&EPA EPA821-R-09-007 Technical Support Document for the Annual Review of Existing Effluent Guidelines and Identification of Potential New Point Source Categories U.S. Environmental Protection Agency Engineering and Analysis Division Office of Water 1200 Pennsylvania Avenue, NW Washington, D.C. 20460 October 2009 ------- CONTENTS Page 1. INTRODUCTION 1-1 1.1 Introduction References 1-2 2. DEVELOPMENT OF TRIRELEASES2007 2-1 2.1 TRI 2-1 2.1.1 Utility of TRI 2-3 2.1.2 Constraints and Limitations of TRI 2-3 2.2 Overview of TRI Databases 2-4 2.3 TRIRawData2007 2-5 2.4 TRICalculations2007 2-6 2.4.1 Modifications to TRI-Reported Data 2-8 2.4.2 POTW Removals 2-9 2.4.3 TWFs 2-9 2.4.4 Metal Compounds 2-22 2.4.5 Determination of "Basis of Estimate" of Reported TRI Releases 2-22 2.5 TRIReleases2007 2-23 2.5.1 NAICS/Point Source Category Crosswalk 2-24 2.5.2 Development of 2007 TRI Rankings 2-24 2.6 Results of the Preliminary Analysis of the TRIReleases200 7 Database 2-25 2.6.1 Metals Analysis 2-27 2.7 Development of TRIReleases2007 References 2-31 3. DMRLOADS2007: DEVELOPMENT AND CATEGORY RANKINGS 3-1 3.1 Overview of DMRLoads2007 3-1 3.1.1 NPDES Permitting and Reporting Requirements 3-1 3.1.2 Overview of PCS and ICIS-NPDES 3-3 3.1.3 PCS and ICIS-NPDES Data Structure 3-5 3.2 DMRLoads2007: Database Development and Methodology 3-7 3.2.1 Data Sources used in the Development of DMRLoads2007 3-7 3.2.2 PCSLoadCalculator2007 3-9 3.2.3 ICIS-NPDES Pollutant Loading Tool 3-25 3.2.4 DMRLoadsAnalysis2007 3-42 3.2.5 DMRNutrients2007 3-50 3.2.6 DMRLoads2007 3-51 3.2.7 Database Corrections 3-52 3.3 Results of the Preliminary Analysis 3-59 3.4 Data Quality Review 3-59 3.4.1 Mercury Discharges Reported Using PRAM 50092 3-65 3.4.2 Facility Reviews 3-65 3.5 DMRLoads2007 References 3-72 4. IDENTIFICATION OF POINT SOURCE CATEGORIES 4-1 4.1 Background on NAIC Sand SIC Codes 4-2 4.2 SIC Code to Point Source Category Crosswalk 4-3 i ------- CONTENTS (Continued) Page 4.2.1 SIC Codes Related to Existing Point Source Categories 4-4 4.3 NAICS Code to Point Source Category Crosswalk 4-10 4.3.1 NAICS Codes Related to Existing Point Source Categories 4-11 4.4 Potential New Point Source Categories 4-20 4.4.1 Direct Discharges 4-21 4.4.2 Indirect Discharges 4-21 4.5 Identification of Point Source Category References 4-21 5. Toxic WEIGHTING FACTORS 5-1 5.1 TWF Background and Development 5-1 5.2 New Toxic Weighting Factors Developed During the 2009 Annual Review 5-1 5.3 Chemicals without Toxic Weighting Factors 5-2 5.4 Toxic Weighting Factor References 5-8 6. QUALITY REVIEW 6-1 6.1 Overview of Quality Review Steps 6-1 6.1.1 Completeness Checks 6-2 6.1.2 Accuracy Checks 6-3 6.1.3 Reasonableness Checks 6-3 6.2 Quality Review of the DMRLoads200 7 Database 6-4 6.3 Quality Review of the TRIReleases2007 Database 6-5 6.4 Quality Review References 6-5 7. RESULTS OF 2009 SCREENING-LEVEL ANALYSIS 7-6 7.1 Preliminary Results of the Screening-Level Review 7-6 7.2 Prioritization of Categories 7-10 7.2.1 Categories for Which EPA is Currently Developing or Revising ELGs7-10 7.2.2 Categories for Which EPA Recently Promulgated or Revised ELGs ... 7-11 7.2.3 Discharges Not Categorizable 7-11 7.2.4 Categories with One Facility Dominating the TWPE 7-12 7.2.5 Combining the Final DMR and TRI Rankings 7-12 7.3 Identification of Categories With Existing Effluent Guidelines for Further Review 7-14 7.4 Results of 2009 Screening-Level Analysis References 7-18 11 ------- LIST OF TABLES Page 2-1 TRI 2007 Tables Downloaded from EPA 2-5 2-2 Tables Imported or Created in TRICalculations2007 2-6 2-3 TWF Modifications 2-10 2-4 Dioxins Congeners and Their Toxic Weighting Factors 2-11 2-5 EPA Facility-Specific Dioxin Congeners Distribution Corrections 2-12 2-6 Chemical Composition of Creosote and TWF 2-13 2-7 Definition of Polycyclic Aromatic Compounds 2-14 2-8 PAC Concentrations in Petroleum Products 2-16 2-9 PAC Concentrations in Crude Oils (mg/kg) 2-17 2-10 Supply and Disposition of Crude Oil and Petroleum Products 2-18 2-11 Products for Which PAC Composition Is Not Available 2-19 2-12 Calculation of Toxic Weighting Factor for Petroleum PACs 2-19 2-13 Calculation of Toxic Weighting Factor for Wood Preserving PACs 2-20 2-14 Calculation of Toxic Weighting Factor for Pulp, Paper, and Paperboard PACs 2-22 2-15 Tables Created in TRIReleases2007 2-23 2-16 Point Source Category Rankings 2-25 2-17 Metals Discharged by TWPE in TRI 2007 2-28 2-18 Facilities Reporting Discharges of Metals with the Highest TWPE in TRI 2007 2-29 3-1 States and Territories Included in DMRLoads2007 3-3 3-2 Data Types in PCS and ICIS-NPDES Used for DMRLoads2007 Development 3-6 3-3 PCS CNVRT Module Output 3-10 3-4 Example Calculation for DL Option Loads 3-16 3-5 NODI Code Descriptions 3-16 3-6 PCSLoadCalculator2007 Output 3-19 iii ------- LIST OF TABLES (Continued) Page 3-7 Results of NODI Code Excluded from EST=YES Revision Analysis for PCS 3-23 3-8 Actual Number of Days per Monitoring Period 3-28 3-9 Example Type 1 Flow Correction 3-29 3-10 Example Type 2 Flow Correction 3-30 3-11 Measurement Value Selection Priorities and Calculations 3-36 3-12 Flow Value Selection Priorities 3-38 3-13 ICIS-NPDES Load Calculator Module Output 3-42 3-14 Tables Imported or Created in DMRLoadsAnalysis2007 3-43 3-15 TWF Assignment for Chemical Mixtures 3-46 3-16 Comparison of Alternative Load Calculation Methods 3-47 3-17 Results of DL Sensitivity Analysis 3-49 3-18 Conversion Factors for Nitrogen Compounds 3-50 3-19 Tables Imported or Created in DMRLoads2007 3-51 3-20 Case-by-Case Point Source Category Reassignments in DMRLoads2007 3-55 3-21 Comparison of Load and Concentration Ranges for Common Parameters 3-60 3-22 DMR 2007 Point Source Category Rankings by TWPE 3-61 3-23 Summary of DMRLoads2007 Facility Review 3-67 4-1 SIC Codes Counted in Multiple Point Source Categories 4-5 4-2 Facility-Level Point Source Category Assignment SIC Codes 4-6 4-3 Example NAICS/Point Source Category Crosswalk Development 4-11 4-4 NAICS Codes Counted in Multiple Point Source Categories 4-12 4-5 Facility-Level Point Source Category Assignment NAICS Codes 4-14 5-1 Revised Boron TWF 5-2 5-2 Chemicals with no TWFs in TRIReleases2007 5-2 5-3 Chemicals with no TWFs in DMRLoads2007 5-3 iv ------- LIST OF TABLES (Continued) Page 6-1 Number of Facilities in Categories Selected for Preliminary Category Review 6-2 7-1 TRIReleases2007 Point Source Category Rankings 7-6 7-2 DMRLoads2007 Point Source Category Rankings 7-8 7-3 Point Source Categories That Have Undergone a Recent Rulemaking or Review 7-11 7-4 Point Source Categories with One Facility Dominating the TWPE Discharges 7-13 7-5 Final DMRLoads2007 and TRIReleases2007 Combined Point Source Category Rankings 7-15 ------- LIST OF FIGURES Page 2-1 Relationship Between the Three TRI2007 Databases 2-5 2-2 Basic Structure of the TRIReleases200 7 Database 2-24 3-1 Relationship Between Data Sources and Database Development Tools for the Development of DMRLoads2007 3-8 3-2 Example PCS STATS Code 3-12 3-3 Flow Diagram for PCS Load Calculator Routine 3-13 3-4 Relationship Diagram for Convert Module Output 3-32 3-5 Example STATS Code in ICIS-NPDES CONVERT_DMR Table 3-34 3-6 DMRLoadsAnalysis2007 Inputs Used to Create DMR2007 Table 3-44 3-7 DMRLoads2007 Database Structure 3-52 VI ------- Section 1 - Introduction 1. INTRODUCTION Under the Clean Water Act (CWA), EPA establishes national technology-based regulations known as effluent guidelines and pretreatment standards to reduce pollutant discharges from categories of industry discharging directly to waters of the United States or discharging indirectly through publicly owned treatment works (POTWs). The CWA sections 301(d), 304(b), 304(g), and 307(b) require EPA to annually review these effluent guidelines and pretreatment standards. This document supports EPA's 2009 review of its existing effluent guidelines and pretreatment standards. It also presents EPA's evaluation of categories of indirect dischargers without pretreatment standards to identify potential new categories for pretreatment standards, as required under CWA sections 304(g) and 307(b). Currently there are 56 point source categories that have effluent guidelines and pretreatment standards, which include over 450 subcategories. Additionally, CWA section 304(m) requires EPA to biennially publish an effluent guidelines program plan and provides for public notice and comment on such plan. Therefore, this document also supports the Preliminary 2010 Effluent Guidelines Program Plan (Preliminary 2010 Plan). Included in the Preliminary 2010 Plan is a solicitation for comments and data on industry categories currently not subject to effluent guidelines that are discharging non-trivial amounts of toxic or non-conventional pollutants. EPA's annual review of effluent guidelines and pretreatment standards has several components. First, EPA reviews all industrial categories subject to existing effluent limitations guidelines and standards to identify potential candidates for revision, as required by the CWA sections 304(b), 301(d), 304(g) and 307(b). The findings of this review are discussed in Section 7 of this report and are called the screening level analysis. Second, EPA reviews direct discharging industries not currently subject to effluent limitations guidelines and standards to identify potential candidates for effluent limitations guidelines development, as required by section 304(m)(l)(B) of the CWA. Finally, EPA reviews indirect discharging industries not currently subject to pretreatment standards to identify potential candidates for pretreatment standards development, as required by section 307(b). In conducting a screening level analysis, EPA uses readily available information from the Toxics Release Inventory (TRI), the Permit Compliance System (PCS), and the Integrated Compliance Information System for the National Pollutant Discharge Elimination System (ICIS- NPDES) to estimate the magnitude and relative toxicity of discharges from these industrial wastewater discharges. Section 1 discusses how EPA uses Standard Industrial Classification and North American Industrial Classification System codes to relate these discharge data to the 56 point source categories. EPA estimates the relative toxicity of these pollutant discharges in terms of toxic-weighted pound equivalents (TWPE). EPA estimates TWPE based on toxic weighting factors, which are discussed in detail in Section 5. EPA also uses available data to estimate discharges of pollutants in pounds, such as nutrients. For its 2009 review, EPA used information as reported to TRI, PCS, and ICIS-NPDES for 2007. EPA used 2007 data because these were the most recent TRI data available at the time it began the 2009 annual review. EPA used 2007 PCS and ICIS-NPDES data to reflect the same reporting year. EPA's 2009 screening level review is similar to that used for previous annual reviews (U.S. EPA, 2004; U.S. EPA, 2006; U.S. EPA, 2008). EPA used the 2009 review to confirm the identification of the three industrial categories prioritized for further review in the Final 2008 Effluent Guidelines Program Plan (73 FR 53218, ------- Section 1 - Introduction September 15, 2008) and to list the industrial categories currently regulated by existing effluent guidelines that cumulatively comprise 95 percent of the reported hazard (reported in units of TWPE) for preliminary category reviews. This report describes the development of the databases that EPA used in conducting its 2009 screening-level analysis. This report is a companion report for the Technical Support Document for the Preliminary 2010 Effluent Guidelines Program Plan (U.S. EPA, 2009). It also presents the results of the 2009 screening-level analysis. The remainder of this report is divided into the following sections: • Section 2 - Development of TRIReleases2007; • Section 3 - Development of DMRLoads2007; • Section 1 - Identification of Point Source Categories; • Section 5 - Toxic Weighting Factors (TWFs); • Section 6 - Quality Review; and • Section 7 - Results of 2009 Screening-Level Analysis. Section 7 provides a summary of the TWPE calculated from the TRI data and PCS/ICIS- NPDES data. EPA used the combined TWPE from the 2007 TRI and PCS/ICIS-NPDES data to prioritize its review of industry sectors that offer the greatest potential for reducing hazard to human health or the environment. 1.1 Introduction References 1. U. S. EPA. 2004. Technical Support Document for the 2004 Effluent Guidelines Program Plan. EPA-821-R-04-014. Washington, DC. (August). EPA-HQ-OW-2003-0074-1346 through 1352. 2. U. S. EPA. 2006. Technical Support Document for the 2006 Effluent Guidelines Program Plan. EPA-821-R-06-018. Washington, DC. (December). EPA-HQ-OW-2004-0032- 2782. 3. U. S. EPA. 2008. Technical Support Document for the 2008 Effluent Guidelines Program Plan. EPA-821-R-08-015. Washington, DC. (August). EPA-HQ-OW-2006-0771-1701. 4. U. S. EPA. 2009. Technical Support Document for the Preliminary 2010 Effluent Guidelines Program Plan. EPA-821-R-09-006. Washington, DC. (October). EPA-HQ- OW-2007-0571 DCN 06703. 1-2 ------- Section 2 - Development of TRIReleases2007 2. DEVELOPMENT OF TRIRELEASES2007 As discussed in Section 1, EPA annually reviews promulgated effluent limitations guidelines and pretreatment standards (ELGs) by investigating available information on industrial pollutant discharges. EPA identified that the Toxics Release Inventory (TRI) contains readily available and relevant data on industrial pollutant discharges, specifically that TRI data has information on industrial pollutant discharges to surface waters ("direct discharges") and to publicly owned treatment works (POTWs) ("indirect discharges"). Consequently, EPA was able to use TRI data for its review of: (1) promulgated effluent guidelines ("direct discharges"); (2) promulgated pretreatment standards ("indirect discharges"); and (3) direct and indirect industrial pollutant discharges not currently subject to effluent guidelines or pretreatment standards. As discussed in Section 7, EPA combined the toxic-weighted pound equivalent (TWPE) calculated from the TRI data and the discharge monitoring report data contained in the Permit Compliance System (PCS) and the Integrated Compliance Information System for the National Pollutant Discharge Elimination System (ICIS-NPDES) (see Section 3 for information about PCS and ICIS-NPDES). EPA used this combined TWPE to prioritize its review of industry sectors that offer the greatest potential for reducing hazard to human health or the environment. This section discusses the methodology EPA used to create TRIReleases2007, a database created by the EPA to analyze 2007 TRI data. It also presents the unweighted annual pollutant load (i.e., pounds) and the relative toxicity of these discharges using toxic weighting factors (TWFs) (i.e., TWPE) for all facilities reporting discharges to TRI for the year 2007 and for the point source categories that these facilities represent. Tables A-l and A-2 in Appendix A present the annual pollutant load and TWPE from TRIReleases2007 on a six-digit North American Industrial Classification System (NAICS) code and chemical basis, respectively. This section is organized in the following subsections: • Section 2.1 - General TRI information; • Section 2.2 - Overview of TRI databases; Section 2.3 - TRIRawData2007; • Section 2.4 - TRICalculations2007; • Section 2.5 - TRIReleases2007; • Section 2.6 - TRIReleases2007; and • Section 0 - TRIReleases2007 References. 2.1 TRI TRI is the common name for Section 313 of the Emergency Planning and Community Right-to-Know Act. Each year, facilities that meet certain criteria must report their releases and other waste management activities of listed toxic chemicals (i.e., the quantities of toxic chemicals recycled, collected and combusted for energy recovery, treated for destruction, or disposed by the facility). A separate report must be filed for each chemical that exceeds the reporting threshold. For the 2009 annual review of effluent guidelines, EPA used data for reporting year 2007, because they were the most recent data available at the time the review began. The TRI list of chemicals for reporting year 2007 includes more than 600 chemicals and chemical categories. Prior to 2006, facilities were required to identify their operations using Standard Industrial Classification (SIC) codes. Starting with reporting year 2006, EPA modified 2-1 ------- Section 2 - Development of TRIReleases2007 the requirement for facilities to report SIC codes, so that facilities now use the NAICS code (73 FR 324666). A facility must submit a TRI report if it meets the following three criteria (U.S. EPA, 2001): 1. NAICS Code Determination: Most facilities in NAICS codes 11, 21, 22, 31 through 33, 42, 48 through 49, 51, 54, 56 and 81, and federal facilities are potentially subject to TRI reporting. EPA generally relies on facility claims regarding the NAICS code identification. The primary NAICS code determines if TRI reporting is required. The primary NAICS code is associated with the facility's revenues, and may not relate to their pollutant discharges (73 FR 12045, March 6, 2008). 2. Number of Employees: Facilities must have 10 or more full-time employees or their equivalent. EPA defines a "full-time equivalent" as a person who works 2,000 hours in the reporting year (there are several exceptions and special circumstances that are well defined in the TRI reporting instructions). 3. Activity Thresholds: If the facility is in a covered NAICS code and has 10 or more full-time employee equivalents, it must conduct an activity threshold analysis for every chemical and chemical category on the current TRI list. The facility must determine whether it manufactures, processes, or otherwise uses each chemical at or above the appropriate activity threshold. Reporting thresholds are not based on the amount of release. All TRI thresholds are based on mass, not concentration. Thresholds for persistent bioaccumulative toxic (PBT) chemicals are lower than for non-PBT chemicals. In TRI, facilities report annual releases to the environment of each toxic chemical or chemical category that meets reporting requirements. TRI requires facilities to report on-site releases to air, receiving streams, disposal to land, underground wells, and several other categories. Facilities must also report the amount of toxic chemicals in wastes transferred to off- site locations, including discharges to POTWs and other off-site locations, such as commercial waste disposal facilities. For this review, EPA focused on facility reports of chemical discharges directly to a receiving stream or transfers to a POTW. For discharges directly to a stream ("direct discharges"), EPA took the annual loads directly from the reported TRI data for calendar year 2007. For transfers of chemicals to POTWs ("indirect discharges"), EPA first adjusted the TRI pollutant loads to account for pollutant removal at the POTW prior to discharge to the receiving stream (see Section 2.4.2 for more details). TRI does not require facilities to sample and analyze wastestreams to determine the quantities of toxic chemicals released. Facilities may estimate releases based on mass balance calculations, published emission factors, site-specific emission factors, or other approaches. Facilities must indicate the basis of their release estimate using a reporting code. According to TRI's reporting guidance, facilities should use one-half the detection limit to estimate mass releases of chemicals that are measured below their detection limit and are reasonably expected ------- Section 2 - Development of TRIReleases2007 to be present. Nondetects of dioxin and dioxin-like compounds, however, may be reported as zero. TRI allows facilities to report releases as specific numbers or as ranges, if appropriate. Specific estimates are encouraged if data are available to ensure the accuracy; however, EPA allows facilities to report releases in the following ranges: 1 to 10 pounds, 11 to 499 pounds, and 500 to 999 pounds. For this review, if a facility reported releases in a range, EPA used the mid- point of each reported range to represent a facility's releases. 2.1.1 Utility of TRI The data collected in TRI are particularly useful for the 304(m) review process for the following reasons: • TRI includes data from all 50 states and U.S. territories; • TRI includes transfers to POTWs, not just direct discharges; • TRI includes discharge data from manufacturing NAICS codes and some other industrial categories which may handle significant quantities of toxic chemicals; and • TRI includes releases of many chemicals, not just those already identified as problems and limited in facility discharge permits. 2.1.2 Constraints and Limitations of TRI TRI provides comprehensive data for direct and indirect discharging facilities. However, EPA identified the following constraints and limitations to using TRI for the screening-level analysis: • Small establishments (less than 10 employees) are not required to report, nor are facilities that do not meet the reporting thresholds. Therefore, facilities reporting to TRI may not provide a complete picture of the industry. • Release reports are, in part, based on estimates, not measurements, which may result in inaccurately reported releases. For example, TRI encourages facilities to report some compounds as present at one-half the detection level if a facility suspects that the compound has the potential to be present, even if measured data show the compound is below its detection level. As a result, many companies are conservative and adopt this approach. For facilities with large flows, this can result in large estimates of pounds or TWPE of pollutant released with no measurements to support that the compound was ever present above the detection level. • Certain chemicals (polycyclic aromatic compounds (PACs), dioxin and dioxin- like compounds, and metal compounds) are reported as a class, not as individual compounds. Because the individual compounds in the class have widely varying toxic effects, the potential toxicity of chemical releases can be inaccurately estimated. • Facilities are identified by NAICS code, not point source category. For some NAICS codes, it may be difficult or impossible to identify the point source category that is the precise source of the toxic wastewater releases (see Section 1 for additional information). 2-3 ------- Section 2 - Development of TRIReleases2007 • The list of chemicals covered by TRI is not all-inclusive and changes over time. • Only facilities in certain NAICS codes are required to report; therefore, some sources of water pollutant discharges are not included. • A facility is not required to report releases if the releases do not exceed the reporting threshold. • Information in TRI does not represent national estimates because not all facilities are required to report to TRI. Despite TRI's limitations and constraints, EPA has determined that it is appropriate to be used for an initial screening-level review and prioritization of the pollutant loadings discharged by industrial categories. EPA will further evaluate the prioritized categories in a second level of review which may include additional data collection and verification of data reported in TRI. 2.2 Overview of TRI Databases EPA developed the end-user database, TRIReleases2007, in three steps: 1. Downloaded relevant data from TRI to create TRIRawData2007 (see Section 2.3). 2. Estimated relative toxicity of discharges, set up groupings of facilities (by NAICS code and discharge type), and made corrections and adjustments to create TRICalculations2007 (see Section 2.4). 3. Grouped the pollutant discharges in TRICalculations2007 by NAICS code, point source category, and other groupings to create TRIReleases2007 for rankings and other analyses (see Section 2.5). Figure 2-1 shows how these three databases are related and the following sections describe the creation and particulates of each database in greater detail. 2-4 ------- Section 2 - Development of TRIReleases2007 EPA TRI Web Site Files File Type 1 File Type 2B File Type 3A File Type 3B File Type 4 (Downloaded) TRIRawData2007 Tables >US_l_2007_v07 US_2b_2007_v07 US_3a_2007_v07 US_3b_2007_v07 US 4 2007 v07 (Copied) Tables . TRI Raw Data 2007 Tables Priority Pollutants PSC Codes NAICS/PSC Crosswalk TWFs POTW Removals Dioxin Chemicals TRI Master List TRI Master Facility List (Linked) Tables . TRI Calculations 2007 Tables Counts of Facilities by NAICS Counts of Facilities by PSC NAICS Code Rankings Point Source Rankings Figure 2-1. Relationship Between the Three TRI 2007 Databases 2.3 TRIRawData2007 EPA created TRIRawData2007 using the 2007 TRI data for all of the United States, which are available from the EPA Web site (www.epa.gov/tri). Table 2-1 lists the relevant TRI 2007 files that EPA imported into the Microsoft Access™ database. Table 2-1. TRI 2007 Tables Downloaded from EPA Table Name "File Type 1 : Facility, Chemical, Releases and Other Waste Management Summary Information" "File Type 2B: Detailed On- Site Waste Treatment Methods and Efficiency" Description of File Contents Facility information (Part I on Form R and Form A), as well as most chemical information (Part II on Form R and Form A). Data elements are reported individually. The information is also disaggregated based on Waste Management code (i.e., Management "M" code reported on TRI Form R), and aggregated up to On-site Releases, Off-site Releases, Other On-site Waste Management, and Transfers Off Site for Further Waste Management categories. Facility information (Part I on Form R and Form A) and On-site Waste Treatment Methods and Efficiency data (Part II, Section 7 A on Form R). 2-5 ------- Section 2 - Development of TRIReleases2007 Table 2-1. TRI2007 Tables Downloaded from EPA Table Name "File Type 3A: Details of Transfers Off Site" "File Type 3B: Details of Transfers to POTW" "File Type 4: Details of Facility Information" Description of File Contents Facility information (Part I on Form R and Form A) as well as details of individual transfers off-site (Part II, Section 6.2 on Form R). Facility information (Part I on Form R and Form A) as well as a list of POTWs (Part II, Section 6. l.B on Form R). Facility information (Part I on Form R and Form A) for all facilities that have ever reported to the TRI program. The "reporting year" field at the beginning of each record identifies the last year the facility reported to the TRI program. Source: http://www.epa.gov/tri/tridata/tri07/data/index.htm. 2.4 TRICalculations200 7 As the second step in developing TRIReleases2007, EPA created TRICalculations2007 by copying raw data tables from TRIRawData2007', omitting unrelated data (e.g., air emissions and source reduction activities), and performing the following actions: • Corrected NAICS code classification for certain facilities and chemicals and corrected certain reported chemical quantities (Section 2.4.1); • Estimated POTW removals for indirect discharges (Section 2.4.2); • Estimated the mass-based and toxic-equivalent pollutant loadings (Section 2.4.3); • Combined releases of parent metals and their associated compounds (Section 2.4.4); and • Determined basis of TRI release and transfer estimates (Section 2.4.5). To perform the calculations listed above, EPA imported tables from previous versions of EPA's TRICalculations databases containing Chemical Abstract Service (CAS) numbers, TWFs, and POTW removal rates. Table 2-2 lists the database tables that EPA imported or created in TRICalculations2007. Table 2-2. Tables Imported or Created in TRICalculations2007 Table Name "All Water Releases" "Dioxin Chemicals" "Dioxin Distributions" "Manual Data Changes" "Manual Dioxin Distribution Changes" Created or Imported? Created using VBA code Imported from TRIReleases2005 Created using VBA Code Created Created a Description Lists calculated TWPE for every chemical discharge reported to TRI in 2007 for which EPA has calculated a TWF. EPA developed this table using data from TRIRawData2007 and TWF tables. This table serves as an intermediate table between the TRIRawData2007 tables and the TRI Master List Table. Lists the 17 dioxin congeners and the TRI congener number associated with each. Lists the dioxin distributions (see Section 2.4.3) of all facilities reporting dioxin discharges in 2007. EPA developed this table using data from TRIRawData2007. Documents any changes that EPA made to the data from TRIRawData2007. Documents changes that EPA made to the dioxin distributions from TRIRawData2007. 2-6 ------- Section 2 - Development of TRIReleases2007 Table 2-2. Tables Imported or Created in TRICalculations2007 Table Name "Manual Load Changes" "Manual NAICS Code Changes" "NAICS Code Changes" "NAICS Code Changes (for facilities with no NAICS Codes)" "NAICS_Codes" "OCPSF Pesticides" "Parent Metals and Compounds" "Pesticides Chemical List" "POTW Removals" "Priority Pollutants" "PSC Codes" "PSC/NAICS Crosswalk" "PSC/NAICS Crosswalk without M PM_fixed" "Pulp and Paper Dioxin Distribution" "Pulp and Paper Phases" "TRI Raw Data 2007 Tables" "Point Source Category Codes" Created or Imported? Created a Created a Created a Created a Created Created using VBA code Imported from TRIReleases2005 Imported from TRIReleases2005 Imported from TRIReleases2005 Imported from TRIReleases2000 Imported from TRIReleases2005 Created Created Imported from TRIReleases2005 Imported from TRIReleases2005 Imported from TRIRawData2007 Imported from TRIReleases2005 Description Documents changes that EPA made to the magnitude of the discharges in the data from TRIRawData2007. Documents changes that EPA made to the NAICS codes in the data from TRIRawData2007 '. Documents changes that EPA makes every year to the NAICS codes of the discharges in the data from TRIRawData2007 '. Documents the NAICS codes that EPA assigned to facilities with no NAICS codes in the data from TRIRawData2007 '. Lists U.S. Economic Census definitions of the NAICS codes. EPA developed this table using information from the U.S. Economic Census Web site (www.census.gov). Lists all pesticide discharges reported for 2007 classified under the Pesticide Chemicals Category. Links parent metals to the appropriate metal compound groups (e.g., nickel and nickel compounds). Lists all chemicals classified as pesticides under the Clean Water Act. Lists all 612 TRI chemicals and chemical compounds and their chemical-specific average POTW percent removal. See "POTW Percent Removals Used for the TRIReleases2002 Database" (Codding, 2005) (see Section 2.4.2), Lists priority pollutants (CAS number and chemical name). Defines all codes for point source categories. EPA used the "SIC/Point Source Category Crosswalk" table from TRIReleases2000 and the NAICS/SIC Crosswalk developed for the 2002 U.S. Economic Census to develop this table (see Section 1). Identifies NAICS codes that could have discharges subject to the Metal Products and Machinery (MP&M) ELGs. EPA created this table from information in the MP&M rulemaking. Contains the dioxin distribution used for facilities in the Pulp, Paper, and Paperboard Category that do not report a facility- specific distribution. EPA developed this table using information obtained from the pulp and paper industry (Matuzko et al., 2006). Lists the NAICS code placeholders used to identify facilities in the Pulp, Paper, and Paperboard Category by regulatory phase. Copy of all original TRI tables stored in the TRIRawData2007 database and deleted information not needed for the 2009 annual review. Lists point source categories and corresponding point source category codes. 2-7 ------- Section 2 - Development of TRIReleases2007 Table 2-2. Tables Imported or Created in TRICalculations2007 Table Name "TWFs" "TRI Chemicals with MP&My/n" "TRI Master List" "TRI Master Facility List" "Wood Preserving Dioxin Distribution" Created or Imported? Imported from TRIReleases2005 Imported from TRIReleases2005 Created using VBA code Created using VBA code Imported from TRIReleases2005 Description Lists TWF for chemicals based on the Office of Water references. EPA originally created this table using TWFs as of December 2004 and updates it with TWFs created or revised after 2004 (see Section 5). Lists chemicals regulated by the MP&M rulemaking. Lists calculated pounds and TWPE for every chemical released by every facility reporting to TRI in 2007. EPA developed this table using data from "All Water Releases" and "TWFs" tables. Complete and unique list of all facilities reporting to TRI, relevant facility information (address, contacts, etc.), and corresponding primary NAICS codes. EPA developed this table using data from TRIRawData2007. Contains the dioxin distribution used for facilities in the Wood Preserving Category that do not report a facility-specific distribution. EPA developed this table using information obtained from the wood preserving industry. a Most of these changes are based on previous knowledge about the facility's operations from previous annual reviews. EPA also added changes discovered as part of the 2009 annual review. See Section 2.4.1 for additional information. VBA - Visual Basic for Applications. 2.4.1 Modifications to TBI-ReportedData Modifications to TRI-Reported data include the following facility-specific changes: • Pollutant loading changes; • Dioxin distribution changes; and • NAICS code changes. During the screening-level reviews of the 2000 through 2007 TRI data, EPA made corrections to TRIReleases databases based on information received from stakeholders, including industry trade associations, facilities, and pretreatment coordinators. The SIC code corrections identified for past years of review were converted to NAICS code corrections using the U.S. Economic Census linkages and similarly applied to the 2007 data, as appropriate. In addition, EPA conducted a quality review of the TRIReleases2007 database (described in Section 6). As a result of this review, EPA made 58 corrections1 to the 2007 releases. Table A-3 in Appendix A, lists the corrections EPA made to the TRIReleases2007 database. EPA assigned pollutant loadings to point source categories based on the primary NAICS code that facilities reported (see Section 1). A facility reports up to six NAICS codes to TRI and specifies one primary NAICS code. In cases where EPA was able to identify that chemical 1 In addition to the 58 changes that were made to individual releases reported to TRI, EPA made 795 NAICS code changes to account for facilities that did not report a NAICS code and SIC code changes that EPA had made to facilities in past years. EPA identified the appropriate NAICS code for facilities that did not report a NAICS code using Envirofacts. 2-8 ------- Section 2 - Development of TRIReleases2007 releases to surface water or a POTW were related to activities covered by a different NAICS code, EPA corrected the NAICS code assigned to the facility and/or chemical. For example, a facility may report their primary NAICS code as 325110, Petrochemical Manufacturing. The facility may also perform pesticide manufacturing, which is covered under NAICS code 325320, Pesticide and Other Agricultural Chemical Manufacturing. If this facility reported a pesticide release, EPA assigned the pesticide release to the Pesticide Chemicals Category, because these pollutant discharges are regulated under the Pesticide Chemicals Category, not the Organic Chemicals, Plastics, and Synthetic Fibers Category. Section 1 in this report provides a detailed discussion of the development of the crosswalk between the NAICS code and point source category. 2.4.2 POTW Removals For facilities that reported transfers of chemicals to POTWs, EPA first adjusted the reported pollutant loads to account for pollutant removal that occurs at the POTW prior to discharge to the receiving stream. EPA estimated the pounds of facilities' waste released to the surface water after POTW removal using the following equation2: Release to Stream (Ibs/yr) = [Transfer to POTW (Ibs/yr)] x [1 - POTW Removal (%)] EPA developed a POTW removal hierarchy for the TRIReleases2002 database, described in the memorandum entitled "POTW Percent Removals Used for the TRIReleases Databases" (Codding, 2005). The TRIReleases2007 database uses the same POTW removal hierarchy. In short, EPA used removal efficiencies from the following data sources, listed in order of preference: 1. Recent effluent guidelines rulemakings; 2. EPA/Office of Research and Development's National Risk Management and Research Laboratories treatability database; and 3. EPA/Office of Prevention, Pesticides, and Toxic Substances' Risk Screening Environmental Indicators model. Table A-4 in Appendix A lists the POTW Removals and their data sources, in alphabetical order. 2.4.3 TWFs To identify potential impacts on human health and the environment, EPA estimated toxic equivalent mass discharge through the use of TWFs. EPA used the "TWFs" table, which lists TWFs by CAS number, in TRICalculations2007 to calculate TWPE for chemical discharges. If the table did not list a TWF for a specific parameter, EPA did not include pollutant discharges for this chemical in its TWPE estimates. Section 5 describes TWFs in more detail. See also Toxic Weighting Factor Development in Support of the 304(m) Planning Process (ERG, 2005). In some cases, EPA calculated industry-specific TWFs for certain chemical compound categories reported in TRI. These TWFs were not used to calculate TWPE for chemical For example, the POTW removal efficiency for lithium carbonate is 1.85 percent. That is if 10,000 Ibs of lithium carbonate discharged to a POTW, only 9,815 Ibs of lithium carbonate will be discharged from the POTW to surface waters as this amount is untreated by the POTW [9,815 Ibs = 10,000 Ibs x (1 - 0.0185)]. 2-9 ------- Section 2 - Development of TRIReleases2007 discharges in PCS/ICIS-NPDES. EPA created specific TRI TWFs when it had additional information about the composition of the compound category, as released from specific industries. Table 2-3 lists the calculated TWFs. Table 2-3. TWF Modifications Chemical Dioxins Creosote PACs PACs PACs PACs Point Source Category All All All point source categories, except those in the Petroleum Refining; Wood Preserving; and Pulp, Paper, and Paperboard Categories Petroleum Refining Category Wood Preserving Category Pulp, Paper, and Paperboard Category TWF Apply individual dioxin compound TWF using the following dioxin congener distribution: 1) facility-specific, 2) industry specific, 3) NAICS-code-average, or 4) median dioxin TWF for all dioxin congeners. 1.3577 a 100.66 26.28 b 8.36 b 34.21 c a Calculations of TWF and TWPEfor Creosote from Wood Preserving Facilities (Bicknell, 2004). b TRI 2002 PACs TWF for Petroleum Refining, Creosote, and Wood Preserves (Finseth, 2005). c NCASI SARA Handbook - Table 5 PAC Concentrations in Pulp Mill Effluents (H.C. Lavallee, Inc., 2005). The remainder of this subsection describes how EPA developed the TWFs, in the following order: • Dioxins for all categories; • Creosote for all categories; • Wood Preserving Category creosote; • Polycyclic Aromatic Compounds (PACs) for all categories; • Petroleum Refining Category PACs; • Wood Preserving Category PACs; and • Pulp, Paper, and Paperboard Category PACs. Dioxins The term 'dioxins' refers to poly chlorinated dibenzo-p-dioxins (CDDs) and poly chlorinated dibenzofurans (CDFs), which constitute a group of PBT chemicals. There are 17 CDDs and CDFs congeners with chlorine substitution of hydrogen atoms at the 2, 3, 7, and 8 positions on the benzene rings, the most toxic of which is 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). The 17 compounds (called congeners) are referred to as 'dioxin-like,' because they have similar chemical structure, similar physical-chemical properties, and invoke a common battery of toxic responses (U.S. EPA, 2000), though the toxicity of the congeners varies greatly. In this report, EPA uses the term "dioxin and dioxin-like compounds" to refer to all 17 of the 2,3,7,8-substituted CDDs and CDFs. EPA developed TWFs for each of the 17 dioxin congeners, ranging from 703,584,000 for 2,3,7,8-TCDD to 2,021 for octachlorodibenzofuran. Due to their toxicity and ability to bioaccumulate, the various dioxin congeners have high TWFs relative to most chemicals. 2-10 ------- Section 2 - Development of TRIReleases2007 Consequently, even small mass amounts of dioxin and dioxin-like compound discharges translate into high TWPEs. Table 2-4 presents the dioxin congener-specific TWFs used in the screening- level analysis. Table 2-4. Dioxins Congeners and Their Toxic Weighting Factors CAS Number Chemical Name Abbreviated Name Toxic Weighting Factor CDDs 1746-01-6 40321-76-4 39227-28-6 57653-85-7 19408-74-3 35822-46-9 3268-87-9 2,3,7,8-tetrachlorodibenzo-p-dioxin 1,2,3,7,8-pentachlorodibenzo-p-dioxin 1,2,3,4,7,8-hexachlorodibenzo-p-dioxin 1,2,3,6,7,8-hexachlorodibenzo-p-dioxin 1,2,3,7,8,9-hexachlorodibenzo-p-dioxin 1,2,3,4,6,7,8-heptachlorodibenzo-p-dioxin 1,2,3,4,6,7,8,9-octachlorodibenzo-p-dioxin 2,3,7,8-TCDD 1,2,3,7,8-PeCDD 1,2,3,4,7,8-HxCDD 1,2,3,6,7,8-HxCDD 1,2,3,7,8,9-HxCDD 1,2,3,4,6,7,8-HpCDD 1,2,3,4,6,7,8,9-OCDD 703,584,000 692,928,000 23,498,240 9,556,480 10,595,840 411,136 6,586 CDFs 51207-31-9 57117-41-6 57117-31-4 70648-26-9 57117-44-9 72918-21-9 60851-34-5 67562-39-4 55673-89-7 39001-02-0 2,3 ,7,8-tetrachlorodibenzofuran 1 ,2,3 ,7,8-pentachlorodibenzofuran 2,3,4,7,8-pentachlorodibenzofuran 1,2,3,4,7,8-hexachlorodibenzofuran 1,2,3,6,7,8-hexachlorodibenzofuran 1,2,3,7,8,9-hexachlorodibenzofuran 2,3,4,6,7,8-hexachlorodibenzofuran 1,2,3,4,6,7,8-heptachlorodibenzofuran 1,2,3,4,7,8,9-heptachlorodibenzofuran 1,2,3,4,6,7,8,9-octachlorodibenzofuran 2,3,7,8-TCDF 1,2,3,7,8-PeCDF 2,3,4,7,8-PeCDF 1,2,3,4,7,8-HxCDF 1,2,3,6,7,8-HxCDF 1,2,3,7,8,9-HxCDF 2,3,4,6,7,8-HxCDF 1,2,3,4,6,7,8-HpCDF 1,2,3,4,7,8,9-HpCDF 1,2,3,4,6,7,8,9-OCDF 43,819,554 7,632,640 557,312,000 5,760,000 14,109,440 47,308,800 51,204,160 85,760 3,033,984 2,021 Source: Toxic Weighting Factor Development in Support of the 304(m) Planning Process (ERG, 2005). Beginning with reporting year 2000, facilities meeting certain reporting criteria were required to report to TRI the total mass, in grams, of the 17 dioxin and dioxin-like compounds released to the environment every year. This reporting method does not account for the relative toxicities of the 17 compounds. However, reporting facilities are given the opportunity to report a facility-specific congener distribution. Yet even if dioxins are released to more than one medium, the facility can report only one distribution. EPA cannot know if the single dioxin congener distribution reported by a facility accurately reflects the dioxin distribution in wastewater3. Nevertheless, it is the best available information and EPA uses it to calculate the reporting facility's dioxin TWPE. To account for the relative toxi cities of the different dioxin congeners, EPA first converted the reported dioxin releases from grams to pounds to be consistent with the units used for other chemicals. EPA then calculated dioxin TWPE estimates using the facility-specific congener distributions for all facilities that reported a distribution. Based on information Beginning with reporting year 2008, facilities will be required to submit information on the amount of each individual dioxin congener where that information is available. Facilities that cannot quantify dioxin releases by congener may continue to report an aggregate number (72 FR 26544; May 10, 2007). 2-11 ------- Section 2 - Development of TRIReleases2007 provided by the facilities that were contacted as part of previous and current quality reviews, EPA made corrections to the reported dioxin distributions for the facilities presented in Table 2-5. Table 2-5 also includes the reason for the correction. Table 2-5. EPA Facility-Specific Dioxin Congeners Distribution Corrections Facility Name Cahaba Pressure Treated Forest Products Inc. Du Pont Memphis Plant Louisiana Pigment Co LP Colfax Treating Co LLC Eastman Kodak Co Kodak Park Facility Location Brierfield, AL Memphis, TN Westlake, LA Pineville, LA Rochester, NY Reason for Dioxin Congener Distribution Change The facility did not use the industry -provided dioxin distribution (Woodruff, 2007). The facility provided more detailed information about the dioxin measurements (Zweig, 2000). The facility provided more detailed information about the dioxin measurements (Kashyap, 2009). The facility did not use the industry-provided dioxin distribution based on pentachlorophenol distribution (Johnston, 2004). The facility provided more detailed information about the dioxin measurements (Moeller, 2009) EPA calculated an average dioxin distribution for each NAICS code which had reported dioxin releases. For facilities that did not report a dioxin distribution, EPA used the average NAICS code distribution to calculate the facility's dioxin TWPE. EPA calculated industry- specific dioxin distributions for the Petroleum Refining and the Pulp, Paper, and Paperboard Categories based on information received from industry trade groups. For facilities that did not report a congener distribution and did not have any facilities within its NAICS code that reported a congener distribution, EPA used a TWF equal to 10,595,840 (the median of the 17 dioxin congener TWFs). Creosote Creosote is a commonly used wood preservative, comprising many different chemicals. EPA did not develop a TWF for creosote using creosote toxicity data. Instead, EPA used the chemical composition of creosote, provided in IARC Monographs, Vol 35, "Coal Tar and Derived Products," (WHO, 1998) and the TWFs for these individual chemicals to calculate a TWF for creosote. EPA made the following assumptions in developing the TWF for creosote: 1. Chemicals will be present in wastewater in the same proportion that they are present in the creosote. 2. If no TWF was available for a specific chemical, its concentration in creosote was assumed to be zero. Using the data provided in IARC Monographs, Vol 35, EPA calculated the average percentage that the chemical represents in creosote based on the high and low value (WHO, 1998). EPA calculated an adjusted TWF for each chemical by multiplying its chemical-specific TWF by its average percentage in creosote. EPA summed these values to calculate a new overall TWF for creosote discharges. Table 2-6 lists the chemical composition of creosote, along with the associated TWF of the various chemicals. 2-12 ------- Section 2 - Development of TRIReleases2007 Table 2-6. Chemical Composition of Creosote and TWF Pollutant Acenaphthene Antracene Benz(a)anthracene Benzo(a)pyrene Benzofluourenes Biphenyl Carbazole Chrysene Dibenz(a,h)anthtracene Dibenzofuran Dimethylnaphthalenes Fluoranthene Fluorene Methylantracenes Methylfluorenes 1 -Methy Inaphthalene 2-Methylnaphthalene Methy Iphenanthrenes Naphthalene Phenanthrene Pyrene Chemical Percentage (%) 11.85 4.50 0.21 0.05 1.50 1.20 1.60 2.80 0.03 5.75 2.15 5.25 8.65 3.95 2.65 6.45 6.60 3.00 9.65 18.50 4.75 TWF 0.0325697 2.5455945 30.695 100.66 0.1555556 0.0365558 0.709071 31.01 30.772 0.49215 1.2846944 0.70105 0.0486957 0.0062222 0.1930493 0.1037037 0.0158701 0.2947368 0.0932033 Total Adjusted TWF 0.0038595 0.1145518 0.0644595 0.05033 0.0023333 0.0004387 0.0113451 0.86828 0.0092316 0.0282986 0 0.0674465 0.0606408 0 0.0012904 0.0004013 0.0127413 0.0031111 0.0015315 0.0545263 0.0044272 1.36 Creosote Releases from Wood Preserving Facilities EPA received information from the Southern Pressure Treaters Association in 2005 that indicates creosote discharges are estimated based on a surrogate analyte, such as oil and grease or total phenols. The Southern Pressure Treaters Association also indicated that TRI-reported PAC discharges are usually estimated based on the creosote estimates, but there is no standard approach for making these estimates (H.M. Rolling Company, 2005). PACs and creosote contain many of the same chemicals (compare Table 2-7 and Table 2-6, respectively). Consequently, if EPA estimated the TWPE for both the PACs and the creosote in the same discharge, then the discharges of some toxic chemicals would be double counted. For this reason, if a wood preserving facility reports PACs and creosote in the same discharge (e.g., both are reported in direct discharges to surface water), EPA included the TWPE for the PAC discharges, but not the creosote discharges. If the wood preserving facility reports only creosote releases (and not PACs), EPA used the calculated creosote TWF of 1.36 to calculate the TWPE. 2-13 ------- Section 2 - Development of TRIReleases2007 Polycyclic Aromatic Compounds (PACs) PACs, sometimes known as polycyclic aromatic hydrocarbons (PAHs), are a class of organic compounds consisting of two or more fused aromatic rings. Table 2-7 lists the 21 individual compounds in the PAC category for TRI reporting, CAS number, and TWF, if available. EPA has TWFs for only eight of the 21 PACs chemicals. Table 2-7. Definition of Polycyclic Aromatic Compounds PAC Compound Benzo(a)anthracene Benzo(a)phenanthrene (chrysene) Benzo(a)pyrene Benzo(b)fluoranthene Benzo(j)fluoranthene Benzo(k)fluoranthene Benzo(j,k)fluorene (fluoranthene) Benzo(r,s,t)pentaphene Dibenz(a,h)acridine Dibenz(aj )acridine Dibenzo(a,h)anthracene Dibenzo(a,e)fluoranthene Dibenzo(a,e)pyrene Dibenzo(a,h)pyrene Dibenzo(a,l)pyrene 7H-Dibenzo(e,g)carbazole 7, 12-Dimethylbenz(a)anthracene Indeno( 1 ,2,3 -cd)pyrene 3 -Methy Icholanthrene 5 -Methy Ichrysene 1-Nitropyrene CAS Number 56-55-3 218-01-9 50-32-8 205-99-2 205-82-3 207-08-9 206-44-0 189-55-9 226-36-8 224-42-0 53-70-3 5385-75-1 192-65-4 189-64-0 191-30-0 194-59-2 57-97-6 193-39-5 56-49-5 3697-24-3 5522-43-0 Toxic Weighting Factor 36.2600 31.0100 100.6600 30.6600 NA 30.6600 0.8290 NA NA NA 30.6600 NA NA NA NA NA NA 30.6600 NA NA NA NA - Not available. EPA has not developed TWFs for these compounds. PACs are classified as PBTs. They are likely present in petroleum products such as crude oil, fuel oil, diesel fuel, gasoline, and paving asphalt (bituminous concrete) and refining by- products such as heavy oils, crude tars, and other residues. PACs form as the result of incomplete combustion of organic compounds. PACs and closely related compounds are major constituents of creosote, a commonly used wood preservative. For TRI, facilities that manufacture, process, or use more than 100 pounds of PACs per year must report the combined mass of PACs released; they do not report releases of individual compounds. In the development of TRIReleases2007 EPA assigned the TWF of benzo(a)pyrene to PACs, with the exception of releases reported by facilities in the Petroleum Refining; Wood Preserving; and Pulp, Paper, and Paperboard Categories (for which EPA has more detailed information). Because the TWF for benzo(a)pyrene (100.66) is higher than any other PAC, this 2-14 ------- Section 2 - Development of TRIReleases2007 represents a worst-case scenario. For PAC discharges that are not completely benzo(a)pyrene, this method overestimates the relative toxicity of the discharges. Petroleum Refining PACs EPA used available data for the Petroleum Refining Category to calculate TWPE for PACs reported by petroleum refining facilities. Facilities report to TRI the combined mass of PACs released, but for this industry EPA used information on the distribution of PACs in refinery products from the American Petroleum Institute (API, 1994). EPA assumed that the composition of PACs released by refineries is proportional to the composition of raw materials (crude oil) and products throughput at U.S. refineries. EPA developed this methodology for the Petroleum Refining Detailed Study supporting the 2004 Effluent Guidelines Program Plan (U.S. EPA, 2004). After the methodology was developed, the calculated refinery PAC TWF changed due to the changes in TWFs for individual PAC chemicals. PACs can occur in a number of petroleum products and crude oils; this information is available in literature (see Table 2-8 and Table 2-9). In addition, the Energy Information Administration (EIA) publishes a yearly report of the amount of petroleum products produced in all U.S. petroleum refineries as well as the amount of crude oil consumed (see Table 2-10). EPA made the following assumptions in developing the TWF for Petroleum Refining Category PACs: 1. PACs will be present in wastewater in the same proportion that they are present in the crude oil and products throughput at U.S. refineries. Table 2-10 presents these proportions. 2. If EPA did not have literature data available for a specific PAC compound, its concentration in the crude oil or product was assumed to be zero. If a PAC compound was reported as not detected, its concentration in the crude oil or product was assumed to be zero. 3. Where PAC composition is not available, it can be estimated using the composition from similar products. Table 2-11 lists the products for which PAC composition is not available and the similar product used to estimate the composition. 4. For crude oil, representative domestic and foreign oils can be used to calculate a weighted average PAC composition for crude oil. According to EIA4, 39.1 percent (volumetric basis) of the total consumed crude oil in the United States in the year 2000 was domestic while 60.9 percent (volumetric basis) was imported. EPA selected South Louisiana Oil, for which PAC composition is available, as a representative domestic oil and Alberta Oil as a representative foreign oil. EPA assumed that a weighted average of the composition of these two crude oils is a reasonable representation of crude oil composition for the purpose of this study. EPA also used a specific weight of 0.92 for crude oil to convert PAC concentrations reported as mg/kg to mg/L. 5. For refined products, EPA assumed a specific weight of 1.0 to simplify the calculation (i.e., no need to convert between mg/kg and mg/L). ' EIA: Petroleum Supply Annual 2000, Vol 1, Page 6 (EIA, 2001). 2-15 ------- Section 2 - Development of TRIReleases2007 Based on the above assumptions, EPA calculated the overall TWF using Equation 2-1 where the concentration of each of the 21 TRI PACs in each crude or finished petroleum product is multiplied by its respective TWF. The concentration of each PAC in petroleum crude oil or products is represented by the variable [m;]j and listed in Table 2-8 for products or Table 2-9 for crude oils. The products in each product produced by U.S. refineries are represented by the variable Q and listed in Table 2-10 while the respective TWFs are listed in Table 2-12. This calculation resulted in an equivalent TWF for each type of product supplied to U.S refineries. EPA then multiplied the mass of each type of product by the respective equivalent TWF and summed this quantity for all products received by U.S. refineries. Dividing this sum by the total sum of all products received by U.S. refineries resulted in an equivalent TWF for the PACs present in wastewaters from U.S. refineries. Overall TWF = (Eqn. 2-1) where: Quantity of Product V Supplied to U.S. Refineries Estimated concentration of PAC compound 'j' in Product T This calculation resulted in a TWF value of 25.417. The TWPE of the combined mass of PACs reported to TRI by petroleum refineries can then be calculated by multiplying the reported PAC releases by 25.417'. Table 2-8. PAC Concentrations in Petroleum Products PAC Chemical Name Benzo(a)anthracene Benzo(a)phenanthrene (chrysene) Benzo(a)pyrene Benzo(b)fluoranthene Benzo(j)fluoranthene Benzo(k)fluoranthene Benzo(j,k)fluorene (fluoranthene) Benzo(r,s,t)pentaphene Dibenz(a,h)acridine Dibenz(aj )acridine Dibenzo(a,h)anthracene Dibenzo(a,e)fluoranthene Dibenzo(a,e)pyrene Dibenzo(a,h)pyrene Dibenzo(a,l)pyrene 7H-Dibenzo(e,g)carbazole Gasoline mg/L 4.30 2.00 1.80 NP NP NP 6.50 NP NP NP NP NP NP NP NP NP Kerosene ppm (wt/vol) 0.75 2.00 0.50 0.75 NP 0.50 4.00 NP 0.20 NP 0.75 NP 0.45 1.00 NP NP No. 2 Diesel Fuels Bunker C No. 6 Oil Paving Asphalt mg/L or mg/kg 0.80 3.40 NP NP NP NP 2.80 NP NP NP NP NP NP NP NP NP 90.00 196.00 44.00 NP NP NP 240.00 NP NP NP NP NP NP NP NP NP 90.00 80.00 1.30 NP NP 1.80 NP NP NP NP 4.60 NP NP NP NP NP Lube Oil6 mg/kg 0.68 3.20 0.23 0.627 NP NP 2.00 NP NP NP NP NP NP NP NP NP 2-16 ------- Section 2 - Development of TRIReleases2007 Table 2-8. PAC Concentrations in Petroleum Products PAC Chemical Name 7, 12-Dimethylbenz(a)anthracene Indeno(a,2,3 -cd)pyrene 3 -Methy Icholanthrene 5-Methylchrysene 1-Nitropyrene Gasoline mg/L NP NP NP NP NP Kerosene ppm (wt/vol) NP 2.00 0.10 NP NP No. 2 Diesel Fuels Bunker C No. 6 Oil Paving Asphalt mg/L or mg/kg NP NP NP 6.00 NP NP NP NP NP NP NP NP NP NP NP Lube Oil6 mg/kg NP NP NP NP NP Source: Data compiled in the American Petroleum Institute's Transport and Fate of non-BTEXPetroleum Chemicals in Soil and Groundwater (API, 1994). NP - Not present. Table 2-9. PAC Concentrations in Crude Oils (mg/kg) PAC Chemical Name Benzo(a)anthracene Benzo(a)phenanthrene (chrysene) Benzo(a)pyrene Benzo(b)fluoranthene Benzo(j)fluoranthene Benzo(k)fluoranthene Benzo(j,k)fluorene (fluoranthene) Benzo(r,s,t)pentaphene Dibenz(a,h)acridine Dibenz(aj )acridine Dibenzo(a,h)anthracene Dibenzo(a,e)fluoranthene Dibenzo(a,e)pyrene Dibenzo(a,h)pyrene Dibenzo(a,l)pyrene 7H-Dibenzo(c,g)carbazole 7, 12-Dimethylbenz(a)anthracene Indeno(a,2,3 -cd)pyrene 3 -Methy Icholanthrene 5 -Methy Ichrysene 1-Nitropyrene South Louisiana Crude Oil 1.7000 17.5600 0.7500 0.5000 0.9000 1.3000 5.0000 NP NP NP NP NP NP NP NP NP NP NP NP NP NP Alberta Crude Oil NP 30.0000 NP 4.0000 NP NP 6.0000 NP NP NP NP NP NP NP NP NP NP NP 3.0000 NP NP Weighted Average 0.6645 25.1372 0.2932 2.6319 0.3518 0.5082 5.6091 NP NP NP NP NP NP NP NP NP NP NP 1.8273 NP NP Source: Data compiled in the American Petroleum Institute's Transport and Fate of non-BTEX Petroleum Chemicals in Soil and Groundwater (API, 1994). NP - Not present. 2-17 ------- Section 2 - Development of TRIReleases2007 Table 2-10. Supply and Disposition of Crude Oil and Petroleum Products Crude and Finished Petroleum Products Finished Motor Gasoline Reformulated Oxygenated Other Finished Aviation Gasoline Jet Fuel Naphtha-Type Kerosene-Type Kerosene Distillate Fuel Oil 0. 05% Sulfur and under Greater than 0. 05% sulfur Residual Fuel Oil Naphtha For Petroleum Feed Use Other Oils For Petroleum Feed Use Special Naphthas Lubricants Waxes Petroleum Coke Asphalt and Road Oil Still Gas Miscellaneous Products Total Products Crude Oil TOTAL VOLUME OF PRODUCTS & CRUDE OIL 1,000 bbl/year 2,910,056 939,493 42,221 1,928,342 6,543 587,974 75 587,899 23,860 1,310,158 905,064 405,094 254,843 74,039 71,762 21,868 65,687 6,478 266,107 192,223 241,365 19,957 6,052,920 5,514,395 11,567,315 % (Products Only) 48.08 NP NP NP 0.11 9.71 NP NP 0.39 21.65 NP NP 4.21 1.22 1.19 0.36 1.09 0.11 4.40 3.18 3.99 0.33 100 — — Volume % (Total) 25.16 NP NP NP 0.06 5.08 NP NP 0.21 11.33 NP NP 2.20 0.64 0.62 0.19 0.57 0.06 2.30 1.66 2.09 0.17 52.33 47.67 100 Source: Petroleum Supply Annual 2000, NP - Not present. Vol. 1, Page 6(EIA, 2001). 2-18 ------- Section 2 - Development of TRIReleases2007 Table 2-11. Products for Which PAC Composition Is Not Available Product Finished Aviation Gasoline Jet Fuel Miscellaneous Products Naphtha For Petroleum Feed Use Other Oils For Petroleum Feed Use Petroleum Coke Special Naphtha Still Gas Waxes PAC Composition Taken from: Gasoline Gasoline Gasoline Gasoline Gasoline Paving Asphalt Gasoline Gasoline Lube Oil Table 2-12. Calculation of Toxic Weighting Factor for Petroleum PACs Pollutant Benzo(a)anthracene Benzo(a)phenanthrene (Chrysene) Benzo(a)pyrene Benzo(b)fluoranthene Benzo(j)fluoranthene B enzo (k)fluoranthene Benzo(j,k)fluorene (Fluoranthene) Benzo(r,s,t)pentaphene Dibenz(a,h)acridine Dibenz(aj )acridine Dibenzo(a,h)anthracene Dibenzo(a,e)fluoranthene Dibenzo(a,e)pyrene Dibenzo(a,h)pyrene Dibenzo(a,l)pyrene 7H-Dibenzo(c,g)carbazole 7, 12-Dimethylbenz(a)anthracene Indeno( 1,2,3 -cd)pyrene 3 -Methy Icholanthrene 5 -Methy Ichrysene 1-Nitropyrene TWF 36.26 31.01 100.66 30.66 NA 30.66 1.2847 NA NA NA 30.66 NA NA NA NA NA NA 30.66 NA NA NA Chemical Percentage (%) 17.47 46.29 4.17 2.74 0.36 0.70 24.32 NP NP NP 0.43 NP NP NP NP NP NP 0.01 NP 3.50 NP Total Adjusted TWF 5.36 14.35 4.20 0.84 0.21 0.31 0.13 0.00 25.417 NA - Not applicable (No TWF Available). NP - Not present. 2-19 ------- Section 2 - Development of TRIReleases2007 Wood Preserving PACs EPA used data available from wood preserving facilities to calculate TWPE for discharges of PACs from wood preserving facilities (NAICS 321114, Wood Preservation). In 2005, 10 wood preserving facilities participated in a sampling program to determine the PACs released with their stormwater runoff. Over the period of a few months, the facilities collected grab samples of runoff during rainfall events. The 10 facilities collected a total of 74 samples. In 37 of these samples, at least one PAC was measured above the detection limit. EPA used the concentrations in these 37 samples to calculate a TWF for the PACs discharged from wood preserving facilities (H.M. Rollins, 2005). For all PAC concentrations reported as not detected, EPA assumed the concentration to be zero. Using the data provided, EPA calculated the average concentration of the six PAC compounds measured. EPA calculated the percentage of each compound relative to the total PACs. EPA calculated an adjusted TWF for each compound by multiplying its chemical-specific TWF by its percentage relative to the total PACs. EPA summed these values to calculate a new overall TWF value for PACs discharged by facilities in the wood preserving NAICS code. Table 2-13 lists TWFs for all PACs, the percent of total PACs, and the adjusted TWF for each PAC. Table 2-13. Calculation of Toxic Weighting Factor for Wood Preserving PACs Chemical Name Benzo(a)anthracene Benzo(a)phenanthrene(chrysene) Benzo(a)pyrene Benzo(b)fluoranthene Benzo(j)fluoranthene B enzo (k)fluoranthene Benzo(j,k)fluorene(fluoranthene) Benzo(r,s,t)pentaphene Dibenz(a,h)acridine Dibenz(aj )acridine Dibenzo(a,h)anthracene Dibenzo(a,e)fluoranthene Dibenzo(a,e)pyrene Dibenzo(a,h)pyrene Dibenzo(a,l)pyrene 7H-Dibenzo(e,g)carbazole 7, 12-Dimethylbez(a)anthracene Indeno(a,2,3 -cd)pyrene 3 -Methy Icholanthrene 5 -Methy Ichrysene 1-Nitropyrene Toxic Weighting Factor 36.2600 31.0100 100.6600 30.6600 NA 30.6600 1.2847 NA NA NA 30.6600 NA NA NA NA NA NA 30.6600 NA NA NA Chemical Percentage (%) 6.73 9.73 0.49 4.98 NP 0.78 77.29 NP NP NP NP NP NP NP NP NP NP NP NP NP NP Total PACs TWF Adjusted TWF 2.44 3.02 0.49 1.53 NP 0.24 0.989 NP NP NP NP NP NP NP NP NP NP NP NP NP NP 8.33 NA - Not available. NP - Not present. 2-20 ------- Section 2 - Development of TRIReleases2007 Pulp, Paper, and Paperboard PACs EPA used data available from pulp and paper mills to calculate TWPE for discharges of PACs from facilities in the Pulp, Paper, and Paperboard Category. The National Council of the Paper Industry for Air and Stream Improvement (NCASI) has provided guidance to the pulp, paper, and paperboard industry for PAC discharges (NCASI, 1988). The NCASI guidance includes a table listing the concentrations of PACs found in wastewaters for several pulping types (kraft, bisulfite, chemithermomechanical (CTMP), and thermal (TMP)). EPA determined that in the United States, there are few bisulfite, CTMP, and TMP mills compared to the number of kraft mills. Therefore, EPA used the kraft mill concentrations to calculate the PAC TWF. Since the NCASI guidance does not distinguish between effluents from mills with or without bleaching, the calculated TWF was used for mills in all pulp, paper, and paperboard ELG phases. NCASI calculated the emission factors for the industry based on six PACs: • Benzo(a)anthracene; • Benzo(a)pyrene; • Benzo(b+k) fluoranthene; • Dibenzo(a,h)anthracene; • Fluoranthene; and • Indeno(l,2,3-c,d)pyrene. For the kraft mills, only fluoranthene was detected above the method detection limit (MDL); however, four of the other five compounds were detected above the MDL for the other pulping types. Because the calculated TWF will be used for all facilities in the Pulp, Paper, and Paperboard Category, EPA used l/2 the detection limit for compounds that were not detected in kraft mill wastewaters. NCASI also calculated the emission factor using !/2 the detection limit for compounds that were not detected. EPA used the concentrations of six PACs to calculate a Pulp, Paper, and Paperboard Category PAC TWF. EPA summed the measured concentrations to calculate the total concentration of PACs in the effluent. EPA then calculated the percentage of each chemical relative to the total PACs in the effluent. EPA calculated an adjusted TWF for each compound by multiplying its chemical-specific TWF by its percentage relative to the total PACs. EPA summed these values to calculate a new overall TWF value for PACs discharged by facilities in the Pulp, Paper, and Paperboard Category. Table 2-14 presents the TWFs for the six PACs, the percentage of total PACs, and the adjusted TWF for each PAC. 2-21 ------- Section 2 - Development of TRIReleases2007 Table 2-14. Calculation of Toxic Weighting Factor for Pulp, Paper, and Paperboard PACs Chemical Name Benzo(a)anthracene Benzo(a)pyrene Benzo(b+k)fluoranthene Benzo(j,k)fluorene(fluoranthene) Dibenzo(a,h)anthracene Indeno( 1,2,3 -cd)pyrene Toxic Weighting Factor 36.2600 100.6600 30.6600 1.2847 30.6600 30.6600 Chemical Percentage (%) 11.74 11.74 11.74 17.84 23.47 23.47 Total PACs TWF Adjusted TWF 4.25 11.81 3.60 0.229 7.20 7.20 33.66 2.4.4 Metal Compounds For TRI reporting, facilities report metal compounds on a single reporting form for each parent metal and do not specify the individual compound(s) released. In addition, if the facility is required to report for a metal (e.g., zinc) and its compounds (e.g., zinc compounds), the facility may report both the metal and metal compound on a single form (reported as the metal compound). For metal compound reporting, the release quantities are based on the mass of the parent metal, only. To calculate TWPEs for metal compounds, EPA used the TWF for the parent metal. EPA then combined the TWPEs for the metal and metal compounds for ranking purposes (i.e., TWPE reported for "zinc and zinc compounds," rather than one TWPE for "zinc" and one TWPE for "zinc compounds"). This analysis does not double count metal discharges because all discharges are separated until the rankings are created. For example, if a facility reported 5 pounds of zinc and 10 pounds of zinc compounds, the discharges would be kept separate in the database. When the rankings are created however, the database would display that the facility has one entry of 15 pounds of "zinc and zinc compounds." 2.4.5 Determination of "Basis of Estimate " of Reported TRI Releases When reporting releases and transfers to TRI, facilities also indicate the basis for their estimate using six reporting codes: Ml: continuous monitoring data or measurements; M2: periodic or random monitoring data or measurements; C: mass balance calculations, such as calculation of the amount of the toxic chemical in streams entering and leaving process equipment; E: published emission factors; E2: site-specific emission factors; and O: other approaches, such as engineering calculations. EPA developed a table in TRICalculations2007 that contains the basis of estimate for direct discharges and indirect discharges (i.e., transfers to POTWs). This table is separate from the "TRI Master List" table. EPA used this table in TRIReleases2007 to summarize how releases are reported for certain NAICS codes and point source categories. 2-22 ------- Section 2 - Development of TRIReleases2007 2.5 TRIReleases2007 As the final step in developing TRIReleases2007, EPA grouped discharges from the "TRI Master List" table to create the point source category rankings and to perform other analyses. The remainder of this subsection describes the development of TRIReleases2007 and discusses preliminary results in the following order: • Section 2.5.1 discusses the NAICS/Point Source Category Crosswalk; and • Section 2.5.2 describes the development of the 2007 TRI rankings, including analysis of facilities with the highest TWPE, pollutants with the highest TWPE, and category prioritization. Table 2-15 lists the database tables that EPA created in TRIReleases2007. Table 2-15. Tables Created in TRIReleases2007 Table Name "Counts of Facilities by NAICS" "Counts of Facilities by PSC" "Point Source Rankings" "NAICS Code Rankings" Description Includes counts of direct dischargers, indirect dischargers, facilities that discharge both directly and indirectly, total dischargers, and total facilities reporting to TRI by NAICS code. Similar to table "Counts of Facilities by NAICS"; however, it reports the counts by point source category. Presents rankings for all point source categories based on calculated TWPEs. TWPEs were calculated using the total discharges to surface water by direct dischargers and transfers to POTWs by indirect dischargers, taking into account pollutant removal occurring at the POTWs. Presents rankings for all NAICS codes based on calculated TWPEs. TWPEs were calculated using the total discharges to surface water by direct dischargers and transfers to POTWs by indirect dischargers, taking into account pollutant removal occurring at the POTW. EPA also imported or linked the following tables from TRICalculatiom2007: • "Dioxin Distributions"; "NAICS_Codes"; • "Parent Metals and Compounds"; • "Pesticide Chemical List"; "PSC Codes"; "PSC/NAICS Crosswalk"; "TRI Master List"; • "TRI Master Facility List"; and "TWFs" 2-23 ------- Section 2 - Development of TRIReleases2007 2.5.1 NAICS/Point Source Category Crosswalk EPA has developed ELGs for 56 specific categories of industrial dischargers. The categories, which may be divided into subcategories, are generally defined in terms of combinations of products made and the processes used to make these products. Facilities with data in TRI are identified by NAICS code. Thus, to use TRI data to estimate the pollutants discharged by each point source category, EPA assigned each 6-digit NAICS code to an appropriate point source category using the "NAICS/Point Source Category Crosswalk" table. Section 1 of this report discusses the crosswalk in more detail. 2.5.2 Development of 2007 TRI Rankings Figure 2-2 presents the TRIReleases2007 database structure, including fields used from each data source. The NAICS codes in the "TRI Master List" table are specific to each facility and each discharge. This allows EPA to make NAICS adjustments to differentiate between various operations at one facility. The default NAICS code is the primary facility NAICS code reported in TRI. For the development of the rankings, EPA associated the NAICS codes with the appropriate point source categories using the "NAICS/Point Source Category Crosswalk" and the "Point Source Category Codes" tables. The TWPE for each discharge was calculated previously in TRICalculations2007 (see Section 2.4). TRIReleases2007 groups releases by chemical, facility, and point source category to allow EPA to perform the following analyses. TRI Master List NAICS Code TRI Facility ID CAS Number Chemical Name Total Pounds Released TWPE Direct/Indirect Discharge Indicator k NAICS/Point Source Category Crosswalk NAICS Code Point Source Category Code Point Source Category Codes Point Source Category Code Category Name Type of Group Figure 2-2. Basic Structure of the TRIReleases2007 Database Top Facilities Analysis. EPA created a table that ranks facilities according to the TWPE discharged by the entire facility. This table also identifies the chemical that contributed the greatest amount of TWPE to the total facility TWPE. EPA used the table to identify facilities with unusually high reported discharges relative to other facilities in an industrial category. As discussed in Section 6, EPA contacted these facilities to learn more about their reported releases. Section 6 also presents EPA's findings about the top facilities' reported releases. Top Pollutants Analysis. EPA created a table that ranks pollutants discharged according to the TWPE discharged by all facilities reporting in TRIReleases2007. The table also includes the number of facilities that report releasing the chemical. Using this analysis, EPA identified pollutants or pollutant categories for further analysis (e.g., metals). Category Prioritization. EPA uses point source category rankings to identify categories that may warrant further review. 2-24 ------- Section 2 - Development of TRIReleases2007 2.6 Results of the Preliminary Analysis of the TRIReleases2007 Database This section presents the results of the analysis of TRIReleases2007 database. Table 2-16 presents the point source category rankings by TWPE. Table A-l in Appendix A presents the six-digit NAICS code rankings by TWPE. Table A-2 in Appendix A presents the total TWPE for chemicals in TRI. See Section 5 of the 2009 TSD for EPA's 2009 annual review (U.S. EPA, 2009). Table 2-16. Point Source Category Rankings 40CFR Part 414.1a 414 423 430 419 420 433 415 440 421 432 458 455 429 417 NA 471 463 439 428 425 469 NA 464 468 NA 418 437 413 Point Source Category Chlorine And Chlorinated Hydrocarbons Organic Chemicals, Plastics And Synthetic Fibers Steam Electric Power Generating Pulp, Paper And Paperboard Petroleum Refining Iron And Steel Manufacturing Metal Finishing Inorganic Chemicals Manufacturing Ore Mining And Dressing Nonferrous Metals Manufacturing Meat and Poultry Products Carbon Black Manufacturing Pesticide Chemicals Timber Products Processing Soap And Detergent Manufacturing National Security & International Affairs Nonferrous Metals Forming And Metal Powders Plastics Molding And Forming Pharmaceutical Manufacturing Rubber Manufacturing Leather Tanning And Finishing Electrical And Electronic Components Miscellaneous Foods And Beverages Metal Molding And Casting (Foundries) Copper forming Tobacco Products Fertilizer Manufacturing Centralized Waste Treatment Electroplating Number of Facilities 28 594 284 198 232 190 2047 142 28 107 144 7 67 107 58 43 105 121 96 182 19 87 133 184 116 21 29 34 352 Total Discharge before POTW Removal 1,500,000 72,500,000 2,160,000 34,900,000 16,600,000 41,500,000 25,800,000 26,900,000 324,000 3,560,000 45,100,000 356 2,250,000 210,000 675,000 15,000,000 12,200,000 15,000,000 5,750,000 1,880,000 634,000 11,300,000 9,520,000 1,690,000 288,000 203,000 3,240,000 2,340,000 8,670,000 Total Pounds Released b 835,000 19,200,000 2,150,000 15,800,000 13,700,000 39,500,000 3,980,000 5,870,000 319,000 2,670,000 41,400,000 356 1,450,000 32,500 69,300 14,900,000 1,330,000 2,140,000 1,510,000 865,000 318,000 3,210,000 5,810,000 204,000 35,500 189,000 3,190,000 448,000 886,000 TWPE 7,270,000 575,000 542,000 460,000 172,000 104,000 62,000 54,700 44,400 38,900 35,900 32,400 24,700 16,300 14,600 14,500 8,830 8,780 8,000 7,860 7,800 7,550 6,580 6,110 4,950 4,760 4,460 3,790 3,210 2-25 ------- Section 2 - Development of TRIReleases2007 Table 2-16. Point Source Category Rankings 40CFR Part 407 467 436 405 410 406 461 438 426 434 411 424 422 443 465 408 466 446 NA 445 NA 454 444 NA 409 447 457 NA NA NA NA NA NA NA NA NA Point Source Category Canned And Preserved Fruits And Vegetables Processing Aluminum forming Mineral Mining And Processing Dairy products processing Textile Mills Grain mills Battery Manufacturing Metal Products And Machinery Glass Manufacturing Coal Mining Cement Manufacturing Ferroalloy Manufacturing Phosphate Manufacturing Paving And Roofing Materials (Tars And Asphalt) Coil Coating Canned And Preserved Seafood Processing Porcelain Enameling Paint Formulating Printing & Publishing Landfills Justice, Public Order, & Safety Gum And Wood Chemicals Manufacturing Waste Combustors Independent And Stand Alone Labs Sugar Processing Ink Formulating Explosives Manufacturing Apparel & Other Textile Products Miscellaneous Retail Wholesale Trade- Nondurable Goods Wholesale Trade- Durable Goods Coal Mining Engineering & Management Services Business Services Drinking Water Treatment Trucking & Warehousing Number of Facilities 20 115 60 243 63 23 62 32 64 14 36 4 11 19 50 8 5 49 65 13 1 10 8 7 o J 8 9 2 1 1 5 1 1 2 2 1 Total Discharge before POTW Removal 4,370,000 2,000,000 2,410,000 20,700,000 2,830,000 10,700,000 1,180,000 116,000 1,510,000 245,000 27,900 2,350 16,200 1,330 67,300 312,000 3,430 1,130,000 370,000 69,500 31.2 3,020 18,300 9,660 72,900 4,500 17,300 6,710 7 44,600 2,990 16.6 720 95 681 66 Total Pounds Released b 3,760,000 304,000 1,800,000 3,170,000 1,170,000 1,800,000 120,000 15,700 185,000 245,000 3,410 2,300 16,100 227 21,600 312,000 2,180 91,500 31,800 22,400 31.2 507 18,300 2,930 23,700 573 16,200 4,090 1.58 4,460 307 16.6 371 9.46 171 40.3 TWPE 2,960 2,710 2,420 2,400 2,390 2,080 1,640 917 546 493 452 340 250 249 241 234 164 140 110 82.7 69.9 54.8 39.6 30.0 25.5 20.0 13.6 4.61 3.54 3.33 2.51 0.458 0.441 0.294 0.29 0.0447 2-26 ------- Section 2 - Development of TRIReleases2007 Table 2-16. Point Source Category Rankings 40CFR Part NA NA Point Source Category Misc. Manuf. Industries Food & Kindred Products Number of Facilities 1 1 Total Discharge before POTW Removal 5 0.004 Total Pounds Released b 5 0.000784 TWPE 0.0281 0.00013 Source: TRIReleases2007_v2. a 414.1 refers to the chlorinated hydrocarbon segment of the Organic Chemicals, Plastics, and Synthetic Fibers Category (40 CFR Part 414) and the Chlor-Alkali Subcategory of the Inorganic Chemicals Manufacturing Category (40 CFR Part 415). b Discharges include transfers to POTWs and account for POTW removals. NA - Not applicable; no existing ELGs apply to discharges. 2.6.1 Metals Analysis For the 2009 screening-level analysis, EPA gave special consideration to reported discharges of metals. Releases of metals from industrial facilities may be associated with current operations or may be from cleanup actions for past practices. If releases are not related to current operations, they are not useful in reviewing the ELGs intended to control discharges from current operations. EPA identified the following metals for further analysis as part of the 2009 annual review, based on total TWPE calculated by TRIReleases2007: • Manganese; • Arsenic; and • Copper. Table 2-17 presents all the metals reported in TRI 2007 ranked by TWPE, including the number of facilities reporting discharges and the pounds discharged. In 2007, 4,428 facilities reported discharging 17 metals5. The total metals discharges after accounting for POTW removals, as appropriate, was 1,240,000 TWPE, which represented 12.9 percent of total TRI TWPE for 2007. Manganese discharges were the largest metals discharges, as measured by TWPE, accounting for almost 25 percent of the total metals TWPE. Arsenic and copper were also significant contributors, with discharges of each accounting for greater than 18 percent of the total metals TWPE. 5 TRI 2007 includes release information for the following metals, including their metal compounds: antimony, arsenic, barium, beryllium, cadmium, chromium, cobalt, copper, lead, manganese, mercury, nickel, selenium, silver, thallium, vanadium, and zinc. 2-27 ------- Section 2 - Development of TRIReleases2007 Table 2-17. Metals Discharged by TWPE in TRI 2007 Chemical Name Manganese and Manganese Compounds Arsenic and Arsenic Compounds Copper and Copper Compounds Lead and Lead Compounds Mercury and Mercury Compounds Zinc and Zinc Compounds Selenium and Selenium Compounds Silver and Silver Compounds Cadmium and Cadmium Compounds Nickel and Nickel Compounds Cobalt and Cobalt Compounds Vanadium and Vanadium Compounds Chromium and Chromium Compounds Barium and Barium Compounds Thallium and Thallium Compounds Antimony and Antimony Compounds Beryllium and Beryllium Compounds Total Number of Facilities 965 145 1,688 2,401 331 1,297 49 42 46 1,368 204 148 1,144 353 10 163 10 4,428 Total Pounds Released a 4,330,000 55,800 354,000 77,000 790 865,000 32,000 2,170 1,420 254,000 194,000 376,000 74,600 860,000 1,630 19,600 89.9 7,490,000 TWPE after POTW Removals (Ib-eq/yr) 305,000 226,000 225,000 173,000 92,500 40,600 35,800 35,700 32,900 27,600 22,100 13,200 5,650 1,710 1,680 240 95.0 1,240,000 TWPE percent of Total Metals TWPE 24.7 18.2 18.1 13.9 7.48 3.28 2.89 2.88 2.66 2.23 1.79 1.06 0.46 0.14 0.14 0.02 0.008 100 TWPE percent of Nationwide TWPE 3.19 2.36 2.35 1.80 0.97 0.42 0.37 0.37 0.34 0.29 0.23 0.14 0.06 0.02 0.02 0.003 0.001 12.9 Source: TRIReleases2007_v2. a Discharges include transfers to POTWs and account for POTW removals. Table 2-18 lists the facilities reporting discharges of greater than 6,000 TWPE of manganese, arsenic, or copper in TRI 2007. Manganese is commonly found in discharges from pulp and paper facilities. All but one of the top facilities (Tronox, LLC) discharging manganese are pulp and paper mills. In the Final Report on the Pulp, Paper, and Paperboard Detailed Study (U.S. EPA, 2006), EPA identified manganese and aluminum as the top metals of concern from pulp and paper mills. EPA reviewed manganese discharges from Tronox LLC as part of the 2009 annual review (Freeze, 2009a; Freeze, 2009b) and determined that manganese in the discharges from Tronox LLC originated in the titanium dioxide manufacturing process onsite known as the chloride process (see Section 7 of the Preliminary 2010 Effluent Guidelines Program Plan Technical Support Document (U.S. EPA, 2009)). Of the 17 metals reported to TRI, manganese is ranked 14th in terms of relative toxicity. Manganese TWPE discharges contributed 24.7 percent to the total metals TWPE to TRI in 2007. 2-28 ------- Section 2 - Development of TRIReleases2007 Table 2-18. Facilities Reporting Discharges of Metals with the Highest TWPE in TRI 2007 Chemical Manganese and Manganese Compounds Arsenic and Arsenic Compounds Facility Domtar Industries Inc. Ashdown Mill Rayonier Performance Fibers Jesup Mill Alabama River Pulp Co In C. Tronox LLC Brunswick Cellulose Inc Georgia-Pacific Crossett Operations Meadwestvaco Texas L.P. Georgia-Pacific Consumer Products LP Bowater Inc -Catawba Operations Georgia-Pacific Consumer Products LLC Georgia-Pacific Brewton LLC Chesterfield Power Station U.S. TVA Johnsonville Fossil Plant U.S. TVA Widows Creek Fossil Plant Duke Energy Corp Wabash River Generating Station U.S. TVA Kingston Fossil Plant Detroit Edison Monroe Power Plant Gaston Steam Plant Kentucky Utilities Co. - E. W. Brown Station Kentucky Utilities Co Ghent Station Eastman Chemical Co Tennessee Operations American Electric Power Cardinal Plant Barry Steam Plant Cliffside Steam Station Point Source Category Pulp, Paper and Paperboard Pulp, Paper and Paperboard Pulp, Paper and Paperboard Inorganic Chemicals Manufacturing Pulp, Paper and Paperboard Pulp, Paper and Paperboard Pulp, Paper and Paperboard Pulp, Paper and Paperboard Pulp, Paper and Paperboard Pulp, Paper and Paperboard Pulp, Paper and Paperboard Steam Electric Power Generating Steam Electric Power Generating Steam Electric Power Generating Steam Electric Power Generating Steam Electric Power Generating Steam Electric Power Generating Steam Electric Power Generating Steam Electric Power Generating Steam Electric Power Generating Organic Chemicals, Plastics and Synthetic Fibers Steam Electric Power Generating Steam Electric Power Generating Steam Electric Power Generating CFR Citation 40 CFR Part 430 40 CFR Part 430 40 CFR Part 430 40 CFR Part 415 40 CFR Part 430 40 CFR Part 430 40 CFR Part 430 40 CFR Part 430 40 CFR Part 430 40 CFR Part 430 40 CFR Part 430 40 CFR Part 423 40 CFR Part 423 40 CFR Part 423 40 CFR Part 423 40 CFR Part 423 40 CFR Part 423 40 CFR Part 423 40 CFR Part 423 40 CFR Part 423 40 CFR Part 4 14 40 CFR Part 423 40 CFR Part 423 40 CFR Part 423 Total TWPE (Ib- eq/yr) 15,700 8,450 8,450 8,110 7,750 7,350 7,160 6,830 6,740 6,470 6,000 18,200 17,000 15,800 13,700 10,900 9,700 9,300 9,110 8,850 8,810 8,490 8,080 6,540 2-29 ------- Section 2 - Development of TRIReleases2007 Table 2-18. Facilities Reporting Discharges of Metals with the Highest TWPE in TRI 2007 Chemical Copper and Copper Compounds Facility Great River Energy Stanton Station Georgia Power Scherer Steam Electric Generating Plant Chesterfield Power Station Bowen Steam Electric Generating Plant U.S. TVA Paradise Fossil Plant American Electric Power Kammer / Mitchell Plants Point Source Category Steam Electric Power Generating Steam Electric Power Generating Steam Electric Power Generating Steam Electric Power Generating Steam Electric Power Generating Steam Electric Power Generating CFR Citation 40 CFR Part 423 40 CFR Part 423 40 CFR Part 423 40 CFR Part 423 40 CFR Part 423 40 CFR Part 423 Total TWPE (Ib- eq/yr) 12,700 10,800 10,200 7,620 6,980 6,380 Source: TRIReleases2007 v2. 2-30 ------- Section 2 - Development of TRIReleases2007 Arsenic and copper are commonly found in discharges from steam electric power generating facilities. In EPA's analysis of pollutants found in discharges from the steam electric industry, EPA found that arsenic and copper are two of the top pollutants reported to PCS/ICIS- NPDES and TRI. All but one of the top facilities (Eastman Chemical Co Tennessee Operations) discharging arsenic are steam electric power generating facilities. Of the 17 metals reported to TRI, arsenic is ranked 4th in terms of relative toxicity. Arsenic TWPE discharges contributed 18.2 percent to the total metals TWPE. All of the top facilities discharging copper are steam electric power generating facilities. Of the 17 metals reported to TRI, copper is ranked 9th in terms of relative toxicity. Copper TWPE discharges contributed 18.1 percent to the total metals TWPE. Conclusions • A total of 4,428 facilities reported discharging 7,490,000 pounds, including transfers to POTWs and accounting for POTW removals, and 1,240,000 TWPE of metals, accounting for 12.9 percent of total nationwide TRI 2007 TWPE. • Manganese, arsenic, and copper are the metals with the highest TWPE contributions to the total nationwide metals TRI 2007 TWPE. • All but one of the top 11 facilities discharging manganese (ranked by TWPE) are pulp and paper mills. • All but one of the top 13 facilities discharging arsenic (ranked by TWPE) are steam electric power generating facilities. • All of the top 6 facilities discharging copper (ranked by TWPE) are steam electric power generating facilities. 2.7 Development of TRIReleases2007 References 1. American Petroleum Institute (API). 1994. Transport and Fate of non-BTEX Petroleum Chemicals in Soil and Groundwater - Appendix A. API Publication Number 4593. (September). EPA-HQ-OW-2003-0074-0312. 2. Bicknell, Betsy. 2004. Calculation of TWF & TWPE for Creosote from Wood Preserving Plants. (August). EPA-HQ-OW-2003-0074-123 8. 3. Codding, Ellie and Deb Bartram. ERG. 2005. Memorandum to 2006 Effluent Guidelines Program Plan Docket, EPA Docket Number OW-2004-0032. RE: Publicly Owned Treatment Works (POTW) Percent Removals Used for the TRIReleases2002 Database. (August 12). EPA-HQ-OW-2004-0032-0967. 4. Eastern Research Group, Inc. (ERG). 2005. Draft Toxic Weighting Factor Development in Support of the 304(m) Planning Process. (July 29). EPA-HQ-OW-2004-0032-0857. 5. Energy Information Administration (EIA). Office of Oil and Gas. 2001. Petroleum Supply Annual 2000, Vol. 1, Page 6. (June). EPA-HQ-OW-2004-0032-0242. 6. Finseth, Thomas. 2005. TRI 2002 PACs TWF for Petroleum Refining, Creosote, and Wood Preservers. (January). DCN 01552. 2-31 ------- Section 2 - Development of TRIReleases2007 1. Freeze, Terry. 2009a. Notes from Email Correspondence between Chris Krejci, Eastern Research Group, Inc., and Terry Freeze, Tronox Hamilton. (March 23). EPA-HQ-OW- 2008-0517 DCN 06406. 8. Freeze, Terry. 2009b. Notes from Telephone Communication between Chris Krejci, Eastern Research Group, Inc., and Terry Freeze, Tronox Hamilton. RE: Verification of Magnitude and Basis of Estimate for Manganese Discharges Reported to TRI. (March 12). EPA-HQ-OW-2008-0517 DCN 6405. 9. H.C. Lavallee, Inc., 2005. Table 5 PAC Concentrations in Pulp Mill Effluents, Excerpt from NCASI's SARA Handbook, Enclosure 10. (April). EPA-HQ-OW-2004-0032-0498. 10. H.M Rollins. 2005. Letter to Carey Johnston, U.S. EPA. RE: Creosote Wood Treating Industry Storm Water Runoff Study Conducted on Behalf of the Southern Pressure Treaters Association and Creosote Council III. (May 26). EPA-HQ-OW-2004-0032- 0301. 11. H.M. Rollins Company, Inc. 2005. Study ofPolycyclic Aromatic Compounds in Storm Water Runoff. Prepared for: Southern Pressure Treaters Association and Creosote Council III. Gulfport, MS. (April 29). EPA-HQ-OW-2004-0032-0294. 12. Johnston, Carey. USEPA. 2004. Memorandum to Public Record for the Effluent Guidelines Program Plan for 2006/2007. RE: November 18, 2004, Meeting Between EPA and Wood Preserving Industry Representatives. Washington, D.C. (November 18). EPA- HQ-OW-2004-0032-0306. 13. Kashyap, Vikram. 2009. Notes from Telephone Communication and E-mail Correspondence between Vikram Kashyap, Louisiana Pigments, and Eleanor Codding, Eastern Research Group, Inc. RE: 2007 TRI Dioxin Water Releases. (May 8). EPA-HQ- OW-2008-0517 DCN 06849. 14. Moeller, Greg. 2009. Notes from Telephone Communication and E-mail Correspondence between Greg Moeller, Eastman Kodak, and Christopher Krejci, Eastern Research Group, Inc. RE: Basis of Estimate for TRI Discharges Reported to EPA. (March 16). EPA-HQ- OW-2008-0517 DCN 06409. 15. National Council of the Paper Industry for Air and Stream Improvement (NCASI). 1988. Handbook of Chemical-Specific Information for SARA Section 313 Form R Reporting. Table 5 PAC Concentrations in Pulp Mill Effluents. EPA-HQ-OW-2004-0032-0498. 16. U.S. EPA. 2000. Guidance for Reporting Toxic Chemicals Within the Dioxins and Dioxin-Like Compounds Category. Washington, D.C. (December). EPA-HQ-OW-2003- 0074. 17. U.S. EPA. 2001. The Emergency Planning and Community Right-to-Know Act: Section 313 Release and Other Waste Management Reporting Requirements. EPA 260-K-01-001. Washington, D.C. (February). EPA-HQ-OW-2008-0517 DCN 01996. 2-32 ------- Section 2 - Development of TRIReleases2007 18. U. S. EPA. 2004. Technical Support Document for the 2004 Effluent Guidelines Program Plan. EPA-821-R-04-014. Washington, DC. (August). EPA-HQ-OW-2003-0074-1346 through 1352. 19. U.S. EPA. 2006. Final Report: Pulp, Paper, and Paperboard Detailed Study. Washington, D.C. (November). EPA-821-R-06-016. EPA-HQ-OW-2004-0032-2249. 20. U. S. EPA. 2009. Preliminary 2010 Effluent Guidelines Program Plan Technical Support Document. Washington, D.C. EPA-821-R-09-006. (October). EPA-HQ-OW-2008-0517 DCN 06703. 21. Woodruff, Al. 2007. Notes from Telephone Communication and E-mail Correspondence between Al Woodruff, Cahaba Pressure Treated Forest Products, Inc., and Christopher Krejci, Eastern Research Group, Inc. RE: Dioxin Distribution Reported at Cahaba Pressure Treated Forest Products. (December 12). EPA-HQ-OW-2006-0771-1165. 22. World Health Organization (WHO). International Agency for Research on Cancer (IARC). 1998. Volume 35 Polynuclear Aromatic Compounds, Part 4, Bitumens, Coal- tars and Derived Products, Shale-oils and Soots. "Summary of Data Reported and Evaluation." (April 20). EPA-HQ-OW-2004-0032-2394. 23. Zweig, Greg. 2000. DuPont Memphis Dioxin Analytical Data. (February). EPA-HQ-OW- 2004-0032-1151. 2-33 ------- Section 3 - DMRLoads2007: Development and Category Rankings 3. DMRLOADS2007: DEVELOPMENT AND CATEGORY RANKINGS As discussed in Section 1, EPA annually reviews promulgated effluent limitations guidelines and pretreatment standards (ELGs) and investigates available information on industrial pollutant discharges. EPA identified that the discharge monitoring reports (DMRs) in the Permit Compliance System (PCS) and the Integrated Compliance Information System for the National Pollutant Discharge Elimination System (ICIS-NPDES) contain readily available and relevant data on industrial pollutant discharges to surface waters ("direct discharges"). Neither PCS nor ICIS-NPDES has information on pollutant discharges to Publicly Owned Treatment Works (POTWs) ("indirect discharges"). Consequently, EPA was able to use DMR data in PCS and ICIS-NPDES for its review of: (1) promulgated effluent guidelines ("direct discharges"); and (2) direct industrial pollutant discharges not currently subject to effluent guidelines. Due to the limitations of PCS and ICIS-NPDES, EPA was unable to use these two data systems to review promulgated pretreatment standards or indirect industrial pollutant discharges not currently subject to pretreatment standards. As discussed in Section 7, EPA combined the toxic-weighted pound equivalent (TWPE) calculated from the DMR data contained in PCS and ICIS-NPDES and Toxics Release Inventory (see Section 2 for information about TRI). EPA used this combined TWPE to prioritize its review of industry sectors to offer the greatest potential for reducing hazard to human health or the environment. This section describes how EPA compiled DMR data from PCS and ICIS-NPDES into the database DMRLoads2007 to estimate the mass and relative toxicity of pollutants discharged by industry categories. DMRLoads2007 compiles information for all facilities classified as major dischargers in PCS and ICIS-NPDES for reporting year 2007 and for the point source categories that these facilities represent. Tables B-l and B-2 in Appendix B list annual loads and TWPE calculated by DMRLoads2007 presented by 4-digit Standard Industrial Classification (SIC) code and pollutant, respectively. The remainder of Section 3 is organized in the following subsections: • Section 3.1- Overview ofDMRLoads2007; • Section 3.2 - DMRLoads2007: Database Development and Methodology; • Section 3.3 - Results of the Preliminary Analysis; • Section 3.4 - Data Quality Review; and • Section 3.5- DMRLoads2007 References. 3.1 Overview of DMRLoads2007 EPA's Office of Enforcement and Compliance Assistance (OECA) stores DMR data in national databases. EPA has used these DMR data as a part of its screening level review of existing effluent guidelines since the 2003 annual review (68 FR 75515, December 31, 2003). Historically, OECA stored DMR data in PCS, but in 2006 began storing certain states' data in ICIS-NPDES. Therefore the 2009 annual review of nationwide discharges required two sets of data, which EPA merged to create DMRLoads2007'. 3.1.1 NPDES Permitting and Reporting Requirements As authorized by the CWA, the NPDES program controls water pollution by regulating point sources that discharge pollutants directly into waters of the United States. Specifically, Federal Water Pollution Control Act Title IV, Permits and Licenses, of the Federal Water Pollution Control Act created the NPDES system for permitting wastewater discharges (CWA ------- Section 3 - DMRLoads2007: Development and Category Rankings Section 402). The Water Permits Division within EPA's Office of Wastewater Management leads and manages the NPDES permit program in partnership with EPA Regional Offices, states, tribes, and other stakeholders. Industrial, municipal, and other facilities must obtain NPDES permits if they discharge directly to surface waters. In most cases, authorized states administer the NPDES permit program. More than 65,000 industrial facilities and municipal wastewater treatment plants have obtained permits for discharges of regulated pollutants. Permitted facilities are required to file DMRs that include permit monitoring data (e.g., pollutant concentration/quantity, flow) to the appropriate regulating authority. The majority of NPDES permits are issued to direct point source dischargers (i.e., those entities that discharge directly into the receiving water body). PCS and ICIS-NPDES contain only limited data for indirect dischargers (i.e., those entities that discharge to POTWs). To provide an initial framework for setting permit issuance priorities, EPA developed a major/minor classification system for industrial and municipal wastewater dischargers. Facilities are classified as major based on an assessment of six characteristics: 1. Toxic pollutant potential; 2. Flow/stream flow volume; 3. Conventional pollutant loading; 4. Public health impact; 5. Water quality factors; and 6. Proximity to coastal waters. Each permitting authority establishes its own specific definitions based on the above characteristics, but major dischargers have the capability to impact receiving waters if not controlled, and, therefore, receive more regulatory attention than minor dischargers. DMR data for approximately 6,600 major facilities are in PCS and ICIS-NPDES for 2007. Facilities with major discharges must demonstrate compliance with NPDES permit limits by submitting monthly DMRs to the permitting authority. The permitting authority enters the reported DMR data into PCS or ICIS-NPDES, including the type of violation (if any), measured concentration and quantity values, and Quarterly Non-Compliance Report indicators. EPA does not require permitting authorities to enter DMR data for minor dischargers into PCS and ICIS- NPDES. Therefore, these databases contain only very limited data for minor dischargers. Table 3-1 identifies states and territories with data in PCS versus ICIS-NPDES at the time DMRLoads2007 was created. Note that three states were in the process of migrating from PCS to ICIS-NPDES in 2007 and had data in both systems. EPA created the database DMRLoads2007 to combine the two systems (PCS and ICIS-NPDES) and generate industrial category rankings for all U.S. states and territories. 3-2 ------- Section 3 - DMRLoads2007: Development and Category Rankings Table 3-1. States and Territories Included in DMRLoads2007 State/Province a Alabama Alaska American Samoa Arizona Arkansas ° California Colorado Connecticut Delaware District of Columbia Florida Georgia Guam Hawaii Idaho Illinois b Indiana Iowa Kansas Kentucky Louisiana Maine Maryland Massachusetts Michigan Minnesota Missouri Northern Mariana Islands Database PCS ICIS-NPDES ICIS-NPDES PCS PCS PCS PCS ICIS-NPDES PCS ICIS-NPDES PCS ICIS-NPDES ICIS-NPDES ICIS-NPDES ICIS-NPDES ICIS-NPDES /PCS ICIS-NPDES PCS PCS PCS PCS PCS ICIS-NPDES ICIS-NPDES PCS PCS PCS ICIS-NPDES State/Territory a Montana Mississippi North Dakota Nebraska13 North Carolina New Hampshire New Jersey New Mexico New York Nevada Ohio Oklahoma Oregon Pennsylvania Rhode Island South Carolina South Dakota Puerto Rico Tennessee Texas Utah Vermont Virgin Islands of the U.S. Virginia Washington Wisconsin West Virginia Wyoming Database ICIS-NPDES PCS PCS ICIS-NPDES / PCS PCS ICIS-NPDES PCS ICIS-NPDES ICIS-NPDES ICIS-NPDES PCS PCS PCS ICIS-NPDES ICIS-NPDES PCS ICIS-NPDES ICIS-NPDES PCS PCS ICIS-NPDES PCS ICIS-NPDES PCS PCS PCS PCS PCS a 2007 DMR data were not available for the following territories/tribes and were not included in DMRLoads2007: Atlantic Offshore, Canal Zone, Federal Micronesia, George's bank, Gulf of Mexico East, Johnson Atoll, Marshall Islands, Midway Islands, Palau, Saint Regis Tribe, Trust Territory, and U.S. Minor Islands. b Indicates states that were in the process of migrating from PCS to ICIS-NPDES in the year 2007. Some facilities in these states have DMR data in either PCS or ICIS-NPDES, while some facilities have DMR data in both PCS and ICIS-NPDES. For facilities with data in both databases for 2007, EPA used the DMR data from ICIS-NPDES (see Section 3.2.4.1). c Indicates states that had DMR data in PCS for 2007 and have since migrated their DMR data to ICIS-NPDES. 3.1.2 Overview of PCS and ICIS-NPDES Both PCS and ICIS-NPDES automate entering, updating, and retrieving NPDES data and tracking permit issuance, permit limits, monitoring data, and other data pertaining to facilities regulated by the NPDES program. Major dischargers are required to submit effluent monitoring data to the permitting authority on DMR. The permitting authority then enters these data into PCS or ICIS-NPDES and evaluates them for compliance with the NPDES permit requirements. ------- Section 3 - DMRLoads2007: Development and Category Rankings Permit limits include water quality parameters (e.g., dissolved oxygen and temperature), specific chemicals (e.g., phenol), bulk parameters (e.g., biochemical oxygen demand), and flow. Facilities report pollutant discharges in their DMR as mass-based quantities and/or concentrations using a wide variety of units. PCS and ICIS-NPDES also include information on the facility's permit requirements, such as monitoring frequency. 3.1.2.1 Utility of PCS and ICIS-NPDES The data stored in PCS and ICIS-NPDES are particularly useful for the annual review process for the following reasons: • PCS and ICIS-NPDES are national in scope, including data from all 50 states and 19 U.S. territories/tribes6; • Discharge reports included in PCS and ICIS-NPDES are based on effluent chemical analysis and metered flow; • PCS and ICIS-NPDES collectively include direct discharging facilities in all point source categories; and • PCS and ICIS-NPDES include data on conventional pollutants for most facilities and for the nutrients nitrogen and phosphorous for many facilities. 3.1.2.2 Limitations of PCS and ICIS-NPDES Limitations of the data collected in the PCS and ICIS-NPDES databases include the following: • The databases contain data only for pollutants a facility is required by permit to monitor; the facility is not required to monitor or report all pollutants actually discharged. • The databases include very limited discharge monitoring data from minor dischargers. • The databases include very limited data characterizing indirect discharges from industrial facilities to POTWs. • Many of the pollutant parameters included in PCS and ICIS-NPDES are not chemical compounds (e.g., "total Kjeldahl Nitrogen," "oil and grease") and cannot have toxic weighting factors (TWFs). • In some cases, the databases identify the type of wastewater being discharged; however, most reported flow rates do not indicate the type of wastewater and therefore, total flow rates reported to PCS and ICIS-NPDES may include stormwater and noncontact cooling water, as well as process wastewater. • Facilities are identified by SIC code, not point source category. For some SIC codes, it may be difficult or impossible to identify the point source category that is the source of the reported wastewater discharges7. 6 The following territories' data/regions are not available in PCS or ICIS-NPDES for 2007: Atlantic Offshore, Canal Zone, Federal Micronesia, George's Bank, Gulf of Mexico East, Johnson Atoll, Marshall Islands, Midway Islands, Palau, Saint Regis Tribe, Trust Territory, and U.S. Minor Islands. 7 ICIS-NPDES includes a data field for entering the applicable ELG (i.e. 40 CFR Part 423 for the Steam Electric Power Generating Category). However, entering data into this field is not required and therefore this field is typically not populated. 3-4 ------- Section 3 - DMRLoads2007: Development and Category Rankings • Some facilities in PCS and ICIS-NPDES do not provide information on applicable SIC codes. Additionally, facilities in PCS do not provide information on applicable North American Industrial Classification System (NAICS) codes, while only a few facilities in ICIS-NPDES provide information on applicable NAICS codes. • Although facilities may provide more than one SIC code to describe their operations, EPA uses only the primary SIC code to classify facilities. • PCS and ICIS-NPDES were designed as permit compliance tracking systems and do not contain production information. • DMR data may be entered into the PCS or ICIS-NPDES database manually, which leads to data-entry errors. Despite the limitations and constraints of the PCS and ICIS-NPDES databases, EPA has determined that they are appropriate for an initial screening-level review and prioritization of the pollutant loads discharged by industrial categories. EPA will further evaluate the prioritized categories in a second level of review, which may include additional data collection and additional verification of data reported in PCS and ICIS-NPDES. 3.1.3 PCS and ICIS-NPDES Data Structure The PCS and ICIS-NPDES databases contain more than 5 million records organized by individual permit files. Each permit file contains information about the following elements: • The permit and the permitted facility, including permit number, dates of issue and expiration, facility name, location, and type of facility; • Permit events, including date application was received, scheduled, and achieved dates for completion of compliance schedule; • Identity of outfalls within the facility and a description of the associated monitoring requirements; • Parameters to be measured at each outfall and the corresponding limitations; and • Inspections performed at the facility, such as type of inspection, inspector identity, and inspector comments. To develop DMRLoads2007, EPA developed two pollutant loading tools: the database PCSLoadCalculator2007 and the ICIS-NPDES Pollutant Loading Tool. These loading tools start with DMR data stored in their respective databases and use similar methodologies to calculate annual mass discharges from DMR data. PCS and ICIS-NPDES store data in a series of tables. Table 3-2 lists the PCS and ICIS-NPDES data types that EPA used to create DMRLoads2007. EPA uses data in the Permit Facility, Pipe Schedule, Measurement Violation, and Permit Event data types to develop PCSLoadCalculator2007'. In ICIS-NPDES, these types of data are stored in several relational database tables, also shown in Table 3-2. In addition to the four data types used to develop PCSLoadCalculator2007, EPA also used data in the Parameter Limits data type to develop the ICIS-NPDES Pollutant Loading Tool. EPA did not use the following nine data types in PCS and ICIS-NPDES in developing its load calculators: • Compliance schedule data; • Enforcement action data; • Evidentiary hearing data; ------- Section 3 - DMRLoads2007: Development and Category Rankings Table 3-2. Data Types in PCS and ICIS-NPDES Used for DMRLoads2007 Development PCS Data Type Permit-Facility Data Pipe-Schedule Data Parameter-Limits Data Measurement- Violation Data Permit Events Data Description General descriptive information on each permitted facility (such as its name, address, classification and design flow rate). It contains the basic information regarding a permit, permit-facility data is the one data type that belongs to all of the families of logically related data types. Detailed information describing each outfall within a permitted facility and the discharge monitoring requirements associated with each (such as effluent waste types, treatment types, and limit start and end dates-initial, interim, or final). Detailed information specifying the monitoring requirements associated with each outfall within a permitted facility (such as monitoring location, the parameter to be monitored, the required frequency of analysis, the units in which the measurements are expressed, and the quantity and concentration limits for each parameter). Detailed information on the reported measurement values for effluent parameters including those that are in violation of established limits for the permit, the type of violation, the reported number of excursions, the actual measurement values, and the percentage by which a measurement exceeds quantity and/or concentration limits. Information tracking the events relating to the issuance of a permit, from initial receipt of the application for a permit through actual permit issuance. Included in PCSLoadCalculator200 7 Yes Yes No Yes Yes Relational Tables used in ICIS NPDES Pollutant Loading Tool Facility interest Permit Facility interest SIC Facility interest NAICs Permitted Feature Permitted Feature Coordinates Limit Limit Set Limit Set Schedule Limit Value DMR DMR Event DMR Form DMR Form Parameter DMR Form Value DMR Parameter DMR Value REF_Parameter Permit Source: Permit Compliance System Generalized Retrieval Training Manual, Table 1-1, pg 1-4 (U.S. EPA, 2001); Results of ICIS-NPDES Pollutant Loading Tool Convert Module Development and DMR Data Review - Update 1 (Camp, 2009). 3-6 ------- Section 3 - DMRLoads2007: Development and Category Rankings • Grant data; • Inspection data; • Inspection scheduling data; • Pretreatment audit/PCI data; • Pretreatment performance data; and • Schedule violation data. 3.2 DMRLoads2007: Database Development and Methodology To develop DMRLoads2007, EPA developed two pollutant loading tools: the database PCSLoadCalculator2007 and the ICIS-NPDES Pollutant Loading Tool. These loading tools start with DMR data stored in their respective databases and use similar methodologies to calculate annual mass discharges from DMR data. Due to differences in the PCS and ICIS-NPDES data structures, EPA's analysis required two separate loading tools - one for PCS and one for ICIS- NPDES8. EPA created DMRLoadsAnalysis2007 to merge data from the two systems, evaluate the impacts of calculation assumptions, and track database corrections. EPA also created DMRNutrients2007 to evaluate point source discharges of nitrogen and phosphorus. The remainder of this subsection describes the methodology and assumptions used in creating the DMRLoads2007 database to generate point source category rankings: • Section 3.2.1 describes the data sources used to create DMRLoads2007'; • Section 3.2.2 describes PCSLoadCalculator2007; • Section 3.2.3 describes the ICIS-NPDES Pollutant Loading Tool; • Section 3.2.4 describes DMRLoadsAnalysis2007; • Section 3.2.5 describes DMRNutrients2007; and • Section 3.2.6 describes DMRLoads2007'. 3.2.1 Data Sources used in the Development of DMRLoads200 7 Figure 3-1 shows the relationship between PCS, ICIS-NPDES, and DMRLoadsAnlaysis2007, DMRNutrients2007, andDMRLoads2007. In future years of DMR data analysis, the ICIS NPDES Pollutant Loading Tool functions will include PCS load calculations, and only one loading tool will be required. 3-7 ------- Section 3 - DMRLoads2007: Development and Category Rankings ICIS-NPDES ICIS-NPDES Convert Module PCS Facility Data NPDES Permit ID (NPID) Standard Industrial Classification (SIC) NAME, CITY, STATE Major Facility Indicator (MADI) Facility Latitude/Longitude Data Permit Issuance Date (PERD) Create Tables Facility Permit Feature Convert DMR PRAM/CAS Crosswalk ICIS-NPDES Load Calculator Module Imported Tables Convert DMR Create Tables ICIS-NPDES Annual Loads PCS CNVRT Module PCSLoadCalculator2007 Imported Tables Manual Database Corrections Create Tables PCS Annual Loads DMRLoadsAnalysis2007 Imported Tables Annual Loads (ICIS & PCS) PCSFAC (PCS) ICISFAC (ICIS) Parameter Groupings (ICIS) Create Tables DMR2007 DMRFAC2007 Sensitivity Analysis (PCS only) •DL (PCS & ICIS) •EST (PCS) •NOMAX (PCS) •NOCNC (PCS) Other Linked Tables SIC/PSC Crosswalk SUPERCAS TWFs DMRLoads2007 Linked Tables DMR2007 DMRFAC2007 Create Tables Category Rankings - Toxic Weight Other Linked Tables SIC/Point Source Category Crosswalk SIC Codes Point Source Category Codes DMRNutrients2007 Linked Tables DMR2007 DMRFAC2007 Create Tables Nutrient Loads by Discharge Category Rankings - Total Nitrogen Category Rankings- Total Phosphorus Figure 3-1. Relationship Between Data Sources and Database Development Tools for the Development of DMRLoads2007 EPA used the following data sources and database development tools to create DMRLoads2007: • PCS: This mainframe database is the source of the pollutant discharge data and facility information used in the development ofPCSLoadCalculator2007. EPA used year 2007 data from PCS to develop PCSLoadCalculator2007'. • ICIS-NPDES: This web-based Oracle™ database is the source of the pollutant discharge data and facility information used in the development of the ICIS- NPDES Pollutant Loading Tool. EPA used year 2007 data from ICIS-NPDES to develop the ICIS-NPDES Pollutant Loading Tool. EPA obtained ICIS-NPDES data directly from OECA, as it is not yet available through a public download system. • EPA's Convert Programs: Two EPA programs convert pollutant concentrations and loads in DMR data into standard units and match them with flows and permit limits. The PCS convert program (CNVRT) is a mainframe computer program developed and maintained by OECA. The ICIS-NPDES Convert Module is a ColdFusion™ based program developed by EPA for the 2009 and future annual reviews that extracts DMR data from ICIS-NPDES and stores the converted data to an Oracle™ database. 3-8 ------- Section 3 - DMRLoads2007: Development and Category Rankings • EPA's Load Calculator Routines: EPA developed its PCS and ICIS-NPDES Load Calculator Routines based on OECA's Effluent Data Statistics (EDS) System for PCS: — The PCS Load Calculator Routine uses a series of Microsoft Access™ database queries in PCSLoadCalculator2007 to compute annual pollutant loads. In addition, PCSLoadCalculator2007 tracks database corrections for monthly flow, quantity, concentration, reporting frequency, and internal monitoring locations. — The ICIS-NPDES Load Calculator Routine uses the ICIS-NPDES Load Calculator Module, a ColdFusion™-based program that extracts converted DMR data from the ICIS-NPDES Pollutant Loading Tool Oracle™ database and calculates annual pollutant loads. The ICIS-NPDES Load Calculator Module differs from PCSLoadCalculator2007 in that it is part of a dynamic web application that allows users to selectively query loads and specify calculation assumptions. Both load calculator routines produce five alternative loads by applying variations in calculations assumptions (see Sections 3.2.2.2 and 3.2.3.2). • DMRLoadsAnalysis2007: This PC-based Microsoft Access™ database standardizes and then combines the annual loads data from PCSLoadCalculator2007 and the ICIS-NPDES Loading Tool. The database also examines the impact of the alternative load calculations (see Section 3.2.4.2 for additional details). The database uses the calculation assumptions that EPA selected based on the results of the data sensitivity analyses conducted for the 2007 annual review, and creates the DMR2007 table, which provides one annual load per pollutant discharge. Additionally, this database calculates the toxic- weighted pound equivalent (TWPE) for each pollutant discharge. This database applies several database corrections, based on findings during previous annual reviews and the 2009 annual review quality review (see Section 3.4), to correct errors related to facility categorization, pollutant discharge categorization, parameter groupings, intermittent discharges, and internal monitoring locations. See Section 3.2.4 for additional details onDMRLoadsAnalysis2007. • DMRNutrients2007: This PC-based Microsoft Access™ database uses the annual loads for nitrogen and phosphorus compounds from the DMR2007 table to calculate aggregate "nitrogen as N" and "phosphorus as P" loads for each facility outfall. The database sums the aggregate nitrogen and phosphorus loads by facility and by point source category and exports the aggregated loads to DMRLoads2007. 3.2.2 PCSLoadCalculator2007 EPA developed PCSLoadCalculator2007 to process CNVRT output into a structure usable to calculate annual loads. PCSLoadCalculator2007 is a Microsoft Access™ database that implements EPA's PCS Load Calculator routine. As depicted in Figure 3-1, PCSLoadCalculator2007 uses CNVRT output and calculates annual loads for each pollutant and discharge point using PCS Load Calculator. The output from PCSLoadCalculator2007 is the ------- Section 3 - DMRLoads2007: Development and Category Rankings "PCS Annual Loads" table, which is exported to DMRLoadsAnalysis2007 for combination with ICIS-NPDES Pollutant Loading Tool annual loads for further calculations and analyses. The PCS Load Calculator routine is based on OECA's mainframe computer program, called the EDS system. This system establishes how to calculate annual loads from the CNVRT output and was used in the 2003 and 2005 annual reviews (U.S. EPA, 1997). In 2005, EPA developed the PCS Load Calculator to duplicate the EDS system and to address data processing difficulties when using EDS (U.S. EPA, 2005). EPA continues to use the PCS Load Calculator routine instead of the EDS system because it allows EPA flexibility and control over the annual load calculations and provides transparent documentation of the calculations. 3.2.2.1 CNVRT Module Input for PCSLoadCalculator2007 EPA used CNVRT module output to create PCSLoadCalculator2007 (see Figure 3-1). From the PCS mainframe, the CNVRT module performs units conversions, matches flow rates with pollutant measurements, assigns a statistical basis of measurement, and performs formatting changes to convert the PCS data into a format that is usable for annual load calculations. Table 3-3 presents the CNVRT module output that EPA used as a starting point for its annual load calculations for PCSLoadCalculator2007'. Table 3-3. PCS CNVRT Module Output PCS Field NPID SIC2 DSCH DRID NRPU PRAM MLOC SEAN MODN LIPQ STAT MVDT MVIO NODI LMQAV LMQMX LMCMN LMCAV LMCMX MQAV MQMX MCMN MCAV Description NPDES Number Standard Industrial Classification Code Discharge Pipe Report Designator Number of Units in Reporting Period Parameter Code Monitoring Location Season Number Modification Number Limit Pipe Set Qualifier Statistical Base Code Measurement/Violation Monitoring Period End Date Measurement/Violation Code No Data Indicator Measurement/Violation Quantity Average BDL Indicator Measurement/Violation Quantity Maximum BDL Indicator Measurement/Violation Concentration Minimum BDL Indicator Measurement/Violation Concentration Average BDL Indicator Measurement/Violation Concentration Maximum BDL Indicator Measurement/Violation Quantity Average Measurement/Violation Quantity Maximum Measurement/Violation Concentration Minimum Measurement/Violation Concentration Average 3-10 ------- Section 3 - DMRLoads2007: Development and Category Rankings Table 3-3. PCS CNVRT Module Output PCS Field MCMX FMQAV FMQMX FMCMN FMCAV FMCMX Description Measurement/Violation Concentration Maximum Measurement/Violation Quantity Average Flow Measurement/Violation Quantity Maximum Flow Measurement/Violation Concentration Minimum Flow Measurement/Violation Concentration Average Flow Measurement/Violation Concentration Maximum Flow The following describes the functions of the CNVRT module: • Unit Conversions: The CNVRT module converts the PCS measurement data into standard units of kg/day for mass quantities, mg/L for concentrations, and million gallons per day (MOD) for flow rates. • Matching Flows with Pollutant Discharges: Quantities and concentrations are reported to PCS using five pollutant parameter measurement fields (MQAV, MQMX, MCMN, MCAV, MCMX (see Table 3-3 for measurement field descriptions). Wastewater flow rates are reported to PCS as a pollutant parameter using the same five measurement value fields. CNVRT matches wastewater flow rates with pollutant measurements using identifying fields in PCS, such as monitoring period end date, monitoring location, discharge pipe number, report designator, and season number. CNVRT creates five new columns for each pollutant discharge record and stores the matching flow information in these fields (FMQAV, FMQMX, FMCMN, FMCAV, and FMCMX). • Assigning Statistical Basis: The statistical basis of measurements in PCS is identified by the statistical base code. CNVRT categorizes the 150 statistical base codes in PCS as representing average, maximum, minimum, or total measured values. CNVRT then simplifies the statistical base code by assigning a number from 0 to 4 to each measurement value field. The assigned numbers are as follows: — 0 - No Value Reported, — 1 - Average, — 2 - Total Monitoring Period Value, — 3 - Maximum, and — 4 - Minimum. CNVRT combines the numbers assigned to each of the five measurement values into one five-digit code called STATS. Each of the five digits in STATS corresponds to one of the five measurement fields for pollutant loads or concentrations. Figure 3-2 shows an example of a possible STATS code. In this figure, the measurements reported for MQAV, and MCAV are based on average values, MQMX and MCMX are based on maximum values, and MCMN is based on the minimum value. 3-11 ------- Section 3 - DMRLoads2007: Development and Category Rankings Formatting Changes: For pollutants measured at concentrations below their detection limit (BDL), facilities report the detection limit concentration to PCS and indicate that the measurement is BDL using a less-than sign (<). CNVRT pulls the less-than signs from the measurement value fields and places them in a separate field. MQAV MQMX MCMN MCAV MCMX (Average quantity (Maximum quantity (Minimum concentration (Average concentration (Maximum concentration measurement) measurement) measurement) measurement) measurement) Figure 3-2. Example PCS STATS Code 3.2.2.2 PCS Annual Load Calculation Routine This section describes the calculations used to produce annual loads from CNVRT output files in PCSLoadCalculator2007'. Figure 3-3 presents a flow diagram for the PCSLoadCalculator2007 routine. Files obtained from the CNVRT module are the starting point for the PCS Load Calculator routine. PCS Data Selection Some monitoring data in the CNVRT output are not relevant to calculating effluent loads, and the PCS Load Calculator selects relevant CNVRT output. Irrelevant information includes pollutant discharges for internal monitoring locations, pollutant discharges reported for certain measurement fields, and flows reported for certain measurement fields. For example, for a certain monitoring location, pollutant discharges may be reported as both a mass quantity and a concentration. However, EPA does not use concentration data if the quantity is also reported. The PCS Load Calculator routine selects relevant PCS data for the following parameters: 1) monitoring location, 2) measurement value, and 3) flow value, as described below. Monitoring Location Selection. Permits often require a facility to monitor at multiple locations. The monitoring location is indicated in PCS in the MLOC field. Two of the many PCS MLOC codes designate effluent discharges: • MLOC 1 - Effluent gross discharge; and • MLOC 2 - Effluent net discharge. For its screening level review, EPA estimates annual loads that represent effluent discharges. Therefore, the PCS Load Calculator searches the monitoring field location (MLOC) in PCS to find effluent data only (MLOC 1 or MLOC 2). When both types of effluent data are present for an outfall, MLOC 2 is used in preference to MLOC 1. 3-12 ------- Section 3 - DMRLoads2007: Development and Category Rankings Includes only Effluent Monitoring Locations (MLOC=1 or MLOC=2) NOMAX: Sets all maxima to zero (MQMX and MCMX=01 Calculate Monitoring Period Load: Quantity * 30 days/month * NRPU; or Concentration * Flow * 30 days/month * NRPU Calculate Annual Load using EST=YES: Sum(Monitoring Period Load) * 12/(SUM(NRPU)) NOMAX KGY01, JNOMAX KGYE1 Apply Hybrid Method T NOMAX_KGYH1 For use in NOMAX Analysis CNVRT Output Monitoring Location Selection • Includes Internal and Effluent Monitoring Locations FMQAV FMQMX >• Flow Values: FMCMN FMCAV FMCMX Flow Value Selection Flow (MGD) Measurement Values: MQAV MQMX MCMN MCAV MCMX Measurement Value Selection NOCONC: Sets all concentrations to zero (MCMN, MCAV, and MCMX=0) Normal Hierarchy Calculate Monitoring Period Load: Quantity * 30 days/month * NRPU; or Concentration * Flow * 30 days/month * NRPU Calculate Monitoring Period Load: Quantity * 30 days/month * NRPU Apply DL Options Apply DL Options BDL = Calculate Annual Load using EST=YES: Sum(Monitoring Period Load)* 12/(SUM(NRPU)) [ KGY01 KGYH01 For use in DL Analysis I Calculate Annual Load using EST=YES: Sum(Monitoring Period Load) * 12/(SUM(NRPU)) NOCONC KGY01 Calculate Annual Load using EST=NO: Sum(Monitoring Period Loads) KGYE-I KGYOO Apply Hybrid Method Apply Hybrid Method KGYEO Apply Hybrid Method NOCONC_KGYH1 For use in NOCONC Analysis KGYH1 For generating Category Rankings KGYHO For use in EST Analysis Figure 3-3. Flow Diagram for PCS Load Calculator Routine 3-13 ------- Section 3 - DMRLoads2007: Development and Category Rankings Measurement Value Selection. PCS contains five measurement value fields for measured pollutant data (MQAV, MQMX, MCMN, MCAV, and MCMX). The PCS Load Calculator uses a two-step process to select which of these measurement values to use to calculate the annual loads. In the first step, the PCS Load Calculator attempts to identify an average value using STATS and a measurement field hierarchy. (See Section 3.2.2.1 for how CNVRT develops STATS number using statistical base codes in PCS data.) The PCS Load Calculator first searches each STAT digit corresponding to the PCS measurement fields in the following sequence, or hierarchy: • Average Load (MQAV); • Maximum Load (MQMX); • Average Concentration (MCAV); • Maximum Concentration (MCMX); or • Minimum Concentration (MCMN). A measurement must meet two criteria to be selected for loads calculation: 1) the mass quantity or concentration must be nonzero, and 2) the corresponding STAT digit for the measurement value field must equal 1. If the PCS Load Calculator cannot identify a measurement that meets these two criteria, then the PCS Load Calculator selects measurement values based on which field they populate without considering the STATS digit. In this step, the following hierarchy is used: • The average load (MQAV) field is used if it contains a non-zero value; • If MQAV cannot be used, and a flow rate is reported, the concentration fields are searched in the following order and the first nonzero concentration is multiplied by the flow to calculate the load: — Average Concentration (MCAV); — Maximum Concentration (MCMX); — Minimum Concentration (MCMN); and • If flow and concentration cannot be used to calculate the load, the maximum load (MQMX) is used. For sensitivity analyses, EPA calculated two sets of alternative loads ("NOMAX" and "NOCONC") using variations on the measurement value selection hierarchy. Figure 3-3 shows how these alternative loads relate to the loads calculated using the normal hierarchy. Section 3.2.4.2 describes the purpose of the alternative calculations and EPA's analysis of NOMAX and NOCONC annual loads. Flow Value Selection. To select the appropriate flow data to use to calculate annual loads, the PCS Load Calculator uses a hierarchy that is similar to the measurement value selection hierarchy. The PCS Load Calculator searches the flow measurement fields in the following sequence and selects the first non-zero value it finds: 3-14 ------- Section 3 - DMRLoads2007: Development and Category Rankings • Average Quantity Flow (FMQAV); • Average Concentration Flow (FMCAV)9; • Maximum Concentration Flow (FMCMX); • Minimum Concentration Flow (FMCMN); and • Maximum Quantity Flow (FMQMX). While conducting the flow selection process, the PCS Load Calculator attempts to identify and correct flows that have misreported units, which is a common problem for flows in PCS. The PCS Load Calculator attempts to correct this problem by assuming that any reported flow rate greater than 5,000 MGD is actually gallons per day (GPD), and divides the reported flow by one million. For flows ranging from 1,300 to 5,000 MGD, EPA compares units for flow permit limits to verify the units reported in PCS and makes corrections on a case-by-case basis10. This is a change from the EDS methodology, which divides all flows that are greater than 1,300 MGD by one million. Section 3.2.2.3 discusses EPA's basis for this change in methodology. Calculate Monitoring Period Load After completing the monitoring location, measurement value, and flow selection hierarchies, the PCS Load Calculator has identified one mass quantity or one concentration and flow to calculate a load for each pollutant discharge for each monitoring period. The duration of discharge that each monitoring period represents depends on the reporting frequency required by a facility's NPDES permit. For example, if a facility is required to report monthly, then the reported discharge for the monitoring period will represent one month of discharges (assuming continuous discharges). If a facility is required to report quarterly, then the reported discharge for the monitoring period will represent three months of discharges. EPA assumes that an outfall discharges continuously for 30 days per month, and the PCS Load Calculator calculates the monthly load using one of the following equations: • Calculation of monthly load from daily load (MQAV or MQMX): Monthly Load (kg/mo) = Daily Load (kg/day) x 30 (days/mo) • Calculation of monthly load from concentration and flow (MCAV, MCMX, or MCMN): Monthly Load (kg/mo) = Cone. (mg/L) x Flow (MGD) x 3.785 (L/gal) x 30 (days/mo) As Figure 3-3 shows, the PCS Load Calculator then adjusts the monthly load to represent quarterly, semiannual, or annual loads where appropriate by multiplying each monthly load by the number of reporting units (NRPU). The NRPU data element is a numeric code that indicates whether a pollutant is monitored monthly (NRPU = 1), quarterly (NRPU = 3), semiannually (NRPU = 6), or annually (NRPU = 12). For example, if a facility reported a 30-day average load 9 A "concentration flow" is a flow measurement that was reported to a concentration measurement field. Facilities may report flows in any of the five measurement value fields. However, all flows are reported in units of MGD whether they are reported in a quantity field or a concentration field. 10 EPA determined that all flows between 1,300 and 5,000 MGD reported by facilities in Ohio were flows in GPD. EPA automatically divided these flows by 1,000,000. However, because power plants are known to have high flows, EPA made flow corrections to Ohio facilities reporting SIC code 4911 (Electrical Services) on a case-by-case basis. 3-15 ------- Section 3 - DMRLoads2007: Development and Category Rankings of 25 kg/day for its required quarterly report (NRPU=3), the PCS Load Calculator calculates the load for the quarter as 25 kg/day x 30 days/month x 3 month/quarter = 2,250 kg/quarter. Apply DL Options As shown in Figure 3-3, the PCS Load Calculator produces two monitoring period loads by using different calculation assumptions for pollutants that were measured BDL. Using the BDL indicator field from the CNVRT output, the PCS Load Calculator identifies pollutants that were measured BDL. If the BDL indicator field contains a less-than sign (<), the PCS Load Calculator calculates two period loads: one by setting the monitoring period load to zero (BDL = 0) and a second by dividing the monitoring period load in half (BDL = 1A DL). If the BDL indicator field is blank, then the PCS Load Calculator uses the calculated period load for both options. Table 3-4 shows an example calculation of loads for the two DL options. Table 3-4. Example Calculation for DL Option Loads Calculated Monitoring Period Load (kg/period) 100 100 BDL Indicator Field Blank < Option BDL = 0 Load (kg/period) 100 0 Option BDL = Vz DL Load (kg/period) 100 50 Calculate Annual Load Scenarios The output from the monitoring period load calculation step should include the following data for each pollutant discharge: • Twelve loads for monthly reports; • Four loads for quarterly reports; • Two loads for semiannual reports; and • One load for annual reports. However, in some cases, PCS does not contain a complete set of discharges for the year. If a facility does not report a pollutant concentration or mass quantity on its DMR, then the facility uses the no data indicator (NODI) field to explain why no discharge is reported. NODI is a single character code in PCS, which corresponds to a no data indicator description. Table 3-5 presents descriptions of all the NODI codes. Table 3-5. NODI Code Descriptions NODI Code 1 2 4 5 7 8 9 NODI Description Wrong flow Operations shutdown Discharge to lagoon/groundwater Frozen conditions No influent Other Conditional monitoring 3-16 ------- Section 3 - DMRLoads2007: Development and Category Rankings Table 3-5. NODI Code Descriptions NODI Code A B C D E F G H I J K L Q R S V w X NODI Description General permit exemption Below detection limit/no detection No discharge occurred for the monitoring period Lost sample Analysis not conducted Insufficient flow for sampling Sampling equipment failure Invalid test Land applied Recycled - water-closed system Flood disaster DMR received but not entered Not quantifiable Administratively resolved Fire conditions Weather related Dry lysimeter/well Parameter/value not reported The PCS Load Calculator includes two options for calculating the annual load when PCS does not contain a complete set of monitoring period loads for the year: 1) sum the existing monitoring period loads to calculate the annual load (EST=NO); or 2) estimate loads for the missing monitoring periods (EST=YES). The following sections describes the calculation of EST=YES and EST=NO loads. For the 2009 annual review, EPA used only the EST=YES loads. Calculate EST=YES Annual Loads. The PCS Load Calculator uses the sum of NRPU values to identify annual loads that do not include a complete set of monitoring period loads. First, the PCS Load Calculator sums the NRPU values for the monitoring periods that have calculated pollutant loads. In addition, the PCS Load Calculator sums the NRPU values for blank records with NODI codes that indicate no discharge occurred for the monitoring period. As part of the 2009 annual review, EPA reviewed all NODI codes and determined that the following NODI codes represent "no discharge" events: 2: Operations Shutdown; 4: Discharge to Lagoon/Groundwater; 7: No Influent; 9: Conditional Monitoring; C: No Discharge; I: Land Applied; J: Recycled - Water-Closed System; and W: Dry Lysimeter/Well. 3-17 ------- Section 3 - DMRLoads2007: Development and Category Rankings Note that EPA updated the list of NODI codes for the 2009 annual review. For previous annual reviews, EPA used a different list of "no discharge" NODI codes. See Section 3.2.2.3 for additional information. The PCS Load Calculator then combines the sum of NRPU values for monitoring period loads and monitoring periods with no discharge. If all monitoring periods for the annual data set either have discharge data or indicate no discharge, then the sum of NRPU will equal 12. For example, if a facility is required to monitor quarterly, the NRPU assigned to each quarterly report is 3. If four quarterly reports are present, the total NRPU is 12 (3+3+3+3), indicating all required reports are present. However, if the annual data set includes blanks for any of the monitoring periods and does not indicate that no discharge occurred for the monitoring period, then the sum of NRPU will be less than 12. As shown in Figure 3-3, the input to the Calculate EST=YES Annual Loads step includes two sets of monitoring period loads from the Calculate DL Options step: BDL = 0 and BDL = 1A DL. To calculate the EST=YES load, the PCS Load Calculator sums monitoring period loads for the DL = 0 option and separately sums the monitoring period loads for the DL = 1A DL option. For each sum, the PCS Load Calculator then extrapolates the calculated annual load to account for blank records using the following equation: (EST=YES) Annual Load (kg/yr) = Sum(Monitoring Period Load x NRPU) x (12/Sum(NRPU)) Calculate EST=NO annual loads. During the EST=YES calculation step, the PCS Load Calculator also calculates an alternative annual load using the EST=NO option. The calculation for EST=NO is the same as the EST=YES calculation except EST=NO does not multiply the sum of the period loads by the ratio of 12 and the sum of NRPU values. The EST=NO annual load is shown in the following equation: (EST = NO) Annual Load (kg/yr) = ^(MonitoringPeriodLoadxNRPU) Apply Hybrid Method As shown in Figure 3-3, the output from the Calculate EST=YES Annual Loads step includes two annual loads for the DL options: BDL = 0 and BDL = 1A DL. During this calculation step, the PCS Load Calculator applies the following logic to select which calculated load to use to represent the final annual load: • If the BDL = 0 load equals zero, use the BDL = 0 load (all monitoring period loads for 2007 are zero); and • If the BDL = 0 load is greater than zero, use the BDL = !/2 DL load (at least one monitoring period was not zero, i.e., the pollutant was detected at least once during 2007). As shown in Figure 3-3, the PCS Load Calculator calculates alternative annual loads starting at the Measurement Value Selection step. During this step, the PCS Load Calculator calculated two sets of alternative monitoring period loads using variations on the measurement value selection hierarchy: 1) set all maximum concentrations and loads to zero (NOMAX); and 2) set all average, maximum, and minimum concentrations to zero (NOCONC). The PCS Load Calculator ------- Section 3 - DMRLoads2007: Development and Category Rankings then applied the DL options to these alternative loads and calculated EST=YES and EST=NO annual loads for the NOMAX and NOCONC alternatives. As a final step the PCS Load Calculator applies the Hybrid Method to the calculated alternative loads. See Section 3.2.4.2 for discussion of the alternative annual loads. PCSLoadCalculator2007 Output The PCS Load Calculator produces 12 calculated annual loads for each pollutant discharge. Table 3-6 lists the 12 calculated annual loads and describes the purpose of each load. Seven of the loads use various assumptions for pollutant measurements reported as BDL, which are used to calculate final loads using the Hybrid Method. Five of the loads are final loads, which are used for category rankings and sensitivity analyses (see Section 3.2.4.2). The five final annual loads are included in the PCSLoadCalculator2007 output to DMRLoadsAnalysis2007'. Table 3-6. PCSLoadCalculator2007 Output Annual Load EST Option DL Option Measurement Selection Hierarchy Purpose Interim Loads KGYE1 KGYOO KGYEO NOMAX JCGY01 NOMAX JCGYE1 NOCONC JCGY01 NOCONC JCGYEl Yes No No Yes Yes Yes Yes BDL = !/2DL BDL = 0 BDL=!/2DL BDL = 0 BDL = !/2DL BDL = 0 BDL = !/2 DL Normal Normal Normal All maxima set to zero All maxima set to zero All concentrations set to zero All concentrations set to zero Used with KGY01 to calculate Hybrid (KGYH1) Used with KGYEO to calculate Hybrid (KGYHO) Used with KGYOO to calculate Hybrid (KGYHO) Used with NOMAX KGYE1 to calculate Hybrid (NOMAXJCGYHl) Used with NOMAX KGY01 to calculate Hybrid (NOMAXJCGYHl) Used with NOCONC KGYE1 to calculate Hybrid (NOCONCJCGYHl) Used with NOCONC KGY01 to calculate Hybrid (NOCONCJCGYHl) Final Loads KGYH1 KGYHO KGY01 NOMAXJCGYHl NOCONCJCGYHl Yes No Yes Yes Yes Hybrid Hybrid BDL = 0 Hybrid Hybrid Normal Normal Normal All maxima set to zero All concentrations set to zero Category Rankings EST Analysis DL Analysis No Max Analysis No Cone Analysis 3.2.2.3 Changes to EDS Methodology As stated previously, EPA followed the EDS methodology to develop the annual load calculation methodology for PCSLoadCalculator2007. This section discusses changes that EPA made to the methodology including the reason for the change. 3-19 ------- Section 3 - DMRLoads2007: Development and Category Rankings NRPU Correction. Monitoring frequencies may vary for certain pollutants or outfalls depending on a facility's permit requirements. Discharges may be reported monthly, quarterly, semiannually, or annually. As discussed previously, the NRPU data element is a numeric code that indicates whether a pollutant is monitored monthly (NRPU = 1), quarterly (NRPU = 3), semiannually (NRPU = 6), or annually (NRPU = 12). As described in Section 3.2.2.2, the PCS Load Calculator uses the NRPU value for two steps in the annual load calculation. • The first step that uses the NRPU value is the monitoring period load calculation. During this step, the PCS Load Calculator calculates a monthly load by multiplying a mass quantity by 30 days per month, and then multiplies the monthly load by the NRPU value to calculate a quarterly, semi-annual, or annual load. • The second step that uses the NRPU value is the calculation of annual loads using the EST=YES option. During this step, the PCS Load Calculator uses the sum of the NRPU values associated with the reported discharges to determine if all DMR data for the pollutant are present in PCS. If the sum of the NRPU values equals 12, then all required discharge data are present for that reporting year. During the development ofPCSLoadCalculator2007, EPA observed that the sum of NRPU values for several annual loads was greater than 12, indicating that discharge data for more than the required number of DMRs were present in PCS. Following are two scenarios that resulted in the sum of NRPU exceeding 12. • Scenario 1: Incorrect NRPU reported. The first scenario is a data-entry error where the NRPU in PCS was incorrect for the frequency of the reported discharges. For example, a quarterly discharge report should have an NRPU value of 3, but the NRPU value in PCS was 6. As a result, the monthly load for each quarter was multiplied by 6 instead of 3 during the quarterly load calculation, which double-counted the quarterly loads. The EST=YES calculation automatically corrects this error by multiplying the annual load by the ratio of 12 to the sum of the NRPU values. For this example, the sum of NRPU values for the four quarterly reports would be 24 instead of 12. Therefore, using EST=YES, the annual load would be multiplied by 12/24 (0.5), which eliminates the double- counting. For EST=NO, however, this error results in double-counting the annual load since the EST=NO calculation does not multiply the annual load by the ratio of 12 to the sum of NRPU values. EPA corrected the NRPU values for the Scenario 1 cases by changing the NRPU values in the monthly data to correctly reflect the monitoring frequency. • Scenario 2: Multiple monthly measurements. The second scenario occurred if a facility reported discharges twice in one month. For example, a facility reports a discharge monthly to PCS (NRPU = 1), but reported two discharges for September (one on September 15 and one on September 30). The NRPU values for both September reports are 1. Similar to Scenario 1, the double-counting that results from this error is corrected during the EST=YES calculation but not during the EST=NO calculation. In addition to double-counting, this error also causes the discharges reported for September to account for a disproportionate amount of the annual load. For example, the monthly load calculation multiplies both the 3^20 ------- Section 3 - DMRLoads2007: Development and Category Rankings September 15th and September 30th loads by 30. As a result, September discharges account for 2 out of 13 months instead of 1 out of 12 months. EPA corrected the NRPU values for the Scenario 2 cases by dividing the NRPU values for months with multiple discharges by the number of discharges reported for the month. For this example, the September NRPU value of 1 was divided by 2 because there were two discharge reports for September (corrected NRPU = 0.5). As a result, the monthly load calculation multiplies each September discharge by 30 days per month and 0.5, making each discharge account for one half of a month (15 days). NODI Codes Excluded from EST=YES Assumption. As stated in Section 3222, EPA updated the list of NODI codes that indicate that no discharge occurred. Prior to the 2009 Annual Review, EPA used the NODI codes shown below: • C: No discharge; • D: Lost sample; • E: Analysis not conducted; • F: Insufficient flow for sampling; • G: Sampling equipment failure; • H: Invalid test; • K: Flood disaster; • 5: Frozen conditions; and • 8: Other. EPA revised this list to the following using information obtained from OECA during the 2009 annual review: • 2: Operations shutdown; • 4: Discharge to Lagoon/Groundwater; • 7: No Influent; • 9: Conditional Monitoring; • C: No discharge; • I: Land Applied; • J: Recycled - Water-Closed System; and • W: Dry Lysimeter/Well. EPA assumed that the above NODI codes represent "no discharge" for the 2009 annual review and will continue to use them for subsequent reviews. EPA evaluated the effect of revising the NODI codes used in the EST=YES calculation on the category rankings by comparing two sets of annual loads calculated using PCSLoadCalculator200 7: 1. Annual loads calculated using the NODI codes for previous annual reviews (C, D, E, F, G, H, K, 5, and 8); and 2. Annual loads calculated using the revised NODI codes for the 2009 annual review (2, 4, 7, 9, C, I, J, and W). 3-21 ------- Section 3 - DMRLoads2007: Development and Category Rankings Table 3-7 presents a summary of the results of the NODI analysis for the 10 point source categories showing the highest absolute increase in TWPE from the NODI changes and the total for PCS-portion ofDMRLoads2007. As shown in Table 3-7, revising the NODI codes changed the total TWPE for the PCS-portion of DMRLoads2007 by only 0.086 percent (794,000 Ib-eq). The categories showing the greatest sensitivity to the NODI revisions include Copper Forming (40 CFR Part 468); Justice, Public Order, and Safety (SIC Group 92), and Grain Mills (40 CFR Part 406) Categories. Table B-3 in Appendix B presents the pounds and TWPE using the two NODI code options by pollutant parameters. Pollutant parameters showing the highest sensitivity to the NODI revisions include mercury, 2,3,7,8-Tetrachlorodibenzo-p-dioxin, and polychlorinated biphenyls. Flow Correction. As described in Section 3.2.2.2, the PCS Load Calculator attempts to identify and correct flows that have misreported units using a two-step process. First, the PCS Load Calculator assumes that any flow rate that is greater than 5,000 MGD should actually be reported as GPD, and divides the flow by one million. EPA also reviews reported flows ranging from 1,300 to 5,000 MGD by comparing reporting units to permit limits to verify the reporting units and makes corrections on a case-by-case basis. This is a change from the EDS methodology, which divides all flows that are greater than 1,300 MGD by one million. The 1,300 MGD cutoff was based on the maximum flow rate identified at the time that EDS was developed. EPA has identified several facilities that currently discharge wastewater at flows exceeding 1,300 MGD. The 1,300 MGD cutoff used by EDS would underestimate loads for these facilities by a factor of one million if the facilities report pollutant discharges as concentrations in PCS. During the development of PCSLoadCalculator2004 as part of the 2007 annual review, EPA queried the Envirofacts Data Warehouse11 Web page for design flows. The design flow rate is the average flow, in MGD, that a facility is designed to accommodate. The highest design flow identified by this query was 4,453 MGD for the DC Water and Sewer Authority (DC0000221). EPA based the new 5,000 MGD cutoff on this design flow. To be consistent with the methodology used in the 2007 annual review, EPA used this cutoff again for the 2009 annual review. NODI B. The following is a discussion of a methodology change that EPA considered, but decided not to implement. NODI (no data indicator) is a single character code that indicates why pollutant measurements are blank for a reporting period. NODI = B means that the pollutant was measured BDL for that monitoring period. Typically, facilities report BDL measurements by reporting the detection limit concentration (or a mass quantity that was calculated using the detection limit concentration) and indicate the measurement is BDL using a less-than (<) sign. However, some facilities report BDL measurements by leaving the measurement value field blank and reporting B in the NODI field. Because the detection limit concentration is not provided in PCS, EPA cannot calculate period loads when the NODI B reporting method is used. 11 Envirofacts is a web-based system that allows the public to access PCS data for recent years. 3-22 ------- Section 3 - DMRLoads2007: Development and Category Rankings Table 3-7. Results of NODI Code Excluded from EST=YES Revision Analysis for PCS Point Source Category Copper Forming (40 CFR Part 468) Justice, Public Order, and Safety (SIC Group 92) Grain Mills (40 CFR Part 406) Nonferrous Metals Manufacturing (40 CFR Part 421) Canned and Preserved Seafood Processing (40 CFR Part 408) Meat and Poultry Products (40 CFR Part 432) Metal Molding and Casting (Foundries) (40 CFR Part 464) Aluminum Forming (40 CFR Part 467) Tobacco Products (PNC) Mineral Mining and Processing (40 CFR Part 436) Total DMRLoads2007 PCS Annual Load, Ib/yr a 2,110,000 1,230,000 21,400,000 152,000,000 9,900,000 53,800,000 5,860,000 13,900,000 10,700 265,000,000 35,800,000,000 PCS Annual Load with NODI Revisions, Ib/yr b 2,300,000 1,270,000 27,500,000 139,000,000 13,100,000 80,700,000 5,340,000 14,100,000 9,990 271,000,000 39,700,000,000 Difference in PCS Annual Load, Ib/yr 191,000(9.1%) 39,300 (3.2%) 6,130,000(29%) 13,000,000 (8.5%) 3,200,000 (32%) 26,900,000 (50%) 517,000(8.8%) 150,000(1.1%) 755 (7%) 6,200,000 (2.3%) 3,860,000,000 PCS TWPE, Ib- eq/yr a 77.9 49.5 437 262,000 3,120 445,000 4,940 11,900 2.95 26,700 918,000,000 PCS TWPE with NODI Revisions, Ib-eq/yr b 1,080 214 1,600 529,000 5,670 152,000 6,040 13,600 2.53 29,400 918,000,000 Difference in PCS TWPE, Ib-eq/yr 1,000 (1,290%) 164 (332%) 1,160 (265%) 267,000 (102%) 2,540 (82%) 292,000 (66%) 1,100(22%) 1,740 (15%) 0.422 (14%) 2,710 (10%) 794,000 Source: DMRLoads2007_v3. a The Total Annual Load and Total TWPE include the revised NODI codes for the EST=YES assumption and were used in generating the category rankings. b The Total Annual Load and Total TWPE with NODI Revisions include the pre-2009 NODI codes for the EST=YES assumption. PNC - Potential new category. 3-23 ------- Section 3 - DMRLoads2007: Development and Category Rankings If the pollutant is measured BDL for all 12 months of the year, then the outcome using NODI B is the same as the Hybrid Method - the total annual load is zero. However, if the pollutant is detected at least once during 2007, the EST=YES option will estimate loads for the months when the pollutant was reported as NODI B based on the detected value. For example, if a pollutant is reported as NODI B for 11 months but is measured at a concentration above its detection limit for one month, then the effect of the EST=YES option would be to multiply the detected concentration by 12 to account for the months when the facility reported NODI B. This is an overestimation of the Hybrid Method, which would use a concentration equal to half the detection limit for months when the pollutant was measured BDL. EPA considered three options for correcting the overestimation of loads for NODI B: • Option 1: Make no change. • Option 2: Exclude NODI B from the EST=YES estimation option. The EST function currently excludes a list of NODI characters that indicate that no discharge occurred for the monitoring period. Adding NODI B to the list would result in setting all BDL measurements that use the NODI B reporting method to zero, which is an underestimation of the Hybrid Method. • Option 3: Use a concentration of one-half the method detection limit (MDL) for BDL measurements if the pollutant was detected at least once for 2007. This option most closely resembles the Hybrid Method, but it would require EPA to identify MDLs for all pollutant parameters with NODI B values. Based on 2007 data, NODI B was reported for more than 250 parameters. EPA conducted an analysis to determine the impact of using the EST function for NODI B on the category rankings. EPA ran the PCS Load Calculator and generated category rankings first using EST=YES for NODI B and then using EST=NO for NODI B. EPA's analysis found that estimating for NODI B using EST=YES accounts for 708,000 Ib-eq (0.08 percent) of the TWPE from facilities mPCSLoadCalculator2007. The top 12 categories generated using EST=YES for NODI B and using EST=NO for NODI B are identical. Therefore, EPA concluded that, because using EST=YES for NODI B did not have a significant impact on the screening- level analysis, no correction was necessary for the NODI B estimation. As a result, EPA did not make any changes to the EST=YES calculation methodology for NODI B. NODI Q. Similar to the NODI B analysis discussed above, EPA evaluated the potential effects of including NODI Q in the EST=YES assumption on the category rankings. NODI Q means that the measurement was not quantifiable. A measurement is not quantifiable if the concentration was above the detection limit but the laboratory has determined that the value cannot be accurately determined. As in the NODI B discussion above, EPA considered three options for correcting the overestimation due to NODI Q: • Option 1: Make no change. • Option 2: Exclude NODI Q from the EST=YES estimation option. The EST function currently excludes a list of NODI characters that indicate that no discharge occurred for the monitoring period. Adding NODI Q to the list would result in setting all BDL measurements that use the NODI Q reporting method to zero, which is an underestimation of the Hybrid Method. 3-24 ------- Section 3 - DMRLoads2007: Development and Category Rankings • Option 3: Use a concentration of one-half the method detection limit (MDL) for BDL measurements if the pollutant was detected at least once for 2007. This option most closely resembles the Hybrid Method, but it requires EPA to identify MDLs for over 300 pollutant parameters. For the same reasons described in the NODI B section above, EPA does not have the detection limits for NODI Q records and cannot apply the EST=YES assumption using the Hybrid Method for outfalls in which a pollutant is detected at least once during the year. Because the number of records reporting NODI Q in PCS was small (0.02 percent of the total), EPA determined that there should be no change to EDS methodology. 3.2.3 ICIS-NPDES Pollutant Loading Tool The ICIS-NPDES Pollutant Loading Tool is a web-based application consisting of a user interface, business logic layer, and an Oracle™ database. The purpose of the pollutant loading tool is to calculate annual loads, similar to PCS CNVRT and PCSLoadCalculators, but for ICIS- NPDES data instead of PCS data. The ICIS-NPDES Pollutant Loading Tool contains two calculation modules: • A Convert Module that extracts ICIS-NPDES DMR data, processes and formats the data for loads calculations, and stores the converted data in an Oracle™ database; and • A Load Calculator Module that queries the Oracle™ database and calculates annual pollutant loads. To allow for consistency between the calculated PCS and ICIS-NPDES loads, the Convert Module mimics the functions of the PCS CNVRT Module while the ICIS-NPDES Load Calculator Module mimics the methodology EPA developed for the PCS Load Calculator routine that is used in PCSLoadCalculator2007'. 3.2.3.1 ICIS-NPDES Convert Module Development and Verification The ICIS-NPDES Convert Module extracts data from ICIS-NPDES tables, stores the extracted data into five denormalized tables, converts DMR measurements into standardized units of measure, identifies the statistical basis of the permit limits, and matches DMR measurements with wastewater flows and permit limits. The following describes the functions of the ICIS-NPDES Convert Module. Step 1: Extract Data and Create Denormalized Tables: The Convert Module downloads and stores the ICIS-NPDES data from 19 extracted tables into the following five interim tables: DMR, FACILITY, LIMITS, PERMIT FEATURE, and PRAM_CAS CROSSWALK. Additionally, the Convert Module also creates three lookup tables to perform the Convert Module functions: • The UNIT_CONVERSIONS table provides conversion factors for unit codes to convert concentrations into units of mg/L, loads into kg/day, and flows into MOD; 3-25 ------- Section 3 - DMRLoads2007: Development and Category Rankings • The STATS table assigns approximately 160 statistical base codes from the LIMITS table in ICIS to one of four categories: 1 = Average; 2 = Total; 3 = Maximum; and 4 = Minimum; and • The FLOW_PRAM_CODES table identifies 24 parameter codes for wastewater flow and assigns priorities that the Convert Module uses to match one flow per outfall and monitoring period for load calculations. Step 2: Convert to Standard Units: DMR data and permit limits are stored in ICIS- NPDES in the measurement units specified by facilities' NPDES permits. The ICIS-NPDES database then converts the DMR measurements and limits into standard units. The Convert Module verifies the ICIS standard units conversion in the following steps: • Identify Units of Measure - Unit codes are provided in the DMR and LIMITS tables. However, the unit code field in the DMR table is blank for most records. Therefore, the Convert Module selects unit codes from the LIMITS tables if the DMR unit code fields are blank. • Verify ICIS Unit Conversions - In this step, the Convert Module back-calculates the ICIS conversion factors by dividing the standard units values by the original values. EPA then compared the ICIS conversion factors to conversion factors that EPA specified in the UNIT_CONVERSIONS look-up table (Table B-4 in Appendix B). EPA corrected the ICIS-NPDES conversions that did not match the look-up table. The Convert Module corrected approximately 0.3 percent of the DMR records in ICIS-NPDES. Step 3: Assign Statistical Base Codes: ICIS data contain approximately 160 statistical base codes to describe the statistical basis of the DMR measurements (e.g., 30-day geometric mean or rolling average).These codes are stored in the LIMITS table. The Convert Module uses the STATS look-up table (Table B-5 in Appendix B) to assign each statistical base code to one of five categories: • 0 = Statistical Base Code is Null; • 1 = Average; • 2 = Total; • 3 = Maximum; and • 4 = Minimum. Although specific information regarding the statistical basis of the measurement is lost during this step, the simplification is necessary for efficient calculation of loads. The Convert Module creates a STATS code (see Figure 3-2 for an example STATS code) consisting of five characters. Each character of the STATS code corresponds to one measurement value field. Step 4: Select Flows: The Convert Module selects DMR measurements for parameters that are identified as wastewater flows in the FLOW_PRAM_CODES look-up table, and matches flows with pollutant measurements. EPA identified 24 flows that are appropriate for loads calculations12. EPA assigned priorities to the PRAM codes in the FLOW PRAM CODES 12 ICIS-NPDES contains other flow parameters such as recirculation flow, flow into well, and flows reported as percentages or number of occurrences. EPA determined that these flows were not appropriate to use in calculating mass discharges to receiving streams. 3-26 ------- Section 3 - DMRLoads2007: Development and Category Rankings look-up table (Table B-6 in Appendix B). If a facility reports more than one type of flow PRAM for the same outfall, then the Convert Module selects the PRAM code with the highest priority in the look-up table. As a result, the Convert Module selects only one flow for each outfall, monitoring location, and monitoring period end date. Step 5: Select Temperature and pH: The Convert Module creates two new columns in the CONVERT_DMR table for wastewater stream temperature and pH. ICIS-NPDES contains two parameter codes for temperature and one parameter code for pH: • Temperature Degrees C = PRAM 00010; • Temperature Degrees F = PRAM 00011; and pH = PRAM 00400. The Convert Module uses a measurement value selection hierarchy, based on the STATS codes created in Step 4, to select one pH and one temperature for each permitted feature, monitoring location, and monitoring period end date. Step 6: Identify Number of Report: The Convert Module identifies the number of days per monitoring period using the NMBR_OF_REPORT field and the MONITORING_PERIOD_END_DATE field. The NMBR_OF_REPORT field indicates the number of months of discharges represented on each DMR. For example, a NMBR_OF_REPORT of 1 indicates a monthly report and a NMBR_OF_REPORT of 3 is a quarterly report (i.e., three months are in a quarter). EPA reviewed the ICIS-NPDES data and identified the following valid NMBR_OF_REPORT values: • 1 = Monthly Report; • 2 = Bi-monthly Report; • 3 = Quarterly Report; • 4 = Triannual Report (typically for April, August, and December); • 6 = Semi-annual Report; and • 12 = Annual Report. However, due to a data-entry rule in ICIS, some DMR records have invalid NMBR_OF_REPORT values, such as 5, 7, 8, 9, 10, 11 or greater than 12. Because ICIS does not allow users to enter a monitoring period start date that is earlier than the permit effective date, facilities whose permits are renewed part-way through the year cannot enter valid NMBR_OF_REPORT values. For example, if a facility submitted a semi-annual DMR in June, which covered discharges from January to June, but their NPDES permit was renewed in February, then ICIS will not allow the facility to enter a monitoring period start date earlier than the effective date of the permit (February). As a result the NMBR_OF_REPORT field in ICIS is 5 instead of 6. The Convert Module addresses this issue by rounding up invalid NMBR_OF_REPORT values to the next valid value. In addition, NMBR_OF_REPORT values that are greater than 12 are converted to 12. Once all NMBR_OF_REPORT have been converted to valid values, the Convert Module uses Table 3-8 to assign the actual number of days for the monitoring period. This table presents the actual number of days for all possible MONITORING_PERIOD_END_DATE and NMBR_OF_REPORT combinations. 3^27 ------- Section 3 - DMRLoads2007: Development and Category Rankings Table 3-8. Actual Number of Days per Monitoring Period Monitoring Period End Date Month January February a March April May June July August September October November December Number of Report 1 31 28 31 30 31 30 31 31 30 31 30 31 2 62 59 59 61 61 61 61 62 61 61 61 61 3 92 90 90 89 92 91 92 92 92 92 91 92 4 123 120 121 120 120 122 122 123 122 123 122 122 6 184 181 182 181 182 181 181 184 183 184 183 184 12 365 365 365 365 365 365 365 365 365 365 365 365 "Does not account for the number of days in February during leap years. As a final step for assigning the number of days per monitoring period, the Convert Module identifies and corrects monitoring periods with multiple reported measurements. For example, if a facility's NPDES permit requires them to report wastewater selenium discharges on both January 15 and January 30, the Loading Tool would overestimate the annual selenium load because it would multiply both the January 15 and January 31 discharges by 31 days per month according to Table 3-8. To eliminate this overestimation, the Convert Module divides the NMBR_OF_REPORT and the NMBR_OF_DAYS by the number of DMRs submitted per monitoring period. Using the above example, the Convert Module calculates the NMBR_OF_REPORT and NMBR_OF_DAYS for the January DMRs as follows: NMBR OF REPORT = _ NMBR_OF_REPORT (l) _ 2 Reports per month = 0.5 NMBR OF DAYS NMBR_OF_DAYS (31) _ 2 Reports per month = 15.5 Step 7: Correct Flows: The Convert Module corrects flows. The methodology corrects all flows exceeding 5,000 MGD, and applies more conservative criteria to correct flows from 1,000 to 5,000 MGD. The Convert Module uses three types of erroneous flow indicators to correct flows: 1. Type 1: Month-to-Month Variability: In this step, the Convert Module compares flows reported for each month and identifies variations greater than three orders of magnitude using the following procedure: 3-28 ------- Section 3 - DMRLoads2007: Development and Category Rankings a. Create a field that identifies the magnitude of each flow (e.g., 62,800 MOD has a magnitude of 10,000); b. Group flow magnitudes by LIMIT_ID13; c. Find the minimum flow magnitude that is >1,000; d. Find the maximum flow magnitude that is <1,000; e. Calculate a flow correction factor by dividing Step C/Step D; f If the correction factor indicates a difference of three orders of magnitude or more, and the actual measured flow is 1,000 to 5,000 MOD, then correct the flow as follows: Corrected Flow (MGD) = Actual Measured Flow (MGD): Maximum Flow Magnitude < 1,000 ^| Minimum Flow Magnitude > 1,000 J g. If the correction factor indicates a difference of one order of magnitude or more, and the actual measured flow is > 5,000 MGD, then correct the flow using the equation in step f). Table 3-9 presents an example of a Type 1 flow correction that the Convert Module identified. As shown in the table, the September 2007 flow is three orders of magnitude higher than the flows reported for other monitoring periods. Therefore, the Convert Module divided the September flow by 1,000. Table 3-9. Example Type 1 Flow Correction External Permit Number GA0037648 GA0037648 GA0037648 GA0037648 GA0037648 GA0037648 GA0037648 GA0037648 GA0037648 GA0037648 GA0037648 GA0037648 Permitted Feature Number OBO OBO OBO OBO OBO OBO OBO OBO OBO OBO OBO OBO Monitoring Period End Date 31-Mar-07 30-Apr-07 31-May-07 30-Jun-07 31-Jul-07 31-Aug-07 30-Sep-07 31-Oct-07 31-Dec-07 28-Feb-07 31-Jan-07 30-Nov-07 Original Flow 0.74 0.54 0.67 1.31 1.02 1.06 2,554.00 1.24 1.29 0.96 1.02 0.85 Maximum Flow Magnitude < 1,000 Minimum Flow Magnitude > 1,000 Correction Factor Flow Magnitude 0.1 0.1 0.1 1 1 1 1000 1 1 0.1 1 0.1 1 1000 1000 New Flow 0.74 0.54 0.67 1.31 1.02 1.06 2.55 1.24 1.29 0.96 1.02 0.85 Correction Applied? No No No No No No Yes No No No No No The LIMIT_ID is a unique identifier in the ICIS_LIMITS table. It is a primary key for each unique set of parameter code, limit set, season, and permitted feature. 3-29 ------- Section 3 - DMRLoads2007: Development and Category Rankings 2. Type 2: Variations from Design Flows and Actual Average Flows in FACILITIES: The FACILITIES table contains information for facility design flow and actual average flow in MGD. These fields are not required and therefore are not populated for all records. However, when populated, these fields can be used to help evaluate the reasonableness of the flows reported in the DMR data. The Convert Module compares the design flow and actual average flow in FACILITIES to the reported flows in CONVERT_DMR using the following procedure: a. Use ACTUAL_AVG_FLOW if reported. If ACTUAL_AVG_FLOW is not reported, then use DESIGN_FLOW. b. Use similar procedure as Type 1 to calculate the magnitude of the reported flows and the actual/design flow magnitudes. c. Divide the reported flow (e.g., FQ1) magnitude by the actual/design flow magnitude to calculate the correction factor. d. If the correction factor indicates a difference of three orders of magnitude or more, and the actual measured flow is 1,000 to 5,000 MGD, then correct the flow as follows: Corrected Flow (MGD) = Actual Measured Flow (MGD) ( Design Flow Magnitude ^ ^Actual Measured Flow Magnitude J e. If the correction factor indicates a difference of one order of magnitude or more, and the actual measured flow is > 5,000 MGD, then correct the flow using the equation in step d. Table 3-10 presents an example of a Type 2 flow correction that the Convert Module identified. As shown in the table, the reported flows (FQ1) were six orders of magnitude higher than the facility design flow. Therefore, the Convert Module divided all flows by 1,000,000. Table 3-10. Example Type 2 Flow Correction External Permit Number NHO 100692 NHO 100692 NHO 100692 NHO 100692 NHO 100692 NHO 100692 NHO 100692 NHO 100692 NHO 100692 NHO 100692 Permitted Feature Number 001 001 001 001 001 001 001 001 001 001 Monitoring Period End Date 31-Jan-07 28-Feb-07 31-Mar-07 30-Apr-07 31-May-07 30-Jun-07 31-Jul-07 31-Aug-07 30-Sep-07 31-Oct-07 Original Flow 250,038 131,243 203,087 308,359 382,444 460,524 308,488 154,491 161,996 158,444 Flow Magnitude 100,000 100,000 100,000 100,000 100,000 100,000 100,000 100,000 100,000 100,000 Design Flow 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 Design Flow Magnitude 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 Correction Factor 1,000,000 1,000,000 1,000,000 1,000,000 1,000,000 1,000,000 1,000,000 1,000,000 1,000,000 1,000,000 New Flow 0.25 0.13 0.20 0.31 0.38 0.46 0.31 0.15 0.16 0.16 3-30 ------- Section 3 - DMRLoads2007: Development and Category Rankings Table 3-10. Example Type 2 Flow Correction External Permit Number NHO 100692 NHO 100692 Permitted Feature Number 001 001 Monitoring Period End Date 30-Nov-07 31-Dec-07 Original Flow 183,168 190,775 Flow Magnitude 100,000 100,000 Design Flow 0.5 0.5 Design Flow Magnitude 0.1 0.1 Correction Factor 1,000,000 1,000,000 New Flow 0.18 0.19 3. Type 3: Flows Exceeding the 5,000 MGD Cap: If a reported flow exceeds 5,000 MOD and is not identified for the Type 1 or Type 2 corrections, then the Convert Module assumes that the flow was incorrectly entered in units of GPD and divides the flow by 1,000,000. Step 8: Select Effluent Monitoring Location: Permits often require a facility to monitor at multiple locations. The monitoring location is indicated in the CONVERT_DMR table in the MONITORING_LOCATION_CODE (MLOC) field. Five monitoring location codes in ICIS- NPDES represent effluent discharges, seen below. For its screening level review, EPA estimates annual loads that represent effluent discharges. Like PCSLoadCalculator2007, the ICIS-NPDES Load Calculator searches the monitoring location field to identify effluent data only. When more than one type of effluent data are present for an outfall, MLOC is selected in the following hierarchy: MLOC 2 - Effluent gross discharge; MLOC 1 - Effluent net discharge; MLOC A - After disinfection; MLOC B - Before disinfection; and MLOC SC - See comments. For example, if a facility reports both MLOC 1 and MLOC 2, MLOC 2 is used in preference to MLOC 1. tables: In executing the above steps, the Convert Module creates the following four output CONVERT DMR; FACILITY; PERMIT FEATURE; and PRAM CAS CROSSWALK. Figure 3-4 shows the relationship diagram for the Convert Module Output. The CONVERT_DMR table contains year 2007 DMR measurements for over 10,000 permits, of which approximately 80 percent are individual NPDES permits, 15 percent are general permits, and the remaining five percent include industrial user permits and state-issued non-NPDES permits. 3-31 ------- Section 3 - DMRLoads2007: Development and Category Rankings FACILITY Table CONVERT DMR Table PERMIT FEATURE Table ICIS FACILITY INTEREST ID EXTERNAL PERMIT NMBR FACILITY UIN MAJOR MINOR STATUS FLAG CITY STATE CODE COUNTY CODE ZIP HUC_CODE EPA_REGION_CODE STATE_WATER_B OD Y_NAME TOTAL_DE SIGN_FLOW_NMBR ACTUAL AVERAGE FLOW NMBR PERMIT TYPE CODE FACILITY TYPE CODE GEOCODE LATITUDE GEOCODE LONGITUDE - ACTIVITYJD 4 > ACTIVITY ID EXTERNAL PERMIT NMBR PERM FEATURE ID PERM FEATURE NMBR LIMIT SET DESIGNATOR LIMIT_SET_ID LQ1 LQ2 LCI LC2 LC3 LIMIT ID STAT5 MONITORING PERIOD END DATE NMBR OF REPORT NMBR OF SUBMISSION NMBR OF DAYS MONITORING LOCATION CODE PARAMETER CODE NODI CODE Ql QUAL Cl QUAL C2 QUAL C3 QUAL MQ1 MQ2 MCI MC2 MC3 FSAT5 FQ1 FQ2 FC1 FC2 FC3 AVG_TEMP AVG_PH LIMIT_SET_SCHED ULE_ID DMR_FORM_ID DMR_EVENT_ID DMR_FORM_PARAMETER_ID DMR_PARAMETER_ID EXTERNAL PERMIT NMBR PERM FEATURE NMBR PERM FEATURE ID DESIGN FLOW NMBR ACTUAL AVERAGE FLOW NMBR WATER BODY NAME REACH ID STATE WATER BODY NAME LATTTUDE_MEASURE LONGITUD E_MEASURE PARAMETER_CODE PARAMETER DESC POLLUTANT DESC CHEMICAL ABSTRACT SERVICE NMBR Figure 3-4. Relationship Diagram for Convert Module Output 3-32 ------- Section 3 - DMRLoads2007: Development and Category Rankings 3.2.3.2 ICIS-NPDES Load Calculator Module Annual Load Calculation This section describes the calculation steps used by the ICIS-NPDES Load Calculator Module to produce annual loads from the ICIS-NPDES Convert Module output tables. As stated in Section 3.2.3, the ICIS-NPDES Load Calculator was developed to mimic the methodology EPA developed for the PCS Load Calculator routine (see Section 3.2.2.2). EPA developed the ICIS-NPDES Pollutant Loading Tool as a user-guided, web-based module that includes functions beyond calculating category rankings for the annual review. The following is a description of the steps taken by the ICIS-NPDES Load Calculator for selecting monitoring location, flows, and measurement values and calculating annual loads. Step 1: Measurement Value Selection. The CONVERT_DMR table (depicted in Figure 3-4) stores DMR data extracted from ICIS-NPDES in five measurement value fields. These include: Quantity 1 (MQ1); • Quantity 2 (MQ2); • Concentration 1 (MCI); • Concentration 2 (MC2); and • Concentration 3 (MC3). These measurement value fields correspond to the five DMR fields where quantity and concentration data are stored: 1) Average Quantity (Quantity 1 or 2); 2) Maximum Quantity (Quantity 1 or 2); 3) Minimum Concentration (Concentration 1, 2, or 3); 4) Average Concentration (Concentration 1, 2, or 3); and 5) Maximum Concentration (Concentration 1, 2, or 3). Note that unlike PCS, the measurement value fields in ICIS-NPDES are not specific to average, maximum, or minimum. The statistical basis of the measurements in ICIS-NPDES is determined by the five-digit statistical base code associated with each measurement field. Facilities may use a variety of measurements to populate the five measurement value fields. For example, a facility can use a monthly average, daily average, 30-day geometric average, etc. to represent the average quantity. The CONVERT_DMR table contains a five-digit statistical base code (STATS). The following codes are used for the types of measurements that may be reported: • Average (STAT5=1); Total (STAT5=2); Maximum (STAT5=3); • Minimum (STAT5=4); and Null (STAT5=0). Each of the five digits in the STATS field corresponds to one of the five measurement value fields. Figure 3-5 shows an example of a possible STATS code. In this figure, the measurements reported for MQ1, MC2, and MC3 are average values, MQ2 is a maximum value, and no value was reported for MCI. The ICIS-NPDES STATS example shown in Figure 3-5 is the same as that shown for PCS in Figure 3-2 except that it was updated to show how the STATS digits correspond to ICIS-NPDES measurement value fields. 3^33 ------- Section 3 - DMRLoads2007: Development and Category Rankings 13011 Quantity 1 Quantity 2 Concentration 1 Concentration 2 Concentration 3 (MQ1) (MQ2) (MCI) (MC2) (MC3) Figure 3-5. Example STATS Code in ICIS-NPDES CONVERT DMR Table The ICIS-NPDES Load Calculator selects measurements for loads calculations using a hierarchy that prioritizes average values and quantities. In the first step, the ICIS-NPDES Load Calculator attempts to identify an average value (STAT=1) by searching the STATS digits from left to right. By scanning left to right, the ICIS-NPDES Load Calculator searches the STATS digits that correspond to measurement fields in the following sequence: • Quantity 1 (MQ1); • Quantity 2 (MQ2); • Concentration 1 (MCI); • Concentration 2 (MC2); and • Concentration 3 (MC3). If the ICIS-NPDES Load Calculator finds an average value (STAT5=1), then it selects the corresponding measurement for load calculation and performs the following calculations: • If the selected measurement is a quantity (MQ1 or MQ2): — Average daily load (kg/day) = MQ — Average concentration (mg/L) = MQ/(Flow (MOD) x3.785 (L/gal)) — Monitoring Period Load (kg/monitoring period) = MQ xNMBR OF DAYS — Monitoring Period Load Over Limit (LOL) = (MQ - LQ (Quantity Limit)) xNMBR OF DAYS • If the selected measurement is a concentration (MCI, MC2, or MC3): — Average daily load (kg/day) = MC * Flow * 3.785 — Average concentration (mg/L) = MC — Monitoring Period Load (kg/monitoring period) = MC x Flow (MGD) x 3.785 (L/gal) x NMBR_OF_DAYS — Monitoring Period LOL = (MC - LC (Concentration Limit)) x Flow x 3.785 x NMBR_OF_DAYS If the ICIS-NPDES Load Calculator does not find an average value (STAT=1), then it scans STATS from left to right for a total value (STAT=2). "Total" values apply only to quantity measurements, and because these measurements represent the total mass discharge for the monitoring period, the ICIS-NPDES Load Calculator cannot use the same calculations used for average, maximum, and minimum values. If the ICIS-NPDES Load Calculator identifies a total value, it selects the value and performs the following calculations: 3-34 ------- Section 3 - DMRLoads2007: Development and Category Rankings Average Daily Load (kg/day) = MQ / NMBR_OF_DAYS • Average Concentration (mg/L) = MQ/(Flow x NMBR_OF_DAYS x 3.785) • Monitoring Period Load (kg/monitoring period) = MQ • Monitoring Period LOL = MQ - LQ If the ICIS-NPDES Load Calculator does not find an average value (STAT=1) or a total value (STAT=2), then it scans STATS from left to right for a maximum value (STAT=3). If the Load Calculator identifies a maximum value, then it selects that value and performs the same calculations used for the average values (STAT=1). If the ICIS-NPDES Load Calculator does not find an average value (STAT=1), total value (STAT=2), or maximum value (STAT=3), then it scans STATS from left to right for a minimum value (STAT=4). If the ICIS-NPDES Load Calculator identifies a minimum value, then it selects that value and performs the same calculations used for the average values (STAT=1). Finally, if all measurement value fields are blank, then the ICIS-NPDES Load Calculator sets the average daily load, average concentration, monitoring period load, and load-over-limit fields to null. Table 3-11 presents the measurement value selection priorities and calculations. Step 2: Flow Value Selection. The ICIS-NPDES Load Calculator uses a similar hierarchy for selecting flow rates and the FSTAT5 code. The FSTAT5 code applies the same concept as the STATS code, and provides information about the statistical basis of wastewater flow values. Similar to the measurement value selection hierarchy, the flow selection hierarchy prioritizes average flows. First, the ICIS-NPDES Load Calculator attempts to find an average flow (FSTAT=1) by scanning the FSTATS code from left to right. By scanning from left to right, the ICIS-NPDES Load Calculator searches the F STATS digits corresponding to the flow values in the following sequence: • Flow Quantity 1 (FQ1); • Flow Quantity 2 (FQ2); • Flow Concentration 1 (FC1)14; • Flow Concentration 2 (FC2); and • Flow Concentration 3 (FC3). If the ICIS-NPDES Load Calculator finds an average value (STAT5=1), then it selects the corresponding flow for load calculation and performs the following calculations: • Average Daily Flow (MGD) = Flow • Monitoring Period Flow (MG/monitoring period) = Flow x NMBR_OF_DAYS 14 A "flow concentration" is a flow measurement that was reported to a concentration measurement field. Facilities may report flows in any of the five measurement value fields. However, all flows are reported in units of MGD whether they are reported in a quantity field or a concentration field. 3-35 ------- Section 3 - DMRLoads2007: Development and Category Rankings Table 3-11. Measurement Value Selection Priorities and Calculations Priority 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Value Type MQ1 MQ2 MCI MC2 MC3 MQ1 MQ2 MQ1 MQ2 MCI MC2 MC3 MQ1 MQ2 MCI MC2 STAT Code 1 1 1 1 1 2 2 o J o J o J o J o J 4 4 4 4 Average Daily Load (KGD) MQ1 MQ2 MCI x Flow x 3.785 MC2 x Flow x 3.785 MC3 x Flow x 3.785 MQ1/ NMBR_OF_DAYS MQ2/ NMBR_OF_DAYS MQ1 MQ2 MCI x Flow x 3.785 MC2 x Flow x 3.785 MC3 x Flow x 3.785 MQ1 MQ2 MCI x Flow x 3.785 MC2 x Flow x 3.785 Average Concentration (MP MGL) MQl/(Flow x 3.785) MQ2/(Flow x 3.785) MCI MC2 MC3 MQ1/(MP_MGD x 3.785) MQ2/(MP_MGD x 3.785) MQl/(Flow x 3.785) MQ2/(Flow x 3.785) MCI MC2 MC3 MQl/(Flow x 3.785) MQ2/(Flow x 3.785) MCI MC2 Monitoring Period DMR Load (KGMP) MQ1 x NMBR_OF_DAYS MQ2 x NMBR_OF_DAYS MCI x Flow x 3.785 x NMBR_OF_DAYS MC2 x Flow x 3.785 x NMBR_OF_DAYS MC3 x Flow x 3.785 x NMBR_OF_DAYS MQ1 MQ2 MQ1 x NMBR_OF_DAYS MQ2 x NMBR_OF_DAYS MCI x Flow x 3.785 x NMBR_OF_DAYS MC2 x Flow x 3.785 x NMBR_OF_DAYS MC3 x Flow x 3.785 x NMBR_OF_DAYS MQ1 x NMBR_OF_DAYS MQ2 x NMBR_OF_DAYS MCI x Flow x 3.785 x NMBR_OF_DAYS MC2 x Flow x 3.785 x NMBR_OF_DAYS MP LOL (MQ1-LQ1) x NMBR_OF_DAYS (MQ2 - LQ2) x NMBR_OF_DAYS (MCI -LCI) x Flow x 3.785 x NMBR_OF_DAYS (MC2 -LC2) x Flow x 3.785 x NMBR_OF_DAYS (MC3 -LC3) x Flow x 3.785 x NMBR_OF_DAYS MQ1-LQ1 MQ2 - LQ2 (MQ1-LQ1) x NMBR_OF_DAYS (MQ2 - LQ2) x NMBR_OF_DAYS (MCI -LCI) x Flow x 3.785 x NMBR_OF_DAYS (MC2 -LC2) x Flow x 3.785 x NMBR_OF_DAYS (MC3 -LC3) x Flow x 3.785 x NMBR_OF_DAYS (MQ1-LQ1) x NMBR_OF_DAYS (MQ2 - LQ2) x NMBR_OF_DAYS (MCI - LCI) x Flow x 3.785 x NMBR_OF_DAYS (MC2 -LC2) x Flow x 3.785 x NMBR_OF_DAYS 3-36 ------- Section 3 - DMRLoads2007: Development and Category Rankings Table 3-11. Measurement Value Selection Priorities and Calculations Priority 17 18 Value Type MC3 No Data (NODI is not null) STAT Code 4 Any Average Daily Load (KGD) MC3 x Flow x 3.785 NULL Average Concentration (MP MGL) MC3 NULL Monitoring Period DMR Load (KGMP) MC3 x Flow x 3.785 x NMBR_OF_DAYS NULL MP LOL (MC3 -LC3) x Flow x 3.785 x NMBR_OF_DAYS NULL 3-37 ------- Section 3 - DMRLoads2007: Development and Category Rankings If the ICIS-NPDES Load Calculator does not find an average flow (FSTAT=1), then it scans FSTAT5 from left to right for a total flow (FSTAT=2). Because "total" flows represent the total wastewater discharge for the monitoring period, the ICIS-NPDES Load Calculator cannot use the same calculations used for average, maximum, and minimum flows. If the ICIS-NPDES Load Calculator identifies a total flow, it selects the value and performs the following calculations: • Average Daily Flow (MOD) = Flow/MNBR_OF_DAYS • Monitoring Period Flow (MG/monitoring period) = Flow If the ICIS-NPDES Load Calculator does not find an average flow (FSTAT=1) or a total flow (FSTAT=2), then it scans FSTAT5 from left to right for a maximum flow (FSTAT=3). If the Load Calculator identifies a maximum flow, then it selects that flow and performs the same calculations used for the average flows (FSTAT=1). If the ICIS-NPDES Load Calculator does not find an average flow (FSTAT=1), total flow (FSTAT=2), or maximum value (FSTAT=3), then it scans FSTAT5 from left to right for a minimum flow (FSTAT=4). If the ICIS-NPDES Load Calculator identifies a minimum flow, then it selects that flow and performs the same calculations used for the average flows (FSTAT=1). Table 3-12 presents the flow value selection priorities and calculations. Table 3-12. Flow Value Selection Priorities Priority 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 Value Type Quantity 1 Quantity 2 Concl Cone 2 Cone 3 Quantity 1 Quantity 2 Quantity 1 Quantity 2 Concl Cone 2 Cone 3 Quantity 1 Quantity 2 Cone 1 Cone 2 Cone 3 FSTAT5 1 1 1 1 1 2 2 3 3 3 3 3 4 4 4 4 4 Average Daily Flow (MGD) FQ1 FQ2 FC1 FC2 FC3 FQ1 / NMBR_OF_DAYS FQ2 / NMBR_OF_DAYS FQ1 FQ2 FC1 FC2 FC3 FQ1 FQ2 FC1 FC2 FC3 Monitoring Period Flow (MGMP) FQ1 * NMBR_OF_DAYS FQ2 * NMBR_OF_DAYS FC1 * NMBR_OF_DAYS FC2 * NMBR_OF_DAYS FC3 * NMBR_OF_DAYS FQ1 FQ2 FQ1 * NMBR_OF_DAYS FQ2 * NMBR_OF_DAYS FC1 * NMBR_OF_DAYS FC2 * NMBR_OF_DAYS FC3 * NMBR_OF_DAYS FQ1 * NMBR_OF_DAYS FQ2 * NMBR_OF_DAYS FC1 * NMBR_OF_DAYS FC2 * NMBR_OF_DAYS FC3 * NMBR_OF_DAYS 3-38 ------- Section 3 - DMRLoads2007: Development and Category Rankings Step 3: Detection Limit Options (DL). When pollutants are not detected, their concentrations are presumed to be below their detection limit (BDL). Permittees may report the detection limit with a less-than sign (<) to indicate that the pollutant was BDL. The CONVERT_DMR table stores the less-than signs for nondetects in the data qualifier field that corresponds to the measurement value (i.e., MQl_Qual, MQ2_Qual, MCl_Qual, MC2_Qual, or MC3_Qual). If a pollutant is BDL, the pollutant concentration may be between zero and the detection limit. The ICIS-NPDES Load Calculator calculates three versions of each monitoring period load and concentration using each of the following assumptions: • BDL equals zero; • BDL equals the detection limit; or • BDL equals one-half the detection limit. Step 4: Calculate Load-Over-Limit. The ICIS-NPDES Load Calculator performs some calculations that are not performed on PCS data, but may be used as part of EPA's screening- level review. The Load-Over-Limit (LOL) compares the monitoring period loads to the NPDES permit limits on a mass basis. LOL is not used in rankings categories but may be used in future reviews. LOL is calculated using the following steps: • Select the limits from the limit value fields that correspond to the selected measurement value fields; • Calculate a monitoring period load over limit using one of the following equations: — Calculation of monitoring period LOL from mass quantity (MQ1 or MQ2): Monitoring Period LOL (kg/period) = [(MQ (kg/day) - LQ (kg/day)] x NMBR_OF_DAYS — Calculation of monitoring period LOL from concentration and flow (MCI, MC2, orMC3): Monitoring Period LOL (kg/period) = [(MC (mg/L) - LC (mg/L)] x Flow (MGD) x 3.735 (L/gal) x NMBR_OF_DAYS • Create two options for Monitoring Period Load-Over-Limit Calculations: — LOL1 - If the Monitoring Period LOL is negative, then set the LOL to zero; — LOL2 - If the Monitoring Period LOL is negative, then retain the calculated negative value. For example, if the DMR Monitoring Period Load is 200 and the Limit Monitoring Period Load is 205, then the result for LOL1 would be 0 and the result for LOL2 would be -5. If the DMR Monitoring Period Load is 210 and the limit is 205 then the results for LOL1 and LOL2 would both equal 5. In Step 5, the ICIS-NPDES Load Calculator sums the LOL1 and LOL2 values for the year to calculate the annual load over limit. The LOL1 method provides the total annual kilograms of pollutant discharges that exceeded permit limits, but does not give credit for monitoring periods 3^39 ------- Section 3 - DMRLoads2007: Development and Category Rankings where the discharges were below the permit limit. Summing the LOL2 values provides the net permit limit exceedances for the year. Step 5: Calculate Annual Totals and Averages. In this step, the ICIS-NPDES Load Calculator calculates the following annual averages and totals for each facility, outfall, monitoring location, limit set designator, and parameter: • Total Annual Pollutant Load (kg/yr) = Sum of Monitoring Period Loads for three DL Options: — BDL = 0 — BDL = !/2 DL; and — BDL = DL; • Annual Average Pollutant Load (kg/day) = Avg of Average Daily Loads for three DL Options: — BDL = 0; — BDL = !/2 DL; and — BDL = DL; • Total Annual Wastewater Flow (MG/yr) = Sum of Monitoring Period Flows; • Annual Average Wastewater Flow (MGD) = Avg of Average Daily Flows; • Annual Average Concentration (mg/L) = Avg of Concentrations for three DL Options: — BDL = 0; — BDL = !/2 DL; and — BDL = DL; • Total Load-Over-Limit 1 = Sum of Monitoring Period LOL1; • Total Load-Over-Limit 2 = Sum of Monitoring Period LOL2; and • Annual Average Temperature and pH. Step 6: Estimation Function (EST). Like PCSLoadCalculator2007 (see Section 3.2.2.2), the ICIS-NPDES Load Calculator estimates discharges for monitoring periods with NODI codes that indicate discharge did not occur. The ICIS-NPDES Load Calculator assumes no discharge for the same NODI codes as PCSLoadCalculator2007: • 2: Operations shutdown; • 4: Discharge to Lagoon/Groundwater; • 7: No Influent; • 9: Conditional Monitoring; • C: No discharge; • I: Land Applied; • J: Recycled - Water-Closed System; and • W: Dry Lysimeter/Well. 3^40 ------- Section 3 - DMRLoads2007: Development and Category Rankings The ICIS-NPDES Load Calculator uses an identical methodology as the EST=YES function used by PCSLoadCalculator2007; the ICIS-NPDES Load Calculator normalizes the calculated annual load to 12 months per year using the following equation: 12 Annual Load (kg/yr) = Sum of Monitoring Period Loads x' Sum of NMBR_OF_REPORT EPA developed the ICIS-NPDES Load Calculator to allow users to selectively query loads calculated using this estimation function (EST=YES) or without the estimation function (EST=NO). Therefore, the ICIS-NPDES Load Calculator differs fromPCSLoadCalculator2007 in that only one set of annual loads is included in the final ICIS-NPDES annual loads output table. EPA ran the ICIS-NPDES Load Calculator using EST=YES to develop the annual loads for the 2009 annual review. Step 7: Parameter Grouping. An NPDES permit may require a facility to measure a pollutant in more than one way. For example, a facility may report both total lead and dissolved lead. Because total lead includes dissolved lead, adding the two measurements together overestimates the mass of lead discharged from the facility. To avoid double counting, the ICIS- NPDES Load Calculator can group parameters that represent a single pollutant more accurately15. The ICIS-NPDES Load Calculator grouping function uses a hierarchy to determine which parameter best represents the total pollutant discharge. For example, copper has eight parameter codes. If a facility reports multiple parameter codes for copper, the ICIS-NPDES Load Calculator uses the following "grouping" hierarchy: 1. Total copper; 2. Copper; 3. Total copper per batch; 4. Total recoverable copper; 5. Dry weight copper; 6. Potentially dissolved copper; and 7. Sum of (dissolved copper and suspended copper). Table B-7 in Appendix B presents the parameter grouping hierarchy. Load Calculator Module Output. The ICIS-NPDES Pollutant Loading Tool output contains loads, concentrations, flows, and wastewater stream conditions for each facility, outfall, monitoring location, and parameter as seen in Table 3-13. 15 EPA also groups parameters in PCS using the same parameter grouping as ICIS-NPDES. The PCS parameter grouping occurs inDMRLoadsAnalysis2007. See Section 3.2.4.1 for additional details. 3-41 ------- Section 3 - DMRLoads2007: Development and Category Rankings Table 3-13. ICIS-NPDES Load Calculator Module Output Output Parameter EST Option DL Option Used in Category Rankings? Purpose Total Annual Pollutant Loads KGYOO KGYEO KGY10 KGY01 KGYE1 KGY11 No No No Yes Yes Yes BDL = 0 BDL = !/2 DL BDL = DL BDL = 0 BDL = !/2 DL BDL = DL No No No Yes Yes No Not used in Rankings or Sensitivity Analysis Not used in Rankings or Sensitivity Analysis Not used in Rankings or Sensitivity Analysis Used with KGYE1 to calculate Hybrid Used with KGY01 to calculate Hybrid DL Sensitivity Analysis Annual Average Daily Loads AVGJCGDO AVGJCGDE AVGJCGD1 NA NA NA BDL = 0 BDL = !/2DL BDL = DL No No No Not used in Rankings or Sensitivity Analysis Not used in Rankings or Sensitivity Analysis Not used in Rankings or Sensitivity Analysis Annual Average Concentrations AVG_MGLO AVG_MGLE AVG_MGL1 NA NA NA BDL = 0 BDL = !/2DL BDL = DL No No No Not used in Rankings or Sensitivity Analysis Not used in Rankings or Sensitivity Analysis Not used in Rankings or Sensitivity Analysis Annual Load-Over-Limit SUM_LOL1 SUM_LOL2 NA NA NA NA No No Not used in Rankings or Sensitivity Analysis Not used in Rankings or Sensitivity Analysis Average Wastewater Stream Conditions AVG_TEMP AVG_PH NA NA NA NA No No Not used in Rankings or Sensitivity Analysis Not used in Rankings or Sensitivity Analysis Total Annual Wastewater Flow MGY NA NA No Not used in Rankings or Sensitivity Analysis Annual Average Daily Wastewater Flow AVG_MGD NA NA No Not used in Rankings or Sensitivity Analysis NA = Not Applicable 3.2.4 DMRLoadsAnalysis2007 As depicted in Figure 3-6, the DMRLoadsAnalysis2007 database imports annual load tables from PCSLoadCalculator2007 and the ICIS-NPDES Load Calculator and facility information from PCS and ICIS-NPDES. PCSLoadCalculator2007 creates annual loads from using the hybrid methodology but without grouping the parameters while ICIS-NPDES Load Calculator groups the parameters but does not apply the hybrid methodology. To create one set of annual loads from the two data sets, DMRLoadsAnalysis2007 applies the hybrid methodology to the output from the ICIS-NPDES Load Calculator and groups the parameters in the annual load output from PCSLoadCalculator2007'. DMRLoadsAnalysis2007 then uses information facility information from PCSFAC and ICIS Facilities and Chemical Abstract Services (CAS) numbers to calculate TWPE and create the "combined" annual loads table ("DMR2007") that is used by DMRLoads2007 to generate category rankings. In 2007, 64 percent of the records in the DMR2007 table were from PCS, while the remaining 36 percent were from ICIS-NPDES. Additionally, DMRLoadsAnalysis2007 uses annual loads from the PCSLoadCalculator2007 and 3-42 ------- Section 3 - DMRLoads2007: Development and Category Rankings ICIS-NPDES Load Calculator outputs to perform a sensitivity analysis on the various calculation assumptions. Table 3-14 describes the function of each table inDMRLoadsAnalysis2007. Table 3-14. Tables Imported or Created in DMRLoadsAnalysis2007 Table Name PRAM Codes - PCS PRAM Codes - ICIS PCS to ICIS-NPDES Parameter Crosswalk Point Source Category/SIC Crosswalk Point Source Category Codes SIC Codes SUPERCAS Category TWFs ICIS Facilities PCSFAC DMRFAC2007 DMR2007 DMR2007 Sensitivity Analysis Parameter Groupings Created or Imported Imported from PCS Imported from ICIS-NPDES Created Imported from TRICalculations2007 Imported from TRICalculations2007 Imported from TRICalculations2007 Imported from ICIS-NPDES Imported from TRICalculations2007 Imported from ICIS-NPDES Imported from PCS Created Created using queries Created using queries Imported from ICIS-NPDES Description Lists pollutants and corresponding parameter codes used for them in PCS. Lists pollutants and corresponding parameter codes used for them in ICIS-NPDES. Links PCS parameter codes to ICIS-NPDES parameter codes Links SIC codes with point source categories using a numeric code assigned in the Point Source Category Codes table. Assigns a numeric code to industrial categories using their 40 CFR Part number or 2-digit or 4-digit SIC code. Lists SIC codes and their descriptions. Links CAS numbers to pollutant parameter codes. Assigns TWF values to chemicals by CAS number. Presents information on permitted facilities, such as facility name, location and major/minor discharge status. Presents information on permitted facilities, such as facility name, location, major/minor discharge status, and date of most recent permit issuance. Combines PCSFAC and ICISFAC into one table. Presents the annual loads in pounds per year and TWPE for each pollutant discharge for each outfall at major permitted facilities. Presents the annual loads in pounds per year for each pollutant discharge for each outfall for the five annual loads calculated by PCSLoadCalculator2007 and three annual loads calculated by the ICIS-NPDES Load Calculator (see Section 3.2.4.2 and Table 3-16). Lists ICIS-NPDES pollutant parameter codes and their hierarchies for grouping parameters. This table is used to group parameters in both PCS and ICIS-NPDES. 3-43 ------- Section 3 - DMRLoads2007: Development and Category Rankings PCS Annual Loads NPDES Permit Number Major Indicator Discharge Pipe Number Monitoring Location Report Designator Parameter Code KGY01 KGYHO KGYH1 NOMAX KGYH1 NOCONC KGYH1 ICIS-NPDES Annual Loads NPDES Permit Number Major Indicator Discharge Pipe Number Monitoring Location Report Designator Parameter Code KGY01 KGYE1 KGY11 Step 1: Parameter Grouping Hierarchy PCS to ICIS-NPDES Parameter Crosswalk PCS Parameter Code ICIS-NPDES Parameter Code Parameter Groupings ICIS-NPDES Parameter Code Group Code (if grouped) Grouped Annual Loads Step 3: Combine Annual Loads Combine Loads LBY = KGYH1 x 2.205 DMR2007 SIC Code NPDES Permit Number Major Indicator Discharge Pipe Number Monitoring Location Report Designator Parameter Code Total Annual Pounds TWPE Major Indicator = M (Majors Only) Combined Facility Data (DMRFAC2007) NPDES Permit ID SIC Code NAME, CITY, ZIP Major Facility Indicator TWPE KGYH1 Step 2: Hybrid Methodology Apply Hybrid Methodology Step 4: Calculate TWPE SUPERCAS ICIS-NPDES Parameter/Group Code Parameter Name CAS Number Figure 3-6. DMRLoadsAnalysis2007 Inputs Used to Create DMR2007 Table 3-44 ------- Section 3 - DMRLoads2007: Development and Category Rankings 3.2.4.1 Creation of DMR2007 Annual Loads Table The following is a description of the steps EPA took to combine the annual loads from ICIS-NPDES Load Calculator and PCSLoadCalculator2007 into one table, "DMR2007". Step 1: Apply Parameter Grouping Hierarchy to PCSLoadCalculator2007 Annual Loads. The first step in applying the parameter grouping hierarchy to the annual loads from PCSLoadCalculator2007 is to convert the PCS parameter codes to ICIS-NPDES parameter codes because there are some parameters that have different parameter codes between the databases. EPA created the PCS to ICIS-NPDES Parameter Crosswalk table, which links the PCS parameter code to the ICIS-NPDES parameter code. EPA then updated the PCS parameter codes in the annual loads tables from PCSLoadCalculator2007 to the ICIS-NPDES parameter codes. As discussed in Section 3.2.3.2, aNPDES permit may require a facility to measure a pollutant in more than one way. The annual loads from PCSLoadCalculator2007 include one load for every parameter reported. To avoid double-counting pollutants, EPA applied the same hierarchy used in the ICIS-NPDES Load Calculator to group the pollutants in the PCSLoadCalculator2007 annual loading tables (see Table B-7 Appendix B). Step 2: Apply Hybrid Methodology to ICIS-NPDES Load Calculator Annual Loads. As discussed in Section 3.2.2.2, the output from the ICIS-NPDES Load Calculator includes two annual loads, KGYE1 and KGY01, calculated using BDL=l/2 DL and BDL=0, respectively. EPA applied the Hybrid Method that was used to calculate the PCSLoadCalculator2007 annual loads to the KGYE1 and KGY01 annual loads from the ICIS-NPDES Load Calculator. See Section 3.2.2.3 for more details on the hybrid methodology. Step 3: Combine Annual Loads into DMR2007. In 2007, annual loads for 72 facilities were in both PCSLoadCalculator2007 and the ICIS-NPDES Load Calculator. Because states are currently migrating from PCS to ICIS-NPDES, EPA chose to use the annual loads from ICIS- NPDES Load Calculator for the overlapping outfalls and pollutants16. Table B-8 in Appendix B presents the list of facilities, outfalls, and pollutants in both PCS and ICIS-NPDES in 2007. EPA used this methodology to combine the two sets of loads into the annual loads table, DMR2007. The DMR2007 table indicates the data source for the calculated annual load. Step 4: Calculate Toxic Weighted Pound Equivalent. To identify potential impacts on human health and the environment, EPA estimated toxic equivalent mass discharge through the use of TWFs. Section 5 of this report discusses TWFs in more detail. Chemicals for which EAD (Engineering and Analysis Division) has developed TWFs are identified by CAS number. To assign TWFs to reported discharges, EPA used the "SUPERCAS" table, developed in earlier work with PCS and TRI data, to link CAS numbers to pollutant parameters reported in PCS. EPA updated the SUPERCAS table to include ICIS-NPDES parameter codes. EPA has expanded the SUPERCAS list of chemicals by identifying CAS numbers for priority pollutants and chemicals that are frequently reported. EPA obtained the CAS numbers from 16Facilities may have some outfalls/pollutants that are in both PCS and ICIS-NPDES and other outfalls/pollutants in PCS or ICIS-NPDES only. For example, chlorine data for outfall 001 may be in PCS and ICIS-NPDES, but aluminum data for outfall 001 is only in PCS. In this example, EPA would use the chlorine load reported to ICIS- NPDES and the aluminum load reported to PCS. 3-45 ------- Section 3 - DMRLoads2007: Development and Category Rankings www.ChemFinder.com. During the 2009 annual review, as was done during previous annual reviews, EPA made the following assumptions to assign CAS numbers to pollutant parameters: • All forms of a pollutant were assigned the same CAS number (e.g., Dissolved Copper, Total Recoverable Copper, and Total Copper (as Cu) were all assigned the CAS number for Copper); and • Chemicals that were reported in different ways were assigned only one CAS number (e.g., Nitrate (as NOs) and Nitrate (as N) were both assigned the CAS number for Nitrate. EPA did not identify CAS numbers for chemicals infrequently reported. In addition, there are no CAS numbers for non-chemical parameters reported in ICIS-NPDES and PCS (e.g., total suspended solids, BOD5, COD, etc.). Once the CAS numbers were assigned to each parameter using the expanded SUPERCAS table, the TWFs were assigned by matching the CAS numbers. EPA did not assign TWFs to all parameters reported in ICIS-NPDES and PCS. For the 2009 annual review, EPA continued to estimate the TWFs for certain parameters that were reported as chemical groups based on transfers from existing TWFs, as was done during previous annual reviews. Table 3-15 lists these parameters and the method of TWF assignment. Table 3-15. TWF Assignment for Chemical Mixtures Parameter Code 78216 82699 30383 34034 74105 Parameter Description Aldrin + Dieldrin Endrin + Endrin Aldehyde (Sum) Benzene, Ethylbenzene, Toluene, and Xylene Chlorinated Phenols Phenols, Chlorinated Method of TWF assignment Average of aldrin and dieldrin TWFs Average of endrin and endrin aldehyde TWFs Average of benzene, ethylbenzene, toluene, and xylene TWFs Average of the TWFs for PCS parameters 2,4,6-trichloro- phenol, pentachlorophenol, 2,4-dichlorophenol, and 2- chlorophenol (most common chlorinated phenols) Average of the TWFs for PCS parameters 2,4,6-trichloro- phenol, pentachlorophenol, 2,4-dichlorophenol, and 2- chlorophenol (most common chlorinated phenols) 3.2.4.2 Sensitivity Analyses As described in Section 3.2.2.2, EPA developed queries in PCSLoadCalculator2007 and used annual loads output from ICIS-NPDES Load Calculator to calculate annual loads using the DL=0 alternative method. For this method, the Load Calculators assume a discharge of zero for pollutants that are labeled BDL. EPA combined the annual loads calculated using the DL=0 alternative method from PCSLoadCalculator2007 and ICIS-NPDES Load Calculator in DMRLoadsAnalysis200 7. During previous annual reviews EPA also calculated annual loads using the following alternative methods: 3-46 ------- Section 3 - DMRLoads2007: Development and Category Rankings • EST=NO. PCSLoadCalculator2007 assumes a discharge of zero for monitoring periods where discharge data are missing. • No Maximum (NOMAX). PCSLoadCalculator2007 used an alternative measurement selection hierarchy, which set maximum concentrations (MCMX) and maximum quantities (MQMX) to zero during the measurement value selection process. • No Concentration (NOCONC). PCSLoadCalculator2007 used an alternative measurement selection hierarchy, which set average concentrations (MCAV), minimum concentrations (MCMN), and maximum concentrations (MCMX) to zero during the measurement value selection process. EPA did not calculate the annual loads using these alternative methods as part of the 2009 annual review based on the findings of the previous reviews. Section 6.1.2.3 of the Technical Support Document for the Preliminary 2008 Effluent Guidelines Program Plan describes the results of the 2007 annual review sensitivity analyses for EST=NO, NOMAX, and NOCONC (U.S. EPA, 2007). Table 3-16 compares the assumptions and calculation options that the ICIS-NPDES Load Calculator and PCSLoadCalculator2007 used to calculate each set of annual loads, including the alternative loads that were not calculated as part of the 2009 annual review. Table 3-16. Comparison of Alternative Load Calculation Methods Annual Load Set EST Option DL Option Measurements Included in Selection Hierarchy a Standard Load Calculation DMR 2007 (PCS and ICIS-NPDES) EST=YES Hybrid (DL=0 or DL=l/2) MQAV MQMX MCMN MCAV MCMX Alternative Load Calculations DL=0 (PCS and ICIS-NPDES) EST=NO b NOMAX b NOCONC b EST=YES EST=NO EST=YES EST=YES DL=0 Hybrid Hybrid Hybrid MQAV MQMX MCMN MCAV MCMX MQAV MQMX MCMN MCAV MCMX MQAV MQMX=0 MCMN MCAV MCMX=0 MQAV MQMX MCMN=0 MCAV=0 MCMX=0 aFor the standard load calculation and DL=0 alternative load calculation in ICIS-NPDES, the measurements included in the selection hierarchy are the five ICIS-NPDES measurements (MQ1, MQ2, MCI, MC2, and MC3). bEPA did not calculate these annual loads as part of the 2009 annual review. Section 6.1.2.3 of the Technical Support Document for the Preliminary 2008 Effluent Guidelines Program Plan describes the results of these alternative load calculation methods (U.S. EPA, 2007). 3-47 ------- Section 3 - DMRLoads2007: Development and Category Rankings EPA examined the impact of each calculation method, shown in Table 3-16, on the calculated pollutant loads in a series of sensitivity analyses. To conduct each sensitivity analysis, EPA calculated TWPE for loads calculated with each alternative method, and compared TWPE calculated using the standard and alternative load calculation methods. EPA made this comparison for total discharge and for the discharges separated into categories. EPA then identified categories and individual facilities within a category that show a large difference between DMR 2007 TWPE and alternative TWPE using the calculations shown below: Amount of TWPE Based on Calculation Alternative (Ib-eq/yr) = Standard Load TWPE (Ib-eq/yr) - Alternative Load TWPE (Ib-eq/yr) Percent of TWPE Based on Calculation Alternative = Amount of TWPE Based on Calculation Alternative (Ib-eq/yr) / Standard Load TWPE (Ib-eq/yr) The following sections discuss the results of the DL sensitivity analyses based on combined 2007 PCS and ICIS-NPDES data. DL Sensitivity Analysis The purpose of the DL sensitivity analysis is to evaluate the impact of EPA's use of the Hybrid Method, which estimates loads for some pollutants reported to PCS and ICIS-NPDES as BDL, on the screening-level analysis. Table 3-17 presents a summary of the results of the DL analysis for the point source categories showing the highest sensitivity to the DL options and the total forDMRLoads2007. As shown in Table 3-17, only 0.12 percent (1,110,000 Ib-eq) of the TWPE in DMRLoads2007 are based on BDL assumptions using the Hybrid Method. The categories showing the greatest sensitivity to the DL options include Superfund Sites, the Pulp, Paper and Paperboard Category, and the Petroleum Refining Category. For the complete results of the DL sensitivity analysis see Table B-9 in Appendix B. Table B-10 in Appendix B presents the results of the sensitivity analysis by pollutant. Pollutant parameters showing the highest sensitivity to the DL options include 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD), TCDD equivalents, and chlorine. 3-48 ------- Section 3 - DMRLoads2007: Development and Category Rankings Table 3-17. Results of DL Sensitivity Analysis Point Source Category Sanitary Services (SIC 4959) Pulp, Paper, and Paperboard (40 CFR Part 430) Trucking and Warehousing (SIC Group 42) Pesticide Chemicals (40 CFR Part 455) Airport Deicing (PNC) Petroleum Refining (40 CFR Part 4 19) Nonferrous Metals Manufacturing (40 CFR Part 421) Organic Chemicals, Plastics, and Synthetic Fibers (40 CFR Part 4 14) Non-classifiable Establishments (SIC Group 99) Independent and Stand Alone Labs (PNC) Total DMRLoads2007 Total Number of Facilities 2 217 2 147 5 108 36 219 10 6 2,177 Number of Facilities Affected by DL 1 78 1 28 2 65 14 89 2 4 1,025 Total Annual Load, Ib/yr 653,000 2,450,000,000 83,300 3,840,000,000 1,160,000 1,950,000,000 188,000,000 1,480,000,000 24,800,000 465,000 43,100,000,000 Total Annual Load Based on DL, Ib/yr 9,450 (1.5%) 70,800,000 (2.9%) 613 (0.7% 109,000 (0.002%) 30,000 (2.6%) 153,000,000 (7.8%) 30,700,000 (16%) 114,000,000(7.7%) 1,560 (0.01%) 10,100 (2.2%) 3,830,000,000 Total TWPE, Ib- eq/yr 2.69 2,730,000 57.6 180,000 265 403,000 343,000 413,000 2,070 5,360 942,000,000 Total TWPE Based on DL, Ib- eq/yr 1.35 (50%) 347,000 (13%) 5.24(9.1%) 8,980 (5%) 9.02 (3.4%) 13,000 (3.2%) 10,200 (3%) 6,310(1.5%) 19 (0.9%) 27.7 (0.5%) 1,110,000 Source: DMRLoads2007_v3. PNC - Potential new category. 3-49 ------- Section 3 - DMRLoads2007: Development and Category Rankings 3.2.5 DMRNutrients2007 DMRNutrients2007 uses the annual loads for nitrogen and phosphorus compounds from the DMR2007 table to calculate aggregate nitrogen as N and phosphorus as P loads for each facility outfall. The database sums the aggregate nitrogen and phosphorus loads by facility and by point source category. Table B-l 1 of Appendix B presents the category rankings for total nitrogen as N loads and Table B-l2 presents the category rankings for total phosphorus as P loads. DMR data include discharges of nitrogen and phosphorus in various chemical forms. For example, nitrogen may be reported in its elemental form (as N), total Kjeldahl nitrogen (TKN), organic nitrogen, ammonia as N, ammonia as NH3 or NH4, un-ionized ammonia, nitrite, or nitrate. EPA developed a series of hierarchies to select the appropriate combination of nitrogen and phosphorus compounds to calculate the total nitrogen and total phosphorus loads. These hierarchies, summarized below, are described in detail in "Point Source Category Rankings by Nitrogen and Phosphorus Loads Calculated Using 2002 PCS Data" (Kandle, 2005a). use): Total Nitrogen Load EPA calculated total nitrogen using one of the following equations (presented in order of • Total Nitrogen Load = Total Nitrogen as N; • Total Nitrogen Load = TKN + Nitrite (NO2) + Nitrate (NO3); or • Total Nitrogen Load = Organic Nitrogen + Ammonia + Nitrite + Nitrate. Nitrogen compounds that are reported as NH3, NH4, NC>2, or NO3 were converted to N based on molecular weight, then summed to calculate Total Nitrogen Load. Table 3-18 presents the conversion factors EPA used for nitrogen compounds. Table 3-18. Conversion Factors for Nitrogen Compounds Nitrogen Compound Ammonia as NH3 Nitrite as NO2 Nitrate as NO3 Conversion Factor 14N/17NH3 14 N / 46 NO2 14 N / 62 NO3 Total Phosphorus Load Loads for phosphorus parameters were grouped by EPA's grouping hierarchy described in Section 3.2.2 and assigned to a grouped parameter code. As a result, DMRNutrients2007 includes only two parameters for phosphorus compounds. EPA used the following hierarchy to calculate total phosphorus load: • If loads of phosphorus (PRAM PHOSP) were available, EPA used the PRAM PHOSP load to represent total phosphorus. EPA assumed that the majority of the loads were reported as phosphorous and did not apply a conversion factor to calculate pounds of phorphosour. 3^50 ------- Section 3 - DMRLoads2007: Development and Category Rankings • If loads of phosphorus (PRAM PHOSP) were not available, EPA used loads of phosphate (PRAMs PO4). EPA multiplied the load by 31/95 to convert the reported phosphate load to pounds of phosphorous. 3.2.6 DMRLoads2007 As the final step in developing DMRLoads2007, EPA grouped discharges from DMRLoadAnalysis2007 to create the point source category rankings and to perform other analyses. Section 3.2.6.1 discusses the tables and table structure ofDMRLoads2007, and Section 3.2.6.2 discusses the SIC/Point Source Category Crosswalk in relation to DMRLoads2007 tables. 3.2.6.1 DMRLoads200 7 Structure Table 3-19 lists and describes the tables in DMRLoads2007'. Table 3-19. Tables Imported or Created in DMRLoads2007 Table Name PRAM Codes SIC/Point Source Category Crosswalk Point Source Category Codes SIC Codes SUPERCAS Category TWFs DMRFAC DMR2007 Manual ICIS Loads Corrections Category Rankings - Nitrogen Category Rankings - Phosphorus SIC Code Rankings Category Rankings - Toxic Weight Created or Imported Linked from DMRLoadsAnalysis2007 Linked from DMRLoadsAnalysis2007 Linked from DMRLoadsAnalysis2007 Linked from DMRLoadsAnalysis2007 Linked from DMRLoadsAnalysis2007 Linked from TRICalculations2007 Linked from DMRLoadsAnalysis2007 Linked from DMRLoadsAnalysis2007 Created Linked from DMRNutrietns2007 Linked from DMRNutrients2007 Created using queries Created using queries Description Lists pollutants and corresponding parameter codes. Links SIC codes with point source categories using a numeric code assigned in the Point Source Category Codes table. Assigns a numeric code to industrial categories using their 40 CFR part number or 2-digit or 4-digit SIC Code. Lists SIC codes and their descriptions. Links CAS numbers to pollutant parameter codes. Assigns TWF values to chemicals by CAS number. Presents information on permitted facilities, such as facility name, location, major/minor discharge status, and date of most recent permit issuance Presents the annual loads in pounds per year and TWPE for each pollutant discharge for each outfall at major permitted facilities. Lists ICIS-NPDES loads corrections identified by manual review. Presents rankings of categories based on aggregated nitrogen load. Presents rankings of categories based on aggregated phosphorus load. Presents rankings of SIC codes based on calculated TWPE. Presents rankings of categories based on calculated TWPE. 3-51 ------- Section 3 - DMRLoads2007: Development and Category Rankings 3.2.6.2 SIC/Point Source Category Crosswalk DMRLoads2007 assigns a facility's discharge to an industrial category using 4-digit SIC codes. Point source categories are not generally defined by SIC codes. As a result, EPA developed a crosswalk that links point source categories to 4-digit SIC codes, described in Section 1 of this document. EPA has developed ELGs for point source discharges from 56 specific categories. The point source categories, which may be divided into subcategories, are generally defined in terms of combinations of products made and the processes used to make these products. Facilities with data in PCS and ICIS-NPDES are identified by SIC code. Thus, to use the PCS and ICIS-NPDES data to estimate the pollutants discharged by each point source category, EPA assigned each 4-digit SIC code to an appropriate point source category using the "SIC/Point Source Category Crosswalk" table. See Section 4.2 for additional information on the development of the SIC/Point Source Category Crosswalk. As shown in Figure 3-7, DMRLoads2007 links information from the DMR2007 Table, DMRFAC, and the SIC/Point Source Category Crosswalk to create point source category rankings. The SIC codes in the DMR2007 Table are specific to each parameter, discharge pipe (outfall), and facility (NPDES permit number). This allows EPA to make SIC adjustments to differentiate between various operations/outfalls at one facility and assign discharges at the pollutant level to different point source categories, as described in Section 4.2.1.2. Sum Annual Loads and TWPE by Point Source Category Category Rankings - Toxic Weight Point Source Category Code Category Name Type of Group Majors Total Annual Pounds Figure 3-7. DMRLoads2007 Database Structure 3.2.7 Database Corrections EPA reviewed DMRLoads2007 output for reasonableness, as described in Section 3.4. Also, during previous screening-level analyses, EPA identified facility-specific corrections for PCS data. Several of these corrections similarly apply to the 2007 DMR data. In addition, EPA's quality review (see Section 3.4) identified other corrections to the 2007 DMR data, (e.g., units incorrectly reported as gallons per day were corrected to MOD). Table B-13 in Appendix B lists all corrections made to the 2007 DMR data. In addition to the facility-specific data corrections, DMRLoadsAnalysis2007 performs the following modifications to the annual loads: 3-52 ------- Section 3 - DMRLoads2007: Development and Category Rankings Categorization of Discharges. Section 1 of this report describes the development of the SIC/Point Source Category Crosswalk, which EPA uses to link between facility SIC codes and categories with existing ELGs. Because most point source categories are not defined by SIC code, the relationship between SIC code and point source category is not a one-to-one correlation. A single SIC code may include facilities in more than one point source category, and associating an SIC code with only one category may be an over simplification. Also, many facilities have operations subject to more than one point source category. Further, facilities in some categories cannot be identified by SIC code (e.g., Centralized Waste Treatment facilities). The database changes are summarized below: — Facility-Level Point Source Category Assignment. For some SIC codes that include facilities subject to guidelines from more than one point source category, EPA was able to assign each facility to the category that best applied to the majority of its discharges. EPA reviewed information available about each facility to determine which point source category applied to the facility's operations. — Pollutant-Level Point Source Category Assignment. Many facilities have operations subject to more than one point source category. For most of these facilities, EPA cannot divide the pollutant discharges among the applicable point source categories. Two exceptions where EPA was able to assign wastewater discharges of certain chemicals to the appropriate point source category include Organic Chemicals, Plastics, and Synthetic Fibers (OCPSF) /Pesticides and MP&M/Metal Finishing (see Section 4.2.1.2 for additional discussion). Table 3-20 shows the facilities for which EPA revised SIC codes to link to an appropriate point source category based on known plant operations. Internal Monitoring. As described in Sections 3.2.2.2 and 3.2.3.2, PCSLoadCalculator2007 and the ICIS-NPDES Load Calculator calculated loads only for monitoring locations that are labeled as effluent (MLOC 1 or 2 in PCS and MLOC 1, 2, A, B, or SC in ICIS-NPDES). For the 2009 annual review, EPA included only MLOC 1 and 2 for ICIS-NPDES. EPA will include MLOC A, B, and SC in future annual reviews. As a result, the Load Calculators exclude discharges for internal monitoring locations such as intake water, influent to treatment, and intermediate points in the wastewater treatment system. However, during previous category reviews and detailed studies, EPA identified instances of double counting that resulted from including certain internal monitoring points in the loads database. For example, a facility monitors for Pollutant A at the effluent from its wastewater treatment system (Internal Outfall 101). Outfall 101 wastewater is later combined with other plant discharges at final Outfall 001 and is discharged to a receiving stream. The facility also monitors for Pollutant A at Final Outfall 001. Both outfalls are effluent monitoring points identified as MLOC 1 or MLOC 2; however, Outfall 101 is upstream of the final outfall. Calculating loads for Pollutant A at both the internal and final outfalls results in double counting Pollutant A discharges. EPA identified instances where pollutant discharges are reported for multiple monitoring locations along the same 3^53 ------- Section 3 - DMRLoads2007: Development and Category Rankings discharge line, and eliminated the discharges for the upstream monitoring locations. EPA made these corrections in PCSLoadCalculator2007 for the PCS data and in DMRLoadsAnalysis2007 for the ICIS-NPDES data (see Table B-13 in Appendix B). • Intermittent Discharges. As described in Sections 3.2.2.2 and 3.2.3.2, in PCSLoadCalculator2007 and the ICIS-NPDES Load Calculator EPA assumes that all discharges in PCS and ICIS-NPDES are continuous. During previous Annual Reviews, EPA identified facility discharges that are intermittent and therefore overestimated by the Load Calculator. EPA calculated annual loads for these discharges based on information obtained from the facility on the frequency and duration of wastewater discharges. EPA made these corrections in PCSLoadCalculator2007 for the PCS data and in DMRLoadsAnalysis2007 for the ICIS-NPDES data (see Table B-13 in Appendix B). • Pollutant Parameters Excluded from DMRLoads2007. Parameters in PCS and ICIS-NPDES include water quality parameters (e.g., dissolved oxygen and temperature), specific chemicals (e.g., phenol), bulk parameters (e.g., biochemical oxygen demand), and flow. As described in Sections 3.2.2.2 and 3.2.3.2, facilities report pollutant mass quantities, pollutant concentrations, and wastewater flow rates to PCS and ICIS-NPDES using a variety of units. EPA's PCS CNVRT program and the ICIS-NPDES convert module convert the discharges into standard units of kilograms per day for mass quantities, milligrams per liter for concentrations, and millions of gallons per day for flow rates. However, some parameters are reported in units that cannot be converted into kg/day or mg/L (e.g. temperature and pH). EPA excluded these parameters from the screening- level analysis. Table B-14 of Appendix B lists the excluded parameters. • ICIS-NPDES Load Corrections. In some cases, EPA identified that loads are not estimated correctly because of errors in units, number of reporting periods, detection limit estimation, improperly-coded outfalls, or other data entry errors. For data in PCS, EPA made corrections in the PCSLoadCalculator database. For data from ICIS-NPDES, EPA corrected loads in DMRLoadsAnalysis2007'. Table B-13 in Appendix B lists these case-by-case ICIS Loads corrections. 3-54 ------- Section 3 - DMRLoads2007: Development and Category Rankings Table 3-20. Case-by-Case Point Source Category Reassignments in DMRLoads2007 NPID TN0002968 OH0048836 MD0000060 MO0002356 MS0046931 MS0002941 AL0000213 AL0001945 AL0003514 DE0050911 KY0003484 KYO 102083 LA0005231 LA0005983 LA0029769 ME0000639 NV0020923 NY0001635 NY0003328 NY0003336 OHO 11 5401 TN0002461 TX0007412 TX0008150 WI0003565 WV0000108 WV0004359 Name US DOE-Oak Ridge Y12 Plant Duke Energy, Ohio, Inc. Perdue Farms, Inc. BCP Ingredients, Inc Scott County Lawrence County Occidental Chemical Corp Olin Chlor Alkali Products Occidental Chemical Corp Diamond Shamrock Chemicals Co. Westlake Ca&O Corp USEC PDGDP Pioneer Chlor Alkali Co., Inc Occidental Chemical Corp IMC-Phosphates Company Holtachem Mfg Pioneer Americas-BMI Complex Olin Corp - Niagara Falls Pit E I Dupont De Nemours & Co, Inc Occidental Chemical Corp Us Enrichment Corp Ports Gaseo Olin Chemicals Corp. Deer Park Plant Oxy Vinyls, Lp, Harris County Erco Worldwide Usa Inc Pt Edw Kincaid Enterprises Ppg Industries, Inc. Old SIC Code 9611 4932 2048 2048 2048 2861 2813 2812 2812 2812 2812 2819 2812 2812 2819 2812 2812 2812 2812 2812 2819 2812 2812 2812 2812 2819 2812 Old Point Source Category Administration Of Economic Programs Electric, Gas, & Sanitary Services Food & Kindred Products Food & Kindred Products Food & Kindred Products Gum And Wood Chemicals Manufacturing Inorganic Chemicals Manufacturing Inorganic Chemicals Manufacturing Inorganic Chemicals Manufacturing Inorganic Chemicals Manufacturing Inorganic Chemicals Manufacturing Inorganic Chemicals Manufacturing Inorganic Chemicals Manufacturing Inorganic Chemicals Manufacturing Inorganic Chemicals Manufacturing Inorganic Chemicals Manufacturing Inorganic Chemicals Manufacturing Inorganic Chemicals Manufacturing Inorganic Chemicals Manufacturing Inorganic Chemicals Manufacturing Inorganic Chemicals Manufacturing Inorganic Chemicals Manufacturing Inorganic Chemicals Manufacturing Inorganic Chemicals Manufacturing Inorganic Chemicals Manufacturing Inorganic Chemicals Manufacturing Inorganic Chemicals Manufacturing New SIC Code a 3499 4911 2048GRAIN 2048GRAIN 2048MPP 2621-2 VCCA VCCA VCCA VCCA VCCA 2819NMM VCCA VCCA 2873 9999 VCCA VCCA VCCA VCCA 2819NMM VCCA VCCA VCCA VCCA 2879 VCCA New Point Source Category Metal Finishing Steam Electric Power Generating Grain mills Grain mills Meat and Poultry Products Pulp, Paper And Paperboard Chlorine And Chlorinated Hydrocarbons Chlorine And Chlorinated Hydrocarbons Chlorine And Chlorinated Hydrocarbons Chlorine And Chlorinated Hydrocarbons Chlorine And Chlorinated Hydrocarbons Nonferrous Metals Manufacturing Chlorine And Chlorinated Hydrocarbons Chlorine And Chlorinated Hydrocarbons Fertilizer Manufacturing Non Classifiable Establishments Chlorine And Chlorinated Hydrocarbons Chlorine And Chlorinated Hydrocarbons Chlorine And Chlorinated Hydrocarbons Chlorine And Chlorinated Hydrocarbons Nonferrous Metals Manufacturing Chlorine And Chlorinated Hydrocarbons Chlorine And Chlorinated Hydrocarbons Chlorine And Chlorinated Hydrocarbons Chlorine And Chlorinated Hydrocarbons Pesticide Chemicals Chlorine And Chlorinated Hydrocarbons 3-55 ------- Section 3 - DMRLoads2007: Development and Category Rankings Table 3-20. Case-by-Case Point Source Category Reassignments in DMRLoads2007 NPID AR0037800 LA0038245 LA0065501 MOO 108472 TN0074225 TX0005941 TX0030937 TX0091855 ME0001872 IL0001724 TN0002488 TN0003671 AL0002658 AL0026832 CO0042480 MI0004227 NY0006548 OH0004219 OH0004260 OH0098540 AL0054704 DE0000612 IL0001350 IN0002101 Name Clean Harbors El Dorado, LLC Clean Harbors Baton Rouge, LLC Clean Harbors White Castle LLC Front St Remedial Action Ettp-Central Neutraliz. Fac Clean Harbors Deer Park WWTP Vopak Logistics Services USA Stolthaven Houston, Inc. Domtar Maine Corporation American Nickeloid Co-Peru State Ind- Ashland Cty Usa Holston Army Ammo Pit Area Anniston Army Depot Golden Rod Broilers Eagle Mine Remediation WWTP Dsc Ltd Owl Wire & Cable Inc - Rome Fac Timken Company - Canton AK Steel Coshocton Stainless Reserve Environmental Services Sabic Innovative Plactics Formosa Plastics Corporation Formosa Plastics-Illinois Sabic Innovative Plastics Mt Ve Old SIC Code 4953 4953 4953 4953 4953 4953 4953 4953 2411 3471 3639 9711 9999 9999 9999 9999 9999 9999 9999 9999 2821 2821 2821 2821 Old Point Source Category Landfills Landfills Landfills Landfills Landfills Landfills Landfills Landfills Lumber & Wood Products Metal Finishing Metal Finishing National Security & International Affairs Non Classifiable Establishments Non Classifiable Establishments Non Classifiable Establishments Non Classifiable Establishments Non Classifiable Establishments Non Classifiable Establishments Non Classifiable Establishments Non Classifiable Establishments Organic Chemicals, Plastics And Synthetic Fibers Organic Chemicals, Plastics And Synthetic Fibers Organic Chemicals, Plastics And Synthetic Fibers Organic Chemicals, Plastics And Synthetic Fibers New SIC Code a 4953WC CWT CWT SUPER 4953WC 4953WC CWT CWT 2411-1 3471CC 3639PE 2892 3795 2015 9999 3316 3351 3562 3312 CWT VCCA VCCA VCCA VCCA New Point Source Category Waste Combustors Centralized Waste Treatment Centralized Waste Treatment Superfund Sites Waste Combustors Waste Combustors Centralized Waste Treatment Centralized Waste Treatment Pulp, Paper And Paperboard Coil Coating Porcelain Enameling Explosives Manufacturing Metal Finishing Meat and Poultry Products Non Classifiable Establishments Iron And Steel Manufacturing Copper forming Metal Finishing Iron And Steel Manufacturing Centralized Waste Treatment Chlorine And Chlorinated Hydrocarbons Chlorine And Chlorinated Hydrocarbons Chlorine And Chlorinated Hydrocarbons Chlorine And Chlorinated Hydrocarbons 3-56 ------- Section 3 - DMRLoads2007: Development and Category Rankings Table 3-20. Case-by-Case Point Source Category Reassignments in DMRLoads2007 NPID LA0000761 LA0002933 LA0003301 LA0006149 LA0006220 LA0007129 LA0041025 LA0056171 NJ0004286 NJ0004391 OH0007269 TX0002798 TX0006335 TX0006483 TX0070416 TX0085570 Name PPG - Lake Charles Occidental Chemical Corp. Dow Chemical - Plaquemine Formosa Plastics Corp Crompton Manufacturing Co. Georgia Gulf Corporation Certainteed Corporation Occidental Chemical Corporatio Polyone Corporation Colorite Polymers Company Dover Chemical Subsidiary Of I Wwtp Oxy Vinyls, Lp Dow Chemical Vinyl Chloride Monomer Plant Formosa Point Comfort Plant Old SIC Code 2869 2869 2869 2869 2869 2869 2821 2869 2821 2821 2869 2869 2821 2869 2869 2869 Old Point Source Category Organic Chemicals, Plastics And Synthetic Fibers Organic Chemicals, Plastics And Synthetic Fibers Organic Chemicals, Plastics And Synthetic Fibers Organic Chemicals, Plastics And Synthetic Fibers Organic Chemicals, Plastics And Synthetic Fibers Organic Chemicals, Plastics And Synthetic Fibers Organic Chemicals, Plastics And Synthetic Fibers Organic Chemicals, Plastics And Synthetic Fibers Organic Chemicals, Plastics And Synthetic Fibers Organic Chemicals, Plastics And Synthetic Fibers Organic Chemicals, Plastics And Synthetic Fibers Organic Chemicals, Plastics And Synthetic Fibers Organic Chemicals, Plastics And Synthetic Fibers Organic Chemicals, Plastics And Synthetic Fibers Organic Chemicals, Plastics And Synthetic Fibers Organic Chemicals, Plastics And Synthetic Fibers New SIC Code a VCCA VCCA VCCA VCCA VCCA VCCA VCCA VCCA VCCA VCCA VCCA VCCA VCCA VCCA VCCA VCCA New Point Source Category Chlorine And Chlorinated Hydrocarbons Chlorine And Chlorinated Hydrocarbons Chlorine And Chlorinated Hydrocarbons Chlorine And Chlorinated Hydrocarbons Chlorine And Chlorinated Hydrocarbons Chlorine And Chlorinated Hydrocarbons Chlorine And Chlorinated Hydrocarbons Chlorine And Chlorinated Hydrocarbons Chlorine And Chlorinated Hydrocarbons Chlorine And Chlorinated Hydrocarbons Chlorine And Chlorinated Hydrocarbons Chlorine And Chlorinated Hydrocarbons Chlorine And Chlorinated Hydrocarbons Chlorine And Chlorinated Hydrocarbons Chlorine And Chlorinated Hydrocarbons Chlorine And Chlorinated Hydrocarbons 3-57 ------- Section 3 - DMRLoads2007: Development and Category Rankings Table 3-20. Case-by-Case Point Source Category Reassignments in DMRLoads2007 NPID TXO 104876 CT0003212 OK0040827 MS0001309 TX0062677 LA0004847 MS0003115 GA0046973 OK0034321 WI0001848 OR0020834 CO0042064 LA0038245 LA0065501 LA0066214 MOO 108472 NJ0005240 TX0030937 TX0091855 MO0001716 Name Organic Chemical Manufacturing Kimberly-Clark Corporation Kimberly-Clark Corp-Jenks Fac Adams County North Regional Treatment Plant Mosaic Fertilizer, LLC Jackson County Fort James Operating Company Fort James Oprating Co-Muskoge Georgia Pacific Consumer Prod St. Helens STP/Boise Cascade Treatment, Storage & Disposal Clean Harbors Baton Rouge, LLC Clean Harbors White Castle LLC NPC Services-Alsen Front St Remedial Action Safety-Kleen - Bridgeport Vopak Logistics Services USA Stolthaven Houston, Inc. Basf Hannibal Plant Old SIC Code 2869 2676 2676 2911 2911 2874 2874 2621 2621 2621 4952 4953 4953 4953 4953 4953 4953 4953 4953 5191 Old Point Source Category Organic Chemicals, Plastics And Synthetic Fibers Paper & Allied Products Paper & Allied Products Petroleum Refining Petroleum Refining Phosphate Manufacturing Phosphate Manufacturing Pulp, Paper And Paperboard Pulp, Paper And Paperboard Pulp, Paper And Paperboard Sewerage Systems Waste Combustors Waste Combustors Waste Combustors Waste Combustors Waste Combustors Waste Combustors Waste Combustors Waste Combustors Wholesale Trade- Nondurable Goods New SIC Code a VCCA 2621-2 2611-2 2611-2 CWT 2874FER 2874FER VCCA VCCA VCCA 2621-1 4953L CWT CWT 4953L SUPER 4953L CWT CWT 2879 New Point Source Category Chlorine And Chlorinated Hydrocarbons Pulp, Paper And Paperboard Pulp, Paper And Paperboard Pulp, Paper And Paperboard Centralized Waste Treatment Fertilizer Manufacturing Fertilizer Manufacturing Chlorine And Chlorinated Hydrocarbons Chlorine And Chlorinated Hydrocarbons Chlorine And Chlorinated Hydrocarbons Pulp, Paper And Paperboard Landfills Centralized Waste Treatment Centralized Waste Treatment Landfills Superfund Sites Landfills Centralized Waste Treatment Centralized Waste Treatment Pesticide Chemicals Source: DMRLoads2007_v3. a Because some point source categories correspond to multiple SIC codes, some changes to SIC codes did not result in a change in point source category assignment. These SIC changes are not shown in the table. 3-58 ------- Section 3 - DMRLoads2007: Development and Category Rankings 3.3 Results of the Preliminary Analysis This section presents the results oftheDMRLoads2007 database. Table 3-22 presents the categories ranked from highest to lowest TWPE. Table B-l of Appendix B presents the four- digit SIC code rankings by TWPE. Table B-2 of Appendix B presents the total TWPE for pollutant parameters reported in DMR. 3.4 Data Quality Review EPA evaluated the quality of the PCS and ICIS-NPDES DMR data for use in the 2009 screening-level review. This evaluation considered data completeness, accuracy, reasonableness, and comparability. The Quality Assurance Project Plan for the 2009 Annual Screening-Level Analysis of TRI and PCS Industrial Category Discharge Data (ERG, 2009) describes the quality objectives in more detail. EPA conducted quality reviews for four stages of the development of DMRLoads2007: PCS CNVRT program output; ICIS-NPDES Convert Module output; PCSLoadCalculator2007 and the ICIS-NPDES Pollutant Loading Tool output; and DMRLoads2007 results. The following discussion provides an overview of the quality review steps for each stage: • PCS CNVRT program output. EPA's quality review of the CNVRT output files included reasonableness checks of pollutant quantities and concentrations. EPA reviewed the CNVRT program output (i.e., the pollutant discharges stored in PCS converted into standard units of kg/day and mg/L) to identify possible errors in recording units of measure. EPA reviewed ranges of pollutant quantities and concentrations and identified pollutant measurements and flows that were unreasonably high. EPA then compared these measurements with measurements available on EPA's Envirofacts web page. If the measurements were similar EPA concluded that the CNVRT file output was acceptable. This review resulted in two types of systematic corrections to the CNVRT output: — Corrections to 1,015 flows ranging from 1,300 MOD to 5,000 MOD17 (see Section 3.2.2.2); and — Corrections to 284 mercury concentrations reported to PCS using PRAM 50092 (Mercury Total Low Level) (see Section 3.4.1). • ICIS-NPDES Convert Module output. EPA conducted an initial quality review of the extracted ICIS-NPDES DMR data to evaluate its completeness, reasonableness, and comparability. For completeness, EPA compared the number of major facilities and the universe of SIC codes in the 2007 ICIS-NPDES DMR data to the PCS DMR data in 2004, the last complete DMR data set for ICIS- NPDES states. The 2007 ICIS-NPDES data had at least as many majors and SIC codes as PCS in 2004. Additionally, EPA verified that, while PCS 2004 had more parameter codes than ICIS-NPDES in 2007, all commonly reported parameters are present in the 2007 ICIS-NPDES DMR data. 17 In addition to these systematic flow corrections, EPA determined that all flows between 1,000 and 5,000 MOD reported by facilities in Ohio were flows in GPD. EPA automatically divided these flows by 1,000,000. However, because power plants are known to have high flows, EPA made flow corrections to Ohio facilities reporting SIC code 4911 (Electrical Services) on a case-by-case basis. 3-59 ------- Section 3 - DMRLoads2007: Development and Category Rankings EPA reviewed the DMR data for reasonableness to identify any data quality issues, such as misreported units that the ICIS-NPDES Convert Module did not correct. EPA identified several wastewater flows that exceeded the reasonable range. EPA reviewed these flows and developed the proposed flow correction function for the ICIS-NPDES Convert Module (described in Section 3.2.3.1). This function is designed to identify data entry errors for flows greater than 1,000 MOD. The ICIS-NPDES Convert Module corrects all flows exceeding 5,000 MOD, and applies more conservative criteria to correct flows from 1,000 to 5,000 MOD. The ICIS-NPDES Convert Module made following corrections to ICIS- NPDES wastewater flows: — 1,113 corrections based on month-to-month variations; — 1,605 corrections based on comparing flows to design flows; and — 142 corrections based on assuming that flows exceeding 5,000 MGD were reported in units of GPD. EPA also evaluated the comparability of the extracted 2007 ICIS-NPDES DMR data to the 2004 PCS data. As shown in Table 3-21, most of the average loads and concentrations in ICIS-NPDES are within one order of magnitude of the 2004 PCS data. However, the maximum loads and concentrations indicate that there may be some unreasonable values in the 2007 ICIS-NPDES DMR data. EPA verified the unit conversions used in the ICIS-NPDES Convert Module and for this reason concluded that the unreasonable flows and pollutant measurements are likely the result of data entry errors and are not the result of any errors in the ICIS-NPDES Convert Module functions. Table 3-21. Comparison of Load and Concentration Ranges for Common Parameters Pollutant Aluminum Ammonia BOD Chlorine Copper Iron Nitrogen Oil and Grease Phosphate Phosphorus TKN TSS Zinc Average 2004 PCS Quantity (kg/day) 77 154 1,296 0.90 0.42 241 53 41 43 583 65 457 1.89 Maximum 2004 PCS Quantity (kg/day) 4,755 1,873 275,456 692 244 51,812 19,985 5,953 5,953 131,464 5,117 131,429 1,046 Average 2004 PCS Cone. (mg/L) 19 2.8 1,773 0.5 2.7 2.7 6.9 3.2 3.3 20 15 838 1.3 Maximum 2004 PCS Cone. (mg/L) 3,333 116 6,690,000 1,420 1,042 19,450 2,701 9,400 9,400 8,104 3,400 73,500 1,360 Average 2007 ICIS- NPDES Quantity (kg/day) 15 343 802 1.5 0.20 165 818 195 6.5 39 298 574 0.67 Maximum 2007 ICIS- NPDES Quantity (kg/day) 72 7,082 3,490,205 1,647 123 46,530 43,584 10,651 30 152,101 9,403 3,405,402 1,308 Average 2007 ICIS- NPDES Cone. (mg/L) 10 8.8 134 0.46 0.11 3.6 20 5.4 4.4 2.1 2,348 93 0.46 Maximum 2007 ICIS- NPDES Cone. (mg/L) 5,620 147 290,171 60,000 850 2,800 114,598 380 93 20,990 240,000 380,800 769 Sources: PCSLoadCalculator2004 and the ICIS-NPDES Pollutant Loading Tool. 3^60 ------- Section 3 - DMRLoads2007: Development and Category Rankings Table 3-22. DMR 2007 Point Source Category Rankings by TWPE 40CFR Part NA 423 433 430 414.1 a 418 420 432 414 419 415 421 440 455 471 NA 410 429 417 NA 444 445 409 436 439 463 422 467 464 428 454 437 469 411 NA NA NA Point Source Category Superfund Sites Steam Electric Power Generating Metal Finishing Pulp, Paper And Paperboard Chlorine And Chlorinated Hydrocarbons Fertilizer Manufacturing Iron And Steel Manufacturing Meat and Poultry Products Organic Chemicals, Plastics And Synthetic Fibers Petroleum Refining Inorganic Chemicals Manufacturing Nonferrous Metals Manufacturing Ore Mining And Dressing Pesticide Chemicals Nonferrous Metals Forming And Metal Powders Drinking Water Treatment Textile Mills Timber Products Processing Soap And Detergent Manufacturing National Security & International Affairs Waste Combustors Landfills Sugar Processing Mineral Mining And Processing Pharmaceutical Manufacturing Plastics Molding And Forming Phosphate Manufacturing Aluminum forming Metal Molding And Casting (Foundries) Rubber Manufacturing Gum And Wood Chemicals Manufacturing Centralized Waste Treatment Electrical And Electronic Components Cement Manufacturing Engineering & Management Services Miscellaneous Foods And Beverages Independent And Stand Alone Labs Number of Facilities 1 550 111 217 40 21 90 44 219 108 55 36 54 139 14 13 48 5 2 35 10 10 21 34 28 6 12 12 7 17 2 6 5 6 1 8 6 Total Pounds 1,331,644 25,138,490,268 77,793,914 2,449,186,965 1,583,223,789 125,646,884 672,371,411 673,799,975 1,484,731,242 1,952,697,634 1,173,945,339 187,923,634 470,835,865 3,843,462,966 5,496,943 1,135,551,072 29,467,857 99,888,774 230,007 92,583,865 19,162,733 18,668,498 698,918,657 264,924,182 43,720,508 88,969,293 62,276,423 15,781,323 6,019,649 8,947,786 838,168 120,470,939 2,665,896 63,110,706 3,284,525 94,045,452 465,432 Total TWPE 909,115,642 20,374,829 3,361,768 2,726,865 1,220,744 1,095,046 730,252 535,913 413,226 402,506 393,523 342,747 184,455 179,697 119,244 119,190 79,934 51,552 47,815 38,983 38,412 35,804 32,520 26,719 24,937 24,626 18,459 12,182 11,271 11,195 10,478 10,403 9,350 8,960 5,978 5,842 5,355 3-61 ------- Section 3 - DMRLoads2007: Development and Category Rankings Table 3-22. DMR 2007 Point Source Category Rankings by TWPE 40CFR Part 424 408 468 434 NA 406 407 443 461 NA NA NA 457 NA 412 NA NA 426 NA NA NA NA 435 NA 465 NA 405 NA NA 460 NA 466 425 451 438 NA 4959 NA Point Source Category Ferroalloy Manufacturing Canned And Preserved Seafood Processing Copper forming Coal Mining Non Classifiable Establishments Grain mills Canned And Preserved Fruits And Vegetables Processing Paving And Roofing Materials (Tars And Asphalt) Battery Manufacturing Amusement & Recreation Services Printing & Publishing Environmental Quality & Housing Explosives Manufacturing General Building Contractors CAFO Justice, Public Order, & Safety Educational Services Glass Manufacturing Special Trade Contractors Construction And Development Lumber & Wood Products Airport Deicing Oil & Gas Extraction Real Estate Coil Coating Executive, Legislative, & General Dairy products processing Trucking & Warehousing Wholesale Trade- Durable Goods Hospital Pipelines, Except Natural Gas Porcelain Enameling Leather Tanning And Finishing Concentrated Aquatic Animal Production Metal Products And Machinery Tobacco Products Sanitary Services Transportation Services Number of Facilities 3 8 9 9 10 14 11 4 1 1 2 5 5 1 1 9 5 o 6 i 2 1 5 5 9 1 2 o 6 2 2 2 1 1 1 23 2 1 2 1 Total Pounds 7,905,371 124,735,909 2,928,183 44,228,933 24,794,788 28,624,175 7,177,438 494,518 136,061 118,566 1,039,175 5,849 21,980,426 41,817 10,812,796 1,351,009 4,934,978 2,715,981 8,073,573 28,460,736 8,975,046 1,162,405 531,118 4,857,073 445 53,109 262,241 83,278 538,559 9,134 289,497 13,507 33,076 5,310,357 1,187,703 10,740 653,114 713,322 Total TWPE 4,349 3,232 2,310 2,294 2,066 1,984 1,757 1,280 1,096 1,025 999 972 785 645 617 505 410 353 330 324 283 265 256 214 166 77 76 58 30 15 12 11 8 5 3 o J 3 o 5 3-62 ------- Section 3 - DMRLoads2007: Development and Category Rankings Table 3-22. DMR 2007 Point Source Category Rankings by TWPE 40CFR Part NA 459 442 NA Point Source Category Photo Processing Photographic Transportation Equipment Cleaning Wholesale Trade- Nondurable Goods Number of Facilities 1 1 2 1 Total Pounds 34,136 34,136 326,427 33,166 Total TWPE 1 1 0 0 Source: DMRLoads2007_v3. a414.1 refers to the chlorinated hydrocarbon segment of the Organic Chemicals, Plastics, and Synthetic Fibers Category (40 CFR Part) 414 and the Chlor-Alkali Subcategory of the Inorganic Chemicals Manufacturing Category (40 CFR Part 415). NA - Not applicable; no existing ELGs apply to discharges. • Load Calculator routines. EPA's quality review for the Load Calculator routines included accuracy checks for database queries onPCSLoadCalculator2007. EPA reviewed the programming code used to develop each query to verify the logic and verified that the number of records in the output table equaled the number of records in intermediate queries to ensure that no data were missing and that there were no duplicate data. EPA also verified the Load Calculator routine in the ICIS- NPDES Pollutant Loading Tool. EPA created a query-based system and compared the annual loads calculated by the queries to those calculated by the ICIS-NPDES Pollutant Loading Tool. The output from the queries was identical to that of the ICIS-NPDES Pollutant Loading Tool. In addition, EPA performed hand calculations to verify the accuracy of the PCSLoadCalculator2007 and ICIS-NPDES Load Calculator Module outputs during reviews of facility discharges for DMRLoads2007 results. • DMRLoads2007 results. EPA's quality review of the DMRLoads2007 results included the following: — Completeness checks: EPA compared counts of dischargers in DMRLoads2007 to PCSLoads2004 to describe the completeness of the database. There were 2,027 facilities that reported a load to PCSLoads2004 and 2,018 facilities that reported a load to DMRLoads2007'. Therefore, EPA determined DMRLoads2007 was complete. — Accuracy of facility discharges. EPA reviewed the accuracy of facilities' discharges that had the greatest impact on total category loads and category rankings to identify possible calculation errors. EPA reviewed monthly information reported in PCS and ICIS-NPDES, measurement data available on EPA's Envirofacts Web page, and information from the facility's NPDES permit. In some cases, EPA contacted facilities to verify the monthly measurements in their DMR. Section 3.4.2 describes EPA's review of facility discharges in more detail. — Accuracy of category discharges. EPA reviewed the accuracy of category discharges by verifying that pollutant discharges in PCS and ICIS-NPDES 3-63 ------- Section 3 - DMRLoads2007: Development and Category Rankings were assigned to the appropriate point source category. EPA used engineering judgment to determine if the pollutant discharge was reasonably associated with the point source category. Section 3.2.7 discusses facility-level and pollutant-level category assignments. Accuracy of database queries. EPA's quality review for the development of DMRLoads2007 included accuracy checks for database queries in DMRLoadsAnalysis2007, DMRNutrients2007, andDMRLoads2007. Documentation of accuracy checks is provided in a QC table in each Microsoft Access database. Reasonableness of pollutant loads. EPA reviewed the Load Calculator output (i.e., the calculated kg/year for each pollutant at each discharge pipe and monitoring location) for those pollutant discharges with the highest toxic-weighted loads (e.g., dioxins, PCBs, and mercury). To identify possible errors in recording units of measure, EPA identified calculated discharges that were orders of magnitude higher than previous years' discharges and other facilities within the same category. EPA reviewed quantities or concentrations and flows that the PCSLoadCalculator2007 and ICIS-NPDES Pollutant Loading Tool databases used to calculate the annual discharge. EPA compared these measurements with measurements available on EPA's Envirofacts web page. If the measurements were similar then EPA concluded that the output was acceptable. If the data did not match between the databases and Envirofacts, EPA corrected the data to match Envirofacts. When EPA was unsure of the correct data, EPA contacted the facility for more information (see Section 3.4.2). Reasonableness of facility loads. EPA identified facility discharges with the highest TWPE and nutrient pollutant loads. EPA identified facilities for review whose pollutant discharges accounted for more than 95 percent of the TWPE for its point source category. Similarly, EPA identified facilities for review whose nitrogen and phosphorus discharges account for the majority of nutrient discharges in DMRLoads2007'. EPA compared 2007 PCS and ICIS-NPDES data to other available information, such as information from EPA's Envirofacts web page, the facility's NPDES permit, and discussion with the facility contact. EPA made several facility- level corrections, as shown in Table 3-23. Comparability. EPA compared DMRLoads2007 to PCSLoads2004 and PCSLoads2002 to identify pollutant discharges or wastewater flows that differ more than the year-to-year variation of other chemicals and facilities. EPA used this comparison to determine if quantity, concentration, or flow corrections were needed for facility discharges with the highest TWPE. If the comparison was unavailable (e.g., the pollutant was not previously reported) EPA contacted the facility (see Table 3-23). 3-64 ------- Section 3 - DMRLoads2007: Development and Category Rankings The following sections discuss EPA's quality review for the development of DMRLoads2007: • Section 3.4.1 describes EPA's review of mercury using PRAM 50092 (Mercury Total Low Level); and • Section 3.4.2 describes EPA's facility review. 3.4.1 Mercury Discharges Reported Using PRAM 50092 During the reasonableness checks of the PCS CNVRT output, EPA identified unusually high mercury concentrations reported to PCS by facilities located in Ohio in the PCS CNVT output. These facilities reported mercury discharges using PRAM 50092 (Mercury Total Low Level). The PRAM 50092 concentrations in the 2004 CNVRT output ranged from 0.2 to 673 mg/L and from 0.0035 to 260,000 mg/L in the 2007 CNVRT output. EPA contacted the Ohio Environmental Protection Agency (Ohio EPA) to determine the correct reporting units for PRAM 50092 (Finseth, 2007c). An Ohio EPA representative explained that Ohio EPA started requiring low level mercury analyses in 2002. At that time, some facilities had limits in micrograms per liter (ug/L). Currently, all of the limits are in nanograms per liter (ng/L). As a result of this contact, EPA concluded that the units for the PRAM 50092 concentrations for the 2004 PCS data should be ng/L, not mg/L. The PRAM 50092 concentrations in the 2007 CNVRT output ranged from 0.0035 to 260,000 mg/L with greater than 99 percent of these concentrations between 0.5 and 800 mg/L. Based on this distribution, EPA concluded that the error for the 2004 data persisted in 2007. Therefore, EPA corrected the concentrations by dividing all concentrations for PRAM 50092 reported by facilities in Ohio in PCSLoadCalculator2007 by one million. EPA did not make any corrections to the ICIS-NPDES Pollutant Loading Tool because Ohio 2007 DMR data are only in PCS. 3.4.2 Facility Reviews EPA reviewed the accuracy of facility discharges that had the greatest impact on total category loads and category rankings. EPA used the following criteria to select facilities for review: • Facilities with the highest toxic-weighted discharges of individual pollutant parameters; • Facilities with the highest discharges of nutrients; and • Facilities with relatively high percent of their discharges based on estimates for missing DMR data (EST). For the identified facilities, EPA used the following steps to review the accuracy of the loads calculated from PCS and ICIS-NPDES data. 1. Reviewed database corrections for PCSLoads2004, PCSLoads2002, and PCSLoads2000 to determine whether corrections were made during previous reviews and evaluated whether these corrections should be applied to the 2007 DMR discharges. 3-65 ------- Section 3 - DMRLoads2007: Development and Category Rankings 2. Reviewed 2007 DMR data, hand calculated annual pollutant loads, and compared results to loads calculated by PCSLoadCalculator2007and the ICIS-NPDES Pollutant Loading tool, and stored in DMRLoads2007. 3. Reviewed PCS and ICIS-NPDES pipe description information available in PCS, EPA's on-line Envirofacts data system, ICIS-NPDES supporting tables, or from the facility's NPDES permit to identify monitored pollutant discharges that are: a. Intermittent (e.g., tidal, seasonal, or occur after a storm event) b. Internal monitoring locations from which wastewater is combined with other waste streams and monitored again, resulting in double counting loads, and c. Not representative of category discharges (e.g., storm water runoff from non-process areas, non-contact cooling water, or wastewater related to operations in another point source category). Table 3-23 presents EPA's review of facilities in DMRLoads2007 and the resulting corrections made to the database. 3-66 ------- Section 3 - DMRLoads2007: Development and Category Rankings Table 3-23. Summary of DMRLoads2007 Facility Review Facility Blue Heron Paper Company Cargill Fertilizer, Inc. - Riv CF Industries - Donaldsonville Clean Harbors White Castle LLC Climax Mine Doe Run Resources Co Location Oregon City, OR Hillsborough County, FL Donaldsonville, LA Iberville Parish, LA Summit County, CO Viburnum, MO Point Source Category Pulp and Paperboard Phosphate Manufacturing Fertilizer Manufacturing CWT Ore Mining and Dressing Ore Mining and Dressing Pollutant(s) in Question Methylmercury Phosphorous Nitrogen, Ammonia Benzidine Molybdenum Lead Review Findings Methylmercury concentrations in PCSLoadCalculator2007 are 1,000 times higher than the concentrations in Envirofacts. Envirofacts methylmercury concentrations are in ng/L but were entered into PCSLoadCalculator2007 as ug/L. Facility contact verified units should be ng/L (McCuutchen, 2009). Facility reports DRID 1 (monthly cone.) and A (annual quan.) with annual loads that do not equal. DMR is counting both DRIDs instead of just one also. Unable to determine the correct DRID to use based on Envirofacts. Maximum quantities are less than average quantities. Suspect that some average quantities should be divided by 10. Envirofacts has the same quantities. The permitted benzidine limit is three orders of magnitude lower than the concentrations in PCSLoadCalculator2007. Facility contact said that benzidine was ND (Ourso, 2009). This is a molybdenum mine. Units are consistent with Envirofacts and permit reporting limits. Permit/fact sheet contains self-monitoring data that agrees with the values reported to PCS (CO DH, 2004; CO DPHE, 2004). This is a lead or zinc mine based on SIC code. Units are consistent with Envirofacts and permit reporting limits (0.005 mg/L to 0.8 mg/L). Action Taken/ Database Correction Database Change: Correct methylmercury concentrations None None Database Change: Revise benzidine concentrations to zero None None 3-67 ------- Section 3 - DMRLoads2007: Development and Category Rankings Table 3-23. Summary of DMRLoads2007 Facility Review Facility Dyno Nobel, Inc. Envirosystems Incorporated Front St. Remedial Action GE Silicones, LLC General Electric - Erie Golden Eagle Refinery Location Carthage, MO Hampton, NH Kansas City, MO Friendly, WV Erie, PA Martinez, CA Point Source Category Explosives Manufacturing Independent And Stand Alone Labs Waste Combustors OCPSF Metal Finishing Petroleum refining Pollutant(s) in Question Nitrogen, Ammonia Cadmium Dioxin Copper Mercury TCDD Equivalents Review Findings For pram 00610 (Nitrogen, ammonia total (as N)), each outfall reports 6 months under DRID B and 6 months under DRID C. Flows for some months are 1 ,000 times greater than other months. Review of fact sheet shows that facility incorrectly reported flows in GPD instead of MGD for certain months (U.S. EPA Region 1,2006). Facility is a superfund site, and operated in the past as both a waste combustor and CWT. Currently treating groundwater contaminated by organics and inorganics. Three of four dioxin concentrations in 2007 were above the detection limit and the MDL. Concentrations were provided by permitting authority. Detected dioxin in Q2 2007 and Q3 and Q4 were ND. Lab did not analyze wastewater for dioxin for Ql (Archterlonie, 2009). Suspected copper concentrations units error because the permit reporting requirements are in ug/L instead of mg/L. Facility confirmed the units error and provided correct concentrations for two quarters. Data was reported as ug/L not mg/L (Martin, 2009a). Facility reported 3.3 mg/L in December 2007, reported annually. Verified units in OTIS. Facility said mercury should be ng/L instead of mg/L (Verderese, 2009). TCDD Equivalents measurements in database are 1 ,000 times larger than the concentrations in Envirofacts. The units for concentrations in Envirofacts are in pg/L. Action Taken/ Database Correction Database Change: Change DRID B and D to C for PRAM 00610 and divide affected flows by 1,000. Database Change: Correct flows for the affected monitoring periods Database Change: Revise SIC code to link to superfund category Future Database Change: Revise copper concentrations Future Database Change: Revise mercury concentration Database Change: Correct TCDD Equivalents measurements 3-68 ------- Section 3 - DMRLoads2007: Development and Category Rankings Table 3-23. Summary of DMRLoads2007 Facility Review Facility IMC - Phosphates Company Innovia Films Jackson County LAC Minerals Morgan's Point Plant Northshore Mining/Silver BayP PEPCO-Benning Location Donaldsonville, LA Tecumseh, KS Pascagoula, MS Central City, SD Morgan's Point, TX Silver Bay, MN Washington, DC Point Source Category Fertilizer Manufacturing Plastics Molding and Forming Fertilizer Manufacturing Ore Mining OCPSF Ore Mining and Dressing Steam Electric Power Generation Pollutant(s) in Question Phosphorous Carbon Disulfide Phosphorous Cyanide Chlorine Copper Arochlor 1260 Review Findings Highest phosphorous loads are from outfall 002. Loads are approximately the same using the quantity and the concentration calculations. Loads also are comparable to PCSLoads2004. One monthly concentration appears to be 100 times higher than the other months in 2007 and 2004. Facility contact provided corrected concentrations for April and May that were units errors (Martin, 2009b). Concentrations in Envirofacts match concentrations in PCSLoadCalculator200 7. A review of the permit and fact sheet indicated that the outfall SIR is an in- stream monitoring location and therefore should be excluded from the facility's loads (Fuller, 2005). The monthly average flow for March 2007 was 10,000 times higher than the monthly maximum flow for that month and the flows for the rest of the year. This is a taconite mine. Units are consistent with Envirofacts and permit reporting limits. The calculation relies on only one reported measurement when the permit shows facility must monitor monthly. A review of OTIS data shows that all PCBs were reported as BDL with "<" and a concentration. The data in ICIS-NPDES did not include the less-than signs. Because all monthly values are BDL, using the Hybrid Method all PCB loads should be zero. Action Taken/ Database Correction None Database Change: Correct carbon disulfide concentrations None Database Change: Change MLOC to Z (excluded from database) outfall STR Database Change: Correct March 2007 flow None Database Change: Zero all PCB (PRAM codes 39508, 39504, and 39496) loads 3-69 ------- Section 3 - DMRLoads2007: Development and Category Rankings Table 3-23. Summary of DMRLoads2007 Facility Review Facility Prasa El Yunque Filtration Plant Rhone-Poulenc Basic Chemicals Sabic Innovate Plastics SIGECO FB Cully Station Tampa Bay Desal Tosco Refinery (Rodeo) USA Holston Army Ammo Plant Area Location Rio Grande, PR Baton Rouge, LA Ottawa, IL Newburgh, IN Tampa Bay, FL Rodeo, CA King sport, TN Point Source Category Drinking Water Treatment Inorganic chemicals manufacturing OCPSF Steam Electric Power Generation Drinking Water Treatment Petroleum refining Explosives Manufacturing Pollutant(s) in Question Copper Phosphorus, Total (asP) Hexachlorobenzene Aluminum Chloride TCDD Equivalents RDX, Total Review Findings Review of the 2007 concentration data in OTIS indicated that February through August, November, and December copper concentrations were reported in ug/L but were in the ICIS-NPDES database as mg/L. A review of the facility's discharges and Envirofacts data shows the phosphorous concentrations should be in pg/L rather than ug/L. Review of concentration data for OTIS showed that the data were missing '<' signs for every month reported for all parameters except for copper. For aluminum, the concentration for 10 months is 1,000 times higher than the Form 2C data (2006) and 2006/2008 data in OTIS. Silver, arsenic, and cadmium concentrations are suspected units error based on the Form 2C data. Corrected concentrations to correspond to Form 2C data (SIGECO, 1994). Previous review identified a mismatch between flows and concentrations. NPDES permit fact sheet indicated the flow is diluted by 70 percent from the plant outfall to the final outfall (State of Florida, 2001). TCDD equivalents measurements in database are 1 ,000 times larger than the concentrations in Envirofacts. The units for concentrations in Envirofacts are in pg/L. Facility contact said the December 2007 value was RDX, Total production instead of effluent concentration. Contact provided correct concentration (House, 2009). Action Taken/ Database Correction Database Change: Revise affected copper concentrations by 1,000 Database Change: Revise Phosphorus, Total (As P) concentrations Database Change: Zero all loads except for copper Database Change: Revise metal concentrations Database Change: Divide monthly flows by 70 Database Change: Revise TCDD equivalents concentrations Database Change: Revise RDX, Total December 2007 concentration 3-70 ------- Section 3 - DMRLoads2007: Development and Category Rankings Table 3-23. Summary of DMRLoads2007 Facility Review Facility Westvaco Texas, L.P. Wise Alloys LLC Location Evadale, TX Muscle Shoals, AL Point Source Category Pulp, paper and paperboard Aluminum Forming Pollutant(s) in Question TCDD Equivalents Nitrogen, Nitrate Total (as N) Review Findings Concentrations in PCSLoadCalculator2007 are 1,000 times larger than the concentrations in Envirofacts. The units for concentrations in Envirofacts are in pg/L. Facility contact also said all quarters were ND, even though the fourth quarter did not have a '<' indicator (Davis, 2009). The facility reported two DRIDs: 1 (monthly cones.) and Q (quarterly quan.). Unable to determine the difference between DRIDs. Envirofacts does not have the permit/fact sheet. Action Taken/ Database Correction Database Change: Revise TCDD equivalents concentrations Future Database Change: Add < indicator to fourth quarter 2007 TCDD equivalents concentration None 3-71 ------- Section 3 - DMRLoads2007: Development and Category Rankings 3.5 DMRLoads2007 References 1. Auchterlonie, Steve. 2009. Notes from Telephone Conversation between Chris Krejci, Eastern Research Group, Inc. and Steve Auchterlonie, Front St. Remedial Action. RE: Verification of Magnitude and Basis of Estimate for Dioxin and Dioxin-Like Compound Discharges in PCS. (March 13). EPA-HQ-OW-2008-517 DCN 06636. 2. Camp, Meghan. 2009. Memorandum to Carey Johnston, U.S. EPA from Meghan Camp, Eastern Research Group, Inc. RE: Results of ICIS-NPDES Pollutant Loading Tool Convert Module Development and DMR Data Review. (May 29). EPA-HQ-OW-2008- 0517 DCN 06357. 3. Colorado Department of Health (CO DH). 2004. NPDES Permit for Climax Molybdenum Company, Summit County, Colorado. (September 13). EPA-HQ-OW- 2008-0517 DCN 05548. 4. Colorado Department of Public Health and Environment (CO DPHE). 2002. NPDES Permit for Climax Molybdenum Company, Summit County, CO. (December 31). EPA- HQ-OW-2008-0517 DCN 05972. 5. Davis, Katherine. 2009. Notes from Telephone Conversation between Elizabeth Sabol, Eastern Research Group, Inc. and Katherine Davis, Westvaco Texas, L.P. RE: Basis of TCDD Equivalent Concentrations Reported in 2007. (July 7). EPA-HQ-OW-2008-0517 DCN 06547. 6. Eastern Research Group, Inc. (ERG). 2009. Revised Quality Assurance Project Plan for the 2009 Annual Screening-Level Analysis ofTRI, ICIS-NPDES, and PCS Industrial Category Discharge Data. (September). EPA-HQ-OW-2008-0517 DCN 06558. 7. FL DEP. 2001. State of Florida Department of Environmental Protection. Save our Bays, Air and Canals, Inc. vs Tampa Bay Desal and Department of Environmental Protection. (October 17). EPA-HQ-OW-2008-0517 DCN 06699. 8. Fuller, Gene. South Dakota Department of Environment and Natural Resources. 2005. NPDES Permit for LAC Minerals, Inc., Lawrence, SD. (March 22). EPA-HQ-OW-2008- 0517 DCN 06638. 9. House, Nigel. 2009. Notes from Telephone Conversation between Jessica Wolford, Eastern Research Group, Inc. and Nigel House, USA Holston Army Ammo Plant Area. RE: Discussion of Total RDX Discharges in PCS for 2007. (April 13). EPA-HQ-OW- 2008-0517 DCN 06696. 10. Kandle, Meghan. 2005. Memorandum to 2006 Effluent Guidelines Program Plan Docket, EPA Docket Number OW-2004-032. RE: Point Source Category Rankings by Nitrogen and Phosphorus Loads Calculated Using 2002 PCS Data. (August 9). EPA-HQ-OW- 2004-0032-908. 3-72 ------- Section 3 - DMRLoads2007: Development and Category Rankings 11. Martin, Jason. 2009a. Notes from Telephone Conversation between Elizabeth Sabol, Eastern Research Group, Inc. and Jason Martin, MPM Silicones LLC. RE: Basis of Copper (Total Recoverable) Concentrations Reported in 2007. (July 1). EPA-HQ-OW- 2008-0517 DCN 06549. 12. Martin, Tony. 2009b. Notes from Telephone Conversation between Tony Martin, Innovia Films Inc. and Jessica Wolford, Eastern Research Group, Inc. RE: Discussion of Carbon Disulfide Discharges for Innovia Films Inc. (March 13). EPA-HQ-OW-2008-0517. DCN 06704. 13. McCuutchen, Kate. 2009. Notes from Telephone Conversation between Elizabeth Sabol, Eastern Research Group, Inc. and Kate McCuutchen, Blue Heron Paper Co. RE: Basis of Methylmercury Concentration Reported in 2007 in DMR. (July). EPA-HQ-OW-2008- 0517 DCN 06546. 14. Ourso, Lisa Jo. 2009. Notes from Telephone Conversation between Chris Krejci, Eastern Research Group, Inc. and Lisa Jo Ourso, Clean Harbors White Castle LLC. RE: Verification of Magnitude of Benzidine Discharges Reported in PCS. EPA-HQ-OW- 2008-0517 DCN 06403. 15. Southern Indiana Gas & Electric Company (SIGECO). 1994. Form 2C NPDES for SIGECO's F.B. Culley Generating Station. DCN 04496A34. 16. U.S. EPARegion 1. 2006. Fact Sheet for NPDES Permit NH0022055, EnviroSystems, Inc., Hampton, NH. (April 11). EPA-HQ-OW-2008-0517 DCN 06635. 17. U.S. EPA. 1997. Guidance and Standards for Calculating Point Source Pollutant Loads using the Permit Compliance System: Point Source Load Reductions as an Indicator of Water Quality Improvements. Washington, DC. (August). EPA-HQ-OW-2004-0032- 0891. 18. U.S. EPA. 2001. Permit Compliance System Generalized Retrieval Training Manual. Washington, DC. (February). EPA-HQ-OW-2003-0074 DCN 00357. 19. U. S. EPA. 2005. 2005 Annual Screening-Level Analysis: Supporting the Annual Review of Existing Effluent Limitations Guidelines and Standards and Identification of New Point Source Categories for Effluent Limitations Guidelines and Standards. EPA-821-B- 05-003. Washington, DC. (August). EPA-HQ-OW-2004-0032-0901. 20. U. S. EPA. 2007. Technical Support Document for the Preliminary 2008 Effluent Guidelines Program Plan. EPA-821-R-07-007. Washington, DC. (October). EPA-HQ- OW-2004-0032-1410. 21. Verderese, Jim. 2009. Notes from Telephone Conversation between Elizabeth Sabol, Eastern Research Group, Inc. and Jim Verderese, General Electric Erie. RE: Basis of Mercury Concentration Reported in December 2007 in DMR. (July). EPA-HQ-OW- 2008-0517 DCN 06548. 3-73 ------- Section 4 - Identification of Point Source Categories 4. IDENTIFICATION OF POINT SOURCE CATEGORIES The purpose of EPA's screening-level analysis is to use existing environmental data reported in discharge monitoring reports (DMRs) and in the Toxics Release Inventory (TRI) to investigate discharges from industrial point source categories and prioritize these categories for additional review. Specifically, EPA prioritizes its review of the industrial categories currently regulated by existing effluent limitations guidelines and standards (ELGs) that cumulatively compose 95 percent of the reported hazard (reported in units of toxic-weighted pound equivalent or TWPE). EPA focuses its efforts on collecting and analyzing data to identify industrial categories whose pollutant discharges potentially pose the greatest hazard to human health or the environment because of their toxicity (i.e., highest estimates of toxic-weighted pollutant discharges). The term "point source category" refers to an industry as a whole based on similarity of product produced or service provided, and is not meant to refer to specific industrial activities or processes involved in generating the product or service. EPA therefore identifies in its biennial Effluent Guidelines Program Plan only those new industries that it determines are properly considered stand-alone "categories" within the meaning of the CWA - not those that are properly considered potential new subcategories of existing categories based on similarity of product or service. As part of existing effluent guidelines and pretreatment standard annual reviews, EPA considers whether there are industrial activities not currently subject to effluent guidelines or pretreatment standards that should be included with these existing categories, either as part of existing subcategories or as potential new subcategories. Pursuant to CWA section 304(b), which requires EPA to establish ELGs for "classes and categories of point sources," EPA has promulgated ELGs for 56 industrial "categories." Each of these "categories" consists of a broad array of facilities that produce a similar product or perform a similar service - and is broken down into smaller subsets, termed "subcategories," that reflect variations in the processes, treatment technologies, costs and other factors associated with the production of that product that EPA is required to consider in establishing ELGs under section 304(b). For example, the Pulp, Paper and Paperboard Point Source Category (40 CFR Part 430) encompasses a diverse range of industrial facilities involved in the manufacture of a like product (paper); the facilities range from mills that produce the raw material (pulp) to facilities that manufacture end-products such as newsprint or tissue paper. EPA's classification of this "industry by major production processes used many of the statutory factors set forth in CWA Section 304(b), including manufacturing processes and equipment (e.g., chemical, mechanical, and secondary fiber pulping; pulp bleaching; paper making); raw materials (e.g., wood, secondary fiber, non-wood fiber, purchased pulp); products manufactured (e.g., unbleached pulp, bleached pulp, finished paper products); and, to a large extent, untreated and treated wastewater characteristics (e.g., BOD loadings, presence of toxic chlorinated compounds from pulp bleaching) and process water usage and discharge rates."18 Each subcategory reflects differences in the pollutant discharges and treatment technologies associated with each process. Similarly, the Iron and Steel Manufacturing Point 18 Supplemental Technical Development Document for Effluent Limitations Guidelines and Standards for the Pulp, Paper, and Paperboard Category, Page 5-3, EPA-821-R-97-011, October 1997 (U.S. EPA, 1997). 4-1 ------- Section 4 - Identification of Point Source Categories Source Category (40 CFR Part 420) consists of various subcategories that reflect the diverse range of processes involved in the manufacture of iron and steel, ranging from facilities that make the basic fuel used in the smelting of iron ore (Subpart A - Cokemaking) to those that cast the molten steel into molds to form steel products (Subpart F - Continuous Casting). An example of an industry category based on similarity of service provided is the Transportation Equipment Cleaning Point Source Category (40 CFR Part 442), which is subcategorized based on the type of tank (e.g., rail cars, trucks, barges) or cargo transported by the tanks cleaned by these facilities, reflecting variations in wastewaters and treatment technologies associated with each. Finally, Section 304(m)(l)(B) of the CWA directs EPA to use the biennial Effluent Guidelines Program Plans to identify categories of sources discharging non-trivial amounts of toxic or non-conventional pollutants for which EPA has not published ELGs under section 304(b)(2) or new source performance standards (NSPS) under section 306. EPA uses DMR and TRI data to assist in the identification of any new point source categories that meet the criteria of Section 304(m)(l)(B). EPA also uses TRI data to help identify indirect dischargers without categorical pretreatment standards to identify potential new categories for pretreatment standards under CWA Sections 304(g) and 307(b). EPA assesses whether industrial operations not currently regulated by existing effluent guidelines or pretreatment standards should be addressed as a potential new subcategory under an existing point source category rather than as a new industrial category. EPA uses Standard Industrial Classification (SIC) and North American Industrial Classification System (NAICS) codes to relate discharge data in DMR and TRI to the point source categories. DMR data are contained in EPA's Permit Compliance System (PCS) and the Integrated Compliance Information System for the National Pollutant Discharge Elimination System (ICIS-NPDES). As part of the 2009 annual review, EPA created DMRLoads2007 to combine DMR data from PCS and ICIS-NPDES (see Section 3). Facilities with data in PCS and/or ICIS-NPDES are identified by a four-digit SIC code, while facilities with data in TRI are identified by a six-digit NAICS code. To use the DMR and TRI data to estimate the pollutants discharged by each industrial point source category, EPA linked each four-digit SIC code and six-digit NAICS code to an appropriate point source category. EPA's linkages are summarized in the "SIC/Point Source Category Crosswalk" and "NAICS/Point Source Category Crosswalk" tables (Tables C-l and C-2 in Appendix C, respectively). These crosswalks are key elements of both the DMRLoads2007 and TRIReleases2007 databases. The remainder of this section presents the following information: • Section 4.1 - Background on NAICS and SIC Codes; • Section 4.2 - SIC Code to Point Source Category Crosswalk; • Section 4.3 - NAICS Code to Point Source Category Crosswalk; • Section 4.4 - Potential New Point Source Categories; and • Section 4.5 - Crosswalk References. 4.1 Background on NAICS and SIC Codes Starting in 2006, facilities reporting to EPA's TRI were required to provide the NAICS code(s) that describe their actions. The NAICS system is the current statistical classification standard underlying all establishment-based federal economic statistics classified by industry (U.S. Census Bureau, 2009). The Office of Management and Budget (OMB) first developed the ------- Section 4 - Identification of Point Source Categories NAICS system in 1997 to replace SIC codes and streamline economic statistics throughout North America. The first set of NAICS codes were updated in 2002 (referred to as the 2002 NAICS system), and the second set were updated in 2007 (referred to as the 2007 NAICS system). Although it was developed by OMB, the NAICS system is used by other government agencies, including EPA, to promote data comparability. In the NAICS system, each establishment is classified according to its primary economic activity, which is determined by its principal product or group of products. An establishment may have activities in more than one NAICS code. Some data collection organizations (e.g., the U.S. Economic Census) assign one NAICS code per establishment. On June 6, 2006, EPA published a final rule requiring facilities to use 2002 NAICS codes, instead of SIC codes, for reporting to the 2006 and 2007 TRI (see 71 FRN 32464). EPA is requiring facilities to use 2007 NAICS codes for reporting to the 2008 TRI and future years (see 73 FRN 32466, June 9, 2008). TRI allows facilities to identify their primary NAICS code and up to five additional NAICS codes. These codes reflect the principal activity causing environmental releases at a facility and other activities, respectively. Facilities with data in PCS and ICIS-NPDES are classified by SIC code. EPA has not announced plans to change its PCS and ICIS-NPDES databases to NAICS codes. As with the NAICS system, an establishment may have activities in more than one SIC code (OMB, 1987). PCS allows facilities to report one SIC code, while ICIS-NPDES allows facilities to report a primary SIC code and up to two additional SIC codes. The primary SIC code reflects the principal activity causing the discharge at each facility and the additional SIC codes represent other activities at the facility. As part of the 2009 annual review, EPA reviewed its existing SIC/Point Source Category Crosswalk to determine if revisions were necessary because facilities reported new SIC codes or additional information about their discharges. Because the TRI data for 2007 is classified by NAICS code, EPA created a NAICS/Point Source Category Crosswalk for TRIReleases2007. For a given facility, the operations covered by the SIC code in PCS and/or ICIS-NPDES may differ from operations covered by the primary NAICS code identified in TRI. 4.2 SIC Code to Point Source Category Crosswalk EPA first developed the SIC code to point source category crosswalk (SIC/Point Source Category Crosswalk) as part of the 2003 and 2004 screening-level analyses (U.S. EPA, 2005a). Since then, EPA has continued to update this crosswalk. Specifically for the 2009 screening- level review, EPA updated this crosswalk for use with DMRLoads2007'. For the 2009 annual review, as well as previous reviews, EPA divided the SIC codes into four groups defined as follows: • Existing Point Source Category. Discharges from most facilities in the SIC code meet the applicability requirements of an existing point source category. • Potential New Subcategory of an Existing Point Source Category. Discharges from most facilities in the SIC code may be considered part of a potential new subcategory of an industrial category subject to an existing ELG. EPA based this 4-3 ------- Section 4 - Identification of Point Source Categories determination on the similarity of processes and operations at facilities in the SIC code to those at facilities in the existing category. • Potential New Point Source Category. Discharges from facilities in the SIC code are similar to each other but do not meet the applicability requirements of and are not similar to a point source category subject to an existing ELG. • Category Not Identifiable. Facilities in the SIC code engage in a variety of industrial operations and likely meet the applicability requirements of several existing point source categories. However, EPA is not able to identify a coherent stand-alone point source category based on the SIC code description. Most SIC codes reported by facilities with DMR discharge information meet the applicability of an existing point source category and fall into the first group. The following sections describe the development and review of the SIC/Point Source Category Crosswalk. 4.2.1 SIC Codes Related to Existing Point Source Categories As part of its 2003 and 2004 screening-level analyses, EPA related SIC codes to existing point source categories. During the development of the existing ELGs for these categories, EPA studied demographic and economic data, including SIC code data, for the facilities to which the ELGs apply. EPA developed the relationship, or "crosswalk," between SIC codes and point source categories by consulting, as necessary, the documentation for the development on the existing ELGs. This crosswalk is included as Table C-l in Appendix C. Because most point source categories are not defined by SIC code, the relationship between SIC code and point source category is not a one-to-one correlation. A single SIC code may include facilities in more than one point source category, so associating an SIC code with only one category may be an over simplification. Also, many facilities have operations subject to more than one point source category. Further, facilities in some categories cannot be identified by SIC code. The following subsections discuss how EPA reconciled these inconsistencies to cross-reference appropriate point source categories to specific SIC codes. EPA reviewed each of these inconsistencies as part of the 2009 annual review and further refined the SIC/Point Source Category Crosswalk. 4.2.1.1 SIC Codes Counted in More than One Point Source Category A single SIC code may include facilities subject to more than one point source category. For example, SIC code 3357, Drawing and Insulating of Nonferrous Wire, includes facilities that draw wire made from aluminum, copper, and other nonferrous metals such as nickel and silver. Depending on the type of metal, ELGs from three categories may apply to the discharges from these operations. EPA included the loads discharged by facilities in SIC code 3357 in each of the three applicable categories: Aluminum Forming, Copper Forming, and Nonferrous Metals Forming. In order to make a "worst case" estimate of the TWPE discharged by every category, EPA included the loads from SIC codes associated with multiple point source categories in the load for each associated category, double- or triple-counting the loads from these SIC codes. Table 4-1 presents the SIC codes associated with multiple point source categories, and identifies the applicable point source categories. ------- Section 4 - Identification of Point Source Categories Table 4-1. SIC Codes Counted in Multiple Point Source Categories SIC Code 3357 3363 3482 3483 3463 4953 7221 7335 7336 7384 SIC Description Drawing and Insulating of Nonferrous Wire Aluminum Die Casting Small Arms Ammunition Ammunition, Except for Small Arms Nonferrous Forgings Refuse Systems Photographic Studios, Portrait Commercial Photography Commercial Art and Graphic Design Photofinishing Laboratories Applicable Point Source Categories Aluminum Forming (40 CFR 467), Copper Forming (40 CFR 468), and Nonferrous Metals Forming & Metal Powders (40 CFR 471) Aluminum Forming (40 CFR 467) and Nonferrous Metals Forming & Metal Powders (40 CFR 471) Metal Finishing (40 CFR 433) and Nonferrous Metals Forming & Metal Powders (40 CFR 471) Metal Finishing (40 CFR 433) and Nonferrous Metals Forming & Metal Powders (40 CFR 471) Aluminum Forming (40 CFR 467), Copper Forming (40 CFR 468), and Nonferrous Metals Forming & Metal Powders (40 CFR 471) Landfills (40 CFR 445) and Waste Combustors (40 CFR 444) Photographic (40 CFR 459) and Photoprocessing (2005 Annual Review Potential New Subcategory) a Photographic (40 CFR 459) and Photoprocessing (2005 Annual Review Potential New Subcategory) a Photographic (40 CFR 459) and Photoprocessing (2005 Annual Review Potential New Subcategory) a Photographic (40 CFR 459) and Photoprocessing (2005 Annual Review Potential New Subcategory) a a As part of the Final 2006 Plan, EPA determined that categorical pretreatment standards were not warranted for the Photoprocessing industry (U.S. EPA, 2006b). 4.2.1.2 SIC Codes Divided Among Point Source Categories As noted previously, some facilities are subject to regulations from more than one point source category. EPA was able to assign discharges from some of these SIC codes to the appropriate category and avoid double counting. EPA made some of these assignments at the facility level and some at the pollutant level, as discussed below. Facility-Level Point Source Category Assignment For some SIC codes with facilities subject to more than one point source category, EPA was able to assign each facility to a category. EPA reviewed information available about each facility to determine which point source category applied to the facility's operations. As part of the 2005 and 2006 annual reviews, EPA contacted facilities to understand which facility operations were the source of reported wastewater discharges if publically available information did not indicate the appropriate category. For example, for the 2005 annual review, EPA located information about facilities in SIC codes associated with both the Porcelain Enameling and Metal Finishing Categories. EPA used this information to determine the category most likely to ------- Section 4 - Identification of Point Source Categories apply to each facility's discharge (Wolford, 2005). As part of the 2009 screening-level review, EPA updated the SIC/Point Source Category Crosswalk based on this review (see Section 4.2.1.2). Facilities reporting these SIC codes for the first time in 2007 (e.g., previously reported a different SIC code or are new facilities) were reviewed to link the facility's discharges to the appropriate point source category as part of the 2009 annual review. Table 4-2 presents the SIC codes that EPA assigned to point source categories at the facility level. In future databases, as new facilities report SIC codes that do not link directly to a point source category (e.g., SIC code 2048 does not link to a point source category), EPA will review facility's operations and identify the appropriate point source category. EPA will also review operations of new facilities with significant TWPE in each of these SIC codes to determine if they are assigned to the appropriate point source category. As part of the 2009 screening-level review, EPA reviewed available information about pollutant loads and facility information for facilities reporting SIC code 4953, Refuse Systems, to determine if the facility's discharges were primarily associated with operations regulated by the Landfill Category (40 CFR Part 444) or by the Waste Combustor Category (40 CFR Part 445). EPA incorporated these changes into the SIC/Point Source Category Crosswalk in the 2007 DMR database based on this review (see Section 4.2.1.2). In future databases, as new facilities report SIC code 4953, EPA will individually review their operations to determine the category that most likely applies to the facility's discharges. Table 4-2. Facility-Level Point Source Category Assignment SIC Codes SIC 2048: Prepared Feed and Feed Ingredients for Animals and Fowl, Except Dogs and Cats 2819: Industrial Inorganic Chemicals, NEC 2874: Phosphatic Fertilizers 3341: Secondary Smelting and Refining of Nonferrous Metals 3431: Metal Sanitary Ware 3469: Metal Stampings, NEC 3471: Plating and Polishing Primary Associated Point Source Category None. In future databases as new facilities report this SIC code, EPA will review facility's operations and identify the appropriate point source category. Inorganic Chemicals Manufacturing (40 CFR Part 415) Phosphate Manufacturing (40 CFR Part 422) Nonferrous Metals Manufacturing (40 CFR Part 421) Metal Finishing (40 CFR Part 433) Metal Finishing (40 CFR Part 433) Metal Finishing (40 CFR Part 433) Other Associated Point Source Categories Grain Mills Manufacturing (40 CFR Part 406) Meat and Poultry Products (40 CFR Part 432) Pharmaceutical Manufacturing (40 CFR Part 43 9) Nonferrous Metals Manufacturing (40 CFR Part 471) Phosphate Manufacturing (40 CFR Part 422) Fertilizer Manufacturing (40 CFR Part 418) Coil Coating (40 CFR Part 465) Porcelain Enameling (40 CFR Part 467) Porcelain Enameling (40 CFR Part 467) Coil Coating (40 CFR Part 465) Expanded SIC Code (Assigned at Facility Level) 2048GRAIN 2048MPP 2048PH 2819NMM 2819PHOS 2874FER 3341CC 343 1PE 3469PE 3471CC 4-6 ------- Section 4 - Identification of Point Source Categories Table 4-2. Facility-Level Point Source Category Assignment SIC Codes SIC 3624: Carbon and Graphite Products 3633: Household Laundry Equipment 3639: Household Appliances, NEC 4953: Refuse Systems Primary Associated Point Source Category Metal Finishing (40 CFR Part 433) Metal Finishing (40 CFR Part 433) Metal Finishing (40 CFR Part 433) Landfills (40 CFR 445) and Waste Combustors (40 CFR 444) a Other Associated Point Source Categories Carbon Black Manufacturing (40 CFR Part 458) Porcelain Enameling (40 CFR Part 467) Porcelain Enameling (40 CFR Part 467) Landfills (40 CFR Part 445) Waste Combustors (40 CFR Part 444) Expanded SIC Code (Assigned at Facility Level) 3624CB 3633PE 3639PE 4953L 4953WC NEC - Not elsewhere classified. a In future databases as new facilities report this SIC code, EPA will review facility's operations and identify the appropriate point source category. EPA is currently considering revisions to a subset of the ELGs for Organic Chemicals, Pesticides, and Synthetic Fibers (OCPSF) (40 CFR 414) and the Inorganic Chemicals Manufacturing (40 CFR 415) for facilities that produce chlorine and chlorinated hydrocarbons (CCH). Because the CCH rulemaking is underway, for the 2009 annual review and previous reviews, EPA assigned the SIC code "VCCA" to the CCH facilities in the SIC/Point Source Category Crosswalk to separately identify these facilities (note VCCA, Vinyl Chloride and Chloralkali, is the previous name for the CCH rulemaking). The list of CCH facilities in DMRLoads2007 is included in Table C-3 in Appendix C. As part of the Pulp, Paper, and Paperboard Category (Pulp and Paper Category) (40 CFR Part 430) Detailed Study (U.S. EPA, 2006a), EPA reviewed the operations of facilities reporting SIC codes 2611: Pulp Mills, 2621: Paper Mills, and 2631: Paperboard Mills to determine the applicable subpart for each facility. A 1988 legal suit obligated EPA to address discharges of poly chlorinated dibenzo-(p)-dioxins and poly chlorinated dibenzofurans from 104 bleaching pulp mills, including nine dissolving pulp mills. During its response to the 1988 legal suit, EPA decided to review and revise the Pulp and Paper Category regulations in three phases. EPA addressed Phase I first, chose not to revise the ELGs for Phase II, and chose to support NPDES permit writers individually in developing permit-specific effluent limitations to control discharges of these chemicals from the remaining operating mills in Phase III. Because the Pulp and Paper Category regulations regulate facilities by process used and product produced they do not correspond to SIC codes. Therefore, EPA added "-1" to the SIC codes of facilities that met the applicability of Phase I: • Subpart B (Bleached Papergrade Kraft and Soda); and • Subpart E (Papergrade Sulfite). EPA added "-2" to the SIC codes of facilities that met the applicability of Phase II: 4-7 ------- Section 4 - Identification of Point Source Categories • Subpart C (Unbleached Kraft); • Subpart F (Semi-Chemical); • Subpart G (Groundwood, Chemic-Mechanical, and Chemi-Thermo-Mechanical); • Subpart H (Non-Wood Chemical Pulp); • Subpart I (Secondary Fiber Deink); • Subpart J (Secondary Fiber Non-Deink); • Subpart K (Fine and Lightweight Papers from Purchased Pulp); and • Subpart L (Tissue, Filter, Non-Woven and Paperboard from Purchased Pulp). EPA added "-3" to the SIC codes of facilities that met the applicability of Phase III: • Subpart A (Dissolving Kraft); and • Subpart D (Dissolving Sulfite). As part of the 2009 annual review, EPA continued incorporating the updated SIC codes identified during previous annual reviews. EPA did not review operations for new facilities reporting the SIC codes 2611, 2621, and 2631 to assign the regulatory phase to the facilities. Outfall-Level Point Source Category Assignment EPA was able to divide the pollutant discharges for selected facilities that discharge wastewater subject to more than one point source category by outfall. As part of the 2007 annual review, EPA reviewed discharges, permits, and permit fact sheets for facilities with high TWPE. EPA determined that one , Radford Army Ammunition Plant in Montgomery County, VA, had selected outfall that were regulated under OCPSF Category (40 CFR Part 414) while other outfall were regulated but the Explosives Manufacturing Category (40 CFR Part 457). EPA assigned the outfalls associated with OCPSF manufacturing to the OCPSF category by appending "OCPSF" to the facility's outfall-level SIC codes. EPA continued this assignment as part of the 2009 annual review. Pollutant-Level Point Source Category Assignment For most facilities that discharge wastewater subject to more than one point source category, EPA was not able to divide the pollutant discharges between applicable point source categories. The following subsections discuss two exceptions where EPA was able to assign wastewater discharges of certain chemicals to the appropriate point source category. Organic Chemicals, Plastics, and Synthetic Fibers/Pesticides The OCPSF ELGs (40 CFR Part 414) may apply to discharges from facilities in the following SIC codes: • 2821: Plastics Materials, Synthetic Resins, and Nonvulcanizable Elastomers; • 2823: Cellulosic Manmade Fibers; • 2824: Manmade Organic Fibers, Except Cellulosic; • 2865: Cyclic Organic Crudes and Intermediates, and Organic Dyes and Pigments; and • 2869: Industrial Organic Chemicals, Not Elsewhere Classified. ------- Section 4 - Identification of Point Source Categories In addition, EPA is considering including operations from the following five SIC codes as potential new subcategories of the OCPSF Category: • 2842: Specialty Cleaning, Polishing, and Sanitation Preparations; • 2844: Perfumes, Cosmetics, and Other Toilet Preparations; • 2891: Adhesives and Sealants; • 2899: Chemicals and Chemical Preparations, Not Elsewhere Classified; and • 5169: Chemicals and Allied Products, Not Elsewhere Classified. Some facilities in the regulated SIC codes and SIC codes of the potential new subcategory manufacture and/or formulate pesticides as well as other organic chemicals. Regulations for the Pesticide Chemicals Category (40 CFRPart 455) control discharges from pesticide operations. For the 2009 screening-level analysis of discharges from existing categories, and previous reviews, EPA subtracted all pesticide discharges from OCPSF and counted them as discharges from the Pesticides Chemicals Category, by appending a "P" to the facility's pollutant-level SIC code (e.g., EPA revised pesticide discharges from SIC code 2869 to SIC code 2869P). EPA used a table containing a list of pesticides and their CAS numbers to identify the pesticide releases from the OCPSF Category for both the DMR and TRI databases. In developing the list of pesticides, EPA started with the list of 272 pesticide active ingredients that was created during the most recent pesticides rulemaking. Some of the pesticides in the list of 272 active ingredients were multiple compounds, for example "2,4 D salts and esters" and "organo-tin pesticides," and were not identified by CAS number. EPA identified individual chemicals and CAS numbers for active ingredients in these groups and added them to the pesticides list. All of the chemicals identified from the list of 272 pesticide active ingredients were included in the pesticides list, except for biphenyl and dichlorobenzene. Biphenyl and dichlorobenzene were not included because EPA determined that OCPSF facilities use these chemicals for specific manufacturing uses not related to pesticides. EPA identified additional pesticide active ingredients by using the 1988 FIFRA and TSCA Enforcement System (FATES) Database and a list created in 2003 by the Office of Pesticide Programs (OPP). EPA combined the two lists and determined which of the pesticide active ingredients facilities reported discharging to the DMR databases in 2007. For reported discharges, EPA determined whether the pesticide active ingredient had significant non-pesticide related manufacturing uses. EPA did not add chemicals, such as acrolein, trichlorofuoromethane, silver, and sulfuric acid, whose primary use was non-pesticide-related, to the list, while EPA added chemicals whose primary purpose was pesticide-related to the list. The list of chemicals reported in the DMR and TRI databases that EPA considered pesticides for the purpose of its screening-level analysis of discharges from existing categories contains 415 chemicals. MP&M/Metal Finishing Regulations for the Metal Finishing Category (40 CFR Part 433) may apply to discharges from facilities in 179 SIC codes for which discharges were reported in DMR in 2007. Regulations for the Metal Products and Machinery (MP&M) Category (40 CFR Part 438) may apply to some of the pollutants directly discharged by facilities in 136 of these SIC codes. The final MP&M rule at 40 CFR Part 438. l(b) specifically excludes both metal-bearing wastewaters and wastewaters subject to other effluent guidelines (e.g., Metal Finishing). For its screening- level analysis of discharges from existing categories, EPA developed methodologies to apportion ------- Section 4 - Identification of Point Source Categories pollutant loads between the MP&M and Metal Finishing Categories. EPA applied this methodology to the 2009 screening-level analysis and previous reviews. The MP&M rule as promulgated regulates oil and grease (O&G) and total suspended solids (TSS) in direct discharges from certain facilities that generate oily wastewater; it does not specifically regulate any other chemicals. EPA used the list of organic "pollutants of concern" it had developed for the MP&M rule and identified 103 pollutants in the DMR databases, including O&G and TSS. For the 2009 screening-level analysis, EPA counted all discharges of these pollutants from the 136 MP&M SIC codes in DMRLoads2007 as MP&M discharges. EPA counted discharges of all other chemicals from these facilities in the Metal Finishing Category in DMRLoads2007'. EPA believes that the identified pollutants are those that are most likely associated with the non-metal bearing oily waste streams subject to the MP&M regulations, and that this apportionment, which avoids double counting pollutant loads, is a reasonable approach for screening-level analysis of discharges from existing categories. Table C-4 in Appendix C lists the 88 organic "pollutants of concern" for the MP&M rule. For the 2009 Annual Review, as for previous reviews, EPA matched DMR pollutants to the list of 88 MP&M chemicals using CAS numbers and the SUPERCAS table (described in Section 3). Using the SUPERCAS table, EPA matched 104 pollutant parameters to the list of 88 organic "pollutants of concern" for the MP&M rule that are discharged by facilities in the 136 MP&M SIC codes. EPA identified these 104 pollutant parameters as "Controlled by MP&M." Table C-5 in Appendix C presents the list of DMR parameters allocated to MP&M for the 2009 Annual Review. 4.2.1.3 Categories Not Identified By SIC Code (Centralized Waste Treaters) The Centralized Waste Treaters (CWT) Category (40 CFR Part 437) is not linked to specific SIC codes; therefore, the SIC/Point Source Category Crosswalk does not assign any SIC codes to the CWT Category. As part of the 2008 annual review, EPA reviewed the list of CWTs developed as part of the CWT rulemaking and assigned these facilities the SIC code of "CWT" and linked it to Part 437 in the SIC/Point Source Category Crosswalk. EPA also reviewed the facilities reporting SIC code 4953, Refuse Systems, and assigned CWT facilities reporting this SIC code the SIC code of "CWT" that links to Part 437. As part of the 2009 annual review, EPA reviewed the operations of all facilities reporting SIC code 4953 that were not previously assigned the "CWT" SIC code to determine if their operations were applicable to the CWT Category, Landfills Category (40 CFR Part 444), or Waste Combustors Category (40 CFR Part 445). 4.3 NAICS Code to Point Source Category Crosswalk The 2007 TRI data was the first reporting year that facilities were required to report NAICS codes rather than SIC codes. Therefore, as part of the 2009 screening-level analyses, EPA developed the NAICS code to point source category crosswalk (NAICS/Point Source Category Crosswalk) to link NAICS codes to appropriate point source categories for use with TRI data. EPA divided the NAICS codes into four groups, the same four groups as EPA used to develop the SIC/Point Source Category Crosswalk: 4-10 ------- Section 4 - Identification of Point Source Categories • Existing Point Source Category. Discharges from most facilities in the NAICS code meet the applicability requirements of an existing point source category. • Potential New Subcategory of an Existing Point Source Category. Discharges from most facilities in the NAICS code may be considered part of a potential new subcategory of an industrial category subject to an existing ELG. EPA based this determination on the similarity of processes and operations at facilities in the NAICS code of concern to those at facilities in the existing category. • Potential New Point Source Category. Discharges from facilities in the NAICS code are similar to each other but do not meet the applicability requirements of and are not similar to a point source category subject to an existing ELG. • Category Not Identifiable. Facilities in the NAICS code engage in a variety of industrial operations and likely meet the applicability requirements of several existing point source categories. However, EPA is not able to identify a coherent stand-alone point source category based on the NAICS code description. Most NAICS codes reported by facilities in TRI meet the applicability of an existing point source category and fall into the first group. 4.3.1 NAICS Codes Related to Existing Point Source Categories As part of its 2009 screening-level analysis, EPA related NAICS codes to existing point source categories. EPA developed this crosswalk by using TRI facilities' point source category assignments from previous years of review. For example, for the 2005 annual review SIC code 2023 is linked to the Dairy Products Processing Category (40 CFR Part 405), shown in Table 4-3. In 2007 TRI, Dietrich's Milk Products reported its pollutant discharges were from activities in the NAICS code 311514. EPA thus assigned NAICS 311514 to the Dairy Products Processing Category in the NAICS/Point Source Category Crosswalk. Using this method, EPA assigned the point source categories to all but 46 of the NAICS codes reported in TRIReleases2007'. These 46 NAICS codes were reported by facilities that did not report to TRI in 2005. For these, EPA assigned these NAICS codes to the appropriate point source category based on NAICS descriptions and point source category applicability. The resulting NAICS/Point Source Category Crosswalk is included as Table C-2 in Appendix C. Table 4-3. Example NAICS/Point Source Category Crosswalk Development Facility Name Dietrich's Milk Products 2005 SIC Code 2023 : Condensed and Evaporated Milk Point Source Category Dairy Products Processing (40 CFR Part 405) 2007 NAICS Code 3 115 14: Dry, Condensed, and Evaporated Dairy Product Manufacturing Because most point source categories are not defined by NAICS code, the relationship between NAICS code and point source category is not a one-to-one correlation. This is also the case for the SIC codes (see Section 4.2.1.1). A single NAICS code may include facilities in more than one point source category, and associating a NAICS code with only one category may be an over simplification. Also, many facilities have operations subject to more than one point source ------- Section 4 - Identification of Point Source Categories category. Further, some categories cannot be identified by NAICS code. The following subsections discuss how EPA reconciled these inconsistencies to cross-reference appropriate point source categories to specific NAICS codes. As part of previous annual reviews, EPA remedied some of these issues in the TRI databases. EPA carried these fixes over to the 2007 TRI database as part of the 2009 annual review. 4.3.1.1 NAICS Codes Counted in More than One Point Source Category A single NAICS code may include facilities subject to more than one point source category. For example, NAICS code 562211, Hazardous Waste Treatment and Disposal, includes facilities that operate treatment and/or disposal facilities for hazardous waste. Depending on the type of treatment and/or disposal, ELGs from three different categories may apply to the discharges from these operations. EPA included the loads discharged by facilities in NAICS code 562211 in each of the three applicable categories: Centralized Waste Treatment, Waste Combustors, and Landfills. In order to make a "worst case" estimate of the TWPE discharged by every category, EPA included the loads from NAICS codes associated with multiple point source categories in the load for each category, double- or triple-counting the loads from these NAICS codes. Table 4-4 presents the NAICS codes associated with multiple point source categories, and identifies the applicable point source categories. Table 4-4. NAICS Codes Counted in Multiple Point Source Categories NAICS Code 331521 332112 332992 332993 332999 335921 335929 562211 NAICS Description Aluminum Die-Casting Foundries Nonferrous Forging Small Arms Ammunition Manufacturing Ammunition (except Small Arms) Manufacturing All Other Miscellaneous Fabricated Metal Product Manufacturing Fiber Optic Cable Manufacturing Other Communication and Energy Wire Manufacturing Hazardous Waste Treatment and Disposal Applicable Point Source Categories Nonferrous Metals Manufacturing (40 CFR Part 421) and Aluminum Forming (40 CFR Part 467)a Aluminum Forming (40 CFR Part 467), Copper Forming (40 CFR Part 468), and Nonferrous Metals Forming and Metal Powders (40 CFR Part 471)b Metal Finishing (40 CFR Part 433) and Nonferrous Metals Forming and Metal Powders (40 CFR Part 471) Metal Finishing (40 CFR Part 433) and Nonferrous Metals Forming and Metal Powders (40 CFR Part 471) Metal Finishing (40 CFR Part 433), Aluminum Forming (40 CFR Part 467), Copper Forming (40 CFR Part 468), and Nonferrous Metals Forming and Metal Powders (40 CFR Part 471)° Glass Manufacturing (40 CFR Part 426) and Plastics Molding and Forming (40 CFR Part 463) Aluminum Forming (40 CFR Part 467), Copper Forming (40 CFR Part 468), and Nonferrous Metals Forming and Metal Powders (40 CFR Part 471) Centralized Waste Treatment (40 CFR Part 437), Waste Combustors (40 CFR Part 444), and Landfills (40 CFR Part 445)d 4-12 ------- Section 4 - Identification of Point Source Categories Table 4-4. NAICS Codes Counted in Multiple Point Source Categories NAICS Code 562219 NAICS Description Other Nonhazardous Waste Treatment and Disposal Applicable Point Source Categories Centralized Waste Treatment (40 CFR Part 437), Waste Combustors (40 CFR Part 444), and Landfills (40 CFR Part 445)d aEPA reviewed publicly available information for these facilities and determined that some reporting this NAICS code have operations applicable to the Metal Molding and Casting Category (40 CFR Part 464). EPA assigned these facilities the NAICS code 331521MMC to link to the Metal Molding and Casting Category in the NAICS/Point Source Category Crosswalk (see Section 4.3.1.2). b EPA reviewed publicly available information for these facilities and determined that some reporting this NAICS code have operations applicable to the Iron and Steel Manufacturing Category (40 CFR Part 420) and the Metal Finishing Category (40 CFR Part 433). EPA assigned the NAICS code 332112IRON to facilities generating wastewater to which the Iron and Steel ELGs apply and the NAICS code 332112MF to facilities generating wastewater to which the Metal Finishing ELGs apply (see Section 4.3.1.2). 0 EPA reviewed publicly available information for these facilities and determined that some reporting this NAICS code have operations that are applicable to the Metal Finishing Category (40 CFR Part 433) and the Nonferrous Metals Forming and Metal Powders Category (40 CFR Part 471). EPA assigned the NAICS code 332999DC to these facilities (see Section 4.3.1.2). EPA determined that some facilities reporting this NAICS code have operations that are applicable to the Aluminum Forming Category (40 CFR Part 467), Copper Forming Category (40 CFR Part 468), and Nonferrous Metals Forming and Metal Powders Category (40 CFR Part 471). EPA assigned the NAICS code 332999TC to these facilities (see Section 4.3.1.2). d EPA reviewed publicly available information for these facilities and identified facilities with operations applicable to the Centralized Waste Treatment Category (40 CFR Part 437) with the NAICS code CWT. EPA assigned the NAICS code WC to facilities with operations applicable to the Waste Combustor Category (40 CFR Part 444). EPA assigned the NAICS code LNDFLL to facilities with operations applicable to the Landfills Category (40 CFR Part 445). Facilities with multiple operations were counted in all the applicable categories. 4.3.1.2 NAICS Codes Divided Among Point Source Categories As noted previously, some facilities are subject to regulations from more than one point source category. EPA was able to assign some of these discharges to the appropriate category and avoid double counting by carrying over changes made during previous annual reviews. EPA made some of these assignments at the facility level, the pollutant level, and the discharge level, as discussed below. Facility-Level Point Source Category Assignment For NAICS codes that include facilities subject to more than one point source category, EPA reviewed available information about pollutant loads and manufacturing operations to assign each facility to the category that applied to its discharges. Table 4-5 presents the NAICS codes that EPA assigned to point source categories at the facility level. In future databases, EPA will review facilities with significant TWPE in each of these NAICS codes to determine if they are assigned to the appropriate point source category. 4-13 ------- Section 4 - Identification of Point Source Categories Table 4-5. Facility-Level Point Source Category Assignment NAICS Codes NAICS 311119: Other Animal Food Manufacturing 31 1225: Fats and Oils Refining and Blending 31 1999: All Other Miscellaneous Food Manufacturing 315992: Glove and Mitten Manufacturing 324199: All Other Petroleum and Coal Products Manufacturing 325120: Industrial Gas Manufacturing 325188: All Other Basic Inorganic Chemical Manufacturing 325510: Paint and Coating Manufacturing Primary Associated Point Source Category Food and Kindred Products Potential New Point Source Category a Miscellaneous Foods and Beverages Potential New Point Source Category a Miscellaneous Foods and Beverages Potential New Point Source Category a Textile Mills (40 CFR Part 410) Petroleum Refining (40 CFR Part 4 19) Inorganic Chemicals Manufacturing (40 CFR Part 415) Inorganic Chemicals Manufacturing (40 CFR Part 415) Paint Formulating (40 CFR Part 446) Other Associated Point Source Categories Grain Mills (40 CFR Part 406) Meat and Poultry Products (40 CFR Part 432) Pharmaceutical Manufacturing (40 CFR Part 439) Fertilizer Manufacturing Category (40 CFR Part 4 18) Dairy Products Processing (40 CFR Part 405) Grain Mills (40 CFR Part 406) Organic Chemicals, Plastics, and Synthetic Fibers (40 CFR Part 414) Meat and Poultry Products (40 CFR Part 432) Rubber Manufacturing (40 CFR Part 428) Apparel and Other Textile Products Potential New Subcategory of Textile Mills (40 CFR Part 4 10) Organic Chemicals, Plastics, and Synthetic Fibers (40 CFR Part 414) Organic Chemicals, Plastics, and Synthetic Fibers (40 CFR Part 414) Organic Chemicals, Plastics, and Synthetic Fibers (40 CFR Part 414) Soap and Detergent Manufacturing (40 CFR Part 4 17) Nonferrous Metals Manufacturing (40 CFR Part 421) Phosphate Manufacturing (40 CFR Part 422) Copper Forming (40 CFR Part 468) Pharmaceutical Manufacturing (40 CFR Part 439) Nonferrous Metals Forming and Metal Powders (40 CFR Part 471 ) Cement Manufacturing (40 CFR Part 411) Electroplating (40 CFR Part 4 13) Organic Chemicals, Plastics, and Synthetic Fibers (40 CFR Part 414) Inorganic Chemicals Manufacturing (40 CFR Part 415) Expanded NAICS Code (Assigned at the Facility Level) 311119GRAIN 311119MPP 311119PH 311225FER 311999DPP 311999GRAIN 3119990CPSF 311999MPP 315992RUB 315992AP 324199OCPSF 3251200CPSF 325188OCPSF 325188SD 325188NMM 325188PHOS 325188COP 325188PH 325188NMF 325510CEM 325510ELEC 3255100CPSF 325510INORG 4-14 ------- Section 4 - Identification of Point Source Categories Table 4-5. Facility-Level Point Source Category Assignment NAICS Codes NAICS 32561 1 : Soap and Other Detergent Manufacturing 325998: All Other Miscellaneous Chemical Product and Preparation Manufacturing 326199: All Other Plastics Product Manufacturing 3311 11: Iron and Steel Mills 331221: Rolled Steel Shape Manufacturing 331314: Secondary Smelting and Alloying of Aluminum 331423: Secondary Smelting, Refining, and Alloying of Copper 331491: Nonferrous Metal (except Copper and Aluminum) Rolling, Drawing, and Extruding 331492: Secondary Smelting, Refining, and Alloying of Nonferrous Metal (except Copper and Aluminum) 331521: Aluminum Die-Casting Foundries b Primary Associated Point Source Category Soap and Detergent Manufacturing (40 CFR Part 417) Organic Chemicals, Plastics, and Synthetic Fibers (40 CFR Part 414) Plastics Molding and Forming (40 CFR Part 463) Iron and Steel Manufacturing (40 CFR Part 420) Iron and Steel Manufacturing (40 CFR Part 420) Nonferrous Metals Manufacturing (40 CFR Part 421) Nonferrous Metals Manufacturing (40 CFR Part 421) Nonferrous Metals Forming and Metal Powders (40 CFR Part 471) Nonferrous Metals Manufacturing (40 CFR Part 421) Nonferrous Metals Manufacturing (40 CFR Part 421) and Aluminum Forming (40 CFR Part 467) Other Associated Point Source Categories Organic Chemicals, Plastics, and Synthetic Fibers (40 CFR Part 414) Business Services Potential New Point Source Category Inorganic Chemicals Manufacturing (40 CFR Part 415) Soap and Detergent Manufacturing (40 CFR Part 4 17) Petroleum Refining (40 CFR Part 419) Metal Finishing (40 CFR Part 433) Pharmaceutical Manufacturing (40 CFR Part 439) Pesticide Chemicals (40 CFR Part 455) Nonferrous Metals Forming and Metal Powders (40 CFR Part 471 ) Electroplating (40 CFR Part 4 13) Organic Chemicals, Plastics, and Synthetic Fibers (40 CFR Part 414) Glass Manufacturing (40 CFR Part 426) Metal Finishing (40 CFR Part 433) Metal Finishing (40 CFR Part 433) Nonferrous Metals Forming and Metal Powders (40 CFR Part 471 ) Electroplating (40 CFR Part 4 13) Nonferrous Metals Forming and Metal Powders (40 CFR Part 471 ) Metal Finishing (40 CFR Part 433) Metal Molding and Casting (40 CFR Part 464) Aluminum Forming (40 CFR Part 467) Metal Molding and Casting (40 CFR Part 464) Nonferrous Metals Forming and Metal Powders (40 CFR Part 471 ) Metal Finishing (40 CFR Part 433) Copper Forming (40 CFR Part 468) Nonferrous Metals Forming and Metal Powders (40 CFR Part 471 ) Metal Molding and Casting (40 CFR Part 464) Expanded NAICS Code (Assigned at the Facility Level) 32561 10CPSF 325998BS 325998INORG 325998SD 325998PR 325998MF 325998PH 325998P 325998NMF 326199ELEC 326199OCPSF 326199GLASS 326199MF 331111MF 331111NMF 331221ELEC 331221NMF 331314MF 331314MMC 331314AL 331423MMC 331423NMF 331491MF 331492COP 331492NMF 331521MMC 4-15 ------- Section 4 - Identification of Point Source Categories Table 4-5. Facility-Level Point Source Category Assignment NAICS Codes NAICS 3321 12: Nonferrous Forging b 332618: Other Fabricated Wire Product Manufacturing 332813: Electroplating, Plating, Polishing, Anodizing, and Coloring 332999: All Other Miscellaneous Fabricated Metal Product Manufacturing 336340: Motor Vehicle Brake System Manufacturing 336360: Motor Vehicle Seating and Interior Trim Manufacturing 337215: Showcase, Partition, Shelving, and Locker Manufacturing Primary Associated Point Source Category Nonferrous Metals Manufacturing (40 CFR Part 421); Aluminum Forming (40 CFR Part 467); and Copper Forming (40 CFR Part 468) Metal Finishing (40 CFR Part 433) Electroplating (40 CFR Part 413) Metal Finishing (40 CFR Part 433) Metal Finishing (40 CFR Part 433) Textile Mills (40 CFR Part 410) Metal Finishing (40 CFR Part 433) Other Associated Point Source Categories Iron and Steel Manufacturing (40 CFR Part 420) Metal Finishing (40 CFR Part 433) Iron and Steel Manufacturing (40 CFR Part 420) Nonferrous Metals Forming and Metal Powders (40 CFR Part 471 ) Printing and Publishing Potential New Point Source Category a Iron and Steel Manufacturing (40 CFR Part 420) Metal Finishing (40 CFR Part 433) Plastics Molding and Forming (40 CFR Part 463) Aluminum Forming (40 CFR Part 467) Printing and Publishing Potential New Point Source Category a Metal Finishing (40 CFR Part 433) and Nonferrous Metals Forming and Metal Powders (40 CFR Part 471 ) Aluminum Forming (40 CFR Part 467); Copper Forming (40 CFR Part 468); and Nonferrous Metals Forming and Metal Powders (40 CFR Part 471 ) Electroplating (40 CFR Part 4 13) Metal Finishing (40 CFR Part 433) Timber Products Processing (40 CFR Part 429) Expanded NAICS Code (Assigned at the Facility Level) 332112IRON 332112MF 33261 8IRON 33261 8NMF 33261 8PP 33281 3IRON 332813MF 33281 3PMF 332813AL 33281 3PP 332999DC b 332999TC b 336340ELEC 336360MF 33721 5TIM 4-16 ------- Section 4 - Identification of Point Source Categories Table 4-5. Facility-Level Point Source Category Assignment NAICS Codes NAICS 339999: All Other Miscellaneous Manufacturing Primary Associated Point Source Category Metal Finishing (40 CFR Part 433) Other Associated Point Source Categories Organic Chemicals, Plastics, and Synthetic Fibers (40 CFR Part 414) Mineral Mining and Processing (40 CFR Part 436) Pesticide Chemicals (40 CFR Part 455) Plastics Molding and Forming (40 CFR Part 463) Nonferrous Metals Forming and Metal Powders (40 CFR Part 471 ) Expanded NAICS Code (Assigned at the Facility Level) 3399990CPSF 339999MIN 339999P 339999PMF 339999NMF a As part of the Final 2006 Plan, EPA determined that categorical pretreatment standards were not warranted for the these industries (U.S. EPA, 2006b). b A single NAICS code may include facilities subject to more than one point source category. EPA included the loads from NAICs codes associated with multiple point source categories in the load for each category, double- or triple-counting the loads from these NAICS codes (see Section 4.3.1.1). EPA is currently considering revisions to a subpart of the ELGs for Organic Chemicals, Pesticides, and Synthetic Fibers (OCPSF) (40 CFR 414) and Inorganic Chemicals Manufacturing (40 CFR 415) for facilities that produce chlorine and chlorinated hydrocarbons (CCH). Because the CCH rulemaking is underway, EPA assigned the NAICS code "VCCA" to the CCH facilities in the NAICS/Point Source Category Crosswalk to separately identify these facilities (Note: VCCA, vinyl chloride and chlor alkali, is the former name of the CCH rulemaking). The list of CCH facilities in TRIReleases2007 is included in Table C-6 in Appendix C. As part of the Pulp, Paper, and Paperboard Category (Pulp and Paper Category) (40 CFR Part 430) Detailed Study (U.S. EPA, 2006a), EPA reviewed the operations of facilities reporting SIC codes 2611: Pulp Mills, 2621: Paper Mills, and 2631: Paperboard Mills to determine the applicable subpart for each facility. A 1988 legal suit obligated EPA to address discharges of poly chlorinated dibenzo-(p)-dioxins and poly chlorinated dibenzofurans from 104 bleaching pulp mills, including nine dissolving pulp mills. During its response to the 1988 legal suit, EPA decided to review and revise the Pulp and Paper Category regulations in three phases. EPA addressed Phase I first, chose not to revise the ELGs for Phase II, and chose to support NPDES permit writers individually in developing permit-specific effluent limitations to control discharges of these chemicals from the remaining operating mills in Phase III. Because the Pulp and Paper Category regulations are subcategorized by process used and product produced they do not correspond to SIC codes. Therefore, EPA added "-1" to the SIC codes of facilities that met the applicability of Phase I: • Subpart B (Bleached Papergrade Kraft and Soda); and • Subpart E (Papergrade Sulfite). EPA added "-2" to the SIC codes of facilities that met the applicability of Phase II: Subpart C (Unbleached Kraft); Subpart F (Semi-Chemical); 4-17 ------- Section 4 - Identification of Point Source Categories • Subpart G (Groundwood, Chemic-Mechanical, and Chemi-Thermo-Mechanical); • Subpart H (Non-Wood Chemical Pulp); • Subpart I (Secondary Fiber Deink); • Subpart J (Secondary Fiber Non-Deink); • Subpart K (Fine and Lightweight Papers from Purchased Pulp); and • Subpart L (Tissue, Filter, Non-Woven and Paperboard from Purchased Pulp). EPA added "-3" to the SIC codes of facilities that met the applicability of Phase III: • Subpart A (Dissolving Kraft); and • Subpart D (Dissolving Sulfite). EPA carried the facility-specific changes from the 2004 and 2005 screening-level reviews to the TRIReleases2007 database by appending "-1", "-2", and "-3" to the NAICS codes for facilities in the Pulp and Paper Category. Discharge-Level Point Source Category Assignment Regulations for the Electroplating Category (40 CFR Part 413) apply to discharges from indirect discharging facilities, while direct discharging electroplating operations are regulated by the Metal Finishing Category (40 CFR Part 433). EPA determined facilities reporting the following NAICS codes may have electroplating operations: • 325510: Paint and Coating Manufacturing; • 326199: All Other Plastics Products Manufacturing; • 331221: Rolled Steel Shape Manufacturing; • 332813: Electroplating, Plating, Polishing, Anodizing, and Coloring; and • 336340: Motor Vehicle Brake System Manufacturing. In TRIReleases2007 facilities can report direct and indirect discharges. Therefore, as part of the 2009 screening-level review, for facilities reporting the above NAICS codes EPA assigned direct discharges to the Metal Finishing Category by appending "MF" to the facility's discharge-level NAICS code (e.g., EPA revised direct discharges from NAICS code 332813 to 332813MF). Pollutant-Level Point Source Category Assignment For most facilities that discharge wastewater subject to more than one point source category, EPA was not able to divide the pollutant discharges between applicable point source categories. The following subsections discuss two exceptions where EPA was able to assign wastewater discharges of certain chemicals to the appropriate point source category. Organic Chemicals, Plastics, and Synthetic Fibers/Pesticides The OCPSF ELGs (40 CFR Part 414) may apply to discharges from facilities in the following NAICS codes: • 325132: Synthetic Organic Dye and Pigment Manufacturing; • 325192: Cyclic Crude and Intermediate Manufacturing; • 325199: All Other Basic Organic Chemical Manufacturing; ------- Section 4 - Identification of Point Source Categories • 325211: Plastics Material and Resin Manufacturing; • 325221: Cellulosic Organic Fiber Manufacturing; and • 325222: Noncellulosic Organic Fiber Manufacturing. In addition, EPA is considering including operations from the following 16 NAICS codes as potential new subcategories of the OCPSF Category: • 311999OCPSF: All Other Miscellaneous Food Manufacturing; • 324199OCPSF: All Other Petroleum and Coal Products Manufacturing; • 325510: Petrochemical Manufacturing; • 325120OCPSF: Industrial Gas Manufacturing; • 325188OCPSF: All Other Basic Inorganic Chemical Manufacturing; • 325193: Ethyl Alcohol Manufacturing; • 325510OCPSF: Paint and Coating Manufacturing; • 325520: Adhesive Manufacturing; • 325611 OCPSF: Soap and Other Detergent Manufacturing; • 325612: Polish and Other Sanitation Good Manufacturing; • 325620: Toilet Preparation Manufacturing; • 325998: All Other Miscellaneous Chemical Product and Preparation Manufacturing; • 326199OCPSF: All Other Plastics Product Manufacturing; • 339999OCPSF: All Other Miscellaneous Manufacturing; • 424690: Other Chemical and Allied Products Merchant Wholesalers; and • 562920: Materials Recovery Facilities. Some facilities in the regulated NAICS codes and NAICS codes of the potential new subcategory manufacture and/or formulate pesticides as well as other organic chemicals. Regulations for the Pesticide Chemicals Category (40 CFRPart 455) control discharges from pesticide operations. For the screening-level analysis of discharges from existing categories, EPA therefore subtracted all pesticide discharges from OCPSF and counted them as discharges from the Pesticides Chemicals Category, by appending a "P" to the facility's pollutant-level SIC code (e.g., EPA revised pesticide discharges from NAICS code 325199 to NAICS code 325199P). EPA developed this methodology as part of the 2005 annual review for use with the PCS and TRI data. EPA did not change the methodology for the 2009 annual review. EPA used a table containing a list of pesticides and their CAS numbers in order to identify the pesticide releases from the OCPSF Category for both the DMR and TRI databases. In developing the list of pesticides, EPA started with the list of 272 pesticide active ingredients that was created during the most recent pesticides rulemaking. Some of the pesticides in the list of 272 active ingredients were multiple compounds, for example "2,4 D salts and esters" and "organo-tin pesticides," and were not identified by CAS number. EPA identified individual chemicals and CAS numbers for active ingredients in these groups and added them to the pesticides list. All of the chemicals identified from the list of 272 pesticide active ingredients were included in the pesticides list, except for biphenyl and dichlorobenzene. Biphenyl and dichlorobenzene were not included because EPA determined that OCPSF facilities use these chemicals for specific manufacturing uses not related to pesticides. 4-19 ------- Section 4 - Identification of Point Source Categories EPA identified additional pesticide active ingredients by using the 1988 FIFRA and TSCA Enforcement System (FATES) Database and a list created in 2003 by the Office of Pesticide Programs (OPP). EPA combined the two lists and determined which of the pesticide active ingredients facilities reported having discharged in TRI in 2007. For releases reported in the 2007 TRI, EPA determined whether the pesticide active ingredient had significant non- pesticide related manufacturing uses. Chemicals, such as acrolein, trichlorofuoromethane, silver, and sulfuric acid, whose primary use was non-pesticide-related were not added to the list, while chemicals whose primary purpose was pesticide-related were added to the list. The list of chemicals reported in TRI and DMR that EPA considered pesticides for the purpose of its screening-level analysis of discharges from existing categories contains 415 chemicals. MP&M/Metal Finishing Regulations for the Metal Finishing Category (40 CFR Part 433) may apply to discharges from facilities in 198 NAICS codes for which discharges were reported in TRI in 2007. Regulations for the Metal Products and Machinery (MP&M) Category (40 CFR Part 438) may apply to some of the pollutants directly discharged by facilities in 165 of these NAICS codes. The final MP&M rule at 40 CFR Part 438. l(b) specifically excludes both metal-bearing wastewaters and wastewaters subject to other effluent guidelines (e.g., Metal Finishing). For its screening-level analysis of discharges from existing categories, EPA developed methodologies to apportion pollutant loads between the MP&M and Metal Finishing Categories. The MP&M rule as promulgated regulates oil and grease (O&G) and total suspended solids (TSS) in direct discharges from certain facilities that generate oily wastewater; it does not specifically regulate any other chemicals. EPA used the list of organic "pollutants of concern" it had developed for the MP&M rule and identified 48 pollutants in the TRI databases, including O&G and TSS. For the 2009 screening-level analysis, EPA counted all discharges of these pollutants from the 165 MP&M NAICS codes in TRIReleases2007 as MP&M discharges. EPA counted discharges of all other chemicals from these facilities in the Metal Finishing Category in TRIReleases2007. EPA believes that the identified pollutants are those that are most likely associated with the non-metal bearing oily waste streams subject to the MP&M regulations, and that this apportionment, which avoids double counting pollutant loads, is a reasonable approach for screening-level analysis of discharges from existing categories. For the 2009 annual review, as for previous reviews, EPA matched TRI pollutants to the list of 88 MP&M chemicals using CAS numbers and the SUPERCAS table (described in Section 1). Using the SUPERCAS table, EPA matched 48 chemicals to the list of 88 organic "pollutants of concern" for the MP&M rule that are discharged by facilities in the 165 MP&M NAICS codes. EPA identified these 48 chemicals as "Controlled by MP&M." Table C-4 in Appendix C lists the 88 organic "pollutants of concern" for the MP&M rule. Table C-7 in Appendix C presents the list of TRI chemicals allocated to MP&M for the 2009 annual review. 4.4 Potential New Point Source Categories Concurrent with its review of existing point source categories, EPA also reviews industries not currently subject to effluent guidelines to identify potential new point source categories. EPA conducts a "crosswalk" analysis based on data in DMR and TRI. Facilities with data in DMR and TRI are identified by a four-digit SIC code or six-digit NAICS code (Section 4.1 provides more details on SIC and NAICS codes, respectively). EPA links each four-digit SIC 4^20 ------- Section 4 - Identification of Point Source Categories code and six-digit NAICS code to an appropriate industrial category (i.e., "the crosswalk").19 This crosswalk identifies SIC codes and NAICS codes that EPA associated with industries subject to an existing guideline. The crosswalk also identifies SIC and NAICS codes not associated with an existing guideline. In addition to the crosswalk analysis, EPA relies on stakeholder comments to identify potential new point sources categories. For each industry identified through the crosswalk analysis or stakeholder comments, EPA evaluates whether it constitutes a potential new category subject to identification in the plan or whether it is properly considered a potential new subcategory of an existing point source category. To make this determination, EPA generally looks at whether the industry produces a similar product or performs a similar service as an existing category. If so, EPA generally considers the industry to be a potential new subcategory of that category. If, however, the industry is significantly different from existing categories in terms of products or services provided, EPA considers the industry as a potential new stand-alone category subject to identification in the plan. 4.4.1 Direct Discharges Because the CWA has different requirements for potential new categories of direct and indirect dischargers, EPA examines potential new categories to determine if the category comprises mostly indirect dischargers or if it comprises both direct and indirect dischargers. If a category consists largely of indirect dischargers, EPA evaluates the pass-through and interference potential of the category discharges (see Section 3.4 of the Technical Support Document for the Preliminary 2010 Effluent Guidelines Program Plan (U. S. EPA, 2009)). If a category consists largely of direct dischargers, EPA evaluates the type of pollutants discharged by facilities in the category. 4.4.2 Indirect Discharges For potential new categories with primarily indirect dischargers, EPA evaluates the potential for the wastewater discharges to "interfere with, pass through, or [be] otherwise incompatible with" the operation of POTWs. See 33 U.S.C. § 1371(b)(l). Using available data, EPA reviews the types of pollutants in an industry's wastewater. Then, EPA reviews the likelihood of those pollutants to pass through a POTW. For most categories, EPA evaluated the "pass through potential" as measured by: (1) the total annual TWPE discharged by the industrial sector; and (2) the average TWPE discharge among facilities that discharge to POTWs. EPA also assesses the interference potential of the discharge. Finally, EPA considers whether the pollutant discharges are already adequately controlled by general pretreatment standards and/or local pretreatment limits. 4.5 Identification of Point Source Category References 1. Office of Management and Budget (OMB). 1987. Standard Industrial Classification Manual. (Unknown). EPA-HQ-TRI-2008-0564-0070. 19 For additional information on "the crosswalk," see Section 4.0 of the 2009 Technical Support Document for the Annual Review of Existing Effluent Guidelines and Identification of Potential New Point Source Categories (U.S. EPA, 2009). 4-21 ------- Section 4 - Identification of Point Source Categories 2. U.S. Census Bureau. 2009. North American Industry Classification System (NAICS) Introduction. (Unknown). EPA-HQ-OW-2007-0571 DCN 06730. 3. U. S. EPA. 2005. 2005 Annual Screening-Level Analysis: Supporting the Annual Review of Existing Effluent Limitations Guidelines and Standards and Identification of New Point Source Categories for Effluent Limitations and Standards. EPA-821 -B-05-003. Washington, DC. (August). EPA-HQ-OW-2004-0032-0901. 4. U. S. EPA. 2006a. Final Report: Pulp, Paper, andPaperboard. EPA-821R-06-016. Washington, DC. (November). EPA-HQ-OW-2004-0032-2249. 5. U.S. EPA. 2006b. Technical Support Document for the 2006 Effluent Guidelines Program Plan. EPA-82l-R-06-018. Washington, DC. (December). EPA-HQ-OW-2004- 0032-2782. 6. U. S. EPA. 2009. Preliminary 2010 Effluent Guidelines Program Plan Technical Support Document. Washington, D.C. EPA-82l-R-09-009. (October). EPA-HQ-OW-2008-0517 DCN 06703. 7. Wolford, Jessica. Eastern Research Group, Inc. 2005. Memorandum to 2006 Effluent Guidelines Program Plan Docket. RE: Identification of Facilities for the Porcelain Enameling Point Source Category. (August). EPA-HQ-OW-2004-0032-0945. 4-22 ------- Section 5 - Toxic Weighting Factors 5. Toxic WEIGHTING FACTORS DMRLoads2007 and TRIReleases2007 provide chemical discharge information in the form of mass loads. In order to rank the human health and environmental hazard potential of these loads, EPA estimates toxic-equivalent mass discharges using toxic weighting factors (TWFs). EPA's Engineering and Analysis Division (EAD) developed TWFs for use in its effluent limitations guidelines and standards (ELGs) development program to allow comparison of pollutants with varying toxicities. The toxic-weighted pound equivalent (TWPE) is the mass of a pollutant or chemical discharged that accounts for its relative toxicity. EPA calculates TWPE by multiplying the mass (in pounds) of the chemical by its TWF. The remainder of this section is divided into the following subsections: • Section 5.1 - TWF background and development; • Section 5.2 - New TWFs developed during the 2009 Annual Review; • Section 5.3 - Chemicals for which EPA has not developed TWFs; and • Section 5.4 - TWF References. 5.1 TWF Background and Development In developing ELGs, EPA developed a wide variety of tools and methodologies to evaluate effluent discharges. EPA's Office of Water, EAD maintains a Toxics Database compiled from over 100 references for more than 1,900 pollutants. The Toxics Database includes aquatic life and human health toxicity data, as well as physical and chemical property data. The pollutants in this database are identified by a unique Chemical Abstract Service (CAS) number. EPA calculates TWFs from these data to account for differences in toxicity across pollutants and to provide the means to compare mass loadings of different pollutants. In its analyses, EPA multiplies a mass loading of a pollutant in pounds per year (Ib/yr) by a pollutant-specific weighting factor to derive a "toxic-equivalent" loading (Ib-equivalent/yr). Throughout this document, the toxic-equivalent is also referred to as toxic-weighted pound equivalents, or TWPE. The development of TWFs is discussed in detail in the Draft and Final TWF Development Documents (U.S. EPA, 2005; U.S. EPA, 2006). EPA derives TWFs from chronic aquatic life criteria (or toxic effect levels) and human health criteria (or toxic effect levels) established for the consumption offish. In the TWF method for assessing water-based effects, these aquatic life and human health toxicity levels are compared to a benchmark value that represents the toxicity level of a specified pollutant. EPA selected copper, a metal commonly detected and removed from industrial effluent, as the benchmark pollutant. The Final TWF Development Document contains details on how EPA developed its TWFs (U.S. EPA, 2006). Table D-l in Appendix D lists the TWFs for those chemicals in the DMRLoads2007 and TRIReleases2007 databases for which EPA has developed TWFs. 5.2 New Toxic Weighting Factors Developed During the 2009 Annual Review During the 2009 annual review, EPA revised the TWF for boron to reflect updated information. EPA did not revise any other TWFs or develop TWFs for any chemicals that had not previously had TWFs as part of the 2009 annual review (Abt, 2008). Table 5-1 lists the revised boron TWF. Boron is reported in both DMRLoads2007 and TRIReleases2007'. ------- Section 5 - Toxic Weighting Factors Table 5-1. Revised Boron TWF Pollutant Boron CAS Number 7440428 Old TWF 0.177 New TWF 0.0083 Source: Memorandum to Josh Hall, U.S. EPA. Subject: Revised Draft - Updating the Boron TWF (Abt, 2008). 5.3 Chemicals without Toxic Weighting Factors BAD has not yet developed TWFs for all chemicals in the DMRLoads2007 and TRIReleases2007 databases. Table 5-2 lists the 29 chemicals in TRIReleases2007 that do not have TWFs. The total discharge of the chemicals in Table 5-2 for TRIReleases2007 is 17,100,000 pounds. Table 5-3 lists the chemicals inDMRLoads2007that do not have TWFs. The total discharge of the chemicals in Table 5-3 for DMRLoads2007 is 9.52 billion pounds. Of these discharges, 3 percent relate to nitrogen- and phosphorous- containing compounds that may act as nutrients. TWFs are not good indicators of the impact of nutrients on water quality. While nutrients may have toxic effects that can be reflected in TWFs, their more important effect on water quality occurs through their promotion of eutrophication20. EPA conducted a screening- level analysis of nutrient discharges, which ranked point source categories based on DMRLoads2007 nitrogen and phosphorous compound loads. The results of this analysis are presented in Section 3.2.5. Table 5-2. Chemicals with no TWFs in TRIReleases2007 CAS Number 872504 N503 7782414 N120 306832 191242 1344281 75456 149304 2837890 554132 94360 N583 64755 28407376 Chemical Name N-METHYL-2-PYRROLIDONE NICOTINE AND SALTS FLUORINE DIISOCYANATES 2,2-DICHLORO- 1,1,1 -TRIFLUOROETHANE BENZO(G,H,I)PERYLENE ALUMINUM OXIDE (FIBROUS FORMS) CHLORODIFLUOROMETHANE 2-MERCAPTOBENZOTHIAZOLE 2-CHLORO- 1,1,1 ,2-TETRAFLUOROETHANE LITHIUM CARBONATE BENZOYL PEROXIDE POLYCHLORINATED ALKANES TETRACYCLINE HYDROCHLORIDE C.I. DIRECT BLUE 2 18 Total Pounds Released a 13,999,796 2,818,643 97,777 38,774 37,940 34,819 34,495 33,565 20,573 17,219 11,444 2,996 2,705 804 302 Eutrophication occurs when nitrogen, phosphorous, and other nutrients in a body of water stimulate the growth of algae. Nutrients flow through ecosystems constantly and eutrophication is a natural process that gradually turns ponds into wetlands and wetlands into meadows. However, when human activity introduces additional nutrients to the natural system, algal growth can become extreme and overwhelm the ecosystem's capacity. This over- fertilization can cause increased turbidity, nuisance, or toxic, algal blooms, changes in biota, and anoxia. All of these effects reduce the level and value of ecosystem services provided by water bodies. 5-2 ------- Section 5 - Toxic Weighting Factors Table 5-2. Chemicals with no TWFs in TRIReleases2007 CAS Number 422560 924425 79947 764410 612839 71751412 354143 26628228 26471625 75683 1928434 98884 7664939 7647010 Chemical Name 3 ,3 -DICHLORO- 1,1,1 ,2,2-PENTAFLUOROPROPANE N-METHYLOLACRYL AMIDE TETRABROMOBISPHENOL A 1 ,4-DICHLORO-2-BUTENE 3 ,3 '-DICHLOROBENZIDINE DIHYDROCHLORIDE ABAMECTIN 1 , 1 ,2,2-TETRACHLORO- 1 -FLUOROETHANE SODIUM AZIDE TOLUENE DIISOCYANATE (MIXED ISOMERS) 1 -CHLORO- 1 , 1 -DIFLUOROETH ANE 2,4-D 2-ETHYLHEXYL ESTER BENZOYL CHLORIDE SULFURIC ACID (1994 AND AFTER "ACID AEROSOLS" ONLY) HYDROCHLORIC ACID (1995 AND AFTER "ACID AEROSOLS" ONLY) Total Total Pounds Released a 239 158 23 21 9 7 5 5 0.4 0.02 0.0002 0 0 0 17,100,000 Source: TRIReleases2007_v2. a Includes transfers to POTWs and accounts for POTW removals. Table 5-3. Chemicals with no TWFs in DMRLoads2007 CAS Number 00900 TSS 00515 BODS 79855 78470 00340 78477 51503 81017 00300 00335 CARBN 80103 CBOD 00181 PHOSP PRAM Code 471341 7727379 10043524 7782447 7440440 Chemical Name Hardness, total (as CaCO3) Total Suspended Solids Residue, totfltrble (dried at 105 C) BOD, 5-day Adsorbable organic halides (AOX) Nitrogen, sludge, tot, dry wt. (as N) Oxygen demand, chem. (high level) (COD) Solids, sludge, tot, dry weight Calcium Chloride Chemical Oxygen Demand (COD) Oxygen, dissolved (DO) Oxygen demand, chem. (low level) (COD) Total Carbon Chemical oxygen demand (COD) Carbonaceous BOD, 5-day Oxygen demand, ultimate Phosphorus Total Pounds Released 3,043,209,486 3,039,742,485 1,069,455,416 384,518,697 281,109,733 270,173,870 269,728,827 212,062,831 173,744,369 158,007,360 138,838,601 134,291,135 115,774,628 36,770,720 30,325,446 24,249,650 21,673,603 5-3 ------- Section 5 - Toxic Weighting Factors Table 5-3. Chemicals with no TWFs in DMRLoads2007 CAS Number 03594 TKN 00410 SIO2 00440 46570 80108 78240 00341 NOX 00343 34044 ORGN 80087 32017 51450 78115 00640 00319 TTC1A 51404 71872 34045 78733 70353 PO4 PO4 82209 HC 00415 51360 39942 78157 U238 00551 03773 PO4ASP 61194 51521 PRAM Code 7727379 471341 7631869 71523 7727379 7727379 7647145 7727379 13863417 14265442 308067530 98486 63231516 1338245 7440611 14265442 335671 Chemical Name Halogens, adsorbable organic Total Kjeldahl Nitrogen Alkalinity, total (as CaCO3) Silica Bicarbonate ion- (as HCO3) Hardness, Ca Mg Calculated (mg/L as CaCO3) Chemical oxygen demand (COD) Metals, total Oxygen demand, chem. (COD), dissolved Nitrogen, oxidized Oxygen demand, total (tod) Oxidants, total residual Nitrogen, organic BOD, carbonaceous, 20 day, 20 C Sodium chloride (salt) Nitrite Plus Nitrate Total Halogen, total organic Nitrogen, inorganic total BOD, (ult. all stages) Static 4Day Chronic Selen. Capricornutum Solids, total suspd. non-volatile Bromine chloride Oxidants, free available Volatile fraction organics (EPA 624) Organic halides, total Phosphate Phosphate Chlorides & sulfates Total Hydrocarbons Alkalinity, phenolphthaline method m-Benzenedisulfonic acid Hydrocarbons, aromatic Naphthenic acid Uranium 23 8 Hydrocarbons, in H2O, IR, CC14 extractible chromatograph Chlorine produced oxidants Phosphate as P Halogen, total residual Perfluorooctanoic Acid Total Pounds Released 11,634,478 7,890,284 7,490,718 5,617,674 5,265,556 5,048,590 4,916,094 4,099,529 3,137,358 3,027,222 2,893,253 1,802,491 1,456,316 1,448,164 1,077,614 854,958 557,377 343,917 330,952 318,731 271,229 259,617 220,001 189,068 185,054 173,386 173,386 165,812 111,168 108,716 107,605 106,144 93,603 78,333 66,577 60,594 58,401 56,810 46,552 5-4 ------- Section 5 - Toxic Weighting Factors Table 5-3. Chemicals with no TWFs in DMRLoads2007 CAS Number 45501 CFA 04370 FLORB 77517 51340 82214 31667 80279 78218 80996 80126 00740 71845 78239 72035 00314 71870 50008 H2O2 49922 51065 04251 00664 78724 51526 03604 22456 47021 51523 00141 77066 51522 82180 01210 78221 51525 85789 DMDS PRAM Code 479618 16872110 98113 98679 8002059 999 14265453 14798039 24959679 7722841 3825261 41663847 999 130498292 497267 7440053 754916 1563388 624920 Chemical Name Petrol hydrocarbons, total recoverable Chlorophyll A Sum BOD and ammonia, water Fluoroborates Benzenesulphonic acid p-Phenolsulfonic acid pH change (range) Oil petroleum, total recoverable CBOD5/NH3-N Phenolic compounds, unchlorinated Spray irrigation BOD, carbonaceous, 5 day, 5 C Sulfite (as SO3) Nitrogen, ammonia total (as NH4) Metals, tox priority pollutants, total Pump hours BOD, nitrogen inhib 5-day (20 deg. C) Bromide (as Br) Priority pollutants total effluent Hydrogen peroxide Diesel range organics diesel, total, wtr Ammonium perfluoroctanoate Clamtrol CT-1, Total Water a Dock discharge of phosphorus b 4-Nitro-N-methylphthalimide, total Perfluorooctanesulfonate Total phenols Polynuc aromatic HC per Method 610 Methylene blue active substances Perfluorobutanoicsulfonate Solids, total susp per production 2-Methyl- 1 ,3 -dioxolane Perfluorobutanoic Acid Hydrocarbons, petroleum Palladium, total (as Pd) Organic pesticide chemicals (40 CFR455) Perfluorooctanesulfonamide 2,2-Dimethyl-2,3-dihydro-7-benzofuranol Dimethyl disulfide Total Pounds Released 40,436 33,343 32,830 30,938 30,034 30,034 29,986 23,635 20,650 18,149 18,004 15,707 14,980 12,154 11,888 11,177 9,199 8,510 7,946 5,713 3,556 3,129 2,600 2,267 2,057 1,867 1,784 1,599 1,475 1,426 1,287 1,224 1,185 1,082 842 809 792 689 649 5-5 ------- Section 5 - Toxic Weighting Factors Table 5-3. Chemicals with no TWFs in DMRLoads2007 CAS Number 01277 01142 39117 51493 00988 49875 HCCB 82560 78456 51524 CLPHN 00144 74052 34283 78155 81559 85795 39084 84085 U308 77247 85812 49491 00741 74053 77102 34521 ABS 82195 51051 00696 77226 77672 51437 51438 51165 78143 81512 77542 PRAM Code 7440213 999 5131668 27154445 1336352 39638329 30498352 683534 7440611 2809214 14265453 872504 191242 42615292 108678 120616 211578040 88164 95169 87854 Chemical Name Total agg concentration #1 Silicon, total Phthalate esters Phenolic Compounds, Total Iron and manganese, soluble Propylene glycol monobutyl ether Hexachlorocyclohexane Total pesticides Halomethanes, sum Perfluorobutanesulfonamide Chlorinated phenols Combined metals sum Chlorinated hydrocarbons, general Bis(2-chloroisopropyl) ether Dichlorobenzyl trifluoride Bromodichloroethane Xylene, meta & para in combination Total purgeable halocarbons Volatile organics detected Uranium 308 Benzoic acids, total 1 -Hy droxy ethy lidene BTEX Sulfite (as S) Pesticides, general N-Methyl-2-pyrrolidone Benzo(ghi)perylene Alkyl benzene sulfonates Thiocarbamates Tin, tri-organo- Nitrofurans 1 ,3 ,5-Trimethylbenzene Dimethyl terephthalate N-Hexane SAS-310, Total SAS - 305, total Monochlorobenzyl trifluoride Benzothiazole Hexamethylbenzene Total Pounds Released 624 588 520 507 497 496 484 465 374 321 313 282 276 223 157 155 137 122 96 77 74 60 41 37 35 32 24 24 16 16 13 12 12 11 11 8 6 5 4 5-6 ------- Section 5 - Toxic Weighting Factors Table 5-3. Chemicals with no TWFs in DMRLoads2007 CAS Number 51202 71910 01168 34102 45097 78721 81611 73525 49702 38579 70027 39379 84103 77086 51009 82181 01279 76025 34679 00143 00987 00973 01289 00177 01278 01288 00148 78232 77625 77666 77676 77889 77983 78028 61916 78222 77081 78237 78732 PRAM Code 7440575 7440746 628966 98839 26523648 1338234 131748 108996 136677093 74931 1318098 103333 77929 30583336 706785 29797408 12408105 497187 144627 Chemical Name Sulfide-hydrogen sulfide (undissociated) Gold, total (as Au) Indium Ethylene glycol dinitrate Methylstyrene Phthalates, total Trichlorotrifluoroethane 2-Butanone peroxide Ammonium picrate Benzene, halogenated COD, 25N K2Cr207, tot DDT/DDD/DDE, sum of p,p' & o,p' isomers Dioxin laboratory - alpha code 3 -Methy Ipy ridine RDX+HMX Hydrocarbons, total petroleum Total agg concentration #3 Chlorinated dibenzo-p-dioxins, effluent 2,3,7,8-TCDD TEC Methyl mercaptan Iron and manganese, total Asbestos, total amphibole Biocides Oxygen demand, dissolved Total agg concentration #2 Foaming agents Herbicides, total Total toxic organics (TTO) (40 CFR469) Azobenzene Citric acid Trichlorotoluene Octachlorocyclopentene Dichlorotoluene Tetrachlorobenzene 1,3-Diaminourea Organic active ingredients (40 CFR455) Oxalic acid Organics, volatile (NJAC reg. 7:23-17e) Volatile compounds, (GC/MS) Total Pounds Released 4 4 4 4 3 2 2 2 1 1 0.3 0.2 0.1 0.1 0.004 0.001 0.0003 0.0001 0.000000004 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 5-7 ------- Section 5 - Toxic Weighting Factors Table 5-3. Chemicals with no TWFs in DMRLoads2007 CAS Number 79817 81328 82080 82602 DDAC 78171 51497 32015 34103 34730 38925 45670 49699 49886 77540 51132 77295 51539 51540 61026 70015 76028 76029 03768 51030 PRAM Code 95772 25323302 7173515 576249 13560899 84764 193700059 583788 108805 108430 4901513 Chemical Name 3 ,4-Dichlorophenol Dichloroethene, total Trihalomethane, tot. Produced sand, weight CalgonH-130M Aromatics, total purgeable Spectrus OX 1200 Base/neutral compounds Benzene, toluene, xylene in combination 2,3-Dichlorophenol, total Dechlorane plus Dinonyl phthalate Betz slimicide C-31, total Betz clam-trol CT-2 2,5-Dichlorophenol Cyanuric acid 3-Chlorophenol Nonpurgeable Organic Halides Purgeable Organic Halides 2,3,4,5-Tetrachlorophenol Freon, total Base neutrals & acid (Method 625), efflnt Organics, tot purgeables (Method 624) Purgeable hydrocarbons, Meth. 601 Spectrus CT 1300 Total Total Pounds Released 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 9,520,000,000 Source: DMRLoads2007_v3. aFrom the ZM Control Guide (Sprecher, 2000), Clam-Ttol CT-1 is a liquid substance that is 8 percent n-alkyl (C12- 40 percent, C14-50 percent, C16-10 percent) dimethylbenzyl ammonium chloride and 5 percent dodecylguanidine hydrochloride. b Dock discharge of phosphorous is required for facilities that operate a ship dock used primarily for loading and unloading solids containing some compound of phosphorus (e.g., phosphate rock, ammoniated phosphates) (State of Louisiana, 2004). TEC - Total equivalent concentration. 5.4 Toxic Weighting Factor References 1. Abt Associates Inc. 2008. Memorandum to Josh Hall, U.S. EPA. RE: Revised Draft - Updating the Boron TWF. Cambridge, MA. (December 5). EPA-HQ-OW-2008-0517 DCN 06729. 5-8 ------- Section 5 - Toxic Weighting Factors 2. Sprecher, Susan, Kurt Getsinger. 2000. Zebra Mussel Chemical Control Guide. ERDC/EL TR-00-1. U.S. Army Engineer Research and Development Center. Vicksburg, MS. (January). EPA-HQ-OW-2004-0032-2171. 3. State of Louisiana Department of Environmental Quality (LA DEQ). 2004. NPDES Permit for IMC Phosphates Co, Faustina Plant. (May 24). EPA-HQ-OW-2004-0032-1134 and 1135. 4. U.S. EPA. 2005. Draft Toxic Weighting Factor Development in Support ofCWA 304(m) Planning Process. Washington, DC. (June). EPA-HQ-OW-2004-0032-0857. 5. U.S. EPA. 2006. Toxic Weighting Factor Development in Support ofCWA 304(m) Planning Process. Washington, DC. (June). EPA-HQ-OW-2004-0032-1634. 5-9 ------- Section 6 - Quality Review 6. QUALITY REVIEW EPA's screening-level analysis involves the collection and use of existing environmental data for purposes other than those for which they were originally collected. Pollutant Compliance System (PCS) and Integrated Compliance Information System for the National Pollutant Discharge Elimination System (ICIS-NPDES) were designed to automate entering, updating, and retrieving NPDES data and to track permit issuance, permit limits and monitoring data, and other data pertaining to facilities regulated under NPDES. The primary purpose of the Toxic Release Inventory (TRI) is to collect and make public annual data on releases and transfers of certain toxic chemicals from industrial facilities to inform communities and citizens of chemical hazards in their areas. Sections 2 and 3 of this report describe how EPA used the data in PCS, ICIS- NPDES, and TRI to calculate annual pollutant loadings to prioritize industrial category discharges for further review. To use data in PCS and ICIS-NPDES, EPA first combined the two datasets to form DMRLoads2007', as described in Section 3. This section describes the quality review steps that EPA uses to determine if the DMRLoads2007 and TRIReleases2007 data are suitable for EPA's use in a screening-level analysis. The remainder of this section is divided into the following subsections: • Section 6.1- Overview of Quality Review Steps; • Section 6.2 - Summary of DMRLoads2007 Quality Review; • Section 6.3 - Summary of TRIReleases2007 Quality Review; and • Section 6.4 - Quality Review References. 6.1 Overview of Quality Review Steps EPA considered the following factors in its quality review of the PCS, ICIS-NPDES, and TRI data: • Completeness. The following information is needed to analyze the toxic weighted pollutant loadings discharged by industrial categories: — Facility identity, — Industrial category under which the facility is regulated, — Identity of parameters discharged and corresponding toxic weighting factors (TWFs), — Mass of pollutants discharged (or pollutant concentration and discharge flow, from which the mass can be calculated), and — Understanding of how available information represents the discharger population and pollutant population. • Accuracy. Analyzed data should accurately categorize and aggregate the underlying database. • Reasonableness. Pollutant identities must be reasonably related to the operations in the category. Reported or calculated loads and facility wastewater flows should reflect the range of flows and loads known to exist in the United States. The following subsections discuss each of these factors in more detail. ------- Section 6 - Quality Review 6.1.1 Completeness Checks In previous years of review, EPA compared the number of facilities listed in the 2007 U.S. Economic Census to the number of facilities in the PCS and TRI databases, as described in the report entitled 2005 Annual Screening-Level Analysis: Supporting the Annual Review of Existing Effluent Limitations Guidelines and Standards and Identification of New Point Source Categories for Effluent Limitations Guidelines and Standards to determine the extent to which the facilities in the databases represent the entire industry (U.S. EPA, 2005). In 2009 for categories selected for preliminary category review21, EPA compared the following statistics in DMRLoads2007 and TRIReleases2007 to the 2002 U.S. Economic Census: the total number of facilities, the number of facilities reporting wastewater discharges (direct or indirect) in TRI, and the number of major and minor facilities in DMR. Table 6-1 lists EPA's findings for the 2007 databases. Table 6-1. Number of Facilities in Categories Selected for Preliminary Category Review Point Source Category Fertilizer Manufacturing (40 CFR Part 4 18) Inorganic Chemicals Manufacturing (40 CFR Part 4 15) Nonferrous Metals Manufacturing (40 CFR Part 421) Ore Mining and Dressing (40 CFR Part 440) Organic Chemicals, Plastics and Synthetic Fibers (40 CFR Part 414) Petroleum Refining (40 CFR Part 419) Pulp, Paper and Paperboard (40 CFR Part 430) NAICS Codes 311225FER, 325312, 325311, 325314 325120, 325131, 325181, 325188, 325998INORG, 331311, 325510INORG 325188NMM, 331312, 331314, 331411, 331419, 331423, 331492, 331521 21220, 212234, 212231, 212221, 212222, 212291, 212299, 213114 311999OCPSF, 324199OCPSF, 325110, 325120OCPSF, 325132, 325188OCPSF, 325192, 325193, 325199, 325211, 325221, 325222, 3255 10OCPSF, 325520, 32561 1OCPSF, 325612, 325620, 325998, 326199OCPSF, 339999OCPSF, 424690, 562920 324110,324191,324199, 325998PR, 474710, 486110 321113-1, 322110, 322121, 322122, 322130, 322211, 322212, 322214, 322215, 322221, 322222, 322224, 322231,322291,322299 2002 U.S. Economic Census >723 >1,335 >937 510 >17,125 >5,785 >4,876 2007 DMRa 85 394 114 57 903 1,393 448 2007 TRI b 110 414 338 76 2,032 780 464 Source: U.S. Economic Census, 2002 (U.S. Census, 2002), TRIReleases2007_v2, DMRLoads2007_v3. a - Major and minor dischargers. Also, DMR data is reported by SIC code; therefore EPA used an NAICS to SIC crosswalk for comparison purposes. b - Releases to any media. See Section 5.3 of the Technical Support Document for the Preliminary 2010 Effluent Guidelines Program Plan (U.S. EPA, 2009) for information on how the categories were selected for preliminary category review. 6-2 ------- Section 6 - Quality Review EPA also considered the pollutant discharges that do not contribute to the category rankings. As discussed in Section 5, EPA identified and profiled the pollutant parameters that do not have an assigned TWF. Table D-l in Appendix D lists the TWFs for those chemicals in the DMRLoads2007 and TRIReleases2007 databases for which EPA has developed TWFs. Table 5-2 and Table 5-3 list the chemicals in the databases that do not have assigned TWFs, as well as the total pounds of these pollutants estimated as discharged. This quality review showed that 52.6% of 18.1 billion pounds of pollutants discharged in DMRLoads2007 are not currently assigned TWF, while 22.2% of 195 million pounds of pollutants released in TRIReleases2007 are not currently assigned a TWF. 6.1.2 Accuracy Checks EPA verified the accuracy of database queries used to analyze DMRLoads2007 and TRIReleases2007 data and generate output tables. As one team member created queries, a second team member reviewed the logic of the programming code, and compared the number of records in the output table to the number of records in intermediate queries. This ensured that no data were missing and that there were no duplicate records. EPA documented the quality checks in a database table that describes the function of each query created, the quality checks that were performed, the name of the reviewer, the date the query was reviewed, and any errors that were identified. Tables A-5 in Appendix A and B-15 in Appendix B present the quality check tables for the TRIReleases2007 and DMRLoads2007 databases. 6.1.3 Reasonableness Checks EPA ranked pollutant discharges and facilities by toxic weighted loadings to identify discharges and loadings that are unusually high. EPA then conducted reasonableness checks on the unusually high pollutant discharges and facility loads to determine if the unusual values were misreported or miscalculated. The reasonableness checks are described in the following subsections. 6.1.3.1 Pollutant Identity EPA ranked the pollutants discharged from each point source category and, using engineering understanding of industrial processes, verified that the pollutants composing the majority of the load could be reasonably related to operations in the industry. For unexpected results, EPA compared the reported releases to information in the facility's NPDES permit and other available resources, such as facility descriptions and discussions with the facility contacts. EPA corrected errors in DMRLoads2007 and TRIReleases2007 and documented the corrections. For example, in the quality review of the TRIReleases2007 database, EPA identified a petroleum refining facility that was reporting dioxin discharges that resulted in a large discharge estimate, in terms of toxic weighted pound-equivalents (TWPE). EPA contacted the facility to verify that the estimated discharge was based on actual measured dioxin, instead of measurements below detection limits, because facilities often use non-detect values when estimating dioxin discharges. The facility verified that the reported discharges were actually an overestimate of their actual dioxin discharge, which follows EPA's guidance on TRI reporting22. This method is The Office of Pollution Prevention and Toxic Substances provides guidance on how to report dioxin to TRI in the document entitled Guidance for Reporting Toxic Chemicals within the Dioxin and Dioxin-like Compounds Category (http://www.epa.gov/tri/guide docs/index.htm#chemicalsp). "For purposes of threshold determinations and the 6-3 ------- Section 6 - Quality Review appropriate for TRI reporting purposes, but for the screening-level databases, EPA adjusted the estimated dioxin discharge to represent the actual dioxin measured in wastewater. 6.1.3.2 Facility Loads EPA reviewed the toxic weighted loadings of facilities to ensure that they compose a reasonable percentage of the total national discharge. Facilities that compose a very high percentage of the national discharge have a large impact on the point source category rankings. EPA reviewed NPDES permit data or other available data to identify where a facility may have made a calculation error or reported the incorrect units of measure, and contacted facilities to confirm suspected errors. EPA corrected confirmed errors and documented the corrections. For example, in the quality review of the DMRLoads2007 database, EPA identified a facility whose calculated TWPE for dioxin was over a billion pound-equivalents. EPA contacted the facility's regulatory authority and learned there were units errors as well as misinterpreted laboratory data. The units error caused DMRLoads2007 to overestimate the dioxin load by a factor of 162 (Auchterlonie, 2009). 6.2 Quality Review of the DMRLoads2007 Database As discussed in Section 3, to identify potential anomalous loads, EPA ranked DMRLoads2007 facilities by total TWPE. For those facilities that ranked the highest for total TWPE, EPA reviewed them carefully to verify the accuracy of the database. The DMRLoads2007 review included the following tasks: • Comparison of DMRLoads2007 to PCSLoads2004; • Comparison of DMRLoads2007 loads to TRIReleases2007; • Review of flow and concentration data for units errors; • Review of reported discharge data and the estimated load for missing data; • Review of permit limits; • Verification of proper SIC code/point source category classification; • Review of NPDES permit or fact sheet where available; and • Discussion with facility contacts. These steps were taken for each facility that seemed to have an unusually high TWPE. Once a possible mistake was identified through the process listed above, EPA contacted the facilities for verification of changes made to the database. Table B-13 in Appendix B presents the resulting corrections identified. reporting of releases and other waste management quantities for dioxin and dioxin-like compounds under EPCRA Section 313, either with monitoring data, or by using the emission factor approach, non-detects are treated as 'zero' if that is how the method being used treats non-detects (e.g., Method 1613, Method 23). However, facilities should use their best readily available information to report, so if a facility has better information than provided by these methods then that information should be used. For example, if a facility is not detecting dioxin or a particular dioxin-like compound using a particular method but has information that shows that they should be detecting them the facility should use this other information and it may be appropriate to estimate quantities using one half the detection limit." This guidance results in many facilities using one-half the detection limit to estimate discharges for years where no dioxin were detected in wastewater. 6-4 ------- Section 6 - Quality Review 6.3 Quality Review of the TRIReleases2007 Database EPA ranked TRI facilities by total TWPE released to surface waters to identify potential anomalous loads. The review of TRIReleases2007 included the following: • Comparison of TRIReleases2007 loads to TRIReleases2006; • Comparison of TRIReleases2007 loads to DMRLoads2007; • Review of reported discharge data and the estimated load for missing data; • Review of the basis of estimate used for reporting the pollutant discharge; • Review of reported dioxin congener distributions; • Verification of proper NAICS code/point source category classification; • Discussions with facility contacts. This review process was carried out for each facility that ranked among the highest for total TWPE released to surface waters or transferred to POTWs in 2007. Comparing databases and publically available discharge information made it possible to identify potential errors in the database, which would result in a high TWPE for a facility. Facilities were contacted to verify that the correct change to the data was taking place. Table A-3 in Appendix A presents the resulting corrections identified. 6.4 Quality Review References 1. Auchterlonie, Steve. 2009. Notes from Telephone Conversation between Chris Krejci, Eastern Research Group, Inc. and Steve Auchterlonie, Front St. Remedial Action. RE: Verification of Magnitude and Basis of Estimate for Dioxin and Dioxin-Like Compound Discharges in PCS. (March 13). EPA-HQ-OW-2008-517 DCN 06636. 2. U.S. Economic Census. 2002. Available online at: http://www.census.gov/econ/census02. 3. U. S. EPA. 2005. 2005 Annual Screening-Level Analysis: Supporting the Annual Review of Existing Effluent Limitations Guidelines and Standards and Identification of Potential New Categories for Effluent Limitations Guidelines and Standards. EPA-821-B-05-003. Washington, DC. (August). EPA-HQ-OW-2004-0032-0901. 6-5 ------- Section 7 - Results of 2009 Screening-Level Analysis 7. RESULTS OF 2009 SCREENING-LEVEL ANALYSIS This section describes the results of the 2009 screening-level analysis and the methodology used by EPA to prioritize categories for further review. This section also discusses the identification of categories warranting detailed studies. The remainder of this section is divided into the following subsections: • Section 7.1 - Preliminary Results of the Screening-Level Review; • Section 7.2 - Prioritization of Categories; and • Section 7.3 - Identification of Categories for Further Review. 7.1 Preliminary Results of the Screening-Level Review The purpose of the screening-level review is to evaluate the amount and toxicity of the pollutants in an industrial category's discharges. Using TRIReleases2007 and DMRLoads2007, EPA ranked point source categories according to their discharges of toxic and non-conventional pollutants (reported in units of toxic-weighted pound equivalent or TWPE). As described earlier in this report, EPA multiplied the pounds of pollutants discharged by toxic weighting factors (TWFs) resulting in an estimate of TWPE. Discharges were assigned to industrial categories on the basis of facility Standard Industrial Classification (SIC) and North American Industrial T\ Classification System (NAICS) codes . Categories included both facilities subject to the existing effluent guidelines for the category and those belonging to potential new subcategories of existing effluent guidelines and to potential new categories. Table 7-1 and Table 7-2 present, for categories for which EPA has promulgated effluent guidelines and pretreatment standards (ELGs), the preliminary rankings using TRIReleases2007 and DMRLoads2007, respectively. Discharges from facilities that produce chlorine or chlorinated hydrocarbons (CCH) are listed on these tables as a separate category. See Section 7.2.1 for further discussion. Table 7-1 and Table 7-2 include discharges associated with facilities subject to the point source category applicability, as well as facilities that are associated with potential new subcategories of existing categories. Table 7-1. TRIReleases2007Point Source Category Rankings Rank 1 2 3 4 5 6 7 40CFR Part 414.1 414 423 430 419 420 433 Point Source Category Chlorine And Chlorinated Hydrocarbons Organic Chemicals, Plastics And Synthetic Fibers Steam Electric Power Generating Pulp, Paper And Paperboard Petroleum Refining Iron And Steel Manufacturing Metal Finishing a TWPE 7,270,000 575,000 542,000 460,000 172,000 104,000 62,000 23 DMR data from PCS and ICIS-NPDES in the DMRLoads2007 has facility SIC codes, while TRI data has NAICS codes. See Section 5 - Identification of Point Source Categories for additional information on how EPA linked SIC and NAICS codes to point source categories. 7-6 ------- Section 7 - Results of 2009 Screening-Level Analysis Table 7-1. TRIReleases2007Point Source Category Rankings Rank 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 40CFR Part 415 440 421 432 458 455 429 417 471 463 439 428 425 469 NA 464 468 NA 418 437 413 407 467 436 405 410 406 461 438 426 434 411 424 422 443 465 408 466 Point Source Category Inorganic Chemicals Manufacturing Ore Mining And Dressing Nonferrous Metals Manufacturing Meat and Poultry Products Carbon Black Manufacturing Pesticide Chemicals Timber Products Processing Soap And Detergent Manufacturing Nonferrous Metals Forming And Metal Powders Plastics Molding And Forming Pharmaceutical Manufacturing Rubber Manufacturing Leather Tanning And Finishing Electrical And Electronic Components Miscellaneous Foods And Beverages Metal Molding And Casting (Foundries) Copper forming Tobacco Products Fertilizer Manufacturing Centralized Waste Treatment Electroplating Canned And Preserved Fruits And Vegetables Processing Aluminum forming Mineral Mining And Processing Dairy products processing Textile Mills Grain mills Battery Manufacturing Metal Products And Machinery Glass Manufacturing Coal Mining Cement Manufacturing Ferroalloy Manufacturing Phosphate Manufacturing Paving And Roofing Materials (Tars And Asphalt) Coil Coating Canned And Preserved Seafood Processing Porcelain Enameling TWPE 54,700 44,400 38,900 35,900 32,400 24,700 16,300 14,600 8,830 8,780 8,000 7,860 7,800 7,550 6,580 6,110 4,950 4,760 4,460 3,790 3,210 2,960 2,710 2,420 2,400 2,390 2,080 1,640 917 546 493 452 340 250 249 241 234 164 7-7 ------- Section 7 - Results of 2009 Screening-Level Analysis Table 7-1. TRIReleases2007Point Source Category Rankings Rank 46 47 48 49 50 51 52 53 54 55 56 40CFR Part 446 NA 445 454 444 NA 409 447 457 406 NA Point Source Category Paint Formulating Printing And Publishing Landfills Gum And Wood Chemicals Manufacturing Waste Combustors Independent And Stand Alone Labs Sugar Processing Ink Formulating Explosives Manufacturing Hospitals Drinking Water Treatment TOTAL TWPE 140 110 83 55 40 30 26 20 14 1 0 9,550,000 Source: TRIReleases2007_v2. NA - Not applicable. These are potential new categories. Table 7-2. DMRLoads2007 Point Source Category Rankings Rank 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 40CFR Part 423 433 430 414.1 418 420 432 414 419 415 421 440 455 NA 471 410 429 417 444 445 Point Source Category Steam Electric Power Generating a Metal Finishing b Pulp, Paper And Paperboard ° Chlorine And Chlorinated Hydrocarbons Fertilizer Manufacturing Iron And Steel Manufacturing Meat and Poultry Products Organic Chemicals, Plastics And Synthetic Fibers d Petroleum Refining Inorganic Chemicals Manufacturing Nonferrous Metals Manufacturing Ore Mining And Dressing Pesticide Chemicals Drinking Water Treatment Nonferrous Metals Forming And Metal Powders Textile Mills Timber Products Processing Soap And Detergent Manufacturing Waste Combustors Landfills Total TWPE 20,400,000 3,360,000 2,730,000 1,220,000 1,100,000 730,000 536,000 413,000 403,000 394,000 343,000 184,000 180,000 119,000 119,000 79,900 51,600 47,800 38,400 35,800 7-8 ------- Section 7 - Results of 2009 Screening-Level Analysis Table 7-2. DMRLoads2007 Point Source Category Rankings Rank 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 40CFR Part 409 436 439 463 422 467 464 428 454 437 469 411 NA NA 424 408 468 434 406 407 443 461 NA 457 412 426 NA NA 435 465 405 460 466 425 451 NA 438 NA Point Source Category Sugar Processing Mineral Mining And Processing Pharmaceutical Manufacturing Plastics Molding And Forming Phosphate Manufacturing Aluminum forming Metal Molding And Casting (Foundries) Rubber Manufacturing Gum And Wood Chemicals Manufacturing Centralized Waste Treatment Electrical And Electronic Components Cement Manufacturing Miscellaneous Foods And Beverages Independent And Stand Alone Labs Ferroalloy Manufacturing Canned And Preserved Seafood Processing Copper forming Coal Mining Grain mills Canned And Preserved Fruits And Vegetables Processing Paving And Roofing Materials (Tars And Asphalt) Battery Manufacturing Printing And Publishing Explosives Manufacturing CAFO Glass Manufacturing Construction And Development Airport Deicing Oil & Gas Extraction Coil Coating Dairy products processing Hospital Porcelain Enameling Leather Tanning And Finishing Concentrated Aquatic Animal Production Tobacco Products Metal Products And Machinery Photo Processing Total TWPE 32,500 26,700 24,900 24,600 18,500 12,200 11,300 11,200 10,500 10,400 9,350 8,960 5,840 5,360 4,350 3,230 2,310 2,290 1,980 1,760 1,280 1,100 999 785 617 353 324 265 256 166 76 15 11 8 5 3 3 1 7-9 ------- Section 7 - Results of 2009 Screening-Level Analysis Table 7-2. DMRLoads2007 Point Source Category Rankings Rank 59 60 40CFR Part 459 442 Point Source Category Photographic Transportation Equipment Cleaning TOTAL Total TWPE 1 0 32,700,000 Source: DMRLoads2007_v3. a EPA corrected a suspected units error in DMRLoads2007_v3 for FB Culley Station in Newburgh, IN (IN0002259) in the Steam Electric Power Generating Category. EPA attempted to contact the facility but the facility never returned calls. Therefore, EPA was unable to verify the correction. bEPA contacted General Electric in Erie, PA (PA0000183) in the Metal Finishing Category and identified a units error in DMRLoads2007_v3 (Verderese, 2009). The new LEY and TWPE reported for this facility were recalculated and are now 0.024 and 2.790, respectively. The new Metal Finishing Category TWPE is 571,500. 0 For the Pulp, Paper, and Paperboard Category, EPA contacted facilities to verify the concentrations of dioxin and dioxin-like compounds in PCS and ICIS-NPDES. EPA found that, for all facilities contacted, there were either units errors (e.g., reported as ng/L but in the database as mg/L) or missing non-detect indicators. The new Pulp, Paper, and Paperboard Category total TWPE is 252,163. See Section 12.2.2.1 in the Technical Support Document for the Preliminary 2010 Effluent Guidelines Program Plan (U.S. EPA, 2009) for additional details on the facilities-specific corrections. dEPA contacted GE Silicones in Friendly, WV (WV0000094), in the OCPSF Category and identified a units error in DMRLoads2007_v3 (Martin, 2009). The new LEY and TWPE reported for this facility were recalculated and are now 158 and 100.3, respectively. The new OCPSF Category total TWPE is 308,721. NA - Not applicable. These are potential new categories. 7.2 Prioritization of Categories For the 2009 screening-level review, EPA combined the results of the TRIReleases2007 and the DMRLoads2007 databases, which are presented in Sections 2 and 3 of this document, respectively. When combining the results of these databases, EPA made adjustments to the rankings for the following: • Discharges from industrial categories for which EPA is currently developing or revising ELGs; • Discharges from point source categories for which EPA has recently promulgated or revised ELGs; and • Discharges from facilities determined not to be representative of their category. Sections 7.2.1 through 7.2.4 discuss the rationale for these decisions. The final combined database rankings represent the results of the 2009 screening-level review and are presented in Section 7.2.5. 7.2.1 Categories for Which EPA is Currently Developing or Revising ELGs EPA is currently considering revisions to ELGs for Organic Chemicals, Pesticides, and Synthetic Fibers (OCPSF) (40 CFR 414) and the Inorganic Chemicals Manufacturing (40 CFR 415) Point Source Categories for facilities that produce Chlorine and Chlorinated Hydrocarbons (CCH). Because the CCH rulemaking is underway, EPA excluded discharges from these facilities from further consideration under the current planning cycle. EPA subtracted the Toxic Weight Pollutant Equivalent (TWPE) loads from facilities that produce chlorine or chlorinated hydrocarbons from the Organic Chemicals, Pesticides, and Synthetic Fibers (OCPSF) and ------- Section 7 - Results of 2009 Screening-Level Analysis Inorganic Chemicals Manufacturing Point Source Category loads. Because facilities that produce chlorine and chlorinated hydrocarbons are only a subset of the OCPSF and Inorganic Chemicals Manufacturing Categories, EPA included loads for all other facilities in these two categories in the prioritization of categories for further review24. 7.2.2 Categories for Which EPA Recently Promulgated or Revised ELGs For the 2009 annual review and development of category rankings, EPA excluded point source categories for which ELGs were recently established or revised but not yet fully implemented, or were recently reviewed in a rulemaking context where EPA decided to withdraw the proposal and select the "no action" option. In general, EPA removes a category from further consideration during a review cycle if EPA established, revised, or reviewed the category's ELGs within seven years prior to the annual review. This seven-year period allows time for the ELGs to be incorporated into NPDES permits. For the 2009 annual review EPA excluded from the development of category rankings any categories with ELGs established, revised, or recently reviewed after August 2002. Table 7-3 lists these categories. Removing a point source category from further consideration in the development of the rankings does not mean that EPA eliminates the category from annual review. In cases where EPA is aware of the growth of a new segment within such category, or where new concerns are identified for previously unevaluated pollutants discharged by facilities in the category, EPA would apply closer scrutiny to the discharges from the category in deciding whether to consider it further during the current review cycle. For example, EPA conducted the detailed study of the coal mining industry based on comments received on the 2006 Preliminary Plan, although the coal mining ELGs were revised in January 2002. Table 7-3. Point Source Categories That Have Undergone a Recent Rulemaking or Review 40 CFR Part Number 122 and 4 12 451 432 413, 433, and 438 420 Point Source Category Concentrated Animal Feeding Operations (CAFOs) Concentrated Aquatic Animal Production (or Aquaculture) Meat and Poultry Products Metal Products and Machinery (including Metal Finishing and Electroplating) Iron and Steel Manufacturing Date of Rulemaking November 20, 2008 August 23, 2004 September 8, 2004 May 13, 2003 October 17, 2002 Source: "Guidelines: Final, Proposed, and Under Development" at http://www.epa.gov/waterscience/guide. 7.2.3 Discharges Not Categorizable EPA identified discharges that are not categorizable into new point source categories or subcategories. In particular, due to the high TWPE discharges EPA reviewed reported discharges from a Superfund site (Auchterlonie, 2009).25 Direct discharges from Superfund sites, whether 24 EPA is also currently revising the concentrated animal feeding operations ELG (Part 412); however, the TWPE associated with this category is low and does not affect the prioritization of categories based on TWPE. For more information on industries currently undergoing rulemakings, see http://www.epa.gov/guide/industry.html. 25 The Comprehensive Environmental Response, Compensation, and Liability Act, commonly known as Superfund, was enacted by Congress on December 11, 1980. 7-11 ------- Section 7 - Results of 2009 Screening-Level Analysis made onsite or offsite, are subject to NPDES permitting requirements (U.S. EPA, 1988a; U.S. EPA, 1988b). For the reasons discussed below EPA determined that these discharges do not represent a point source category and excluded these TWPE from the point source category rankings. EPA identified that discharges from Superfund sites are too varied to be categorized into a point source category. In particular, these discharges vary by: • Contaminants (e.g., metals, pesticides, dioxin); • Treatment technologies (e.g., air stripping, granular activated carbon, chemical/ultra-violet oxidation, aerobic biological reactors, chemical precipitation); and • Types of facilities causing groundwater contamination (e.g., wood treatment facilities, metal finishing and electroplating facilities, drum recycling facilities, mine sites, mineral processing facilities, radium processing facilities). Moreover, the duration and volume of these direct discharges vary significantly due to differences in aquifer characteristics and the magnitude, fate, and transport of contaminants in aquifers and vadose zones. Currently at Superfund sites, permit writers determine technology- based effluent limits using their best professional judgment (BPJ). EPA selects the remedial technology and derives numerical effluent discharge limits. The permit must also contain more stringent effluent limitations when required to comply with state water quality standards. EPA finds that the current site-specific BPJ approach is workable and flexible within the context of a Superfund cleanup. 7.2.4 Categories with One Facility Dominating the TWPE EPA identified point source categories with significant TWPE where only one facility was responsible for most of the TWPE reported to be discharged (i.e., where one facility's TWPE accounted for more than 95 percent of the category TWPE, but was not the only facility reporting discharges for the category). Table 7-4 lists these categories. EPA identified 10 facilities that dominated the TWPE in the category to which they belonged. EPA investigated these facilities to determine if their discharges were representative of the category. If they were not, EPA subtracted the facility's TWPE from the total category TWPE and recalculated the category's ranking. EPA performed this analysis separately for both of the databases. Based on EPA's knowledge of these industries and the review of the pollutant discharges for these facilities, EPA determined that all of the pollutant discharges are representative of the industry and therefore, EPA did not remove the discharges from the category. 7.2.5 Combining the Final DMR and TRI Rankings After adjusting the category TWPE totals and rankings as described in Sections 7.2.1 through 7.2.4, EPA consolidated the DMRLoads2007 and TRIReleases2007 rankings into one set using the following steps: 7-12 ------- Section 7 - Results of 2009 Screening-Level Analysis Table 7-4. Point Source Categories with One Facility Dominating the TWPE Discharges Point Source Category Textile Mills (Part 4 10) Independent and Stand Alone Labs (Potential New Category) Canned and Preserved Seafood Processing (Part 408) Plastics Molding and Forming (Part 463) Timber Products Processing (Part 429) Soap and Detergent Manufacturing (Part 417) Ferroalloy Manufacturing (Part 424) Construction and Development (Potential New Category) Soap and Detergent Manufacturing (Part 417) Tobacco Products (Potential New Category) Facility with Over 95% of Category TWPE Deroyal Textiles Brookhaven National Laboratory Campbell Soup Company Innovia Films, Inc Stimson Lumber Co Bonner Mill Stepan Company - Elwood Eramet Marietta Inc Aeroquip - Vickers Crodia Inc Philip Morris Park 500 Site Facility Location Camden, SC Upton, NY Napoleon, OH Topeka, KS Bonner, MT Elwood, IL Marietta, OH Joplin, MO New Castle, DE Chester, VA Data Source DMR 2007 DMR 2007 DMR 2007 DMR 2007 DMR 2007 DMR 2007 DMR 2007 DMR 2007 TRI 2007 TRI 2007 Pollutant Driving TWPE Aldrin PCBs Hexavalent Chromium Carbon Disulfide Chlorine Hexachlorobenzene Cadmium Cadmium Bis(2-chloroethyl) Ether Chlorine Facility TWPE 76,469 5,166 3,123 24,219 51,374 47,795 4,349 324 14,453 4,730 Percent of Total Category TWPE 95.6% 96.5% 96.6% 98.3% 99.7% 99.96% 99.99% 99.99% 99.1% 99.4% Action Did not remove load from category TWPE Did not remove load from category TWPE Did not remove load from category TWPE Did not remove load from category TWPE Did not remove load from category TWPE Did not remove load from category TWPE Did not remove load from category TWPE Did not remove load from category TWPE Did not remove load from category TWPE Did not remove load from category TWPE Source: TRIReleases2007 v2\DMRLoads2007 v3. 7-13 ------- Section 7 - Results of 2009 Screening-Level Analysis • EPA combined the two lists of point source categories by adding each category's DMRLoads2007 TWPE and TRIReleases2007 TWPE26. • EPA then ranked the point source categories based on total DMRLoads2007 and TRIReleases2007 TWPE. Table 7-5 presents the combined DMRLoads2007 and TRIReleases2007 rankings. These are the final category rankings accounting for all corrections made to the databases during the 2009 screening-level review and removal of any categories and discharges as discussed in Sections 7.2.1 through 7.2.4. 7.3 Identification of Categories With Existing Effluent Guidelines for Further Review After completing the development of the prioritized list, shown in Table 7-5, EPA selected for further review the point source categories that cumulatively discharge 95 percent of the total DMRLoads2007 and TRIReleases2007 TWPE. The cutoff point is shown as a bold line in Table 7-5. EPA performed detailed studies on three point source categories as part of its 2009 annual review based on the results of its 2007 and 2008 annual reviews. Because EPA data collection was not finished in 2008, EPA continued detailed studies of the Steam Electric Generating Category (Part 423), Oil and Gas Extraction (Part 435) (to assess whether to revise the limits to include coalbed methane extraction as a new subcategory), and the Health Care Industry (includes Hospitals (Part 460)). EPA did not identify additional categories for detailed study as part of the 2009 annual review. EPA's detailed studies generally examine the following: (1) wastewater characteristics and pollutant sources; (2) the pollutants driving the toxic-weighted pollutant discharges; (3) availability of pollution prevention and treatment; (4) the geographic distribution of facilities in the industry; (5) any pollutant discharge trends within the industry; and (6) any relevant economic factors. First, EPA attempts to verify the screening-level results and fill in data gaps. Next, EPA considers costs and performance of applicable and demonstrated control technology, process change, or pollution prevention alternatives that can effectively reduce the pollutants remaining in the industrial category's wastewater. Last, EPA considers the affordability or economic achievability of the technology, process change, or pollution prevention measures identified above. Types of data sources that EPA may consult in conducting its detailed studies include, but are not limited to: (1) the U.S. Economic Census; (2) TRI, PCS, and ICIS-NDPES data; (3) trade associations and reporting facilities to verify reported releases and facility categorization; (4) regulatory authorities (states and EPA regions) to understand how category facilities are permitted; (5) NPDES permits and their supporting fact sheets; (6) EPA effluent guidelines technical development documents; (7) relevant EPA preliminary data summaries or study reports; and (8) technical literature on pollutant sources and control technologies. 26 EPA notes that this may result in "double-counting" of chemicals a facility reported to both PCS/ICIS-NPDES and TRI, and "single-counting" of chemicals reported in only one of the databases. The combined databases do not count chemicals that may be discharged but are not reported to PCS/ICIS-NPDES or TRI. ------- Section 7 - Results of 2009 Screening-Level Analysis Table 7-5. Final DMRLoads2007 and TRIReleases2007 Combined Point Source Category Rankings 40CFR Part 423 430 418 414 419 415 421 440 455 471 410 429 417 444 445 463 439 409 458 436 428 422 464 469 467 437 Point Source Category Steam Electric Power Generating Pulp, Paper And Paperboard Fertilizer Manufacturing Organic Chemicals, Plastics And Synthetic Fibers Petroleum Refining Inorganic Chemicals Manufacturing Nonferrous Metals Manufacturing Ore Mining And Dressing Pesticide Chemicals Nonferrous Metals Forming And Metal Powders Textile Mills Timber Products Processing Soap And Detergent Manufacturing Waste Combustors Landfills Plastics Molding And Forming Pharmaceutical Manufacturing Sugar Processing Carbon Black Manufacturing Mineral Mining And Processing Rubber Manufacturing Phosphate Manufacturing Metal Molding And Casting (Foundries) Electrical And Electronic Components Aluminum forming Centralized Waste Treatment DMRLoads2007 TWPE 20,374,829 a 2,726,865 b 1,095,046 413,226 c 402,506 393,523 342,747 184,455 180,117 119,244 79,934 51,552 47,815 38,412 35,804 24,626 24,937 32,520 26,719 11,195 18,459 11,271 9,350 12,182 10,403 TRIReleases2007 TWPE 541,508 459,959 4,462 574,742 171,756 54,657 38,885 44,437 24,693 8,834 2,389 16,301 14,585 40 83 8,781 7,996 26 32,375 2,416 7,864 250 6,115 7,551 2,707 3,785 Total TWPE 20,916,337 3,186,823 b 1,099,509 987,968 c 574,262 448,181 381,632 228,892 204,810 128,077 82,323 67,852 62,401 38,451 35,887 33,407 32,934 32,545 32,375 29,135 19,059 18,709 17,386 16,902 14,889 14,189 Cumulative Percent of Total TWPE 72.64 83.71 87.53 90.96 92.96 94.51 95.84 96.63 97.35 97.79 98.08 98.31 98.53 98.66 98.79 98.90 99.02 99.13 99.24 99.34 99.41 99.47 99.54 99.59 99.65 99.69 Rank 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 7-15 ------- Section 7 - Results of 2009 Screening-Level Analysis Table 7-5. Final DMRLoads2007 and TRIReleases2007 Combined Point Source Category Rankings 40CFR Part NA 454 411 425 468 NA NA 407 424 406 408 434 461 405 443 NA 426 457 465 435 466 446 447 460 NA Point Source Category Miscellaneous Foods And Beverages Gum And Wood Chemicals Manufacturing Cement Manufacturing Leather Tanning And Finishing Copper forming Independent And Stand Alone Labs Tobacco Products Canned And Preserved Fruits And Vegetables Processing Ferroalloy Manufacturing Grain mills Canned And Preserved Seafood Processing Coal Mining Battery Manufacturing Dairy products processing Paving And Roofing Materials (Tars And Asphalt) Printing & Publishing Glass Manufacturing Explosives Manufacturing Coil Coating Oil & Gas Extraction Porcelain Enameling Paint Formulating Ink Formulating Hospital Photo Processing DMRLoads2007 TWPE 5,842 10,478 8,960 8 2,310 5,355 3 1,757 4,349 1,984 3,232 2,294 1,096 76 1,280 999 353 785 166 256 11 15 1 TRIReleases2007 TWPE 6,576 55 452 7,802 4,951 30 4,756 2,960 340 2,084 234 493 1,642 2,402 249 110 546 14 241 164 140 20 Total TWPE 12,418 10,532 9,412 7,809 7,261 5,385 4,759 4,717 4,689 4,068 3,467 2,787 2,738 2,479 1,529 1,109 899 798 407 256 175 140 20 15 1 Cumulative Percent of Total TWPE 99.74 99.77 99.81 99.83 99.86 99.88 99.89 99.91 99.93 99.94 99.95 99.96 99.97 99.98 99.99 99.99 99.99 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 Rank 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 7-16 ------- Section 7 - Results of 2009 Screening-Level Analysis Table 7-5. Final DMRLoads2007 and TRIReleases2007 Combined Point Source Category Rankings 40CFR Part 459 442 Point Source Category Photographic Transportation Equipment Cleaning Total DMRLoads2007 TWPE 1 0 26,719,348 TRIReleases2007 TWPE 2,073,457 Total TWPE 1 0 28,792,806 Cumulative Percent of Total TWPE 100.00 100.00 Rank 52 53 Source: TRIReleases2007_v2; DMRLoads2007_v3. a EPA corrected a suspected units error in DMRLoads2007_v3 for FB Culley Station in Newburgh, IN (IN0002259) in the Steam Electric Power Generating Category. EPA attempted to contact the facility but the facility never returned calls. Therefore, EPA was unable to verify the correction. b For the Pulp, Paper, and Paperboard Category, EPA contacted facilities to verify the concentrations of dioxin and dioxin-like compounds in PCS and ICIS- NPDES. EPA found that, for all facilities contacted, there were either units errors (e.g., reported as ng/L but in the database as mg/L) or missing non-detect indicators. The new Pulp, Paper, and Paperboard Category total DMR TWPE is 252,163, while the new DMR and TPJ combined total TWPE is 712,122. See Section 12.2.2.1 in the Technical Support Document for the Preliminary 2010 Effluent Guidelines Program Plan (U.S. EPA, 2009) for additional details on the facilities-specific corrections. dEPA contacted GE Silicones in Friendly, WV (WV0000094), in the OCPSF Category and identified a units error in DMRLoads2007_v3 (Martin, 2009). The new LEY and TWPE reported for this facility were recalculated and are now 158 and 100.3, respectively. The new OCPSF Category total DMR TWPE is 308,721, while the new DMR and TRI combined total TWPE is 883,463. NA - Not applicable; no existing ELGs apply to discharges. 7-17 ------- Section 7 - Results of 2009 Screening-Level Analysis Preliminary category reviews are similar to detailed studies and have the same purpose. During preliminary reviews, EPA generally examines the same items listed above for detailed studies. However, EPA's preliminary review of a category and available pollution prevention and treatment options is less rigorous than its detailed studies. While EPA collects and analyzes hazard and technology-based information on categories undergoing preliminary review, it assigns a higher priority to investigating categories undergoing detailed studies. EPA identified for preliminary review those industrial categories currently regulated by existing effluent guidelines that cumulatively compose more than 95 percent of the combined DMRLoads2007 and TRIReleases2007 total TWPE. EPA also reviewed the Ore Mining and Dressing Category (40 CFR Part 440) because during previous annual reviews, EPA has concluded that there are not sufficient data available to determine whether wastewater discharges from the Ore Mining and Dressing Category warrant a detailed study. In addition to the Steam Electric Power Generating Category this list includes the following point source categories: • Fertilizer Manufacturing; • Inorganic Chemicals Manufacturing; • Nonferrous Metals Manufacturing; • Ore Mining and Dressing; • Organic Chemicals, Plastics, and Synthetic Fibers; • Petroleum Refining; and • Pulp, Paper and Paperboard. EPA recently conducted detailed studies or preliminary reviews of many of the categories listed above. For each of these categories, because EPA's annual review builds on previous reviews, EPA primarily looked at the pollutants reported in 2007 and their contribution to their category's TWPE. After considering the results of the studies and preliminary category reviews, EPA will determine whether further study or development or revision of an effluent guideline is appropriate. Final determinations will be presented in the 2010 Effluent Guidelines Plan. 7.4 Results of 2009 Screening-Level Analysis References 1. Davis, Katherine. 2009. Notes from Telephone Conversation between Elizabeth Sabol, Eastern Research Group, Inc. and Katherine Davis, Westvaco Texas, L.P. RE: Basis of TCDD Equivalent Concentrations Reported in 2007. (July 7). EPA-HQ-OW-2008-0517 DCN 06547. 2. Martin, Jason. 2009. Notes from Telephone Conversation between Elizabeth Sabol, Eastern Research Group, Inc. and Jason Martin, MPM Silicones LLC. RE: Basis of Copper (Total Recoverable) Concentrations Reported in 2007. (July 1). EPA-HQ-OW- 2008-0517 DCN 06549. 3. McCuutchen, Kate. 2009. Notes from Telephone Conversation between Elizabeth Sabol, Eastern Research Group, Inc. and Kate McCuutchen, Blue Heron Paper Co. RE: Basis of Methylmercury Concentration Reported in 2007 in DMR. (July). EPA-HQ-OW-2008- 0517 DCN 06546. ------- Section 7 - Results of 2009 Screening-Level Analysis 4. Verderese, Jim. 2009. Notes from Telephone Conversation between Elizabeth Sabol, Eastern Research Group, Inc. and Jim Verderese, General Electric Erie. RE: Basis of Mercury Concentration Reported in December 2007 in DMR. (July 1). EPA-HQ-OW- 2008-0517 DCN 06548. 5. U. S. EPA. 2009. Technical Support Document for the Preliminary 2010 Effluent Guidelines Program Plan. EPA-821-R-09-006. Washington, DC. (October). EPA-HQ- OW-2007-0571 DCN 06703. 7-19 ------- |