                                       
                    Economic Analysis for the Final Rule: 
          Federal Aluminum Aquatic Life Criteria Applicable to Oregon
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                 December 2020
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                 United States Environmental Protection Agency
                                Office of Water
                       Office of Science and Technology
                        1200 Pennsylvania Avenue, N.W. 
                             Washington, DC 20460


































                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                               Table of Contents
                                       
                                       
Executive Summary	ii
Background and Final Criteria	ii
Point Source Compliance Costs	iv
Incremental Impairments	vi
Total Costs	vi
1.	Introduction	1-1
1.1	Background	1-1
1.2	Purpose and Scope of the Analysis	1-2
1.3	Report Organization	1-2
2.	Baseline for the Analysis	2-1
2.1	Water Quality Criteria	2-1
2.2	Sources of Toxic Pollutants to Surface Waters	2-2
2.2.1	Municipal and Industrial Dischargers	2-3
2.2.2	Urban Stormwater	2-4
2.2.3	Agriculture and Forestry	2-4
2.2.4	Mining	2-5
2.2.5	Atmospheric Deposition	2-6
2.3	Water Quality	2-6
3.	Final Criteria	3-1
3.1	Economic Analysis Criteria	3-2
3.2	Criteria Used for Sensitivity Analysis	3-3
3.3	Expected Benefits	3-4
4.	Methods for Estimating Potential Costs: Point Sources	4-1
4.1	Identification of Potentially Affected Permittees	4-1
4.2	Reasonable Potential Analysis	4-2
4.3	Results of RPA	4-3
4.4	Supplemental Point Source Analysis: Identifying Control Scenarios	4-3
4.4.1	Process Optimization	4-3
4.4.2	Technology Change	4-4
4.4.3	End-of-Pipe Treatment	4-5
4.5	Supplemental Point Source Analysis: Unit Costs	4-5
4.5.1	Aluminum Anodizing Facilities	4-5
4.5.2	Drinking Water Treatment Plants	4-8
4.5.3	Wastewater Treatment Facilities	4-12
4.6	Supplemental Point Source Analysis: Total Costs	4-14
5.	Methods for Identifying Potential Costs: Incremental Impairments	5-1
5.1	Available Data	5-1
5.2	Identifying Exceedances	5-2
5.3	Identifying Potential Control Actions and Costs	5-4
6.	Potential Compliance Costs	6-1
6.1	Point Sources	6-1
6.2	Nonpoint Sources	6-2
6.3	Government Regulatory Costs	6-3
6.4	Total Estimated Costs	6-3
6.5	Uncertainties in the Analysis	6-5
6.5.1	Data Limitations	6-5
6.5.2	Baseline	6-5
6.5.3	Final Criteria	6-5
6.5.4	Point Sources	6-5
6.5.5	Nonpoint Sources	6-6
7.	References	7-1
                                       
Appendix A: Facility Analyses
Appendix B: Net Present Value of Costs 

                                       
                               List of Exhibits
Exhibit 3-1. Map of Nine Level III Ecoregions in Oregon	3-2
Exhibit 3-2. Acute and Chronic Aluminum Criteria Values by Ecoregion that EPA Used to Evaluate Potential Impacts and Costs, also known as Economic Analysis Criteria	3-3
Exhibit 3-3. Alternative Economic Analysis Criteria used for Sensitivity Analysis	3-4
Exhibit 4-1. NPDES Permitted Facilities Potentially Affected by the Final Rule	4-2
Exhibit 4-2. Costs for Aluminum Anodizers, Both Options (2017$)	4-8
Exhibit 4-3. WTPs in Oregon by Population Served	4-10
Exhibit 4-4. Annual Costs of Chemicals and Sludge Disposal for WTPs by Range of Population Served (2017$)	4-11
Exhibit 4-5. Annual Costs of Chemicals and Sludge Disposal for POTWs (2017$)	4-13
Exhibit 4-6. Total Annual Costs for Point Source Compliance (Supplemental Analysis) (2017$)	4-14
Exhibit 5-1. Summary of Available Aluminum Monitoring Data by Ecoregion	5-2
Exhibit 5-2. Number of Potential Impairments by Ecoregion Based on Aluminum Monitoring Data	5-2
Exhibit 5-3. Location of Potential Impairments Based on Aluminum Monitoring Data	5-4
 

                          Acronyms and Abbreviations
Alum		Aluminum sulfate
AWQC	Ambient water quality criteria
AWQMP	Agricultural water quality management program
BLM		Biotic ligand model
BMP		Best management practice
CCC		Criterion continuous concentration
CERCLA	Comprehensive Environmental Response, Compensation and Liability Act
CMC		Criterion maximum concentration
CWA		Clean Water Act
DOC		Dissolved organic carbon
EPA		Environmental Protection Agency
EQC		Environmental Quality Commission
ESA		Endangered Species Act
GDP		Gross Domestic Product
ICIS-NPDES	Integrated Compliance Information System for NPDES program
M		Multiplier
MEP		Maximum extent practicable
mgd		Million gallons per day
mg/L		Milligrams per liter
MLR		Multiple Linear Regression model
MP&M	Metal Products and Machinery
MRL		Minimum reporting level
MS4		Municipal separate storm sewer system
NMFS		National Marine Fisheries Service
NPDES	National Pollutant Discharge Elimination System
NPV		Net Present Value
NWAS		Northwest Aluminum Specialties	
ODA		Oregon Department of Agriculture
ODEQ		Oregon Department of Environmental Quality
O&M		Operation and Maintenance
POTW		Privately and/or Publicly owned treatment works	
RO		Reverse osmosis
RP		Reasonable potential
RPA		Reasonable potential analysis
SB1010	Oregon 1993 Agricultural Water Quality Management Act, (Senate Bill 1010)
SWPCP	Stormwater pollution control plan
SWMP		Stormwater management program
TMDL		Total maximum daily load
ug/L		micrograms per liter
USFWS	U.S. Fish and Wildlife Service	
USGS		U.S. Geological Survey
WET		Whole effluent toxicity
WQBEL	Water quality based effluent limit
WQS		Water quality standards
WTP		Water treatment plant


Executive Summary
The United States Environmental Protection Agency (EPA) is finalizing federal Clean Water Act (CWA) aquatic life criteria applicable to fresh waters under jurisdiction of the State of Oregon to protect freshwater aquatic life from acute and chronic exposure to aluminum. This report provides estimates of the potential incremental control actions and their costs, as well as a qualitative description of benefits, that may be associated with the final regulation. 

 Background and Final Criteria
The CWA directs States to adopt water quality standards (WQS) to protect the public health and welfare, enhance the quality of water, and serve the purposes of the CWA. Under CWA Section 303(c)(2)(A), State WQS must include: (1) designated uses for the navigable waters involved and (2) water quality criteria for such waters based upon such uses. The CWA also requires States to hold public hearings once every three years for the purpose of reviewing applicable WQS and, as appropriate, modifying and adopting standards. EPA must approve or disapprove any new or revised standards. 
On July 8, 2004, the Oregon Department of Environmental Quality (ODEQ) submitted 89 revised aquatic life criteria for 25 toxic pollutants to EPA for review under CWA Section 303(c), including acute and chronic criteria for aluminum. Oregon's Environmental Quality Commission (EQC) resubmitted the revised WQS with corrections to EPA for review on April 23, 2007, and again on July 21, 2011. EPA disapproved the acute and chronic aluminum criteria in January 2013. The CWA directs EPA to promptly propose water quality standards (WQS) that meet CWA requirements if a state does not adopt WQS addressing the Agency's disapproval. On April 20, 2015, EPA was sued for failing to promptly prepare and publish replacement criteria for seven of the aquatic life criteria disapproved in its January 31, 2013 action (Northwest Environmental Advocates v. U.S. EPA, 3:15-cv-663-BR (D. Or. 2015)). This lawsuit was resolved in a consent decree entered by the District Court on June 9, 2016, which established deadlines for EPA to address the disapproved aquatic life criteria by either approving replacement criteria submitted by Oregon or by proposing and promulgating federal criteria. The preamble to the final rule contains more detail regarding the litigation and consent decree. The State has not yet addressed EPA's 2013 disapproval of its freshwater criteria for aluminum, therefore, EPA is finalizing freshwater acute and chronic criteria for aluminum in accordance with CWA Section 303(c)(3) and (c)(4). EPA will withdraw its final rule for freshwater acute and chronic aluminum if Oregon adopts and submits protective criteria for aluminum that the Agency approves as meeting CWA requirements.
In December 2018, EPA revised its federally recommended ambient water quality criteria (AWQC) for aluminum to reflect the latest science, and published an updated final federal recommendation, Final Aquatic Life Ambient Water Quality Criteria for Aluminum 2018 (USEPA, 2018). A key aspect of the updated federal recommendation is the shift from applying a single AWQC (as in EPA's 1988 recommendation) for all waters to a model-based AWQC that incorporates the water chemistry parameters (i.e., pH, dissolved organic carbon (DOC), and total hardness) that better reflect the toxicity of aluminum to aquatic life. Instead of setting AWQC to a single value that would be protective everywhere (and therefore more conservative than necessary in many circumstances), the new AWQC is a calculator which is built upon an underlying peer-reviewed model that results in more stringent criteria only where the local water chemistry conditions warrant. Under most conditions, the final AWQC are less stringent than EPA's 1988 recommendation. EPA's final rule for Oregon is based on the 2018 final federal recommendation, which reflects that latest scientific knowledge. 
To protect aquatic life in Oregon's fresh waters, EPA is finalizing aluminum criteria for Oregon based on Multiple Linear Regression (MLR) models for aluminum, as described in EPA CWA Section 304(a) recommended freshwater aquatic life criteria for aluminum (USEPA, 2018). Although the State has considerable discretion in how it implements the final criteria, for the purposes of this economic analysis, EPA used ecoregional-based economic analysis criteria to estimate potential impacts and costs. The economic analysis criteria were calculated by inputting available data for measured water chemistry conditions for each ecoregion into the model and selecting the 10[th] percentile of the model outputs for each ecoregion. The resulting criteria were used to estimate potential economic impacts of this rule, and these are referred to as the economic analysis criteria. This document provides an estimate of the potential impacts of the final criteria by comparing the "economic analysis criteria" (a proxy for the final criteria) to the baseline, which is the State's narrative criteria. Actual impacts and costs will depend on the water quality criteria values that Oregon develops for each water body using the aluminum MLR models, which may differ from these economic analysis criteria. 
As Oregon currently has narrative criteria for aluminum and no numeric acute or chronic aquatic life criteria for aluminum, it is difficult to compare the numeric economic analysis criteria used in this analysis to the baseline narrative criteria without significant additional analysis. Absent such analysis, all costs associated with meeting the final criteria identified by this economic analysis are attributed to the final rule. Also, although no Oregon waters are listed as impaired with respect to the narrative aluminum criteria, it is possible that some waters are in fact impaired (but have not been listed). Thus, if any waters are in fact impaired under the narrative criteria in the baseline, the costs estimated here for those waters would be more properly considered baseline costs, and not incremental costs of the final rule. To the extent this is the case, the costs estimated here are an overestimate. State implementation would result in new or revised National Pollutant Discharge Elimination System (NPDES) permit conditions for point source dischargers found to have reasonable potential (RP) to cause or contribute to an excursion of the final criteria, and additional controls on nonpoint sources of pollutant loadings where incremental impairments are identified. This analysis provides information on the expected incremental costs associated with such controls based on the information available to EPA. 
This final rule is anticipated to result in environmental benefits. There is both the potential for and observed occurrences of aluminum toxicity adverse impacts on aquatic biota in Oregon surface waters. Establishing protective criteria as water quality standards helps ensure that aquatic life is protected and that impaired waters are addressed, and conditions improved. Numeric criteria allow measurement and quantification for purposes of monitoring and assessment, determining the need for discharge limitations, modeling source reduction needs on a watershed basis, and setting targets for allowable loads from nonpoint sources. Without such numeric criteria, it is difficult to be proactive in addressing both the potential for adverse effects from aluminum and the need for restoration efforts. The benefits that accrue are expected to be both in use (e.g., from enhanced recreational fishing) and nonuse benefits (e.g., from the satisfaction of knowing that ecological integrity is maintained and protected). However, there is no attempt to quantify or monetize these benefits in this analysis for a variety of reasons. The scope of the rule is limited, the costs are speculative and not large, and lacking the measurement and quantification efforts previously noted hampered the selection of metrics that would represent the full range of likely benefits.

 Point Source Compliance Costs
To identify facilities in Oregon whose NPDES permits contain effluent limitations or monitoring requirements for aluminum, EPA used the Integrated Compliance Information System NPDES program (ICIS-NPDES) database. EPA examined data from the past ten years. Minor facilities (i.e., facilities that have a design flow of less than 1 million gallons per day (mgd), serve a population of less than 10,000, are not required to have a pretreatment program, and do not have the potential to cause significant water quality impacts) are listed in the database, though their effluent data and monitoring requirements are not. This includes one minor facility with effluent limits for aluminum. General permits are also listed in the database. None of them included aluminum effluent limits or monitoring requirements. EPA identified two major facilities with aluminum limits and data. One of these facilities has closed and the property has been redeveloped, so EPA evaluated the remaining facility, which discharges to the Columbia River. EPA evaluated existing baseline permit conditions, RP to exceed economic analysis criteria, and potential to exceed projected effluent limitations based on the last five years of effluent monitoring data. Analysis of the available data for the facility indicates that no exceedances of projected effluent limits are expected when using the economic analysis criteria. 
Because of the lack of data for aluminum in point source discharges in the State in conjunction with some waters being potentially incrementally impaired under the final criteria, EPA additionally identified potential costs for point source dischargers that utilize aluminum in their operations and, therefore, could potentially be affected by the final rule by presuming a need for additional controls or product substitution, without facility-specific data. This analysis supplements the analysis of potential point source costs based on reasonable potential analysis (RPA) using facility-specific data. 
For this supplemental point source analysis, EPA evaluated potential costs to three types of facilities (aluminum anodizing facilities, drinking water treatment plants, and wastewater treatment facilities) that could incur costs under the final rule (if the facilities were found to have RP). Because this analysis requires EPA to make a number of assumptions, this analysis is only as accurate as those assumptions. Because EPA is uncertain as to the number of facilities that would require additional technology to comply, the results of this supplemental analysis are highly speculative compared to costs estimated made using the facility-specific, data-based RPA approach. First, several aluminum anodizing facilities discharge to either privately owned or publicly owned treatment works (POTWs). This analysis assumed that the final criteria would result in the POTWs establishing local limits for these aluminum anodizers. EPA identified two options for potential treatment upgrades that may be required: countercurrent cascade rinsing and countercurrent cascade rinsing plus chemical precipitation/flocculation. EPA developed cost estimates for each technology. Second, drinking water treatment plants using surface water, or groundwater under the influence of surface water as source water, often use aluminum sulfate (alum) in treatment processes as a coagulant. The final criteria may result in some of the State's drinking water systems needing to reduce aluminum concentrations in their wastewater discharges. For this analysis, EPA assumed that all water treatment plants in Oregon with surface, or groundwater under the influence of surface water, as source water that discharge directly to surface waters currently use alum as a coagulant. EPA then estimated costs to the plants if they were to reduce aluminum in their wastewater discharges, diverting that aluminum to sludge disposal. Third, wastewater treatment facilities often use chemical precipitation followed by filtration to remove phosphorus from the wastewater prior to discharge. EPA examined the NPDES permits in Oregon with permit limits for total phosphorus, which could indicate use of alum for phosphorus removal. Unless EPA could confirm otherwise, EPA assumed that these facilities would substitute ferrous coagulants for the aluminum coagulants and estimated the effect on costs of that change.

EPA made both a low-end and a high-end estimate for the costs to the State's 14 aluminum anodizers, based on two different technology upgrade options. Capital costs were annualized over the expected 20-year life of the technology (Metcalf & Eddy, 2014; USEPA, 1979) and combined with annual O&M costs to produce an estimate of annual costs. Without facility-specific information to shed light on which option each facility would choose if it had to upgrade, EPA estimated that if all 14 facilities upgraded to countercurrent cascade rinsing technology, the total annual cost would be $61,600 using a 3% discount rate and $84,000 using a 7% discount rate. On the high end, EPA estimated that if all 14 facilities upgraded to countercurrent cascade rinsing technology plus chemical precipitation and settling, the total annual cost would be $6.88 million using a 3% discount rate and $6.95 million using a 7% discount rate. For the 49 drinking water treatment plants assumed to use alum as a coagulant, EPA estimated the annual costs for chemicals and sludge disposal at $0.90 million; there are no capital costs, so this estimate pertains to both the 3% and 7% discount rates. For the seven wastewater treatment facilities currently or potentially using alum as a coagulant, EPA found that if they were to switch to a ferrous coagulant, each would realize cost savings. EPA assumes that, in absence of the final rule, the facilities would possess full information regarding the prices of both coagulants and other factors affecting the choice of coagulant (of which EPA is not aware, lacking facility-specific data), and would already be using the lowest cost treatment. Therefore, although the analysis would suggest potential cost savings, EPA estimated that the final rule would result in no change in cost for these facilities. For all three of these discharger types, EPA based these estimates on an assumed need for control strategies simply based on the potential presence of aluminum in various operations, without specific knowledge of actual levels in any waste stream; thus, these supplemental cost estimates are speculative.
 Incremental Impairments
If the final criteria result in an incremental increase in waters that are identified as impaired, requiring the development of a Total Maximum Daily Load (TMDL), there could be costs to point and nonpoint sources of aluminum loadings under a TMDL. According to ODEQ's assessment practices, a water body is impaired when there are two or more exceedances of a numeric criterion, and it is of "potential concern" when there is a single exceedance. Using available ambient surface water monitoring data collected between August 2000 and May 2017 (described in a technical support document in the docket), EPA compared total aluminum concentrations to the ecoregion-specific economic analysis criteria values to identify water bodies that may be identified as impaired, pursuant to the State's assessment practices. 
Based on available monitoring data, EPA had sufficient data for 260 stations, of which EPA identified 54 that would be potentially impaired under the economic analysis criteria. 
If ODEQ develops TMDLs for aluminum impairments as a result of the final rule, then it will incur incremental costs. If, for example, average TMDL development costs range from $38,000 to $41,000 (USEPA, 2001a; adjusted to 2018 dollars), then Oregon's incremental administrative costs would range from $2.05 million to $2.21 million for 54 impairments. TMDL development costs are one-time costs that EPA assumed would be spread uniformly over several years because it is not realistic to assume that Oregon would conduct all these analyses at once. EPA is assuming a 10-year period for TMDL development. This time period covers the time it takes to develop a TMDL and assumes Oregon will stagger the start dates of the TMDLs' development. As such, annual costs over the initial 10-year period will range from $205,000 to $221,000.  
This may overestimate the government administrative costs. For example, the State may be able to combine TMDLs for common water bodies (i.e., if the State decides to combine development of TMDLs for a class of waters with impairments for similar causes) and reduce development costs, though EPA has no way to predict whether the State will do this, or for how many TMDLs. Additionally, it is possible that after listing a water as impaired with respect to aluminum based on the final criteria, Oregon could conduct additional monitoring before developing a TMDL. Additional monitoring could result in the State identifying the water as unimpaired with respect to aluminum, and the State would therefore not develop a TMDL. On the other hand, the dated unit costs for TMDL development likely underestimate costs. The overall impact of these uncertainties on estimated costs of TMDL development is unknown. 
 Total Costs
Based on the initial RPA, there would be no compliance costs for point source dischargers under the final criteria. However, based on the supplemental point source analysis that EPA performed because of the lack of effluent data on aluminum concentrations in NPDES discharges, EPA estimated that there could be costs associated with compliance for 14 aluminum anodizers and 49 drinking water treatment facilities. The total annual costs to those facilities are estimated between $1.0 million and $7.8 million at a 3% discount rate over the 20-year expected life of the capital (Metcalf & Eddy, 2014; USEPA, 1979) and between $1.0 (costs appear to be the same due to rounding) and $7.9 million at a 7% discount rate. 
If the final criteria result in additional waters being identified as impaired, then the State will incur incremental costs for TMDL development. Using available monitoring data, EPA estimated that the annual costs for TMDL development would range from $205,000 to $221,000 for the 54 TMDLs over 10 years, dropping to $0 thereafter.  
Combining the potential costs for point source compliance from the supplemental point source analysis with the incremental cost of TMDL development, the total cost would range from $1.2 million to $8.0 million for the first ten years at a 3% discount rate, and $1.2 million to $8.1 million at a 7% discount rate. The cost would be slightly less in subsequent years after the TMDL development is complete.  

The point source cost estimates are from a supplemental analysis based on very limited data; the estimates are highly speculative and may err on the high side. The number of TMDLs, and therefore costs, could be higher or lower than these estimates depending on choices made by the State (as noted above). Further, if nonpoint sources are the primary cause of some impairments, then the final criteria may result in some costs to nonpoint sources to implement best management practices (BMPs) to reduce aluminum loadings to affected waters. The magnitude of cost impacts to nonpoint sources depends on the extent to which additional practices are needed to meet the final criteria. While the total effect of these uncertainties on costs are ambiguous, EPA expects that its estimates are likely to be overestimated, particularly at the high end of the range, because the point source costs account for the majority of costs.

The aluminum criteria being finalized for Oregon are designed to protect aquatic life from detrimental impacts that excess aluminum can cause. Aluminum's toxicity to aquatic life depends on the water chemistry (e.g., pH, dissolved organic carbon (DOC), and total hardness) at a specific site. For example, at acidic pH levels, aluminum can lead to the demise of biotic communities, such as the impacts of aluminum toxicity observed in the 1980s in the northeastern U.S. due to acid rain. Numeric criteria allow measurement and quantification for purposes of monitoring and assessment, determining the need for discharge limitations, modeling source reduction needs on a watershed basis, and setting targets for allowable loads from point and nonpoint sources. Without such numeric targets, it is difficult to address both the potential for adverse effects from aluminum and the need for restoration efforts with an acceptable level of certainty. People hold value for protecting aquatic life from the impacts of high levels of aluminum, though this value cannot be fully ascertained through readily available and observable market (e.g., commodity purchases) or behavioral data (e.g., recreational trips). EPA has not attempted to quantify or monetize the benefits associated with implementation of the final criteria but includes a more in-depth qualitative discussion of benefits in this economic analysis.

Introduction
The United States Environmental Protection Agency (EPA) is proposing to promulgate revisions to the current federal aquatic life criteria applicable to waters under jurisdiction of the State of Oregon to protect freshwater aquatic life from acute and chronic exposure to aluminum. This report provides estimates of the potential incremental compliance actions and costs that may be associated with the final regulation. 
Background
The Clean Water Act (CWA) directs States to adopt water quality standards (WQS) to protect the public health and welfare, enhance the quality of water, and serve the purposes of the CWA. Under CWA Section 303(c)(2)(A), State WQS must include: (1) designated uses for the navigable waters involved and (2) water quality criteria for such waters based upon such uses. The CWA also requires States to hold public hearings once every three years for the purpose of reviewing applicable WQS and, as appropriate, modifying and adopting standards. EPA must approve or disapprove any new or revised standards. 
On July 8, 2004, the Oregon Department of Environmental Quality (ODEQ) submitted 89 revised aquatic life criteria for 25 toxic pollutants to EPA for review under CWA Section 303(c), including acute and chronic criteria for aluminum. Oregon's Environmental Quality Commission (EQC) resubmitted the revised WQS with corrections to EPA for review on April 23, 2007, and again on July 21, 2011. EPA disapproved the acute and chronic aluminum criteria in January 2013. The CWA directs EPA to promptly propose WQS that meet CWA requirements if a state does not adopt WQS addressing the Agency's disapproval. On April 20, 2015, EPA was sued for failing to promptly prepare and publish replacement criteria for seven of the aquatic life criteria disapproved in its January 31, 2013 action (Northwest Environmental Advocates v. U.S. EPA, 3:15-cv-663-BR (D. Or. 2015)). This lawsuit was resolved in a consent decree entered by the District Court on June 9, 2016 which established deadlines for EPA to address the disapproved aquatic life criteria by either approving replacement criteria submitted by Oregon or by proposing and promulgating federal criteria. The deadlines for aluminum criteria were subsequently amended. The preamble to the final rule contains more detail regarding the litigation and consent decree. The State has not yet addressed EPA's 2013 disapproval of its freshwater criteria for aluminum, therefore, EPA is proposing freshwater acute and chronic criteria for aluminum in accordance with CWA Section 303(c)(3) and (c)(4). EPA will not proceed with final rulemaking (or will withdraw its final rule, as applicable) for freshwater acute and chronic aluminum if Oregon adopts and submits protective criteria for aluminum that the Agency approves as meeting CWA requirements.
EPA recently finalized its federally recommended ambient water quality criteria (AWQC) for aluminum to reflect the latest science. EPA's proposal for Oregon is based on the federal recommendation (USEPA, 2018). A key aspect of the updated federal recommendation is the shift away from applying a single AWQC to all waters to basing AWQC on models that incorporate the water chemistry parameters (i.e., pH, dissolved organic carbon (DOC), and total hardness) that better reflect the toxicity of aluminum to aquatic life. Under most conditions, the AWQC derived from site specific conditions are higher (less stringent) than EPA's prior 1988 numeric criteria recommendation for aluminum. While the criteria finalized here for Oregon will not necessarily be higher for all subject waters in Oregon, this final rule is generally not expected to result in more stringent requirements in Oregon. Instead of setting AWQC to a single value that would be protective everywhere (and therefore be more conservative than necessary in many circumstances), the new AWQC is a calculator which is built upon an underlying peer-reviewed model that results in more stringent criteria only where the local water chemistry conditions warrant. This analysis provides information on the potential for incremental costs to be associated with such incremental controls to ensure attainment of State water quality designated uses protected by the criteria in the final rule. 
Purpose and Scope of the Analysis
Oregon can establish effluent limits of its own on the basis of ambient data and reasonable potential analysis. The purpose of this economic analysis is to identify, using available water quality and discharge data and information, the incremental compliance actions and costs that indirect dischargers, sewerage systems (i.e., privately or publicly owned wastewater treatment works, or POTWs) and industrial point source dischargers in Oregon may incur as a result of EPA's final criteria. State implementation may result in new or revised National Pollutant Discharge Elimination System (NPDES) permit conditions for point source dischargers to incorporate revised water quality based effluent limits (WQBELs) based on reasonable potential (RP) to exceed the water quality criteria. 
The revised standards may also result in incremental determinations that waters exceed WQS. As such, ODEQ may need to develop additional total maximum daily loads (TMDLs) for incrementally impaired waters. There may also be incremental controls and costs associated with load allocations for nonpoint sources under such TMDLs to attain standards. However, the data and information needed to evaluate potential control needs are more limited. For this analysis, EPA identified the potential for incremental impairment, and thus the need for TMDLs to be developed, as well as the potential for incremental controls and costs for nonpoint sources but did not develop statewide cost estimates for nonpoint source controls. 
Report Organization 
The remainder of this report is organized as follows:
Section 2: Baseline for the Analysis  -  describes the current applicable aquatic life criteria and sources of aluminum to surface waters, water quality impairments from aluminum, and ongoing efforts to reduce and eliminate these impairments. 
Section 3: Final Criteria  -  outlines the changes to existing WQS, as well as qualitatively describing the benefits associated with implementing the final criteria. 
Section 4: Method for Estimating Potential Costs: Point Sources  -  describes the methods for estimating compliance costs associated with the final criteria for point sources in terms of revisions to NPDES permits. 
Section 5: Method for Identifying Potential Costs: Incremental Impairments  -  describes the methods for identifying potential for incremental impairment and compliance costs associated with the final criteria for nonpoint sources, as well as potential administrative costs for TMDL development. 
Section 6: Potential Compliance Costs  -  provides estimates of potential costs to comply with the final WQS and discusses the uncertainties associated with the estimates. 
Section 7: References  -  provides the references used in the analysis. 
Appendix A: provides data and information on individual sample facilities. 
Appendix B: provides estimates of costs related to Executive Order (EO) 13771.
Baseline for the Analysis
This section describes the applicable baseline for evaluating the incremental costs associated with the final revised WQS, including current water quality criteria and associated implementation procedures. This section also discusses potential sources of aluminum, the current level of impairment, and listing procedures. 
The baseline is a reference point that reflects the current conditions without the final regulation, serving as a starting point for conducting an economic analysis. According to EPA's Guidelines for Preparing Economic Analyses (USEPA, 2014), the baseline is defined as the world without the final regulation or policy action. As a general rule, EPA typically assumes full compliance with existing and newly promulgated rules (even if they are not yet fully implemented) as a basis for estimating the costs and benefits of a proposed or final regulation. This baseline approach ensures that the costs and benefits of existing rules are not double counted. Therefore, the baseline reflects EPA's expectations of how current regulations are implemented by permitting authorities. The costs associated with the final revised water quality criteria represent incremental actions above those associated with the baseline. 
Water Quality Criteria 
On July 8, 2004, Oregon submitted 89 revised aquatic life criteria for 25 toxic pollutants to EPA for review under CWA Section 303(c), including acute and chronic criteria for aluminum. A subsequent consent decree between EPA and Northwest Environmental Advocates established deadlines for EPA to complete its CWA Section 303(c) review of Oregon's aquatic life criteria. Prior to taking a final action on the aquatic life criteria, EPA requested formal consultation with the National Marine Fisheries Service (NMFS) and the U.S. Fish and Wildlife Service (USFWS) on its proposed approval of the State's criteria, consistent with Section 7 of the Endangered Species Act (ESA). EPA initiated this consultation on January 14, 2008, by submitting a biological evaluation to the NMFS and USFWS, which contained an analysis of the potential effects of EPA's proposed approval of Oregon's criteria, including criteria for aluminum, on threatened and endangered species in Oregon. 

EPA contemplated approving Oregon's aluminum criteria but realized that its initial understanding was based on understanding that Oregon's criteria were entirely equivalent to EPA's 1988 CWA Section 304(a) recommended criteria. However, while EPA's 1988 CWA Section 304(a) recommended aluminum criteria "apply at pH values of 6.5 - 9.0," EPA later identified a footnote to Oregon's revised aluminum criteria table specifying that Oregon's aluminum criteria applied "to waters with pH values less than 6.6 and hardness values less than 12 milligrams per liter (mg/L) (as CaCO3)." The State had not supplied a scientific rationale for the difference between Oregon's statement of the conditions under which the criteria would be valid and EPA's specified pH range for the criteria, so EPA prepared to disapprove the aluminum criteria. EPA sent a letter to the NMFS and USFWS identifying this change. The USFWS had already completed and transmitted its non-jeopardy opinion to EPA by that point, and EPA was therefore unable to withdraw the consultation request for aluminum. The USFWS biological opinion (provided to EPA on July 31, 2012) found that EPA's proposed approval of Oregon's aquatic life criteria would not jeopardize the continued existence of endangered species for which USFWS was responsible, including the aluminum criteria as applied to waters with pH 6.5  -  9. 

The NMFS had not yet transmitted its analysis to EPA at that time, so EPA sent a letter to the NMFS withdrawing its request for consultation on Oregon's acute and chronic aluminum criteria. The NMFS acknowledged EPA's request to withdraw the aluminum criteria from consultation in the biological opinion; however, the NMFS did not have time to modify the document to exclude the acute and chronic aluminum criteria. On August 14, 2012, the NMFS concluded in its biological opinion that seven of Oregon's revised freshwater criteria would jeopardize the continued existence of endangered species in Oregon for which the NMFS was responsible, including acute and chronic aluminum (applied to waters with pH 6.5-9). The NMFS acknowledged EPA's request to withdraw the aluminum criteria from consultation and indicated that it would await a further request from EPA relating to EPA's potential future actions regarding Oregon's aluminum criteria. 

On January 31, 2013, EPA disapproved these seven revised aquatic life criteria under CWA Section 303(c). EPA disapproved the State's aluminum criteria because the State had not supplied a scientific rationale for the conditions under which the criteria would apply. The State and EPA have addressed the disapprovals for five of the criteria, but the State has not yet addressed EPA's 2013 disapproval of its freshwater criteria for aluminum. As a result, EPA is promulgating numeric freshwater acute and chronic criteria for aluminum in Oregon in this final rule in accordance with CWA Section 303(c)(3) and (c)(4). 

Sources of Toxic Pollutants to Surface Waters
According to the Agency for Toxic Substances and Disease Registry (ATSDR, 2008; p.184), aluminum is the most abundant metal in the Earth's crust, and it occurs ubiquitously in natural waters as a result of the weathering of aluminum-containing rocks and minerals. In addition to point source discharges, aluminum concentrations in surface waters can also be increased by several nonpoint sources, such as urban stormwater runoff, agriculture, forestry, acid mine drainage, and atmospheric deposition. 
Municipal and Industrial Dischargers
EPA's Integrated Compliance Information System NPDES (ICIS-NPDES) database identifies individual NPDES permits for 330 industrial dischargers and POTWs in Oregon, with 69 of those facilities being major facilities (USEPA, 2017c). For POTWs, major facilities are those that have a design flow of one million gallons per day (mgd) or greater, serve a population of 10,000 or more, are required to have a pretreatment program, or have the potential to cause significant water quality impacts. Industrial dischargers are classified as a major facility when the point value reaches a specific level which is based on several criteria, including toxic pollutant potential, flow volume, and water quality factors such as the impairment of the receiving water. One major facility and one minor facility (both industrial dischargers) in Oregon currently have effluent limitations for aluminum in their individual NPDES permits. 
ODEQ issues NPDES general permits for a wide variety of discharger categories, including cooling water/heat pumps, filter backwash, fish hatcheries, log ponds, boiler blowdown, seafood processing, suction dredges, stormwater from gravel mining, stormwater construction activities, industrial stormwater, oily stormwater runoff, tank cleanup and treatment of groundwater, washwater, noncontact geothermal water, and small municipal separate storm sewer systems (MS4). Currently there are no specific effluent limits or monitoring requirements for aluminum in these general permits. 
If the criteria are promulgated or otherwise adopted by the State, facilities not currently regulated for aluminum in their discharge may have new requirements added to future NPDES permit reissuances. Currently, there are 63 drinking water facilities with NPDES permits in Oregon (USEPA, 2017c). Aluminum sulfate (alum) and aluminum chlorohydrate/polyaluminum chloride are often used as coagulants in the treatment of drinking water. Coagulation/flocculation is a process used to remove turbidity, color, and some bacteria from water. In a 2011 national study of drinking water treatment plant residuals, including approximately half a dozen plants in Oregon, EPA estimated that long-term average aluminum concentrations discharged from drinking water plants ranged from 0.177 to 2.16 mg/L based on discharge monitoring report data (USEPA, 2011). 
In addition, metal salts (alum, sodium aluminate, or polyaluminum chloride) are commonly utilized in chemical treatment to remove phosphorus from wastewater. Chemical treatment is the most common method used for phosphorus removal to meet effluent concentrations below 1.0 mg/L. The required chemical dose is related to the liquid phosphorus concentration. Seven individually permitted wastewater treatment facilities in Oregon currently have effluent limitations for phosphorus and potentially use metal salts that contain aluminum in the treatment process.
Aluminum anodizing facilities may generate wastewater containing aluminum. In Oregon, there are no direct dischargers and 14 aluminum anodizing facilities that are indirect dischargers. The anodizing facilities may be given pretreatment requirements before discharging to POTWs as a result of the final aluminum criteria.
Urban Stormwater
Stormwater discharges are generated by precipitation runoff from land, pavement, building rooftops, and other surfaces. Urban stormwater from municipal and industrial areas may contribute metals, including aluminum, to surface waters. 
ODEQ regulates stormwater discharges from MS4s through NPDES permits. The MS4 permits require the discharger to develop and implement a stormwater management program (SWMP), with the goal of controlling pollutant discharges to the maximum extent practicable (MEP). In the State's Phase II general permits, ODEQ interprets the MEP requirement as being consistent with requiring all controls that are reasonable and available per ORS 468.020(2)(b). The management programs specify the best management practices (BMPs) that will be used in the following areas: public education and outreach; public participation and involvement; illicit discharge detection and elimination; construction runoff control; post-construction runoff control; and pollution prevention and good housekeeping. Permittees implement BMPs through an adaptive management or iterative approach. 
ODEQ has issued four area-wide "Phase I" MS4 permits covering Clackamas County Group; Portland Group; Multnomah County, Gresham Group, City of Salem, City of Eugene; and the Oregon Department of Transportation (ODEQ, 2017a). Phase I MS4s are those systems serving populations equal to or greater than 100,000. "Phase II" (or "small") MS4s are those systems serving municipalities with populations less than 100,000 located within Census Bureau-defined urbanized areas. There are 19 jurisdictions in Oregon covered by 15 Phase II MS4 permits (with some jurisdictions covered as co-permittees). These Phase II MS4s were previously covered by individual NPDES permits; however, ODEQ is in the process of developing a general permit to cover all current and future Phase II permittees (ODEQ, 2017b). Federal regulations also provide States the discretion to require other MS4s outside of urbanized areas to apply for a permit. 
Industrial dischargers, including those engaged in manufacturing, transportation, mining, and steam electric power generation, as well as scrap yards, landfills, certain sewage treatment plants, and hazardous waste management facilities may have stormwater requirements in their NPDES permits. ODEQ requires industrial stormwater dischargers to develop a Stormwater Pollution Control Plan (SWPCP) that includes a monitoring plan and discussion of the site controls that the discharger will implement to prevent stormwater pollution (ODEQ, 2017c). 
Agriculture and Forestry
Aluminum is the most plentiful metal in the Earth's crust (ATSDR, 2008). As such, activities that increase soil erosion may facilitate transport of aluminum to surface waters (see discussion in Canadian Council of Ministers of the Environment, 2003). Because agriculture and forestry often involve disturbances to substantial land areas, these sectors represent potentially significant sources of aluminum through erosion and runoff to surface waters. 
Aluminum may also be present in pesticides used in agriculture. Aluminum phosphide is a fumigant that is primarily used to control insects and rodents (USEPA, 1998). While data from the United States Geological Survey (USGS) (USGS, 2015) show that aluminum phosphide was used widely on Oregon agricultural lands as late as 1996, there was little to no use of the pesticide in the State from 1997 through 2014. 
The Oregon Department of Agriculture (ODA) Agricultural Water Quality Management Program (AWQMP) is charged with assisting the agriculture industry in reducing pollution from agricultural sources. The legal foundation of the AWQMP is the 1993 Agricultural Water Quality Management Act, also known as Senate Bill 1010 (SB1010). SB1010 provides the department with enforcement authority to address situations where corrective action is needed but is not voluntarily being taken by an operator. In those cases where a farmer or rancher refuses to take action, the law authorizes ODA to require corrective measures or use civil penalties to address the issue. 
ODA has identified 38 watershed-based agricultural water quality management areas. Each management area has a local advisory committee consisting of water quality specialists, local farmers, ranchers, and community leaders. All 38 local advisory committees have approved AWQMPs along with enforceable administrative rules to ensure all landowners do their part to avoid and resolve water quality problems. These plans must describe a program to achieve WQS and ensure that landowners will prevent and control water pollution from agricultural activities and soil erosion. See ODA (2016). 
Oregon's Forest Practices Act (FPA) regulates operations in the commercial forestry, and a large portion of the act aims to protect surface waters. Landowners are required to implement BMPs such as maintenance of riparian buffers, restricted use of forest roads during wet weather, and construction of roads using placement and methods to minimize runoff. See Oregon Department of Forestry (2017) and Oregon Forest Resources Institute (2011).
Mining
Active and abandoned or inactive mine sites, whether they are for the extraction of aluminum or other materials, can contribute to elevated aluminum concentrations in surface waters. Mining activities require large amounts of water to process ore, which results in discharged effluent, seepage from tailings storage facilities, and potential accidental releases. In addition to this direct contamination, exploratory activities such as road building and erosion of mineralized overburden during mine construction can also contribute to the contamination of surface water by heavy metals (Jennings et al., 2008). The leeching of heavy metals from excavated rock is especially problematic when mine wastes with high acid-generating potential are oxidized through exposure to surface conditions. 
Abandoned and inactive mines exist throughout Oregon, including mining districts in eastern Oregon, southwest Oregon, and the Willamette Basin. Over the last century, minerals and metals such as gold, silver, copper, zinc, and mercury have been extracted from these mines. State and federal agencies have identified over 150 mines in Oregon for possible further investigation or cleanup and have initiated assessment of about 95 mines (ODEQ, 2012a). For example, at the Formosa Mine Superfund site, acid mine drainage contributes to elevated heavy metal concentrations, negatively affecting downstream water quality and fisheries in the Middle Creek (USEPA, 2008). 
Atmospheric Deposition
Atmospheric deposition is a potential nonpoint source of aluminum to surface waters (ASTDR, 2008) and can contribute to surface water loads through either direct or indirect deposition. Direct deposition occurs when pollutants are deposited directly on surface waters from the atmosphere. Indirect deposition reflects the process by which metals deposited on the land surface are washed off during storm events and enter surface water through stormwater runoff. Atmospheric deposition is not directly addressed through any existing regulation and if it were a significant source could be indirectly addressed through TMDLs by inclusion in the load allocation portion of the TMDLs. 
 Water Quality
ODEQ (2014) classifies all surface waters into one of the following categories, based on available monitoring data:
 Category 2: Available data and information indicate that some designated uses are supported, and the water quality standard is attained. 
 Category 3: Insufficient data to determine whether a designated use is supported. Oregon further sub-classifies waters if warranted as:
 3B: Potential concern when data are insufficient to determine use support but some data indicate non-attainment of a criterion. 
 Category 4: Data indicate that at least one designated use is not supported but a TMDL is not needed. This includes:
 4A: TMDLs that will result in attainment of water quality standards have been approved. 
 4B: Other pollution control requirements are expected to address pollutants and will result in attainment of WQS. 
 4C: Impairment is not caused by a pollutant (e.g., flow or lack of flow are not considered pollutants). 
 Category 5: Data indicate a designated use is not supported or a water quality standard is not attained and a TMDL is needed. This category constitutes the Section 303(d) list that EPA approves or disapproves under the CWA. 
Oregon does not currently have any waters identified as impaired for aluminum, nor any aluminum TMDLs. Three sections of the Lower Columbia River were placed into Category 2 for aluminum (pursuant to the narrative criteria for toxic substances) based on an assessment conducted in 1998. No other aluminum classifications are itemized in ODEQ's 2012 Integrated Report database. 

Final Criteria
EPA is promulgating the federal criteria for aluminum to be applicable to Oregon's fresh waters. To protect aquatic life in Oregon's fresh waters, EPA is finalizing aluminum criteria for Oregon that incorporate by reference the calculation of the freshwater acute criterion, known as Criterion Maximum Concentration (CMC), and the freshwater chronic criterion, known as Criterion Continuous Concentration (CCC), values for a site using the final 2018 recommended national criteria. That means that the final CMC and CCC freshwater aluminum criteria values for a site shall be calculated using the 2018 Aluminum Criteria Calculator V.2.0 (Aluminum Criteria Calculator V.2.0.xlsx) or a calculator in R or other software package using the same 1985 Guidelines calculation approach and underlying model equations as in the Aluminum Criteria Calculator V.2.0.xlsx as established in EPA's Final Aquatic Life Ambient Water Quality Criteria for Aluminum 2018 (EPA 822-R-18-001). Consistent with the final 2018 recommended national criteria, EPA proposes to express the CMC as a one-hour average total recoverable aluminum concentration (in ug/L) and the CCC as a four-day average total recoverable aluminum concentration (in ug/L), and that the CMC and CCC are not to be exceeded more than once every three years. Calculating concentration expressions for the acute and chronic aluminum criteria using the MLR-based models requires input parameters for pH, dissolved organic carbon (DOC), and total hardness which characterizes ambient water conditions. The numeric outputs of the models for a given set of conditions will depend on the specific pH, DOC, and total hardness concentrations entered into the models. 
While criteria values for metals are typically expressed as dissolved metal concentrations, the CMC and CCC are expressed as total recoverable aluminum concentrations, because total recoverable aluminum is a more appropriate measurement of effects on aquatic life than dissolved aluminum. For additional information, refer to the Final Aquatic Life Ambient Water Quality Criteria for Aluminum 2018 (EPA 822-R-18-001). Using site-specific data for all MLR-based model input parameters results in the most accurate aquatic life criteria. EPA recommends using measured site-specific values for pH, DOC, and total hardness when data are available. EPA recommends three methods to identify criteria values based on site-specific data. The first method is to obtain criteria values by selecting one or more individual model outputs that will be protective of aquatic life under the full range of ambient conditions, including conditions of high aluminum toxicity. The second method is to calculate the criteria values by selecting the 10[th] percentile of the distribution of individual model outputs. The third recommended method to calculate the criteria would be to select the lowest model outputs (the lowest CMC and the lowest CCC) as the criteria values. All three recommended calculation methods assume sufficient site-specific data are available to characterize the site. The three methods above are just that, methods and not values. In order to analyze the impacts of the final criteria, and in the absence of sufficiently representative site-specific data to use the MLR-based models, EPA based its economic analysis on what is referred to below as the "economic analysis criteria," calculated using methodologies Oregon developed to implement its NPDES program and the second method, based on the 10th percentile of the distribution of individual model outputs, given data available to EPA at the time of this analysis.
Economic Analysis Criteria
EPA derived the economic analysis criteria for each of Oregon's Level III ecoregions based on MLR model results for ecoregion-specific measurements of pH, DOC, and total hardness. There are nine Level III ecoregions in Oregon: 1, 3, 4, 9, 10, 11, 12, 78, and 80 as shown below in Exhibit 3-1. 
Exhibit 3-1. Map of Nine Level III Ecoregions in Oregon
                                       
Source: https://www.epa.gov/eco-research/level-iii-and-iv-ecoregions-continental-united-states (USEPA, 2017b)

Exhibit 3-2 shows the economic analysis criteria, in both acute and chronic aluminum criteria values. In this report, EPA refers to these values as "economic analysis criteria" because the final rule provides Oregon with discretion on how to develop criteria values for different waters. To conduct this analysis, EPA used the second method, described above, to derive Level III ecoregional economic analysis criteria. These ecoregional economic analysis criteria were used to estimate potential impacts and costs, although the State may base criteria for a particular water body on the MLR model outputs using site-specific data. The economic analysis criteria values were calculated based on publicly available data from ODEQ and the USGS National Water Information System. The CMC and CCC for each Level III ecoregion were calculated from the 10[th] percentile of model output distributions of both measured and transformed data. EPA used these values to evaluate RP and potential effluent limits for each affected discharger. EPA also used these values to evaluate the potential for identifying waters as incrementally impaired. 
Actual impacts and costs will depend on the water quality criteria values that Oregon develops, which may differ from the economic analysis criteria. The economic analysis criteria were calculated to be protective under most conditions, including conditions of high aluminum toxicity, in each ecoregion. Therefore, actual criteria values, as implemented by Oregon, are likely to be higher (less stringent), leading to lower impacts and lower costs than are estimated here. 
In addition, as stated in Section 2, it is difficult to compare the numeric criteria to the baseline criteria without significant additional analysis. Without that analysis, EPA was only able to attribute costs identified by the economic analysis as being costs to comply with the numeric criteria (here, based on the economic analysis criteria). Although no Oregon waters are listed as impaired under the narrative criteria, if any waters are in fact impaired under the narrative criteria in the baseline, the costs for those waters are actually baseline costs and should be excluded from the incremental costs of the final criteria. Thus, the costs of the final rule could be overestimated.

Exhibit 3-2. Acute and Chronic Aluminum Criteria Values by Ecoregion that EPA Used to Evaluate Potential Impacts and Costs, also known as Economic Analysis Criteria
Ecoregion (code)
Aluminum Acute Criteria, CMC
(ug/L)
Aluminum Chronic Criteria, CCC
(ug/L)
Coast Range (1)
680
350
Willamette Valley (3)
870
440
Cascades (4)
600
350
Eastern Cascade Slopes (9)
1,100
600
Columbia Plateau (10)
1,400
840
Blue Mountains (11)
1,300
780
Snake River Plain (12)
3,000
1,200
Klamath Mountains (78)
1,300
780
Northern Basin & Range (80)
1,400
790

Criteria Used for Sensitivity Analysis
To illustrate the potential costs of the final aluminum criteria, EPA conducted the analysis using the economic analysis criteria, which are based on the 10[th] percentile of the model output distributions, as sufficient statewide site-specific data are not available for this analysis. This approach may result in conservative (i.e., low) criteria compared to criteria based on site-specific ambient data. To test the sensitivity of the economic analysis to the conservative estimate, EPA calculated alternative economic analysis criteria based on the 50[th] percentile of the model output distributions (listed in Exhibit 3-3). This sensitivity analysis attempts to illustrate criteria for sites with site-specific water quality parameters that lie in the median range of their respective ecoregion datasets. EPA also used the alternative economic analysis criteria to re-analyze the NPDES dischargers, as discussed in Section 4, and the potential impairments, as discussed in Section 5. 
Exhibit 3-3. Alternative Economic Analysis Criteria (50[th] percentile) used for Sensitivity Analysis
Ecoregion (code)
Acute Aluminum Criteria, CMC
(Al, ug/L)
Chronic Aluminum Criteria, CCC
(Al, ug/L)
Coast Range (1)
1,100
560
Willamette Valley (3)
1,400
650
Cascades (4)
1,000
630
Eastern Cascade Slopes (9)
1,700
910
Columbia Plateau (10)
2,300
1,200
Blue Mountains (11)
2,400
1,300
Snake River Plain (12)
3,400
1,400
Klamath Mountains (78)
2,000
1,000
Northern Basin & Range (80)
2,400
1,200

Expected Benefits
In the early 1980s, the impacts of acid rain and aluminum toxicity were observed in aquatic and terrestrial environments in specific regions of the U.S., most notably in the northeastern part of the country where aquatic systems had limited buffering capacity to prevent pH changes. Researchers observed that aluminum can be a major factor responsible for the demise of biotic communities since the toxicant becomes more soluble and potentially more toxic to aquatic biota at acidic pH (Gensemer and Playle 1999). Aluminum has no biologically important functions or beneficial properties to aquatic life and is therefore considered a non-essential metal (Eichenberger 1986; Exley 2003; Tchounwou et al. 2012; Williams 1999; Wood 1984, 1985). It has been identified as the cause of harmful effects on fish and wildlife, but is not a known teratogen, carcinogen or mutagen (Leonard and Gerber 1988). Overall, aquatic plants are generally insensitive to aluminum. Thus, the aluminum criteria being finalized are designed to protect aquatic life from the detrimental impacts from aluminum, in combination with other water chemistry conditions (i.e., pH, DOC, and total hardness). 

Bioavailability and thus toxicity of aluminum is affected by water chemistry parameters such as pH, DOC, and total hardness. The pH of waters affects aluminum speciation and solubility. Aluminum can sorb to DOC, such as humic and fulvic acids, and form organic aluminum complexes. An increase in DOC in waters reduces the bioavailability of aluminum to aquatic organisms as a result of this binding (Wilson 2012). Total hardness also has an effect on the toxicity of aluminum, as the cation Al[+3] competes with other cations present in water such as calcium (Ca[+2]) for uptake (Gensemer and Playle 1999). The observed effect of total hardness may be due to one or more of a number of usually interrelated ions, such as hydroxide, carbonate, calcium, and magnesium.

Numeric criteria allow measurement and quantification for purposes of monitoring and assessment, determining the need for discharge limitations, modeling source reduction needs on a watershed basis, and setting targets for allowable loads from point and nonpoint sources. Without such numeric targets, it is difficult to address both the potential for adverse effects from aluminum and the need for restoration efforts with a level of confidence.
  
This final rule will provide Oregon with the necessary tools to ensure aluminum levels in fresh waters in Oregon are maintained at a level to support a broad range of aquatic life, and in some cases reduce the levels of aluminum to protect aquatic ecosystems currently under stress. People hold value for these effects, but typically cannot express this value in the form of a market transaction. The maintenance and decrease in aluminum loadings is expected to ensure and improve protection of resident species, keep and enhance the general health of fish and invertebrate populations, and maintain and expand fisheries for both commercial and recreational purposes or use values. EPA expects that the final rule will increase values of the affected water resources. The expected benefits of water quality improvements fall into two broad categories: use benefits and nonuse benefits.  

Use benefits include the value of improved environmental goods and services used and valued by people. Degradation of water quality resulting from aluminum toxicity to aquatic life can reduce owner satisfaction with the property and the residential area in general. It can also adversely affect recreational opportunities. Improved (or maintained) water quality from reducing aluminum loadings is expected to enhance (or retain) the quality of living in the areas affected (or potentially affected), thereby resulting in welfare gain (or maintenance) to the resident populations. The final rule is also expected to enhance recreational uses of water resources affected by (or potentially affected by) aluminum toxicity, thereby resulting in welfare gain (or maintenance) to recreational users of these resources. Improved water quality may translate into two components of recreational benefits: (1) an increase in the value of a recreational trip resulting from a more enjoyable experience, and (2) an increase in recreational participation.

Even if no human activities (or uses) are affected by aluminum toxicity to aquatic life, such environmental impacts may still affect social welfare. For a variety of reasons, including bequest, altruism, and existence motivations, individuals may value the knowledge that water quality is being maintained, that ecosystems are being protected, and that populations of individual species are healthy completely independent of their use value. There are challenges associated with quantifying the relationship between changes in pollutant discharges and the improvements in societal well-being that are not associated with current use of the affected ecosystem or habitat. However, there is ample evidence to suggest these values exist, as evidenced, for example, by society's willingness to contribute to organizations whose mission is to purchase and preserve lands or habitats (although some portion of these donations may be motivated by use values). Recent economic literature provides strong support for the hypothesis that nonuse values are greater than zero for many types of environmental improvements. Moreover, when a substantial fraction of the population holds even small per capita nonuse values, these nonuse values can be very large in the aggregate. As stated by Freeman (1993), "there is a real possibility that ignoring non-use values could result in serious misallocation of resources." Both EPA's own Guidelines for Preparing Economic Analysis and OMB's Circular A-4, governing regulatory analysis, support the need to assess nonuse values (USEPA 2010a; USOMB 2003).



Methods for Estimating Potential Costs: Point Sources
This section describes the methods for estimating the potential costs to point sources associated with compliance with the final aquatic life criteria. Compliance costs for municipal and industrial point sources may result from changes to NPDES permit requirements and associated effluent limitations. The results of applying this methodology appear in Section 6. See Appendix A for additional details and calculations. 
EPA estimated costs to municipal and industrial dischargers under the final criteria. This section describes the identification of potentially affected permittees, the reasonable potential analysis (RPA), identification of limits under the final criteria, and estimation of costs to meet final criteria, using the "economic analysis criteria" as a proxy for the final criteria. 
Because of the lack of data for aluminum concentrations in point source discharges in the State, and because there is a potential for some waters to exceed the final criteria (discussed in Section 5), EPA decided to identify potential costs for point source dischargers that utilize aluminum in their operations (aluminum anodizing facilities, drinking water treatment plants, and wastewater treatment facilities) and, therefore, could potentially be affected by the final rule by presuming a need for additional controls or product substitution. The EPA supplements the typical analysis of potential point source costs based on RPA and effluent limit calculations using facility-specific data. This supplemental analysis appears in Sections 4.4, 4.5, and 4.6.
Identification of Potentially Affected Permittees
Factors that may affect the potential magnitude of control costs include flow and facility type. Larger flows are typically associated with the largest treatment costs, although per-unit costs may decrease due to economies of scale. An industrial facility may incur different costs than municipal wastewater facilities. EPA uses the designation of "major" for municipal discharges with a design flow of one mgd or more, and industrial discharges with a major rating point value (based on, among other factors, the presence of toxics and volume of discharge) over a specified value (USEPA, 2010a). Minor dischargers typically do not have monitoring requirements for toxic pollutants and may not contribute significantly to instream loads even if such pollutants were present in the effluent from these facilities. Thus, the potential for minor facilities to incur costs as a result of the final criteria is low.
EPA used the ICIS-NPDES database to identify facilities in Oregon whose NPDES permits contain effluent limitations or monitoring requirements for aluminum. As shown in Exhibit 4-1, there are two major (industrial) individually permitted facility that discharge wastewater to waters for which the final criteria would apply, and only one of those facilities (Northwest Aluminum Company) has available effluent monitoring data for aluminum. EPA also identified one minor (industrial) facility with effluent limits for aluminum: the Fujimi Corporation (OR0040339). However, this facility currently does not have any effluent data available. EPA was unable to further evaluate the two facilities (Georgia-Pacific Consumer Operations, LLC) with no available effluent data for aluminum.

Exhibit 4-1. NPDES Permitted Facilities Potentially Affected by the Final Rule
                               Facility NPDES ID
                                  Permit Name
                                  Major/Minor
                                     Type
OR0001708
Northwest Aluminum Company
                                     Major
                                  Industrial
OR0000795
Georgia-Pacific Consumer Operations, LLC
                                     Major
                                  Industrial
  Source: EPA's ICIS-NPDES database (USEPA, 2019).
Reasonable Potential Analysis
For the facility with available data, EPA conducted an RPA to determine whether there is RP for the discharge to cause or contribute to an excursion above any State water quality standard. ODEQ (2012b) provides permit writers with instructions for determining whether dischargers have RP to cause or contribute to an exceedance of WQS. ODEQ developed RPA and WQBEL calculation tools to assist permit writers in conducting the RPA. EPA used these tools to conduct the RPA for the potentially affected facility using the economic analysis criteria for the ecoregion where the facility is located as proxy values for the final criteria for aluminum. EPA incorporated approved regulatory mixing zone dilution factors specified in the fact sheet of the facility's NPDES permit.  
EPA used data on facility-specific effluent aluminum concentrations, receiving water characteristics, ambient parameter concentrations, and dilution factors in its RPA. EPA gathered these data from a variety of sources, including:
 Facility-specific permit fact sheets, available from ODEQ; and
 Discharge Monitoring Report data available through EPA's ICIS-NPDES. 
For the RPA, EPA projected the maximum effluent concentration of aluminum for the facility. For this projection, EPA applied multipliers to the maximum observed effluent concentration to determine a statistical maximum expected concentration of the pollutant (USEPA, 1991; Tables 3-1 and 3-2). The fewer the number of data points available for the analysis, the larger the multiplying factor. For the one facility analyzed, the multiplier was 1.0. 
When the projected maximum effluent concentration is less than both the acute and chronic criteria, there is no RP. If the concentration is greater than either the acute or chronic criteria, the permit writer must determine whether a mixing zone is appropriate. 
Where a regulatory mixing zone is appropriate, the ODEQ's RPA tool allows for two methods of incorporating the available dilution. The first is to use a factor derived from a mixing zone study and the second is to use a default factor calculated directly from the ratio of discharge and stream flow rates. The facility conducted a mixing zone study to determine the available dilutions at the boundary of the zone of initial dilution and the boundary of the mixing zone. Appendix A provides facility-specific information and the RP results (under economic analysis criteria conditions). 
Results of RPA
The economic analysis criteria for aluminum did not trigger RP for the NPDES permitted facility listed in Exhibit 4-1, and therefore did not result in new WQBELs. As the economic analysis criteria are more stringent than the alternative economic analysis criteria, which are based on the 50[th] percentile of model output distributions instead of the 10[th] percentile, the alternative economic analysis did not trigger RP for any NPDES permitted facilities. Thus, the results of this economic analysis are not sensitive to the percentile used from the model output distributions.
Supplemental Point Source Analysis: Identifying Control Scenarios
EPA's analysis of the available data for Northwest Aluminum Company indicates no RP for new effluent limits for aluminum. However, EPA is aware that some facilities may be using aluminum chemicals for nutrient removal, especially phosphorus. In addition, TMDLs for aluminum-impaired waters may result in new effluent limitations for certain facilities. Therefore, EPA conducted a supplemental point source analysis and identified the following potential control scenarios for such facilities: process optimization, technology change, and end-of-pipe treatment. Dischargers will typically pursue the most cost-effective means available for meeting effluent limits. Dischargers would most likely first evaluate whether they could optimize current treatment processes to be able to meet new effluent limits, prior to considering technology change, or additional end-of-pipe-treatment technologies. To conduct this analysis, EPA made many assumptions in the absence of facility-specific data; the results of this supplemental analysis are therefore highly speculative as compared to costs estimated using the RPA approach which is a facility-specific, data-based approach. The discussion below is provided for informational purposes. 
Process Optimization
The most cost-effective option is likely the adjustment of existing treatment (process optimization). This option would be most feasible where relatively small pollutant reductions are needed or monitoring data indicate that pollutant loads increase throughout the treatment process as a result of chemical additions or treatment techniques. 
Process optimization usually involves process analysis and process adjustment. Process analysis is an investigation of the performance-limiting factors of the treatment process and is a key factor in achieving optimum treatment efficiency. Performance-limiting factors for common wastewater treatment processes (e.g., sedimentation, activated sludge, filtration) may include a response to changes in wastewater quality, preventative maintenance activities, on-line analyzers, supervisory control and data acquisition system (SCADA) controls, process control testing, and additional operator training. The cost of process analysis includes the cost of additional or continuous monitoring throughout the treatment process, and a treatment performance evaluation. These costs vary based on the number of treatment processes analyzed and the magnitude of the reductions needed. 
Process adjustment includes activities short of adding new treatment technology units (conventional or unconventional) to the treatment train. For increasing pollutant removal efficiencies, process adjustment could include adjusting coagulant doses to increase settling, equalizing flow if pollutant concentrations change during wet weather events, optimizing filter operations and backwash cycles, installing on-line analyzers and process control systems or SCADA controls, and additional operator training. Several months of adjustments may be necessary to achieve a desired level of process optimization. In practice, the process adjustment necessary would be determined by the process analysis study. 
Process optimization costs depend on the pollutant needing reductions and efficiency of the existing treatment processes and operations. The effectiveness of process optimization largely depends on the efficiency of current operations, the existing treatment processes, and the fate and transport of aluminum through the treatment train. For example, if a facility is already well maintained and operated, implementing process optimization may not result in any significant additional aluminum reductions because the existing treatment processes are already performing at their optimum level. Also, because most conventional treatment technologies are designed to maximize removal of suspended solids, process optimization aimed at increasing those removal efficiencies may not result in significant reductions for aluminum if it exists primarily in dissolved form. Given the available information for the affected facilities, it is generally not possible to determine the reductions achievable on a plant by plant basis. However, EPA estimates below (see Section 4-5) a potential range of costs to dischargers by assuming all dischargers implement one of two possible control methods.
Technology Change
As discussed in Section 2.2.1, metal salts such as alum, sodium aluminate, and polyaluminum chloride are commonly used in chemical treatment to remove phosphorus from wastewater. If metal salt addition to the treatment process is the primary issue, replacing it with another flocculant, such as iron sulfate, may be sufficient to allow the facility to meet its new effluent limits on aluminum. This would still require completing bench scale, and possibly pilot scale, studies to select an alternative coagulant and establishing a dose that achieves the solids or phosphorus removal goal while not causing other difficulties for the treatment process (e.g., oxidized iron could foul process components).

Costs associated with a change in chemical in the treatment process depend on the pollutant needing reductions and efficiency of the existing treatment processes and operations with the replacement chemical. Costs also depend on the relative costs of the existing and replacement chemicals. Iron sulfate is a potential byproduct of certain industrial activities and may be readily available locally. Given the available information for the affected facilities, it is generally not possible to determine the reductions achievable with a switch in chemical; rather, a detailed bench-scale or pilot study would be necessary. 
End-of-Pipe Treatment
Aluminum removal can also be achieved through end-of-pipe treatment technologies, including ion exchange and reverse osmosis (RO). While the one NDPES-permitted discharger analyzed is not expected to require such additional treatment, these technologies are briefly described for informational purposes. 

The RO process removes pollutants from water using a semi-permeable membrane that allows water, and not dissolved compounds, to pass through its pores. The contaminated water is subjected to high pressures that force pure water through the membrane, leaving pollutants behind in a brine solution (NDWC, 1997). Theoretically, RO can remove nearly all dissolved metals and inorganic contaminants from water, as well as viruses and most organic contaminants, including some pesticides. However, the use of RO for treating wastewater is limited in practice, and therefore performance data for full-scale operations are scarce and limited, so pilot testing would be required to confirm removal efficiencies for aluminum. In addition, regulatory and climatic brine disposal constraints in Oregon could make this option economically infeasible for many dischargers, who would likely seek alternative compliance mechanisms. These constraints are described in Appendix C of EPA's Economic Analysis for the Proposed Rule: Aquatic Life Criteria for Copper and Cadmium in Oregon (USEPA, 2016). 

Ion exchange technology has been designed to remove dissolved and ionic forms of pollutants from wastewater using resins selected specifically for their contaminant removal properties. The treatment system relies on adsorption and chemical bonding of specific pollutant ions through ion exchange resins and works best when waste streams are pretreated to remove suspended and dissolved solids to enhance pollutant removal and eliminate competing ions (NDWC, 1997). Aluminum can be removed with strong acid cationic resins, week acid cationic resins, or selective chelating resins, depending on the pH and other pollutants targeted for removal. Resins should be selected and pilot-tested on a case-by-case basis. 
Supplemental point source analysis: Unit Costs 
Because of the lack of data for aluminum concentrations in point source discharges in the State, and because some waters were identified in Section 5 as being potentially incrementally impaired, EPA also identified potential costs for point source discharges that use aluminum in their operations (aluminum anodizing facilities, drinking water treatment plants, and wastewater treatment facilities) and, therefore, could potentially be affected by the final rule by presuming a need for additional controls or product substitution. This analysis supplements the typical analysis of potential point source costs based on RPA based on facility-specific data. 
Aluminum Anodizing Facilities
EPA identified 14 aluminum anodizing facilities in the State of Oregon, including two businesses operating from more than one location. These facilities discharge to wastewater systems, and the analysis below applies to all 14 facilities in the absence of facility-specific information to contradict the applicability. 
Background
Anodizing is an electrochemical process that converts the metal surface into a finish and aluminum works well with this process. Aluminum anodizing involves applying a protective oxide film to the base metal, usually by passing an electric current through an electrolyte bath in which the metal is immersed. These oxide coatings provide corrosion protection, decorative surfaces, a base for painting and other coating processes, and special electrical and mechanical properties. Wastewater generated during anodizing includes spent anodizing solutions, sealants, and rinse waters. 

The aluminum anodizers in Oregon all discharge wastewaters to local publicly owned treatment works (POTWs). In addition, all are subject to pretreatment requirements found in either 40 CFR Part 413 (Electroplating) or Part 433 (Metal Finishing), the distinction being when the facility began operation and if changes have occurred. Neither regulation contains pretreatment standards for aluminum; however, POTWs could potentially require the aluminum anodizer facilities to meet local limits (local limits for indirect dischargers are similar to water quality based effluent limits for point source dischargers). Depending on the operations conducted at each facility, treatment could consist of neutralization only. If other metals are utilized, the facilities may have chemical precipitation and sedimentation treatment systems in place. In addition, flow reduction practices such as countercurrent cascade rinsing can greatly reduce wastewater volume. Spray rinses and drag-out rinses are also effective flow reduction technologies.  

During development of the Metal Products & Machinery (MP&M) effluent limitations, guidelines, and standards (40 CFR Part 438), EPA conducted comprehensive wastewater sampling of aluminum anodizing facilities and other industries processing metal products. EPA also developed pollutant loadings and cost estimates based on model treatment technologies for operations within the MP&M industrial categories (USEPA, 2003).

The final criteria for aluminum in Oregon may result in Oregon POTWs establishing local limits for aluminum anodizers. Local limits are intended to prevent site-specific POTW and environmental problems due to non-residential dischargers. If local limits are applied to the aluminum anodizers, wastewater treatment upgrades may be necessary for compliance. 
Potential Treatment Upgrades
EPA identified two options for potential treatment upgrades that may be required based on the final aluminum criteria and developed cost estimates for these upgrades. 
 The first upgrade option assumes that the facility already has a chemical precipitation/flocculation system in place so that only flow reduction technologies and countercurrent cascade rinsing are necessary. 
 The second upgrade option assumes that facilities only have a neutralization system in place and would require a wastewater treatment system incorporating flow reduction technologies, as well as chemical precipitation/flocculation. 
Assumptions 
Option 1 
EPA used the cost equations derived for countercurrent cascade rinsing technologies (USEPA, 2003), and adjusted the costs from 2001 to 2018 dollars using the Gross Domestic Product (GDP) Implicit Price Deflator. The original equations (USEPA, 2003) are listed below for reference purposes.

Annual Costs=0.004xTankVol+0.2243xDPYxHPDx0.047
Capital Costs=6.047xTankVol+3784.3 (tank, piping, and pump)

Where, 
   TankVol = tank volume 
   DPY = days per year
   HPD = hours per day

EPA was unable to identify site-specific assumptions for the 14 anodizing facilities, and assumed the same conditions across all facilities, as identified below:  
 Tank volume of 800 gallons, based on best professional judgment for anodizers; 
 Five rinse tanks per line, based on best professional judgment for anodizers; and
 Operation 8 hours per day, 5 days per week, with a two-week shut-down per year based on expected business operation. 

Option 2 
EPA used the capital and operation and maintenance (O&M) costs derived for non-chromate anodizers during development of the MP&M ELG (USEPA, 2003) and adjusted the costs from 2001 to 2018 dollars using the Gross Domestic Product (GDP) Implicit Price Deflator. Aggregate costs in that document are provided for different subcategories, including non-chromate anodizing facilities. EPA divided the total costs for the subcategory by the number of facilities in that subcategory to determine an average facility cost for the non-chromate anodizer subcategory. EPA was unable to identify site specific information for the 14 aluminum anodizers in Oregon, therefore the average MP&M costs for the non-chromate anodizing subcategory was applied to all 14 anodizing facilities in Oregon.  

EPA's cost estimates for Option 2 assume the following: 
 Facilities are only using aluminum. If they are using other metals (e.g., chromium or zinc), the facilities could already be subject to pre-treatment standards, and costs would be lower than estimated here.
 Facilities currently only neutralize their wastewater and require a treatment system upgrade for countercurrent cascade rinses, chemical precipitation, and settling. If current wastewater treatment already includes some of these elements, upgrade costs could be lower than estimated here. 
Cost Estimates 
Exhibit 4-2 presents the capital, O&M, and total costs per facility, in 2018$. Total annual cost per facility is calculated as the capital cost per facility, annualized over the expected life of the capital equipment of 20 years (Metcalf & Eddy, 2014; USEPA, 1979) using 3% and 7% discount rates, plus the O&M costs.

        Exhibit 4-2. Costs for Aluminum Anodizers, Both Options (2018$)
                                    Option
                               Treatment Upgrade
                                 Capital Cost
                                 Per Facility
                           O&M Cost Per Facility
                        Total Annual Cost per Facility
                              (3% discount rate)
                        Total Annual Cost per Facility
                              (7% discount rate)
                                       1
Countercurrent cascade rinsing only
                                   $ 59,700
                                     $350
                                    $4,400
                                    $6,000
                                       2
Countercurrent cascade rinsing, chemical precipitation, and settling
                                   $ 180,200
                                   $ 479,600
                                   $ 491,700
                                   $ 496,600

Applying the average costs to all 14 anodizing facilities in Oregon, the total annual cost for all 14 facilities ranges from $61,600 to $6.88 million using a 3% discount rate over a 20-year period, and from $84,000 to $6.95 million using a 7% discount rate over a 20-year period. The cost estimates here likely reflect a larger range than would be present in reality because facilities are more likely to employ a mix of options rather than all facilities using either Option 1 or Option 2.
Drinking Water Treatment Plants
EPA identified 61 water treatment plants (WTPs) in Oregon covered under a general permit issued by the ODEQ. EPA then determined that four of the 61 facilities were duplicates and excluded the duplicate entries from the analysis, which leaves 57 WTPs potentially affected by the final aluminum criteria. 
Background
Drinking water sources are subject to contamination and require appropriate treatment to provide safe water to the communities being served. Public drinking water systems use various methods of water treatment including coagulation and flocculation, sedimentation, filtration, and disinfection.  

Coagulation and Flocculation
Coagulation and flocculation are often the first steps in water treatment. Chemicals with a positive charge (e.g., alum, ferric chloride) are added to the water. The positive charge of these chemicals neutralizes the negative charge of dirt and other dissolved particles in the water. When this occurs, the particles bind with the chemicals and form larger particles, called floc. 

Sedimentation
During sedimentation, floc settles to the bottom of the water supply, due to its weight. This settling process is called sedimentation.

Filtration
Once the floc has settled to the bottom of the water supply, the clear water on top will pass through filters of varying compositions (sand, gravel, and charcoal) and pore sizes, in order to remove dissolved particles, such as dust, parasites, bacteria, viruses, and chemicals.

Disinfection
After the water has been filtered, a disinfectant (e.g., chlorine, chloramine) may be added in order to kill any remaining parasites, bacteria, and viruses, and to protect the water from pathogens when it is distributed to homes and businesses.

Wastewater discharges from WTPs utilizing the treatment system described above are often comprised of filter backwash, residuals from solids settling, and flushing of intakes and equipment. 
 
Drinking water systems utilizing this method of treatment often use aluminum salts as a coagulant. The final aluminum criteria in Oregon may result in Oregon drinking water systems needing to reduce aluminum concentrations in their wastewater discharges. 

Per the surface water treatment rule, surface water systems (including systems using groundwater under the influence of surface water) are required to filter and disinfect the source water prior to distribution (40 CFR § 141.70). However, the filtration requirements do not apply to groundwater systems. Of the 57 WTPs potentially affected by the final aluminum criteria, eight systems are groundwater systems. If a system is not filtering, it is unlikely to be using aluminum or other coagulants prior to distribution. These systems were excluded from the analysis, resulting in 49 systems potentially affected by the final aluminum criteria. 
Potential System Modifications
WTPs have several options for reducing aluminum concentrations in their wastewater discharges. Typically, wastewaters are routed to settling basins prior to discharge and additional treatment such as filtration could be added prior to discharge. In addition to aluminum salts, iron salts are also used as coagulants in drinking water treatment, usually at a lower cost than additional treatment. WTPs are likely to pursue the most cost-effective compliance mechanism, therefore, EPA costed product substitution for these facilities. This substitution from aluminum to ferrous coagulants uses the same drinking water treatment mechanisms. Ferrous salts do produce more sludge than aluminum salts, so the additional costs associated with sludge disposal were examined in addition to the raw material costs.
Assumptions
EPA assumed the same conditions across all facilities, as identified below:  
 All WTPs currently use alum as a coagulant;
 Facilities are backwashing filters and discharging 365 days per year;
 Alum and iron (ferric chloride) salt coagulants are used at a concentration of 20 milligrams (mg) per liter (L) of water treated (http://www.filtronics.com/coagulation); 
 Conversion from alum to ferric chloride will result in a 20% increase in sludge production (Michigan DEQ, Undated); 
 Facilities use a dewatering process to reduce the amount of liquid sludge disposal, which will result in disposal of only 10% of the total wet sludge. 

EPA conducted a study on drinking water plant residuals (USEPA, 2011) that provides cost information for systems based on the population range served by the WTP. The study provides average treatment plant flows and average sludge generated for each population range. EPA classified the WTPs in Oregon using the same population ranges and tabulated the total number of WTPs for each population range. Exhibit 4-3 provides the number of WTPs and average treatment plant flow for each population range. 
                                       
               Exhibit 4-3. WTPs in Oregon by Population Served

                                       A
                                       B
                                       C
Population Range
                           Number of WTPs in Oregon
                     Average Treatment Plant Flow (mgd)[1]
                                    Average
                              Sludge Volume (gpd)
Less than 3,300
                                      27
                                                                           0.23
                                                                            770
3,301 to 10,000
                                       7
                                                                            0.7
                                                                          2,000
10,001 to 25,000
                                       7
                                                                            2.1
                                                                          5,300
25,001 to 50,000
                                       4
                                                                              5
                                                                         12,100
50,001 to 75,000
                                       2
                                                                            8.8
                                                                         19,800
75,001 to 100,000
                                       1
                                                                             13
                                                                         28,600
100,001 to 500,000
                                       1
                                                                             27
                                                                         56,200
      [1] EPA used the average treatment plant flows for each population range from EPA drinking water plant residuals study (USEPA, 2011).
Cost Estimates
EPA used coagulant costs of $409 and $520 per ton of alum and ferric chloride, respectively when adjusted from 2015 to 2018$ using the GDP Implicit Price Deflator (City of Lima, 2015). EPA also estimated sludge disposal costs at a rate of $75/wet ton (Kelley et al., 2018). EPA calculated total costs for chemical use and sludge disposal for both alum and ferric chloride for each facility representative of the average treatment flow within its population range (Exhibit 4-4). 

Exhibit 4-4. Annual Costs of Chemicals and Sludge Disposal for WTPs by Range of Population Served (2018$)
                                       
                                       A
                                       B
                                       C
                                       D
                                       E
                                       F
Population Range
                                Alum Chemical 
                                      (1)
                    Sludge Disposal for WTP Using Alum (2)
                           Ferric Chloride Chemical
                                     (3) 
               Sludge Disposal for WTP Using Ferric Chloride (4)
                 Total Cost Per Facility in Each Category (5)
                   Total Cost for All Facilities in Category
Less than 3,300
                                                                        $2,900 
                                                                        $8,800 
                                                                        $3,600 
                                                                       $10,500 
                                                                        $2,400 
                                                                       $64,800 
3,301 to 10,000
                                                                        $8,700 
                                                                       $22,800 
                                                                       $11,100 
                                                                       $27,400 
                                                                        $7,000 
                                                                       $49,000 
10,001 to 25,000
                                                                       $26,100 
                                                                       $60,500 
                                                                       $33,200 
                                                                       $72,600 
                                                                       $19,200 
                                                                      $134,400 
25,001 to 50,000
                                                                       $62,200 
                                                                      $138,100 
                                                                       $79,000 
                                                                      $165,800 
                                                                       $44,500 
                                                                      $178,000 
50,001 to 75,000
                                                                      $109,400 
                                                                      $226,000 
                                                                      $139,100 
                                                                      $271,200 
                                                                       $74,900 
                                                                      $149,800 
75,001 to 100,000
                                                                      $161,600 
                                                                      $326,500 
                                                                      $205,500 
                                                                      $391,800 
                                                                      $109,200 
                                                                      $109,200 
100,001 to 500,000
                                                                      $335,700 
                                                                      $641,500 
                                                                      $426,800 
                                                                      $769,900 
                                                                      $219,500 
                                                                      $219,500 
                                                                   Grand Total:
                                                                       $904,700
 Calculated as Average Treatment Plant Flow (column B of Exhibit 4-3) * 20 mg/L chemical addition rate * $409 cost per ton (with 10[6] gallons per million gallons, 3.785 L/gallon, 1.102 x 10[-9] tons/mg, and 365 days/year unit conversion factors), and rounded to nearest $100.
 Calculated as Average Sludge Volume (column C of Exhibit 4-3) * 8.34 pounds/gallon water * 10% liquid sludge reduction factor * $75 cost per ton of wet sludge disposed (with 365 days/year and 2,000 lbs/ton unit conversion factors) and rounded to nearest $100.
 Calculated as Average Treatment Plant Flow (column B of Exhibit 4-3) * 20 mg/L chemical addition rate * $520 cost per ton (with 10[6] gallons per million gallons, 3.785 L/gallon, 1.102 x 10[-9] tons/mg, and 365 days/year unit conversion factors) and rounded to nearest $100.
 Calculated as Average Sludge Volume (column C of Exhibit 4-3) * 8.34 pounds/gallon water * 10% liquid sludge reduction factor * $75 cost per ton of wet sludge disposed * 1.2 to reflect a 20% increase in sludge production using ferric chloride (with 365 days/year and 2,000 lbs/ton unit conversion factors) and rounded to nearest $100.
 Calculated as ([column C + column D]  -  [column A + column B])
 Calculated as Number of WTPs in Oregon by category (column A of Exhibit 4-3) * (column E of Exhibit 4-4)

EPA applied the costs in Exhibit 4-4 to all Oregon facilities by population range and estimated that changing from alum to ferric chloride could result in approximate total annual costs of $904,700 for the 49 WTPs analyzed.
Wastewater Treatment Facilities
In the State of Oregon, EPA identified eight POTWs, including two minor facilities, that have effluent limits for total phosphorus in their discharge permits and may use alum for phosphorus removal. 
Background
Wastewater treatment plants seeking to treat for total phosphorus often use chemical precipitation followed by filtration to remove phosphorus from the wastewater prior to discharge. Chemical treatment for phosphorus removal involves the addition of metal salts to react with soluble phosphate to form solid precipitates that are removed by solids separation processes including clarification and filtration. EPA analyzed POTWs in Oregon with effluent limitations for total phosphorus. EPA reviewed the treatment descriptions in the available NPDES permit fact sheets and facilities plans and identified five POTWs that specified use of alum for phosphorus removal. One major facility relies on biological rather than chemical phosphorus removal prior to discharge to a wetland-based Natural Treatment System, so the facility is not relying on alum, and EPA did not include it in the analysis. EPA assumed that the two remaining facilities (the two minors) may substitute ferrous coagulants for aluminum coagulants for phosphorus removal.

Advantages for using chemical treatment for phosphorus removal include reliability, low effluent concentrations, and ability to retrofit existing treatment systems. Disadvantages are the costs of the chemical feed system, chemicals, and increased sludge production.

The final aluminum criteria in Oregon may result in Oregon POTWs re-examining the use of aluminum salts in the phosphorus removal systems at their facilities. 
Potential System Modification
For this analysis, EPA costed product substitution for the seven facilities potentially relying on alum for phosphorus removal. Iron salts can be used as alternative coagulant to alum in phosphorus removal. This substitution from aluminum to ferrous coagulants utilizes the same existing chemical precipitation and filtration treatment mechanisms. Ferrous salts produce more sludge than aluminum salts, but the dosage of ferric chloride is less than that of alum due to the stoichiometry of the phosphorus removal chemical process (Kruger, Undated). EPA estimated costs associated with sludge disposal as well as raw material costs.
Assumptions
EPA assumed the same conditions across all facilities, as identified below:  
 The POTWs are treating for phosphorus removal 365 days per year;
 Alum coagulant is used at a concentration of 40 mg/L of water treated (Michigan DEQ, Undated); 
 Ferric chloride is used at a concentration of 20 mg/L of water treated; (Michigan DEQ, Undated);
 Conversion from alum to ferric chloride for a drinking water treatment facility will result in a 20% increase in sludge production (Michigan DEQ, Undated); and 
 For every pound of alum that is used by a POTW, 0.4 pounds of sludge is generated, and for every pound of ferric chloride that is used, 0.6 pounds of sludge is generated (Kruger, Undated). 
Cost Estimates
EPA used the design flow for each POTW to determine potential costs, and estimated sludge disposal costs at a rate of $75/wet ton (Kelley et al., 2018). EPA used coagulant costs of $409 and $520 per ton of alum and ferric chloride (City of Lima, 2015), respectively, when adjusted from 2015 to 2018$ using the GDP Implicit Price Deflator. EPA calculated costs for the use of alum and disposal of the sludge generated with alum and compared those to ferric chloride. Exhibit 4-5 provides cost estimates for each of the seven POTWs. 

 Exhibit 4-5. Annual Costs of Chemicals and Sludge Disposal for POTWs (2018$)
                                   NPDES ID
                                     NAME
                               Design Flow (mgd)
                                      (1)
                              Alum Chemical Cost
                                      (2)
                        Sludge Disposal Cost Using Alum
                                      (3)
                         Ferric Chloride Chemical Cost
                                      (4)
                  Sludge Disposal Cost Using Ferric Chloride
                                      (3)
                             Total Cost Using Alum
                       Total Cost Using Ferric Chloride
OR0020559
City of Cottage Grove
                                      1.8
                                                                       $44,800 
                                                                        $3,300 
                                                                       $28,500 
                                                                        $2,500 
                                                                        $48,100
                                                                        $31,000
OR0020729
City of Canyonville
                                      0.5
                                                                        $12,400
                                                                           $900
                                                                         $7,900
                                                                           $700
                                                                        $13,300
                                                                         $8,600
OR0022730
City of Glendale
                                     0.999
                                                                        $24,800
                                                                         $1,800
                                                                        $15,800
                                                                         $1,400
                                                                        $26,600
                                                                        $17,200
OR0026255
City of Ashland
                                      2.3
                                                                       $57,200 
                                                                        $4,200 
                                                                       $36,400 
                                                                        $3,100 
                                                                        $61,400
                                                                        $39,500
OR0028118
CWS-Durham STP
                                     22.6
                                                                      $561,900 
                                                                      $416,200 
                                                                      $357,200 
                                                                       $30,900 
                                                                       $603,100
                                                                       $388,100
OR0029777
CWS - Rock Creek STP
                                      39
                                                                      $969,700 
                                                                       $71,100 
                                                                      $616,400 
                                                                       $53,300 
                                                                     $1,040,800
                                                                       $669,700
OR0034002
City of McMinnville
                                      5.6
                                                                       $139,200
                                                                        $10,200
                                                                        $88,500
                                                                         $7,700
                                                                       $149,400
                                                                        $96,200
 Design flow.
 Calculated based on design flow (Column 1), alum coagulant usage at 40 mg/L of water treated, and estimated cost of chemical per ton.
 Calculated based on an annual cost for disposal of $75/wet ton.
 Calculated based on design flow (Column 1), ferric chloride coagulant usage at 20 mg/L of water treated, and estimated cost of chemical per ton.

Exhibit 4-5 implies that changing from alum to ferric chloride may result in annual cost savings at each POTW based on the lower chemical costs due to the lower dosing amounts for ferric chloride. This is a counterintuitive result, because facilities have an incentive to change their processes to take advantage of cost savings that does not depend on this final rule. Although the analysis would suggest potential cost savings by switching to ferric chloride, EPA assumes that, in the absence of the final rule, the facilities would already be using the lowest cost treatment. For example, facility-specific factors that EPA is unable to take into consideration may cause alum to be the lowest cost treatment approach. In addition, the facilities may have modified their treatment system since the permits were issued, and EPA could not confirm whether the two minors are using alum as part of their phosphorus removal process. Therefore, EPA is not subtracting these cost savings from the other costs associated with this final rule and assumes that the final rule would result in no change in cost for POTWs. 
Supplemental Point Source Analysis: Total Costs
Total annual cost estimates for point source compliance under the supplemental analysis are summarized in Exhibit 4-6.
                                       
Exhibit 4-6. Total Annual Costs for Point Source Compliance (Supplemental Analysis) (2018$)
                                 Facility Type
                     Total Annual Cost for all Facilities
                             (3% discount rate)[1]
                     Total Annual Cost for all Facilities
                             (7% discount rate)[1]
Aluminum Anodizing Facilities
                            $0.06 to $6.88 million
                            $0.08 to $6.95 million
WTPs
                                 $0.90 million
                                 $0.90 million
POTWs
                                     $0[2]
                                     $0[2]
                                                                       Total[3]
                             $1.0 to $7.8 million
                             $1.0 to $7.9 million
[1] Discount rate does not affect costs for WTPs or POTWs because EPA does not expect those facilities to incur capital costs. 
[2] Zero is used as the estimate in lieu of the estimated cost savings.
[3] Totals may vary due to rounding. 
            

Methods for Identifying Potential Costs: Incremental Impairments 
EPA compared available water quality measurements for aluminum against the economic analysis criteria (and the alternative economic analysis criteria) to estimate the potential water quality impairment for aluminum. As noted in Section 2.3, there are no current CWA Section 303(d) listed aluminum impairments in Oregon; as such, any waters that would be newly identified as potentially impaired for aluminum under the economic analysis criteria are considered as incremental impairments. As discussed in Section 2, the regulatory baseline for evaluating the potential impact of the final criteria includes existing requirements for nonpoint sources and stormwater dischargers to implement BMPs as well as wasteload allocations for point sources as part of future TMDLs to address sources of impairment other than aluminum. Therefore, the aluminum concentrations in existing water quality data may not reflect future conditions following full implementation of baseline BMPs. If any waters identified as incrementally impaired in this analysis are in fact impaired under the narrative criteria in the baseline due to aluminum exposure (for example identified as a result of Whole Effluent Toxicity (WET) testing), the costs estimated here are an overestimate.
Available Data
EPA used available instream monitoring data to assess the potential impacts that the economic analysis criteria may have on water quality attainment. EPA obtained total recoverable aluminum monitoring data between August 2000 and May 2017 (described in a technical support document in the docket for this rule). This is a reasonable timeframe to assess the current conditions of all of the State's regularly monitored waters. 
EPA's data review and validation included conversions to 1) transform the sample results into units comparable to the economic analysis criteria, and 2) map each sample station into one of the State's nine Level III ecoregions based on latitude and longitude. Additionally, some monitoring stations had multiple samples reported in one day. In those cases, EPA took only the highest available concentration for that day and used that concentration as a single observation (dropping all lower concentrations that occurred at the same station on the same day). 
Exhibit 5-1 summarizes the available aluminum monitoring data pursuant to these data conversions and filtering. There are 826 samples across 260 stations, including 33 nondetects. For these observations, EPA assigned a concentration equal to the minimum reporting level (MRL).
Exhibit 5-1. Summary of Available Aluminum Monitoring Data by Ecoregion
Ecoregion[1]
Stations
Samples[2]
Detects
Code
Name



1
Coast Range
                                                                             49
                                                                            103
                                                                             92
3
Willamette Valley
                                                                            104
                                                                            400
                                                                            396
4
Cascades
                                                                              6
                                                                             18
                                                                             17
9
Eastern Cascade Slopes
                                                                             18
                                                                             60
                                                                             52
10
Columbia Plateau
                                                                             12
                                                                             27
                                                                             22
11
Blue Mountains
                                                                             25
                                                                             75
                                                                             75
12
Snake River Plain
                                                                              6
                                                                             20
                                                                             20
78
Klamath Mountains
                                                                             23
                                                                             68
                                                                             64
80
Northern Basin & Range
                                                                             17
                                                                             55
                                                                             55
Total

                                                                            260
                                                                            826
                                                                            793
[1] Ecoregions correspond to Level III Ecoregions of the Continental United States from EPA (USEPA, 2017b). 
[2] Includes only monitoring data for total recoverable concentrations and the highest observed concentration for a single station on a single day.

Identifying Exceedances
The ODEQ's 303(d) listing policy requires at least two exceedances of water quality standards with sufficient data available for a water to be classified as impaired and needing a TMDL (Category 5). EPA compared the water quality monitoring data to the ecoregion-specific economic analysis aluminum criteria estimates for the final rule. Based on this comparison, EPA assigned each sampling station to one of the following categories:
Potentially impaired: two or more sample results collected on different sampling dates exceeded the economic analysis criteria;
Impairment uncertain: only one sample result exceeded economic analysis criteria; and
Impairment unlikely: no sample result exceeded economic analysis criteria. 
Since the CCC is the more stringent of the criteria, EPA compared the concentrations to those ecoregion-specific criteria to determine the number of excursions per station. Exhibit 5-2 summarizes the results of this analysis. 
Exhibit 5-2. Number of Potential Impairments by Ecoregion Based on Aluminum Monitoring Data
Ecoregion[1]
Potentially Impaired
Impairment Uncertain
Impairment Unlikely 
Code
Name



1
Coast Range
                                                                              3
                                                                             15
                                                                             31
3
Willamette Valley
                                                                             36
                                                                             20
                                                                             48
4
Cascades
                                                                              0
                                                                              1
                                                                              5
9
Eastern Cascade Slopes
                                                                              3
                                                                              3
                                                                             12
10
Columbia Plateau
                                                                              1
                                                                              1
                                                                             10
11
Blue Mountains
                                                                              0
                                                                             11
                                                                             14
12
Snake River Plain
                                                                              5
                                                                              0
                                                                              1
78
Klamath Mountains
                                                                              2
                                                                              8
                                                                             13
80
Northern Basin & Range
                                                                              4
                                                                              7
                                                                              6
Total
                                                                             54
                                                                             66
                                                                            140
[1] Ecoregions correspond to Level III Ecoregions of the Continental United States from EPA (USEPA, 2017b). 

Based on this comparison, there is a potential impairment status for 54 stations, as shown in Exhibit 5-2. Without additional information about how Oregon might categorize stations into individual water bodies for the purpose of defining reaches impaired for aluminum, it is possible that the 54 stations represent the upper bound on the number of incremental TMDLs for the set of waters with monitoring data. One reason for this assumption is the likelihood that water bodies have multiple monitoring stations, which would mean that there are fewer than 54 impaired water bodies. For example, 13 identified impairments are on the Tualatin River, so the ODEQ potentially could develop a single TMDL for the combined river stretches, rather than 13 separate TMDLs corresponding to each monitoring station. Another reason EPA views this assumption as an upper bound is that the State could choose to combine the development of TMDLs for waters with similar causes of impairment into a single effort, thereby achieving efficiencies and reducing costs. However, EPA does not have a method to estimate the potential for such savings at this time. Additionally, after listing a water body as impaired with respect to aluminum based on ecoregional criteria, Oregon potentially could conduct additional monitoring before developing a TMDL. Because the economic analysis criteria are conservative, the additional monitoring could result in the State identifying the water body as unimpaired with respect to aluminum, and the State would therefore not develop a TMDL. On the other hand, waters in the "impairment uncertain" category may eventually end up in the impaired category, which could result in the estimate of 54 TMDLs being an underestimate. In summary, the actual number of incremental TMDLs could be higher or lower than the estimated 54 TMDLs. After considering all of these factors, EPA has concluded that the factors that would result in 54 TMDLs being an overestimate are more persuasive than the factor which would result in the range being an underestimate, and therefore, EPA considers the 54 TMDLs to be its best estimate of waters incrementally impaired under the final rule.
Exhibit 5-3 shows the location of the stations identified as "potentially impaired," "impairment uncertain," and "impairment unlikely." 

Exhibit 5-3. Location of Potential Impairments Based on Aluminum Monitoring Data
                                       


Identifying Potential Control Actions and Costs
The causes of elevated aluminum levels in surface water may be numerous and variable and could be natural or anthropogenically influenced. One major reason for increased aluminum in surface waters may be increased acidification of catchment basin soils through atmospheric deposition of acidic compounds. This can have the effect of remobilizing aluminum that had precipitated out (Driscoll, 1985). The large number of potentially impaired locations in eastern Oregon are generally within a region of acid sensitive waters. Elevated acidity also enhances the toxicity of aluminum. Thus, a principal control strategy would likely be about reducing acidity in these waters and watersheds. This could result in load allocations to acid deposition in TMDLs or possibly acknowledgement of water quality limitations through use attainability analysis. A first step in any effort would be to collect monitoring data to evaluate aluminum levels in context of the full capabilities of the aluminum criteria rather than to rely on the conservative values derived for this study. 
Nonpoint sources and stormwater may be at least partially responsible for these potential impairments. Typical control strategies for these sources include: 
Agricultural and forest lands  -  changes in sediment and erosion controls beyond those specified under existing State and federal regulations and plans;
Mining  -  cleanup and remediation beyond those actions being taken under existing State and federal plans, including excavation and onsite capping of contaminated soils, capping of onsite solid waste mining debris, regrading of tailings to mitigate mass wasting and off-site migration, and abatement and mitigation of physical hazards; and
Stormwater discharges  -  increased or additional nonstructural BMPs (e.g., institutional, educational, or pollution prevention practices designed to limit generation of runoff or reduce the pollutant load in runoff); and structural controls (e.g., engineered and constructed systems designed to provide water quantity or quality control). 
However, these strategies typically address overland flow of particulates with adsorbed metals. These types of strategies may or may not be relevant to the underlying cause of elevated aluminum in surface waters, which may or may not reflect levels of concern for aquatic life depending on levels of unmeasured pH, DOC, and total hardness. It is likely that the underlying cause of potential impairments, should they be confirmed, are related to atmospheric deposition of acidic compounds and limited buffering capacity of soils. To the extent that these nonpoint source control strategies described above would be implemented, it is also most likely to be for situations where there are elevated levels of multiple metals, and thus compliance with exiting water quality criteria would put these costs within the baseline. For these reasons, EPA did not quantify additional costs to attribute to this proposal for nonpoint source controls. 
If there is incremental water quality impairment under the final criteria, there would be incremental costs for TMDL development. EPA reports that the national average cost to develop a TMDL for a single source of impairment ranges from $27,000 to $29,000 in 2000 dollars (USEPA, 2001a), or $38,000 to $41,000 when updated to 2018 dollars. The inflation adjustment does not account for any increased cost of TMDL development resulting from increased sophistication and expectations surrounding TMDL development that have accrued since the time of this estimate. 

The USEPA (2001b) report then estimated an additional set of unit burdens and associated costs to represent efficiencies associated with States coordinating the development of multiple TMDLs for multiple water bodies and similar pollutants. The USEPA (2001b) report validated the estimates by comparing the estimated burden and costs to the actual or estimated burden or costs for 131 TMDLs, the Washington State Workload Model, a May 1996 EPA report containing 14 case studies of TMDL cost, the Gap State Water Quality Management Resource Needs Model and commonly cited high-cost TMDLs (see USEPA 2001b for more details). Finally, the original range of $27,000 to $29,000 reflects a 5- to 10-year transition period over which States are assumed to fully achieve the cost efficiencies that can be realized by clustering water bodies and impairment causes when they develop TMDLs.
Using the economic analysis criteria, EPA determined that there was a potential for an increase in the number of monitoring stations meeting criteria for Category 5  -  impaired waters. If the estimated number of impairments is an indication of the potential increase in the number of TMDLs, then the total costs for TMDL development would range from $2.05 million (54 TMDLs x $38,000) to $2.21 million (54 TMDLs x $41,000). TMDL development costs are one-time costs that would be spread out over several years, because it is not realistic to assume that Oregon would conduct all these analyses simultaneously. 
EPA assumed a uniform distribution of costs over a 10-year time period for TMDL development, covering both the time it takes to develop a TMDL and the assumption that Oregon would not start work on all TMDLs simultaneously. The annual costs for TMDL development would range from $205,000 to $221,000 for the development of 54 TMDLs over 10 years, dropping to $0 thereafter. Given that TMDL costs are one-time costs, the fully annualized costs (into perpetuity) are $62,000 to $66,000 at the 3% discount rate, and $144,000 to $155,000 at a 7% discount rate. TMDLs could be combined for common water bodies (i.e., if the State decides to combine development of TMDLs for a class of waters with impairments for similar causes) to reduce development costs. However, EPA has no way to predict whether the State would do this, or for how many TMDLs. Additionally, it is possible that after listing a water as impaired with respect to aluminum based on ecoregional criteria, Oregon could conduct additional monitoring before developing a TMDL. Because the economic analysis criteria are conservative, the additional monitoring could result in the State identifying the water as unimpaired with respect to aluminum, and the State would therefore not develop a TMDL.


Potential Compliance Costs
This section summarizes the potential compliance costs to point sources and nonpoint sources, and discusses the limitations and uncertainties associated with the analyses. EPA used certain assumptions to operationalize the methodologies described in the preamble of the final rule for the purposes of this economic analysis. In this document, EPA refers to the calculated criteria output values used in this analysis as the "economic analysis criteria." This recognizes that the State has discretion in how it selects criteria output values base on the final criteria model, which may differ from the economic analysis criteria EPA generated. The State also has discretion in how it will implement its resulting criteria output values once established. To calculate the economic analysis criteria, EPA used measured pH and measured or estimated DOC and total hardness data from each of Oregon's Level III Ecoregions as inputs to the criteria models. EPA then selected the 10[th] percentile of the model output distributions for each Ecoregion, consistent with Method Two as described in the preamble of the final rule to reconcile model outputs. Estimates of the impacts based on these economic analysis criteria are conservative because more stringent criteria estimates tend to result in overestimates of impacts and costs. 
Point Sources
The RPA showed no RP for the discharge from the one facility for which aluminum monitoring data were available, thus no compliance costs were associated with this specific facility using the economic analysis criteria. However, for this economic analysis, which shows waters that would be potential incrementally impaired for aluminum, EPA elected to conduct a supplemental point source analysis of point source costs. To do so, EPA made many assumptions in the absence of facility-specific data; the results of this supplemental analysis are therefore highly speculative as compared to costs estimated using the RPA approach which is a facility-specific, data-based approach.

Because of the lack of data for aluminum from point source discharges in the State, EPA additionally identified potential costs for point source dischargers that utilize aluminum in their operations (aluminum anodizing facilities, drinking water treatment plants, and wastewater treatment facilities) and, therefore, could potentially be affected by the rule by presuming a need for additional controls or product substitution. This analysis supplements the analysis of potential point source costs based on RPA using Oregon facility-specific data.
EPA developed both a low-end and a high-end estimate for the costs to the State's 14 aluminum anodizers, based on two different technology upgrade options. Without facility-specific information on which option each facility would choose if they were to make an upgrade, EPA estimated that if all 14 facilities upgraded to countercurrent cascade rinsing technology, the total annual cost would be $61,600 using a 3% discount rate and $84,000 using a 7% discount rate, both over a 20-year period to match the expected life of the capital equipment (Metcalf & Eddy, 2014; USEPA, 1979). On the high end, EPA estimated that if all 14 facilities upgraded to countercurrent cascade rinsing technology plus chemical precipitation and settling, the total annual cost would be $6.88 million using a 3% discount rate and $6.95 million using a 7% discount rate, again, over the 20-year expected life of the capital equipment (Metcalf & Eddy, 2014; USEPA, 1979). For the 49 drinking water treatment plants assumed to use alum as a coagulant, EPA estimated the annual costs for chemicals and sludge disposal at $0.90 million. For the seven wastewater treatment facilities currently or potentially using alum as a coagulant, EPA found that if they were to switch to a ferrous coagulant, each would realize cost savings. EPA assumes that, in absence of the final rule, the facilities would possess full information regarding the prices of both coagulants and would already be using the lowest cost treatment. To the extent that a facility is not using the lowest cost treatment, EPA assumes that some other circumstance, potentially related to availability, engineering or process requirements, or some other issue that EPA is unaware of, is the reason why the ferrous coagulant is not the preferred alternative. Therefore, although EPA analysis would suggest potential cost savings, EPA concluded that the final rule would result in no change in cost for these facilities. Because these estimates are based on assumed need for control strategies simply based on the potential presence of aluminum in various operations, with no facility-specific knowledge of actual levels in any waste stream, these costs are highly speculative.

Based on the supplemental point source analysis that EPA performed due to the lack of effluent data on aluminum concentrations in discharges, EPA estimated that there could be costs associated with compliance for 14 aluminum anodizers and 49 drinking water treatment facilities. The total annual costs to those facilities are estimated to be between $1.0 million and $7.8 million at a 3% discount rate and between $1.0 million and $7.9 million at a 7% discount rate over a 20-year period. 
Nonpoint Sources 
Under the baseline, control costs for nonpoint sources (e.g., agricultural and forest operations; contamination from historic mining sites) and municipal stormwater include those needed to reduce instream pollutant levels to baseline criteria or TMDL targets. Although there are no baseline numeric aluminum criteria for aquatic life, there are ongoing baseline efforts to reduce nonpoint sources contributing to other types of impairment sources such as soil erosion or industrial stormwater. Full implementation of these baseline efforts may reduce aluminum loadings as well. Furthermore, waters potentially exceeding the AWQC being finalized here may not be the result of nonpoint sources of aluminum (for which there are no BMPs other than soil erosion controls). It is plausible that the exceedance of the AWQC is due to sources of acidity, and it may be more appropriate to address those sources directly.  
Government Regulatory Costs
The ODEQ already has implementation procedures for aquatic life criteria, but the State may adopt new implementation procedures to support implementation of the final criteria. While there may be one-time adjustment costs to switch from narrative to numeric criteria, Oregon has experience implementing other multi-parameter equations and models, including the ammonia criteria and copper biotic ligand model (BLM)-based criteria, and thus would be well-positioned to implement the final aluminum criteria. EPA assumed that the ODEQ would rely largely on existing technical and guidance documents as well as past and on-going ODEQ analyses of aquatic life criteria. Therefore, EPA assumed the incremental costs of implementing the final criteria would be de minimis for the purpose of this cost analysis. 
If the ODEQ develops incremental TMDLs for aluminum as a result of the final rule, then it may incur associated costs. If, for example, average TMDL development costs range from $38,000 to $41,000 (USEPA 2001a; updated to 2018$ using the Implicit Price Deflator for Gross Domestic Product), then incremental costs would range from $2.05 million to $2.21 million for 54 incremental listings of impaired stations. TMDL development costs are one-time costs that would be spread out over several years, because it is not realistic to assume that Oregon could conduct all these analyses simultaneously. EPA is assuming a 10-year time period for the TMDLs to be developed, covering both the time it takes to develop a TMDL and the assumption that Oregon will not start work on all TMDLs simultaneously. As such, annual costs will range from $205,000 to $221,000 for the first 10 years after implementation, falling to $0 thereafter. Given that TMDL costs are one-time costs, the fully annualized costs (into perpetuity) are $62,000 to $66,000 at the 3% discount rate, and $144,000 to $155,000 at a 7% discount rate. 
It may be possible to combine TMDLs for common water bodies (i.e., if the State decides to combine development of TMDLs for a class of waters with impairments for similar causes) and reduce development costs, though EPA has no way to predict whether the State would do this, or for how many TMDLs. Additionally, it is possible that after listing a water as impaired with respect to aluminum based on ecoregional criteria, Oregon could conduct additional monitoring before developing a TMDL. Because the economic analysis criteria are conservative, the additional monitoring could result in the State identifying the water as unimpaired with respect to aluminum, and the State would therefore not develop a TMDL.
Total Estimated Costs
Based on the RPA, there would be no compliance costs for the one point-source discharger analyzed under the final criteria. Based on the supplemental point source analysis that EPA performed because of the lack of effluent data on aluminum in discharges, EPA estimated that there could be costs associated with compliance for 14 aluminum anodizers and 49 drinking water treatment facilities. The total annual costs to those facilities are estimated at $1.0 million to $7.8 million using at the 3% discount rate, and $1.0 million to $7.9 million at the 7% discount rates over a 20-year period.
If the final criteria result in additional waters being identified as impaired, then the State will incur incremental costs for TMDL development. Using available monitoring data, EPA estimated that 54 stations would be identified as impaired, at a total cost of $2.05 million to $2.21 million. TMDL development costs are one-time costs that would be spread out over a 10-year time period. As such, annual costs will range from $205,000 to $221,000 over the first 10 years for the development of 54 total TMDLs, falling to $0 thereafter. 
Combining the potential costs for point source compliance from the supplemental point source analysis with the incremental cost of TMDL development, the total cost annualized at a 3% discount rate would range from $1.2 million to $8.0 million for the first ten years. Using a 7% discount rate, the total cost would range from $1.2 million to $8.1 million for the first 10 years. The cost would be slightly less in subsequent years after the TMDL development is complete. 

The number of TMDLs, and therefore costs, could be lower than these estimates depending on choices made by the State (as noted above). Further, if nonpoint sources are the primary cause of some impairments, then the final criteria may result in some costs to nonpoint sources to implement BMPs to reduce aluminum loadings to affected waters. The magnitude of cost impacts to nonpoint sources depends on the extent to which additional practices are needed to meet the final criteria.
The total costs presented in this document are a product of a series of assumptions and subsequent analyses that are intended to be both conservative and as comprehensive as possible. Where there is uncertainty in whether and/or how the application of the aluminum water quality standard would generate costs, the assumptions and analysis err on the side of generating estimates of costs, in the interest of transparency. The final rule includes provisions to ensure that aluminum criteria would be calculated for a given water body to identify only actual, as opposed to speculative, water quality problems in practice. In addition, there are procedures to implement water quality standards that can minimize costs such as identifying the most cost-effective mix of point and nonpoint source controls for TMDLs and identifying aluminum levels that are either natural or cannot be remedied (at least in a short-term horizon). 


Uncertainties in the Analysis
There are a number of uncertainties in the analysis associated with data limitations, potential pollutant load reductions achievable, and how dischargers would respond to revised requirements and permit conditions that affect the estimated costs. 
Data Limitations
The lack of available facility-specific data for both point and nonpoint sources adds significant uncertainties to the analysis of potential costs associated with meeting the final criteria. 
Baseline
As Oregon has no current acute or chronic numeric aquatic life criteria for aluminum (there are narrative toxics criteria), it is difficult to compare the numeric criteria finalized here to the baseline criteria without additional analysis. EPA therefore assumed that the narrative criteria are fully implemented, and in the absence of information to the contrary, was only able to attribute costs identified by the economic analysis as being costs to comply with the numeric aluminum criteria in this final rule. If any waters are in fact impaired under the narrative criteria in the baseline, the costs associated with those waters are baseline costs and should be excluded from costs for the final criteria. EPA cannot determine whether any of the incrementally impaired waters would be considered impaired based on the narrative criteria.
Final Criteria
EPA used a conservative approach to estimate the economic analysis criteria values from available data, which may differ from the approach the ODEQ would employ. Therefore, the ODEQ may apply the criteria differently from the assumptions EPA used for the RPA and water quality impairment analyses. The likely effect of EPA's assumption is that the costs here are over-estimated. 
Point Sources
Data are not available for aluminum for all NPDES-permitted facilities, making it difficult to determine which facilities are potentially affected by the final criteria. Specifically, whether an affected facility is meeting baseline standards and whether incremental controls would be needed to meet final criteria. Thus, costs could be higher or lower than estimated in this economic analysis. 
For the one facility with effluent data, the data available are limited. If the available data do not reflect typical loadings of aluminum, then the RPA results could be either overstated or understated. 
Because of the lack of data for aluminum from point source discharges in the State, EPA additionally identified potential costs for point source discharges that utilize aluminum in their operations (aluminum anodizing facilities, drinking water treatment plants, and wastewater treatment facilities) and, therefore, could potentially be affected by the final rule by presuming a need for additional controls or product substitution. This analysis supplements the analysis of potential point source costs based on RPA using data from specific facilities in Oregon. The supplemental analysis is highly speculative because it is unknown whether any of the facilities included would have RP under the final criteria.
Nonpoint Sources and TMDLs
Uncertainty in estimating the impacts on nonpoint sources may arise from uncertainty in identifying potential incrementally impaired waters. The lack of available monitoring data for some water bodies and the use of economic analysis criteria values instead of site-dependent aluminum criteria prevent reliable estimation of impairments for aluminum. Additionally, the monitoring data used in this analysis is based on a more conservative analytical method, which requires a pH of 2.0 for extraction and may overestimate bioavailable aluminum concentration. Further, an assumption that the net increase in the estimate of impaired stations represents the net increase in the number of TMDLs may overstate the impact if several stations can be addressed by one TMDL. 
As described in Section 3, for the nonpoint source component of the economic analysis, EPA used the economic analysis criteria, based on the 10[th] percentile of model output distributions, to derive the final water body-specific criteria. This is a conservative approach in that it likely errs on the side of overestimating impacts because it assumes lower criteria will apply to all sites within an ecoregion. EPA conducted a sensitivity analysis using the alternative economic analysis criteria, calculated using the 50[th] percentile of model output distributions. The sensitivity analysis results in 42 stations identified as impaired rather than 54 based on the 10[th] percentile analysis. Estimated total costs under the sensitivity analysis are reduced from a range of $2.05 to $2.21 million to a range of $1.60 to $1.72 million, while annual costs are reduced from a range of $205,000 to $221,000 to a range of $160,000 to $172,000. 



References
ATSDR. 2008. Toxicological Profile for Aluminum. Agency for Toxic Substances and Disease Registry. https://www.atsdr.cdc.gov/toxprofiles/tp22.pdf.

Ballentine, R. K., and L. J. Guarraie (eds.). 1975. The integrity of water." a symposium. U.S. Environmental Protection Agency, Washington, D.C. 230 pp.

Canadian Council of Ministers of the Environment. 2003. Canadian Water Quality Guidelines for the Protection of Aquatic Life: Aluminum. 
City of Lima, OH. Utilities 2015 Chemical Bids Tabulation.  

City of McMinnville, 2009. McMinnville Water Reclamation Facilities Plan

Driscoll, Charles. 1985. Aluminum in Acidic Surface Waters: Chemistry, Transport, and Effects. Environmental Health Perspectives, Vol. 63, pp. 93-104. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1568493/pdf/envhper00446-0095.pdf.

Eichenberger, E. 1986. The interrelation between essentiality and toxicity of metals in the aquatic ecosystem. In: H. Sigel (Ed.), Metal Ions in Biological Systems, Vol. 20. Concepts on Metal Ion Toxicity. Marcel Dekker, NY, 67-100. 
Exley, C. 2003. A biogeochemical cycle for aluminum. J. Inorg. Biochem. 97: 1-7. 
Freeman III, A. M. 1993. "Non-use values in natural resource damage assessment. "In R. J. Kopp, V. K. Smith (Eds.), Valuing Natural Assets. Washington, D.C.: Resources for the Future.
Gensemer, R.W. and R.C. Playle. 1999. The bioavailability and toxicity of aluminum in aquatic environments. Crit. Rev. Environ. Sci. Technol. 29(4): 315-450. 
Jennings, S.R., D.R. Neuman, and P.S. Blicker. 2008. Acid Mine Drainage and Effects on Fish Health and Ecology: A Review. Reclamation Research Group Publication, Bozeman, MT. 
Karr, J.R. and Dudley, D.R. 1981. Ecological Perspective on Water Quality Goals.  Environmental Management, Vol. 5, No. 1, pp. 55-68.
Kelley, E. and Twohig, E. 2018. Vermont Department of Environmental Conservation Waste Management & Prevention Division Residual Waste & Emerging Contaminants Program. Wastewater Treatment Sludge and Septage Management in Vermont. https://dec.vermont.gov/sites/dec/files/wmp/residual/RMSWhitePaper20180507.pdf 

Kopp, R. J., Smith, V. K. 1993. Valuing Natural Assets: The Economics of Natural Resource Damage Assessment. Washington D.C.: Resources for the Future.

Kruger. Undated. Physical Chemical Phosphorus Removal. http://www.acsawater.com/sites/default/files/websitefiles/SVWWTPN/Part%202%20Physical%20Chemical%20Phosphorous%20removal%20Compatibility%20M.pdf  

Leonard, A. and G.B Gerber. 1988. Mutagenicity, carcinogenicity and teratogenicity of aluminium. Mutat. Res. 196(3): 247-57. 

Metcalf & Eddy. 2014. Wastewater Engineering: Treatment and Resource Recovery. 5[th] Edition. McGraw-Hill, 2014. 

Michigan Department of Environmental Quality Operator Training and Certification Unit. The Basics of Phosphorus Removal.  

Minnesota Pollution Control Agency. 2006. Phosphorus Treatment and Removal Technologies. Municipal Division, Wastewater Program wq-wwtp9-02. https://www.pca.state.mn.us/sites/default/files/wq-wwtp9-02.pdf. 
NDWC. 1997. Tech Brief: Ion Exchange and Demineralization. National Drinking Water Clearinghouse. http://www.nesc.wvu.edu/ndwc/pdf/OT/TB/TB4_IonExchange.pdf. 
ODA. 2016. A Landowner's Guide to Oregon's Agricultural Water Quality Management Program. Oregon Department of Agriculture, Salem, OR. http://www.oregon.gov/ODA/shared/Documents/Publications/NaturalResources/WaterQualityandAgriculture.pdf. 
ODA. 2002. Water Quality and Agriculture: SB 1010. Oregon Department of Agriculture, Salem, OR. http://extension.oregonstate.edu/landmarksdvd/documents/ODA_SB_1010.pdf. 
ODEQ. 2017a. Municipal Stormwater Permits. Oregon Department of Environmental Quality, Portland, OR. https://www.oregon.gov/deq/wq/wqpermits/Pages/MS4-Permits.aspx. 
ODEQ. 2017b. MS4 Phase II General Permit Development. Oregon Department of Environmental Quality, Portland, OR. https://www.oregon.gov/deq/wq/wqpermits/Pages/MS4-PhaseII-General-Permit.aspx. 
ODEQ. 2017c. Industrial Stormwater Permits. Oregon Department of Environmental Quality, Portland, OR. https://www.oregon.gov/deq/wq/wqpermits/Pages/Stormwater-Industrial.aspx. 
ODEQ. 2014. National Pollutant Discharge Elimination System Waste Discharge Permit. Issued to: Northwest Aluminum Specialties, Inc. EPA Reference No: OR000170-8
ODEQ. 2012a. Fact Sheet: Cleanups of Abandoned Mines in Oregon Complicated by High Cost and Lack of Funding. Oregon Department of Environmental Quality, Portland, OR. 
ODEQ. 2012b. Internal Management Directive, Reasonable Potential Analysis Process for Toxic Pollutants Version 3.1. Oregon Department of Environmental Quality, Portland, OR. Updated February 13, 2012. 
ODEQ. 2014. Methodology for Oregon's 2012 Water Quality Report and List of Water Quality Limited Waters. Oregon Department of Environmental Quality, Portland, OR.
Oregon Department of Forestry. 2017. Forest Practices Act. http://www.oregon.gov/ODF/Working/Pages/FPA.aspx.
Oregon Forest Resources Institute. 2011. Oregon's Forest Protection Laws: An Illustrated Manual: Revised, Second Edition.
Sparling, D.W. and T.P. Lowe. 1996a. Environmental hazards of aluminum to plants, invertebrates, fish, and wildlife. Rev. Environ. Contam. Toxicol. 145: 1-127. 

Tchounwou, T.B., C.G. Yedjou, A.K. Patlolla and D.J. Sutton. 2012. Heavy metals toxicity and the environment. EXS. 101: 133-164. 

USEPA. 2019. ICIS-NPDES. U.S. Environmental Protection Agency. 

USEPA. 2018. Final Aquatic Life Ambient Water Quality Criteria for Aluminum in Freshwater. U.S. Environmental Protection Agency, Office of Water, Washington, DC. Pre-Publication Notice. https://www.epa.gov/sites/production/files/2018-12/documents/aluminum-prepub.pdf

USEPA. 2017b. Level III Ecoregions of the Continental United States. U.S. Environmental Protection Agency. https://www.epa.gov/eco-research/level-iii-and-iv-ecoregions-continental-united-states. 

USEPA. 2016. Economic Analysis for the Proposed Rule: Aquatic Life Criteria for Copper and Cadmium in Oregon. U.S. Environmental Protection Agency, Office of Water, Washington, DC 

USEPA. 2014. Guidelines for Preparing Economic Analyses. U.S. Environmental Protection Agency, Office of Water, Washington, DC. December 17, 2010, updated May 2014. https://yosemite.epa.gov/ee/epa/eerm.nsf/vwAN/EE-0568-50.pdf/$file/EE-0568-50.pdf.

USEPA. 2011. Drinking Water Treatment Plant Residuals Management Technical Document. U.S. Environmental Protection Agency, Office of Water, Washington, DC EPA-820-R-11-003. https://www.epa.gov/sites/production/files/2015-11/documents/dw-treatment-residuals-mgmt-tech-report-sept-2011.pdf.

USEPA. 2010a. Guidelines for Preparing Economic Analyses.

USEPA. 2010b. NPDES Permit Writer's Manual. U.S. Environmental Protection Agency, Office of Water, Washington, DC EPA-833-K-10-001. https://www3.epa.gov/npdes/pubs/pwm_2010.pdf.

USEPA. 2003. Development Document for the Final Effluent Limitations Guidelines and Standards for the Metal Products & Machinery Point Source Category. U.S. Environmental Protection Agency, Office of Water, Washington, DC EPA-821-B-03-001. https://www.epa.gov/sites/production/files/2015-11/documents/mp-m_dd_2003.pdf. 

USEPA. 2008. Superfund Fact Sheet: Formosa Mine Superfund Site, Douglas County Oregon. U.S. Environmental Protection Agency, Seattle, WA. 

USEPA. 2003. Development Document for the Final Effluent Limitations Guidelines and Standards for the Metal Products & Machinery Point Source Category. U.S. Environmental Protection Agency, Office of Water, Washington, DC EPA-821-B-03-001. https://www.epa.gov/sites/production/files/2015-11/documents/mp-m_dd_2003.pdf. 

USEPA. 2001a. The National Costs of the Total Maximum Daily Load Program (Draft Report). U.S. Environmental Protection Agency, Office of Water, Washington, DC EPA-841-D-01-003. https://nepis.epa.gov/Exe/ZyPDF.cgi/901K0800.PDF?Dockey=901K0800.pdf. 

USEPA. 2001b. The National Costs to Develop TMDLs (Draft Report) Support Document #1. U.S. Environmental Protection Agency, Office of Water, Washington, DC EPA 841-D-01-004. https://nepis.epa.gov/Exe/ZyPURL.cgi?Dockey=P1008JEP.txt

USEPA. 1998. R.E.D. Facts: Aluminum and Magnesium Phosphide. U.S. Environmental Protection Agency, Office of Water, Washington, DC EPA-738-F-98-015. https://archive.epa.gov/pesticides/reregistration/web/pdf/0025fact.pdf.

USEPA. 1991. Technical Support Document for Water Quality-based Toxics Control. U.S. Environmental Protection Agency, Office of Water, Washington, DC EPA/505/2-90-001. http://www.waterboards.ca.gov/water_issues/programs/tmdl/docs/303d_policydocs/338.pdf.

USEPA. 1979. Development Document for Existing Source Pretreatment Standards for the Electroplating Point Source Category. U.S. Environmental Protection Agency, Office of Water, Washington, DC EPA 440/1-79/003. https://www.epa.gov/sites/production/files/2015-10/documents/electroplating_dd_1979.pdf.

USGS. 2015. Estimated Annual Agricultural Use for Aluminum Phosphide, 1996 (EPest-High). Pesticide National Synthesis Project. U.S. Geological Survey. https://water.usgs.gov/nawqa/pnsp/usage/maps/show_map.php?year=1996&map=ALUMINUMPHOSPHIDE&hilo=H. 

Williams, R.J.P. 1999. What is wrong with aluminium? The J.D. Birchall memorial lecture. 
J. Inorg. Biochem. 76: 81-88.

Wilson, R.W. 2012. Chapter 2: Aluminum. In: C.M Wood, A.P. Farrel and C.J. Brauner (Eds.), Homeostasis and toxicology of non-essential metals: Volume 31B. Elsevier Inc.

Wood, J.M. 1985. Effects of acidification on the mobility of metals and metalloids: An 
overview. Environ. Health Perspect. 63: 115-119.

Wood, J.M. 1984. Microbial strategies in resistance to metal ion toxicity. In: H Sigel (Ed.), Metal Ions in Biological Systems, Vol. 18. Circulation of Metals in the Environment. Marcel Dekker, NY, 333-351. 


Woodwell, G.M. 1975. Biological integrity--1975. Pages 141-147 in R. K. Ballentine and L.J. Guarraie, eds. The Integrity of Water. U.S. Environmental Protection Agency. Washington,
D.C.



Facility Analyses
This appendix provides detailed analyses for the point source facilities. 
Northwest Aluminum Company 
Facility Description
Northwest Aluminum Company (NPDES permit number OR0001708) owns the site in the City of The Dalles, Oregon, where they had previously operated a primary aluminum smelting facility, which ceased operation and was demolished by July 2009. The site currently has a recycling plant operated by Northwest Aluminum Specialties (NWAS) and a Comprehensive Environmental Response, Compensation and Liability Act (CERCLA) tank leachate collection system owned by Lockheed Martin. The recycling plant and CERCLA tank leachate collection system discharge into the Columbia River at river mile 189. The facility's sanitary wastewater and deburring wastewater is collected and treated by the City of The Dalles. 
Effluent and Treatment Process
The NWAS recycling plant discharge consists of 0.6-1.8 mgd river water used for non-contact cooling water and 0.2-0.5 mgd well water for cooling the metal jackets around the electro-melt furnaces and the door jambs on the homogenizing furnaces. The CERCLA tank leachate discharge consists of leachate from the CERCLA and Resource Conservation and Recovery Act landfills. The leachate is treated while in the collection system through use bioremediation technology for the reduction of cyanide. The outfall to the Columbia River total about 2.3 mgd currently and 6 mgd maximum production. The facility permit has limits for the following pollutants: total suspended solids, benzo(a)pyrene, antimony, nickel, fluoride, aluminum, oil and grease, pH, and chlorine residual. 
Reasonable Potential Analysis
EPA had available data to conduct RPA for aluminum. Dilution factors used in the RPA were provided in the facility permit. The results of the RPA indicate that there was no RP under the final criteria. As such, no effluent limitations were calculated. 
References and Sources

ODEQ. 2012. Internal Management Directive, Reasonable Potential Analysis Process for Toxic Pollutants Version 3.1. Oregon Department of Environmental Quality. 
ODEQ. 2009 and 2014. Northwest Aluminum Specialties, Inc. Fact Sheet and NPDES Wastewater Discharge Permit Evaluation. NPDES permit number OR0001708. Oregon Department of Environmental Quality.
ODEQ. RPAspreadsheetInd  -  excel spreadsheet. Oregon Department of Environmental Quality.
ODEQ. RPAdomR36(1)  -  excel spreadsheet. Oregon Department of Environmental Quality.
     Appendix B - Net Present Value of Costs
This appendix presents the net present value (NPV) of the costs of this final rule and the need to have an accounting across regulations to meet EPA's regulatory budget, using both a 3% and a 7% discount rate. 

EPA chose a period of three years for the NPV calculation because the State is required to undertake a triennial review of its water quality standards and criteria every three years. If a State or tribe chooses not to adopt new or revised criteria for any parameters for which EPA has published new or updated criteria recommendations under CWA Section 304(a), the State or tribe must explain its decision when reporting the results of the triennial review to EPA under CWA Section 303(c)(1). In addition, based on the cooperative federalism embedded in the CWA regarding State and federal roles in setting State water quality standards, it is not EPA's intention for a federally promulgated rule to remain in place for very long. In most circumstances, EPA is only promulgating a State water quality standards rule because the CWA imposes deadlines on EPA when States are unable to promulgate in a timely manner water quality standards that EPA can approve, based on the federal water quality standards regulations. When EPA issues such regulations, States frequently take action soon after EPA's action, and once EPA approves the State action, EPA acts to remove the federal regulation. This process generally takes less time than the three years associated with the triennial review. However, until EPA can gather data to support a more refined estimate of the time federally-promulgated State water quality standards are in effect, EPA has elected to use three years, based on the triennial review time frame. EPA is also basing the NPV calculation on the lower bound of the cost estimates because the upper bound estimates are much more speculative and require the assumption that facilities currently are not doing anything to control the discharge of aluminum.

The total NPV over three years for the costs of this final rule thus ranges from $3.1 million at a 3% discount rate, to $3.4 million at a 7% percent discount rate. 

