                                                                               
MEMORANDUM


DATE:	February 9, 2015

SUBJECT:	Revised Technology Review for the Off-Site Waste and Recovery Operations Equipment Leaks

FROM:	Lesley Stobert, EC/R Inc.

TO:		Paula Hirtz, EPA/OAQPS/SPPD


The EPA proposed amendments to off-site waste and recovery operations (OSWRO) national emissions standards for hazardous air pollutants (NESHAP) on July 2, 2014, based on residual risk and technology reviews conducted for the source category. The purpose of this memorandum is to present a revised analysis related to the technology review for equipment leaks at OSWRO facilities, considering comments submitted during the public comment period for the proposed OSWRO NESHAP amendments.

1.0	INTRODUCTION	

On July 2, 2014 (79 FR 37850), the EPA proposed amendments to the OSWRO source category following the results of a review required by section 112(d)(6) of the CAA, which directs the EPA to review previously issued NESHAP regulations and to revise them as necessary (taking into account developments in practices, processes and control technologies) no less frequently than every 8 years. Section 112(f)(2) of the CAA requires the EPA to assess, within 8 years of promulgation of the original NESHAP for major sources and area sources subject to MACT, the remaining risks due to emissions of HAP from these source categories and determine whether the emissions standards provide an ample margin of safety to protect public health. These reviews are collectively referred to as residual risk and technology reviews (RTRs).

For the July 2014 proposal, a technology review for equipment leaks at OSWRO facilities was conducted. Potential developments in practices, processes and control technologies were identified through the review of subsequent air toxics regulations and the proposed National Uniform Emission Standards for Storage Vessels and Transfer Operations, Equipment Leaks, and Closed Vent Systems and Control Devices ("Uniform Standards"), which were proposed on March 26, 2012. Under CAA section 112(d)(6), the EPA proposed to require facilities to follow the more stringent leak detection and repair (LDAR) program of 40 CFR part 63, subpart H rather than allowing facilities the choice of complying with either the requirements of that rule or with 40 CFR part 61, subpart V. 

Following the July 2014 proposal, the EPA received comments from one industry trade association and two companies challenging some of the data and assumptions used in the analysis of emissions reductions and costs associated with the LDAR program of 40 CFR part 63, subpart H. We have considered these comments and conducted an analysis to determine the impacts of these comments on our previously estimated emissions reductions and costs associated with compliance with 40 CFR part 63, subpart H over the emissions reductions and costs associated with compliance with 40 CFR part 61, subpart V. 

2.0	EQUIPMENT LEAKS RE-ANALYSIS

2.1	Equipment Leaks at OSWRO Facilities

Equipment leaks are emissions emanating from leaking pieces of equipment, such as pumps, compressors, agitators, pressure relief devices, sampling connection systems, open-ended valves or lines, valves, connectors, surge control vessels, bottoms receivers, and instrumentation systems in organic HAP service; and any control devices or systems required by subpart DD for these pieces of equipment. The amount of HAP emissions from equipment leaks is proportional to the type and number of pieces of equipment and the concentration of HAP constituents of the stream passing through the equipment.

2.2	Summary of Existing MACT Level of Control

The OSWRO MACT standards for equipment leaks apply to the group of all equipment that contains or contacts off-site material having a total HAP concentration of 10 percent by weight or more and intended to operate in off-site material service for 300 hours or more during a calendar year. For all of the equipment subject to the standards, the owner or operator must comply with the LDAR requirements of either 40 CFR part 61, subpart V or 40 CFR part 63, subpart H. The primary provisions of these subparts are shown in Table 1.
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
  Table 1. Primary Provisions from Part 61, Subpart V and Part 63, Subpart H
Equipment Type
                                40 CFR part 61
                                   subpart V
                           40 CFR part 63
subpart H
Valves in gas/vapor service and  light liquid service
Conduct EPA Method 21 monitoring and leak definition of 10,000 ppm (monthly or less often up to quarterly depending on leak rate)
Same as part 61, subpart V but with leak definition of 500 ppm (monthly or less often up to annually depending on leak rate)
Pumps in light liquid service
 Conduct monthly EPA Method 21 monitoring and a leak definition of 10,000 ppm
Same as part 61, subpart V but with leak definition of 1,000 ppm and repair required at 2,000 ppm
Connectors in gas/vapor and light liquid service
                Sensory monitoring (visual, audible, olfactory)
     (type of service is not specified, no monitoring schedule specified)
Conduct EPA Method 21 monitoring with leak definition of 500 ppm annually (or less often up to once every 4 years depending on leak rate)
Pressure relieve devices in gas/vapor service
Operate with no detectable emissions (i.e., less than 500 ppm above background) or equipped with a rupture disk or routed to an air pollution control device
                          Same as part 61, subpart V
Open-ended valves or lines
             Equip with a cap, blind flange, plug or second valve
                          Same as part 61, subpart V

2.3	Proposed Standard for Equipment Leaks

At proposal, during our review of regulatory development efforts for other rules, we found LDAR programs that incorporate more stringent leak detection limits than the OSWRO MACT standard. Leak definitions of 500 parts per million (ppm) for valves in gas and vapor and in light liquid service, 1,000 ppm for pumps in light liquid service and 500 ppm for connectors in gas and vapor and in light liquid service are the leak definitions currently required for facilities subject to certain existing rules and were considered as regulatory options for the proposed Uniform Standards for equipment leaks. These leak definitions are the same as those required by 40 CFR part 63, subpart H. Consequently, we identified the more stringent leak definitions of the subpart H LDAR program as a development in practices, processes or control technologies. We also identified two options for implementing the requirements of the subpart H LDAR program. As it is possible that it may be cost effective to monitor only certain equipment at certain levels, Option 1 included switching from a subpart V LDAR program to a subpart H LDAR program, without the subpart H monitoring requirements for connectors in gas/vapor and light liquid service. Option 2 included switching from a subpart V LDAR program to a subpart H LDAR program, with the subpart H monitoring requirements for connectors in gas/vapor and light liquid service. We estimated the costs and emissions reduction of both options based on information received from the industry as part of a CAA section 114 survey and on information used in the technology review analysis for the Uniform Standards for equipment leaks. Based on the estimated control costs and emissions reductions from this analysis, we proposed to require equipment leaks Option 2. 




2.4 	Revised Estimation of Emissions Reductions and Costs of a 40 CFR part 63, subpart H LDAR Program

Seals on agitators in gas/vapor service and in light liquid service are required by 40 CFR part 63, subpart H to be monitored monthly using Method 21; 40 CFR part 61, subpart V has no requirements for agitators. At proposal, we did not consider the requirements for agitators to be among the primary provisions of 40 CFR part 63, subpart H, and we did not include the emissions reductions or the costs of including an LDAR program in our analysis of the 40 CFR part 63, subpart H LDAR program.

Based on comments received during the public comment period regarding the proposed revised equipment leak requirements from Clean Harbors, a company with several OSWRO facilities, we have revised our analysis of the costs of an LDAR program with the more stringent leak definitions of 40 CFR part 63, subpart H to include monitoring of agitators under both Options 1 and 2. 

Table 2 presents a summary of the equipment leak options that were included in the revised analysis. Other types of equipment not included in this table are largely the same under 40 CFR part 61, subpart V and 40 CFR part 63, subpart H, and were not analyzed in this comparison of emissions reductions and costs for Options 1 and 2.
                                       
   Table 2. Summary of Equipment Leak Options for the OSWRO Source Category
                               Type of Equipment
                                   Subpart V
                                  (Baseline)
                              Alternative Options
                                       
                                       
                                   Subpart H
                                w/o connectors
                                       
                                  (Option 1)
                                       
                                   Subpart H
                                 w/connectors
                                       
                                  (Option 2)
                                       
                                     Pumps
          Monthly Method 21 monitoring; leak definition of 10,000 ppm
          Monthly Method 21 monitoring; leak definition of 1,000 ppm 
                             (repair at 2,000 ppm)
          Monthly Method 21 monitoring; leak definition of 1,000 ppm
                             (repair at 2,000 ppm)
                                    Valves
          Monthly Method 21 monitoring; leak definition of 10,000 ppm
                        Monthly Method 21 monitoring; 
                          leak definition of 500 ppm
                        Monthly Method 21 monitoring; 
                          leak definition of 500 ppm
                                  Connectors
                                   Sensory 
                                  monitoring
                                   Sensory 
                                  monitoring
Instrument monitoring for connectors in gas/vapor and light liquid service at 500 ppm 
                                   Agitators
                                No requirements
          Monthly Method 21 monitoring; leak definition of 10,000 ppm
          Monthly Method 21 monitoring; leak definition of 10,000 ppm

We estimated emissions reductions and costs for the two control options using a model plant approach to represent equipment counts found at OSWRO facilities. The median number of pieces of equipment was estimated from the reported data available from the OSWRO CAA section 114 survey. Table 3 presents the median model plant equipment counts for estimating equipment leak impacts from the OSWRO source category.

                                       
                        Table 3. Model Plant for OSWRO
Equipment Type
                       Number of Pieces of Equipment[a]
Pumps (light liquid service)
                                      27
Valves (light liquid service)[b]
                                      910
Connectors (in gas/vapor and light liquid service)
                                     1,350
Agitators (in gas/vapor and light liquid service)
                                       9
       [a] The number of pieces of equipment is based on the median value reported by facilities in responses to the OSWRO CAA section 114 survey.
       b  The type of service a valve was in (gas/vapor or light liquid) was not discernible from the CAA section 114 survey responses, and based on the process types present at OSWRO facilities, it was assumed the reported valves were in light liquid service.
       
Next, using the number of pieces of equipment in Table 3, emission factors were applied to determine the baseline emissions and the emissions under Options 1 and 2 that would be expected by equipment type. For the baseline, we assumed facilities are complying with 40 CFR part 61, subpart V based on information from the section 114 questionnaire responses, which indicate that most facilities comply with this option. The emission factors were multiplied by the number of pieces of equipment of that type at the OSWRO model plant and converted to tons per year (tpy) to arrive at the total emissions in tpy at the baseline (40 CFR part 61, subpart V) and for Options 1 and 2 (40 CFR part 63, subpart H without and with connector monitoring). These emissions factors were based on those used in the proposed Uniform Standards for equipment leaks analysis, but because the factors used in that analysis are for total organic compounds, which for OSWRO facilities are generally volatile organic compounds (VOC), these factors were modified to reflect the amount of VOC that are expected to be HAP. The VOC emission factors were adjusted to HAP emission factors using the HAP to VOC ratio of 82 percent for equipment leaks from the OSWRO MACT development Background Information Document (BID). The resulting emissions factors used for the OSWRO source category are shown in Table 4. 

        Table 4. Equipment Leak HAP Emission Factors by Equipment Type
Type of Equipment
                    Emission Factor (kg/hr/equipment piece)

                                   Baseline
                               Option 1/Option 2
Pump Seals  -  Light Liquid Service
                                   2.05E-03
                                   5.69E-04
Valves  -  Light Liquid Service
                                   3.19E-04
                                   1.90E-04
Connectors  -  Gas/Vapor and Light Liquid Service
                                   2.51E-04
                               2.51E-04/1.33E-04
Agitators  -  Gas/Vapor and Light Liquid Service
                                   1.01E-02
                                   2.05E-03

To determine the costs of the LDAR program at the baseline and for Options 1 and 2, the approach taken was the approach used for the proposed Uniform Standards for equipment leaks analysis. In determining the costs for the OSWRO source category, the same leak rates, repair times, labor charges and other administrative and miscellaneous costs were assumed (see Appendix 1 for more detailed cost information). Table 5 shows the costs and cost effectiveness by equipment types and as a total for all four equipment types for each option. The cost effectiveness of Option 1 is $1,426, and for Option 2 the cost effectiveness is $3,598. The incremental cost effectiveness of Option 2 compared to Option 1 is $6,715.
                                       
Table 5. Summary of Cost Effectiveness Analysis for Pumps, Valves, Connectors and Agitators

2.5 	Alternate Analysis of Connector Monitoring Costs 

Based on comments from two commenters (the American Chemistry Council (ACC) and Eastman Chemical Company) regarding the costs to monitor connectors, we performed an additional analysis to determine the potential effects increased connector monitoring costs may have on the cost effectiveness of the 40 CFR part 63, subpart H LDAR program as applied to OSWRO facilities. Specifically, the commenters estimate that the ongoing monitoring fee per connector is $6.50, compared with the value of $2.50 used in our analysis, and they estimate the administrative costs associated with connector monitoring are $27,000 per year, compared with the value of $4,636 per year used in our analysis. 

Regarding ongoing monitoring costs, we believe the $2.50 used in our analysis is a reasonable estimate of the monitoring costs per connector. This estimate, which has also been used in previous EPA LDAR analyses, is based on an average monitoring cost per component of $1.00 to $1.50, and then increased to $2.50 to account for industry claims that connectors are more difficult than other components to monitor. However, to determine how a fee of $6.50 per connector, as suggested by the commenters, would affect the cost effectiveness of the provisions, we conducted an additional analysis of costs of an LDAR program using this value.

Regarding the administrative and reporting costs, the commenters estimate these costs to be $27,000 per year for these activities for connectors alone, based on a study Eastman Chemical Company recently undertook at its Kingsport, Tennessee facility to examine its LDAR program costs. The results of this study were summarized in an attachment to the ACC comment letter. Based on information provided in this study, the weighted average labor rate for LDAR administrative activities is $34.56 per hour, which is reportedly based on labor rates from the U.S. Bureau of Labor Statistics data multiplied by a factor of 1.5 to account for overhead and benefits costs. At these labor rates, administrative costs of $27,000 would equate to 781 hours per year. We do not find this amount of time to be reasonable for connector administrative and reporting costs, especially considering that connector monitoring may only be required once every four years. However, it may be possible that our estimate of 50 hours per year at a labor rate of $92.92 per hour overestimates the labor rate and underestimates the amount of time required to complete the necessary administrative requirements. Therefore, we conducted an additional analysis of the costs of an LDAR program assuming twice as many hours as we previously estimated and the labor rates provided by the commenter for these administrative actions, in addition to considering the suggested increased costs for connector monitoring described above. 

Using these alternate connector monitoring and administrative cost values, the incremental cost effectiveness for connectors would be $6,825/ton of HAP reduced. This incremental cost effectiveness is still approximately $7,000/ton HAP reduced, as was calculated without these alternate connector monitoring values. Table 6 provides a summary of the costs and cost effectiveness for the LDAR program as shown in Table 5, but includes these alternate connector monitoring costs. See Appendix 2 for more detailed cost information. 


Table 6. Summary of LDAR Program Cost Effectiveness Analysis with Alternate Connector Monitoring Costs
                                       



                                       
                                  Appendix 1:
                       Equipment Leak Model Plant Costs


                                       
                                       
                                  Appendix 2:
Equipment Leak Model Plant Costs with Alternate Connector Monitoring Cost Values













	
                                       
