                                                                               
MEMORANDUM   

TO:		Jodi Howard, EPA/OAQPS/SPPD

FROM:	Bradley Nelson, EC/R, Inc.

DATE:	April 25, 2016

SUBJECT:	Estimation of Potential Emission Reductions with the Implementation of a Method 21 Monitoring Program 
--------------------------------------------------------------------------------


      The purpose of this memorandum is to estimate the potential emission reduction from the implementation of a Method 21 monitoring and repair program. This analysis estimates the potential Method 21 emission reduction potential for leak definitions of 500 and 10,000 parts per million (ppm) and monitoring frequencies of annual, semiannual and quarterly. The analysis will be used to determine the equivalency of Method 21 monitoring and repair program with fugitive emission monitoring and repair using optical gas imaging (OGI).
Emission Reduction Methodology
	The potential emission reductions from a Method 21 monitoring and repair program were estimated using data and procedures from the EPA Equipment Leak Protocol (referred to in this memorandum as the EPA Protocol). This document provides procedures and approaches for estimating potential emissions from synthetic organic chemical manufacturing industry (SOCMI), petroleum refineries, marketing terminals and oil and gas production. 
      For estimating leak detection and repair (LDAR) control effectiveness, we used the methodology outlined in Chapter 5.3.1 of the EPA Protocol. The control effectiveness is calculated using a stepwise approach that starts from the initial leak frequency and adds monitoring cycles until the leak frequency after monitoring reaches steady state. The difference between the initial leak rate and the final leak rate provides the control effectiveness for the leak monitoring program. The equations for these leak rates at various leak definitions are provided in the EPA Protocol for SOCMI, petroleum refineries, marketing terminals and oil and gas production. The variable for determining the leak rate from these equations is the leak fraction. To determine the leak fraction, the stepwise approach was used to estimate the leak frequency immediately after and immediately preceding a monitoring cycle.
      The equations for determining these values are shown in the equations below.
Yi=Zi-FR-Zi+FR*Zi*R
where:
Yi = Leak fraction immediately after monitoring cycle i;
Zi = Leak fraction immediately after monitoring cycle i (Note that Z1 equals the initial leak frequency);
R = Fraction of repaired components for which a leak immediately recurs; and
FR = Fraction of leaking components successfully repaired.

Zi+1=Oc*1-Yi+YI
where:
Zi*1 = Leak fraction immediately preceding monitoring cycle i+1;
Oc = Fraction of non-leaking components which will leak in the time period between monitoring cycles (e.g., occurrence rate); and
Yi = Leak fraction immediately after monitoring cycle i.
      Other parameters included in the monitoring cycle calculations are the percentage of successfully repair components, the percentage of new leaks and the percentage of leaks that were repaired but have reoccurred. The EPA Protocol only provides data for these other parameters for the SOCMI and refineries. The refinery emissions data are provided in non-methane organic compound (NMOC) units, which would require assumed TOC and methane weight fractions to determine the TOC emission factors, whereas the SOCMI emissions data is already based on TOC. The assumed TOC and methane weight fractions would add another level of uncertainty to the emission reduction percentage calculations if the refinery data were used. Because the objective of this memorandum is to estimate the potential emission reductions from the implementation of a Method 21 monitoring and repair program, we determined that using the SOCMI data would provide the best estimate, and would be comparable to the potential fugitive emission reductions for oil and gas production, if the other parameters were available for this segment. 
      The SOCMI data in the EPA Protocol only included occurrence rates for monthly and quarterly monitoring of 1.0 and 2.97 percent, respectively. To calculate annual and semiannual occurrence rates, a logarithmic equation was derived from the data points. Using this logarithmic equation, the occurrence rate for semiannual and annual monitoring were estimated to be 4.21 and 5.46 percent, respectively. A plot of the equation can be found in the appendix of this memorandum.
      The initial leak fraction for the Method 21 monitoring program was calculated using the average leak equations in the EPA Protocol. Valves were selected because they represent the largest portion of fugitive emissions from oil and gas operations, and would provide comparable emission reduction percentages as other components. Initial leak frequencies were calculated using the EPA Protocol average leak rate equations for 500 and 10,000 ppm gas valves (see Table 5-4 of the EPA Protocol) and the average SOCMI emission factor for gas valves of 0.00597 kilograms per hour per source (see Table 2-1 of the EPA Protocol) and solving for leak fraction. The initial leak frequencies can also extrapolated using the lines in Figure 5-1 in the EPA Protocol. The average leak fraction equation and calculated initial leak frequency are provided in Table 1.

    Table 1. Parameters and Assumptions Used to Calculate Monitoring Cycles
                                   Parameter
                           Parameter Value (500 ppm)
                                Parameter Value
                                 (10,000 ppm)
Occurrence Rate
                                5.46% Annual, 
                              4.21% Semiannual, 
                                2.97% Quarterly
                                5.46% Annual, 
                              4.21% Semiannual, 
                                2.97% Quarterly
Recurrence Rate
                                      14%
                                      14%
Unsuccessful Repair Rate 
                                      10%
                                      10%
Initial Leak Frequency 
                                    13.53%
                                     7.49%
Average Leak Rate (ALR) Equation[a]
                           ALR = 0.044*LF + 0.000017
                           ALR = 0.078*LF + 0.00013
a. the variable LF stands for leak fraction or frequency.
      Using the parameters in Table 1, the estimated emission reductions were calculated using the monitoring cycle approach in the EPA Protocol. The average fugitive emission rate was calculated for each cycle using the average leak rate equation and the average of the Yi and Zi leak fractions for that cycle.  The leak frequency after monitoring reached steady state on the sixth monitoring cycle and the percent reduction was calculated by comparing the average leak rate with the initial leak rate (i.e. Cycle 1). The results of the emission reductions are presented in Table 2. The calculations for these emission reductions are provided in the appendix of this memorandum.
      As shown in Table 2, in all cases the percent reduction for the 500 parts per million Method 21 alternative is equal to or greater than the estimated OGI monitoring and repair percent reduction. The percent reduction for the 10,000 parts per million leak threshold was greater than the OGI option only for annual monitoring. 
Table 2. Percent Reduction in Emissions for EPA Method 21 Monitoring and Repair 
                             Monitoring Frequency
                      Fugitive Emission Percent Reduction
                                       
                          Method 21  Repair Threshold
                                    OGI[a]

                                  10,000 ppm
                                    500 ppm
                                       
Annual
                                      42
                                      68
                                      40
Semiannual
                                      55
                                      75
                                      60
Quarterly
                                      67
                                      83
                                      80
      a. OGI emission reductions percentage based on information from Colorado cost-benefit analysis.
      
















Appendix A
Method 21 Emissions Reduction Calculations





















Graph of leak occurrence and monthly and quarterly monitoring SOCMI data to estimate semiannual and annual leak occurrence.
