is in DTN: SN0208F3409100.009 [DIRS 163079]. The summary data are contained in Vent-Test Phase-II.zip (DTN: MO0306MWDVTPH2.000). There is one file for each test (Phase II Test 1_Q.xls, Phase II Test 2_Q.xls, Phase II Test 3_Q.xls, Phase II Test 4_Q.xls, Phase II Test 5_Q.xls, Phase II Test 6_Q.xls, Phase II Test 7_Q.xls, Phase II Test 8_Q.xls, Phase II Test 9_Q.xls, Phase II Test 10_Q.xls, Phase II Test 11_Q.xls, Phase II Test 12_Q.xls, Phase II Test 13_Q.xls, Phase II Test 14_Q.xls, Phase II Test 15_Q.xls, and Phase II Test 16_Q.xls) that contains: • The data recorded by the datalogger and entered into the Technical Data Management System – sheet name “raw data.” This sheet also contains simple statistical analysis (average, standard deviation, maximum and minimum) for a defined time period. The time period was chosen as a 24-hour period over which data is representative of steady-state conditions where the ventilating air was at or near the desired temperature and relative humidity. The sheet is organized as follows: - Cell A14: DTN: SN0208F3409100.009 associated with the data - Time period chosen for averaging data: Cells C1 and C2 - Rows 4 and 5: addresses corresponding to chosen time period for statistical analysis - Rows 7, 8, 9 and 10: resulting statistical analysis (average, standard deviation, maximum and minimum) for the chosen time period - Row 24: Starting row for the data pulled from the Technical Data Management System • The calculated total power input (summation of the five stations) and line load (total power input divided by the heated length) – sheet name “power.” The sheet is organized as follows: - Columns A through F: Summary of the power data, including the time stamp and recorded power data for the five power stations (taken directly from the “raw data” worksheet) - Column H: Summation of the five recorded power inputs for each time stamp - Column I: Calculated average line load for the test train, defined as the total power input (column H) divided by the total heated length (111’ 4” (33.9 m)), calculated by adding the recorded distance to the leading edge of waste package 25 (98’ 2 1/2” + 105 1/2”) plus the recorded length of waste package 25 (52”) (BSC 2003 [DIRS 160724], p. 2-3). Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 XI-2 October 2004 • The calculated volumetric and mass flow rates – sheet name “flow.” The flow rates for each test were calculated based on air velocity probe differential pressure, relative humidity, barometric pressure, and air temperature measurements. Complete details of the calculation can be found in Testing to Provide Data for Ventilation System Design: Phase 1 (BSC 2003 [DIRS 160724], Section 5.2.1). As a summary, the measured differential pressure was converted to an air velocity. Properties of the ventilating fluid (e.g., the mixture of air and water vapor) were determined using measured relative humidities and temperatures. The air velocity was then combined with the cross-sectional area of the ducting to determine a volume flow rate. The sheet is organized as follows: - Column A through J: Summary of the data required to calculate flow. At station A, there were two differential pressure gauges (VA-VEL-01 and VA-VEL-02), two relative humidity gauges (VA-HUM-H1 and VA-HUM-H2), and nine RTDs measuring air temperature (VA-RTD-01 through VA-RTD-09), that were used in calculating the flow. Each set of measurements was averaged to create the differential pressure, relative humidity, and air temperature needed for the flow calculations. The flow at station D was calculated using measurements from one differential pressure gauge at station D (VD-VEL-01), two relative humidity gauges at station C (VC-HUM-H1 and VC-HUM-H2), and two RTD air temperature gauges at station C (VC-RTD-01 and VC-RTD-02). - Columns L through AD: Calculations of flow rates for Station A - Column AI through AY: Calculations of flow rate for Station D • All constants and dimensions used in the calculation are given in a sheet named “properties.” References for these properties are provided. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 October 2004 APPENDIX XII DOCUMENTATION OF THE VENTILATION PHASE 1 POST-TEST ANSYS ANALYSES FOR MODEL VALIDATION (INPUTS AND OUTPUTS) Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 October 2004 Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 XII-1 October 2004 This appendix documents the Ventilation Test Phase 1 post-test ANSYS modeling for validation purposes, which was developed using the Ventilation Test Phase 1 data, ANSYS software and spreadsheet methods. The input and output files, and Microsoft Excel spreadsheets are contained in DTN: MO0209MWDANS30.017. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 XII-2 October 2004 INTENTIONALLY LEFT BLANK Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 October 2004 APPENDIX XIII ANALYTICAL SOLUTION USING MATHCAD FOR THE CONTRIBUTION OF LATENT HEAT TO THE IN-DRIFT AIR OF A VENTILATED EMPLACEMENT DRIFT USING A SOLUTION FOR STEADY-STATE UNSATURATED FLOW TO MOISTURE POTENTIAL BOUNDARY AT THE DRIFT WALL Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 October 2004 Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 XIII-1 October 2004 Note that the symbol := used throughout this appendix means to assign the right hand value or expression to the left hand variable. Develop a steady solution for radial unsaturated flow to the specified moisture potential conditions. Neglect the gravity component of flow, and consider the van Genuchten constitutive relationships. Soil Physics (Jury et al. 1991 [DIRS 102010], Section 3.4) develops the solution for radial flow under saturated conditions. In the case of steady-state flow under saturated conditions, the water conservation equation for a cylindrical coordinate geometry is given by: ( ) 0 1 = · · r J r dr d r (Eq. XIII-1) where r = Radial Coordinate Jr = Darcy Flux in the Radial Direction Equation XIII-1 can be integrated once to produce the result: 0 2 z Q t tan cons J r r · = = = · p f (Eq. XIII-2) where Q = Steady-state moisture flow z0 = Drift length The radial flux under Darcy’s Law is given by: dr dp K J s r · - = (Eq. XIII-3) where Ks = Saturated Hydraulic Conductivity p = Pressure or Pressure Head Depending on convention adopted for Darcy’s Law Writing Darcy’s Law for radial flow to the tunnel surface: r z Q dr dp Ks · · = · - 0 2 p (Eq. XIII-4) Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 XIII-2 October 2004 This equation can be integrated after placing all factors explicitly for r on the same side of the equation: r dr z K Q dp s · · · - = 0 2 p (Eq. XIII-5) Since p(R1) = p1 and p(R2) = p2 are specified at the boundary, then: . . · · - = 2 1 2 1 0 2 R R s p p r dr z K Q dp p (Eq. XIII-6) from which we calculate: ( ) . .. . . .. . - · · · = 1 2 2 1 0 2 R R ln p p z K Q s p (Eq. XIII-7) This expression agrees with the formulation presented in Soil Physics (Jury et al. 1991 [DIRS 102010], p. 113 Equation 3.92). Now consider the unsaturated flow case. The pressure gradient becomes a moisture potential gradient. For unsaturated flow, the unsaturated hydraulic conductivity is a strong nonlinear function of the moisture potential .. Neglecting the elevation head: 1 1 1 0 . . - = + = g p H w (Eq. XIII-8) 2 2 2 0 . . - = + = g p H w (Eq. XIII-9) where H1 = Total Potential at the Drift Surface R1 H2 = Total Potential at the Outer Boundary R2 .1 = Moisture Potential at Radius R1 Set by the RH in the Drift .2 = Moisture Potential at Radius R2 Set by Undisturbed State of Capillary Equilibrium .w = Unit Weight of Water g = gravitational constant Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 XIII-3 October 2004 Writing Darcy’s Law for unsaturated radial flow: ( ) r z Q dr dH K · · = · - 0 2 p . (Eq. XIII-10) Noting that if we neglect the elevation head: . ˜ H ( ) r z Q dr d K · · = · - 0 2 p . . (Eq. XIII-11) The convention is adopted that moisture potential is in units of head (Jury et al. 1991 [DIRS 102010], p. 51). Now the van Genuchten constitutive relation can be invoked. From Contaminant Hydrogeology (Fetter 1993 [DIRS 102009], p. 182), the constitutive relation is: ( ) ( ) ( ) [ ] { } ( ) [ ]2 2 1 1 1 1 m n m n n s K K a. a. a. . + + - = - - (Eq. XIII-12) where n = 1/(1-m) a = Van Genuchten alpha m = Van Genuchten fitting parameter Substituting in the constitutive relation into Darcy’s Law: ( ) ( ) [ ] { } ( ) [ ] r z Q dr d K m n m n n s · · = · + + - - - - 0 2 2 1 2 1 1 1 p . a. a. a. (Eq. XIII-13) Equation XIII-13 can be integrated in the same manner: ( ) ( ) [ ] { } ( ) [ ] . . . .. . . .. . · · = · = · + + - - - - 2 1 2 1 1 2 0 0 2 2 1 ln 2 2 1 1 1 R R m n m n n s R R z Q r dr z Q d K p p . a. a. a. . . (Eq. XIII-14) Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 XIII-4 October 2004 Note that the sign convention in the constitutive law is positive while in Darcy’s Law it is negative. Note also that .1 and .2 are expressed in units of head consistent with the sign convention presented in Soil Physics (Jury et al. 1991 [DIRS 102010], p. 151). Substituting in the definition of hydraulic conductivity (Fetter 1993 [DIRS 102009], p. 181): µ .gk Ks = (Eq. XIII-15) where k = intrinsic permeability (m2) µ = fluid viscosity (N·s/m2) ( ) ( ) [ ] { } ( ) [ ] . .. . . .. . · + + - · · · - = . - - 1 2 2 2 1 0 2 1 1 1 1 2 R R ln d gk z Q m n m n n . a. a. a. µ . p . . (Eq. XIII-16) Now consider the boundary conditions, and the geometry for the problem. Use an RH of 30 percent (Section 4.1.2) in the ventilated drift and use the Kelvin Equation to calculate moisture potential (Jury et al. 1991 [DIRS 102010], p. 60): . .. . . .. . = RT M exp RH w w . .1 (Eq. XIII-17) Input properties for analysis. The properties for water are obtained from Section 4.1.12 at 350 K: .w = 973.7 kg/m3 = 0.9737 gm/cm3 From Table 4-20: Mw = 18 gm/mol R = 8.315 J/mol·K Substituting into Equation XIII-17: ( ) .. . .. . · = 100 RH ln M RT T , RH w w . . ( ) 2 8 10 896 1 350 30 s m kg . K %, · · - = . Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 XIII-5 October 2004 The moisture potential is expressed in pressure. Calculate the moisture potential in units of head: ( ) cm . m . s m . m kg . s m kg . g K %, w 6 4 2 3 2 8 1 10 985 1 10 985 1 81 9 7 973 10 896 1 350 30 · = · = .. . .. . · .. . .. . · · - = = . . . From Table 4-6 and Table 4-9, the hydrologic properties for the repository host rock unit surrounding the drift (Tptpll or tsw35) are: Matrix permeability = km = 4.48·10-18 m2 Matrix porosity = fm = 0.1486 Van Genuchten matrix alpha = a = 1.08·10-5 Pa-1 Van Genuchten matrix fitting parameter = m = 0.216 Residual matrix saturation = .r = slrm· fm = 0.0178 Satiated matrix saturation = .s = slsm· fm = 0.1486 From Table 4-18 at 350 K: µw = 3.65·10-4 N·s/m2 To convert a from Pa-1 to cm-1, multiply by .wg: 1 3 1 2 3 2 2 5 10 032 1 1032 0 81 9 7 973 1 1 1 1 10 08 1 - - - - · = = .. . .. . .. . .. . . . . . . . . . . . . . · . . . . . . . . . . . . . .. . . .. . · = cm . m . s m . m kg . s m kg N m N Pa Pa . a The retention relationship (Fetter 1993 [DIRS 102009], p. 172, Equation 4.9) is used to calculate the moisture potential: ( ) [ ]m n r s r a. . . . . + - + = 1 (Eq. XIII-18) where m n - = 1 1 (Eq. XIII-19) Solve Equation XIII-18 for moisture potential in terms of volumetric moisture content: Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 XIII-6 October 2004 ( ) a . . . . . n m r r s 1 1 1 .. .. . .. .. . .. . .. . - + - - + - = (Eq. XIII-20) The average saturation of Tptpll (tsw35) given the saturations measured in USW SD-7, USW SD-9, USW NRG-6, USW NRG-7/7A, and USW UZ-7A in Table 4-4 is 0.74. The average volumetric moisture content is: 1100 . 0 74 . 0 = · = s .. Solving Equation XIII-20 then yields: . = 2908 cm = .2 The radius of the drift, R1, is 2.75 m (Table 4-16). Assume a radius of influence, R2, of 6 m. The drift length, z0, is 600 m. Calculate the steady-state moisture flow at the drift wall by solving the integral in Equation XIII-16: s m . Q 3 8 10 002 2 - · = The steady-state moisture flow, expressed as a liquid flux toward the drift wall, is: yr mm 0.061 1m 1000mm 1yr 31556926s 600m 2.75m 2p s m 10 2.002 z R p Q 3 8 = · · · · · = · · - 0 1 2 Calculate the latent heat transfer over the 50-year ventilation period by multiplying the flow by the latent heat of vaporization at 350 K (Table 4-18): ( ) J 10 . 1yr 31556926s 50yr 1kJ 1000J kg kJ 2317 m kg 973.7 s m 10 2.002 10 3 3 8 · = . .. . . .. . · · .. . .. . · . .. . . .. . · .. . .. . · . .. . . .. . · - 130 7 The total waste package heat input over 50 years and 600 meters is 8.60×1014 J (DTN: MO0307MWDAC8MV.000, worksheet “Ventilation Efficiency”). The contribution of latent heat expressed as a percentage of the total waste package heat input is: 0.01% J 10 8.605 J 10 7.130 14 10 = · · Note that the above analysis is based upon a steady state analysis and measured values of moisture potential. At the time of waste emplacement within the repository emplacement drift, a transient flow response will be induced with a higher flow rate that might result in more rapid Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 XIII-7 October 2004 dewatering of the saturated matrix pore space than is predicted on the basis of the steady state analysis presented above. However, the results of Multiscale Thermohydrologic Model (BSC 2004 [DIRS 169565]) show that at selected locations the matrix saturation remains high during the preclosure period. For example, matrix saturation remains high at locations P2ER8C6, P2WR8C8, P2WR5C10, and P3R7C12 (BSC 2004 [DIRS 169565], Figures 6.3-7 through 6.3-10). Further, the volumetric moisture content versus temperature relation as measured from neutron logging of boreholes 79 and 80 during the Drift Scale Test heating phase shows little reduction in volumetric moisture content below the boiling point of water (DTN: MO0406SEPTVDST.000 [DIRS 170616], file “both.xls”). Since the ventilation analysis predicts below boiling conditions, the latent heat of vaporization from the dewatering of the saturated matrix under transient flow is not expected to be significant. The calculation of the farfield moisture potential from the saturation on core measurements (Table 4-4) and the van Genuchten retention relationship may be compared with measurements of water potential made in the ECRB Cross-Drift (Table 4-5). At a depth, R2, of 5.62 m, the measured water potential, .2 , is 10 m. The potential at the drift wall was calculated previously to be .1 (30%, 350K) = 1.985·104 m. The drift length, z0, is again 600 m. Calculate the steady-state moisture flow at the drift wall by solving the integral in Equation XIII-16: s m 10 9.1196 3 8 - · = Q The steady-state moisture flow, expressed as a liquid flux toward the drift wall, is: yr mm 0.278 1m 1000mm 1yr 31556926s 600m 2.75m 2 s m 10 96 11 . 9 p 2 3 8 0 1 = · · · · · = · · - p z R Q Calculate the latent heat transfer over the 50-year ventilation period by multiplying the flow by the latent heat of vaporization at 350 K (Table 4-18): ( ) J 10 3.246 1yr 31556926s 50yr 1kJ 1000J kg kJ 2317 m kg 973.7 s m 10 9.1196 11 3 3 8 · = . .. . . .. . · · .. . .. . · . .. . . .. . · .. . .. . · . .. . . .. . · - The total waste package heat input over 50 years and 600 meters is 8.60·1014 J (DTN: MO0307MWDAC8MV.000, worksheet “Ventilation Efficiency”). The contribution of latent heat expressed as a percentage of the total waste package heat input is: 0.04% J 10 8.605 J 10 3.246 14 11 = · · Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 XIII-8 October 2004 INTENTIONALLY LEFT BLANK Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 October 2004 APPENDIX XIV DOCUMENTATION OF THE ANSYS-LA-COARSE-INSTANTANEOUS-EFFICIENCYAPPLICATION (INPUTS AND OUTPUTS) Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 October 2004 Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 XIV-1 October 2004 This appendix documents the ANSYS-LA-Coarse-Instantaneous-Efficiency-Application model which was developed using the ANSYS software and spreadsheet methods. The input and output files, and Microsoft Excel spreadsheet are contained in the file ANSYS-LA-Coarse- Instantaneous-Efficiency-Application.zip (DTN: MO0306MWDCIEAP.000). Table XIV-1 is a description of the input and output files, and the worksheets contained in the spreadsheet ANSYS-LA-Coarse-Instantaneous-Efficiency-Application.xls. Further documentation of the cell formulas and referencing are found within the electronic copy of the file. Table XIV-1. Contents of ANSYS-LA-Coarse-Instantaneous-Efficiency-Application.zip ANSYS Input and Output Files File Description decay_data_c3.input ANSYS input file containing the waste package heat decay for segment 3, reduced by the ventilation efficiency. decay_data_c8.input ANSYS input file containing the waste package heat decay for segment 8, reduced by the ventilation efficiency. th_data.input ANSYS input file containing the thermal properties of the repository layers and the EBS components. la800.dat ANSYS input file which generates the mesh and assigns thermal properties to each cell within the mesh. la800.db ANSYS output file. la800.grph ANSYS output file. la800.sub ANSYS output file. la800.out ANSYS output file. air_temp_c3 and air_temp_c8 ANSYS input files containing the inlet air temperature of the specified segment (from ANSYS-LA-Coarse.xls). dr_h_c3 and dr_h_c8 ANSYS input file containing the drift wall convection coefficients for the specified segment (from ANSYS-LA-Coarse.xls). wp_h_c0 through wp_h_c10 ANSYS input files containing the waste package convection coefficients for the specified segment (from ANSYS-LA-Coarse.xls). la800c3_ev1.dat and la800c8_ev1.dat Main ANSYS input files. la800c3_ev1.db and la800c8_ev1.db ANSYS output files. la800c3_ev1.grph and la800c8_ev1.grph ANSYS output files. la800c3_ev1.dsub and la800c10.dsub ANSYS output files. la800c3_ev1.mntr and la800c8_ev1.mntr ANSYS output files. la800c3_ev1.osav and la800c8_ev1.osav ANSYS output files. la800c3_ev1.rth and la800c8_ev1.rth ANSYS output files. la800c3_ev1.stat and la800c8_ev1.stat ANSYS output files. la800c3_ev1.s01 to .s21 and la800c8_ev1.s01 to .s21 ANSYS output files. la800c3_ev1.out and la800c8_ev1.out Main ANSYS output files. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 XIV-2 October 2004 Table XIV-1. Contents of ANSYS-LA-Coarse-Instantaneous-Efficiency-Application.zip (Continued) ANSYS Input and Output Files File Description result_c3_ev1 and result_c8_ev1 Temperature results that are cut from the end of the .out files and imported to ANSYS-LA-Coarse.xls. 100m data Contains the data from ANSYS-LA-Coarse.xls at 100 m (segment 3) used in the instantaneous ventilation efficiency application and the output of the ANSYS model. Plot 100m Plots the waste package and drift wall temperatures for ANSYS-LACoarse and ANSYS-LA-Coarse-Instantaneous-Efficiency-Application. c3-t0-19 Contains the temperature results from the result_c3 ANSYS output files. Performs a circumferential weighted average given the temperatures of each element of the drift wall and waste package. 600m data Contains the data from ANSYS-LA-Coarse.xls at 600 m (segment 8) used in the instantaneous ventilation efficiency application and the output of the ANSYS model. Plot 600m Plots the waste package and drift wall temperatures for ANSYS-LACoarse and ANSYS-LA-Coarse-Instantaneous-Efficiency-Application. c8-t0-19 Contains the temperature results from the result_c8 ANSYS output files. Performs a circumferential weighted average given the temperatures of each element of the drift wall and waste package. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 October 2004 APPENDIX XV DOCUMENTATION OF THE ANSYS-LA-COARSE-INSTANTANEOUS-EFFICIENCYTWP- APPLICATION (INPUTS AND OUTPUTS) Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 October 2004 Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 XV-1 October 2004 This appendix documents the ANSYS-LA-Coarse-Instantaneous-Efficiency-Twp-Application model, which was developed using the ANSYS software and spreadsheet methods. The input and output files, and Microsoft Excel spreadsheet are contained in the file ANSYSLA- Coarse-Intstantaneous-Efficiency-Twp-Application.zip (DTN: MO0306MWDCIETA.000). Table XV-1 is a description of the input and output files, and the worksheets contained in the spreadsheet ANSYS-LA-Coarse-Intstantaneous-Efficiency-Twp-Application.xls. Further documentation of the cell formulas and referencing are found within the electronic copy of the file. Table XV-1. Contents of ANSYS-LA-Coarse-Intstantaneous-Efficiency-Twp-Application.zip ANSYS Input and Output Files File Description decay_data_c3.input ANSYS input file containing the waste package heat decay for segment 3, reduced by the ventilation efficiency. decay_data_c8.input ANSYS input file containing the waste package heat decay for segment 8, reduced by the ventilation efficiency. th_data.input ANSYS input file containing the thermal properties of the repository layers and the EBS components. la800.dat ANSYS input file which generates the mesh and assigns thermal properties to each cell within the mesh. la800.db ANSYS output file. la800.grph ANSYS output file. la800.sub ANSYS output file. la800.out ANSYS output file. air_temp_c3 and air_temp_c8 ANSYS input files containing the inlet air temperature of the specified segment (from ANSYS-LA-Coarse.xls). dr_h_c3 and dr_h_c8 ANSYS input file containing the drift wall convection coefficients for the specified segment (from ANSYS-LA-Coarse.xls). wp_h_c0 through wp_h_c10 ANSYS input files containing the waste package convection coefficients for the specified segment (from ANSYS-LA-Coarse.xls). la800c3_ev2.dat and la800c8_ev2.dat Main ANSYS input files. la800c3_ev2.db and la800c8_ev2.db ANSYS output files. la800c3_ev2.grph and la800c8_ev2.grph ANSYS output files. la800c3_ev2.dsub and la800c10.dsub ANSYS output files. la800c3_ev2.mntr and la800c8_ev2.mntr ANSYS output files. la800c3_ev2.osav and la800c8_ev2.osav ANSYS output files. la800c3_ev2.rth and la800c8_ev2.rth ANSYS output files. la800c3_ev2.stat and la800c8_ev2.stat ANSYS output files. la800c3_ev2.s01 to .s21 and la800c8_ev2.s01 to .s21 ANSYS output files. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 XV-2 October 2004 Table XV-1. Contents of ANSYS-LA-Coarse-Intstantaneous-Efficiency-Twp-Application.zip (Continued) ANSYS Input and Output Files File Description la800c3_ev2.out and la800c8_ev2.out Main ANSYS output files. result_c3_ev2 and result_c8_ev2 Temperature results that are cut from the end of the .out files and imported to ANSYS-LA-Coarse.xls. 100m data Contains the data from ANSYS-LA-Coarse.xls at 100 m (segment 3) used in the instantaneous ventilation efficiency application and the output of the ANSYS model. Plot 100m Plots the waste package and drift wall temperatures for ANSYS-LACoarse and ANSYS-LA-Coarse-Instantaneous-Efficiency-Application. c3-t0-19 Contains the temperature results from the result_c3 ANSYS output files. Performs a circumferential weighted average given the temperatures of each element of the drift wall and waste package. 600m data Contains the data from ANSYS-LA-Coarse.xls at 600 m (segment 8) used in the instantaneous ventilation efficiency application and the output of the ANSYS model. Plot 600m Plots the waste package and drift wall temperatures for ANSYS-LACoarse and ANSYS-LA-Coarse-Instantaneous-Efficiency-Application. c8-t0-19 Contains the temperature results from the result_c8 ANSYS output files. Performs a circumferential weighted average given the temperatures of each element of the drift wall and waste package. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 October 2004 APPENDIX XVI CALCULATION FOR ESTIMATING THE IN-DRIFT CROSS-SECTIONAL AREA AVAILABLE FOR AIR FLOW Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 October 2004 Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 XVI-1 October 2004 A B C D E Circle 1 Circle 2 G F From Table 4-16: m 75 2 2 m 5 5 AD ACE AB . . = = = = m 806 0 CE . = From Table 4-15: m 822 0 2 m 644 1 FG . . = = Using the Pythagorean Theorem: 2 2 2 AB BC AC = + Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 XVI-2 October 2004 Then: ( ) m 945 1 806 0 75 2 75 2 BC 2 2 . . . . = - - = .CAB is: ( ) 707 0 75 2 806 0 75 2 AB AC CAB . . . . cos = - = = . Or: ( ) o 45 707 0 CAB 1 = = . - . cos Since ACE bisects BCD, .DAB is twice .CAB, or 90°. Since the sum of all internal angles emanating from the center of a circle is 360°, the pie shaped slice composed of points A, B, and D, and arc BED is ¼ the area of the Circle 1. The area available for flow is: the area of Circle 1; minus the area of the pie shaped slice composed of points A, B, and D, and arc BED; plus the area of the triangle composed of points ABCD; minus the area of the Circle 2. The area of Circle 1 is: 2 2 2 m 758 23 75 2 AB . . = · = p p The area of the pie shaped slice composed of points A, B, and D, and arc BED is: 2 2 2 m 940 5 75 2 4 1 AB 4 1 . . = · = p p The area of the triangle composed of points ABCD is: ( ) 2 m 781 3 806 0 75 2 945 1 2 2 1 AC BCD 2 1 . . . . = - · · · = · · The area of Circle 2 is: 2 2 2 m 123 2 822 0 FG . . = · = p p Therefore, the area available for flow is: 2 m 476 19 123 2 781 3 940 5 758 23 . . . . . = - + - Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 October 2004 APPENDIX XVII CALCULATION OF DITTUS-BOELTER HEAT TRANSFER COEFFICIENTS FOR THE VENTILATION TEST PHASE I CASES 1 THROUGH 5 Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 October 2004 Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 XVII-1 October 2004 Table XVII-1 provides details of the calculations of Dittus-Boelter heat transfer coefficients. These coefficients are presented in Table 7-10 for comparison with the mixed convection correlation. Table XVII-1. Calculating the Convection Heat Transfer Coefficients for the Ventilation Test Phase 1 Cases 1 Through 5 Using the Dittus-Boelter Correlation for Fully Developed Turbulent Flow in a Smooth Cylinder Input Parameter Value Source Constant (pi), dimensionless 3.14 Universal Constant Emplacement Drift Diameter (D), m 1.3716 Table 7-5 (convert in to m) Waste Package Diameter (d), m 0.4064 Table 7-4 (convert in to m) Wetted Perimeter (P), m 5.6 P=Pi · (D+d) Cross Section Area (A), m2 1.35 A=pi/4 · (D2-d2) Hydraulic Diameter (Dh), m 0.9652 Dh=4A/P=D-d Air Density (rho), kg/m3 1.1614 Table 4-17 (for 300K) Air Thermal Conductivity (k), W/m·K 0.0263 Table 4-17 (for 300K) Air Specific Heat (Cp), J/kg-K 1007 Table 4-17 (for 300K) Air Dynamic Viscosity (mu), kg/m-s 1.846E-05 Table 4-17 (for 300K) Air Prandtl Number (Pr), dimensionless 0.707 Table 4-17 (for 300K) Case 3 and Case 5 Air Flow Rate (Q), m3/s per drift 0.5 Table 7-2 Air Flow Velocity (v), m/s 0.37 v=Q/A Reynolds Number (Re), dimensionless 22468.25 Re=rho·v·Dh/mu (Incropera and DeWitt 1996 [DIRS 108184], Section 8.1.2) Nusselt Number (Nu), dimensionless 60.64 Nu=0.023·Re0.8·Pr0.4 (Incropera and DeWitt 1996 [DIRS 108184], Section 8.5) Conv. Heat Transfer Coef. (h), W/m2·K 1.65 h=k·Nu/Dh (Incropera and DeWitt 1996 [DIRS 108184], Section 8.5) Case 1 and Case 4 Air Flow Rate (Q), m3/s per drift 1 Table 7-2 Air Flow Velocity (v), m/s 0.74 v=Q/A Reynolds Number (Re), dimensionless 44936.49 Re=rho·v·Dh/mu (Incropera and DeWitt 1996 [DIRS 108184], Section 8.1.2) Nusselt Number (Nu), dimensionless 105.58 Nu=0.023·Re0.8·Pr0.4 (Incropera and DeWitt 1996 [DIRS 108184], Section 8.5) Conv. Heat Transfer Coef. (h), W/m2·K 2.88 h=k·Nu/Dh (Incropera and DeWitt 1996 [DIRS 108184], Section 8.5) Case 2 Air Flow Rate (Q), m3/s per drift 2 Table 7-2 Air Flow Velocity (v), m/s 1.48 v=Q/A Reynolds Number (Re), dimensionless 89872.98 Re=rho·v·Dh/mu (Incropera and DeWitt 1996 [DIRS 108184], Section 8.1.2) Nusselt Number (Nu), dimensionless 183.8 Nu=0.023·Re0.8·Pr0.4 (Incropera and DeWitt 1996 [DIRS 108184], Section 8.5) Conv. Heat Transfer Coef. (h), W/m2·K 5.01 h=k·Nu/Dh (Incropera and DeWitt 1996 [DIRS 108184], Section 8.5) Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 XVII-2 October 2004 INTENTIONALLY LEFT BLANK Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 October 2004 APPENDIX XVIII QUALIFICATION OF INPUTS OBTAINED FROM OUTSIDE SOURCES Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 October 2004 Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 XVIII-1 October 2004 This appendix demonstrates that inputs from outside sources, including those listed in Table 4-22, are suitable for their uses in this report. Handbooks are considered to be compilations of established facts. However, handbooks in themselves derive or present no new information; they only present what has been published in the open literature, either in textbooks or publications. Thus, when a textbook, source, or a publication is referenced (or cited) by a handbook, the textbook, source, or publication becomes reliable because it is part of the handbook, which in its entirety is established fact. Therefore, some of the following sources are demonstrated to be reliable for the intended use identified in Table 4-22 because the reliability of these sources (per AP-SIII.10Q, Section 5.2.1(k)) is demonstrated by being cited as references in the indicated handbook(s) and thus widely used in standard work practices by engineers and scientists. The other sources are demonstrated as being reliable by other specific methods as described. The extent to which the data (information or equations) demonstrate the properties (information or mathematics) of interest is also addressed. Qualification of the use of information from Bird, R.B.; Stewart, W.E.; and Lightfoot, E.N. 1960 [DIRS 103524]. The referenced source by Bird et al. was first published in 1960 and has been in publication ever since. This source is referenced by handbooks, specifically those by Cho et al. (1998 [DIRS 160802], reference number 10) and Perry et al. (1984 [DIRS 125806], in the general references for Section 10: Heat Transmission). The information from the source by Bird et al. is reliable and qualified for the intended use because it has been in publication for over four decades. This source is cited in two handbooks in the subject area of heat and mass transfer, and thus is widely used in the standard work practices on these topics. The extent to which this source of information addresses the use of equations for annular radiant heat transfer is considered adequate because these topics are well known, as documented here. Qualification of the use of information from Carslaw, H.S. and Jaeger, J.C. 1959 [DIRS 100968]. The referenced source by Carslaw and Jaeger was first published in 1946. The second edition was first published in 1959 and has been reprinted 13 times. This source is referenced in two handbooks, one by Yovanovich (1998 [DIRS 171591], reference number 11 in Chapter 3: Conduction and Thermal Contact Resistances (Conductances)), and one by and Perry et al. (1984 [DIRS 125806], in the general references for Section 10: Heat Transmission). The information from the source by Carslaw and Jaeger is reliable and qualified for the intended use because it has been in publication for over four decades, it is cited in two handbooks in the subject area of heat conduction, and thus is widely used in the standard work practices on these topics. The extent to which this source of information addresses the linearization of radiant heat transfer and analytical/mathematical results for conduction heat transfer is considered adequate because these topics are well known, as documented here. Qualification of the use of information from Conte, S.D. and de Boor, D. 1972 [DIRS 159800]. The referenced source by Conte and de Boor is a text on elementary numerical analysis that was published in 1972 as the second edition. The information in this text as used in this report is corroborated from other publications. The error in linear interpolation is also described by Alenitsyn et al. (1997 [DIRS 171443], Section 10.2.2). The mean value theorem and derivatives are described by Weinberger (1965 [DIRS 163216], Chapter IV, Section 24). The bisection method as used for the approximate solutions of equations is also described by Alenitsyn et al. (1997 [DIRS 171443], Section 10.4.1). The information from the source by Conte and de Boor is considered reliable for its intended use because it has been in publication Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 XVIII-2 October 2004 for over three decades and is in its second edition, and the information is corroborated by other sources as noted. The extent to which this source of information addresses the error in linear interoplation, mean value theorem and derivatives, and the bisection method, is considered adequate because these topics are well known, as documented here. Qualification of the use of information from Fetter, C.W. 1993 [DIRS 102009]. The referenced source by Fetter on the topics of the theory supporting analytical equations for steady-state unsaturated flow in porous medium and the use of the van Genuchten relation was reviewed by the following individuals: J.M. Bahr at the University of Wisconsin – Madison; R.A. Griffin at the University of Alabama; J.I. Hoffman at Eastern Washington University; M. Th. Van Genuchten at the U.S. Department of Agriculture Salinity Laboratory; S. Kornder at the James River Paper Company; G. Sposito at the University of California – Berkeley; N. Valkenburg at Geraghty and Miller, Inc.; and P. Wierenga at the University of Arizona. Noting that the information of interest from Fetter pertains to unsaturated flow and the van Genuchten relation, and the fact that this source was reviewed by Martinus Th. van Genuchten, among others, the source is considered reliable for its intended use. The extent to which this source of information addresses the supporting analytical equations for steady-state unsaturated flow in porous medium and the use of the van Genuchten relation is considered adequate because these topics were extensively reviewed, as documented here. Qualification of the use of information from Hahn, G.J. and Shapiro, S.S. 1967 [DIRS 146529]. The referenced source by Hahn and Shapiro is used for the delta method for investigating the sensitivity, or effect of uncertainty, of key input parameters with respect to the dependent variables. This method of investigating the uncertainty of a result is also described by the American Society of Mechanical Engineers (1998 [DIRS 153195], Section 7). The information from the source by Hahn and Shapiro is reliable and qualified for intended use because it also appears in an ASME standard as ASME PTC 19.1-1998 [DIRS 153195], as referenced here. The extent to which this source of information addresses the use of the delta method to investigate the effect of uncertainty is considered adequate because this topic appears in a standard, as documented here. Qualification of the use of information from Hartman, H.L. 1982 [DIRS 128009]. The referenced source by Hartman for calculating the saturation vapor pressure of water is demonstrated as being reliable by corroborating a value calculated in the report from Hartman. This value is calculated in Section 6.9.1 as a saturation vapor pressure at 42°C of 2.427 in. Hg using equation 21-1 from Hartman. To convert in. Hg to Newton/m2 (which is the pascal) multiply in. Hg by 3376.9 (Perry et al. 1984 [DIRS 125806], Table 1-5 and Table 1-6). Thus, 2.427 in. Hg is 8195.7 Pa. The saturation vapor pressure is also obtained from Haar et al. (1984 [DIRS 105175], Table 1) at 42°C as P = 0.082054 MPa = 8205 Pa (where M denotes mega as a SI prefix (Perry et al. 1984 [DIRS 125806], Table 1-3). These two values for the saturation vapor pressure of water, 8195.7 and 8205 Pa, are sufficiently close to be considered the same. Therefore, the information from the source by Hartman for calculation of the saturation vapor pressure of water and related psychrometric information is considered reliable for the intended purpose. The extent to which this source of information addresses the use of psychrometric information is considered adequate because these topics are well known, as documented here. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 XVIII-3 October 2004 Qualification of the reference by Incropera, F.P. and DeWitt, D.P. 1996 [DIRS 108184]. The referenced source by Incropera and DeWitt is referenced as the third edition in the handbook by Rohsenow et al. (1998 [DIRS 169241], reference number 6 in Chapter 2: Thermophysical Properties). The source cited here is the fourth edition of this publication. The information from this source by Incropera and DeWitt is reliable and qualified for the intended use because it has been in publication through four editions, this source is cited in handbooks, it is a textbook (with exercises), and thus is widely used in the standard work practices on thermophysical properties and heat transfer topics. The extent to which this source of information addresses rock and concrete emissivity, thermophysical properties of air and water, constants, heat transfer correlations, definitions, radiant heat transfer for an annulus, treatment of air as a non-participating medium in radiant heat transfer, identities, boundary layer formation, and radiant heat transfer between concentric cylinders is considered adequate because these topics are well known, as documented here. Qualification of the use of information from Jury, W.A.; Gardner, W.R.; and Gardner, W.H. 1991 [DIRS 102010]. The referenced source by Jury et al. on a the theory supporting analytical solutions supporting stead-state flow in porous media, information regarding vapor diffusion and enhanced vapor diffusion, heat capacity of geologic media, and the volumetric heat capacity of air is described in the fifth edition of this source which was initially published in 1972. The authors of this source have published extensively on these subjects as evidenced by their appearance in the bibliography of Hillel (1998 [DIRS 165404]). W.A Jury is cited seven times, W.R. Gardner is cited 18 times, and W.H. Gardner is cited three times. The qualifications of the personnel generating the source of information is considered adequate through extensive publication over 30 years (from 1972), and thus the information from the source by Jury et al. (1991 [DIRS 102010]) is consider reliable for the intended uses. The extent to which this source of information addresses the topics noted here is considered adequate because these topics are well known, as documented here. Qualification of the reference by Kays, W.M. and Leung. E.Y. 1963 [DIRS 160763]. The referenced source by Kays and Leung is referenced in the handbook by Rohsenow et al. (1998 [DIRS 169241], reference 111 in Chapter 5: Forced Convection, Internal Flow in Ducts). The information from the source by Kays and Leung is reliable and qualified for the intended use because it is a topic-specific paper published in the International Journal of Heat and Mass Transfer and is cited in a handbook. The extent to which this source of information addresses forced convection in annular passages is considered adequate because this topic is well known, as documented here. Qualification of the reference by Kays, W.M. and Perkins, H.C. [DIRS 160782]. The referenced source by Kays and Pekins is referenced in the handbook by Rohsenow et al. (1998 [DIRS 169241], reference number 263 in Chaper 5: Forced Convection, Internal Flow in Ducts). The information from the source by Kays and Perkins is reliable and qualified for the intended use because it is a topic-specific paper published in an earlier edition of the Handbook of Heat Transfer (see reference number 263 as noted) and is (still) cited in a handbook. The extent to which this source of information addresses forced convection, internal flow in ducts, is considered adequate because this topic is well known, as documented here. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 XVIII-4 October 2004 Qualification of the reference by Kern, D.Q. 1950 [DIRS 130111]. The referenced source by Kern is referenced by Perry et al. (1984 [DIRS 125806], general references for Section 10: Heat Transmission). The information from the source by Kern is reliable and qualified for the intended use because it is cited in a handbook. The extent to which this source of information addresses the linearization of radiative heat transfer and definitions is considered adequate because these topics are well known, as documented here. Qualification of the reference by Kuehn, T.H. and Goldstein, R.J. 1976 [DIRS 100675]. The referenced source by Kuehn and Goldstein is referenced in the handbook by Rohsenow et al. (1998 [DIRS 169241], reference number 163 in Chapter 4: Natural Convection). Also, it is referenced in the textbook by Incropera and Dewitt (1996 [DIRS 108184], Chapter 9, Free Convection, reference number 38). The information from the source by Kuehn and Goldstein is reliable and qualified for the intended use because it is based on experimental data, has been in the open literature for over three decades, is cited in a handbook and textbook on this topic, and thus is widely used in the standard work practices on the topic of natural convection. The extent to which this source of information addresses natural convection heat transfer correlations between concentric cylinders is considered adequate because the correlation is well known and based on experimental measurements, as documented here. Qualification of the reference by Kuehn, T.H. and Goldstein, R.J. 1978 [DIRS 130084]. The referenced source by Kuehn and Goldstein is referenced in the handbook by Rohsenow et al. (1998 [DIRS 169241], reference number 164 in Chapter 4: Natural Convection). The information from the source by Kuehn and Goldstein is reliable and qualified for the intended use because it is based on experimental data, has been in the open literature for over three decades, is cited in a handbook on this topic, and thus is widely used in the standard work practices on the topic of natural convection. The extent to which this source of information addresses natural convection heat transfer correlations between concentric cylinders is considered adequate because the correlation is well known and based on experimental measurements, as documented here. Qualification of the reference by Morgan, V.T. 1975 [DIRS 160791]. The referenced source by Morgan is referenced in the handbook by Rohsenow et al. (1998 [DIRS 169241], reference 198 in Chapter 4: Natural Convection). The information from the source by Morgan is reliable and qualified for the intended use because it is cited in a handbook on this topic, and widely used in the standard work practices on the topic of natural convection heat transfer. The extent to which this source addresses natural convection heat transfer is considered adequate because this topic is well known, as documented here. Qualification of the reference by Moyne, C.; Batsale, J.C.; Degiovanni, A; and Maillet, D. 1990 [DIRS 153164]. The referenced source by Moyne et al. is the source of experimental results for vapor diffusion and enhanced vapor diffusion. This particular reference is a summary paper of two previous publications, one by Azizi et al. (1988 [DIRS 154108]), and the other by Moyne et al (1988 [DIRS 154107]). These two publications are detailed descriptions of the experimental and theoretical approach of the thermal conductivity of wet porous media: experiments and theory. The cited reference and the two that precede it more than adequately demonstrate the information of interest, and this is experimental results. The extent to which this source of information addresses the experimental results of vapor diffusion and enhanced vapor Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 XVIII-5 October 2004 diffusion in porous media is adequate because of the detailed documentation available, as noted here. Qualification of the reference by Nagle, R.K. and Saff, E.B. 1994 [DIRS 100922]. The referenced source by Nagle and Saff is the source of mathematics describing the superposition principle. This source is corroborated by Weinberger (1965 [DIRS 163216], Chapter 2) and Zwillinger (1996 [DIRS 152179], Chapter 5). Both of these cited corroborating references discuss the superposition principle as indicated. The Nagle and Saff source is reliable and qualified for the intended use because the superposition principle is widely known and used in the standard work practice of solving differential equations. The extent to which this source of information addresses the superposition principle is considered adequate, as documented here. Qualification of information on the physical properties of air from Reid et al. 1977 [DIRS 130310]. The information used from Reid et al. pertains to the physical properties of air. The physical properties of interest are the compressibility factor (to demonstrated that air behaves as an ideal gas), viscosity and thermal conductivity. The reference by Reid et al. is referenced by Perry et al. (1984 [DIRS 125806]), in Section 3: Physical and Chemical Data, in the general references (on p. 3-5) and as reference number 196 (on p. 3-290). Thus the information on the physical properties of air from Reid et al. is considered reliable and qualified for intended use because Reid et al. is referenced by a handbook on the topic of physical properties. The extent to which this information on the physical properties of air address the properties of interest, applicability of the ideal gas law, viscosity and thermal conductivity, is adequate because this information is well known, as documented here. Qualification of the reference by Sutherland, W.A. and Kays, W.M. 1964 [DIRS 160789]. The referenced source by Sutherland and Kays is referenced in a handbook by Kays and Perkins (1973 [DIRS 160782], reference number 144 in Chapter 7: Internal Flow in Ducts). The information from the source by Sutherland and Kays is reliable and qualified for the intended use because it is a single-topic paper cited in a handbook and thus is widely used in the standard work practices on the topic of heat transfer for internal flow in ducts. The extent to which this source of information addresses this topic is considered adequate because the topic is well known, as documented here. Qualification of the information on the standard atmosphere from White 1986 [DIRS 111015]. The information on the standard atmosphere is atmospheric pressure at two elevations. This information can also be found in Perry et al. (1984 [DIRS 125806], Table 3-214). The pressure given by White for an elevation of 1000 m is 89,889 Pa, and from Perry et al., the pressure is 0.89876 bar. The conversion from bar to Newtons per square meter is to multiply bar by 1 × 105 (Perry et al. 1984 [DIRS 125806], Table 1-6). The conversion factor for Newtons per square meter to Pa (Pascal) is unity (Rohsenow et al. 1998 [DIRS 169241], Table 2.4). Therefore the information from White is considered established fact because it is corroborated in a handbook. The extent to which these data address the topic of interest is adequate because the standard atmosphere is well known, as documented here. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 XVIII-6 October 2004 INTENTIONALLY LEFT BLANK Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 October 2004 APPENDIX XIX CALCULATION OF TOTAL PRESSURE AT THE REPOSITORY ELEVATION Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 October 2004 Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 XIX-1 October 2004 This appendix estimates the total pressure at the elevation of the repository. The minimum and maximum elevations of the repository are 1,039 m and 1,107 m (BSC 2004 [DIRS 164519]). The median elevation is therefore 1,073 m. Atmospheric pressure at this elevation can be determined from the U.S. Standard Atmosphere (White 1986 [DIRS 111015], Figure 2.7 and Table A.6). The atmospheric pressure drops off nearly linearly up to a few thousand meters, as can be ascertained from examining Figure 2.7 in the cited reference. In the cited Table A.6 at elevations of 1,000 and 1,500 m, the pressures are 89,889 and 84,565 Pa. By linear interpolation, the atmospheric pressure at 1,073 m is 89,112 Pa, which is 0.879 atmosphere, using the conversion factors on p. xix of this report. Round this to 0.88 atmosphere. Ventilation Model and Analysis Report ANL-EBS-MD-000030 REV 04 XIX-2 October 2004 INTENTIONALLY LEFT BLANK