Daily Rules, Proposed Rules, and Notices of the Federal Government
The information presented in this preamble is organized as follows:
The EPA is proposing amendments to the emissions standards for hazardous air pollutants (HAP) and to the performance standards for Portland cement plants. These proposed amendments respond to petitions for reconsideration filed by the Portland cement industry and to a decision by the United States Court of Appeals for the District of Columbia Circuit (D.C. Circuit). These amendments, which are consistent with the CAA, if adopted, will also provide less costly compliance options and compliance flexibilities, and thereby result in cost savings for the Portland cement industry. This result would also be consistent with Executive Order 13563. The proposed amendments include a new compliance date for the PM, mercury, HCl, and THC existing source standards.
a. Need for the Regulatory Action. The EPA is proposing amendments to the national emission standards for hazardous air pollutants (NESHAP) for the Portland cement source category and to the new source performance standards (NSPS) for Portland cement plants issued under sections 112(d) and 111(b) of the Clean Air Act (CAA). Section 112 of the CAA establishes a regulatory process to address emissions of HAP from stationary sources. After the EPA identifies categories of sources emitting one or more of the HAP listed in section 112(b) of the CAA, section 112(d) requires the EPA to promulgate technology-based NESHAP for those sources. Section 111 of the CAA requires that NSPS reflect the application of the best system of emission reductions achievable which, taking into consideration the cost of achieving such emission reductions, and any non-air quality health and environmental impact and energy requirements, the Administrator determines has been adequately demonstrated.
This proposal addresses the remand by the D.C. Circuit in
The proposal also addresses technical issues with respect to the standard for PM in both the NESHAP and the NSPS that have emerged since these rules' promulgation. We are proposing to amend the standard for PM, and also proposing to amend various implementation requirements in a way that would provide more compliance flexibilities. In addition, the proposal addresses the issues on which the EPA previously granted reconsideration.
b. Legal Authority for the Regulatory Action. These proposed amendments implement sections 112(d) and 111(b) of the CAA.
a. PM (PM) Emission Standards. The EPA is proposing changes to the emission standards for PM that potentially make available compliance alternatives unavailable under the promulgated existing source standards. The promulgated rule requires compliance to be demonstrated using a Continuous Emission Monitoring System (CEMS) (see section 63.1348 (75 FR 55056)). Based on the information the EPA now has, we believe that it may be problematic for a PM CEMS to meet the mandated Performance Specification 11 (PS 11) correlation requirements complying with the promulgated PM standards. (See section III.D.) As a consequence, the EPA is proposing to amend the existing and new source PM standards in the NESHAP to require manual stack testing in lieu of PM CEMS for compliance determinations. An additional consequence of this
b. Response to Remand. Consistent with the court's remand, the EPA has removed all the CISWI kilns from the database used to set the 2010 existing source standards for PM, mercury, hydrochloric acid and total hydrocarbons (THC). The EPA then recalculated existing source floors for each of these pollutants, and determined what standards to propose in light of that analysis. This analysis informed the level of the proposed standards for PM just discussed. The resulting standards are discussed immediately below.
c. Other Emissions Standards. The EPA is not proposing any changes to the existing source standards for mercury, total THC or hydrogen chloride (HCl). The reasons are set out in sections III A, B and C below.
With respect to new source standards, under section 112(d)(3) of the CAA, new source floors can be based on the performance of the “best controlled similar source.” A CISWI cement kiln is a similar source for purposes of this provision. The EPA, therefore, is not proposing to amend any of the new source floors or standards for mercury, THC or HCl where the best performing source in the database used to set the standards was a CISWI cement kiln.
The EPA is also proposing to amend the alternative standard for organic HAP, whereby organic HAP are measured directly. To avoid a situation where the alternative organic standard level is lower than the practical quantitation limit of the relevant analytic methods, the EPA is proposing to increase the alternative organic HAP standard from 9 parts per million (ppm) to 12 ppm. See additional discussion in section III.H below.
d. Standards during Startup and Shutdown. In the final 2010 NESHAP, the EPA established specific numerical standards for startup and shutdown for each pollutant to be measured using a CEMS over an accumulative 7-day rolling average. Because raw materials (the source of most cement kiln air emissions) are not introduced into cement kilns during startup and shutdown, cement kilns' emissions during these periods should be appreciably lower than the level of the standards. The EPA is, therefore, proposing that sources monitor compliance with these standards via recordkeeping.
e. Proposed Compliance Dates. The EPA is proposing that the compliance date for all existing source standards including standards for PM, mercury, HCl and THC, clinker piles and the standards for startup and shutdown be extended to September 9, 2015. The EPA believes that the proposed change to the PM standard makes possible compliance alternatives unavailable under the promulgated existing source standards) and that an extension until September 9, 2015, is the period in which these new compliance strategies can be implemented most expeditiously.
f. The EPA is also taking action on the remaining issues on which it granted reconsideration on May 17, 2011.
The following table 1 summarizes the costs and emissions reductions of this proposed action.
The cost information in Table 1 is in 2005 dollars at a discount rate of 7 percent. The net change in annualized costs in 2015 is a $12.2 million savings compared to the 2010 rule. The EPA did not have sufficient information to quantify the overall change in benefits or costs for 2013 to 2015 that might arise due to the proposed change in compliance dates.
The following Table 2 shows the proposed standards.
Categories and entities potentially regulated by this final rule include:
This table is not intended to be exhaustive, but rather provides a guide for readers regarding entities likely to be regulated by this action. To determine whether your facility will be regulated by this action, you should examine the applicability criteria in 40 CFR 60.60 (subpart F) or in 40 CFR 63.1340 (subpart LLL). If you have any questions regarding the applicability of this final action to a particular entity, contact the person listed in the preceding
Do not submit information containing CBI to the EPA through
In addition to being available in the docket, an electronic copy of this proposal will also be available on the World Wide Web (WWW) through the EPA's Technology Transfer Network (TTN). Following signature by the EPA Administrator, a copy of this proposed action will be posted on the TTN's policy and guidance page for newly proposed or promulgated rules at the following address:
Section 112 of the CAA establishes a regulatory process to address emissions of HAP from stationary sources. After the EPA has identified categories of sources emitting one or more of the HAP listed in section 112(b) of the CAA, section 112(d) requires us to promulgate NESHAP for those sources. For “major sources” that emit or have the potential to emit 10 tons per year (tpy) or more of a single HAP or 25 tpy or more of a combination of HAP, these technology-based standards must reflect the maximum reductions of HAP achievable (after considering cost, energy requirements and non-air quality health and environmental impacts) and are commonly referred to as maximum achievable control technology (MACT) standards.
The statute specifies certain minimum stringency requirements for MACT standards, which are referred to as “floor” requirements. See CAA section 112(d)(3). Specifically, for new sources, the MACT floor cannot be less stringent than the emission control that is achieved in practice by the best controlled similar source. The MACT standards for existing sources can be less stringent than standards for new sources, but they cannot be less stringent than the average emission limitation achieved by the best-performing 12 percent of existing sources (for which the Administrator has emissions information) in the category or subcategory (or the best-performing five sources for categories or subcategories with fewer than 30 sources.
In developing MACT, we must also consider control options that are more stringent than the floor. We may establish standards more stringent than the floor based on the consideration of the cost of achieving the emissions reductions, any non-air quality health and environmental impacts, and energy requirements. See CAA section 112(d)(2).
Section 111(b) requires the EPA to set standards for emissions that “reflect the
The history of this proposed rule, commencing with the 1999 standards and proceeding through the amendments issued in September 2009, is set out in detail in 75 FR 54970 (Sept 9, 2010). Various parties filed petitions for reconsideration of aspects of those amendments. On May 17, 2011, the EPA granted reconsideration of various issues, and denied the petitions to reconsider as to the remaining issues. See 76 FR 28318 (May 17, 2011). On December 9, 2011, the D.C. Circuit issued an opinion upholding the NESHAP itself (as well as the section 111 NSPS), but found that the EPA had arbitrarily failed to grant reconsideration to consider the effect of the EPA's NHSM rule on the standards (76 FR 15456 (March 21, 2011)), which rule had the effect of reclassifying some cement kilns as commercial and solid waste incinerators. See
In this action, the EPA is responding to the court's remand. For existing sources, the EPA is doing so by removing all kilns classified as commercial and industrial solid waste incinerators from the data used to establish the 2010 NESHAP standards. The EPA is then recalculating each of the floors based on this revised dataset and making beyond-the-floor determinations based on the recalculated floors. The EPA believes that this approach is fully responsive to the court's remand. See 665 F. 3d at 188 where the court referred favorably to this type of recalculation. For new sources, the EPA is basing floors on the performance of the best performing similar source.
As just noted, in
Applying that definition, the EPA has determined that there are 24 cement kilns which combust solid waste. See 76 FR 28322 and Memorandum “Combustion in a Cement Kiln and Cement Kilns' Use of Tires as Fuel” (April 25, 2011) (“April 25 memorandum”); see also 76 FR 80452 (Dec. 23, 2011) where the EPA identified 23 of the 24 kilns as commercial incinerators as were identified in the April 25 memorandum. The 24th kiln was identified as a CISWI kiln after development of the April 25, 2011, memorandum, but the addition of this kiln did not affect the calculations contained in the May 17, 2011 notice (CISWI Data Revisions since Reconsideration Proposal, docket EPA-HQ-OAR-2003-0119). Although the EPA has proposed to reconsider certain narrow aspects of the NHSM rule, see 76 FR 80598 (Dec. 23, 2011), this count remains unchanged by any of the issues being considered in the reconsideration of the NHSM rule. This is because either the types of secondary materials being addressed in that reconsideration are not combusted by cement kilns or the EPA has already accounted for those materials in its April 25 memorandum analysis. See 76 FR 28319 (May 17, 2011). Specifically, in the NHSM reconsideration proposal, the EPA proposed to clarify that clean cellulosic biomass and clean construction and demolition wood are not solid wastes when burned for energy recovery and that unused, off-specification tires are not wastes when burned for energy recovery. The EPA's analysis underlying its April 25, 2011, memorandum already reflects that these non-hazardous secondary materials are not wastes when burned by cement kilns for energy recovery. The EPA expects the reconsideration of the NHSM rule to be completed before taking final action on this proposed rule and the EPA will account forany changes resulting from the reconsidered final NHSM rule when it takes final action here.
1. Existing Source Floors. We removed the 24 CISWI kilns from the database used to establish existing source standards and recalculated floors for existing sources. Under this analysis, the existing source floor for mercury increased from 55 lb/million (MM) tons clinker to 58 lb/MM tons clinker, the existing source floor for PM increased from 0.04 lb/ton clinker to 0.05 lb/ton clinker, the existing source floor for THC decreased to 15 parts per million by volume, dry (ppmvd), and the existing source floor for HCl stayed the same at 3 ppmvd.
As explained in section B below, the EPA is proposing to establish a beyond the floor standard for mercury of 55 lb/MM tons clinker. Moreover, for reasons independent of this analysis, the EPA is proposing to amend the existing and new source NESHAP for PM. See section D below. The EPA is not proposing to amend the HCl standard or the THC standard.
2. New Source Standards. With respect to new source standards, the EPA does not believe that any reclassification and reanalysis is necessary under the court's opinion. New source floors can be based on the performance of “the best controlled similar source”, as opposed to existing source floors which must reflect performance of sources “in the category or subcategory”. See CAA section 112(d)(3) and (d)(3)(A). A CISWI cement kiln is similar to a non-CISWI cement kiln since the device is a cement kiln. Equally important, burning secondary materials for energy recovery does not significantly alter a cement kiln's HAP emission profile. See 76 FR 28320 (May 17, 2011) (documenting both the basis for this conclusion and the cement industry's agreement with it).
1. New Source Standard. As explained above, the new source standard is based on the performance of the best performing similar source.
2. Existing Source Standard. As noted above, the recalculated existing source floor is 58 lb/MM tons clinker produced. The EPA is proposing a beyond-the-floor standard of 55 lb/MM ton clinker produced, the level of the 2010 final standard. As described below, the only difference in cost between the two levels is the incremental cost of removing slightly more mercury, which is estimated at $2,000/lb of mercury removed. This is because the control equipment needed for mercury would not alter, would not need to be sized differently, and would need to perform on average nearly identically at either a 55 lb/MM tons clinker or a 58 lb/MM tons clinker level. That is, in planning compliance, kilns would calibrate to achieve an average performance of 34.1 lb/MM tons clinker for a standard of 58 lb/MM tons clinker, and 31.7 for a standard of 55 lb/MM tons clinker, which translates to an additional reduction of 2.4 lb/MM tons of clinker per year. This equates to an estimated 180 pounds of nationwide mercury emissions per year, incremental to the recalculated floor. To achieve this additional reduction, we estimated an additional cost of approximately $355,000 for the industry, the cost of purchasing additional carbon injection materials. This equates to a cost-effectiveness of $2,000/lb of mercury reduction per year. This is the incremental cost of going from the recalculated floor of 58 lb/MM tons clinker to the proposed 55 lb/MM tons clinker. Because this is the same level as the 2010 rule, there are no incremental costs or emissions impacts when compared to the 2010 rule. See section 8.2, Portland Cement Reconsideration Technical Support Document. Moreover, this reduction is highly cost-effective. A cost effectiveness value of $2,000/lb. mercury is considerably less than values the EPA have found to be cost effective for removal of mercury in other air toxics rules. For example, in the National Emission Standards for Hazardous Air Pollutants: Mercury Emissions from Mercury Cell Chlor-Alkali Plants, the cost effectiveness was found to be between $13,000 to $31,000 per pound for the individual facilities (see Supplemental proposed rule, 76 FR 13858 (March 14, 2011)). The EPA also does not see any adverse energy or non-air quality health or environmental consequences of a 55 lb/MM tons clinker beyond-the-floor standard.
We are not proposing a beyond the floor level below 55 lb/MM tons clinker for the same reasons given in the 2010 final rule—in particular the possibility that a lower standard could force some kilns to find alternative sources of limestone, at enormous cost and disruption. See 75 FR 54980 (September 9, 2010).
The THC data for the 2010 standard consist of CEMS data for 15 kilns. After removing the four CISWI kilns, nine kilns remain. Thus, the MACT floor kilns consisted of 12 percent of these nine kilns, or two kilns. The top two kilns were Suwannee and Holcim. As explained above, when CISWI sources are removed from the database for the 2010 standards, the existing source floor for THC becomes more stringent from 24 ppmvd to 15 ppmvd, and the new source standard would drop from 24 ppmvd to 11 ppmvd. This change results from removing from the database a CISWI cement kiln (the Lehigh Union Bridge kiln) with the lowest daily average performance but with more associated variability than the other kilns with the next highest daily average performance. See also 76 FR 28322 (May 17, 2011) n. 11 and 665 F. 3d at 188. However, notwithstanding this calculation, the EPA is not proposing to reduce the level of either the new source or the existing source THC standard.
1. New Source Standard. As just explained, the new source standard can be based on performance of a “best controlled similar source”, so there is no reason under the statute or the court's remand to amend the new source THC standard. The standard is also technically appropriate. See 75 FR 54981 (September 9, 2010) (explaining basis for the THC new source standard, which discussion is summarized below for the readers' convenience). Removing the CISWI Union Bridge kiln as the best performing new source would leave the Suwannee kiln as the lowest emitter based on its daily average THC emissions. See Portland Cement Reconsideration Technical Support Document (TSD), section 8.4, which is available in this rulemaking docket. This kiln has higher average emissions than the Union Bridge kiln (that is, its daily average emissions are higher than the Union Bridge kiln). This kiln thus emits more THC than the Union Bridge kiln, so the EPA identified the kiln emitting less THC on average—the Union Bridge kiln—to be the best performer. The Suwannee kiln has less measured variability than the Union Bridge kiln, but also has hundreds of fewer observations. For this reason, the EPA considered the Union Bridge kiln to be more representative of variability, and used its 99th percentile performance as the measure of performance of the best performing similar source in establishing the new source standard. See 75 FR 54981 (September 9, 2010).
2. Under the calculation described above, the existing source floor would be reduced from 24 ppmvd to 15 ppmvd. Subject to any comments the EPA receives on this proposed action, the EPA believes that such a floor level would not be technically appropriate. It omits the variability of the similar source with the best average performance for THC (the Union Bridge kiln), and so may not be fully representative of variability of best performing sources. As noted above, cement kiln HAP emissions are not appreciably affected by burning secondary materials for energy recovery so the Union Bridge's variability is representative of cement kiln variability. In addition, as noted above, the number of daily observations for the Union Bridge kiln is among the most robust in the database, containing over 3 times the number of observations as the next best performing cement kiln. Thus, there is a “demonstrated relationship” between the variability of the Union Bridge kiln and the variability of the best performing sources in the existing source floor pool.
If the variability of the Union Bridge kiln is included along with the variability of the two best performing cement kilns, and applied to the two best performing cement kilns' performance, the floor would be 24 ppm, which the EPA is proposing as a floor. See Portland Cement Reconsideration TSD, section 8.4. This is the level of the 2010 standard.
3. Beyond the floor standards. The EPA is not proposing a beyond the floor THC standard for existing cement kiln sources. The reasons given in the rulemaking remain valid. See 75 54983 (September 9, 2010); 74 FR 21153 (May 6, 2009). We especially note that a more stringent standard for THC would force the increased use of energy-intense control technologies like regenerative thermal oxidizers (RTO) which have negative environmental implications, notably increased emission of carbon dioxide (CO
Based largely on developments which have occurred after theEPA granted reconsideration on certain aspects of the NESHAP
On January 17 2012, LaFarge Cement submitted a petition for reconsideration containing no new data or information but arguing that the Holcim petition justified reconsideration of the standards. The EPA believes that this petition is subsumed by the Holcim petition.
In comments to the 2009 proposal, industry commenters maintained that there were several problems with implementing the monitoring requirements to demonstrate compliance using a PM CEMS and with the requirements to conduct a periodic audit of the PM CEMS in accordance with Performance Specification (PS) 11 of appendix B and Procedure 2 of appendix F to part 60. The EPA responded to these comments in the 2010 final rule. See 75 FR 55007 (September 9, 2010); NESHAP Response to Comment Document pp. 163-166. Since that time, the Portland cement industry has identified further technical issues associated with the current PM CEMS technology in satisfying PS 11 correlation requirements that have emerged as the industry has attempted to develop a CEMS-based compliance strategy for PM pursuant to the 2010 NESHAP.
1. PS 11. The EPA has continued to review the application of PM CEMS in relation to the procedures and acceptance criteria of PS 11, the protocol mandated by the promulgated rule. See section 63.1350(b)(1). PS 11 is structured differently than other PS that apply to validating the performance of gaseous pollutant CEMS. This is primarily because the pollutant, PM, is defined entirely by the test method specified by regulation to measure it. As the industry commenters note, there are no independent standard reference materials for PM concentrations as there are for gaseous pollutants (e.g., NIST traceable compressed gases for validating SO
PS 11 provides procedures and acceptance criteria for validating the performance of several types of PM CEMS technologies. Although there are multiple instrument and data reporting operational performance checks in PS 11 that are similar in concept to those for gaseous pollutant CEMS, there is the principal PM CEMS performance requirement that is distinctly different. That requirement is the development of a site-specific PM CEMS correlation or mathematical response curve. There is a key procedural element to developing that correlation. That is, PS 11 requires that the source conduct multiple stack test runs using an EPA PM test method (e.g., Method 5) and simultaneously collect corresponding PM CEMS output data. PS 11, section 8.6, requires at least five test runs at each of three different operating (i.e., low, mid, and high PM concentration) conditions that range from 25 to 100 percent of allowable emissions, if possible, for a total of 15 or more test runs. Then the source must use the test method data and the corresponding PM CEMS output data to develop an equation (i.e., a calculated linear or nonlinear curve) that will be used to define the relationship between the PM CEMS output and the test method measured PM concentrations. Each site-specific correlation must meet several PS 11 acceptance criteria including limits on confidence interval and tolerance interval equating to ±25 percent of the applicable emissions limit.
2. Discussion of Technical Issues. A particular challenge in applying PM CEMS to source emissions monitoring is in measuring the very low PM concentrations associated with a low applicable emissions limits for PM precisely enough to meet the PS 11 correlation requirements. In addition to
As noted above, PS 11 specifies acceptable criteria for a correlation directly related to the applicable emissions limit. The Portland cement NESHAP PM emissions limit for existing sources of 0.04 lb/ton of clinker equates to 5 to 8 mg/dscm, depending on production rate (assuming a typical total gas flow rate per clinker production rate). For a PM CEMS set up to measure compliance with a 5 to 8 mg/dscm equivalent limit, the inherent uncertainty associated with a 1 hour Method 5 measurement (±0.6 to 1.2 mg/dscm) would constitute more than half of the ±25 percent of the applicable PS 11 acceptance threshold (i.e., ±1.2 to 2.0 mg/dscm) of the mid-level PS 11 correlation test (i.e. the correlation for the middle of the three PS 11 correlation points).
Although one can improve the method detection capabilities of the Method 5 or other filterable PM test method by increasing sampling volume and run time, uncertainties in measurement would remain. For example to achieve a practical quantitation limit of 1 mg/dscm, one would need to conduct a test run of 6 hours or longer. The measurement uncertainty associated with a 6-hour Method 5 test runs at this concentration would be ±0.01 to 0.2 mg/dscm. At this level, the uncertainty associated with the PM test method measurements alone would be about half of the correlation limit allowed in PS 11. The PS 11 correlation calculations would also have to account for any PM CEMS measurement uncertainty.
Factoring in the inherent PM CEMS response variability and the uncertainty associated with the representative sampling (e.g., PM and flow stratification), we agree with commenters that trying to satisfy PS 11 at such low concentrations using 1 hour Method 5 test runs could be problematic. The same issue arises for the new source standard because of the lower limit of the new source standard.
The industry also argued that the variable raw feed material and chemical additives used in cement production will lead to changes in particle size, refractive index, particle density, and other physical characteristics of the particulate in the exhaust stream. This is important, according to the comments, because correlations developed for the light scatter and scintillation PM CEMS technology may be adversely affected by these physical changes in particles irrespective of changes in mass emissions rates or concentrations.
In developing the 2010 final rule, the EPA assumed that cement kilns would be using light-scatter or scintillation PM CEMS.
The dominant sources of PM from a cement kiln are not from fuel combustion but from processing raw materials. Cement kilns process mostly limestone with naturally occurring variability in component percentages. See 74 FR 21142 (May 6, 2009); 75 FR 54977 (September 9, 2010). Cement kiln operators also add other chemical additives in variable concentrations to produce certain product characteristics. See 74 FR 21142. As noted in the EPA's technology background documents (e.g.,
This is an issue of special import for cement kilns. One can expect significant variations in particle size distribution and other particle characteristics in Portland cement kiln exhaust because of the complicating effects of variable content feed materials and chemical additives. That means that correlations developed for one set of conditions may not apply with changes in feed materials or under other operating conditions (e.g., different chemical additives).
The EPA has investigated whether PM CEMS that work on principles other than light scattering could effectively measure cement kiln PM and be calibrated per PS 11 requirements. There is at least one other PM CEMS technology, beta attenuation PM CEMS, also referred to as beta gauge technology that is much less sensitive to changes in particle characteristics than are light based detectors. The beta attenuation PM CEMS extracts a sample for the stack gas and collects the PM on a filter tape. The device periodically advances the tape from the sampling mode to an area where the sample is exposed to Beta radiation. The detector measures the amount of beta radiation emitted by the sample and that amount can be directly related to the mass of PM on the filter.
The majority of PM CEMS devices used to date by cement kilns are based on light scatter or scintillation detection. We understand that a few Portland cement operators have applied beta attenuation devices. Since the EPA premised the rule on use of a different type of PM CEMS, since there is minimal operating experience with beta gauge PM CEMS in this industry, and because we are not aware that the experience includes a beta gauge PM CEMS calibrated per PS 11, the EPA believes that some type of research effort involving testing would be needed before predicating a PM standard on use of a beta gauge PM CEMS. Such an effort is likely to take several years to implement.
These issues exacerbate the uncertainties of calibrating PM CEMS at the level of the 2010 p.m. standards noted above. Using data from longer Method 5 test runs will improve the probability of a PM CEMS meeting PS 11 correlation requirements but will also raise practicality concerns potentially without completely resolving the problems. Given the combination of the low emissions concentrations PM CEMS measurement
3. A monitoring approach alternative to PM CEMS and PS 11. To address technical issues associated with PM CEMS meeting PS 11 correlation requirements at low PM emissions concentrations from cement kilns and the potentially variable PM emissions characteristics expected from Portland cement kilns, the EPA is proposing to change the compliance basis for the PM emissions limit from PM CEMS and the 30-day average emissions calculation. For monitoring continuous compliance, the rule would require PM CEMS equipment but, as explained below, that equipment would be used for continuous parametric monitoring rather than for direct measure of compliance with the numerical PM emissions limit.
The EPA is proposing to change the means of demonstrating compliance from PM CEMS to Method 5 stack testing. In applying Method 5, PM is withdrawn isokinetically from the source and collected on a glass fiber filter maintained at a temperature of 120 ± 14 °C (248 ± 25°F). The PM mass, which includes any material that condenses at or above the filtration temperature, is determined gravimetrically after the removal of uncombined water. Compliance with the numerical emissions limit is then based on an average of three 2-hour test runs rather than a 30-day average determined from PM CEMS data. The numerical level of the standard would change to reflect the different averaging period. See 75 FR 54988 (September 9, 2010) (explaining that more measurements of a properly designed and operated control device decreases measured variability since there are likely to be more measurements at the mean of performance); see also 75 FR 54975 (September 9, 2010) (explaining how this phenomenon is reflected in the Upper Prediction Limit (UPL) equation used to project variability, since the m term (i.e., the number of measurements) in the equation becomes larger with more observations resulting in a larger denominator and hence lower ultimate level). By changing from a 30-day average with potentially 720 hourly values to a three-run test average producing three test run values, we reviewed and revised the calculation of the PM emissions floor and standard, and consistent with the court's remand, removed all CISWI kilns from the database in doing so. In calculating the PM MACT floor, the best performing kilns used in the analysis changed as a result of removing the kilns identified as CISWI kilns.
In addition, we realized that in the original analysis PM emissions data for a single kiln were inadvertently treated as test results for three different kilns. After making that correction and after eliminating kilns identified as CISWI, the number of kilns in the data set was reduced from 45 kilns to 28 kilns. Therefore, the best performing 12 percent was represented by four kilns. As a result of removing the CISWI kilns, two kilns which were not best performers in the 2010 dataset are now best performers. See TSD section 8.3 and Appendices E and F.
As in the 2009 proposal, we used individual test run data from our best performing kilns and calculated the 99th confidence UPL. Rather than using m = 30 in the equation as we did in the 2010 final rule where compliance was based on a 30 day rolling average, see 75 FR 54988 (September 9, 2010), we used m = 3 consistent with the proposed requirement to determine compliance using a three run Method 5 test. Under this analysis, we determined the revised proposed PM MACT floor to be 0.07 lb/ton clinker produced when based on the three run Method 5 test. Beyond-the-floor standards do not appear to be justified for the same reasons given in the 2010 final rule. See 75 FR 54988 (September 9, 2010). We are, therefore, proposing this emissions limit for the kiln and clinker cooler and an initial and annual compliance test using Method 5 to demonstrate compliance.
These issues affecting the existing source PM limit also apply to the new source PM limit. Based on this revised compliance regime, the new source floor would change from 0.01 lb/ton clinker produced, to 0.02 lb/ton clinker produced, based on a three run average from a Method 5 stack test. See Portland Cement Reconsideration TSD, section 8.3. The best performing kiln used to set the MACT floor for new sources in the 2010 rule was a cement kiln, not a CISWI kiln, so the same kiln was used for this analysis. The difference is that because a 3-run test would be used to determine compliance rather than a 30-day rolling average, the calculation of the 99th confidence UPL used m = 3 rather than 30, which results in a floor of 0.02 lb/ton clinker. The EPA is not proposing a beyond-the-floor standard for the reasons given at 75 FR 54988 (September 9, 2010).
As indicated above, the EPA is further proposing to use PM CEMS technology for continuous parametric monitoring of the proposed PM standards. The EPA has developed requirements for continuously monitoring operating parameters in instances where compliance is based on non-continuous measurements, as would now be the case for PM. This implements section 114(a)(3) of the CAA which requires major sources to use enhanced monitoring for compliance certifications. The EPA's historic approach has been to require monitoring of a control device operating condition (e.g., electrical power, water flow rate, pH) the limit of which is based on a periodic compliance test with the compliance test method. The use of a continuous parametric monitoring system (CPMS) based on PM CEMS technology (PM CPMS) is a significant step closer to direct measurement of emissions in units of the emissions limit and an improvement over less direct monitoring of a process control device conditions.
Specifically, this proposal recognizes the value of PM monitoring technology sensitive to changes in PM emissions concentrations and use of such a tool to assure continued good operation of PM control equipment. This approach avoids the PM CEMS calibration (i.e., PS 11 correlation) issues that can be exacerbated for Portland cement installations. PM CEMS technology can be effective in monitoring control device performance (see, e.g., 77 FR 9371 (February 16, 2012)) where the EPA established PM CPMS parametric operating limits for electricity generating units).
As a result, this proposed rule would require the installation and operation of a PM CPMS for parametric monitoring associated with the proposed PM standard. The source owner would not have to meet PS 11 requirements but would have to prepare and submit for approval, if requested by a permitting authority, a site-specific monitoring plan to apply sound practices for installing, calibrating and operating the PM CPMS.
Current PM CPMS have an operating principle based on in-stack or extractive light scatter, light scintillation or beta attenuation. The source owner or operator would need to examine the fuel and process conditions of his stack as well as the capabilities of these devices before selecting a particular CPMS technology. The reportable measurement output from the PM CPMS may be expressed as milliamps, stack
We are proposing a number of consequences if the kiln PM monitoring parameter is exceeded. First, the source owner will have 48 hours to conduct an inspection of the control device and to take action to restore