Loading...
HomeMy WebLinkAboutStaff Report 1297 City of Palo Alto (ID # 1297)   City Council Informational Report        Report Type: Informational Report Meeting Date: 2/14/2011    February 14, 2011 Page 1 of 2  (ID # 1297)   Title: EPA Proposed Rule on Sewage Sludge Incinerators  Subject: Palo Alto Comments on the Environmental Protection Agency’s  Proposed Rule for Sewage Sludge Incinerator Units  From: City Manager     Lead Department: Public Works  Executive Summary  Staff submitted the attached comment letter to the Environmental Protection Agency (EPA) on  the proposed new emissions requirements for sewage sludge incinerators and will return to  Council with key information following publication of the final rule.  This is a brief informational  report to Council regarding the proposed rule making.  No action by Council is required.    Discussion  On October 14, 2010, EPA proposed emission limits for new and existing sewage sludge  incineration units.  Due to a court decision in 2007, EPA was directed to regulate sewage sludge  incinerators under Section 129 of the Clean Air Act.   Section 129 of the Clean Air Act directs  EPA to promulgate emission limits that are based on Maximum Achievable Control Technology.    In the proposed rule, EPA determined that the calculated Maximum Achievable Control  Technology emission limits for mercury were not stringent enough, and therefore proposed a  “beyond‐the‐floor” mercury emission limit that is even more stringent.    The Palo Alto Regional Water Quality Control Plant is one of the 97 agencies nationwide that  utilizes sewage sludge incinerators and is thereby affected by the proposed rule. Palo Alto’s  sewage sludge incinerators would not be able to meet the proposed “beyond‐the‐floor”  mercury limits without significant and costly emissions control upgrades.  Staff has tracked this  issue and worked closely with the National Association of Clean Water Agencies in developing  the attached comment letter.  Staff submitted comments on the proposed rulemaking on  November 29, 2010 (Attachment A).  EPA acknowledged that many of the comments submitted  in response to the proposed emissions guidelines are significant and valid, and EPA requested  that the U.S. District Court for the District of Columbia grant a six‐month extension to July 2011  for promulgation of a final rule.  The Court denied the request, instead ordering EPA to finalize  the rule by February 21, 2011.  Staff will return to Council after the final rule is published with a  staff report discussing the implications of the final emissions rule for Palo Alto’s sewage sludge  incinerators.        February 14, 2011 Page 2 of 2  (ID # 1297)   On a longer time frame, staff is studying other alternatives for managing sewage sludge as part  of the Water Quality Control Plant’s Long Range Facilities Planning Process, which began in the  fall of 2010.  As directed by Council, staff will be evaluating energy recovery technologies during  this process, which is scheduled for completion in 2012.   ATTACHMENTS:   PA Comments EPA HQ‐OAR‐2009‐0559_Final (PDF)  Prepared By: Karin North, Associate Engineer    Department Head: J. Michael Sartor, Interim Director    City Manager Approval:    James Keene, City Manager Divisions Administration 650.329.2373 650.329.2299 fax Engineering 650.329.2151 650.329.2299 fax Environmental Compliance 650.329.2598 650.494.3531 fax Equipment Management 650.496.6922 650.496.6958 fax Facilities Management 650.496.6900 650.496.6958 fax Operations 650.496.6974 650.852.9289 fax Regional Water Quality Control 650.329.2598 650.494.3531 fax Ci~ of Palo Alto Public Works Department November 29,2010 VIA ELECTRONIC SUBMISSION EP A Docket Center (EP AlDC) United States Environmental Protection Agency (6102T) Proposed Rulemaking -New Source Performance Standards and Emission Guidelines for Sewage Sludge Incinerators Docket ID No. EP A-HQ-OAR-2009-0559 1200 Pennsylvania Avenue NW Washington, DC 20460 Re: City of Palo Alto Comments on the Proposed Rule on Standards of Performance for New Stationary Sources and Emission Guidelines for Existing Sources: Sewage Sludge Incineration Units, 75 Fed. Reg. 63260 (October 14,2010) Dear Administrator Jackson: The. City of Palo Alto appreciates the opportunity to comment on the proposed Clean Air Act Standards for Sewage Sludge Incinerators (SSIs) referenced above. The City of Palo Alto operates the Regional Water Quality Control Plant (RWQCP), a wastewater treatment plant, for the East Palo Alto Sanitary District, Los Altos, Los Altos Hills, Mountain View, Palo Alto, and Stanford University and serves a sewer population of approximately 230,000. On average, the RWQCP treats 21.4 million gallons per day. Annually, we incinerate . approximately 24,200 wet tons (6,400 dry tons) of sewage sludge solids onsite in '. a multiple hearth incinerator. As set forth below, we believe that EPA has not demonstrated that the proposed "beyond-the-floor" mercury standard is achievable using available stack gas control technologies on multiple hearth incinerators. Further, we believe that source control, not end-of-pipe or stack treatment, is the most environmentally protective and cost effective strategy; it should be fully implemented in this case. The EPA-specified stack gas treatment will cost Palo Alto $7 million dollars (an EPA estimate) in an attempt to annually remove 10 pounds (our estimate) of mercury. EPA has significantly overestimated the amount of mercury emitted by SSIs resulting in a flawed costibenefit analysis. The mercury costibenefit analysis was further deficient in failing to include all the costs identified by EPA, by failing to include any financial benefits for mercury, and by inappropriately including particulate matter benefits. This can be remedied by using the data P.O.Box 10250 Palo Alto, CA 94303 Printed with soy-based inks on 100% recycled paper processed without chlorine. Page 2 of 20 provided in our detailed comments below and relying on source control rather than stack gas treatment for beyond-the-floor mercury control. As you will see in our detailed comments, we find that the cost for Palo Alto to comply with EPA’s proposed beyond-the- floor mercury standard is $436 million per ton of mercury removed. Our calculated cost for multiple hearth incinerators nationwide to comply with the beyond-the-floor mercury standard is $188 million per ton of mercury removed. Source control for mercury, however, can achieve very significant mercury reductions at a cost of approximately $18 million per ton. As an initial matter, we note that it is difficult if not impossible to adequately evaluate and comment on the Agency's beyond-the-floor determinations if the Maximum Achievable Control Technology (MACT) floors themselves were improperly set.1 There are numerous deficiencies in the methods by which EPA reached its MACT floor determination within the proposed standards for sewage sludge incinerator units. As detailed in the attached comments submitted by the National Association of Clean Water Agencies (NACWA), these include, but are not limited to, the lack of sufficient data to establish an appropriate MACT floor and reliance on outdated information in EPA’s analysis. The City endorses the comments submitted by NACWA and rather than repeat them here, our comments focus on our most significant concern, which is EPA’s proposed beyond-the-floor mercury standard. Furthermore, the City is cognizant of the fact that EPA is currently operating under a court ordered deadline to promulgate rules for sewage sludge incinerators. However, EPA is not under any legal obligation to go beyond the floor as part of its proposed rule. Under these time constraints, it may be difficult for EPA to gather sufficient supporting data to adequately inform the establishment of a MACT floor, and it is even more difficult to gather the additional supporting information and analysis needed for establishing a beyond-the-floor mercury standard as EPA has attempted to do under the proposed rule. Given the demonstrated deficiencies in EPA’s mercury data, the failure to analyze factors required to be considered when adopting beyond-the-floor standards, and the abbreviated timeline imposed by court order, EPA should consider delaying the adoption of beyond-the-floor standards rather than attempting to incorporate them into this compressed process. A. Beyond-the-Floor Stack Gas Treatment for Mercury Control Has Not Been Demonstrated to Be Effective for Multiple Hearth (MH) SSIs. EPA acknowledges that there are no MH SSIs that utilize an activated carbon injection system. Yet, even though there are no proven examples that this technology will work on a multiple hearth incinerator, EPA assumed that this technology will consistently achieve an 88 percent reduction in mercury concentrations. EPA has not demonstrated that it is technologically feasible to meet the proposed standards via activated carbon injection on MH SSIs. EPA states in the rule, 1 See Northeast Md. Waste Disposal Auth. v. EPA, 358 F.3d 936, 955 (D.C. Cir. 2004) Page 3 of 20 we believe activated carbon injection is applicable to both types of SSI combustors and do not know of any technical reason that activated carbon injection could not be applied to reduce Hg emissions at MH units. We are requesting comment and additional information on the feasibility of using this technology on MH units. (75 FR 63277) The St. Paul Metro Wastewater Treatment Plant (WWTP) is the only Publicly Owned Treatment Works (POTW) in the United States that injects activated carbon into its exhaust gas system. St. Paul Metro WWTP’s SSI is a fluidized bed incinerator, not a multiple hearth incinerator, and has experienced significant problems with its activated carbon injection system. The carbon is injected into a carbon contact chamber that is followed by a fabric filter. Incinerator exhaust gases entering the contact chamber are first cooled to roughly 350 F by passing through two heat exchangers and a boiler. Since start-up in 2005, the St. Paul Metro WWTP has experienced numerous abrasion and corrosion problems with its carbon system and fabric filter. A review of EPA’s Memorandum Estimation of Baseline Emissions Rates from Existing Sewage Sludge Incinerators indicates that there are no MH SSIs that use fabric filters. EPA’s contention that none of the existing MH systems will need to install new particulate matter control equipment to meet the beyond the MACT floor mercury limit is not supported by the information in the record. Section D of this letter details the reasons why existing wet scrubbers that provide particulate control cannot be used in conjunction with activated carbon injection for mercury control. Palo Alto concurs with the EPA that activated carbon injection cannot be applied alone to control mercury emissions. Activated carbon injection requires particulate control devices to remove the carbon that is injected to adsorb the mercury. In addition to the lack of suitable particulate controls, there are several other factors that are critical for activated carbon injection to work. These include:  Residence time  Flow rates  Operating temperatures  Fuel/flue gas analysis  Mercury levels and mercury speciation in the flue gas The above process parameters will require, at minimum, a fabric filter, duct work changes, and a carbon reactor to provide the needed residence time. Fabric filters are very large and would require additional building space that most facilities do not have. In addition, exhaust gases from a multiple hearth incinerator are commonly in the range of 1000-1500 F, which is too high for adsorption to occur. An exhaust gas conditioner will be required to bring the temperature to 300-400 F. Palo Alto’s engineering consultant has contacted several activated carbon injection system vendors and has confirmed that these vendors have had no experience with activated carbon injection in multiple hearth systems for mercury control. At the Palo Alto RWQCP, a small site is available for any new facilities that may be required; however it is likely that the space is not adequate for all the new air pollution Page 4 of 20 control equipment. It is clear that Palo Alto would lose a critical roadway within the plant to install the needed air pollution control equipment. This space constraint is severe and will make it extremely difficult to receive chemical deliveries, modify existing infrastructure, and install future wastewater treatment technologies. Additionally, the mercury concentration in the flue gas streams of SSIs is typically very low as compared to other incinerators such as municipal solid waste incinerators. At low concentrations of mercury in the flue gas, the adsorption process could become mass transfer limited and reduce the removal efficiency. This would require injection of larger quantities of carbon to remove a relatively small quantity of mercury. EPA has erred in not factoring in these variables, which will greatly increase operating costs, into the beyond-the-floor mercury calculations. The beyond-the-floor mercury reduction of 88 percent for activated carbon injection in multiple hearth systems is arbitrary. While it may be true that data gathered from other combustion processes like coal-fired boilers indicate a mercury control efficiency range of 85-95 percent with activated carbon injection and subsequent particulate removal, the information in the record does not support a conclusion that MH SSIs can achieve significant mercury control to meet the beyond-the-floor limit. Combustion of coal results in flue gas containing relatively stable constant mercury concentrations. In contrast, SSIs combust biosolids with widely variable mercury concentrations, and mercury flue gas concentrations from SSIs are therefore quite variable. To illustrate this point, Figure 1 provides Palo Alto’s monthly sludge cake mercury concentration data and triplicate flue gas mercury concentration data since 2001. In setting the MACT floor standards for mercury and other constituents, EPA used the Upper Prediction Limit (UPL) statistical method to account for variability in the data set. EPA’s selection of 88 percent as a constant mercury reduction achievable using activated carbon injection fails to account for variability in the performance of activated carbon injection and the other air pollution equipment that will be needed in conjunction with activated carbon injection. Variability in the performance of the activated carbon injection process, coupled with varying mercury concentrations in flue gas upstream of activated carbon injection, may result in removal less than 88 percent and flue gas concentrations that exceed the proposed beyond-the-floor mercury standard. EPA should not adopt a stringent mercury standard for MH SSIs based on untested assumptions that activated carbon injection technology can be successfully transferred to MH SSIs. Activated carbon injection has not yet been demonstrated to be a sustainable success in its single fluidized bed SSI application. As described above, activated carbon injection on MH SSIs has not been demonstrated and would require multiple new air pollution control components beyond the activated carbon injection itself. Given the multiple system components that would be required and the variability of the mercury concentrations in SSI flue gas, EPA should not assume that a constant 88 percent mercury removal is achievable by this technology. Page 5 of 20 0 100 200 300 400 500 600 70001-Jan-01 01-Jan-02 01-Jan-03 01-Jan-04 01-Jan-05 01-Jan-06 01-Jan-07 01-Jan-08 01-Jan-09 01-Jan-10 Date Me r c u r y i n S l u d g e C a k e ( u g / k g ) 0.00 20.00 40.00 60.00 80.00 100.00 120.00 140.00 160.00 180.00 Me r c u r y i n A i r (u g / d s c m @ 7 % O 2 ) Sludge Cake ug/kg Emission (ug/dscm @ 7% O2) Figure 1: Variability of Palo Alto mercury air emissions compared to concentration of mercury in sludge cake B. National Source Control is a More Desirable and Cost Effective Way to Achieve Beyond-the-Floor Mercury Reductions for SSIs. In addition to being more economical, source control prevents environmental release of mercury. Stack gas treatment only transfers mercury from one medium to another. Mercury is an element that cannot be destroyed. Because of mercury’s volatility, mercury readily moves from one media to another and it is extremely difficult to actually remove from the environment. Only source control truly prevents environmental release of mercury. We are concerned that the wet scrubbing described in EPA’s control strategy analysis would simply transfer airborne mercury to the water stream. Further, attempting to take mercury to a landfill could result in ultimate volatilization and release. EPA has previously recognized the desirability of using mercury source control for hazardous waste incinerators (EPA Docket HQ: OAR: 2004-0022). EPA should also identify source control as the appropriate technology for beyond-the-floor mercury control for SSIs. Palo Alto and other communities have demonstrated the effectiveness of mercury source control. Palo Alto assisted in authoring California legislation that eliminated mercury in thermometers, certain switches, and novelty items (Chapter 656, Statutes of 2001). Palo Alto was one of the first California POTWs to require amalgam separators at dental Page 6 of 20 facilities; Palo Alto developed training programs for dental office workers which are now in widespread use. Palo Alto has an ongoing drop-off program for all types of mercury- containing equipment including thermostats, thermometers, medical devices, switches, reagents, and medicines containing mercury. More work needs to be done and can be done to eliminate mercury in consumer and commercial products. Mercury has been placed in many products to retard microbial activity. Mercury is also in use in manufacturing seals, measuring devices, switches and reagents, which in turn lead to mercury incorporation in products such as chlorine. Mercury devices are still found in hospitals and laboratories. These uses can and should be eliminated. Palo Alto stands ready to work with the EPA and others on Toxic Substances Control Act regulations, new legislation, and other types of restrictions to eliminate mercury use. When mercury is eliminated in consumer and commercial products, the mercury concentration in wastewater will drop substantially further. Palo Alto’s stack gas emissions have already reached a new low of 0.051 mg/dscm, 70 percent below EPA’s proposed MACT-floor level. EPA should now estimate the mercury level that can be achieved in sewage sludge incinerator stack gas after implementation of full source control. This is the appropriate and best way to develop a beyond-the-floor mercury limit. Details of the Palo Alto Mercury Program can be found in its “2010 Clean Bay Plan.” The appropriate chapter of the Plan is an attachment to these comments. Data showing mercury reductions achieved in Palo Alto to date using source control are summarized in Figure 2 below. Palo Alto has achieved a 63 percent reduction since the 2001-2004 timeframe. The significant decrease since 2004 is attributable to Palo Alto’s dental amalgam program, which required dental offices to install amalgam separators in 2005. Mercury concentrations in Palo Alto’s sludge cake continue to decrease. Other POTWs that have implemented dental amalgam programs that mandate amalgam separators have observed comparable reductions in biosolids mercury concentrations. National source control programs, as opposed to local government programs, will now be needed to continue the work begun by Palo Alto and others. Product and manufacturing restrictions will be needed and are best done by EPA. Palo Alto has estimated the cost effectiveness of beyond-the-floor mercury reductions associated with its dental amalgam program. Our calculation concludes that the total program cost is approximately $18 million per ton of mercury removed. The figure accounts for the amalgam separator capital and annual maintenance costs for the Palo Alto dental offices and Palo Alto staff time, compared to the actual measured reduction in the mass of mercury discharged to the plant. A Technical Memorandum providing the estimate is attached. This $18 million per ton figure is less than one tenth of our estimate of the cost of stack gas treatment (Table 3), and is far more certain of environmental success. Page 7 of 20 0 100 200 300 400 500 600 Ja n - 0 1 Ja n - 0 2 Ja n - 0 3 Ja n - 0 4 Ja n - 0 5 Ja n - 0 6 Ja n - 0 7 Ja n - 0 8 Ja n - 0 9 Ja n - 1 0 Ja n - 1 1 Date Me r c u r y ( µ g / k g ) 6-sample rolling average discrete data points Figure 2: Summary of mercury concentrations in Palo Alto sludge cake’ Dental offices were required to install amalgam separators 2005. (Average mercury concentration in sludge cake before amalgam separators was 397 µg/kg. Current 6-month average mercury concentration in sludge cake is 148 µg/kg, more than a 60 percent reduction.) C. EPA Overestimated Baseline Emissions for Mercury The proposed rule incorrectly estimates that the 218 existing SSIs emit 3.1 tons of mercury per year to the atmosphere. In 2009 the Water Environment Research Foundation released a report titled Minimizing Mercury Emissions in Biosolids Incinerators. The report conservatively determined that SSIs collectively emit less than one ton of mercury to the atmosphere each year. After careful review of the spreadsheets in the attachments to the proposed rule, we found that EPA grossly overestimated the baseline emissions for mercury. EPA’s baseline mercury emissions have been overestimated both for facilities that responded to EPA’s Information Collection Request (ICR) survey and for facilities, such as Palo Alto, for which data from the ICR facilities was used to develop estimates. Central Contra Costa Sanitation District (CCCSD) is one of the nine facilities that responded to EPA’s ICR. We have reviewed both EPA's emission calculation as well as the actual source test data from CCCSD’s December 2009 source test and have concluded that ERG, EPA’s consultant for the proposed rule, overestimated CCCSD’s mercury emissions by 77 percent. ERG multiplied concentration data corrected to 7 percent O2 with uncorrected flue gas flow rate data. For CCCSD, the flue gas flow during the source test contained 13 percent O2. The proper way to calculate emissions is by either (a) multiplying concentration data corrected to 7 percent O2 by flue gas flow rate data corrected to 7 Page 8 of 20 percent O2, or (b) multiplying raw (uncorrected) concentration data by uncorrected flue gas flow rate data. By using corrected concentration and uncorrected flue gas flow rate data, ERG overestimated CCCSD emissions by 77 percent. A detailed spreadsheet illustrating this error is included in CCCSD’s comment letter. ERG’s error in using uncorrected flue gas flow rate data with corrected mercury concentrations is not limited to the facilities that responded to the ICR, because the sludge feed rates and flue gas flow rates from these facilities were used to develop “flow rate factors” in dscfm/dry tons per hour of sludge that were then used in the subsequent calculation of baseline estimates for the facilities that did not receive the ICR. Therefore, it appears that the mercury baseline emissions for all SSIs were calculated using uncorrected flue gas flow rate data and concentration data corrected to 7 percent O2. In estimating baseline mercury emissions from Palo Alto’s SSIs, the flow rate factors that were developed resulted in a flue gas flow rate almost 50 percent greater than the actual measured flowrate. This inflated flowrate was then multiplied by EPA’s assumed concentration of 0.103 mg/dscm at 7% O2, introducing the error of using uncorrected flue gas flow rate with corrected concentration. Additionally, Palo Alto’s measured mercury concentration is about half of EPA’s assumed concentration. These factors combine to result in a baseline mercury emissions estimate for Palo Alto that is about six times higher than the actual emissions, as shown in Table 1 below. Table 1: Comparison of EPA’s annual baseline estimate for mercury to Palo Alto’s measured emissions Emission Estimates Flow rate (dscfm) Operating Hours Concentrati on (mg/dscm) Concentration (mg/dscm at 7% O2) Baseline (lbs/year) Baseline (tons/ year) EPA Estimate 19,458 4,200 per incinerator; total 8,400 NA 0.103 63.2 0.0316 Palo Alto Stack Testing 13,380 8,400 0.024 0.051 10.08 0.00504 * EPA assumed that we operate incinerators half of the year; therefore the EPA estimate for each incinerator was summed for an annual number. Assuming similar errors on baseline emissions across the board, USEPA and ERG have significantly overstated the baseline mercury emissions from SSIs. EPA must reevaluate baseline emissions, the MACT floor assessment, and the beyond the MACT floor assessment. Page 9 of 20 D. Activated Carbon Injection / Wet Scrubber Will Not Achieve Substantial Mercury Reductions to the Environment In the preamble, EPA states: …it is important to note that activated carbon injection cannot be applied alone. It requires particulate control devices to remove the carbon that is injected to adsorb the Hg. Based on our available data, all of these units have some type of PM control device in place so they would not need to install new PM control equipment. (75 FR 63277) We are deeply concerned with EPA’s incorrect assumption that MH SSIs would not need to install new particulate matter control equipment if they already had a wet scrubber in place. Wet scrubbers return the mercury to the water stream and fail to prevent release to the environment. Mercury readily cycles through ecosystems from water to air. As shown in Figure 3 below, 98.2 percent of Palo Alto’s mercury mass loading is emitted to the atmosphere via the incinerator stack and 1.7 percent is discharged to San Francisco Bay via the plant’s effluent. Palo Alto’s lowest mercury discharge limit for the San Francisco Bay discharge is 11 ng/L and our average effluent concentration is 2.5 ng/L. If activated carbon was injected and the particles were collected through the wet scrubber we would soon be in violation of our NPDES discharge permit. EPA’s statement that PM control is not required is not accurate; at a minimum, it is necessary to install fabric filter after a carbon injection system. Activated carbon injection in conjunction with wet scrubbers simply will not work. Figure 3: Mercury mass balance for Palo Alto based on August 2010 flue gas, ash, and effluent testing Page 10 of 20 E. EPA Failed to Include all the Costs Associated with Controlling Mercury Beyond- the-Floor for Multiple Hearth Incinerators MH SSIs utilizing activated carbon injection systems, at a minimum, will have to install carbon contact chambers and fabric filters. In the proposed rule EPA states that The incremental cost of adding activated carbon injection to all MH units is estimated to be $12 million per ton of pollutants (Hg and CDD/CDF) removed (or $6,000 per pound of Hg removed). However, it is important to note that activated carbon injection cannot be applied alone. It requires particulate control devices to remove the carbon that is injected to adsorb the Hg. Based on our available data, all of these units have some type of PM control device in place so they would not need to install new PM control equipment. (75 FR 63277) We disagree with EPA’s conclusion, as discussed in Section D. Palo Alto will need to install, at a minimum, an activated carbon injection system and a fabric filter. A carbon contact chamber will likely also be required. Therefore, we recalculated the cost effectiveness associated with achieving beyond-the-floor control of mercury. Our cost calculation includes EPA’s estimated costs for Palo Alto to install an activated carbon injection system and a fabric filter. EPA’s estimated costs are overly conservative because they do not include a carbon contact chamber or any additional pollution control equipment that may be required. As discussed in Section C, EPA overestimated the mercury emissions from Palo Alto’s SSIs; therefore, our cost calculation uses Palo Alto’s actual measured mercury emissions. EPA’s cost effectiveness calculation and our calculation as described above are both provided in Table 2. Our cost effectiveness calculation indicates that Palo Alto’s cost to control the lower than anticipated amount of mercury using activated carbon injection and fabric filter is a staggering $436 million per ton, far greater than the $12 million per ton that is calculated using EPA’s erroneous assumptions. Again, even this $436 million per ton figure is conservative given that additional air pollution control equipment will likely be needed. In addition to evaluating the cost effectiveness specific to Palo Alto, we similarly recalculated the cost effectiveness for all MH SSIs to include EPA’s estimated costs for an activated carbon injection system and a fabric filter in conjunction with the WERF baseline emissions estimate of 0.9 tons per year. This calculation, as well as EPA’s cost effectiveness calculation for MH SSIs, is presented in Table 3. EPA’s assumption that fabric filters are not necessary is incorrect; therefore EPA needs to include these additional costs in its cost analyses for beyond-the-floor control of mercury. The proposed rule estimated the incremental cost of adding activated carbon injection to all multiple hearth incinerators to be $12 million per ton of mercury removed. When the WERF baseline mercury emissions are used and both activated carbon injection and fabric filters are included in the calculations, the cost effectiveness weakens to $188 million per ton. Even the $188 million dollar per ton figure is conservative, given that additional air pollution control equipment will likely be needed. Page 11 of 20 Table 2: Summary of costs for Palo Alto associated with going beyond-the-floor for mercury control. Cost Baseline Emissions and Incremental Emission Reductions (tons/year) TCI ($) TAC ($/year) EPA Estimate Palo Alto Estimate a Palo Alto Multiple Hearth Incinerators Hg (tons) Incremental Cost- effectiveness ($/ton) Hg (tons) Incremental Cost- effectiveness ($/ton) Baseline Emissions - - 0.0316 - 0.00504 - MACT Floor Total Cost and Emission Reductions 0 - - - Fabric Filter $6,877,542 $1,593,646 - - 0.00444 $358,929,279 Afterburner Retrofit $2,513,280 $714,328 - - - - Packed Bed Scrubber $3,720,948 $777,212 - - - - Additional Costs and Emission Reductions by Control Activated Carbon Injection $80,626 $343,366 0.0278 $12,351,295 0.00444 $77,334,685 Total cost per ton $12,351,295 $436,263,964 a. Palo Alto’s estimate uses baseline emissions from actual stack testing and includes the EPA cost estimates for fabric filter and activated carbon injection. Page 12 of 20 Table 3: Summary of multiple cost estimates for multiple hearth incinerators to achieve beyond-the-floor mercury control. Baseline Emissions and Incremental Emission Reductions (tons/year) Cost ($) EPA Estimate Palo Alto Estimate a Multiple Hearth Incinerators Nationwide TCI TAC Hg (tons) Incremental Cost- effectiveness ($/ton) Hg (tons) Incremental Cost- effectiveness ($/ton) Baseline Emissions - - 3.0536 - 0.9 - MACT Floor Total Cost and Emission Reductions $131,764,712 $40,327,113 0.0315 - 0.0093 b - Fabric Filter $478,373,914 $115,254,825 0 - 0.7838 $147,046,217 Afterburner Retrofit $145,514,140 $43,193,966 0 - - - Packed Bed Scrubber $258,596,495 $54,863,534 0 - - - Additional Costs and Emission Reductions by Control Activated Carbon Injection $6,230,844 $32,335,212 2.6235 $12,325,219 0.7838 $41,254,417 Total cost per ton $12,325,219 $188,300,634 a. Palo Alto’s estimate uses the WERF estimate of 0.69 tons for baseline emissions and includes the EPA cost estimates for fabric filter and activated carbon injection. b. Assumes MACT floor reduction from the baseline to be some percentage as the EPA estimate (1.03%) Page 13 of 20 F. EPA Failed to Analyze the Benefits of Source Control When Setting the Beyond- the-Floor Standard for Mercury Section B of this letter discussed the effectiveness of mercury source control and recommended that EPA establish beyond-the-floor mercury standards based on the reductions that are achievable through source control. Source control is also integral to EPA’s consideration of beyond-the-floor mercury standards utilizing activated carbon injection because mercury source control efforts already being planned will lead to reductions regardless of whether EPA establishes beyond-the-floor standards based on source control. These source control efforts will have a significant negative impact on the cost effectiveness of a beyond-the-floor mercury standard based on activated carbon injection technology. EPA’s Office of Water recently announced that it intends to propose a rule in 2011 addressing discharges of mercury from dental facilities. The Office of Water states that the rule is expected to be finalized in 2012, and that the focus will be on amalgam separators. As discussed previously in Section B, source control of dental amalgam mercury is very effective at reducing the amount of mercury entering wastewater treatment plants. Palo Alto has observed a 60 percent reduction in the concentration of mercury in its sludge cake since amalgam separators were required in 2005, and mercury emissions from Palo Alto’s SSIs have similarly declined. Once amalgam separators are in place at dental offices nationwide, consistent with the anticipated Office of Water rulemaking, a reduction in baseline mercury emissions from SSIs of approximately 60 percent can be expected. Table 4 provides a cost effectiveness calculation for EPA’s proposed beyond-the-floor mercury reductions for MH SSIs assuming that baseline mercury emissions have been reduced by 60 percent through source control programs. The calculation includes EPA’s cost estimates for activated carbon injection and fabric filters, and uses the WERF baseline estimate of 0.9 tons per year reduced by 60 percent. The calculation is analogous to the one provided in Table 3, except that the assumed 60 percent reduction from source control results in a smaller removal of mercury through the addition of activated carbon injection and fabric filters. Once it is assumed that the WERF baseline estimate of 0.9 tons per year is reduced by 60 percent due to source control, the cost effectiveness to remove 88 percent of the remaining mercury through activated carbon injection and fabric filters is $470 million per ton of mercury removed. In previous rulemakings for hazardous waste and boiler regulations, EPA considered source control during the beyond-the-floor analysis. We encourage EPA to use similar analysis in its development of beyond-the-floor standards for SSIs. Page 14 of 20 Table 4: Summary of cost estimates to for multiple hearth incinerators to achieve beyond-the-floor reductions of mercury after implementation of a national dental amalgam program. Based on Palo Alto’s dental amalgam results EPA can assume a 60 percent reduction in the amount of mercury entering the wastewater treatment plants. Baseline Emissions and Incremental Emission Reductions (tons/year) Cost ($) Updated EPA Estimate (using 60 percent reduction in baseline) Palo Alto Estimate Multiple Hearth Incinerators TCI TAC Hg Incremental Cost- effectiveness ($/ton) Hg Incremental Cost- effectiveness ($/ton) Baseline Emissions - - 1.2215 a - 0.36 b - MACT Floor Total Cost and Emission Reductions $131,764,712 $40,327,113 0.0126 c - 0.0037 c - Fabric Filter $478,373,914 $115,254,825 0 - 0.3135 $367,638,995 Afterburner Retrofit $145,514,140 $43,193,966 0 - - - Packed Bed Scrubber $258,596,495 $54,863,534 0 - - - Additional Costs and Emission Reductions by Control Activated Carbon Injection $6,230,844 $32,335,212 1.0638 $30,395,950 0.3135 $103,142,622 Total cost per ton $30,395,950 $470,781,617 a. Baseline emission is EPA estimate baseline of 3.0536 tons reduced by 60 percent. b. Baseline emission is WERF baseline of 0.9 tons reduced by 60 percent c. Assumes MACT floor reduction to be the same percentage as the EPA estimate in Table 3 (1.03%) Page 15 of 20 G. EPA Failed to Identify Any Mercury Benefits in the Proposed Mercury Beyond- The-Floor Standard EPA fails to identify or quantify any benefits resulting from mercury reductions in its cost/benefit analysis for the mercury beyond-the-floor proposed limits. It appears that benefits resulting from reductions in emissions of particulate matter benefits were considered instead. Yet the basis for requiring the beyond-the-floor technology was to control mercury, not to reduce particulate matter. It is not appropriate to use particulate matter benefits in a cost/benefit analysis that is examining the benefits of mercury reductions as compared to the cost of removal. As noted previously, the proposed beyond-the-floor control technologies will not necessarily result in actual mercury reduction benefits because mercury readily moves from one media to another and is extremely difficult to actually remove from the environment. An analysis of reductions in particulate matter cannot serve as a surrogate for actual analysis of the benefits of mercury reduction under various alternatives, including source control. EPA’s failure to identify actual mercury benefits in the beyond-the-floor standard renders its analysis deficient in this respect. H. EPA Failed to Adequately Consider Increased Energy Requirements When Setting the Beyond-the-Floor Standard EPA acknowledges that it must consider energy requirements when considering setting beyond-the-floor standards, and is under a statutory obligation to do so.2 However, the analysis found within the document entitled “Secondary Impacts of Control Options for the Sewage Sludge Incineration Source Category” provides no indication that EPA has actually analyzed and considered energy requirements related to the proposed rule. EPA guidance indicates that when examining energy impacts of the proposed beyond-the-floor standard, EPA should address energy use in terms of penalties or benefits associated with a control system and the direct effects of such energy use on the facility. If such benefits or penalties exist, they should be quantified to the extent possible. While the City of Palo Alto does not necessarily agree that the analysis is limited to these considerations, at a minimum EPA should consider potential benefits and penalties in some way when considering a beyond-the-floor standard. We have reviewed numerous documents made available by the EPA in support of the proposed rule, including the "Analysis of Beyond the Maximum Achievable Control Technology (MACT) Floor Controls for Existing SSI Units." and “Secondary Impacts of Control Options for the Sewage Sludge Incineration Source Category.” These documents fail to adequately discuss increased energy requirements associated with the control mechanisms proposed to meet the beyond-the- floor mercury standard. Increased energy use will occur through the addition of a carbon injection system and a fabric filter (bag house), as would be required under the beyond- the-floor standards associated with Option 2, yet the energy use shown for the “MACT 2 See 75 Fed. Reg. 63275 (“EPA may adopt emission limitations and requirements that are more stringent than the MACT floor (i.e., beyond-the-floor). Unlike the MACT floor methodology, EPA must consider costs, nonair quality health and environmental impacts and energy requirements when considering beyond- the-floor standards.”) Page 16 of 20 Floor Only” option is identical to the energy use shown for the other two options. The City can only assume that EPA failed to consider or conduct any analysis for this increased energy use as direct consequence of the proposed control system. EPA must examine these increased energy requirements before establishing a beyond-the-floor mercury requirement for sewage sludge incinerators. I. EPA Failed to Adequately Consider Non-Air Quality Health and Environmental Impacts When Setting the Beyond-the-Floor Standard EPA acknowledges that it has a statutory obligation to consider non-air quality health and environmental impacts when considering beyond-the-floor standards. However, there is no indication that EPA has analyzed the non-air quality health and environmental impacts in setting the proposed beyond-the-floor standard for mercury. EPA guidance indicates that a consideration of environmental impacts should concentrate on collateral environmental impacts due to control of emissions of the pollutant in question, such as solid or hazardous waste generation and discharges of polluted water from a control device. At a minimum, EPA is required to evaluate any health and environmental impacts that may result directly or indirectly from measures that will achieve the emission reductions.3 As noted in other sections of these comments, EPA has failed to consider a number of significant environmental impacts resulting from the chosen method of mercury control. For example, the proposed wet scrubber control mechanism merely returns mercury to the water stream and fails to prevent release to the environment. As a direct result of implementing the proposed technology, increased levels of mercury will be shifted to water, clearly a non-air quality environmental impact. The beyond-the-floor analysis and other supporting materials do not address this probable effect, nor do they address whether and why release of mercury with effluent discharged to surface waters is a superior alternative to air release with respect to environmental goals. In addition, attempting to dispose of mercury in landfills, as would be required in a number of locations under the proposed rule, could result in ultimate volatilization and release of the pollutant. This is a potentially significant environmental impact. While the impact on landfills was addressed as an alternative and considered when performing the cost-benefit analysis, it was only examined through that lens and failed to consider the environmental impacts and consequences of that option. Increased landfilling could potentially have additional non-air quality environmental impacts that were not considered by EPA. The beyond-the-floor analysis also fails to consider the potential hazardous waste generation that may result from the proposed rule. EPA’s document titled “Secondary Impacts of Control Options for the Sewage Sludge Incineration Source Category” estimates that greater than 10,000 tons per year of activated carbon would be used by the activated carbon injections systems that would be needed to comply with the beyond-the-floor mercury standard. Given its use in adsorbing mercury, this activated carbon and associated fly ash are likely to require management as hazardous waste. EPA should 3 See Sierra Club v. EPA, 353 F.3d 976, 990 (D.C. Cir. 2004) Page 17 of 20 examine these environmental impacts and others before establishing a beyond-the-floor mercury requirement for sewage sludge incinerators. J. Landfilling Undigested Biosolids is Not an Option in California, and Anaerobic Digestion is Not an Economically Preferable Option to Incinerator Upgrades In the proposed rule, EPA assumes that small POTWs that currently incinerate biosolids will decide that it is more economical to simply landfill their biosolids, and will abandon their incineration facilities. Although the proposed rule did not suggest that larger POTWs such as Palo Alto would switch to landfilling, it is important to point out the infeasibility of landfilling biosolids as an alternative to incineration for Palo Alto. In California, the Integrated Waste Management Act of 1989 (AB 939) established solid waste diversion requirements for local jurisdictions. Palo Alto must continue to reduce the volume of its solid waste that is landfilled, and has implemented an extensive and ongoing Zero Waste Program to continue this progress. Landfilling of undigested biosolids as an alternative to incineration would prevent Palo Alto from meeting the diversion requirements. Additionally, most California landfills, including Palo Alto’s own landfill, are unable to accept wastes with a solids content less than 50 percent. While biosolids are currently used in California landfills for alternative daily cover this is not a sustainable practice, in that it is really no different than landfill disposal. Even if Palo Alto were to engage in landfilling biosolids, digestion would be required first. Palo Alto has recently developed a planning level estimate of costs for building anaerobic digestion facilities that could replace incineration of biosolids. The total project capital cost was estimated to be $42.2 million, with an annualized project cost of $2.7 million. Total annual operating cost was estimated to be $0.42 million not including ultimate disposal costs. Clearly, anaerobic digestion of Palo Alto’s biosolids is not an economically attractive option that could easily be implemented in lieu of complying with activated carbon injection system based beyond-the-floor mercury standards. Palo Alto will almost certainly not be able to switch to landfilling for all of the reasons given above. K. New Source Performance Standards Will Limit Future Upgrades to Palo Alto’s Incinerator; New Source Performance Standards Should be Separately Developed for Multiple Hearth and Fluidized Bed SSIs The proposed performance standards for new incinerators are based on a very limited amount of air emissions test data obtained from three POTWs utilizing FB SSIs (i.e., St. Paul Metro Plant; Ypsilanti, MI WWTP; and the Greensboro, NC T.Z. Osborne WWTP). Under the Information Collection Request, EPA obtained air emissions data for four of the five SSIs located at these plants. In developing the proposed new source performance standards, EPA used individual one-hour test runs instead of an average of three one-hour test runs. This methodology contradicts the test methods, and is not valid. The average of three one-hour tests is more representative of SSI steady state operations than any individual one-hour test. Sludge variability, operational variances, and seasonal temperatures have not been factored into the criteria; therefore, the same unit may not Page 18 of 20 meet the limit in future tests. EPA must collect sufficient data to allow the recalculation of the new source performance standards using averages of three tests, as is the intention of the test methods. EPA established the proposed new source performance limits by taking the lowest number from the best performing units. EPA must realize that carbon monoxide and oxide of nitrogen emissions from SSIs are inversely proportional. As a result, by selecting a carbon monoxide limit from one SSI and an oxides of nitrogen limit from a different SSI, EPA is establishing limits that no new SSI will be able to meet. EPA’s approach would discourage incremental improvements at MH units because these improvements would trigger FB-based emission limits that cannot be met. EPA has acknowledged the design differences that make meeting these limits impossible, but has provided no pathway by which modified or reconstructed MH incinerator units may achieve compliance. EPA must retain the separate MH and FB incinerator subcategories for both new and existing sources to avoid subjecting newly-constructed or modified MH SSIs to unachievable emission standards. Establishing separate emission limits for new MH and FB incinerators will also preserve incentives for innovation and for improvements in the operation of MH incinerators currently in use. L. Annual Air Emissions Testing Requirements The proposed rule requires annual compliance testing of each of the 218 sewage sludge incinerators at a cost of $61,000 per unit, in addition to the $91,500 per unit for initial testing. While this cost is accurate, Palo Alto disagrees with the need for annual compliance testing on each incinerator. Palo Alto has two identical incinerators and only operates one incinerator per year. It will be very costly to place an identical, non- operating incinerator into service simply for an air emissions test. Palo Alto recommends that if EPA moves forward with the proposed emissions guidelines that EPA take the following approach: after the initial testing of each incinerator, subsequent testing be limited to once every five years. M. Certain Proposed Operating Standards are Neither Necessary Nor Achievable EPA has proposed a number of operating standards for sewage sludge incinerators that are not practical, achievable, or necessary, as follows: 1. Sludge moisture content EPA has proposed that the sludge moisture content be measured on a daily basis, and that it be limited to a range from 10 percent less than to 10 percent greater than the average sludge moisture content during the most recent air emissions test. The example given is that if the moisture content during the most recent test was 20 percent, then the moisture content of the sewage sludge would have to be within 18 and 22 percent every day. Page 19 of 20 This is a requirement that is impossible to continuously meet, since the moisture content of the sewage sludge varies. The variability of the moisture content does not impact air emissions, and is therefore not an important factor on which to place limitations. 2. Minimum Scrubber Pressure Drops, Scrubber Flow Rates, Scrubber Liquor pH, and Combustion Temperatures. EPA is proposing that the minimum pressure drop across each wet scrubber, the minimum scrubber liquor flow rate, the minimum scrubber liquor pH, and the minimum combustion temperature (or the minimum afterburner temperature) shall be calculated at 90% of the value determined during the most recent air emissions test. These minimums will be impossible to continuously meet and contradict the operating parameters contained in 40 CFR Part 60, Subpart O and 40 CFR Part 503. 3. Maximum Feed Rate EPA is proposing that the maximum dry sludge feed rate shall be calculated at 110 percent of the average dry feed rate during the most recent performance test. EPA must determine at which feed rate each SSI needs to be operated during the test. If the test feed rate is at the SSI’s normal feed rate, then the proposed maximum feed rate will be exceeded on a regular basis. The only way that this operational standard will not be exceeded is if the test feed rate is at or near the maximum permitted capacity. As a result, Palo Alto requests that the proposed operating parameter section of the rule be revised. N. The CEMS Averaging Times Should Remain as 24-Hour Block Averages EPA requested comments on changing the averaging times for all CEMS and CASS from 24-hour block averages to 12-hour rolling averages to be consistent with the averaging times of the PS tests. Palo Alto strongly recommends that EPA retain the current 24-hour block averages because 24-hour block averages are more accommodating of the significant variability in the sludge cake feed to SSIs. Additionally, 40CFR503 requirements for SSIs require 24-hour averages. Changing the averaging period to 12 hours under this rule would increase the administrative burden on SSI operators by requiring calculation and reporting in two different formats. In conclusion, EPA has not demonstrated that the proposed beyond-the-floor requirements for mercury are achievable. Nor has EPA correctly identified the costs and benefits of these controls. Most importantly, EPA has not fairly considered the most cost efficient and more environmentally sound options for source control. Clearly it is source EXECUTIVE COMMITTEE PRESIDENT Jeff Theerman Executive Director Metropolitan St. Louis Sewer District Saint Louis, MO VICE PRESIDENT David R. Williams Director of Wastewater East Bay Municipal Utility District Oakland, CA TREASURER Suzanne E. Goss Government Relations Specialist JEA (Electric, Water & Sewer) Jacksonville, FL SECRETARY Julius Ciaccia, Jr. Executive Director Northeast Ohio Regional Sewer District Cleveland, OH PAST PRESIDENT Kevin L. Shafer Executive Director Milwaukee Metropolitan Sewerage District Milwaukee, WI EXECUTIVE DIRECTOR Ken Kirk November 29, 2010 VIA ELECTRONIC SUBMISSION EPA Docket Center (EPA/DC) United States Environmental Protection Agency (6102T) Proposed Rulemaking – New Source Performance Standards and Emission Guidelines for Sewage Sludge Incinerators Docket ID No. EPA-HQ-OAR-2009-0559 1200 Pennsylvania Avenue NW Washington, DC 20460 Via Electronic Mail: a-and-r-docket@epa.gov Re: Comments of the National Association of Clean Water Agencies on the Proposed Standards of Performance for New Stationary Sources and Emission Guidelines for Existing Sources: Sewage Sludge Incineration Units, 75 Fed. Reg. 63260 (October 14, 2010) The National Association of Clean Water Agencies (“NACWA”) appreciates the opportunity to comment on the United States Environmental Protection Agency’s (“EPA” or “Agency”) request for public comments on the Proposed Standards of Performance for New Stationary Sources and Emission Guidelines for Existing Sources: Sewage Sludge Incineration Units, 75 Fed. Reg. 63260 (October 14, 2010) (“Proposed Rule”) under § 129 of the Clean Air Act (“CAA”). NACWA represents the interests of nearly 300 of the nation’s publicly owned treatment works (“POTWs”), which collectively serve the majority of the sewered population of the United States, including approximately half of the sewage sludge incinerators that will be affected by this rulemaking. For over 40 years, NACWA has maintained a leadership role in legal and policy issues affecting the public agencies responsible for cleaning the nation’s wastewater, and has been at the forefront of the development and implementation of scientifically-based, technically-sound and cost-effective environmental programs for protecting public and ecosystem health. EPA’s Proposed Rule, if finalized, would have a significant negative impact on the NACWA members that use incineration to responsibly manage the thousands of tons of sewage sludge generated on a daily basis from the treatment of the nation’s wastewater. Although wastewater treatment agencies rely on having multiple options for managing their sludge, the list of available options has slowly NACWA Comments on Docket ID No. EPA-HQ-OAR-2009-0559 November 29, 2010 Page 2 shrunk over the years for many communities. EPA’s proposed action, if finalized without significant revision, would severely limit the use of incineration – which is currently used to manage approximately a fifth of the sludge generated annually in the U.S. – into the future and eviscerate progress toward a new, viable source of renewable energy for the country. In fact, the extremely stringent proposed new source performance standards could effectively eliminate the construction of new sewage sludge incinerators (“SSIs”), with no current incinerator manufacturers indicating that they will be able to meet the new limits. As discussed in more detail below, NACWA believes EPA should abandon this rulemaking and return to its previous plan to regulate SSIs under § 112 of the CAA. The technical corrections discussed in the comments below are warranted under any future rulemaking effort, whether pursuant to § 129 or § 112. NACWA encourages EPA to take the time now to develop a more thorough and accurate understanding of this sector. NACWA’s major concerns and recommendations include: 1. EPA should recognize that the CAA requires that POTWs (including their SSIs) be regulated under § 112 and not under § 129. NACWA urges EPA not to proceed with the promulgation of standards for SSIs under CAA § 129 and instead to return to EPA’s earlier approach of addressing SSIs under CAA § 112. Congress directed that POTWs be regulated under § 112, and this includes the incinerators that POTWs operate to manage their sludge. EPA also faces legal obstacles to using § 129 to regulate SSIs because they do not fit the statutory definition of “solid waste incineration unit.” As discussed more completely below, EPA faces a more defensible path under § 112 for regulating POTW-operated incinerators. 2. EPA should acknowledge the value of incineration as a local option for residuals management and as an emerging renewable energy source for generating electricity and steam. Under the Clean Water Act (“CWA”), Congress reserved residuals management decisions to the local POTWs because of their unique ability to balance competing interests and make sound cost determinations on a site-specific basis. See CWA § 405(e) (“The determination of the manner of disposal or use of sludge is a local determination.”). NACWA has long argued that local communities must have the flexibility to choose the residual management approach that works best for them. NACWA’s members have for decades been balancing the complex concerns of local communities over residual management options. As examples, The Metropolitan District’s (Hartford, Connecticut) nearest landfill option is 375 miles away so its choice to incinerate results in less air pollutant emissions than the trucks would emit while transporting its sludge to landfill. The small community of Edmonds, Washington chose incineration because it is 270 miles from the closest landfill and Washington State may not allow that landfill to take sewage sludge. The Northeast Ohio Regional Sewer District (“NEORSD”) recently revisited its choice to incinerate and concluded that it was the “greenest” option available for managing residuals. NEORSD is proceeding with a project to install new fluidized bed incinerators that will generate enough electricity to power 1800 homes, will minimize odors, and will avoid the extra fuel and air emissions from 8-12 trucks per hour NACWA Comments on Docket ID No. EPA-HQ-OAR-2009-0559 November 29, 2010 Page 3 transporting sludge through its surrounding communities to a landfill that is 65 miles away. As Congress has recognized and affirmed, sewage sludge management is a responsibility uniquely suited for local control, where incineration is often the best and cleanest option for residuals management. The Proposed Rule oversimplifies the decision making process involved in choosing a management option for sewage sludge and, as currently drafted, will make incineration a less viable option for all communities in the future. 3. EPA’s rule writers should be given the time, data, and resources to understand POTWs and SSIs sufficiently to regulate them in an appropriate manner within the bounds of existing legal requirements. EPA’s rush to promulgate this rule has been done at the expense of EPA’s understanding of POTWs, the nature and amount of sewage sludge that they manage, and the emissions that these incinerators actually generate. Mistakes in the Proposed Rule reveal the depth of this misunderstanding and necessitate fundamental changes to the rule and the analysis that EPA uses to support it. NACWA has asked repeatedly for EPA to slow the rulemaking pace to allow for a more reasoned approach. We ask that EPA withdraw the Proposed Rule until it can collect sufficient data to support an appropriate rulemaking. 4. EPA’s fundamental misunderstanding of sewage sludge and the relative cost of landfilling undermine the credibility of its analysis of the rule’s impact on small entities. EPA assumed that “dewatered” sludge contains no moisture. In fact, the average moisture content in dewatered sludge is still 70-80 percent. EPA assumed that dry tons of sludge (instead of wet tons) would be transported to landfill, which underestimated the amount of sludge requiring storage, handling, odor control, and transport by a factor of three to five. EPA relied on this error in finding No Significant Impact on a Substantial Number of Small Entities (“No SISNOSE”) under the Regulatory Flexibility Act (“RFA”) and the Unfunded Mandates Reform Act (“UMRA”). Given as much as a five-fold increase in the assumed landfill costs, EPA is not justified in concluding that all small entities will prefer to landfill, and it must reconsider the adverse effects of this rule on small entities. EPA cannot know the local calculus for alternative residual management options, but it should start with correct assumptions regarding the amount of sewage sludge that must be diverted and the costs associated with that diversion. 5. EPA should revisit its beyond the floor mercury control determination after correcting the overestimation of the amount of mercury generated by POTWs and after considering more cost effective pollution prevention alternatives for mercury reductions. EPA significantly overestimated the baseline mercury emissions used in determining the cost effectiveness of beyond the floor maximum achievable control technology (“MACT”) controls for mercury. EPA also significantly underestimated the cost of mercury controls applied to SSIs. As a result, the cost effectiveness calculation used to justify beyond the floor mercury controls for SSIs is flawed and must be revisited. Local pollution prevention efforts targeting dental NACWA Comments on Docket ID No. EPA-HQ-OAR-2009-0559 November 29, 2010 Page 4 offices demonstrate that mercury reduction can be achieved far more reliably and cost effectively upstream of the POTW than by installing add-on control devices at the incinerator. 6. EPA’s decision to limit data collection to stack tests at nine facilities undercuts proper MACT floor determinations for the SSI subcategories. EPA decided to circumvent the Paperwork Reduction Act (“PRA”) and the Office of Management and Budget (“OMB”) review and comment period by limiting its request for information to nine of the 118 POTWs operating SSIs. EPA then applied statistical methods to extrapolate from this limited data set as if it were a random sample of data representative of SSIs as a whole. The data are not random because the nine POTWs from which the data are drawn were selected based on the type of pollution control equipment employed, without considering sludge characteristics and other factors that affect SSI emissions. The data also are not representative because the tests were snap shots that do not reflect seasonal, regional, or inherent variability in the domestic sewage treated by POTWs. EPA should use other available data to understand the variability and to incorporate that variability into the MACT floor determinations. 7. EPA should use all of its available discretion to mitigate the impact of a rulemaking on public entities already financially over-burdened and resource constrained. POTWs are facing significant costs to upgrade the infrastructure necessary to clean the nation’s domestic wastewater to meet federal standards. Resources are not available for public entities to realistically meet all existing and proposed federal mandates. EPA must ensure that every dollar of implementation cost is absolutely necessary and required. Stack testing should be done every five years, not every year. New monitors should not be required when existing sludge sampling can be used to track compliance. Public entities should be entrusted to develop an operation and maintenance plan for their control devices that utilizes existing data to the maximum extent possible to minimize costs. Stack tests are not a reliable basis for setting operating parameter ranges that vary with the sludge characteristics. 8. EPA should use subcategories to set achievable standards and to further relieve the compliance burden on POTWs. EPA has broad discretion to subcategorize SSIs to make the limits more achievable. EPA uses just two subcategories (fluidized bed (“FB”) incinerators (“FBIs”) and multiple hearth (“MH”) incinerators (“MHIs”)) for existing sources and a single subcategory for new sources. EPA inappropriately assumes that no new MHIs will be constructed or modified. By requiring all new sources to meet FBI standards, EPA is precluding upgrades to MHIs that could improve efficiency and performance for some pollutants. EPA should also create subcategories for limited use units, small units, and isolated units that are so far from a usable landfill that emissions of hazardous air pollutants (“HAP”) from incineration are less than HAP emissions from transporting sludge to landfill. NACWA Comments on Docket ID No. EPA-HQ-OAR-2009-0559 November 29, 2010 Page 5 EPA’s errors in the Proposed Rule indicate that the Agency has been rushed to regulate without a clear understanding of POTWs and the incinerators they operate. Additional time is needed to understand the unique regulatory context for POTWs that dictates their regulation under § 112 of the CAA. Additional time is also needed to understand the nature of sewage sludge and its inherent variability and to understand the unique attributes associated with incinerating a material that is 70-80 percent water. NACWA urges EPA to take this SSI rulemaking off the fast track so that it can be done right. Congress under the CWA wisely left POTWs with the power to choose the residual management option that best meets the needs of each local community after directing EPA’s Office of Water to ensure that public health was adequately protected under every option. See 40 CFR Part 503. EPA should use the authority Congress granted under CAA § 112 to regulate the remaining HAP in a way that preserves incineration as a viable residuals management option. I. SEWAGE SLUDGE INCINERATORS MUST BE REGULATED UNDER CAA § 112 Congress has directed EPA to regulate SSIs in a way that is unique from all other incineration source categories. However, EPA’s Proposed Rule fails to recognize these differences and instead treats SSIs like one of the several categories of solid waste incineration units to be regulated under CAA § 129. This failure results in numerous legal and technical flaws in the Proposed Rule that EPA must correct. A. Incinerators at POTWs Historically Have Been Successfully Regulated Under the CWA Since 1993, POTWs that practice incineration have been subject to a comprehensive, risk-based program for reducing the potential environmental risks of sewage sludge pursuant to CWA § 405 and the implementing regulations set forth in 40 CFR Part 503. Section 405(d) of the CWA requires EPA to establish numeric limits and management practices that protect public health and the environment from the adverse effects of toxic pollutants in sewage sludge. Section 405(e) prohibits any person from disposing of sewage sludge from a POTW or other treatment works treating domestic sewage through any use or disposal practice for which regulations have been established pursuant to § 405, except in compliance with the Part 503 regulations. In the Part 503 regulations, EPA has identified the pollutants in sewage sludge that may adversely affect public health or the environment and has specified the management practices for the utilization and disposal of sewage sludge that are protective of public health and the environment. For disposal by incineration, the Part 503 regulations mandate, among other requirements: (i) Numerous management practices and general requirements; (ii) Risk-based, site-specific limits for arsenic, cadmium, chromium, lead, and nickel content in the sewage sludge incinerated; (iii) Compliance with National Emission Standards for Hazardous Air Pollutants (“NESHAPs”) for mercury and beryllium (as discussed below); NACWA Comments on Docket ID No. EPA-HQ-OAR-2009-0559 November 29, 2010 Page 6 (iv) Operational technology-based emission limits for total hydrocarbon (“THC”) or an alternative emission limit for carbon monoxide (“CO”); and (v) Monitoring, recordkeeping and reporting requirements. 40 CFR Part 503, Subpart E. Furthermore, in the course of developing the Part 503 regulations, EPA also proposed to establish requirements for dioxins (including specific congeners of dioxin, dibenzofuran, and coplanar PCBs).1 However, after evaluating the emissions of dioxins from sewage sludge incineration, as well as surface disposal and land application, EPA decided such requirements were not warranted.2 This decision was based on the results of a comprehensive risk assessment that demonstrated that dioxin levels in biosolids and biosolids incinerator exhaust gases do not pose a significant risk to human health or the environment.3 As explained in detail in NACWA’s 1997 comments to EPA (pages 15-17), the numeric emission limits and management practices requirements established under the Part 503 regulations were derived from years of study and evaluation of the potential risks to human health and the environment which could be posed by the incineration of sewage sludge.4 (Attachment A – Compendium of NACWA Correspondence to EPA) The regulation of SSIs under this existing regime is via risk-based standards that were developed to protect human health and the environment from any reasonably anticipated adverse effects from pollutants that may be present in sewage sludge. In fact, the Part 503 regulations for SSIs were developed through a partnership between EPA’s water and air offices – a partnership that continues today as EPA conducts its mandated review of the Part 503 standards. As a result, SSIs can clearly demonstrate that the emissions from their units are not adversely impacting human health and the environment by maintaining compliance with the Part 503 requirements. Moreover, the statutory framework of this regime provides ample means for EPA to identify and regulate additional concerns if supported by scientific evidence. For example, CWA § 405 provides for a biennial review process that was specifically established for identifying and regulating any additional pollutants of concern. EPA has repeatedly emphasized its confidence that the Part 503 regulations are adequately protective of public health and the environment.5 Additionally, since 1975 EPA has imposed NESHAPs for mercury and beryllium emissions that apply to certain SSIs. See 40 CFR Part 61, Subpart E and C. The mercury NESHAP applies, in relevant part, to any source that incinerates sludges from wastewater treatment plants. The NESHAP imposes emission limits for mercury, as well as stack testing, sampling, and monitoring requirements. See 40 CFR Part 61, Subpart E. The beryllium NESHAP applies, in relevant part, to incinerators that process beryllium-containing waste. This 1 See Standards for the Use or Disposal of Sewage Sludge: Proposed Rule, 64 Fed. Reg. 72,045 (Dec. 23, 1999). 2 See Standards for the Use or Disposal of Sewage Sludge: Final Notice, 66 Fed. Reg. 66,028 (Dec. 21, 2001). 3 See id. 4 Prior to 2006, NACWA was known as the Association of Metropolitan Sewerage Agencies (AMSA). 5 See Letter from James A. Hanlon, Director of EPA Office of Wastewater Management, to Greg Kester, State of Wisconsin Department of Environmental Resources (Sept. 20, 2004) (“EPA believes that 40 CFR Part [503] regulations are protective of public health and the environment and we continue to support biosolids management in full compliance with the Part 503 regulation.”) submitted with NACWA’s August 14, 2006 comments on the reconsideration of the final Other Solid Waste Incinerator rule, included in Attachment A. NACWA Comments on Docket ID No. EPA-HQ-OAR-2009-0559 November 29, 2010 Page 7 NESHAP imposes emission limits for beryllium, as well as sampling requirements. See 40 CFR Part 61, Subpart C. These NESHAPs are expressly incorporated into the 40 CFR Part 503 requirements for POTWs. Emissions from SSIs are already regulated by other Congressionally-mandated, comprehensive regulations that are adequately protective of human health and the environment. Accordingly, no public health or environmental benefit will be realized from including SSIs under CAA § 129. B. The 1990 CAA Amendments Direct EPA to Regulate POTWs Under § 112 CAA § 112(e)(5) requires EPA to establish NESHAP for POTWs. Section 112(e)(5) states: The Administrator shall promulgate standards pursuant to subsection (d) of this section [112] applicable to publicly owned treatments works (as defined in title II of the Federal Water Pollution Control Act [33 U.S.C.A. § 1281 et seq.]) not later than 5 years after November 15, 1990. 42 U.S.C. § 7412(e)(5). The definition of “treatment works” contained in Title II is broad and includes: … any devices and systems used in the storage, treatment, recycling, and reclamation of municipal sewage or industrial wastes of a liquid nature to implement [section 201 of the Act], or necessary to recycle or reuse water at the most economical cost over the useful life of the works, including intercepting sewers, outfall sewers, sewage collection systems, pumping, power, and other equipment, and their appurtenances; extensions, improvements, remodeling, additions, and alterations thereof; elements essential to provide a reliable recycled supply such as standby treatment units and clear well facilities; and any works, including site acquisition of the land that will be an integral part of the treatment process (including land used for the storage of treated wastewater in land treatment systems prior to land application) or is used for ultimate disposal of residues resulting from such treatment. 33 U.S.C. § 212(2)(A) (emphasis added). This language clearly encompasses the areas of a POTW used to manage sewage sludge, including the incinerators that are “used for ultimate disposal of residues resulting from” the sewage treatment process. Congress’ intentional use of this well-understood term in CAA § 112(e)(5) has no other conceivable meaning. Likewise, EPA’s regulatory definition of “treatment works” makes clear the expansive meaning of the term under the CWA. The definition at 40 CFR § 35.905 includes: Any devices and systems for the storage, treatment, recycling, and reclamation of municipal sewage, domestic sewage, or liquid industrial wastes used to implement section 201 of the Act, or necessary to recycle or reuse water at the most NACWA Comments on Docket ID No. EPA-HQ-OAR-2009-0559 November 29, 2010 Page 8 economical cost over the useful life of the works. These include intercepting sewers, outfall sewers, sewage collection systems, individual systems, pumping, power, and other equipment and their appurtenances; extensions, improvement, remodeling, additions, and alterations thereof; elements essential to provide a reliable recycled supply such as standby treatment units and clear well facilities; and any works, including site acquisition of the land that will be an integral part of the treatment process or is used for ultimate disposal of residues resulting from such treatment (including land for composting sludge, temporary storage of such compost, and land used for the storage of treated waste water in land treatment systems before land application); or any other method or system for preventing, abating, reducing, storing, treating, separating, or disposing of municipal waste or industrial waste, including waste in combined storm water and sanitary sewer systems. (emphasis added). All SSIs are located within the boundaries of a POTW and nearly all are owned and operated by the municipalities that own and operate the POTW.6 As Figure 1 (below) depicts, incinerators are integrated physically and operationally into the solids management process making them an essential part of the solids treatment process. Incinerators are designed and operated for the specific sludge volume, water removal capabilities (i.e., sludge thickening, chemical or thermal conditioning and dewatering), and other unique characteristics of a particular POTW. 6 NACWA is aware that two or three SSIs may be privately owned and/or operated by sludge management companies. These units are still located at a POTW, but for financial reasons the municipality may have chosen to establish a contract-operation agreement with a private entity or to relinquish ownership to the private entity. These SSIs are still integral to the operation of the POTW and are equally essential to reducing, treating and disposing of the sewage sludge generated by these POTWs. NACWA Comments on Docket ID No. EPA-HQ-OAR-2009-0559 November 29, 2010 Page 9 Figure 1: Furthermore, EPA has long viewed SSIs as included within the CWA “treatment works” definition through its implementation of the CWA Title II construction grant program. EPA consistently approved funding for the construction or upgrade of incinerators through its Title II grant fund, which is specifically limited to “treatment works” as defined above. NACWA has collected information from its members indicating that many SSIs were constructed and/or upgraded using Title II funding. The following NACWA members have confirmed that they received Title II “treatment works” funds to construct and/or upgrade SSIs: • Northeast Ohio Regional Sewer District, Cleveland, Ohio (all seven existing incinerators) • Hampton Roads Sanitation District, Virginia Beach, Virginia • Central Contra Costa Sanitary District, Martinez, California • The Metropolitan District, Hartford, Connecticut • The City of Greensboro Water Resources Department, Greensboro, North Carolina Indeed, it may well be the case that many or all of the 234 SSIs7 operated by municipalities in the US were constructed or upgraded with CWA Title II construction grants. This would have been contrary to the CWA, and EPA would have illegally approved the use of millions of dollars in grant funds, if the Agency did not consider SSIs to be part of the “treatment works.” 7 The Water Environment Research Foundation counted 234 active biosolids incineration units in its 2009 WERF Report at ES-1 (Attachment B). NACWA asks that EPA take the time to confirm the number of incinerators and that it include dormant incinerators with active operating permits in the total number of SSIs considered for this rulemaking. Sludge Processing Train NACWA Comments on Docket ID No. EPA-HQ-OAR-2009-0559 November 29, 2010 Page 10 EPA’s recognition that sewage sludge management, including incineration, is an inherent part of POTW operations for which federal funding was made available is also reflected in the primary regulatory program for sewage sludge management, the Part 503 program.8 This was developed under the authority of CWA § 405 and RCRA. For example, EPA has stated that: Sewage sludge has been an important concern of the Agency since 1972, when EPA, through the Federal Water Pollution Control Act construction grants program, began assisting in the financing of wastewater treatment facilities. 58 Fed. Reg. 9248, 9260 (February 19, 1993). Treatment works treating domestic sewage, as noted above, include facilities dedicated to the disposal of sewage sludge (i.e., surface disposal sites and incinerators). Id. at 9359; see also 40 CFR § 122.2, in which “treatment works” is expressly defined to include sewage sludge treatment systems. The legislative history of CAA § 112 also indicates that Congress intended the air emissions from POTW operations covered by CAA § 112 to include air emissions from SSIs. For example, Congress indicated that [t]he Administrator is specifically directed to include publicly owned treatment works (as defined in the Clean Water Act) and [certain RCRA facilities] among the categories of major sources pursuant to this subsection . . . . The Agency has also indicated that air emission standards for POTWs may be promulgated under the Clean Water Act. There is no standard of protection of either human health or the environment from releases to air under that Act. It is more likely that appropriate standards would survive a legal test, if established pursuant to these new authorities of § 112 of the Clean Air Act.9 Senate Report No. 100-231, Committee on Environment and Public Works, 1990 CAA Legislative History 9436, 9668. The CWA-derived air emission standards that Congress intended to supplement with CAA § 112 include those that control emissions on SSIs. Thus, Congress was well aware that air emissions from SSIs were already regulated by the CWA § 405 requirements and chose § 112 as the means to update these requirements as warranted by the applicable protection requirements. In addition, POTWs and SSIs are both included on a Congressional list of source categories intended to be regulated under CAA § 112.10 By contrast, as NACWA has 8 40 CFR Part 503, Standards for the Use or Disposal of Sewage Sludge. 9 Senate Report No. 100-231, Committee on Environment and Public Works, 1990 CAA Legislative History 9436, 9668. 10 Senate report No. 101-228, Committee on Environment and Public Works, 1990 CAA Legislative History 8338, 8528. NACWA Comments on Docket ID No. EPA-HQ-OAR-2009-0559 November 29, 2010 Page 11 noted in previous correspondence to EPA, the legislative history of CAA § 129 is silent as to both POTWs and SSIs. See Attachment A. Further demonstrating Congress’ clear direction, EPA has already regulated SSIs as intended under CAA § 112 by identifying SSIs as an area source category under this section. EPA examined the issue of the CAA regulation of SSIs in 1992, when it issued its initial list of major and area source categories under § 112. See 57 Fed. Reg. 31576 (July 16, 1992). That initial list included SSIs as a § 112 source category. In this notice EPA expressly states “the Agency does not consider sewage sludge incineration units to be covered under § 129 so it has authority to list and set standards for these units under § 112.” See also 58 Fed. Reg. 9248, 9262, 9276-77 (Feb. 19, 1993) (noting that SSIs are regulated under § 112 of the CAA). In 1999, EPA promulgated a NESHAP under § 112 for POTW treatment plants. See 64 Fed. Reg. 57572 (Oct. 26, 1999). Significantly, while the definition of “POTW treatment plant” is appropriately focused on the treatment part of a POTW, the definition of POTW in that rule is much broader and encompasses everything that is eligible to receive grant assistance under Title II of the CWA. See 40 CFR § 63.1595. While EPA’s rulemaking preamble inexplicably stated that EPA then believed that SSIs were subject to regulation under CAA § 129, EPA nowhere elaborated on the rationale for this change from its prior approach or justified this new interpretation of the statute. See 62 Fed. Reg. 1868 (Jan. 14 1997). NACWA has previously raised this issue with EPA, including in the September 2009 correspondence mentioned in the preamble to the Proposed Rule. See Attachment A. Now EPA claims that it “has taken the position in its regulation of POTWs under the CAA that § 112(e)(5) does not apply to SSI units and for this reason did not regulate them in its POTW § 112(d) emission standards.” 75 Fed. Reg. at 63264. Yet EPA still offers nothing but this bare statement to justify its interpretation. EPA does not explain how it reaches the conclusion that SSIs are not covered by the expansive CWA definition of “treatment works” incorporated into CAA § 112(e)(5), how its interpretation is possible in light of the integral role SSIs play in the management of sewage sludge, or how SSIs could have been built and improved using CWA Title II funds if they are not part of the CWA definition of “treatment works.” Further, EPA also fails to recognize its own contrary statements and actions. In fact, shortly after EPA promulgated the POTW NESHAP, EPA reversed its position and expressly stated that SSIs would be regulated under § 112 instead of §129. See Unified Agenda 65 Fed. Reg. 23459-01 (Apr. 24, 2000). Then, in February 2002, EPA revised its list of source categories under § 112 to delete SSIs, not because they were not covered by § 112, but because there were no major sources in that category. See 67 Fed. Reg. 6521 (Feb. 12, 2002). EPA then added SSIs to the list of area source categories under §§ 112(c) and 112(k) of the CAA. See 67 Fed. Reg. 43112 (Jun. 26, 2002); 67 Fed. Reg. 70427 (Nov. 22, 2002). SSIs remain on EPA’s current list of area source categories. See http://www.epa.gov/ttn/atw/area/70list.pdf (last visited Nov. 29, 2010). CAA § 129(h)(2) states that: “no solid waste incineration units subject to performance standards under [§§ 129 and 111] shall be subject to standards under [§ 112(d)].” Thus, the language of CAA § 129(h) makes clear that EPA’s regulation of sources under CAA § 129 or CAA § 112 is mutually exclusive. EPA has consistently recognized that sources regulated under CAA § 112 cannot also be regulated under CAA § 129. Since area source categories are subject to the promulgation of emission standards under CAA § 112(d), SSIs cannot also be regulated under CAA § 129. As the D.C. Circuit directed EPA in NRDC v. EPA, 489 F.3d 1250 NACWA Comments on Docket ID No. EPA-HQ-OAR-2009-0559 November 29, 2010 Page 12 (D.C. Cir. 2007), EPA must follow the plain language of the statute. In the case of SSIs, the plain language of the statute directs EPA to regulate POTWs (including SSIs) under § 112. The court in the NRDC case did not address § 112(e)(5), nor did it specifically address the regulation of POTWs or SSIs, so the decision in that case has no bearing on how SSIs are to be regulated under the CAA. Likewise, nothing in § 129 suggests that SSIs are not to be treated as part of the POTW to be regulated under § 112, and it is axiomatic that the specific direction of Congress contained in § 112(e)(5) is controlling over the general provisions of § 129 dealing with solid waste incineration. “However inclusive may be the general language of a statute, it will not be held to apply to a matter specifically dealt with in another part of the same enactment.” Fourco Glass Co. v. Transmirra Products Corp., 353 U.S. 222, 228 (1957). The bottom line is that CAA § 112(e)(5) unambiguously directs EPA to set emission standards for HAPs from SSIs under § 112(d), and this section does not contain language permitting EPA discretion to regulate SSIs under § 129. While Congress under § 112(d)(1) allows EPA to promulgate differing standards for major and area sources at POTWs and for differing categories of sources at POTWs, the Agency emphatically does not have the authority to set standards under § 129 for POTWs or parts of POTWs, including SSIs. Indeed, EPA has already adopted regulations based on this conclusion. While EPA can change its mind on policy issues if the Agency develops a reasoned analysis to support the change, See Motor Vehicle Mfrs. Ass'n. v. State Farm Mutual, 463 U.S. 29 (1983), it cannot act contrary to a statutory directive that compels a singular outcome, as the CAA does here. The D.C. Circuit decision in NRDC does not alter these fundamental principles of statutory construction and, indeed, is based on an admonition to EPA to apply the CAA as written. Any action seeking to regulate SSIs under CAA § 129 is contrary to the plain meaning of §§ 112 and 129. C. EPA Has No Authority to Regulate Sewage Sludge Incinerators Under § 129 1. Sewage Sludge is Not a Solid Waste11 The CAA defines solid waste by referencing the definition of solid waste under RCRA: “The term ‘solid waste’ . . . shall have the meaning established by the Administrator pursuant to [RCRA].” CAA § 129(g)(6). RCRA defines “solid waste” as: any garbage, refuse, sludge from a waste treatment plant, water supply treatment plant, or air pollution control facility and other discarded material, including solid, liquid, semisolid, or contained gaseous material resulting from industrial, commercial, mining, and agricultural operations, and from community activities, but does not include solid or dissolved material in domestic sewage . . . . 42 U.S.C. § 6903(27) (emphasis added). This expressly excludes the sewage sludge in domestic sewage from the definition of “solid waste” in what is commonly referred to as the “Domestic Sewage Exclusion” (“DSE”). This concept of a broad POTW exemption was established by Congress as early as 1965 in the Solid Waste Disposal Act.12 There was an early recognition that a comprehensive solid waste program, designed 11 See also NACWA’s August 3, 2010, comments on the proposed definition of solid waste (75 Fed. Reg. 31844) (Jun. 4, 2010), included in Attachment A. 12 Solid Waste Disposal Act, Pub. L. 89-272, 79 Stat. 992 (1965). NACWA Comments on Docket ID No. EPA-HQ-OAR-2009-0559 November 29, 2010 Page 13 primarily to address hazardous wastes, did not apply to POTWs as long as they were effectively regulated under the CWA, which has always been and remains the primary statutory authority for comprehensive regulation of POTW operations. The Solid Waste Disposal Act and RCRA included the DSE in explicit recognition of this critical policy choice.13 Indeed, as EPA began its long efforts to define “solid waste” and “hazardous waste” for purposes of Subtitle C of RCRA, the Agency explicitly understood and discussed the importance of the comprehensive federal sewage sludge management program. For example, in the 1980 preamble to EPA’s development of the Subtitle C regulations, EPA describes the importance, scope and ultimate supremacy of the to-be developed CWA § 405 program, indicating that, once this program was in place, it would serve as the comprehensive regulatory scheme for use and disposal of sewage sludge. See 45 Fed. Reg. 33084, 33102 (May 19, 1980) (“Once such a regulation is in place, sewage sludge will be exempted from coverage under other sets of regulations.”). EPA has similarly interpreted the scope of the DSE to include sewage sludge generated by POTWs in the preamble to the Agency’s 1990 Final Rule to identify and list hazardous wastes for petroleum refinery process wastewaters. EPA concluded that POTW sewage sludge falls within the DSE: These wastes [P038 and K048 wastes] are being added to the list of [hazardous] wastes . . . in order to regulate sludges generated at wastewater treatment facilities on site at petroleum refineries as well as sludges generated at off-site wastewater treatment facilities. It should be noted that if wastewaters generated at petroleum refineries are discharged to a POTW and such wastewaters are mixed with domestic sewage from nonindustrial sources, the sludges generated in the POTW are covered under the domestic sewage exclusion and are not included in today’s listings. It should be noted that if wastewaters generated at petroleum refineries are discharged to a POTW and such wastewaters are mixed with domestic sewage from nonindustrial sources, the sludges generated in the POTW are covered under the domestic sewage exclusion and are not included in today’s listings. 55 Fed. Reg. 46354, 46364 (Nov. 2, 1990) (emphasis added). Thus, there has been a clear recognition for over 30 years that sewage sludge is different than solid waste for regulatory purposes, and that sewage sludge is primarily regulated under the CWA, not RCRA. Furthermore, when Congress incorporated RCRA’s definition of “solid waste” in CAA § 129 in 1990, Congress was well aware that the DSE was encompassed in the definition of “solid waste” and that CAA § 129 would not apply to sewage sludge. This statutory exemption for sewage sludge – the subject of broad consensus and reliance in the regulated community – can not be abrogated by subsequent rule making or preamble statements. Moreover, the 1987 CWA amendments and the subsequent Part 503 rules established a management program for sewage sludge dependent on its exclusion from RCRA regulation. 13 Solid Waste Disposal Act, Pub. L. No. 89-272, § 203(4), 79 Stat. 992, 998 (1965) (defining the term “solid waste” to exclude “solids or dissolved material in domestic sewage or other significant pollutants in water resources . . . .”); accord Resource Conservation and Recovery Act, Pub. L. No. 94-580, § 1004, 90 Stat. 2795, 2801 (1976). NACWA Comments on Docket ID No. EPA-HQ-OAR-2009-0559 November 29, 2010 Page 14 Lastly, the statutory and regulatory provisions that implement the sewage sludge management program distinguish between sewage sludge and solid waste, and thereby demonstrate that they are different types of material. For example, the preamble to the Part 503 rules states that: The standards also do not apply to sewage sludge that is co-incinerated with large amounts of solid waste . . . . However, the standards established in the rule do apply to sewage sludge that is incinerated in a sewage sludge incinerator with incidental amounts of solid waste use as an auxiliary fuel (i.e., 30 percent or less solid waste by weight). 58 Fed. Reg. 9248, 9253. In keeping with the distinctions drawn by Congress between sewage sludge and solid waste, EPA’s careful regulatory approach in the Part 503 regulations distinguishes between sewage sludge and solid waste. 2. SSIs Are Not Solid Waste Incineration units Per § 129(a)(1)(A) and (b)(1), EPA is directed to set emission standards “for each category of solid waste incineration units.” Section 129(b) directs EPA to establish emission guidelines for existing solid waste incineration units. Thus, the definition of solid waste incineration unit serves a gate-keeping function – if the incinerator at issue is a solid waste incineration unit, then it is subject to standards under § 129. However, as EPA has recently asserted, sewage sludge combusted in SSIs is a newly generated solid waste derived from the treatment of domestic and industrial sewage within the POTW, a local government-owned and operated facility, and therefore is not a solid waste material “from commercial and industrial establishments or the general public” as defined under § 129(g)(1). See 75 Fed. Reg. 31844 (June 4, 2010). SSIs cannot be regulated under § 129 because they are combusting a material that is generated by the POTW, which is neither a commercial or industrial establishment nor the general public. Section 129(a)(1)(B)-(C) also directs EPA to set standards for “solid waste incineration units [of specified sizes] combusting municipal waste ... .” But to qualify as a unit combusting “municipal waste” the unit must first be a “solid waste incineration unit,” which does not include SSIs. Furthermore, the term “municipal waste” is defined as: … refuse (and refuse derived fuel) collected from the general public and from residential, commercial, institutional, and industrial sources consisting of paper, wood, yard wastes, food wastes, plastics, leather, rubber, and other combustible materials and non-combustible materials such as metal, glass and rock, provided that: (A) the term does not include industrial process wastes or medical wastes that are segregated from such other wastes; and (B) an incineration unit shall not be considered to be combusting municipal waste for purposes of section 111 or this section if it combusts a fuel feed stream, 30 percent or less of the weight of which is comprised, in aggregate, of municipal waste. CAA § 129(g)(5). Although some of the types of wastes listed above (e.g., wood, yard wastes, rock) enter the sanitary sewer system, most of these types of waste are screened out at the POTW headworks and disposed of in landfills. See Figure 1 (above). The screened wastes and grit are separated to protect the POTW treatment system and are not contained in the solids that are combusted in SSIs. NACWA Comments on Docket ID No. EPA-HQ-OAR-2009-0559 November 29, 2010 Page 15 With all of these legal impediments to proceeding under Section 129, including Congress’ clear statutory intent to regulate SSIs under § 112 and not § 129, EPA should abandon this rulemaking and return to its previous plan to regulate SSIs under § 112 of the CAA, likely as area sources. The technical corrections discussed in the comments below are warranted under any future rulemaking effort. NACWA encourages EPA to take the time now to develop a more thorough and accurate understanding of SSIs. II. EPA’S REGULATORY FLEXIBILITY ACT ANALYSIS IS FLAWED AND THE IMPACT ON SMALL ENTITIES MUST BE REVISITED. A. EPA’s Small Entity Analysis Assumes All Units Will Abandon Incineration EPA is required under the RFA to prepare a regulatory flexibility analysis of this rule unless the agency certifies that the proposed rule will not have a significant economic impact on a substantial number of small entities. See 75 Fed. Reg. at 63292. EPA concluded that none of the 18 small entities14 identified in the database have cost-revenue ratios greater than one percent because they would all find it cost effective to divert sewage sludge to landfills. Based on this conclusion, the Administrator certified that the Proposed Rule resulted in no significant impact for a significant number of small entities (“No SISNOSE”). In the Regulatory Impacts Analysis (“RIA”), EPA nonetheless identified the following small entity impacts if they were required to install one of the three control technology options under consideration for the final MACT standard: o Option 1 (MACT floor) ƒ 9 of 18 small entities affected at greater than 1 percent cost-revenue ratio ƒ 2 of 18 affected at greater than 3 percent cost-revenue ratio o Option 2 (MACT floor + afterburner for MH) ƒ 13 of 18 small entities affected at greater than 1 percent ƒ 2 of 18 affected at greater than 3 percent o Option 3 (Option 2 + fabric filter for MH) ƒ 16 of 18 small entities affected at greater than 1 percent ƒ 3 of 18 affected at greater than 3 percent See RIA Tables 4-5, 4-6, and 4-7. EPA’s Option 3 with its beyond the floor mercury controls will significantly affect 16 of 18 small entities at greater than 1 percent of their cost-revenue ratios and three of those at greater than 3 percent. These levels of impact would require a more thorough consideration of small entity impacts based on EPA’s Final Guidance for EPA Rulewriters: Regulatory Flexibility Act as Amended by the Small Business and Regulatory Enforcement Fairness Act (November 2006) (disallowing a presumption of No SISNOSE at these levels of impact). Only by assuming the cost-effective landfill alternative for small entities could the Administrator presume No SISNOSE under this guidance and shortcut the administrative protections that the RFA and Small 14 NACWA invites EPA to reassess the number of small entities operating SSIs when it confirms the number of active SSIs. Regional sewer districts may serve across municipal boundaries making the traditional 50,000 population criterion for small entities difficult to apply. In the absence of an ICR survey, we understand that the small entity determinations were made based on internet searches of population data for the closest related municipal entity. This is not a reliable methodology. For the purpose of this discussion, we rely on EPA’s numbers. NACWA Comments on Docket ID No. EPA-HQ-OAR-2009-0559 November 29, 2010 Page 16 Business Regulatory Fairness Act (“SBREFA”) provide for small governments. EPA’s rushed analysis, however, has a significant flaw as discussed below. EPA also relied on this flawed small entity analysis as a central component of its UMRA cost-benefits assessment. See 75 Fed. Reg. at 63293 (claiming no Section 203 UMRA obligation because “EPA’s analysis shows that for the more likely scenario that small governmental entities switch to landfilling, none of the ratios was greater than 1 percent.”) Due to the central flaw discussed below, EPA must also revisit its UMRA analysis. B. EPA Misunderstood that Sewage Sludge is Wet When Fed to an Incinerator EPA’s rush to propose this rule provided insufficient time for the Agency to gain the fundamental understanding of sewage sludge necessary to this rulemaking. EPA presumes incorrectly that the material fed into an SSI is dry and does not contain moisture. In fact, sewage sludge is typically only 20-30 percent solids and 70-80 percent moisture. While SSIs are rated based on the number of dry tons per day they can combust, the sewage sludge being fed into the incinerator is not dry. When sewage enters the POTW headworks it is over 99 percent water. POTW processes increase the solids content to about 4 percent before it is “dewatered.” The dewatering process increases solids content to 20-30 percent before it is fed into the incinerator. The incinerator drives off the rest of the moisture and it combusts the volatile solids. Thus, an SSI rated at 100 dry tons per day will feed 400-500 tons per day of dewatered sewage sludge into the incinerator to combust that 100 dry tons of sewage sludge. EPA incorrectly presumed that the 100 dry ton/day SSI would feed and combust 100 tons of sewage sludge per day. C. EPA’s Error Undercuts its Small Entities Analysis EPA’s fundamental misconception about the nature of sewage sludge affects many aspects of the Proposed Rule. For instance, EPA’s cost analysis for the landfill alternative is based on dry tons of sludge, which underestimates the amount of sludge being sent to a landfill by a factor of three to five. The landfill alternative requires as many as five times more truck loads,15 five times more landfill tipping fees, and five times as much on-site storage and loading capacity. EPA assumed that onsite storage capacity would require a cement pad with a railing, instead of the more costly tankage necessary to contain sludge that is 70-80 percent water. EPA also failed to consider the cost and limitations associated with landfills rejecting wet sludge due to capacity restrictions and moisture limitations. POTWs will have to transport sludge farther in search of landfill capacity willing and able to take wet sewage sludge. NACWA is confident that when EPA corrects its cost analysis for small entities, the landfill alternative will not be cost-effective for many if not all of the small entities. As an example, the City of Edmonds, a small government entity located in Washington State, has reported to NACWA that it has no intention of trucking its sewage sludge to a landfill, the closest of which is 270 road miles away. EPA should abandon the landfill alternative presumption that it used to shortcut small government relief under RFA and UMRA. Instead, EPA should consider the full cost of its proposed control technology on small entities and engage in the appropriate RFA/UMRA processes to evaluate ways to mitigate the burden of this rule on these small entities. 15 EPA also overestimates the amount of sludge a single truck can hold. As an example, EPA estimated that 34 tons of sewage sludge can be hauled in each truck, when in reality only 15-20 tons can be hauled per truck based on the 80,000 pound total truck weight limit for roadways in Ohio. This alone doubles the cost that EPA assumed to be associated with trucking sewage sludge to landfill. NACWA Comments on Docket ID No. EPA-HQ-OAR-2009-0559 November 29, 2010 Page 17 III. EPA’S ERROR ALSO UNDERCOUNTS THE ENVIRONMENTAL BENEFITS OF INCINERATION COMPARED WITH OTHER SLUDGE DISPOSAL ALTERNATIVES When EPA develops a better understanding of the amount of sewage sludge to be handled, the Agency will also gain an appreciation for the environmental benefits that incineration can offer compared with other residuals management options. At a certain driving distance, the air emissions associated with trucking sewage sludge to landfill exceed the air emissions from incineration. Additional environmental benefits accrue when incineration is used to generate steam and/or electricity that offsets the need to burn fossil fuels. Incineration also avoids the generation of methane, a potent greenhouse gas, produced when sewage sludge is placed in a landfill and it begins to biodegrade. Incineration also reduces odors, which is generally the environmental issue of greatest interest to those living near POTWs. When all of these environmental attributes are evaluated, incineration can be the most environmentally beneficial option for residuals management. The NEORSD evaluated residuals management options based on both cost and environmental benefits when considering whether to invest in new fluidized bed incinerators. In the final analysis, incineration was significantly “greener” than landfilling for managing sludge from its Southerly POTW. The NEORSD analysis estimated 11,300 to 16,200 truckloads of sewage sludge per year, making a 130-mile round trip to the closest current landfill. Criteria pollutants from the diesel trucks were comparable to the permitted emission levels from the incinerators. Significantly, the organic compounds from the diesel exhaust were over three times higher than the SSI emissions. See NEORSD Comments – Section D – on the Proposed Rule. NEORSD concluded that it would emit more air pollution by sending its sewage sludge to landfill 65 miles away than it would operating onsite incinerators. The air emissions benefit from incineration would be even greater for the POTWs that face a greater distance to landfill (e.g., The Metropolitan District of Hartford, Connecticut (375 miles to landfill); Edmonds, Washington (270 miles to landfill)). The comparison is even more favorable for incineration when considering the electricity generation component of the NEORSD incinerator project. The new fluidized bed incinerators will use excess heat to generate steam, which will be sent to steam turbines to generate electricity for internal use. The electricity will offset approximately 25 percent of the current NEORSD demand for electricity each year, which is equivalent to the electricity needed to power 1800 homes. In Ohio, where most of the electricity on the grid is generated by coal combustion, this demand reduction reduces the mercury emissions and other air contaminants associated with utility coal combustion. Electricity generation from sewage sludge is not an isolated phenomenon. Ohio, like many states, recognizes the generation of electricity from sewage sludge as eligible for renewable energy credit toward the state’s renewable portfolio standard. Utilities required to demonstrate that they generate the required portion of their electricity using renewable or alternate energy means will purchase the renewable energy credits generated from the SSI-electricity project. This provides additional economic incentives to invest in the electric-generating equipment for SSIs. Kern County, California recently announced its intent to install sewage sludge incinerators with 13.5 megawatts of electric-generating capacity to replace land application. This is a trend that is likely to continue because SSIs present a reliable renewable fuel source for baseload electricity generation that utilities will help fund to secure renewable energy credits to meet renewable portfolio standards. NACWA Comments on Docket ID No. EPA-HQ-OAR-2009-0559 November 29, 2010 Page 18 EPA cannot accurately assess the relative benefits of incineration without understanding the true costs of landfilling. The Metropolitan St. Louis Sewer District notes that EPA’s estimate of its additional cost to landfill without incineration is $3.8 million per year, when the District’s actual landfill option cost estimate is over $11 million per year. NEORSD similarly estimates that its annual cost to landfill is $9 million more than incineration. This does not include the additional cost of a loading and storage facility that can accommodate up to 12 trucks per hour that NEORSD estimates will cost over $50 million. These costs are difficult to justify when they do not reduce emissions and they eliminate the opportunity to use sewage sludge as a viable alternative energy source for electricity and steam generation. EPA’s miscalculation of the amount of sewage sludge to be landfilled is not just a cost issue; it goes to whether EPA properly evaluated the feasibility of the landfill alternative. The Metropolitan St. Louis Sewer District would need to dispose of an additional 600 tons per day of sewage sludge that is 75 percent water after pressing/dewatering. This would fill the current landfill owned by the District well before it could locate and permit another landfill. NEORSD has been advised that its closest landfill (65 miles away) will not be able to handle all of its sludge. Furthermore, NEORSD has not been able to identify a landfill that will accept the necessary volume or type (wet) of biosolids. EPA has also failed to consider the feasibility of adding the necessary on-site storage, truck loading operations, and odor control measures in local jurisdictions. Landfilling and land application alternatives are just not feasible for many POTWs. EPA needs to revisit every aspect of its Proposed Rule that relied on a landfill alternative and correct its assessment of the feasibility and cost associated with the landfill option. IV. EPA LACKS THE DATA NECESSARY TO EVALUATE BEYOND THE FLOOR MACT CONTROLS FOR MERCURY In order to justify the proposed beyond the floor controls for mercury, EPA must make a proper cost effectiveness determination. This requires accurate data for the baseline level of mercury being emitted today, accurate information on the potential for mercury emission control, and accurate information regarding the cost of that control. EPA’s data is lacking in each of these respects. A. EPA’s Mercury Assumptions Are Inconsistent with Available Data and Past Reports EPA starts with the inaccurate presumption that baseline emissions from the 21816 existing SSIs in the U.S. include 3.1 tons of mercury emissions. This is inconsistent with EPA’s 1997 Mercury Study Report to Congress in which it stated that 0.95 tons of mercury was emitted from sewage sludge incinerators in 1994. More recently, the Water Environment Research Foundation (“WERF”) released a 2009 report entitled Minimizing Mercury Emissions in Biosolids Incinerators (“2009 WERF Report”) (Attachment B). This report used the weekly sewage sludge samples required under Part 503 to calculate the amount of mercury in the sewage sludge being incinerated in the U.S. in a baseline year. The report conservatively assumed that all of the mercury in the sewage sludge was emitted despite the potential for some forms of mercury to be removed by existing particulate control systems. WERF calculated the average uncontrolled mercury concentration in the feed to SSIs and conservatively rounded up to 1 mg/dry kg. WERF concluded that SSIs collectively emit less than 1 ton of mercury to the atmosphere each year. This is consistent with EPA’s 1997 report to Congress. 16 Again, NACWA asks that EPA take the time to confirm the number of incinerators and that it include dormant incinerators with active operating permits in the total number of SSIs considered for this rulemaking. NACWA Comments on Docket ID No. EPA-HQ-OAR-2009-0559 November 29, 2010 Page 19 The actual amount of mercury in sewage sludge has been going down consistently over time. This is supported by test data from POTWs, including the data from the Upper Blackstone Water Pollution Abatement District (Millbury, Massachusetts) that reveals a significant reduction in the mercury content of its sewage sludge since 1997. See Table 1. Table 1 1997 2001 2010 Mercury content in Upper Blackstone Sewage Sludge 8.9 mg/ dry kg 1.1 mg/dry kg 0.84 mg/dry kg The amount of mercury in sewage sludge is decreasing as communities implement efforts to control sources discharging mercury into the sewage system. Dental offices are one of the most significant sources of mercury to sewer systems. The Metropolitan Council Environmental Services (St. Paul, MN) implemented a dental amalgam separator program that achieved a 50 percent reduction in the mercury content of their biosolids. Other dental amalgam separator programs implemented by the City of Palo Alto Regional Water Quality Control Plant and the Central Contra Costa Sanitary District similarly achieved a 50 percent reduction in mercury flowing to their treatment plants. The total annualized cost per dental site has been estimated to be just $717 per year in Palo Alto, California. See T. Barron and K. North, “Cost of Hg Diversion Through Dental Office Source Control” (November 16, 2010) (Attachment C). This pollution prevention program reduces mercury at an estimated cost of $9,000 per pound, which is far less than NACWA’s estimates that range as high as $190,000 per pound for end-of-pipe mercury control systems added to existing SSIs. These cost effective options for mercury control upstream of the POTW have sustained a declining trend in the content of mercury in sewage sludge. Declining mercury content in sewage sludge is expected to continue. On September 27, 2010, EPA announced that it intends to propose a national rule to reduce the 3.7 tons of mercury waste discharged from dental offices each year by requiring the installation of dental amalgam separators in dental offices. The American Dental Association has been encouraging its members to install dental amalgam separators for some time and has signaled its support for this rulemaking. With no significant opposition, EPA would be expected to finalize a rule in time to significantly reduce mercury in POTW systems before the proposed SSI rules would become effective. Dental amalgams are generally believed to be at least 50 percent of the mercury loadings to POTWs, which means that the mercury in sewage sludge is likely to be cut in half from 2010 baseline amounts. NACWA supports these efforts to remove mercury from the POTW system as a pollution prevention initiative that is far more efficient and cost effective than SSI mercury controls. B. EPA’s Extrapolation from Limited Data Significantly Overestimates Baseline Emissions of Mercury In EPA’s rush to propose the SSI rule, it did not take the time to benchmark its mercury estimate against these other sources of data. The result is an estimate of mercury that is at least three times higher than other more credible estimates. EPA’s baseline 3.1 tons/year mercury estimate relies on the stack test data from the nine units that were the target of the Agency’s request for information. EPA generates a mercury emission factor from this data that it applies to all SSIs in the database. These data are snapshots taken over a few hours NACWA Comments on Docket ID No. EPA-HQ-OAR-2009-0559 November 29, 2010 Page 20 on one day. EPA does not know if these data are representative of normal operating conditions or whether the metals content in the sludge was at the low end or the high end of a typical range. See Proposed Rule, 75 Fed. Reg. at 63268 (requesting comment on whether the metals content in the sludge during the test was representative of low or high points in the typical mercury range). This range can vary significantly with a ten- fold difference between the minimum and maximum mercury values. See Attachment A of NEORSD Comments on Proposed Rule. Thus, EPA may have miscalculated the mercury content of sewage sludge because its data are not representative of typical mercury values. EPA also makes significant incorrect assumptions about operating hours and sewage sludge feed rates that impact its mercury estimate. Rather than take the time to collect actual operating data through a formal Information Collection Request (“ICR”), EPA assumes that each incinerator operates at 75 percent of its rated capacity. POTWs and their incinerators are designed to accommodate significant wet weather events that can significantly increase the flow to the headworks of the POTW with minimal notice. As such, the typical operating conditions for the POTW and its incinerators are significantly below the design capacity. See Table 2. Table 2 NEORSD (3 plants) Township of Wayne Anchorage HRSD Boat Harbor HRSD Army Base EPA Est. 2009 Actual EPA Est. 2009 Actual EPA Est. 2009 Actual EPA Est. 2009 Actual EPA Est. 2009 Actual Sludge incinerated (dry tons per year) 71,000 37,000 4,340 1,560 16,968 6,100 12,306 8,132 8,190 3,337 Hours of operation per year 33,600 27,000 5,430 2,590 8,400 8,200 8,400 8,135 8,400 7,912 These examples indicate that EPA has overstated the amount of dry tons of sludge incinerated per year by a factor of two or more. Hours of operation are also overstated in EPA’s analysis but by a lesser amount. EPA must take the time needed to collect the feed rate and operating hours data from all SSIs before it uses this information to calculate the baseline mercury emissions from SSIs. An accurate assessment of baseline mercury content is critical to a proper beyond the floor analysis of the cost effectiveness of additional mercury controls. The anecdotal information available to NACWA indicates that EPA’s calculation significantly overestimates mercury emissions. NEORSD compared EPA’s mercury estimate for two of its facilities with its own calculation based on the actual sludge samples taken in 2009 to demonstrate compliance with Part 503 requirements. The NEORSD calculation reveals that EPA’s mercury estimate for its Southerly and Westerly plants combined was nearly three times higher than the actual concentration of mercury in the sludge in 2009 based on the Part 503 sampling. Wayne County, New Jersey performed a similar analysis using stack test data NACWA Comments on Docket ID No. EPA-HQ-OAR-2009-0559 November 29, 2010 Page 21 and found that EPA’s mercury estimate was nearly three times the estimate based on site-specific stack test data. Other NACWA members have reached similar conclusions, calling into serious question the reliability of EPA’s baseline mercury emissions estimates. Given these significant errors in the baseline mercury estimate, EPA must revisit its decision that beyond-the-floor mercury controls are cost effective. Cost effectiveness can only be reliably assessed after an accurate baseline is established. EPA assumed at least three times more mercury is being emitted from sewage sludge incineration than has been measured in the sewage sludge. This cannot be right. Moreover, with local and national efforts underway to reduce mercury significantly from dentist offices across the country, the mercury in sewage sludge will be even further below EPA’s estimate by the time a rule is promulgated and effective. The actual mercury in sewage sludge after upstream pollution prevention efforts does not justify the significant cost that this Proposed Rule would impose on public entities. C. EPA Does Not Have Sufficient Data on the Effectiveness of Mercury Control EPA’s mercury control assumptions are also flawed. EPA assumes incorrectly that activated carbon injection (“ACI”) can work without significant changes to the existing particulate control devices. This theoretical assumption does not work in practice for POTWs. Most of the existing SSIs use wet scrubber systems for particulate control. The exhaust temperature before the scrubber is typically 1200-1600 degrees Fahrenheit, which is too hot for injecting activated carbon or for adsorption to occur. Also, a contact chamber is necessary to provide adequate residence time for the mercury to adsorb onto the activated carbon. Some POTWs lack the space for these exhaust train changes. Even if the ductwork could be reconfigured to allow the exhaust temperature to cool and to accommodate a contact chamber, the wet scrubber particulate control device will not consistently remove mercury. This is because the scrubber water is typically sent back to the headworks of the POTW for cleaning, thereby recycling any of the mercury that is captured by ACI back through the process that generates sewage sludge. This increases the likelihood of elevated mercury in the POTW effluent and it concentrates mercury in the sludge. EPA has not performed a life cycle analysis to evaluate the potential for the mercury to be released from the activated carbon in a form that is more difficult to keep out of the environment. Furthermore, when EPA corrects the flawed baseline analysis for mercury, the amount of mercury in the exhaust gas will be significantly reduced. At lower levels of mercury, ACI becomes less efficient at adsorbing and removing mercury resulting in mercury reductions significantly less than EPA’s estimated 85-95 percent control efficiency. By contrast, other options for mercury control, including upstream pollution prevention measures at dentist offices and elsewhere, are expected to be far more effective at reducing mercury emissions from SSIs at far less cost per pound of mercury removed. ACI is not a proven technology on MHIs. NACWA has been unable to locate any MH units utilizing the ACI technology advocated by EPA in the Propose Rule. The only active mercury controls known to NACWA are installed on FBIs: 1. Metropolitan Council Environmental Services (St. Paul, MN) installed an ACI system that consists of a carbon contact chamber followed by a fabric filer. Heat exchangers and a boiler upstream of the carbon injection point lowers the exhaust gas temperature to 350 degrees Fahrenheit, and NACWA Comments on Docket ID No. EPA-HQ-OAR-2009-0559 November 29, 2010 Page 22 2. Ypsilanti, Michigan uses a carbon adsorption system. Both of these facilities reported serious abrasion and corrosion problems shortly after installation. Unless and until this technology is proven effective and reliable on both FBIs and MHIs, EPA should not consider it as a feasible beyond-the-floor control option. D. EPA Significantly Underestimated the Cost of Mercury Control Equipment EPA’s cost estimate for the beyond-the-floor mercury control is unreasonably low. EPA determined that the MACT floor mercury limit would require just two of the 163 multiple hearth incinerators to install mercury controls. EPA proposes beyond-the-floor mercury controls for the other 161 MHIs. EPA estimates that these 161 units could install mercury controls for an additional capital cost of $5,000,000 ($31,056 per incinerator) and an additional annualized cost of $32 million ($198,758 per incinerator). See 75 Fed. Reg. at 63276 (Table 8). These costs are wildly inaccurate. NACWA members estimate actual capital costs for multiple hearth units will exceed $4 million per incinerator. First, EPA’s cost data assumes that activated carbon can be injected upstream of existing particulate control devices, and does not account for the installation of new fabric filters for each ACI installed. As explained above, ACI systems will require baghouses to collect the carbon injected by the ACI system and heat exchangers and/or boilers to reduce the exhaust gas temperature below 400 degrees Fahrenheit. However, no existing MHIs are equipped with baghouses for particulate control. Instead, all existing MHI units employ wet scrubbers. Wet scrubbers can become clogged by the injected carbon, and this carbon can also be released back into the POTW system through the recycling of the scrubber water to the POTW headworks. Any units equipped with ACIs will therefore require the addition of a baghouse, a fan to pull the exhaust through the fabric filters and a bag leak detection monitor. EPA does not account for these capital costs in its beyond-the- floor cost effectiveness analysis for mercury. EPA must also consider the annual operating costs of a baghouse. This includes the significant cost of electricity to operate fans large enough to pull the exhaust through the fabric filters. EPA should also consider the mercury and other air emissions associated with the generation of this electricity when it determines whether this beyond the floor option yields a net environmental benefit. EPA must also consider the cost of replacement bags clogged by activated carbon. EPA also fails to account for other costs associated with ACI mercury controls. The Metropolitan Council Environmental Services (St. Paul, MN) discovered during the design of its incinerators that exhaust gas temperatures leaving the incinerator are so high that carbon injected will combust before it has a chance to adsorb mercury. The Metropolitan Council Environmental Services, as previously reported, installed additional ductwork, along with heat exchangers and boilers to cool the exhaust gas before the activated carbon is injected. They also installed a contact chamber to ensure sufficient adsorption, followed by a fabric filter. These are complex systems that require significant engineering costs for site-specific design and installation oversight. EPA does not account for any of these additional costs in its assumed mercury control cost. NACWA members estimate $3.5 - 4 million for the minimum equipment necessary: a carbon contact chamber and a baghouse for each incinerator. Total capital costs for the 161 MH incinerators would be in the range of $564 - NACWA Comments on Docket ID No. EPA-HQ-OAR-2009-0559 November 29, 2010 Page 23 644 million. However, the costs to add the required heat exchangers and the boilers, reconfigure the existing air pollution control train and expand the incinerator building could easily raise the price to $8 - 10 million/incinerator or higher. This would result in total capital costs of $1.3 - 1.6 billion or higher. Applying these mercury control costs to the actual site-specific mercury content (instead of EPA’s inflated estimate), reveals extremely high cost effectiveness ratios, many times greater than EPA’s $6,000 per pound of mercury control estimate. See ERG, Analysis of Beyond the Maximum Achievable Control Technology Floor Controls for Existing SSIs. NEORSD’s mercury control cost analysis concluded that its incinerators would need a carbon contact chamber and fabric filter, as well as a heat exchanger and boiler, all of which would require engineering design services and construction management. The average mercury removal cost for NEORSD would be more than $100,000 per pound of mercury removed.17 The City of Palo Alto Regional Water Quality Control Plant conducted a similar cost analysis that put the national cost of mercury removal at nearly $190,000 per pound, and over $400,000 per pound for the City based on its actual emissions level. See Comments filed by the City of Palo Alto on the Proposed Rule. These cost figures are not reasonable when demonstrated pollution prevention options are available at $9,000 per pound for dental amalgam separators. Carbon adsorption for mercury control is also a costly option. For example, if POTWs were to procure and install activated carbon adsorbors the cost would be at least $3.5 - 4 million per incinerator just for the equipment, for a total capital cost in of $564 - $644 million for the 161 MH incinerators evaluated in the beyond-the-floor analysis. The cost to expand existing incineration facilities, add new ductwork and controls, and engineering related design and construction management costs could easily raise the cost of these units to $5 - $6 million. The resulting total cost would be $800 million - $1 billion or higher for the 161 MH incinerators. It should also be noted that no one in the U.S. manufacturers carbon adsorbers and they would have to be imported from overseas. Ypsilanti, Michigan’s experience suggests that this technology may not be feasible, on a long-term basis for SSIs due to the significant corrosion problems that they have experienced. At a minimum, carbon adsorption polishing must be placed after the particulate control device. For SSIs, the existing particulate control is a wet scrubber, which generates significant amounts of steam and moisture in its exhaust that are incompatible with carbon adsorption. Therefore, the exhaust must be heated high enough so that moisture does not condense as it moves through the carbon adsorption system. Edmonds, Washington reports that it faces a capital cost of $1,500,000 for the exhaust heater and carbon adsorption system to control a portion of 1.5 lbs of mercury measured in the sewage sludge per year. With a useful life, based on Ypsilanti’s experience, of less than 10 years, the annualized cost per pound for capital alone is over $100,000 per pound of mercury removed. Edmonds is a small entity that does not have a sufficient footprint for the installation of this system. EPA cannot justify carbon adsorption as a beyond the floor mercury control. EPA is required to consider energy impacts as part of its beyond the floor analysis. EPA should consider the projects currently underway that will generate electricity from the heat produced by SSIs. Every kilowatt hour of electricity generated by the SSI reduces the POTW demand for electricity from the grid. NEORSD plans to meet 25 percent of its electricity demand through SSI-generated electricity by 2014. In Ohio, NEORSD’s reduced demand for electricity from the grid means less coal combustion and less of the emissions that coal would have generated, including less mercury emissions and less reliance on fossil fuels. Beyond the floor 17 NEORSD’s detailed cost analysis is included as Attachment D. NACWA Comments on Docket ID No. EPA-HQ-OAR-2009-0559 November 29, 2010 Page 24 controls could discourage this advanced energy option, resulting in more coal combustion and more mercury emissions. EPA’s rush to develop the Proposed Rule has resulted in an improper baseline emissions estimate for SSI mercury, inadequate assessments of the feasibility of mercury control as applied to SSIs, and a significant underestimate of the cost of mercury controls for this sector. EPA is compelled to explore alternative, less costly ways to decrease mercury emissions before imposing beyond the floor controls. Therefore, EPA’s beyond the floor mercury limits are not justified by the record and should be removed from the final rule. V. EPA’S MACT FLOOR METHODOLOGY IS FLAWED In setting emission standards under § 129, EPA is obligated to start with an analysis of the performance of existing sources. From that analysis, EPA is required to determine the “MACT floors,” which for new units is “the level of control that is achieved in practice by the best controlled similar unit” and for existing units is “the average emissions limitation achieved by the best performing 12 percent of units in the category… .” CAA § 129(a)(2). EPA can set standards more stringent than the MACT floor only after considering cost, energy and other factors. EPA’s assessment of existing source performance was insufficient to set accurate MACT floors. A. EPA Does Not Have Enough Valid Data to Establish Lawful and Proper MACT Floors EPA’s rush to regulate SSIs has left the Agency with inadequate data to assess existing performance among SSIs. EPA chose to limit its ICR to just nine entities because collecting information from ten or more entities would have triggered the PRA obligations and a more rigorous OMB review. See 44 U.S.C. 3502(3)(a)(i) (defining the threshold term “collection of information” as a request “imposed on ten or more persons” other than federal agencies.) EPA’s plan to circumvent the PRA and OMB review resulted in an inadequate dataset for this rulemaking that leaves EPA unable to reliably take the first necessary step in a § 129 rulemaking: to determine which of the SSIs are the best performing sources. EPA has built a statistical house of cards on this flawed foundation, which undermines the legality and scientific credibility of the emission standards set forth in the Proposed Rule. As a consequence of EPA’s limited data collection effort, the dataset available for this rulemaking contains actual data from less than 12 percent of the population of sources in each subcategory. Even if all of the data collected could somehow be attributed to top performers, EPA would be unable to determine based on actual data the average performance of the top 12 percent as is required for the existing source MACT floor. See ERG Memo, MACT Floor Analysis for the Sewage Sludge Incinerator Source Category at 6 (June 2010) (“MACT Floor Memo”). In fact, EPA is using actual data from as little as 4.3 percent of a subcategory (7 of 163 MH units for HCl) to determine how the top 12 percent perform. The way that EPA targeted its data collection undermines its ability to use statistics to fill data gaps. EPA targeted its ICR to the nine POTWs “expected to have the lowest emissions based on the type of unit and the installed air pollution controls.” See MACT Floor Memo at p. 6. EPA then uses statistical methods that predict the distribution of all data based on a sample set. This statistical method relies on random sampling of representative data. See Attachment F of the NEORSD Comment on Proposed Rule. EPA’s targeted approach to collecting data from expected top performers undermines its ability to presume the data is a random sample NACWA Comments on Docket ID No. EPA-HQ-OAR-2009-0559 November 29, 2010 Page 25 representative of the entire source category or subcategory. If the data gathered is not representative at the outset, then the data cannot reliably be used in a statistical equation to predict the emissions data across the source category or subcategory. Thus, EPA improperly limited the data it considered and therefore cannot accurately determine the range of performance for the category or subcategory to determine the top 12 percent of the existing source subcategories. NACWA members have determined a number of errors in the data used by EPA in this rulemaking. Specific references to these errors will be in the comments of those members from whom the data were collected. The errors are an indication of inadequate quality assurance and quality control on the data in the database. EPA should not rely on this public comment process for its check on the quality of the data it relies upon in this rulemaking. EPA should take the time to collect sufficient data so that outliers are readily apparent and properly excluded. EPA should also subject the data to a rigorous quality assessment to establish their validity before relying on them to create enforceable limits. B. EPA’s Data are not Sufficiently Representative to Establish Lawful and Proper MACT Floors To make up for its lack of actual data, EPA tries to extrapolate from the data it has by claiming that it is representative of the source category. This approach fails because the data are not representative of the regional, seasonal and day-to-day variability of sewage sludge. EPA inexplicably characterizes sewage sludge as a “homogeneous” material (see MACT Floor Memo at 6), in an apparent attempt to support its extrapolations. EPA’s only claim of support for this characterization is that sewage sludge concentrations are capped by the CWA regulations under Part 503. A closer look at Part 503 data confirms the common sense conclusion that domestic sewage is an unpredictable and highly variable source that generates heterogeneous sewage sludge that cannot be characterized by EPA’s stack test database composed of snap shot measurements over 3-4 hours on a single day during the winter. Unlike other types of industrial and commercial incinerators that EPA regulates under CAA § 129, POTWs – and their SSIs – are statutorily obligated to manage all of the sewage that enters into the sanitary sewer system. Many different entities and individuals have virtually unlimited access to the POTW through thousands of toilets, sinks and drains throughout the system that flow into the treatment works. The screening system at the headworks of the POTW filters out wood and stones and other large debris, but dissolved material and small particles enter the treatment works and ultimately end up in the sewage sludge. This results in highly variable and often unpredictable spikes in concentrations. POTWs are designed to clean and protect water resources, so treatment focuses on removing metals and other compounds from the wastewater and capturing them in the sludge. The result is a highly variable and heterogeneous waste stream. POTW inlet concentrations also vary based on the nature and type of dischargers. POTWs treat wastewater from residential, commercial and industrial dischargers in varying degrees. POTWs that are dominated by residential customers will have a different range of sludge constituents than those with significant commercial and industrial dischargers. Some POTWs have significant industrial dischargers that contribute constituents in their wastewater that result in sludge concentrations unlike other POTWs. Pretreatment opportunities also vary because POTW authority to control discharges into the sewer system is limited and the way that authority is exercised varies. Finally, the nature of sewage entering the POTW changes over time as the character of a community changes, the age of the population changes, and commercial and NACWA Comments on Docket ID No. EPA-HQ-OAR-2009-0559 November 29, 2010 Page 26 industrial dischargers come and go. Sewage sludge constituents and concentrations vary because the discharge to each POTW is unique. Without the use of long-term data to support the level of emission standards, this variability makes numeric technology-based limits impractical and infeasible and should provide EPA strong motivation to look to other regulatory options. POTWs also face significant regional and seasonal variability that is not captured by EPA’s dataset. Initial high flow periods in the spring often scour the sewers and dislodge heavier material that has settled in the sewer system during low-flow periods, which often results in a spike in metals concentrations (e.g., mercury, cadmium, lead) in the sewage sludge. The ICR stack tests in January and February that were used for the EPA database would not have captured these events. In addition, northern cities that use salt for de-icing roadways experience significant increases in chlorides during the winter months. High chloride concentrations are known to improve the effectiveness of mercury control at existing wet scrubbers. See 2009 WERF Report at 2-4 (Attachment B). Stack test results during January and February in these communities may not be replicable during other times of year. EPA has the data it needs to examine sludge variability for metals. Most POTWs have been required to collect weekly sludge samples and test them for the Part 503 metals. In the Table 3 NEORSD has compiled the monthly average sewage sludge content of cadmium, lead, and mercury at two of its facilities. Even when averaged over an entire month, the range of average metal concentrations varies by a factor of three or more. Table 3: 2009 Monthly Average Sewage Sludge Concentrations (mg/dry kg) Cadmium Lead Mercury Southerly Westerly Southerly Westerly Southerly Westerly January 7.95 13.25 62.23 97.00 0.68 0.44 February 8.45 11.95 75.60 123.10 0.77 0.41 March 8.90 14.46 84.94 141.20 0.69 0.34 April 9.11 14.96 67.93 123.00 0.84 0.37 May 10.40 12.97 87.16 145.95 0.78 0.36 June 4.60 7.94 99.54 204.30 1.24 0.77 July 4.46 7.92 123.15 218.55 1.24 0.92 August 4.78 9.08 114.10 188.23 3.12 0.59 September 5.13 8.12 103.41 183.93 2.08 0.87 October 3.99 7.19 94.33 129.27 0.98 0.63 November 4.14 6.22 82.78 104.81 0.91 0.58 December 3.80 5.97 74.66 97.55 1.38 0.40 Data collected in January and February from these sources could not be considered representative of the sewage sludge during other months or seasons. NACWA has every reason to believe that other POTWs face variability at this level or greater and that this degree of variability applies to other constituents as well. EPA should consider the Part 503 data that POTWs submit to the Agency regularly to account for variability when setting emission limits. In the Proposed Rule, EPA bases its MACT floor analysis solely on a limited set of emission test data. However, nothing in the CAA restricts EPA to emission test data when determining MACT emission rates and NACWA Comments on Docket ID No. EPA-HQ-OAR-2009-0559 November 29, 2010 Page 27 intra-source variability. As EPA recognized in its MACT floor analysis, emission tests provide only a snapshot of emissions at a particular point in time. It is therefore appropriate for EPA to consider non-emission test factors, such as metals content in sewage sludge, in evaluating variability and setting MACT floors. EPA has requested additional sewage sludge metals content information, but has restricted its request to metals content information collected during stack tests. See 75 Fed. Reg. at 63269. It is unclear why EPA would deem only metals content during stack tests relevant to variability. POTWs subject to Part 503 are required to track metals data in their sewage sludge year-round, during all periods of operation. See, e.g., Weekly Lead (Pb) and Mercury (Hg) Concentrations from POTWs (EPA-HQ-OAR-2009-0059-0036) (providing lead and mercury variability from 2005 to 2009 at POTWs). True variability can be established by using all metals content data and applying the site-specific control efficiencies determined under Part 503 testing or additional air emissions testing. C. Part 503 Limits Do Not Render Sewage Sludge Homogeneous EPA claims that sewage sludge is homogeneous because the Part 503 regulations cap sludge concentrations. This ignores certain realities associated with the health-based standards established under Part 503. First, the Part 503 regulations do not address all of the § 129 pollutants (PM, opacity, SO2, HCl, NOx and dioxin/furans). Part 503 does not regulate the variability of sulfur, chlorine and nitrogen concentrations in sewage sludge, which have a direct effect on the amount of SO2, HCl, and NOx generated when the sludge is incinerated. Second, Part 503 establishes risk-based limits for lead and cadmium that are different for every POTW based on feed rates, stack heights, and exhaust flow rates. Within these limits, POTW performance varies significantly depending upon the sources of these compounds in the sewage system, including industrial sources, soil content flushed during rainfall events, or residual material within the sewer system that is cleaned out periodically. While the Part 503 regulations have focused attention on reducing lead, cadmium and mercury compounds in sewage sludge over the years, EPA is wrong to claim that they have eliminated the variability of these compounds in sewage sludge. In fact, actual performance for all POTWs is far below the health-based limits established under Part 503. As such, the range of actual concentrations can vary by a factor of 100 or more without approaching the Part 503 limit. For example, compilations of weekly mercury data from POTWs demonstrate that concentrations of mercury can range from less than 0.1 mg/kg to as high as 17.25 mg/kg. See Weekly Hg and Pb Variability at POTWs, Docket ID No. EPA-HQ-OAR-2009-0059-0036. As a result, mercury concentrations can vary by two orders of magnitude without ever exceeding the NESHAPs limit. In the face of significant data supporting the heterogeneous nature of sewage sludge, EPA’s reliance on Part 503 standards as evidence of homogeneity is seriously misplaced. D. EPA’s Misunderstanding of the Variability of Sewage Sludge Leads to Unachievable New Source Limits EPA incorrectly presumes that stack test results account for the full variability of an SSI’s performance. As indicated above, that is based on the erroneous conclusion that sewage sludge is a homogeneous waste with insignificant variability of relevant constituents. This is particularly troubling when applied to the new source standards that are based on a snapshot stack test of the single best performing unit. EPA concedes that it does not know whether the stack test data is even representative of typical operation for the source that was tested. The preamble to the Proposed Rule states: NACWA Comments on Docket ID No. EPA-HQ-OAR-2009-0559 November 29, 2010 Page 28 It is not clear from the data available to EPA whether the sludge burned during the emissions tests (that were used to establish the MACT floor) represent typical sludge composition/concentrations or are closer to minimum or maximum levels. 75 Fed. Reg. at 63268. This is an essential missing step in determining how to use the data to set an emission limit. If the sludge burned during an emissions test was not at or near the maximum constituent concentration level (e.g., due to seasonal variability), a new source emission limit based on this data could not be achieved over the full range of expected normal operating conditions confronted by the best performing source. At a minimum, EPA must consider all available data (including Part 503 data) for the best performing source and use that to establish a variability factor applied to the stack test data. EPA’s request for metals data during the stack test is insufficient to account for the full intra-source variability. Moreover, variability for the compounds not regulated by Part 503 must be accounted for as well before setting the new source limit. E. EPA’s Misunderstanding of the Variability of Sewage Sludge Leads to Erroneous MACT Floors for Existing Sources EPA also mistakenly presumes that snapshot stack test data is sufficient to set limits for existing sources because sludge concentration variability is insignificant. As indicated above, EPA did not collect enough data to set existing source limits based on actual data. As such, EPA can only set emission limits by extrapolating based on the presumption that the data it has is representative of the top performing 12 percent of the source subcategories. The fact that these data were not randomly selected undermines the statistical credibility of this exercise. But even if the data were randomly selected from the full source category, the stack test data is inadequately representative as the basis for a lawful MACT floor calculation. EPA cannot use statistics to make up for its lack of representative data because each step compounds the problem. EPA uses its flawed predicted data set to determine the top 12 percent of existing sources in each source category, and to determine the average emission rate, the intra-source variability, and the upper prediction limit (“UPL”) for the subcategory. The sources that are statistically fabricated to fill the data gaps lack the variability of the actual sources that round out the top 12 percent. EPA’s method creates a more stringent emission limit than would be justified under a proper assessment of the top 12 percent of existing sources. As such, the proposed emission limits constitute unlawful beyond-the-floor MACT standards that have been developed without considering the required criteria in the CAA, including cost. EPA’s request for additional stack test data to bolster its inadequate database is an inappropriate burden at this stage of the rulemaking. EPA has primary responsibility for gathering the data. It is a circumvention of the PRA and OMB’s review to intentionally limit the scope of an ICR to nine entities and then ask NACWA and its members to provide additional stack test data during the 45-day public comment period. EPA’s data analysis is an important part of generating the regulatory options. Waiting until the public comment period to gather new data undermines NACWA’s ability to comment on the proposed options for regulation. EPA is rushing to the finish line with inadequate data and hoping that the public comments will contribute enough data to justify its proposed path. Even if NACWA and its members were able to secure the resources to conduct additional stack testing to supplement the record, EPA offers insufficient time to conduct NACWA Comments on Docket ID No. EPA-HQ-OAR-2009-0559 November 29, 2010 Page 29 the stack tests, quality assure the data and submit it for EPA consideration. NACWA has requested, and EPA has rejected, an extension of time for the comment period on the Proposed Rule. EPA should not now blame POTWs for not generating sufficient data during the comment period. F. EPA Should Not Adjust the Upper Predictive Limit to Compensate for its Lack of Data. EPA seeks comment on whether to use the 99 percent UPL or a modified 95 percent UPL. EPA should not use a 95 percent UPL to compensate for its lack of data. This essentially presumes that sources represented by the database would be expected to be out of compliance 5 percent of the time. This is contrary to the CAA requirement that MACT limits must be met at all times and under all reasonably foreseeable conditions. See Sierra Club v. EPA, 551 F.3d 1019 (D.C. Cir. 2008). EPA should use a 99 percent UPL to set enforceable limits as has been done in other recent rulemakings. Adjusting the percent basis for the UPL to get to a target emission limit that is more palatable to the Agency would turn the MACT process on its head and be vulnerable to legal challenge. G. MACT Floor Methodology Must Reflect an Achieved Emission Limit The proposed MACT standards for SSIs are based on pollutant-by-pollutant analyses that rely on a different set of best-performing sources for each separate § 129 pollutant. See 75 Fed. Reg. at 63270 (“The MACT floor analysis was then conducted using all the emissions information for each pollutant in each subcategory.”). EPA cannot demonstrate that any of these units is capable of meeting all of the proposed emission limits. The Metropolitan Council Environmental Services’ Metropolitan Treatment Plant, which is the best performing source for some pollutants, cannot reliably meet all of the limits under all operating conditions because EPA failed to consider the true operational variability of POTWs. The result is a set of standards that reflect hypothetical performance of a set of sources that simultaneously achieve the greatest emission reductions for each and every § 129 pollutant without regard to whether any such SSI actually exists or whether the resulting standards are in fact achievable by any SSI. This approach is untenable and contrary to the language of § 129. CAA § 129 unambiguously directs EPA to set standards based on the overall performance of incineration units. Section 129 specifies that emissions standards must be established based on the performance of “units” in the category or subcategory and that EPA’s discretion in setting standards for such units is limited to distinguishing among classes, types, and sizes of units. These provisions make clear that standards must be based on actual units and cannot be the product of pollutant-by-pollutant parsing that results in a set of composite standards that do not reflect the overall performance of any actual unit. Congress provided express limits on EPA’s authority to parse units and sources for purposes of setting standards under § 129 and that express authority does not allow EPA to “distinguish” units and sources by individual pollutant as is proposed in this rule. See Sierra Club, 551 F.3d at 1028. Section 129 MACT standards present very real problems in this regard because the statute requires maximum achievable limits for both nitrogen oxides and carbon monoxide. For combustion sources, carbon monoxide is controlled by increasing excess oxygen and combustion temperature and nitrogen oxides are controlled by decreasing excess oxygen and combustion temperature. A unit that is over-controlling its combustion unit for NOx would be expected to increase CO. Similarly, a unit that is focused on CO reduction NACWA Comments on Docket ID No. EPA-HQ-OAR-2009-0559 November 29, 2010 Page 30 in the combustion zone would be expected to increase NOx emissions. If EPA chooses a top performer for NOx and a different top performer for CO, the emission limits may well be unachievable for any source. While afterburners may be added for post-combustion CO control, these units burn fuel that results in a corresponding increase in emissions of NOx and CO. The better alternative is to set limits based on the best performers for both pollutants simultaneously. VI. EPA’ S TWO PROPOSED SUBCATEGORIES ARE INADEQUATE TO ACCOUNT FOR OPERATIONAL DIFFERENCES AMONG TYPES AND CLASSES OF SSIS A. MACT Subcategories for Existing Sources Should Also Apply to New Sources In the Proposed Rule, EPA correctly determined that MHIs and FBIs are distinct types of combustion units that justify separate subcategories. See 75 Fed. Reg. at 63268. FB incinerators have higher turbulences that increase combustion efficiencies and reduce particulate emissions, and many have internal afterburning zones that increase residence time to reduce carbon monoxide, hydrocarbon, and particulate emissions. MH incinerators do not share these characteristics. All of these differences result in lower emissions from FB incinerators that cannot be duplicated by MH incinerators. Based on these distinctions, EPA created distinct MH incinerator and FB incinerator subcategories. However, EPA only recognized this distinction in setting emission limits for existing units. EPA abandoned these subcategories for new units by basing emission limits for all new incinerators, whether FB or MH design, on emission limits achieved by the best performing FB incinerator. EPA concluded that no new MH incinerators have been built recently and none are expected to be built in the future, and that therefore all newly built units would be FB design. EPA cannot make this determination for POTWs. In fact, EPA is essentially setting a beyond-the-floor MACT limit for MHIs without considering any of the criteria that the statute requires. The new source limits in the proposed rule must reflect the best performing similar source for the multiple hearth design. EPA also seems to ignore the “modification” trigger for the new source standards. EPA’s proposed definition of “modification” could make existing MH units subject to the new source FB-based standards. EPA would impose the new standards on units for which the “cumulative cost of the changes over the life of the unit exceeds 50 percent of the original cost of building and installing the SSI unit.” First, EPA must make clear in the final SSI rule that the cumulative costs to be considered are only those costs incurred since the effective date of the final SSI rule. This is the approach taken in the Municipal Waste Combustor § 129 rule at 40 CFR § 60.51b. The “life of the unit” reference in the definition of modification could be misconstrued as a retroactive evaluation of incurred costs that pre-date the rule. Second, modified MH units are placed in the untenable position of having to meet emission limits set by the best performing FB incinerator – an impossible feat due to the inherent design differences already recognized by EPA. This will prevent all existing MH units, which make up 75 percent of existing SSI units in the U.S., from investing in changes to those units due to the risk that they might qualify as modifications of the type that would trigger unachievable new source performance standards. This includes combustion efficiency improvements or changes that increase steam output for electricity generation, when these environmentally beneficial changes meet the definition of modification proposed at 40 CFR § 60.4930. NACWA Comments on Docket ID No. EPA-HQ-OAR-2009-0559 November 29, 2010 Page 31 EPA’s approach would discourage incremental improvements at MH units because they would trigger FB-based emission limits that cannot be met. EPA has acknowledged the design differences that make meeting these limits impossible, and has provided no pathway by which these modified or reconstructed MH incinerator units may achieve compliance. EPA must retain the separate MH and FB incinerator subcategories for both new and existing sources to avoid illegally subjecting newly-constructed or modified MH SSIs to unachievable and unjustified beyond-the-floor emission standards. Establishing separate emission limits for new MH and FB incinerators will also preserve incentives to be innovative and improve the use of the 163 MH incinerators currently in use in the U.S. See 75 Fed. Reg. at 63268. B. EPA Should Consider Additional Subcategories Based on Use, Size, and Class Section 129(a)(2) provides that EPA “may distinguish among classes, types (including mass-burn, refuse-derived fuel, modular and other types of units) and sizes of sources within a category” when establishing MACT standards. These provisions vest EPA with broad authority to group like units for purposes of establishing emissions limitations. Given the significant differences in operating time, size, and input between incinerators in both the MH and FB categories, it is appropriate for EPA to create further subcategories. After choosing to gather data from only nine entities and consider only add-on pollution control technologies, EPA must not use the lack of data as an excuse for not establishing additional subcategories that may make this standard more achievable for SSIs of all sizes and types. 1. Limited Use SSIs EPA should include an additional subcategory for limited use SSIs that function as back-up or emergency units and operate at only 10 percent of their annual capacity. These units may be essential to ensuring that POTWs can continue processing sewage sludge in a safe manner when a primary unit is unavailable. In past rulemakings, EPA has created separate subcategories for limited use and emergency units, recognizing that their unique operating conditions could make it difficult if not impossible to meet the requirements applicable to regularly operating units. In the 2004 Industrial, Commercial, and Institutional Boiler and Process Heater MACT rulemaking, EPA recognized that back-up boilers, which operate “10 percent of the year or less,” “are different compared to typical industrial, commercial, and institutional boilers” and that “such limited use units should have their own subcategory.” 69 Fed. Reg. at 55232. Similarly, in setting NESHAP for Reciprocating Internal Combustion Engines, (“CI RICE”), 75 Fed. Reg. 9648 (Mar. 3, 2010), EPA recognized that stationary existing CI RICE should be divided into non-emergency and emergency categories “in order to capture the unique differences between these types of engines.” Id. at 9650. In that rulemaking, EPA found that as “emissions occur only during emergency situations or for a very short time to perform maintenance checks and operator training,” EPA found that “[e]missions from these units are expected to be low on an annual basis.” 18 18 Subcategorization of Stationary Reciprocating Internal Combustion Engines ≤500 HP at 5 (May 15, 2006) (EPA-HQ-OAR-2005-0030- 0012). While these statements focus on an “emergency use” subcategory, it is important to note that the limited duration of the use, not the purpose for using the CI RICE is the key issue. For example, the same rule also creates a subcategory for “black start” engines (engines used to start a turbine generator), which operate during both “emergency and high demand days.” 75 Fed. Reg. at 9662. NACWA Comments on Docket ID No. EPA-HQ-OAR-2009-0559 November 29, 2010 Page 32 Some SSI units share several similarities with limited use boilers and emergency CI RICE units. Limited use SSIs are put into service during shutdown or curtailment of a primary incinerator and they operate until the primary unit returns to full service. As a result, these units operate for only a small portion of the year, typically 10 percent of the year or less. Because of this limited operation, annual emissions are expected to be low, making the cost of add-on pollution controls for beyond-the-floor mercury control, performance testing and other requirements unjustified. Also, since these units sustain operation for a shorter duration, they spend a larger percentage of their time in startup and shutdown modes, which have different emission characteristics. Furthermore, limited use SSIs cannot feasibly accommodate the Proposed Rule’s testing obligations without running for a considerably longer period than they would typically otherwise operate. Limited use units also may not operate under steady-state conditions for sufficient periods of time to enable testing. Testing would require these units to extend operations beyond routinely scheduled operations to conduct the testing required by the Proposed Rule. Testing is also problematic for limited use units because the Proposed Rule requires submission of a Notification of Intent at least 30 days before any performance test. See 75 Fed. Reg. at 63330 (40 CFR. § 60.5220(a)(8)). Due to this requirement, even if a limited use unit was operated for an entire month after an unplanned start, there would be no time to conduct a properly noticed performance test. NACWA asks that EPA establish a limited use subcategory set at 10 percent of the annual heat input capacity of a unit. These units would have emission limits that are based on their unique emission characteristics from spending a larger percentage of their time in startup and shutdown modes. Also, beyond- the-floor mercury controls would not be justified because their relatively short periods of operation would render the additional controls not cost effective. 2. Space Constraints Justify a Size Subcategory Additional subcategories based on size are also appropriate. EPA should exercise its discretion to establish a subcategory for small entities that operate SSIs to relieve the disproportionate cost burden on these small entities. EPA is required to consider mitigation measures for these small entities under the RFA, and establishing a separate subcategory that is consistent with the definition of small entities allows EPA to target relief to these entities to satisfy that obligation. Also, some POTWs will face significant challenges due to space constraints at their facilities. SSIs must be located within the solids handling portion of the POTW and space constraints in these areas of POTWs make it technically infeasible to install large pieces of control equipment. EPA assumes that control equipment can easily fit into existing sites, but this ignores the fact that many POTWs are located in areas where expansion is impossible or impracticable. At NEORSD, only one of three treatment plants could accommodate the additional 30 feet of space needed to install the ductwork, contact chamber and fabric filter system necessary to accommodate an ACI system; the other two treatment plants cannot accommodate them at all. See NEORSD Comments on the Proposed Rule. Even larger systems may be necessary to control emissions to control SO2, NOx, and HCl under the full range of operating conditions confronted by POTWs. EPA’s control cost assumptions are based on data that do not reflect all operating conditions, so NACWA is skeptical of EPA’s conclusion that most units will not need to install control equipment to meet other emission standards. A subcategory for these small or space-constrained units is warranted to accommodate the unique characteristics of these units. 3. Unique Sewage Sludge Characteristics Justify a Class Subcategory NACWA Comments on Docket ID No. EPA-HQ-OAR-2009-0559 November 29, 2010 Page 33 Unique differences in the content of sewage sludge may also justify additional subcategories based on the “class” of SSI. For example, there are differences in the composition of sludge feed from treatment plant to treatment plant, which can also vary within a plant depending on the time of the year. There are also differences if the sludge is digested prior to incineration and differences between non-digested sludges that are thermally conditioned or chemically conditioned. Sludge that is digested prior to incineration typically has a solids content that is 50 percent volatile solids or less, compared with 55-80 percent volatile solids for non- digested sludge. See Water Environment Federation’s (WEF’s) Manual of Practice: Wastewater Solids Incineration Systems (MOP-30). The low volatile content in digested sludge necessitates more auxiliary fuel to sustain combustion, which changes emission characteristics. Also, thermally conditioned sludge typically has a higher solids content (up to 48 percent solids compared with 20-30 percent solids for chemically conditioned sludge), which results in a higher combustion temperature. Thermally conditioned sludge would be expected to have lower CO emissions and higher NOx emissions due to its higher combustion temperature. These unique attributes justify separate subcategories for digested, thermally conditioned, and other (chemically conditioned) sewage sludge incinerators. EPA should use its discretion to subcategorize by class to ensure all units have a path to compliance. 4. Distance to Landfill Justifies a Type Subcategory As indicated above, some POTWs are so far away from a landfill that they will have higher emissions from diesel exhaust during sewage sludge transport than from incineration. EPA should not create emission standards so stringent that they force an environmental detriment. For example, MACT standards that force sewage sludge to be transported to landfill or land application sites equal to or greater than 65 miles away emit higher quantities of the pollutants regulated under § 129 than the emissions from their incinerator. These units should be placed in a separate subcategory to ensure that beyond the floor MACT limits do not force incinerators at these POTWs to shut down resulting in transportation emissions in excess of the emissions that would have been reduced by the MACT rule. VII. THE PROPOSED RULE PLACES UNJUSTIFIED BURDENS ON OPERATORS A. EPA Should Require Compliance Stack Testing No More Than Once Per Permit Term The Proposed Rule requires annual stack testing to demonstrate compliance with emission limits for particulate matter, hydrogen chloride, dioxins/furans, mercury, nitrogen oxides, sulfur dioxide, cadmium, lead, opacity, and fugitive emissions from ash handling. See 75 Fed. Reg. at 63303. EPA estimates that stack testing for all of these pollutants costs an average of $61,000 per test. See ERG Memo, Cost and Emission Reduction of the MACT Floor Level of Control at 5, Tbl. 6c (June 2010) (“Cost and Emission Reduction Memorandum”). NACWA members report that this number may be significantly higher. Factoring this additional annual cost into already strained municipal budgets will place an extraordinary burden on regulated entities. As EPA recognized, nearly all SSI units are operated by municipal entities. Municipal governments across the country have been struggling with the current economic climate, and many already face budget deficits in the face of declining revenues. Requiring testing on an annual basis would place a significant burden on the municipal entities subject to the SSI rule. NACWA Comments on Docket ID No. EPA-HQ-OAR-2009-0559 November 29, 2010 Page 34 As noted in Part I.A., supra, POTWs are already subject to comprehensive management practices, strict health-based sludge content limits, and NESHAPs under CWA § 405 and the Part 503 regulations. Annual stack testing is not necessary to ensure compliance with emission limits, and such frequent testing would disrupt normal operations that are providing a critical public service.19 The final rule should require stack testing no more than once per permit term. Minimizing the cost and disruption of stack tests will help alleviate the burden on municipal entities and reduce interference with operations without increasing risks to human health and the environment. B. The Relatively Inert Characteristics of Sewage Sludge Do Not Justify Stringent Operating, Monitoring and Recordkeeping Requirements EPA attempts to make the monitoring requirements in the Proposed Rule consistent with those in the Hospital/Medical/Infectious Waste Incinerators (“HMIWI”) rule. See Cost and Emission Reduction Memorandum at 4. However, the waste combusted in HMIWI units is completely different from the waste combusted in SSIs. There are no similarities between the two waste streams that justify basing SSIs rules on rules developed for infectious waste. EPA’s decision to base monitoring and recordkeeping requirements for SSI units on those developed for HMIWI units simply because they are both incinerators is arbitrary and capricious. Daily pressure taps, as proposed by EPA, are not necessary or feasible for SSI units. Disconnecting pressure taps from related parameter monitoring equipment can trip safety mechanisms and cause the system to shut down. Causing daily shutdowns of the system is an unreasonable interference with POTW operations. Furthermore, operators have not encountered problems with clogged taps, making daily checks unnecessary. C. SSI Operations Respond to External Factors Affecting Operating Parameter Values That Cannot Be Kept Within an Arbitrarily Designated Range. EPA has also set operating parameters that will be impossible for most SSI operators to meet. EPA proposes to set operating limits based on the operating parameter values during stack tests. As indicated above, stack tests take a snap shot of a highly variable process. It is unreasonable for operating parameters to remain within plus or minus 10 percent of the minimum or maximum value generated during a stack test. SSI feed rates and moisture contents necessarily vary widely depending upon the amount of wastewater that is coming into the POTW. During storms and other high flow events, the POTW is working hard to keep up with the influent and the sewage sludge feed rates and moisture content are necessarily on the higher end of the normal range. During low flow periods, SSIs may operate significantly below maximum feed rates with sewage sludge moisture at the low end of the normal range. POTWs are responding to external events and cannot control these variables sufficiently to stay within a designated range. To accommodate site-specific variation, operating and maintenance parameters should be established in site-specific operating plans that focus on the parameters that correlate with control device efficiency. D. Use of a Bypass Stack Does Not Indicate a Deviation for All Emission Standards 19 EPA requires performance testing less frequently than annually in several MACT rules. See, e.g., NESHAP for Coke Ovens: Pushing, Quenching, and Battery Stacks, 40 CFR. § 63.7321 (testing twice per permit term); NESHAP for Lime Manufacturing Plants, 40 CFR. § 63.7111 (testing every 5 years); NESHAP for Iron and Steel Foundries, 40 CFR. § 63.7731(a) (testing every 5 years). NACWA Comments on Docket ID No. EPA-HQ-OAR-2009-0559 November 29, 2010 Page 35 During malfunctions or what our industry refers to as emergency conditions (e.g., loss of draft within a multiple hearth incinerator, loss of the induced draft fan, loss of scrubber water, loss of power, etc.), sludge feed to a multiple hearth incinerator is promptly stopped and all of the sludge within the incinerator (approximately 30 - 45 minutes worth of sludge feed) is burned out. It takes approximately 30 minutes for the burn-out to be completed. If a relief stack is not utilized during emergency conditions, the health and safety of operating personnel can be compromised and the incinerator and associated equipment can be damaged beyond repair. Fluidized bed incinerators are not equipped with emergency relief stacks. During emergency conditions, the sludge feed to the incinerator is promptly stopped and the fluidizing air blower is shut down. Combustion within the fluidized bed incinerator will cease within a short amount of time, due to the low volume of non-combusted sludge within the unit. In 40 CFR § 60.4900(d), EPA arbitrarily deems use of a bypass stack when sewage sludge is being charged an “emissions standards deviation for all pollutants listed in Table 1.” Bypass stacks are an essential part of the safety equipment and operators should be allowed to open the bypass stack immediately as part of a continuous series of events that includes stopping the sludge feed without triggering a deviation. Moreover, EPA does not have the authority to presume deviations of emission standards, particularly when the bypassed emission controls may not be necessary to meet the emission standard. Contemporaneous Part 503 data can show that the sludge content of cadmium, lead or mercury were so low that it could be met without operating a control device. Use of a bypass stack must not be deemed a deviation of emission limitations for these compounds in the face of this credible contrary evidence. Assuming that the use of a bypass stack results in emission violations for all pollutants is arbitrary and subsection (d) of 40 CFR § 60.4900 should be removed from the final rule. VIII. THE PROPOSED RULE FAILS TO PRESCRIBE ACHIEVABLE STANDARDS FOR STARTUP, SHUTDOWN AND MALFUNCTION PERIODS NACWA is very concerned about EPA’s proposal to apply the same proposed emission standards for steady-state operating periods to startup, shutdown and malfunction (“SSM”) events. EPA incorrectly claims that its authority to prescribe unique standards for SSM periods is constrained by Sierra Club v. EPA, 551 F.3d 1019 (D.C. Cir. 2008). Although the preamble states that EPA “believes” SSIs will be able to achieve the proposed standards during startup and shutdown (75 Fed. Reg. at 63282), this claim is not supported by the record and NACWA believes this position fails to recognize that some startup and shutdown conditions may temporarily increase emissions, even among the “best performing” SSIs. EPA makes no claim that SSIs will be able to achieve the proposed standards during malfunctions, yet it does not propose unique standards for malfunction periods because they “should not be viewed as a distinct operating mode and, therefore, any emissions that occur at such times do not need to be factored into development of CAA section 129 standards, which, once promulgated, apply at all times.” Id. We believe that EPA’s proposed SSM approach is flawed from the outset because the Agency starts with the premise that the D.C. Circuit’s decision in Sierra Club supports EPA’s proposed SSM approach – applying NACWA Comments on Docket ID No. EPA-HQ-OAR-2009-0559 November 29, 2010 Page 36 the same emission standards for all operating conditions. EPA states that Sierra Club “requires EPA to apply MACT emission standards on a continuous basis, thereby precluding exemptions applied for malfunctions or other singular events.” 75 Fed. Reg. at 63282 (emphasis added). EPA concludes: “Therefore, consistent with Sierra Club v. EPA, EPA is proposing that the standards in this rule [i.e., a single set of standards developed for steady-state operating conditions] apply at all times.” Id. While it is true that a blanket exemption from any standard may be inconsistent with the Sierra Club holding, the opinion does not prohibit EPA from applying different, even non-numerical, standards during SSM events from those standards that apply during steady-state operations. The court only rejected EPA’s decision not to impose any emission standard whatsoever during SSM periods. See 551 F.3d at 1027-28. In fact, Sierra Club acknowledged that the definition of emission standard in § 302(k) indicates that any one standard need not apply at all times. The court noted that the Part 63 General Provisions at issue in Sierra Club were not a design, equipment, work practice, or operational standard under § 112(h) and expressly did not decide whether EPA could promulgate a work practice or engineering standard under CAA § 112(h) (much less under the analogous § 111(h)) instead of the exemption EPA sought to defend. See 551 F.3d at 1028. Far from supporting EPA’s effort to ignore SSM periods, Sierra Club actually highlights the legal weakness of EPA’s proposed SSM approach. Therefore, EPA cannot use Sierra Club to justify not setting different standards for SSM events. A. EPA’s Rationale for its SSM Approach is Inconsistent with the CAA and Factually Unsound Beyond not being supported by Sierra Club, NACWA believes EPA’s SSM approach is contrary to the requirement under § 129(a)(2) that emission standards be achievable in practice by the best performing units. EPA deals with the complex issue of setting standards for SSM with a bare statement that emissions from supplemental fuels during startup “are expected to generally be lower than from burning solid wastes” – presumably meaning sewage sludge – and that emissions during shutdowns “are also generally lower than emissions during normal operations because the materials in the incinerator would be almost fully combusted before shutdown occurs.” 75 Fed. Reg. at 63282-83. Finally, EPA claims that its variability analysis has adequately addressed “any minor variability that may potentially occur during startup or shutdown.” 75 Fed. Reg. at 63283. For the same reasons discussed in Section IV of these comments, EPA’s variability analysis cannot justify the use of the same emission standards for SSM periods. That analysis includes only limited stack test data from the nine ICR POTWs and none of those data include SSM periods. Thus, the data used in the variability analysis are not representative of SSM events, and statistical analysis alone cannot correct this fundamental flaw. Aside from the shortcoming that there is no information supporting EPA’s view that SSIs can comply with the proposed standards during SSM events, there is reason to believe that emissions of some pollutants can be expected to be higher during startups and shutdowns. For example, emissions of CO and PM can be expected to be elevated during startup when oxygen levels are higher due to lower combustion temperatures, resulting in higher pollutant concentrations when corrected to 7 percent oxygen. Similarly, absolute pollutant levels tend to increase during startup and shutdown due to incomplete combustion that is unavoidable at lower temperatures. The influence of unstable combustion may be more pronounced during shutdowns as the incinerator combusts the remaining sewage sludge for 30 minutes or more. EPA should also account for NACWA Comments on Docket ID No. EPA-HQ-OAR-2009-0559 November 29, 2010 Page 37 situations where higher emissions occur during the time it takes to bring control equipment from startup to steady-state operations. In the case of malfunctions, NACWA disagrees with EPA’s newly articulated view that malfunctions are not distinct operating conditions from steady-state operations. SSI operators must treat malfunctions as very distinct events from steady-state operations, depending on the severity of the malfunction requiring anything from shutdown of the unit to emergency fire response actions. Depending on the nature and severity of the malfunction event, emissions often are not capable of being captured and routed to a stack for control and/or measurement and, when they are, test methods do not adequately account for the often short-term and unstable characteristics of the malfunction event. EPA acknowledges that even the best performing SSIs are subject to any number of potential malfunctions and that the factual complexity of differing processes and of the severity, frequency and duration of malfunctions makes standard setting difficult. So it appears EPA’s “determination” that malfunctions are not distinct operating conditions is simply an unjustified decision to ignore the impact that the inherent limitations of combustion and pollution control technologies have on the ability of SSIs to achieve the proposed standards. Compounding this error, EPA declines to use emissions during malfunction periods in its MACT floor analysis for normal operating conditions and fails entirely to address its authority under §111(h) to set alternative work practice and engineering standards. EPA’s only explanation for this decision is its suggestion that applying the MACT floor concept of “best performing” to a source experiencing a malfunction “presents significant difficulties.” 75 Fed. Reg. at 63283. This argument has no basis in the CAA, which requires EPA to distinguish among types and classes of sources in order to set achievable emission standards and which allows EPA to use a variety of alternative work practice standards when setting an emission standard is too difficult. EPA’s thin semantic argument leads it to ignore the fact that there are work practices employed by the best performing SSIs that represent the best practices for minimizing emissions during a malfunction. These practices may include monitoring combustion parameters to identify a malfunction and stopping the charging of materials to an incinerator. While the measures that represent these best practices will depend on facility-specific issues, such as incinerator design, pollution control train, and other factors, they are nonetheless the “best technological system of continuous emission reduction … adequately demonstrated.” NACWA believes the reality of the technological challenges, and the enormous potential cost that would be necessary to monitor SSIs during SSM, give EPA the basis to prescribe alternative design, equipment, work practice or operational standards for SSM. Therefore, NACWA asks EPA to reconsider its SSM approach and to allow for alternative work practice standards for SSM events in the form of a facility-specific SSM plan. EPA should consider a flexible compliance option that allows the source to elect to comply with the MACT floor emission standards for normal operating conditions or with the requirements of the SSM plan for the SSM event. B. EPA’s “Affirmative Defense” Rests on Unsound Legal Footing and Cannot Make Up for a Failure to Set Achievable Standards for SSM Periods Recognizing that SSIs experiencing malfunctions will not always be able to comply with the proposed emission standards, EPA offers an “affirmative defense” that shifts the burden to the POTW to prove that a lengthy list of criteria are met and actions are taken by the source in order to avoid enforcement for civil NACWA Comments on Docket ID No. EPA-HQ-OAR-2009-0559 November 29, 2010 Page 38 penalties. It is unclear where EPA finds the legal authority in the CAA to shift the burden to the regulated community of proving (or disproving) essential elements of an alleged violation. The statute is silent as to the issue and “the ordinary default rule [is] that plaintiffs bear the risk of failing to prove their claims.” Shaeffer v. Weast, 546 U.S. 49 (2005), quoting McCormick on Evidence § 337, at 412 (“The burdens of pleading and proof with regard to most facts have and should be assigned to the plaintiff who generally seeks to change the present state of affairs and who therefore naturally should be expected to bear the risk of failure or proof or persuasion”); C. Mueller & L. Kirkpatrick, Evidence § 3.1, p. 104 (3d ed. 2003) (“Perhaps the broadest and most accepted idea is that the person who seeks court action should justify the request, which means that the plaintiffs bear the burdens on the elements in their claims”). While the Supreme Court has recognized exceptions such as affirmative defenses, courts retain the authority to establish such rules unless Congress acts to delegate that authority. EPA also does not justify why an affirmative defense would not apply to startup and shutdown events, and would be limited only to “civil penalties,” excluding other remedies such as claims for injunctive relief. The proposed regulations also contain elements that are vague and not reasonably connected to whether a source should be penalized for a malfunction event. The following lists some of the most extreme examples: • The condition requiring “all possible steps” to minimize the impact of the excess emissions is unreasonable in terms of its potential operational and economic impact and therefore essentially impossible to satisfy. • The requirement that emissions control systems be “kept in operation if at all possible” and that any bypass be permitted only if “unavoidable to prevent loss of life, severe personal injury, or severe property damage” are both unwise and unreasonable. It is often the best operating practice to shut down an emission control device rather than have the equipment risk serious damage. And the latter requirement suggests that EPA is tacitly encouraging sources to weigh worker safety practices (albeit not those that risk “severe” injury) against the risk of enforcement under the CAA. • Some elements of the affirmative defense are impossibly vague, such as whether “proper design” would have prevented a malfunction. • The requirement to prepare a written root cause analysis “to determine, correct, and eliminate the primary causes of the malfunction” is unreasonably stringent and would result in some categories of malfunctions potentially never satisfying the proposed affirmative defense. For example, the only method a source has to “correct and eliminate” a malfunction caused by a power failure from the grid is to install redundant power generating facilities. The technological and economic impacts of “super engineering” facilities in the way envisioned by EPA are unreasonable and disconnected from the authority EPA has to set standards under the CAA. Due to the legal uncertainty and unreasonably stringent restrictions on the proposed “affirmative defense,” in a future rulemaking NACWA urges EPA to reconsider the need for separate SSM standards and to propose alternative work practice and engineering standards for SSIs. NACWA Comments on Docket ID No. EPA-HQ-OAR-2009-0559 November 29, 2010 Page 39 IX. AVERAGING TIMES FOR EMISSION LIMITS AND OPERATING LIMITS SHOULD BE EXPANDED TO ACCOMMODATE THE SIGNIFICANT VARIABILITY OF SEWAGE SLUDGE EPA requested comment on whether the averaging time using continuous emission monitoring systems (CEMS) should be reduced from 24 hours to 12 hours. See 75 Fed. Reg. at 63281. NACWA asks that EPA use broad averaging times (e.g., 30 days or longer) for CEMS, sludge concentration data, and operating parameters to accommodate the significant and unpredictable variability in the influent that must be processed by POTWs. Spikes in compound concentrations are unavoidable in a sewer system and they are unlikely to be captured in the data that EPA has gathered for its MACT database. High concentrations of mercury may be released when segments of the sewer are periodically cleaned downstream of a dentist office where mercury-containing amalgams have been discharged into the sewer. The sewer system can also be the discharge point for ‘spring- cleaning’ activities, spills, and other periodic discharges that typically arrive at the POTW without warning. The POTW does not have the luxury of refusing material discharged in most instances. It must do its best to respond by cleaning the wastewater before it is discharged to a local water body, which redirects these compounds in higher concentrations to the sewage sludge. Longer averaging times help mitigate the effect of spikes on compliance demonstrations. NEORSD’s Part 503 data from weekly sludge samples taken from 2003 to 2009 reveal that the maximum concentration of lead and mercury were over five times higher than the median concentrations. See NEORSD Comments on the Proposed Rule. Spikes of this magnitude are typical for POTWs and they justify broad averaging periods. It takes a significant number of normal data points to bring the median back from a 5-times spike.20 While the emission standards in the Proposed Rule apply during these spikes, these periodic elevated concentrations are not reflected in EPA’s database. EPA should help mitigate the burden of its lack of data by maximizing the averaging times for demonstrating compliance. EPA should not dictate the averaging times for operating parameters. Facilities should have the flexibility to develop an operations and maintenance plan for their control devices and emission reduction methods. EPA improperly directs that the operating parameter range be limited to the range established during a single performance test. If a stack test confirms that emissions are 50 percent of the emissions limit, the operating parameters derived from that test would be ensuring that emissions remain controlled to 50 percent of the emissions limit during future operations. This more stringent limit is unlawful. It is set at a level that is unquestionably lower than the MACT floor, yet EPA has not conducted any of the analysis necessary to justify a more stringent, beyond-the-floor MACT limit. The operating parameter range should reflect the full range of operating conditions that correlate with emissions up to the emissions limitation. The POTW should be able to develop a continuous compliance plan based on the stack test data and all other available information regarding the correlation of operating parameters to control device performance. This ensures that the operating parameters actually correlate to performance at the emission limit and not at the rate captured during the stack test. X. NACWA SUPPORTS EPA’S APPROACH ON THE FOLLOWING AREAS FOR WHICH EPA REQUESTED COMMENT 20Assuming for illustration purposes that the median concentration is 10 and the limitation is 15, a spike that is five times the median concentration would require eight weeks of normal sample results to bring the average back to the limit. NACWA Comments on Docket ID No. EPA-HQ-OAR-2009-0559 November 29, 2010 Page 40 A. No Numerical Emission Limit for PM2.5 NACWA supports EPA’s determination not to include an emission standard for PM2.5 due to the potential interferences with moisture from scrubber controls. It is reasonable for EPA to control both total and fine particulate by using surrogates because the use of wet scrubbers is not always compatible with OTM 27. Entrained water droplets that occur when stack gas moisture levels exceed vapor capacity can bias PM2.5 particle measurements and provide inaccurate readings of filterable PM2.5. It would be unreasonable for EPA to include emission limits that are not subject to accurate measurement. Furthermore, the same control devices would be required to control both total and final particulate. Cf. Nat’l Lime Assoc. v. EPA, 233 F.3d 625, 637 (D.C. Cir. 2000) (citing Dithiocarbamate Task Force v. EPA, 98 F.3d 1394, 1399 (D.C. Cir. 1996)). Limits on filterable particulate matter and opacity provide an adequate surrogate for direct fine particulate matter contributing to PM2.5. There is also a demonstrated correlation between reductions in SO2 and NOx and reductions in nitrate and sulfate contributions to PM2.5.21 Therefore, the reductions in SO2 and NOx resulting from these emission limits will also serve to reduce contributions to PM2.5. Given the technical difficulties with accurately measuring PM2.5 emissions, and the reductions in PM2.5 contributors already incorporated into the rule, it is reasonable for EPA to use filterable PM10 as a surrogate for PM2.5 emissions. B. No Beyond the Floor Controls for Non-Mercury Pollutants NACWA supports EPA’s decision not to impose beyond the floor controls for non-mercury pollutants, including CO. The 100 ppmv CO limit in Part 503 is not required for all sources; it is one option for demonstrating compliance with the total hydrocarbon limits contained in the rule. See 40 CFR § 503.40(c)(2). Thus, Part 503 does not require sources to install afterburners. Any beyond the floor controls for CO imposed under § 129 would require MH incinerators to retrofit a regenerative thermal oxidizer or other afterburner device onto existing equipment. EPA estimated that installing an afterburner device would increase the cost of MACT compliance by over $145 million, and would have uncertain effects on emissions. See ERG Memo, Analysis of Beyond the Maximum Achievable Control Technology (MACT) Floor Controls for Existing SSI Units at 4-5 & Attachment C.2 (Aug. 2010). Furthermore, afterburners require the use of supplemental fuel, which EPA estimates could require annual combustion of an additional 1700 million cubic feet of natural gas. This would lead to emissions of an additional 84 tons of NOx, 70 tons of CO per year, and an additional amount of greenhouse gas emissions. See 75 Fed. Reg. at 63277. In light of these determinations, EPA was correct to eliminate beyond the floor emission limits from the Proposed Rule. C. Flexible Compliance Options 1. Use of Continuous Emission Monitors in Place of Parametric Monitoring and Annual Testing NACWA supports flexible compliance options allowing, but not mandating, CEMS for demonstrating compliance. NACWA also supports allowing sources to propose site-specific operation and maintenance plans with procedures for addressing missing CEMS data. Providing small entities with a choice of compliance options is particularly important, as it will allow each POTW to independently assess which compliance option 21 Russell R. Dickerson et al., PM2.5 Maryland State Implementation Plan Weight of Evidence Report 4-40 to 6-40 (Jan. 30, 2008) NACWA Comments on Docket ID No. EPA-HQ-OAR-2009-0559 November 29, 2010 Page 41 is most economical for that facility. NACWA also supports EPA’s determination that parametric monitoring and annual testing can be eliminated for those using CEMS. EPA has well-developed guidelines regarding the use of CEMS, and they are relied on by industry and EPA alike to provide accurate emission information. Requiring annual testing and parametric monitoring on top of CEMS would be duplicative and unnecessary. 2. Use of Sewage Sludge Content Monitoring EPA requested comment on whether facilities should be allowed to demonstrate compliance by monitoring the content of sewage sludge entering the SSI unit. See 75 Fed. Reg. at 63281. NACWA strongly encourages EPA to adopt content monitoring of sewage sludge as an alternative to annual testing or continuous emission monitoring for all pollutants for which a correlation can be established between emissions and sludge content. EPA’s Office of Water regulates the average daily sludge content of cadmium, lead and mercury to ensure that SSI emissions stay below the health-based standards set under Part 503. See 40 CFR § 503.43. Part 503 requires a stack test to set the control efficiency of the control device for each metal. POTWs monitor the sludge feed and the moisture content and use this data to calculate the average daily sludge feed rate in dry tons per day. POTWs also calculate a monthly average concentration of each pollutant based on all the sludge samples taken in a month. EPA should offer this approach to demonstrating compliance as an alternative to the PM surrogate limit in the Proposed Rule. The Part 503 sampling procedure is something NACWA members are familiar with and it will decrease the burden associated with complying with this new rule. Sludge sampling is a cost-effective way for units already regulated under Part 503 to demonstrate compliance with mercury emission limits, and is significantly less burdensome than installing and maintaining CEMS or performing annual stack tests. NACWA also supports using content monitoring for other pollutants (SO2, NOx, HCl) for which a correlation can be established between the content of the sewage sludge and the incinerator emissions. For instance, SO2 stack testing with concurrent sulfur content monitoring can be used to establish a correlation between sludge concentration and emission rate. After that correlation is reliably established, stack testing would no longer be necessary to demonstrate compliance. The POTW could demonstrate compliance by monitoring the sulfur content and, if a control device is used, by monitoring an operating parameter that ensures proper control device operation. NACWA strongly encourages EPA to include a sewage sludge content- monitoring option in the final rule. D. Use of Prior Performance Tests for Initial Compliance Demonstration NACWA supports allowing initial compliance demonstrations to be based on earlier performance tests conducted prior to the rule if they represent current operating conditions and used the appropriate test method. EPA should not impose an arbitrary two-year cut-off period for the initial performance test. Any test that meets the above criteria should be accepted, including all emissions testing conducted for the ICR. E. Less Frequent Testing Based on Test Results NACWA supports less frequent emission testing when test results demonstrate that a source’s emissions are less than 75 percent of the applicable emissions limits. However, as explained in Part VII.A, supra, the initial testing frequency should not exceed one test per permit term. The content of sewage sludge is already regulated NACWA Comments on Docket ID No. EPA-HQ-OAR-2009-0559 November 29, 2010 Page 42 by the CWA to risk-based levels, and additional annual testing is not necessary to reduce health risks from SSIs. If EPA seeks to require more frequent testing, then any final rule should provide for less frequent testing when the results are sufficiently below the emission limit. F. A Waste Management Plan Would Be Duplicative of the Waste Management Plans Already Required by Part 503 NACWA supports EPA’s conclusion that requiring waste management plans under § 129 would be duplicative of the waste management practices already in place under the CWA. The numeric emission limits and management practice requirements established under the Part 503 regulations are based on one of the Agency’s largest risk assessments, which was conducted in the late 1980s and early 1990s to protect human health and the environment from any reasonably anticipated adverse effects from pollutants that may be present in sewage sludge.22 As a result, SSIs demonstrate that the emissions from their units are not adversely impacting human health and the environment by demonstrating compliance with the Part 503 requirements. Pursuant to 40 CFR Part 403 (Part 403), POTWs additionally implement, through local regulatory authority, pretreatment standards to prevent discharge of pollutants to the POTW that may pass through or interfere with treatment processes. Pretreatment reduces harmful constituents in the sewage sludge combusted by incinerators. Pretreatment has dramatically reduced the contaminants in sewage sludge and accordingly emissions from SSIs have become cleaner. Comparison of the sewage sludge quality measured in the 1980s23 with the measurements in EPA’s 2006-2007 Targeted National Sewage Sludge Survey shows a clear improvement in sewage sludge quality since Part 403 and Part 503 were implemented. Specifically, NEORSD, which serves the City of Cleveland and 61 suburban communities, has seen significant decreases in the concentrations of heavy metals in both its influent, attributable to the Part 403 regulations, and its effluent, attributable to both Part 403 and Part 503.24 Between 1980 and 2004, NEORSD has seen the concentration of lead in the influent reduced by 95 percent while the concentration of lead in the effluent was reduced by 100 percent.25 POTWs are already employing successful management practices to comply with Part 403 and Part 503 under the CWA. Requiring additional management plans under CAA § 129 would be duplicative of these efforts and needlessly waste resources that can be better allocated to complying with other provision of the Proposed Rule. G. Emission Limits Appropriately Apply Only When Sludge Is Being Fed to the Incinerator NACWA supports an approach that makes emission standards applicable only when an incinerator is being charged. Since an incinerator may only be regulated under § 129 if it is combusting a solid waste, it is appropriate to limit the application of the emission limits to the periods when a solid waste is being combusted. This may also simplify compliance during startup and shutdown, because the sludge is not being fed to the 22 EPA, A Guide to the Biosolids Risk Assessments for the EPA Part 503 Rule, at 107 (1995) (“[T]he risk assessments quantitatively identified allowable concentrations or application rates of pollutants in biosolids that are used or disposed that protect human health and the environment from reasonably anticipated adverse effects.”). 23 See EPA, 40 City Study (1982); EPA, National Sewage Sludge Survey (1988). 24 See Lita Laven, Frank Foley and Robert Dominak, Improvements in Biosolids Quality Due to EPA’s Pretreatment and Biosolids Programs, Residuals and Biosolids Management Conference 2006, at 142-147 25 Id. at 147, Tbl. 1 NACWA Comments on Docket ID No. EPA-HQ-OAR-2009-0559 November 29, 2010 Page 43 combustion chamber during typical startup and shutdown sequences. Most malfunctions are addressed by promptly shutting off the feed to the incinerator. * * * Again, EPA’s Proposed Rule, if finalized, would have a significant negative impact on the POTWs that rely on incineration and further constrain the list of available sludge management options. NACWA encourages EPA to take the time now to develop a more thorough and accurate understanding of this sector before finalizing emission standards for SSIs. We appreciate your consideration of these comments. Please do not hesitate to contact Chris Hornback at (202) 833-9106 if you would like to discuss these issues further. Sincerely, Ken Kirk Executive Director cc: James Hanlon, Director, Office of Wastewater Management, OW, EPA Ephraim King, Director, Office of Science and Technology, OW, EPA Suzanne Rudzinski, Acting Director, Office of Resource Conservation and Recovery, OSWER, EPA Susmita Dubey, Office of General Counsel, EPA Amy Hambrick, Sector Policies and Programs Division, OAQPS, EPA Attachment A: Compendium of NACWA Correspondence with EPA Attachment B: 2009 WERF Report, Minimizing Mercury Emissions in Biosolids Incinerators Attachment C: City of Palo Alto, California – Mercury Source Control Cost Memorandum Attachment D: NEORSD Mercury Control Cost Memorandum ATTACHMENT A Compendium of NACWA Correspondence with EPA (Submitted to the Docket Separately) ATTACHMENT B 2009 WERF Report, Minimizing Mercury Emissions in Biosolids Incinerators (Submitted to the Docket Separately) ATTACHMENT C City of Palo Alto, California – Mercury Source Control Cost Memorandum Technical Memorandum 1 2010-11-16 Cost of Hg Diversion Through Dental Office Source Control For: Brad Eggleston Palo Alto RWQCP Prepared by: Thomas Barron, PE Reviewed by: Karin North Date: November 16, 2010 Summary The Palo Alto RWQCP source control program for dental offices has been active for the past decade. This cooperative effort with the local dental community has been successful in diverting an estimated 5.1 kg/yr from being discharged into the sanitary sewer system. For the entire service area the program is estimated to have an annual cost in the range of $100,000, which consists mostly of expenses for maintaining amalgam separator units. This range in annual costs is equivalent to $18 million per ton of mercury diverted from the sewer system. Estimate Calculations Step 1 - Number of Dental Offices The RWQCP has 135 dental practices in its service area that perform amalgam placement or removal procedures. These practices are served by 116 vacuum systems that each have an amalgam separator installed. RWQCP Dental Practices & Amalgam Separators 135 Dental offices in the service area that perform amalgam procedures -19 Less practices that use a shared vacuum system and amalgam separator provided by another office, or by the landlord 116 Net number of amalgam separators Technical Memorandum 2 2010-11-16 Step 2 - Amount of Hg Diverted In its 2008 evaluation of waste mercury sources, the RWQCP estimated that dental practices in the service area have successfully diverted 5.1 kg/yr of this metal from the sewer system. This diversion is equivalent to 0.0056 tons per year. Step 3 - Cost to Purchase & Install A Separator In 2008 US EPA published a comparative study of amalgam separators, including their effectiveness and costs (See EPA-821-R-08-014). Data from Table 5-4 of the EPA report are used here, together with information on installation costs from local sources. Annualized Amalgam Separator Costs Step 4 - Annual Maintenance Costs Local dental practices indicate that they service their amalgam separators once every 12 to 18 months, with the frequency depending upon the number of practitioners served by the unit and the amount of work that these dentists do. The following table presents estimated total costs for separators that are serviced once per year. This annual total includes: equipment amortization for 10 years; typical annual maintenance costs of about $600; and 1/8th of a FTE for RWQCP staff to conduct on-going inspections. Costs for maintaining the Solmetex Hg5 amalgam separator are used here because this unit is the most common in the service area. $934 Cost to purchase an amalgam separator (i.e., a Solmetex Hg- 5, which is the most common unit in the RWQCP Service Area) $187 Cost to install the separator (20%) $1,121 Total installed cost (2008 $$) 10 Anticipated useful life of the separator, yrs $112 Annualized Cost to Purchase & Install a Separator Technical Memorandum 3 2010-11-16 Cost per Ton of Diverted Hg (If All Separators Are Serviced Once Per Year) $605 Maintenance Service at 12 month intervals ($$/yr) $112 Annual Cost to Purchase & Install a Separator $717 Total Annual Cost per site $83,181 Total Annual Cost for 116 Sites $15,625 Annual RWQCP Staff Cost $98,806 Annual cost for RWQCP Service Area $18 Equivalent cost per ton of Hg diverted ($$ Million/ton) ATTACHMENT D NEORSD Mercury Control Cost Memorandum NEORSD Preliminary Cost Estimate of Complying with the Proposed SSI Rule ************************************************************* The following is the NEORSD’s preliminary estimate of costs to procure and install additional air pollution control devices that would be required to achieve compliance with the proposed SSI rule. 1. Reduction in Particulate Matter and Metal Emissions A. Advanced Scrubbing Systems In order to reduce particulate matter and some of the particulate based cadmium, lead, and mercury, an advanced scrubbing system might be required. Recent bids for the NEORSD’s Southerly WWTP’s new fluidized bed incineration project, the procurement cost of each of the new “Venturi Pak” advanced scrubbing system was $0.75 million. Assuming $0.5 - 1.0 million for delivery and installation costs, related instrumentation, controls and electrical modifications, and engineering related services, the total cost of a single unit installed would be in the range of $1.25 - $1.75 million/unit (2009 $). This does not include the cost of additional space in a new facility and nor the cost to renovate or expand an existing facility. B. Wet Electrostatic Precipitators (ESPs) If wet electrostatic precipitators (ESPs) are required to reduce particulate matter and some of the particulate based metal emissions, costs would be substantially higher than the aforementioned costs for advanced scrubbing systems. In 2001, bids for a new biosolids handling complex at the Metropolitan WWTP in St. Paul, Minnesota which included the installation of three wet ESPs. The bid price for the three units was $2.49 million or $831,000 per unit. The associated design work and the necessary ductwork, piping, structural, instrumentation and controls, electrical, etc., could total $2.6 million ($867,000 per unit). As a result, total costs averaged $1.7 million per wet ESP (2001$). Given substantial increases in corrosion resistant metal costs since 2001, along with a 44% increase in the Engineering News Record’s construction cost index, costs would be $2.5 - $3 million million per wet ESP (2009 $). This does not include the cost of additional space in the new facility nor the cost to renovate or expand an existing facility. For the NEORSD’s Southerly WWTP new fluidized bed incineration project, approximately 15 feet would have to be added onto the new facility to house the wet ESPs and associated equipment and we do not have room at Westerly or Easterly to add this equipment. 2. Reduction in Sulfur Dioxide and Oxide of Nitrogen Emissions A. Sodium Hydroxide To reduce sulfur dioxide and oxides of oxides of nitrogen emissions, sodium hydroxide could be added to the wet scrubbing systems. Costs to design, procure and install a sodium hydroxide addition system at the Southerly WWTP were estimated to be $0.5 million for three incinerators, with estimated annual chemical costs of $0.7 million/year. This would result in capital costs of $0.25 - $ 0.3 million per incinerator (2009$). This does not include the cost of additional space in a new facility and nor the cost to renovate or expand an existing facility. B. Urea or Ammonia To reduce oxides of nitrogen emissions, either urea or ammonia would be added to the hot gases leaving the incinerators. Cost to design, procure and install a urea injection system at the NEORSD’s Southerly WWTP has been estimated to be $ 5 million, with annual operating & maintenance costs of $0.5 million (2013$). This could result in capital costs of $1.5 - $2 million per incinerator. This does not include the cost of additional space in a new facility and nor the cost to renovate or expand an existing facility. The procurement of an ammonia injection system at the Metropolitan WWTP in St. Paul, Minnesota in 2001 had a bid price approximately $400,000/incinerator (2009$). However, the unit is not being used due to the low oxides of nitrogen emissions from the plant’s three fluidized bed incinerators. It should also be noted that the use of urea and ammonia results in the conversion of the oxides of nitrogen to nitrous oxide (N2O) a greenhouse gas that has a global warming potential of 310. This means that each ton of N2O emissions is equivalent to 310 tons of CO2 emissions. 3. Reduction in Mercury Emissions A. Activated Carbon Adsorbing System In 2005, the Ypsilanti, Michigan WWTP installed a new fluidized bed incinerator equipped with a proprietary activated carbon adsorbing system to reduce mercury emissions. Ypsilanti’s Consultant believes that this unit will also reduce dioxin emissions. The cost to procure and install the system was roughly $3 million. This would result in capital costs of $3.5 - $4 million per incinerator (2009$), and does not include a number of other costs which could easily could increase the costs to $ 5 - 6 million per incinerator. In addition, the activated carbon will have to be replaced once every three years. B. Activated Carbon Injection System followed by a Baghouse EPA as estimated that the total cost to add simple activated carbon inject for the 163 multiple hearth incinerators will be roughly $5 million. As previously reported, the cost is unrealistic and an entire activated carbon injection system, consisting of an activated carbon injection system, carbon contact chamber, baghouse and heat exchangers and boilers to reduce the exhaust gas temperature to below 350 deg-F prior to the injection point. In 2001, bids were received for an activated carbon injection system/baghouse for each of the three new fluidized bed incinerators located at the Metropolitan WWTP in St. Paul, Minnesota. The bid price for the three systems, including procurement, installation, appurtenances and engineering related services, was $6 million or $2 million per unit. Given substantial increases in corrosion resistant metal costs since 2001 and a 44% increase in the Engineering News Record’s construction cost index, costs in 2009$ would be approximately $3 million per system. This would result in capital costs of $3.5 - $4.0 million per system. This does not include the cost of additional space in a new facility, the cost to renovate or expand an existing facility, the cost to revise the air pollution control train, nor the cost to procure and install heat exchangers and/or boilers to reduced the exhaust gas temperature prior to the activated carbon injection point below 350 deg-F. Taking all of this into consideration, the actual cost for all of the require improvements could be $8 - 10 million/incinerator or higher. It should be noted that the $3.5 - 4 million/incinerator for a carbon injection system, carbon contact chamber and baghouse, or the carbon adsorber as contained in the docket for the proposed SSI rule (document # 2009-0059-0015), does not address the additional costs listed above and need to be updated. 4. Reduction in Carbon Monoxide Emissions Carbon monoxide emissions can be reduced by utilizing either internal or external afterburners. However, this action, which requires the burning of substantial quantities of fossil fuel, will result in a substantial increase in oxides of nitrogen emissions and greenhouse gas emissions. While the use of internal afterburners will result in reduced carbon monoxide emissions, most if not all of the POTWs that practice incineration will be required to install external afterburners to meet the potential MACT Standards. It has been estimated that the cost to procure and install an external afterburning system, along with all appurtenances, engineering services and building space, will be in the range of $3 - 4 million per unit. This does not include the cost to renovate or expand an existing facility. 5. Reduction in Hydrogen Chloride At this time, we do not know of any method to reduce hydrogen chloride emissions from sewage sludge incinerators, which are negligible. Clean Bay Pollution Prevention Plan 2010 The Pollution Prevention Plan for the City of Palo Alto's Regional Water Quality Control Plant February 2010 For More Information Additional program information is available at the Environmental Compliance Division web site: www.cityofpaloalto.org/depts/pwd/compliance. Questions about this document should be directed to the Regional Water Quality Control Plant, 2501 Embarcadero Way, Palo Alto, CA 94303, (650) 329-2598, cleanbay@cityofpaloalto.org. Acknowledgements The Clean Bay Plan is produced by the City of Palo Alto's Environmental Compliance Division, and describes the pollution prevention activities of the Industrial Waste and Source Control programs. Key staff members in the development and implementation of this plan include Phil Bobel, Maree Doden, Brad Eggleston, Chris Fujimoto, Brian Jones, Jay Kim, Karin North, Jim Stuart, Ken Torke, Julie Weiss, and Margaret Zittle. 4-1 Section 4: Mercury Program Mercury, a 303(d)-listed pollutant for San Francisco Bay, is present in both wastewater and stormwater discharges to the Bay. Figure 4-1 presents the estimate of mercury sources to the San Francisco Bay, based on the Regional Board’s September 2004 TMDL Proposed Basin Plan Amendment and Staff Report (Regional Board, 2004). Bed Erosion and the Central Valley Watershed are the two largest source of mercury to the San Francisco Bay, with the single largest contributor being the historic legacy of abandoned mercury mines. Bed Erosion (38%) Central Valley Watershed (36%) Urban Storm Water Runoff (13%) Direct Atmospheric Deposition (2%) Guadalupe River Watershed (8%) Non-urban Storm Water Runoff (2%) Wastewater (1.6%) Figure 4-1. Estimated Contributions of Mercury to the San Francisco Bay 4-2 A. Wastewater Sources Sources of mercury discharge to wastewater include laboratories, hospitals, dental offices, human waste, food waste, and stormwater inflow. Since 1997, the RWQCP has quantified the relative importance of mercury sources using local sampling information in conjunction with data from other wastewater treatment plants and the scientific literature. As noted later in this chapter, the RWQCP’s dental amalgam program, which required dental offices to install amalgam separator devices in 2005, has significantly reduced mercury loadings from dental offices. Figure 4-2 presents the mass loading estimates as a percent of the total influent load. The mercury loading estimate was updated in 2008 to reflect the decreased contribution from dental offices (“Mercury Headworks Analysis for 2008”, Barron, February 2009). The estimate also accounts for new information about decreases in amalgam procedures at dental offices and the greater importance of mercury in human waste resulting from dental amalgam swallowed by patients during amalgam placement and removal. For comparison, the mercury loading estimates for 2000 and 2008 are both provided in the Mercury Program Evaluation section. Dental Offices, 29.2% People w/ Amalgam Restorations, 51.4% Foods & Products, 8.8% Industrial & Commercial Dischargers, 3.8% Infiltration / Inflow, 6.4% Water Supply, 0.4% Figure 4-2. RWQCP Mercury Loading Estimate, 2008 4-3 B. Stormwater Sources The largest stormwater source in the South Bay is abandoned mercury mines. Other contributors in the watershed include: • Mobile combustion • Stationary combustion • Fluorescent lamps • Construction erosion • Natural erosion and reservoir spills C. Residential Pollution Prevention Plan Table 4-1 summarizes the RWQCP’s current efforts and evaluation criteria used to reduce mercury discharges to both the sanitary sewer and storm sewer. Mercury program highlights are described below. 1. Residential Fluorescent Lamp Recycling The RWQCP initiated a pilot fluorescent light- recycling program at local hardware stores in 2000. In 2004, Santa Clara County adopted and expanded Palo Alto’s fluorescent light recovery program county-wide. In that same year Palo Alto also began collecting fluorescent lamps at its recycling center from Palo Alto residents. The quantity of bulbs collected each fiscal year is tracked and reported in the Residential Mercury Program Evaluation section. RWQCP partner cities Mountain View and Stanford cannot establish fluorescent collection at their recycling centers due to either space constraints or concerns about universal waste storage at unmanned sites. 2. Universal Waste and Producer Responsibility RWQCP staff works with the California Product Stewardship Council (CPSC) to address issues of Extended Producer Responsibility (EPR). The goal of CPSC is to work state-wide with water quality, solid waste, and hazardous waste programs to develop partnerships and/or legislation to change the current system in which local government is tasked with the financial and operational burden of collecting electronic, pharmaceutical, and other wastes. Instead, EPR would involve manufacturers and retailers to reduce the public and environmental risks associated with their products and packaging. 4-4 Table 4-1. Mercury Pollution Prevention Plan Source Audiences Message / Program Implementation Plan & Timeline Evaluation Criteria ALL AUDIENCES Air deposition of mercury Air Quality District Reduce amount of mercury air emissions Continue to urge the Bay Area Air Quality Management District to enact strict controls on atmospheric releases of mercury for the expressed purpose of meeting water quality objectives in the State’s natural waterways. Number of correspondences with air district staff GOVERNMENT AGENCIES Fluorescent lights Purchasing and Utilities Departments Continue recycling fluorescent lights and purchase low- mercury bulbs Continue to recycle spent bulbs and to purchase low mercury replacement bulbs. Collaborate with the California Product Stewardship Council to effect legislation requiring extended producer responsibility to cover universal waste disposal costs and operations. Annual confirmation that low mercury bulbs are being purchased Progress towards legislation. Mercury- containing thermo- meters, switches and manometers Facilities, Utilities, and Wastewater Treatment Operations Identify and replace mercury- containing switches and manometers within City operations Continue to replace mercury-containing thermometers, switches, and manometers with non-mercury alternatives upon failure. Number of locations or uses for which alternative switches are identified. Chapter 4-1. Mercury Pollution Prevention Plan (continued) 4-5 Source Audiences Message / Program Implementation Plan & Timeline Evaluation Criteria COMMERCIAL/INDUSTRIAL Mercury- containing thermostats Building and HVAC contractors Recycle mercury- thermostats. Install non- mercury thermostats Following up on the passage of AB2347, which requires mercury-containing thermostat wholesalers to provide collection effective April 1, 2009, staff will review the local status of this effort and assess next steps. Identification of program reach in RWQCP service area: where are drop off locations, extent of coverage Fluorescent lights Facility managers Recycle fluorescent lamps. Purchase low- mercury lamps. Via SCVURPPP participation, provide outreach and assistance regarding Universal Waste Rule (UWR) requirements for conditionally exempt small quantity generators. Continued participation in SCVURPPP project Scrap amalgam, chairside trap waste, filter waste, and amalgam sludge Dentists, dental assistants and dental hygienists Reduce dental mercury discharges through amalgam separator technology and best management practices Continue implementation of dental amalgam program, including annual self-certification forms and inspection of dental offices to confirm proper amalgam separator maintenance and BMP compliance. Continue to distribute BMP information to community college dental assistant classes. In 2009, 77 students were reached in a dental assistant class and in a dental hygienist class. Receipt of annual self- certification forms, confirmation of amalgam separator maintenance and BMP compliance through inspections. Continued tracking of treatment plant mercury data Number of students reached Chapter 4-1. Mercury Pollution Prevention Plan (continued) 4-6 Source Audiences Message / Program Implementation Plan & Timeline Evaluation Criteria RESIDENTS Consumer products: Mercury- containing thermostats, Thermo- meters, and other consumer products Homeowner, landlords and renters Recycle mercury- containing products Continue to collect and recycle thermometers and thermostats through the RWQCP dropoff program, HHW programs and special events. In 2010, organize and publicize a service area-wide thermometer takeback campaign. Partner with regional take-back efforts as they arise and CPSC on implementing producer responsibility programs. Quantity of thermometers, thermostats, and other mercury-containing products collected Progress of CPSC and contributions of RWQCP staff time. Participation in program. Consumer products: Fluorescent lights Homeowner, landlords and renters Recycle fluorescent lights Monitor and attempt to assist with any funding issues experienced by the County HHW Program to ensure that the County-wide fluorescent lamp collection program continues to serve the RWQCP’s entire service area. Continue to collect and recycle fluorescent lamps accepted at the recycling center and at HHW events. Continuation of County-wide program Quantity of lamps collected at recycling center and HHW events. 4-7 To date, the City of Palo Alto has taken a leadership role in promoting EPR by contributing to CPSC financially, organizing and providing presentations to the water quality community and initiating a 2008 meeting with Assembly member Ira Ruskin to ask for his advocacy for both EPR and CPSC (he later successfully sponsored AB2347 which required wholesalers of mercury-containing thermostats to set up a collection program for mercury thermostats by April 1, 2009). In 2009, RWQCP partnered with the Palo Alto Zero Waste Program and adopted an EPR resolution and policy to further its efforts in incorporating EPR into its operations. RWQCP leads the City’s Green Purchasing efforts and has begun the process of incorporating EPR language for both products and packaging. 3. Residential Thermometer and Thermostat Drop-off Program The RWQCP’s thermometer and thermostat takeback program, which began in 1998, remains in operation. The number of thermometers collected annually has decreased significantly since the conclusion of a very successful campaign in 2004. Thermometer and thermostat takeback will continue at the RWQCP facility, monthly Household Hazardous Waste events, and special events upon request. The RWQCP plans to conduct a reenergized, service area-wide thermometer takeback campaign in 2010 that will involve advertising, multiple locations and tracking of results on the RWQCP’s cleanbay.org website. 4. Residential Mercury Program Evaluation About 98 pounds of mercury, including 8,034 thermometers, have been collected since the program began in January 1998 (Table 4-2). Annual collection varies and can be strongly influenced by residents that occasionally bring in several pounds of bulk mercury at one time. The mass of mercury collected can be compared to the RWQCP’s annual influent loading of approximately 13 pounds. Below is a summary of the amount of mercury products collected at the RWQCP (Table 4-2) and the HHW Program (Table 4-3). Table 4-2: Mercury Products Collected at RWQCP 1 Collected in 2009 Total collected to date (since 1998) Thermometers 34 units 8,034 units Thermostats 2 units 301 units Total weight of mercury collected (includes bulk mercury) 0 pounds 77 pounds 1From CPB Mercury_RX__HHW Tracking Form; Calculation of pounds of mercury assumes 1 gram mercury per thermometer and 5 grams mercury per thermostat 4-8 Table 4-3: Mercury Products Collected at Household Hazardous Waste Events 1 Collected in FY 2008-2009 Total collected to date (beginning FY 2007-08)2 Devices (mercury thermometers, thermostats) 89 pounds 305 pounds Fluorescent Lamps 19,445 pounds 38893 pounds Total weight collected 19,534 pounds 39,198 pounds 1 From monthly HHW events and Palo Alto Recycling Center recorded on CIWMB 2008-2009 303a Form 2 Collection at HHW has occurred since program inception. The format of the 303a form changed in 2009. For this report, totals mercury containing device collection will be tabulated from fy 2009. D. Dental Office Source Control Since 2000, the RWQCP has partnered with the Mid- Peninsula Dental Society (MPDS) to educate dentists about their role in preventing mercury contamination in San Francisco Bay. RWQCP prepared training materials, conducted training seminars, made on-site consultations, and defined voluntary best management practices (BMPs) to prevent mercury amalgam from reaching the environment. In 2004, the RWQCP adopted an ordinance requiring all owners and operators of dental vacuum suction systems to install an ISO 11143 certified amalgam separator device by March 31, 2005. The ISO 11143 certification process for amalgam separators includes demonstrating that the separator removes a minimum 95% of amalgam from the wastestream. In addition to the amalgam separator requirement, the following best management practices (BMPs) that directly affect sewer discharges were required by the ordinance: • Do not rinse chairside traps, vacuum screens, or amalgam separator equipment in a sink or other sanitary sewer connection. • Train staff in the proper handling and disposal of amalgam materials and fixer-containing solutions; training records are available for inspection. • Do not use bleach or other chlorine-containing disinfectants to disinfect the vacuum line system. • Do not use bulk liquid mercury; use only precapsulated dental amalgam. • Store amalgam waste in accordance with recycler or hauler instructions. Other amalgam-related BMPs, though not required by the ordinance, are encouraged. 1. Dental Program Implementation Based upon the information collected by the RWQCP in 2004, it was determined that 134 dental offices in the RWQCP’s service area were required to install amalgam separators. A small 4-9 number of dental offices are exempt from the ordinance because they specialize in fields of dentistry in which amalgam removal or placement is uncommon, and the removal or placement of amalgam fillings occurs at their facility no more than 3 days per year. By late 2006, each of the dental offices subject to the amalgam separator requirements had been inspected at least once. A typical inspection lasted approximately 20 minutes, and addressed the following issues: • Implementation of amalgam best management practices (BMPs); • Presence of an amalgam separator; • Review of separator maintenance records; and • Management of amalgam wastes and x-ray processing wastes. In 2007, the RWQCP developed a new database to help track dental offices and dentists within our service area. The dental database is complex because many amalgam separators are shared by multiple dental practices or managed by a landlord or property management company. The database also tracks when the separator was last maintained from the data collected on the annual amalgam separator form. RWQCP staff sends an amalgam self-certification form to the dental offices, which requires each dental office to self-certify that the office is maintaining its amalgam separator and complying with the required BMPs. The dentists complete this form annually, assisting the RWQCP in tracking new business information and prioritizing inspections. 2. Ongoing Dental Program With all of the dental offices subject to the ordinance having installed amalgam separators, the primary goals of the ongoing program are to verify that amalgam separators are properly maintained, maximize compliance with BMPs, and ensure that new dental offices are captured by the program. The key components of the ongoing dental amalgam program are as follows: • Track new dental offices through the building permit process and ensure that amalgam separators are installed • Require submittal of annual report forms from all dental offices to allow self-certification of BMP compliance and amalgam separator maintenance • Inspect dental offices to verify compliance and to educate dentists • Maintain database using information obtained through inspections and annual report forms • Maintain web site that includes electronic versions of the brochure, forms, ordinance text, list of approved separators (linked to list maintained by Bay Area Pollution Prevention Group) and answers to frequently asked questions. In February 2009, RWQCP staff sent amalgam self-certification forms to 131 dental offices. Based on the response, the RWQCP inspected 25 dental offices to confirm that they are following the required BMPs and maintaining their amalgam separators. Staff also inspected those offices with new owners to ensure compliance with the ordinance requirements. In 2010, RWQCP staff will look for dentists that may not be listed in our database to ensure compliance with the dental ordinance requirements. 4-10 3. Dental Program Evaluation a. Dental Facilities in Compliance with Ordinance In 2009, RWQCP staff confirmed full compliance with separator maintenance requirements through a combination of annual report forms and inspections. Inspections have also confirmed that dental offices are following the BMPs, although a small number of dental offices have been found to continue to place chairside traps in the trash. In these instances, the inspection is used as an opportunity to educate the office and a follow-up letter is sent reminding the office that following BMPs is required by the ordinance. As discussed previously, inspections confirmed that all dental offices have installed an amalgam separator. b. Decrease in estimated loading from dental facilities The RWQCP’s mercury loading estimate was updated in 2008 to reflect lower contributions from dental facilities after implementation of the dental amalgam program. Figures 4-3 and 4-4 illustrate the dramatic reduction in estimated mercury loading from dental offices between 2000 and 20081. The dental office contribution is estimated to have decreased from 56 percent to 29 percent of the RWQCP’s influent mercury loading. During this same period of time, the overall influent mercury loading has decreased by approximately one-third. The methodology employed to estimate the annual mass of mercury discharged to the sanitary sewer for individual dentists can also be used to estimate the mass of mercury captured (e.g. collected for appropriate disposal by amalgam separators and implementation of Best Management Practices) due to the RWQCP’s dental amalgam program. The 2008 analysis estimates that the current loading to the sewer from dental offices in the Palo Alto service area is 5.1 pounds per year, with 47.4 pounds per year captured by use of best management practices and amalgam separators. It is estimated that the dental loading would be 16.3 pounds per year if the dental amalgam program had not been implemented. Therefore, the RWQCP estimates that 11.2 of the total 47.2 pounds captured or collected by dentists is attributable to the dental amalgam program. 1 The 2000 Mercury Loading Estimate displayed in Figure 4-5 differs from the 2000 Estimate provided in previous Clean Bay Plans. As the 2008 Estimate was in process, it was noted that the original 2000 Estimate used a lower number of dentists than the number now known to be practicing in the RWQCP service area. In addition, the “People with Amalgam Restorations” estimate had been understated because excretion of mercury swallowed during dental procedures had not been considered. For details, see “Mercury Headworks Analysis for 2008”, Barron, February 2009. 4-11 Dental Offices, 56.3%People w/ Amalgam Restorations, 30.8% Foods & Products, 4.9% Industrial & Commercial Dischargers, 5.4% Infiltration / Inflow, 2.3% Water Supply, 0.2% Figure 4-3. RWQCP Mercury Loading Estimate, 2000 Dental Offices, 29.2% People w/ Amalgam Restorations, 51.4% Foods & Products, 8.8% Industrial & Commercial Dischargers, 3.8% Infiltration / Inflow, 6.4% Water Supply, 0.4% Figure 4-4. RWQCP Mercury Loading Estimate, 2008 4-12 4. Educating Other Bay Area Wastewater Agencies Since December 2006, staff from the RWQCP spearheaded and helped organize three dental amalgam training sessions that were co-sponsored by the Bay Area Pollution Prevention Group (BAPPG). The overall goals of the workshops were to provide wastewater agencies with the tools and knowledge to start their own dental amalgam programs. The most recent training was held in January 2010, which was attended by 55 inspectors and program managers. Both trainings were well-received and participants now feel more knowledgeable when they inspect a dental office. 5. Educating Dental Hygienists and Dental Assistants Palo Alto has partnered with the Foothill College Dental Hygienist program to educate new hygienists about minimizing mercury pollution from dental offices. In 2009, BAPPG decided to expand this outreach to educate all dental hygienists and dental assistants in the Bay Area. Most recently, Palo Alto educated 77 dental hygienists at Foothill College. In 2009, BAPPG’s contractor also educated over 400 students throughout the Bay Area. BAPPG also created an outreach page for dentists and dental hygienists located on the baywise site. In 2010, BAPPG will continue to fund a contactor to educate dental hygienists and assistants on ways to reduce pollution from dental offices. E. POTW Mercury Use Investigation In 2003, RWQCP conducted a thorough investigation of the uses of mercury at the Plant. The investigation included the following: • An inventory of the mercury used at the Plant, focusing on switches and reagents, but including all uses • A listing of available non-mercury alternative products • Recommendations and a time schedule for action by Palo Alto where appropriate The inventory is presented in Table 4-4, organized into three separate sections: lights, switches, and lab reagents. The list includes currently inventoried mercury-containing products, as well as the plan for action, as appropriate. 4-13 Figure 4-5. Some of the mercury alarm switches that were removed from service or pulled from storage at the RWQCP in 2002. (All such switches are turned in to the City’s municipal hazardous waste collection for recycling.) In 2005, the RWQCP reviewed these results and found no additional mercury-containing components to add to the list. If additional sources are identified in the future, the RWQCP will review applicability of alternative products and recommend any additional actions. As for previously identified products, the schedule is presented in Table 4-4. The primary items for which alternatives have been identified are mercury switches (Figure 4- 5). In those cases, because there is no immediate pathway between the switches and the environment, the schedule involves removing mercury switches upon failure. F. POTW Mercury Replacement Effectiveness Effectiveness of this program is measured by: • Numbers of switches collected and replaced • Adoption of a mercury-free alternative to the COD laboratory procedure • Further identification of other sources and alternatives 1. RWQCP mercury replacement Since 2002, RWQCP has replaced or removed 52 mercury switches (approximately 3.1 pounds) from the plant. Many of the switches that were recycled were not in service but were inventory in the shop and warehouse contributing to the large amount of mercury recycled in a short period of time. In 2010, RWQCP replaced one float switch with contained approximately 50 grams of mercury. In 2008, the RWQCP’s laboratory switched to a mercury-free Chemical Oxygen Demand (COD) laboratory method following the method’s approval by EPA. The RWQCP laboratory typically analyzes approximately 350 samples per year for COD. The actual number of COD vials used is greater than 350, as many are analyzed in duplicate and quality control samples must also be analyzed. 4-14 2. Identification of sources and alternatives City staff identified an alternative float switch, which is a non mercury bulb and is produced by the same manufacturers as the mercury switches. For other applications, such as temperature or pressure switches, there are a variety of techniques for replacement, usually by digitizing the signal and using logic to control a switch. In addition the RWQCP uses a Programmable Logic Controller (PLC) system to make contacts when necessary while monitoring a signal. These are usually a case-by-case modification rather than a universal replacement. Other situations arise where the switch is an integral part of an existing system that does not retrofit, such as the torque alarms on the clarifier sweeps. Nothing short of redesigning the sweep arms can replace the mercury switches; these will not be replaced. Table 4-4. Inventory Results and Current Plans Product Use at RWQCP? Plan for Seeking Alternatives Lights Fluorescent Yes – common Now using low mercury lights – formally initiated in 2001 Metal halide (high intensity discharge lamps w/blue-white light) None found High pressure sodium (yellow-white lights used for street lamps, outdoor security, and some lighting in process areas) Yes – common In 2010, plan to evaluate potential for replacing these lamps with LED lighting Mercury vapor lamps Yes – uncommon use (2-3 only) No current plan Switches “Silent” wall switches (prior to 1991) None found Will replace any overlooked upon failure Wall-mounted office thermostat Yes – 6 to 10 Will replace upon failure Airflow/fan limit controls None found Building security systems None found Pressure control None found Will replace, upon failure, any overlooked devices and replace with Hg-free Torque-arm alarm switches Yes – about 12 to 16 Will not replace; integral part of existing torque arm Float control (used for sump pumps) Yes Will replace with Hg-free upon failure. Lab Reagents Pre-mixed test tubes for monitoring chemical oxygen demand (COD) Yes In 2008, the laboratory switched to a mercury free COD test procedure. G. Mercury Program Evaluation The best measure of the mercury program’s overall effectiveness is the mercury concentrations in influent, effluent and biosolids; figures 4-9 through 4-14 present 4-15 RWQCP data on influent and effluent mercury concentrations and mercury concentrations in sludge cake prior to incineration. The dental amalgam program has proven very effective in reducing mercury loadings to the Plant and to the environment; the average effluent concentration for 2006 through 2009 declined 49 percent compared to the average effluent concentration for 2001 through 2004 (Table 4-5). These reductions are believed to result primarily from the dental amalgam program, not from other residential and commercial programs addressing mercury, because they occurred after the takeback programs for thermometers, thermostats, and fluorescent lamps had been in place for a number of years. Table 4-5. Summary of the average mercury concentration before and after the installation of amalgam separators for influent, sludge cake and effluent Average Mercury Concentration Location 2001-2004 (pre-amalgam separators) 2006-2009 (post-amalgam separators) Percent Reduction (pre- vs. post- separators) Influent 0.31 µg/L 0.20 µg/L 37% Sludge Cake 397 µg/kg 248 µg/kg 37% Effluent 0.006 µg/L 0.003 µg/L 49% As shown in Figures 4-6 and 4-7, influent mercury concentrations have decreased significantly since 2004. Influent samples are collected on a weekly basis. Since the ordinance required installation of amalgam separators by March 31, 2005, data from 2001 through 2004 are used as the “pre-separator” baseline. The average influent concentration for 2006 through 2009 was 37 percent lower than the average influent concentration for 2001 through 2004. With respect to influent data gathering, there is a data gap that must be acknowledged. Mercury amalgam particles are very dense, and are more likely to travel along the bottom of sewer pipes than to be suspended throughout the water column. Due to the configuration of the RWQCP influent sampling location, it is believed that some mercury amalgam particles would not be captured in the influent samples. These amalgam particles would not enter the RWQCP’s primary treatment process, but would be trapped in the “grit” channel at the headworks of the RWQCP. Because the grit is highly irregular, it is not feasible to conduct accurate, representative sampling. The grit is disposed of at a landfill, and its pollutant content has never been included in the various mass balances that have been conducted. Therefore, we recognize that undetected decreases in the mercury content of that material may have occurred. 4-16 Figures 4-8 and 4-9 present effluent mercury concentration data. Effluent samples are collected on a monthly basis. The average effluent concentration for 2006 through 2009 declined 49 percent from the average concentration for 2001 through 2004. The 2009 average effluent concentration is the lowest annual average concentration to date for mercury. As shown in Figure 4-9, the effluent mercury concentration since early 2005 trended steadily downward through early 2006, then upward, then downward again. Since the installation of amalgam separators, the correlation between effluent suspended solids and mercury has become much stronger, and these trends track well with suspended solids removal performance. Mercury concentrations in the RWQCP’s pre-incineration sludge cake are presented in Figures 4-10 and 4-11. Sludge cake samples are collected monthly. Sludge cake mercury concentrations have declined substantially since 2004. The average concentration for 2006 through 2009 is 37 percent lower than the average for 2001 through 2004. 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00 Ja n - 0 1 Ja n - 0 2 Ja n - 0 3 Ja n - 0 4 Ja n - 0 5 Ja n - 0 6 Ja n - 0 7 Ja n - 0 8 Ja n - 0 9 Ja n - 1 0 Date Me r c u r y ( µ g / L ) 6-sample rolling average discrete data points Figure 4-6. RWQCP Mercury Influent Concentrations: 2001-2009 4-17 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 2001 2002 2003 2004 2005 2006 2007 2008 2009 Year Me r c u r y ( µ g / L ) Figure 4-7. RWQCP Average Mercury Influent Concentrations: 2001-2009 0 0.002 0.004 0.006 0.008 0.01 0.012 0.014 Ja n - 0 1 Ja n - 0 2 Ja n - 0 3 Ja n - 0 4 Ja n - 0 5 Ja n - 0 6 Ja n - 0 7 Ja n - 0 8 Ja n - 0 9 Ja n - 1 0 Date Me r c u r y ( µ g / L ) 6-sample rolling average discrete data points Figure 4-8. RWQCP Mercury Effluent Concentrations: 2001-2009 4-18 0.000 0.001 0.002 0.003 0.004 0.005 0.006 0.007 0.008 2001 2002 2003 2004 2005 2006 2007 2008 2009 Year Me r c u r y ( µ g / L ) Figure 4-9. RWQCP Average Mercury Effluent Concentrations: 2001-2009 0 100 200 300 400 500 600 700 Ja n - 0 1 Ja n - 0 2 Ja n - 0 3 Ja n - 0 4 Ja n - 0 5 Ja n - 0 6 Ja n - 0 7 Ja n - 0 8 Ja n - 0 9 Ja n - 1 0 Date Me r c u r y ( µ g / k g ) 6-sample rolling average discrete data points Figure 4-10. RWQCP Sludge Cake Mercury Concentration: 2001-2009 4-19 0 50 100 150 200 250 300 350 400 450 2001 2002 2003 2004 2005 2006 2007 2008 2009 Year Me r c u r y ( µ g / k g ) Figure 4-11. RWQCP Average Mercury Sludge Cake Concentrations: 2001-2009 Technical Memorandum 1 2010-11-16 Cost of Hg Diversion Through Dental Office Source Control For: Brad Eggleston Palo Alto RWQCP Prepared by: Thomas Barron, PE Reviewed by: Karin North Date: November 16, 2010 Summary The Palo Alto RWQCP source control program for dental offices has been active for the past decade. This cooperative effort with the local dental community has been successful in diverting an estimated 5.1 kg/yr from being discharged into the sanitary sewer system. For the entire service area the program is estimated to have an annual cost in the range of $100,000, which consists mostly of expenses for maintaining amalgam separator units. This range in annual costs is equivalent to $18 million per ton of mercury diverted from the sewer system. Estimate Calculations Step 1 - Number of Dental Offices The RWQCP has 135 dental practices in its service area that perform amalgam placement or removal procedures. These practices are served by 116 vacuum systems that each have an amalgam separator installed. RWQCP Dental Practices & Amalgam Separators 135 Dental offices in the service area that perform amalgam procedures -19 Less practices that use a shared vacuum system and amalgam separator provided by another office, or by the landlord 116 Net number of amalgam separators Technical Memorandum 2 2010-11-16 Step 2 - Amount of Hg Diverted In its 2008 evaluation of waste mercury sources, the RWQCP estimated that dental practices in the service area have successfully diverted 5.1 kg/yr of this metal from the sewer system. This diversion is equivalent to 0.0056 tons per year. Step 3 - Cost to Purchase & Install A Separator In 2008 US EPA published a comparative study of amalgam separators, including their effectiveness and costs (See EPA-821-R-08-014). Data from Table 5-4 of the EPA report are used here, together with information on installation costs from local sources. Annualized Amalgam Separator Costs Step 4 - Annual Maintenance Costs Local dental practices indicate that they service their amalgam separators once every 12 to 18 months, with the frequency depending upon the number of practitioners served by the unit and the amount of work that these dentists do. The following table presents estimated total costs for separators that are serviced once per year. This annual total includes: equipment amortization for 10 years; typical annual maintenance costs of about $600; and 1/8th of a FTE for RWQCP staff to conduct on-going inspections. Costs for maintaining the Solmetex Hg5 amalgam separator are used here because this unit is the most common in the service area. $934 Cost to purchase an amalgam separator (i.e., a Solmetex Hg- 5, which is the most common unit in the RWQCP Service Area) $187 Cost to install the separator (20%) $1,121 Total installed cost (2008 $$) 10 Anticipated useful life of the separator, yrs $112 Annualized Cost to Purchase & Install a Separator Technical Memorandum 3 2010-11-16 Cost per Ton of Diverted Hg (If All Separators Are Serviced Once Per Year) $605 Maintenance Service at 12 month intervals ($$/yr) $112 Annual Cost to Purchase & Install a Separator $717 Total Annual Cost per site $83,181 Total Annual Cost for 116 Sites $15,625 Annual RWQCP Staff Cost $98,806 Annual cost for RWQCP Service Area $18 Equivalent cost per ton of Hg diverted ($$ Million/ton)