PRODUCED WATER COMPOSITIONAL DATABASE This version of the produced water database is a subset of a larger database originally provided to the USGS by GeoINFORMATION at the University of Oklahoma. The origin of the database is traced to the Petroleum Laboratory at Bartlesville, Oklahoma that was originally operated by the U.S. Bureau of Mines and subsequently by the Department of Energy. In the 1980s, the database was provided to the University of Oklahoma. The database as received by the USGS included 77,650 records and 102 fields. The review process and complete description of each field within the database are presented below. DESCRIPTION OF DATA REVIEW Four principal data-types were reviewed in detail: location, geologic setting, sample type and major ion composition. Review criteria were applied liberally to retain as many records as possible. Records that were retained include a location within a state and county (except for offshore wells), a reference to a geologic setting (basin or province), a sample type other than drilling, work-over or fluids open to external input, and contain consistent major ion chemical composition data. The resulting database contains 58,706 records that satisfy the criteria described below. Records in the database contain results from test holes as well as producing wells. Samples from the test wells were retained because they generally intersect geologic units with a history or potential of petroleum production. Although this database may be described to contain formation water compositions, the large number of samples collected from the petroleum production system removed from the wellhead is more consistent with a produced water designation. A few of the records are for water collected from water wells. These may have been collected to identify possible sources of water for enhanced recovery operations and were also retained in the database. Location Parameters that identify the geographic location of the sampled well include state, county, latitude and longitude, and oil and gas field name. These parameters were reviewed for consistency. Records that were not consistent and could not be corrected independently were deleted from the database. State and county names were corrected for consistency of spelling and location. Latitude and longitude of the sampled well were retained if they were originally reported in the database or added if the coordinates could be calculated using the public lands survey information. Approximately 20% of the records list field latitude and longitude as entered in the original database for the sample location. Oil and gas field names were corrected for consistency of names, field code and location by comparing with entries in the 1998 Field Code Master List database that is maintained by the Energy Information Agency of the Department of Energy (http://www.eia.doe.gov/pub/oil_gas/natural_gas/data_publications/field_code_master_list/historical/1998/ html/fcml_1998.html). Geologic Setting The geologic setting of the sample is described by the geologic basin (province), the name of the geologic unit from which the fluid was produced and the age of that geologic unit. Entries describing the geologic setting used in the original database were revised to conform to the nomenclature used by the American Association of Petroleum Geologists (AAPG). Geologic names and ages of the sampled units were compared with the USGS Lexicon of Geologic Names (hgmdb.usgs.gov/Geolex/geolex_home.html) and the stratigraphic charts presented in Gautier and others (1995). Names of geologic units varied and included groups, formations, and members. Where possible all three names were included. The position of the sample site relative to the USGS assessment provinces, identified in Gautier and others (1995), was added to the database for all records with a latitude and longitude or county location. All records for onshore wells list a geologic basin and 66% have a geologic unit and age. Sample type Water is collected by several methods and at various points during well completion and petroleum production. The method and location of sample collection can substantially affect the composition of the water sample because of mineral precipitation, degassing, addition of oxygen, contamination by drilling or workover fluids, water mixing, and evaporation. Originally more than 5000 different descriptors were used to describe the method/location of sample collection. These were consolidated into 18 general categories intended to indicate general sample types as described in Table 2. In general, samples collected from a drill stem test or production test yield the best results for formation water analyses. Samples collected from a wellhead are considered the best representation of produced water as they are likely to be water brought to the surface prior to treatment and separation. Chemical composition Review of the database focused on the major dissolved ions and total dissolved solids (TDS) because these values can be checked for internal consistency using simple calculations (Hitchon and Brulotte, 1994). The major dissolved species typically include: sodium, potassium, calcium, magnesium, chloride, bicarbonate, and sulfate. The sum of the ions should equal the TDS for a good water analysis. Records that have water analyses within 15% of the TDS were considered acceptable. The charge balance was also calculated as the sum of cations minus the sum of anions divided by the total ion content (all values in units of equivalents/liter). Records with a charge balance within 15% were considered acceptable. Application of these evaluations to this database is problematic in that sodium in most standard formation water analyses is determined by difference and therefore will likely result in good mass and charge balance. As described by Hitchon and Brulotte (1994), several other chemical parameters can assist in identifying poor analyses or contaminated samples. All samples with pH>9 and <4 and samples with potassium concentrations greater than five times the sodium content were deleted. These compositions are likely the result of contamination due to treatment of a well (Hitchon and Brulotte, 1995). Chemical compositions of the 58,706 records that were retained in the database (Fig. 1) meet the review criteria described above, but may not represent complete water analyses. Water analyses of oil and gas production waters may include 1) incomplete analyses, 2) the sum of sodium and potasssium as sodium only, 3) the sum of calcium and magnesium as calcium, or 4) no report of pH. Complete major ion chemical compositions (pH, Na+, Ca2+, Mg2+, K+, SO42-, Cl- and HCO3-) are given for only 7,722 records in the revised database. Most records lack results of a potassium analysis. For many preliminary assessments, incomplete analyses though not desirable may be suitable for many types of evaluation. Detailed water assessments, however, (particularly geochemical modeling), require complete analyses. FIELD NAMES AND DESCRIPTIONS 1. UNIQID -- Unique identifier (Long Integer). Numeric identifier used as primary key for each record. Original entries were text format. Format was changed to numeric. Values are as originally entered. (Null entries: 0) 2. STATECODE -- Code for the state (Text, 2). Two character designation assigned to conform to the FIPS code. Entry based on the state name in the STATE field after that field was completely reviewed. STATECODE was added during USGS review of the database. (Null entries: 0) 3. STATE -- Name of state (Text, 14). Full spelling of state name. Entry verified to be consistent with latitude, longitude, county and field name for non-null entries in those fields. Records with inconsistent entries among these parameters were removed from this version of the database. Offshore wells located in federally managed waters are also assigned to a state to indicate approximate location. (Null entries: 0) 4. REGDIST -- Regional district (Text, 4). Identifies the oil and gas district within Texas. Entries were verified in 2001 by consistency of REGDIST with COUNTY as defined by the Texas Railroad Commission (web site below). Verification and correction was conducted manually using a database query to list all unique combinations of counties in Texas and their associated REGDIST. Reference to Texas Railroad Commission designations available at: http://www.rrc.state.tx.us/divisions/og/information-data/ogicd.html. (Null entries: 44,657) 5. COUNTY -- County name (Text, 50). Name corrected for spelling using TIGER database. Original records listed multiple counties for some records. Corrections were made manually by comparing TIGER county listing with a table of unique county spellings for each state prepared by a query to the database. Discrepancies in spelling between TIGER and COUNTY were checked by manually comparing with county spellings listed in official state web sites. Entries were revised to make other location parameters (latitude and longitude, field name, township, range and section) consistent with entry. Consistency for records with LATITUDE and LONGITUDE was verified by plotting (ARCVIEW) the record locations on a county boundary basemap (ESRI). Records in which field name, county or latitude and longitude remained inconsistent after review were deleted. (Null entries: 0) 6. CNTYCODE -- Three digit code for the county (Text, 4). Corrected where necessary to be consistent with county FIPS code according to the TIGER database. Entries were reviewed manually using a listing of unique entries prepared by a database query of COUNTY and CNTYCODE entries. (Null entries: 0) 7. GEOBASIN -- Geologic basin (Text, 50). Basin name as originally entered in the database. Minor corrections were made for consistency of spelling. Corrections were made manually by preparing a table of unique GEOBASIN entries and comparing with established spellings of defined geologic basins. (Null entries: 125) 8. USGSPROV -- Geologic province consistent with USGS 1995 National Oil and Gas Assessment assignments (Text, 50). Valid records are limited to those entries with valid county designations. Province was assigned using a GIS overlay operation (ArcInfo IDENTITY command) w/ a national province coverage (available at http://energy.cr.usgs.gov/oilgas/noga/). For records lacking accurate latitude and longitude USGSPROV was assigned manually by comparing COUNTY designation with USGSPROV boundaries. (Null entries: 514 (equal to number of federally managed offshore wells; these areas were not included in the definition of USGS Provinces.)) . 9. USGSREG -- Geologic region consistent with USGS 1995 National Oil and Gas Assessment assignments (Text, 50). Region was assigned using a GIS overlay operation (ArcInfo IDENTITY command) w/ a national region coverage (available at http://energy.cr.usgs.gov/oilgas/noga/) For records lacking accurate latitude and longitude, USGSREG was assigned manually by comparing USGSPROV designation with USGSREG boundaries. Records for wells located in federally managed offshore wells were assigned as MMS wells consistent with agency jurisdiction for assessment. (Null entries: 0) 10. FIELD -- Oil and gas field (Text, 40). Field names in the original entries were checked against the 1998 Field Code Master List (FCML) maintained by the Energy Information Agency of DOE. Comparison of the FIELD entries with the FCML was conducted with a visual basic program written to make corrections to FIELD (for details see FLDFLAG). Many of the corrections were format and spelling corrections that were done manually. Deletion of unrecognized names was part of the correction. The county location listed for FIELD in FCML was also checked against COUNTY in each record to verify consistency. If inconsistent, the record was deleted. Minor changes in formatting relative to the listing in FCML were required for some field names. Entries with apostrophes ( ') were modified to remove the apostrophe so that search routines and supporting visual basic programs could aid in data review. Site of 1998 FCML http://www.eia.doe.gov/pub/oil_gas/natural_gas/data_publications/field_code_master_list/historical/1998/h tml/fcml_1998.html. (Null entries: 8,154) 11. FLDCODE -- Oil and gas field code (Text, 10). The numeric code for the field name as corrected in the above field. Numeric codes used were checked for consistency with the 1999 release of the EIA Field Code Master List (FCML) as described in FIELD and FLDFLAG. (Null entries: 12,238) 12. FLDFLAG -- Field name correction flag (Number, Integer). This field was inserted into the database by the USGS to record the type of correction made to the FIELD entries. Explanation of entries: 0, oil and gas FIELD name and FLDCODE are absent; 1, FLDCODE matches entry of the invalid field name database, FIELD is null; FIELD and FLDCODE corrected according to recommendations of the invalid field name database (part of FCML); 2, FIELD name is presented in the invalid fields database, FIELD and FLDCODE corrected to be consistent with correction recommended in invalid fields database; 3, FIELD name not in FCML but FLDCODE matches FCML entry, changes made to correct FIELD entry to be consistent with FLDCODE; 4, FIELD name changed to conform to change of alias name in FCML; 5, FIELD and FLDCODE are both valid entries in the FCML but are inconsistent, entries manually corrected to be consistent with COUNTY; 6, FIELD and FLDCODE are consistent and match FCML as entered; 7, FIELD is wildcat; 8, FIELD entry spelling or other editorial correction made to make name consistent with FLDCODE; 9, correction was made to spelling or other editorial correction to make FIELD consistent with entries in FCML, for these records FLDCODE was null but added to correspond to FIELD; Where resolution was not possible FIELD and FLDCODE set equal to null. (Null entries: 0) 13. SECTION -- Section (Text, 2). Section number as assigned by the public lands survey grid. Value as entered in the original database. (Null entries: 18,670) 14. TOWNSHIP -- Township (Text, 5). Township location as assigned by the public lands survey grid. Values are as entered in the original database. Minor corrections to format were made manually to standardize the listing of direction and number. (Null entries: 19,010) 15. RANGE -- Range (Text, 5). Range location as assigned by the public lands survey grid. Values are as entered in the original database. Minor corrections to format were made manually to standardize the listing of direction and number. (Null entries: 18,919) 16. LATITUDE -- Latitude location of the well (Number, Double). Entries derived from original well locations in the database, the latitude of the oil and gas field containing the well as listed in the original database, or calculated from a spatial calculator that determines a centroid latitude from the section, township and range coordinates. The spatial calculators were used for: Arizona, California, Colorado, Idaho, Iowa, Kansas, Montana, Nebraska, New Mexico, Oklahoma, Oregon, South Dakota, Utah, and Wyoming. Latitude for other states with township, range, and section information were not calculated. Conversions for Oklahoma were conducted using an online calculator available from Geo Information at the University of Oklahoma (http://www.geo.ou.edu/index.htm). Locations for Iowa were calculated using a spatial calculator available from the Iowa Department of Natural Resources (http://samuel.igsb.uiowa.edu/gettrs ). Locations for other states were calculated using TRS2LL a freeware program written in 1995 by Martin Wefald (http://www.geocities.com/jeremiahobrien/trs2ll.html; see also http://www.esg.montana.edu/gl/trs-data.html). (Null entries: 4,916) 17. LONGITUDE -- Longitude location of the well (Number, Double). Entries derived from original well locations in the database, the latitude of the oil and gas field containing the well as listed in the original database, or calculated from a spatial calculator that determines a centroid longitude from the section, township and range coordinates (PLS, public land survey). The spatial calculators were used for: Arizona, California, Colorado, Idaho, Iowa, Kansas, Montana, Nebraska, New Mexico, Oklahoma, Oregon, South Dakota, Utah, and Wyoming. Latitude for other states with township, range, and section information were not calculated. Conversions for Oklahoma were conducted using an online calculator available from Geo Information at the University of Oklahoma (http://www.geo.ou.edu/index.htm). Locations for Iowa were calculated using a spatial calculator available from the Iowa Department of Natural Resources (http://samuel.igsb.uiowa.edu/gettrs). Locations for other states were calculated using TRS2LL a freeware program written in 1995 by Martin Wefald (http://www.geocities.com/jeremiahobrien/trs2ll.html; see also http://www.esg.montana.edu/gl/trs-data.html). (Null entries: 4,916) 18. LOCGRADE -- Location grade (Text, 2). Values inserted into the database to record the reliability with which the LATITUDE and LONGITUDE of the sampled well are known. Values of LOCGRADE include: B, LATITUDE and LONGITUDE given in original database; C, LATITUDE and LONGITUDE calculated from township, range, and section location; D, LATITUDE and LONGITUDE are for the oil and gas field not for the individual well; values as entered in the original the database (assumed to be the field centroid); G, LATITUDE and LONGITUDE are unknown. (Null entries: 0) 19. API -- Well number assigned according to API crteria(Text, 12). American Petroleum Institute well number assigned in the original database. Entries are unchanged from the original database. (Null entries: 26,478) 20. WELLNAME -- Name of the sampled well (Text, 50). Entries are unchanged from the original database. (Null entries: 3,246) 21. COMPDATE -- Completion date (Date, MM/DD/YYYY). Date that the well was completed. Format changed from text (original database) to date. Entries in which month was absent were manually modified to have "01" as the month; those lacking day likewise are "01". Otherwise entries are as presented in the original database. (Null entries: 55,542) 22. SAMPFORM -- Sampled formation name (Text, 24). Geologic unit(s) from which the sample was produced (screen or perforated interval). Entries are unchanged from the original database. (Null entries: 4,699) 23.REVSAMPFORM -- Revised sample formation name (Text, 50). Geologic units from which the samples was produced. Corrections to SAMPFORM made to be consistent with stratigraphic columns in the USGS 1995 National Oil and Gas Assessment and the USGS Geologic Names Lexicon (http://ngmdb.usgs.gov/Geolex/geolex_home.html Corrections were made in 2000 by manually by listing the unique SAMPFORM entries for each state and comparing those entries to the references listed above. Changes were made using an update query on the database to replace existing entries with those more consistent with standard stratigraphic nomenclature as defined in the above references. Entries are presented in an effort to depict the hierarchy of units. "/" identifies greater unit to the left (group, formation, member) while the unit on the right is contained within the unit on the left (formation, member, unit, bed etc.) A"-" identifies units that are either equivalent in rank or not part of the same higher units. Null entries indicate an null entry in SAMPFORM or the entry in SAMPFORM could not be directly related to the reference listings of geologic units. (Null entries: 17,252) 24. #FORMS -- Number of producing formations (Text, 2). USGS entered parameter indicating the number of formation produced by each well (S, single; M, multiple; ? unknown). Entries were made manually and are based on the entries in RevSAMPFORM. (Null entries: 0) 25. SAMPFORMAGE -- Geologic age of the sample formation (Text, 3). Three digit code assigned to correspond to the age of sample formation. Entry as presented in the original database. (Null entries: 28,343) 26. REVSAMPFORMAGE -- Geologic age of the sample formation (Text, 3) revised to conform to Table 1 (attached). Corrections to SAMPFORM made to be consistent with stratigraphic columns in the USGS 1995 National Oil and Gas Assessment and the USGS Geologic Names Lexicon (http://ngmdb.usgs.gov/Geolex/geolex_home.html. Corrections to this field were completed in 2000. (Null entries: 9,515) 27. UPPERDEPTH -- Depth of the top of the sampled interval (Number, Long integer). Entry as originally entered in the database, with change of format from text to number. (Null entries: 10,853) 28. LOWERDEPTH -- Depth of the bottom of the sampled interval (Number, Long integer). Entry as originally entered in the database, with change of format from text to number. (Null entries: 18,735) 29. METHOD -- Location/method of where the sample was collected from the sampled well (Text, 50). Entry as contained in the original database. (Null entries: 17,838) 30. REVMETHOD -- Revised location/method of sample collection site (Text, 20). USGS added field to simplify entries in METHOD to a simplified list of methods including: wellhead, DST, separator, heater/treater, tank, bailer, swab, bomb, openhole, production, production test, formation test, bleeder, and unknown. A listing of the more common METHOD entries and their assigned REVMETHOD entries are listed in Table 2. (Null entries: 0) 31. MethGRADE -- USGS assigned grade assigned to the location/method of sample collection (Text, 1). Grades intended to reflect the relative ability of the sample to represent the composition of produced water. B: well head, production; C: drill Stem, formation or production tests; D: swab, bailer; E: separator, heater/treater, tank, F: unknown. MethGRADE entries and their corresponding REVMETHOD and METHOD are listed in Table 2. (Null entries: 0) 32. SAMPDATE -- Date sample was collected (Date, MM/DD/YYYY). Date that the water sample was collected. Format changed from text (original database) to date. Entries in which month was absent are listed as "01"; those lacking day likewise are "01". Otherwise entries are as presented in the original database. (Null entries: 16,025) 33. pH -- pH of the water sample. pH was converted from character to numeric format and corrections were made of the type of entry. Records with pH values less than 4 and greater than 9 were excluded from this compilation because values outside the range are typically a product of contamination by chemicals used during drilling and workovers. (Null entries: 18,441) 34. UNITS -- Concentration units mg/L or ppm. Units are as used in the original version of the database. (Null entries: 0) 35. BICARB -- Concentration of bicarbonate ion in UNITS. Converted from character to numeric format. Values of the entries could not be verified independently. Recoding of some entries was according to the following scheme: values of -3 replace original character entries of "ND", "NA" or null entries; -2 replace original character entries of "none", "negative", or "nil", -1 replaces original character entries of "trace" and "minor". (Null entries: 0) 36. CALCIUM -- Concentration of calcium ion in UNITS. Converted from character to numeric format. Values of the entries could not be verified independently. Recoding of some entries was according to the following scheme: values of -3 replace original character entries of "ND", "NA" or null entries; -2 replace original character entries of "none", "negative", or "nil", -1 replaces original character entries of "trace" and "minor".. (Null entries: 0) 37. CHLORIDE -- Concentration of chloride ion in UNITS. Converted from character to numeric format. Values of the entries could not be verified independently. Recoding of some entries was according to the following scheme: values of -3 replace original character entries of "ND", "NA" or null entries; -2 replace original character entries of "none", "negative", or "nil", -1 replaces original character entries of "trace" and "minor". (Null entries: 0) 38. MAGNESIUM -- Concentration of magnesium ion in UNITS. Converted from character to numeric format. Values of the entries could not be verified independently. Recoding of some entries was according to the following scheme: values of -3 replace original character entries of "ND", "NA" or null entries; -2 replace original character entries of "none", "negative", or "nil", -1 replaces original character entries of "trace" and "minor". (Null entries: 0) 39. POTASSIUM -- Concentration of potassium ion in UNITS. Converted from character to numeric format. Values of the entries could not be verified independently. Recoding of some entries was according to the following scheme: values of -3 replace original character entries of "ND", "NA" or null entries; -2 replace original character entries of "none", "negative", or "nil", -1 replaces original character entries of "trace" and "minor". (Null entries: 0) 40. POT-SODIUM -- Concentration of the sum of potassium and sodium ions in UNITS. Converted from character to numeric format. Values of the entries could not be verified independently. Recoding of some entries was according to the following scheme: values of -3 replace original character entries of "ND", "NA" or null entries; -2 replace original character entries of "none", "negative", or "nil", -1 replaces original character entries of "trace" and "minor" (Null entries: 0) 41. SODIUM -- Concentration of sodium ion in UNITS. Converted from character to numeric format. Values of the entries could not be verified independently. Recoding of some entries was according to the following scheme: values of -3 replace original character entries of "ND", "NA" or null entries; -2 replace original character entries of "none", "negative", or "nil", -1 replaces original character entries of "trace" and "minor". (Null entries: 0) 42. SULFATE -- Concentration of sulfate ion in UNITS. Converted from character to numeric format. Values of the entries could not be verified independently. Recoding of some entries was according to the following scheme: values of -3 replace original character entries of "ND", "NA" or null entries; -2 replace original character entries of "none", "negative", or "nil", -1 replaces original character entries of "trace" and "minor". (Null entries: 0) 43. TDS -- Concentration of total dissolved solids in UNITS. Converted from character to numeric format. Values of the entries could not be verified independently but were checked for consistency with the sum of major ions in the MASSBAL field. Entry of –3 indicates records in which TDS was not given and presumably not determined. (Null entries: 0) 44. CHARGBAL -- Calculated value of charge balance (Number, single precision). Calculated as: (sum of equivalents of cations – sum of equivalents of anions) divided by the sum of the equivalents of all ions. Most records have good charge and mass balance due to the determination of sodium by difference. Values outside the range of 0.15 to –0.15 were deleted. (Null entries: 0) 45. MASSBAL -- Calculated value of mass balance (Number, single precision). Calculated as the sum of major ion species (sodium, potassium, potsodium (as sodium), calcium, magnesium, chloride, sulfate, and bicarbonate) divided by the reported total dissolved solids. Values outside the range of 0.85 to 1.15 were deleted. Values were not calculated for TDS entries of –3. (Null entries: 39) 46. SOURCE -- Identifies the source of the record. Typically this lists the name of the petroleum company or the analytical laboratory responsible for the analytical data. These entries are unchanged from the original database. (Null entries: 29,882) References Gautier, D.L., Dolton, G.L., Takahashi, K.I., and Varnes, K.L., 1995, National Assessment of Oil and Gas Resources—Results, Methodology, and Supporting Data, U.S. Geological Survey Digital Data Series DDS-30 Release 2. Hitchon, Brian and Brulotte, M., 1994, Culling criteria for "standard" formation water analyses: Applied Geochemistry v. 9, p. 637-645. 1 7 02/25/02