J. I. Mickalonis
Westinghouse Savannah River Company
Aiken, SC 29808

This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency thereof, nor any of their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process or service by trade name, trademark, manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof.

This report has been reproduced directly from the best available copy.

Available for sale to the public, in paper, from:  U.S. Department of Commerce, National Technical Information Service, 5285 Port Royal Road, Springfield, VA 22161,  phone: (800) 553-6847,  fax: (703) 605-6900,  email:  orders@ntis.fedworld.gov   online ordering:  http://www.ntis.gov/support/ordering.htm

Available electronically at  http://www.osti.gov/bridge/

Available for a processing fee to U.S. Department of Energy and its contractors, in paper, from: U.S. Department of Energy, Office of Scientific and Technical Information, P.O. Box 62, Oak Ridge, TN 37831-0062,  phone: (865 ) 576-8401,  fax: (865) 576-5728,  email:  reports@adonis.osti.gov

Acidic degraded moderator was temporarily stored in Tuff Tanksä located in L-area. The moderator characteristics included a D₂O content of 5.02-5.33 %, a pH of 1.25-1.31, a conductivity of 29,300-31,200 m mhos/cm, tritium activity of 114-141 m Ci/mL, and levels of approximately 6000 ppm for chloride and 500 ppm for chromium. The compatibility of the degraded with AISI Type 304L stainless steel (304L) was investigated in this study. Following ASTM standard practice, coupon immersion tests were conducted in both treated and untreated and moderator. Treatment included the addition of either a 40 wt % NaOH solution, distilled water to serially dilute the chloride, or concentrated nitric acid to increase the nitrate concentration.

Type 304L stainless steel exposed to the Tuff Tank moderator was found from these tests to:

• have a general corrosion rate of less than 5 mils per year (mpy) for 304L plate, which bounds that of the 304L storage drum,
  
  
• passivate at chloride concentrations up to 5000 ppm for 304L sheet,
  
  
• resist corrosion for nitrate/chloride ratios ranging from 0.1 to 1000,
  
  
• and be susceptible to crevice corrosion.

Based on these test results, storage of the degraded moderator in 55-gallon 304L drums (0.065" thick) would not cause failure by general corrosion for up to 5+ years storage.

The chloride concentration, [Cl], in the degraded moderator has been measured up to 6000 ppm. The potential or risk for aggressive localized attack of 304L increases with [Cl] concentration. A qualitative range is as follows [7]:

· [Cl-] < 150 ppm, excellent resistance, very low risk
· 150 < [Cl-] < 350 ppm, good resistance, low risk
· 350 < [Cl-] < 600 ppm, reasonable resistance, medium risk

The degraded moderator should be treated to reduce the chloride concentration to reduce the potential for localized corrosion and the risk for a leakage failure of the drum. A good practice would be to reduce the [Cl] concentration to below 250 ppm, the site standard for stainless steel. The monitoring program of these drums should continue as part of the Moderator Storage Program. More frequent inspections is a good practice as storage time increases. Long-term surveillance testing with creviced samples in actual degraded moderator should be performed in conjunction with the monitoring program, especially if the intended storage time is to exceed several years or chloride concentrations exceed 250 ppm.

Spent Fuel and Storage Division (SFSD) collected low-level waste in a 4000-gallon stainless steel storage trailer which was 15 years old. After several months of storage the trailer leaked near a weld in the lower section of the trailer. The waste was transferred to 13 polyethylene Tuff-Tanksä for temporary storage. The waste was evaluated and found to have the following characteristics: a D₂O content of 5.02-5.33 %, a pH of 1.25-1.31, a conductivity of 29,300-31,200 m mhos/cm, tritium activity of 114-141 m Ci_/mL, and levels of approximately 6000 ppm for chloride and 500 ppm for chromium. The waste was also found to have an organic or oil phase. Immediately after the leak occurred, the waste was classified as a RCRA hazardous waste which required treatment for permanent storage within 90 days. Based on the unexpected D₂O content, the waste was reclassified as degraded moderator which was to be treated for storage in stainless steel drums. This classification does not have the 90-day time limit for treatment. A proposed option was the neutralization of the degraded moderator with subsequent storage in 304L stainless steel 55-gallon drums.

The degraded moderator was considered to be highly corrosive because the storage trailer failed in a short period of time and the characteristics of the degraded moderator were suspect, specifically the high chloride levels and the low pH. The high chromium levels of the degraded moderator also indicated that the stainless steel could have corroded. The Materials Technology Section (MTS) of the Savannah River Technology Center was contacted initially to test the untreated degraded moderator and, more importantly at that time, a neutralized moderator to determine corrosion rates for stainless steel. Subsequent accelerated tests were requested for the Tuff Tank moderator to determine if both a critical chloride level existed below which corrosion was considered benign and a nitrate/chloride ratio above which corrosion was inhibited.

The experimental approach for this study was to use short-term coupon immersion testing to address corrosion concerns about the Tuff Tank moderator because of the fast response requested by the customer. Short-term testing has limitations since long-term corrosion rates can differ. Testing was conducted in three phases. The objective of the initial phase was to determine a corrosion rate for 304L in both untreated and neutralized Tuff Tank moderator. The second phase objective was to determine if a chloride concentration limit existed at which the degradation of 304L would be detrimental to 55-gallon storage drums. The final phase objective was to determine a nitrate/chloride ion ratio where sufficient nitrates were present to inhibit the corrosive effect of chlorides.

The "Conduct of Research & Development – Savannah River Technology Center" was followed prior to initiation of testing [1]. All information was recorded in laboratory notebook WSRC-NB-94-321 [2]. Coupon immersion testing was performed following the ASTM standard practice G31-72 [3].

Several series of tests were performed to assess the compatibility of 304L in Tuff Tank moderator. The initial testing was conducted in untreated degraded moderator at room temperature and 80 ° C and in neutralized moderator at 80 ° C. Glass test vessels with temperature-controlled hot plates were used for the heated tests. The room temperature test was conducted in a 500-ml polyethylene bottle. The moderator was neutralized to a target pH of 5-6 by the addition of 40 wt % NaOH solution. The coupons were welded 304 L stainless steel (1" or 0.75" ´ 2" ´ 0.125"). Prior to immersion, the coupons were cleaned with ethyl alcohol and weighed. Two coupons were placed in each test; one coupon was totally immersed and one was partially immersed to form a moderator/air interfacial zone. The degraded moderator used for testing was taken from samples sent by SFSD to the SRTC/Analytical Development Section for chemical analysis.

Most coupons were removed after one week of exposure. Two coupons were removed at four days and then placed back into test for a total time of one week. When removed the coupons were cleaned by dipping in acetone and ethyl alcohol and then weighed. The coupons, which were exposed to the untreated degraded moderator, were harder to clean than those exposed to the neutralized moderator. This difference was due to the change in the degraded moderator characteristics upon neutralizing. The untreated degraded moderator had fine suspended particles, which settled over time, a dark bluish color, and an immiscible phase. The neutralized moderator completely precipitated out the suspended particles, which resulted in a two-phase system of a clear solution and a dark fine sludge.

This first series was followed by a second test series in which coupon tests were conducted at room temperature to substantiate the test results. The tests were conducted using the same moderator solutions and coupons to minimize waste generation from these tests. The coupons were thoroughly cleaned and weighed prior to use. Two coupons were again immersed in each moderator, one completely and the other partially. The test was conducted for six days.

In the final test series, a 304L U-bend and a 304L flat coupon with a crevice assembly were used to characterize the degraded moderator for stress corrosion cracking (SCC) and crevice corrosion. The crevice assembly is a Teflonä fixture that forms multiple crevices with the coupon. The 304L U-bend was exposed in the untreated degraded moderator for five days, two days at 40 ° C and three days at room temperature. Crevice coupons were exposed for two weeks in both the neutralized and untreated moderator at room temperature. Similar procedures were used for cleaning and weighing as described above.

The chloride limit was investigated with coupon immersion tests performed in a set of serially diluted Tuff Tank moderator solutions. The degraded moderator used for these test was taken from Tuff Tank D0253. Analytical results for this degraded moderator were: 5.02 mol% D₂O, 29400 m mhos/cm, 1.28 pH and 116 m Ci/mL, and 519.4 ppm Cr.

The chloride concentration of the initial Tuff Tank moderator was 4302 ppm, which was taken from a different sample than used in the previous testing. The dilution series was: 1/1, 1/10, 1/100, 1/1000, which corresponds to chloride concentrations of 4300 ppm, 430 ppm, 43 ppm, and 4 ppm, respectively. For each dilution level, 250 mL of solution were set up in two polyethylene bottles. The volumes of degraded moderator and distilled water used in each bottle are shown in Table 1.

Other constituents including the tritium in the degraded moderator were also diluted by magnitudes of ten. Changes in the pH were not adjusted; the pH ranged from 1.4 for the 1/1 solution to 4.0 for the 1/0.001 solution. All testing was conducted at room temperature.

The test coupons were made of 304L sheet, which was 0.032" thick. Two sheets had been welded together by gas tungsten arc with AISI Type 308L stainless steel filler metal. Coupons were cut from the weld such that the weld ran horizontally across the width of the coupon. Coupons were sectioned from regions of both good and bad welds. Additional coupons were sectioned from non-welded portions of the sheet. Final coupon dimensions were 1" ´ 2" ´ 0.032".

Prior to testing the coupons were deburred, cleaned with acetone and ethyl alcohol, and weighed. After weighing, a Teflonä tape was attached and had the unique identification number for the coupon. The tape was used to suspend the coupon in the solution.

For each serial dilution, one bottle contained a good and a bad weld coupon. The other bottle contained two non-welded coupons. One non-welded and both welded coupons were weighed weekly to calculate a time-dependent weight change. The other non-welded coupon was exposed continuously for one month prior to a weight measurement. The coupons were visually examined after weighing to assess for any degradation.

For weight measurements, the coupons were removed singularly from the bottle. They were cleaned by sequentially soaking in acetone and ethyl alcohol. Coupons were thoroughly wiped with a clean rag. To meet radiological control requirements, the coupons were weighed in plastic bags, whose weights were subtracted from the measured coupon weights. The weight of the bag added a degree of error to the data since new bags were required each time. Recorded weights were the average of two measured values.

The nitrate/chloride ratio experiments were set up similar to those for the determination of the chloride limit. The difference was that the serially diluted solutions were prepared with nitric acid instead of distilled water. The nitrate/chloride ratios were based on ion chromatograph measurements of the chloride and nitrate concentrations, which for the undiluted moderator were 3824 and 388 ppm, respectively. The tested ratios were 0.1/1, 100/1, 500/1, 1000/1 with chloride quantities of 3824, 2756, 1301, and 784 ppm, respectively. For each dilution level, solutions of 250 mL were set up in two polyethylene bottles. The volumes of degraded moderator and concentrated nitric acid used for these tests are shown in Table 2.

All tests were conducted at room temperature. The coupons and procedures for preparation and weighing were the same as those described above for the chloride limit determination.

Weight losses were used to evaluate the corrosion resistance of 304L and to calculate a rate for general corrosion. The equation used to calculate the rate (CR) was [4]

CR [mpy] = (3.45 ´ 10⁶)(W) / (A)(T)(D) ,

where W is weight loss (grams), A is surface area (cm²), T is time (hours), and D is the density (g/cm³), which for 304L is 7.94 g/cm³. The area for the completely immersed coupons was either 30.65 or 23.81 cm² and for the partially immersed coupon was 17.1 or 14.3 cm².

The corrosion rates from the two tests are shown in Table 3. The data entries that are shaded were the results from the second tests. As can be seen from a comparison of the data, corrosion rates in the neutralized moderator were greater than those in the untreated moderator. This difference is attributed to the change in solution pH. The oxide on stainless steel is better maintained in the more acidic moderator, so the chlorides would be less effective breaking it down. The rate increased slightly with time as observed for the neutralized degraded moderator at 80 ° C. Temperature had a more significant effect on increasing rates; temperature increases the kinetics of the corrosion process. The rates for the second series dropped from that of the first. This result may have occurred due to the build up of an oxide layer on the coupons since they were the same coupons used in the first tests.

The coupons from the two test series did not show any visual signs of degradation. Although there was some discoloration from the degraded moderator, the coupons still had a metallic finish and the initial grinding marks were distinct. The partially immersed samples had an interfacial line, but did not show any significant corrosion.

The high chloride concentration of the degraded moderator was a concern for the occurrence of either SCC or crevice corrosion in 304L 55-gallon drums that had either poor welds, manufacturing flaws, or degradation effects from previous use. The U-bend, which is a stressed strip of metal, was used to evaluate for SCC. The U-bend showed no significant corrosion and

* I – completely immersed, P – partially immersed

did not crack. Although a U-bend is quick and easy to use, the disadvantage of using a U-bend is that a no cracking result does not eliminate SCC as a possible degradation mode since a longer exposure could eventually lead to cracking. The present result indicated that SCC is not likely for short exposures. Crevice corrosion was observed in one location of the coupon exposed to the neutralized moderator. Crevice corrosion, therefore, may occur. Further testing would be required to obtain quantitative results.

All the coupons had a final weight gain indicating that the coupons passivated after exposure to the serially diluted Tuff Tank moderator. This passivation or oxidation aids in protecting the stainless steel from corrosion by the growth of an oxide film. The data are summarized in Table 4, which show the initial and final weights and a calculated corrosion rate. Note that all the corrosion rates are negative indicating the development of a passive film.

Figure 1 shows the weekly weights at all dilutions for the non-weld coupons. The weight gains were not dependent on the dilution. The observed trend is similar to those observed also for both the good and bad weld coupons. Generally, the coupons experienced a small initial weight loss, which subsequently reversed with weight gains. Corrosion and staining were not detected visually during the exposure.

The 304L corrosion characteristics were variable in this series of solutions as shown by the data presented in Table 5. There was no consistent trend of either oxidation or corrosion of any of the coupon types in a given solution or for a given coupon type in any of the solutions. For example, in the 0.1/1 solution, which was the undiluted solution, oxidation and corrosion were both observed. Oxidation was indicated by a negative corrosion rate and corrosion was indicated by a positive corrosion rate as shown in the table. Again, oxidation resulted in the growth of a protective oxide film. This behavior was characteristic for all the solutions. Also, any coupon type, such as non-weld – week, did not show a constant behavior of oxidation or corrosion for all the solutions. Figure 2 shows the weekly weight measurements for all the solutions of the non-weld – week coupon. A visual evaluation of the coupons showed that there were only slight changes in surface characteristics.

The corrosion testing conducted during this study were short term with the longest exposure period of one month. Short-term testing was used to meet the customer need of a fast response. Short-term corrosion results provide an estimate of the anticipated corrosion behavior. However, conditions from extended periods or initiation of corrosion processes may not develop during short term testing, such as the formation of crevices or growth of oxide layers. Conservative approximations are used generally to minimize this aspect of short-term testing.

The corrosion rate results from the first phase of this study are a conservative estimate for the corrosion rate of 304L in the Tuff Tank moderator. The coupons were standard laboratory 304L test coupons which are characteristic of plate material. The 55-gallon storage drums are made of sheet material, which has a nominal thickness of 0.065" [5]. Surface characteristics of sheet and plate material differ due in part to surface roughness. The rougher plate material is more susceptible to corrosion than sheet, thereby, providing a conservative corrosion rate estimate [6,9]. For ambient storage conditions, the corrosion rate of 304L in Tuff Tank moderator may be expected to vary between 0.5 to 7.0 mils per year.

Corrosion processes are affected by many factors which when vary even slightly can have significant effects on corrosion. Primary factors include the condition of the surface oxide on the material of interest, the oxygen content of the solution, temperature, and the solution composition. The variability observed in this test as for the ambient storage conditions are indicative of changes in these factors. The effect of changes in the solution composition are clearly shown by the test at 80 ° C. The corrosion rate of 304L in the untreated degraded moderator was 4.0 mils per year while in the neutralized degraded moderator the rate varies between 6.4 to 14.8 mils per year.

The chloride limit and nitrate/chloride ratio testing was performed with an objective to obtain actual or realistic limits. Therefore, the test sample was changed to sheet coupons. As explained above, sheet is more resistant to corrosion. The test results for the chloride limit and the nitrate/chloride ratio are indicative of this change. The 304L sheet passivated by the growth of an oxide layer, which experimentally is observed as an increase in coupon weight over the exposure period.

The changes in weights measured during the nitrate/chloride test (Table 5) were small, much smaller that those observed during the chloride limit test. The lower weights could be attributed to the more aggressive conditions in these highly acidic solutions except for the undiluted degraded moderator. The difference may also be attributed to differences in the degraded moderator source or test coupons. The results do show, however, that for this test the range of nitrates and chlorides were not aggressive for the stainless steel. These results indicated that attempts to inhibit the Tuff Tank moderator by the addition of nitric acid are probably unnecessary.

The undiluted degraded moderator was anticipated to be the most aggressive or corrosive solution of those tested because it had the highest chloride concentration. Corrosion was not observed on the sheet coupons. However, since general and crevice corrosions were observed on the plate coupons under similar conditions a critical chloride limit does exist. For storage in 55-gallon drums, crevice corrosion is the primary mode of concern. From the literature, the crevice corrosion resistance of 304L to chloride concentration ([Cl^-]) can be categorized as follows: exceptional for [Cl^-] < 150 ppm; very good for 150 < [Cl^-] < 350 ppm; reasonable for 350 < [Cl^-] < 600 ppm [7].

The Savannah River Site has an accepted standard limit of 250 ppm Cl^- for various items in contact with 304 and 304L stainless steel [8]. This standard was initially established for tapes and markers in contact with the stainless steel piping of the reactors. However, the standard has been applied prudently when possible for general service for 304L stainless steel. Therefore, based on these test results, accepted site wide standards, and prudent engineering judgement, the recommended chloride limit for storage of the Tuff Tank moderator in 55-gallon 304L drum is 250 ppm.

At a chloride concentration of 250 ppm, the crevice corrosion resistance of 304L is considered very good. Storage up to one year for the stainless steel drums should avoid localized leakage failures. A monitoring program for the drums is also highly recommended, especially if storage time exceeds one year. Several factors affect the time to leak and cannot be accurately predicted. These factors include the initial condition of the drum, actual moderator composition in the drum, and storage conditions.

Long-term surveillance testing is highly recommended especially if the intended if storage time is to exceed several years. Better estimates of time to leak may be made by conducting long-term testing with the Tuff Tank moderator using coupons with crevice assemblies. These coupons could either be placed directly into the drum and monitored during the surveillance inspections or set up in a laboratory and monitored at an established frequency. The former approach would yield better results since testing occurs in the actual environment, although the later approach would have better test control.

The corrosion tests performed with the Tuff Tank moderator have shown that 304L stainless steel is fairly resistive to corrosion depending on product form. Corrosion tests consisting of coupon immersion tests were conducted in both untreated and modified degraded moderator. The test results have shown that for the Tuff Tank moderator:

• 304L plate had an average general corrosion rates at room temperature less than 5 mpy
  
  
• 304L sheet, which is the material of construction for the 55-gallon storage drums, passivated at chloride concentrations up to approximately 5000 ppm
  
  
• 304L sheet was resistant to corrosion at nitrate/chloride ratios ranging from 0.1 to 1000
  
  
• Neutralized degraded moderator (pH 4-5) was more aggressive than the untreated moderator
  
  
• Crevice corrosion is a possibility in the degraded moderator

Based on these test results, storage of the degraded moderator in 55-gallon 304L drums (0.065" thick) would not cause failure by general corrosion for up to 5+ years storage.

The chloride concentration, [Cl], in the degraded moderator has been measured up to 6000 ppm. The potential or risk for aggressive localized attack of 304L increases with [Cl] concentration. A qualitative range is as follows [7]:

· [Cl-] < 150 ppm, excellent resistance, very low risk
· 150 < [Cl-] < 350 ppm, good resistance, low risk
· 350 < [Cl-] < 600 ppm, reasonable resistance, medium risk

The degraded moderator should be treated to reduce the chloride concentration to reduce the potential for localized corrosion and the risk for a leakage failure of the drum. A good practice would be to reduce the [Cl] concentration to below 250 ppm, the site standard for stainless steel. The monitoring program of these drums should continue as part of the Moderator Storage Program. More frequent inspections is a good practice as storage time increases. Long-term surveillance testing with creviced samples in actual degraded moderator should be performed in conjunction with the monitoring program, especially if the intended storage time is to exceed several years or chloride concentrations exceed 250 ppm.

The author would like to express his appreciation for the excellent laboratory work performed by Karen Hicks and radiological oversight provided by Matt Weigle. Helpful discussions were held with Mac Louthan and Bob Sindelar. The assistance of SFSD personnel in performing this work is also appreciated.

1. "Conduct of Research & Development – Savannah River Technology Center (U)", WSRC-IM-97-00024, Rev. 1, November 30, 1998.
2. "Corrosion Study of Lawrence Slurry Pump Material (U)," WSRC-NB-99-0233, 1999.
3. ASTM G31-72 (Reapproved 1995), "Standard Practice for Laboratory Immersion Corrosion Testing of Metals", American Society For Testing and Materials, Philadelphia, PA, 1999.
4. R. Baboian (ed.), Corrosion Tests and Standards, American Society For Testing and Materials, Philadelphia, PA, 1995.
5. DPSOL 208-Q-101, Rev. 2, SRP Engineering Standard and Specification, 55-Gallon Stainless Steel Drum, 2/5/76.
6. G. Hultquist and C. Leygraf, Corrosion, Vol. 36, p. 134, 1980.
7. J. W. Oldfield, International Materials Reviews, Vol. 32 (3), p. 153, 1984.
8. Engineering Standard # 05951, Rev. 1, SRS Engineering Standards Manual, WSRC-TM-95-1, 4/26/99.
9. D. Wallinder et al., Corrosion Science, Vol. 41 (2), pp 275-89, 1999.