www.erg.com
www.erg.com

MEMORANDUM

TO:	Tonisha Dawson, U.S. EPA/OAQPS/SPPD  -  EPA Office of Air Quality Planning and Standards

FROM:	Eastern Research Group, Inc.

DATE:	September 2018

SUBJECT:	Cost Impacts for the Asphalt Processing and Asphalt Roofing Manufacturing Risk and Technology Review Proposal
--------------------------------------------------------------------------------

INTRODUCTION
This regulatory action proposes amendments to the Asphalt Processing and Asphalt Roofing Manufacturing National Emission Standards for Hazardous Air Pollutants (NESHAP). The proposed amendments include:
          Eliminating the startup, shutdown, and malfunction (SSM) exemption consistent with the United States Court of Appeals decision in Sierra Club v. EPA 551 F.3d 1019 (D.C. Cir. 2008). Eliminating the SSM exemption will remove the requirements and associated burden for SSM plans and periodic SSM reports. We estimate a one-time burden for reviewing requirements associated with this amendment, as discussed in section 2.0 of this memorandum.
          Requiring periodic performance testing for each air pollution control device used to comply with the PM, total hydrocarbon (THC), opacity, and visible emission standards. We estimate an associated cost for this proposed amendment, as discussed in section 3.0 of this memorandum.
          Adding an option to allow the use of manufacturers' specifications to establish the maximum pressure drop across the control device used to comply with the particulate matter (PM) standards. For this proposed amendment, we estimate cost savings as discussed in section 4.0 of this memorandum.
          Adding flexibility to allow owners and operators to use the performance test average inlet temperature and apply an operating margin of +20 percent to determine maximum inlet gas temperature of a control device used to comply with the PM standards. For this proposed amendment, we estimate cost savings as discussed in section 4.0 of this memorandum.
          Requiring owners and operators to submit electronic copies of performance test reports, performance evaluation reports, compliance reports, and Notification of Compliance Status reports. For these proposed amendments, we estimate no cost impacts. In other words, we assume the burden estimate for paper format submittal of these reports is an appropriate estimate for electronic submittal of these reports.
          Other technical corrections to improve consistency and clarity. For these proposed amendments, we estimate no cost impacts.
We estimate the following costs (in 2017$) over a 5-year period for the proposed amendments: $4,000 in 2019 associated with reviewing the proposed amendments and $138,800 in 2022 for 5-year periodic performance testing. We also estimate the following cost savings (in 2017$) over a 5-year period for the proposed amendments: $16,500 in 2021 associated with the proposed monitoring options in lieu of a performance test and $490,000 in 2023 associated with extending filter replacements by three months. To estimate costs and cost savings, we assume (i) a 5-year period for the analysis and (ii) that the rule becomes final in 2019. This remainder of this memorandum provides details on how these costs and cost savings were determined. For additional details on how the costs and cost savings are summarized over time, please see the memorandum entitled "Economic Impact Analysis for Asphalt Processing and Asphalt Roofing Manufacturing NESHAP RTR Proposal" in the docket (EPA-HQ-OAR-2017-0662).
ONE-TIME COST FOR REVIEWING REQUIREMENTS ASSOCIATED WITH ALL PROPOSED AMENDMENTS 
We estimated a one-time cost of about $500 per facility for reviewing the requirements associated with all proposed amendments described in section 1.0 of this memorandum. We assume this review will occur in 2019. This one-time cost is expected to result in a total industry-wide cost of $4,000 in 2017 dollars. 
This cost assumes that, for each facility, it will require 4 technical hours (plus 0.2 managerial and 0.4 clerical hours) to review and understand the proposed amendments, at a labor rate of $112.98/hour for technical labor (and $149.35/hour for managerial and $54.81/hour for clerical labor).
COST FOR 5-YEAR PERIODIC PERFORMANCE TESTING
For each air pollution control device used to comply with the PM, THC, opacity, and/or visible emission standards, the EPA is proposing periodic performance testing requirements at 40 CFR 63.8691(e). Depending on the date of construction or reconstruction of each affected source, the EPA is proposing the first periodic performance test compliance date to be either 3 years after the final rule is promulgated, or no later than 60 months following the initial performance test. The EPA is also proposing subsequent periodic performance tests be conducted no later than 60 months thereafter following the previous periodic performance test. Because all facilities have already conducted their initial performance test, we estimate the cost of the first periodic performance tests to occur in the third year, or 2022, after the final rule is promulgated.
We estimated a total industry-wide cost of $138,800 in 2017 dollars over a 5-year period  for these proposed testing requirements. Appendix A provides details for this estimate. 
COST SAVINGS DUE TO PROPOSED MONITORING REQUIREMENTS FOR CONTROL DEVICE USED TO COMPLY WITH THE PM STANDARDS
For control devices used to comply with the PM standards of the Asphalt Processing and Asphalt Roofing Manufacturing NESHAP, facilities are required to conduct a performance test to reestablish new pressure drop and temperature operating limits. However, the EPA is proposing to add an option at 40 CFR 63.8689(d) to allow the use of manufacturers' specifications to establish the maximum pressure drop across the control device used to comply with the PM standards. The EPA is also proposing to add a footnote to Table 2 to subpart LLLLL of part 63 of the Asphalt Processing and Asphalt Roofing Manufacturing NESHAP to allow owners and operators to use the performance test average inlet temperature and apply an operating margin of +20 percent to determine maximum inlet gas temperature of a control device used to comply with the PM standards.
These proposed monitoring requirements are projected to impact asphalt roofing manufacturing facilities as follows:
          Alleviate some need for a facility to have to retest the PM control device for the sole purpose of reestablishing new temperature and pressure drop operating limits (see section 4.1); and
          Allow facilities to extend filter replacement by three months (see section 4.2).
Cost Savings for Using the Proposed Monitoring Options In Lieu of Conducting a Performance Test to Reestablish New Temperature and Pressure Drop Operating Limits
The proposed monitoring options described in section 4.0 are estimated to alleviate some need for a facility to have to retest the PM control device for the sole purpose of reestablishing new temperature and pressure drop operating limits. We assumed one of the four asphalt roofing manufacturing facilities subject to the PM standards would have conducted one performance test on one PM control device in the second year, or 2021, after the final rule is promulgated simply for the sole purpose of reestablishing new temperature and pressure drop operating limits. Therefore, we estimated a total industry-wide cost savings for using the proposed monitoring options in lieu of conducting a PM performance test to be $16,500 in 2017 dollars over a 5-year period. 
Cost Savings for Extending Filter Replacements by Three Months
To comply with pressure drop operating limits established during the performance test for control devices used to comply with the PM standards of the Asphalt Processing and Asphalt Roofing Manufacturing NESHAP, companies assert that they replace filters approximately 3-months sooner (e.g., every 12 months instead of every 15 months) than if pressure drop operating limits were allowed to be established using manufacturer specifications (as previously mentioned in section 4.0, facilities are currently required to conduct a performance test to reestablish new pressure drop operating limits). According to these companies, it is difficult to schedule testing dates that would capture a high enough pressure drop to account for the full range allowed by the manufacturer (see Appendix B).
Using air permits and the information received as part of the CAA section 114 request, we identified the asphalt roofing manufacturing facilities subject to the PM standards and estimated the number of filters used at each facility to comply with these PM standards. We determined a replacement cost of $35,000 per filter which is the average of the estimates provided to the EPA by CertainTeed Corporation and Owens Corning (see Appendix B).
Based on this information, we estimated a total industry-wide cost savings for extending filter replacements by three months to be $490,000 in 2017 dollars over a 5-year period . Table 2 summarizes these cost savings.
Table 2. Cost Savings for Extending Filter Replacements by Three Months
                                 Facility Name
                                     City
                                     State
                               Number of Filters
              Cost to Replace every 12-mths over 5-yr period[1,2]
                                      ($)
              Cost to Replace every 15-mths over 5-yr period[1,3]
                                      ($)
                         Net Savings over 5-yr period
                                      ($)
                                 Owens-Corning
                                  Minneapolis
                                      MN
                                       3
                                    525,000
                                    420,000
                                   (105,000)
                               Certainteed Corp.
                                    Oxford
                                      NC
                                       4
                                    700,000
                                    560,000
                                   (140,000)
                       Owens-Corning Fiberglass Trumball
                                    Medina
                                      OH
                                       4
                                    700,000
                                    560,000
                                   (140,000)
                               Certainteed Corp.
                                   Shakopee
                                      MN
                                       3
                                    525,000
                                    420,000
                                   (105,000)
                                                 TOTAL SAVINGS OVER 5-YR PERIOD
                                   (490,000)
[1]	Each filter replacement is assumed to be an average cost of $35,000 based on information provided by CertainTeed Corporation and Owens Corning.
[2]	Assumes filter is replaced every 12-months, or 5 times in 5 years. ($35,000/filter) x (# filters) x (5 replacements/5-year period).
[3]	Assumes filter is replaced every 15-months, or 4 times in 5 years. ($35,000/filter) x (# filters) x (4 replacements/5-year period).

NATIONWIDE COST IMPACTS
For the 8 facilities subject to the Asphalt Processing and Asphalt Roofing Manufacturing NESHAP we estimate the following costs (in 2017$) over a 5-year period for the proposed amendments: $4,000 in 2019 associated with reviewing the proposed amendments and $138,800 in 2022 for 5-year periodic performance testing. We estimate the following cost savings (in 2017$) over a 5-year period for the proposed amendments: $16,500 in 2021 associated with the proposed monitoring options in lieu of a performance test and $490,000 in 2023 associated with extending filter replacements by three months. These costs and costs savings were presented in sections 2.0 through 4.0.


                                       
                                       
                                       
                                       
                                       
                                  APPENDIX A
                                       
DETAILS USED TO ESTIMATE COST OF REQUIRING 5-YEAR PERIODIC PERFORMANCE TESTING
                                       
                                       
                                       
                                 Facility Name
                                     City
                                     State
                           State Required
5-yr Tests
                           Asphalt Storage Tanks[1]
                               Number of Filters
                          Number of Thermal Oxidizers
                          VE or Opacity
Test Cost[1]
                                      ($)
                                PM
Test Cost[2]
                                      ($)
                               THC
Test Cost[3]
                                      ($)
                                Total
Test Cost
                                      ($)
                             Hunt Refining Company
                                  Tuscaloosa
                                      AL
                                      Yes
                                      Yes
                                       0
                                       1
                                     1,500
                                       0
                                    20,750 
                                see footnote 4 
                                 Owens-Corning
                                  Minneapolis
                                      MN
                                      Yes
                                      Yes
                                       3
                                       1
                                     1,500
                                    38,700 
                                    20,750 
                                see footnote 4 
                  Hunt Southland Refining Co. - Sandersville
                                  Heidelberg
                                      MS
                                      No
                                      Yes
                                       0
                                       1
                                     1,500
                                       0
                                    20,750 
                                    22,250 
                               Certainteed Corp.
                                    Oxford
                                      NC
                                      No
                                      Yes
                                       4
                                       1
                                     1,500
                                    49,800 
                                    20,750 
                                    72,050 
                       Owens-Corning Fiberglass Trumball
                                    Medina
                                      OH
                                      Yes
                                      Yes
                                       4
                                       1
                                     1,500
                                    49,800 
                                    20,750 
                                see footnote 4 
                            Wynnewood Refining Co.
                                   Wynnewood
                                      OK
                                      No
                                      Yes
                                       0
                                       1
                                     1,500
                                       0
                                    20,750 
                                    22,250 
                              Valero Refining Co.
                                    Ardmore
                                      OK
                                      No
                                      Yes
                                       0
                                       1
                                     1,500
                                       0
                                    20,750 
                                    22,250 
                               Certainteed Corp.
                                   Shakopee
                                      MN
                                      Yes
                                      Yes
                                       3
                                       1
                                     1,500
                                    38,700 
                                    20,750 
                                see footnote 4 
                                                    TOTAL COST OVER 5-YR PERIOD
                                    138,800
[1]	Cost of opacity testing using EPA Test Method 9 is estimated to be $1,500. For each facility that has asphalt storage tanks, it was assumed that the facility would have at least one Method 9 test.
[2]	Cost of PM performance testing using EPA Test Method 5A is estimated to be $16,500 for first filter and an add-on charge of $11,100 for each additional filter.
[3]	Cost of THC performance testing using EPA Test Method 25A is estimated to range from $16,200 (concentration) to $20,750 (efficiency). We chose to be conservative and apply the higher cost of $20,750.
[4]	Costs not included for these facilities because they are already required to test every 5-yrs by the State permitting agency and will not incur additional testing costs with the new testing requirements.
                                       
                                       

                                       
                                       
                                       
                                       
                                       
                                  APPENDIX B
                                       
                         COMPANY EMAIL CORRESPONDENCE

CertainTeed Corporation Email Correspondence

From: Spittler, Melissa [mailto:Melissa.A.Spittler@saint-gobain.com] 
Sent: Tuesday, August 21, 2018 8:15 AM
To: Dawson, Tonisha <dawson.tonisha@epa.gov>
Subject: RE: Performance Testing

Good morning:

First  -  I want to clarify my email below on 8/13  -  it should read "not solely".

Second, thanks for asking about the filter costs. On average, the cost for replacing filters is approximately $60,000 each time it is done. Right now, we are changing filters 2-3 months earlier than recommended by the manufacturer.

If you need anything else, please let me know.

Thank you,

Melissa Spittler
CertainTeed
--------------------------------------------------------------------------------


From: Dawson, Tonisha [mailto:dawson.tonisha@epa.gov] 
Sent: Tuesday, August 21, 2018 9:00 AM
To: Spittler, Melissa
Subject: RE: Performance Testing

Based on the estimated filter cost you provided, it appears that having to replace filters early is a much higher expense than having to repeat testing. Am I interpreting this correctly?
--------------------------------------------------------------------------------


From: Spittler, Melissa [mailto:Melissa.A.Spittler@saint-gobain.com]
Sent: Friday, August 24, 2018 8:07 AM
To: Dawson, Tonisha <dawson.tonisha@epa.gov>
Subject: RE: Performance Testing

Hi Tonisha  -  I will respond to both of your questions in one email (frequency with filter bed changes and the cost question).

First, the quick one - 12 month interval for filter changes because of the issues with differential pressure being above tested.

Second - Yes, you are correct in interpreting the costs (change filters versus re-test). On the surface, the filter change can be an expensive undertaking compared to repeat testing. However, cost alone is not the only criteria to consider.

Re-testing has the potential to establish new conditions that a site would be subject to. As the group has shared with you previously, there are variables associated with testing that are not within our control, such as ambient temperature. Outside temperature can have an impact on operating conditions and attempts to control temperatures by other means have been unsuccessful. Similarly the atomizing of water may create other unintended process issues For example, one site evaluated the spraying of cool water into the duct work to maintain temperature. This was not effective  -  it shorted out electrical components and got into sealant, compromising the quality of the product.

Oxford did re-test, in an effort to obtain better operating parameters and chose to stop further re-testing because they were told that if they got lower numbers (i.e., more restrictive operating range), that would be their new limit. Summertime temperatures generally reach above 90°F consistently. On the flip side, when the weather is below freezing, ice crystals can form in the device which increases the differential pressure.

Obtaining results from performance testing can take up to 30 days, whereas, completing routine maintenance (filter changes) is performed as needed/scheduled, without delay and allowing greater operational flexibility without compromising operational compliance with operating limits.

Shakopee has not re-tested since their initial performance testing, however, they successfully reasoned with the Region 5 to use of manufacturers recommended values on their mist eliminators for inlet temp (120 degrees) and differential pressure (16 inches of water). I have provided a brief summary of that reasoning below, but in part, is because the MACT is specific in that the differential pressure and temperature requirement are for devices used to meet the particulate standard.

1.	Testing a mist eliminator on an outside storage tank at the maximum inlet temperature cannot be accomplished without manipulating the test. The maximum inlet temperature will follow ambient conditions since the mist eliminator is pulling atmospheric air across the headspace of the tank. It is nearly impossible to reach "worst case" inlet temperature numbers without reducing airflow with a baffle or otherwise manipulating the test. If the industry is going to be held to a maximum inlet temperature requirement based on the test, then the sites are going to artificially increase that number during the testing by manipulating airflow. Scheduling a Method 9 test on the hottest day of the year is problematic at best since the hottest day isn't always a clear enough day for testing, and the required 30 day notice for the testing prevents an accurate prediction of test day conditions.
2.	Differential pressure is a false indicator for the performance of a mist eliminator. The best treatment occurs at the highest differential pressures. Holding sites to a maximum established during testing means that sites can never use a filter that is "dirtier" than the one used during testing even though dirtier means better performance as a filter. This logic is flawed and will result in early filter replacement. In addition, icing occurs on the filter at extreme cold temperatures. For sites in the northernmost regions, maximum differential pressure occurs when the outside air is the coldest. It is impossible therefore to get to the maximum differential pressure during testing as the coldest temperatures occur in the middle of the night when method 9 observations are not possible.
3.	The mist eliminator manufacturers have guaranteed 0 percent opacity at up to 120 degree inlet temperature and 16 inches of differential pressure. 
4.	CertainTeed has never "failed" an opacity test for a mist eliminator controlling emissions from group 2 tanks only. It is clear that there is no issue meeting 0 percent opacity at the maximum manufacturers recommended values of 120 degrees inlet temperature and 16 inches of water differential pressure.
5.	EPA region 5 has already agreed to allow CertainTeed to use manufacturers recommended values or the values obtained during the performance testing (whichever numbers are higher) as they have recognized the inherent issues with the language in the MACT as written. As such, the MACT should clearly differentiate between mist eliminators used to meet the particulate matter standard for production line sources and mist eliminators that are only used to control opacity from group 2 asphalt storage tanks.

Melissa
Owens Corning Email Correspondence

From: Whiteside, Devlin [mailto:Devlin.Whiteside@owenscorning.com] 
Sent: Tuesday, August 28, 2018 8:23 AM
To: Dawson, Tonisha <dawson.tonisha@epa.gov>
Cc: Whiteside, Devlin <Devlin.Whiteside@owenscorning.com>
Subject: RE: Asphalt RTR Question

Hi Tonisha  -  cost depends on the size of the unit and the number of filters per unit. To change the filters in a small unit can be around $10,000 whereas a larger unit can be $60,000 or higher.

Devlin Whiteside
503-358-9234


