15th LDAR Symposium May 19-21, 2015 New Orleans,

15th LDAR Symposium May 19-21, 2015 New Orleans,

15th LDAR Symposium May 19-21, 2015 New Orleans, LA Development of Emission Factors from API 622/624 Test Data Buzz Harris and Bronson Pate Sage Environmental Consulting Standards Certification Education & Training Publishing Conferences & Exhibits Presenter Information Buzz Harris holds a BS in Chemical Engineering with 45 years experience and still learning Bronson Pate holds a degree in Engineering with 8 years of full time experience plus 3 years of internships in LDAR during university studies 2 Overview Low Emission (Low E) Technology Introduction Primary US tests for Low E packing and valves use methane and include Method 21 type readings Tests are based on accelerated wear that should be representative of 5 years or more of field operations EPA is requiring Low E in new Consent Decrees and encouraging voluntary application Emission Factors (EF) and control efficiencies can be developed from the test readings, which, if approved, would provide more incentive to use Low E in any new

construction Conclusions 3 Low E Technology There have always been packings and valves that were more or less likely to leak Only recently, however, have we had guarantees, warranties, and, most importantly, test data to prove manufacturers claims of low emissions Based on hard data, EPA has begun to require Low E in Consent Decrees written over the last 4 to 5 years Several companies had begun their own Low E testing and implementation voluntarily before that The use of Low E is gradually being recognized as cost effective and just good business 4 Fugitive Emissions Testing A number of valve and packing tests have been developed: ISO 15848-1 is an EU and British standard that allows testing by either helium or methane, with most test data to date done with helium API 622 is a packing test using methane API 624 is a valve test using methane ChevronTexaco test procedure is similar to API 622, but it tests packing in a valve and includes more wear cycles Shell also has its own packing test The API test procedures have more industry wide applicability and focus on methane testing 5 API 622 and 624 Details Comparison of Service Parameters: API 622 and API 624

Standard identification Title and edition API STD 622 API STD 624 Type Testing of Rising Stem Valves Equipped Type Testing of Process Valve Packing for with Graphite Packing for Fugitive Emissions, Fugitive Emissions, Second Edition First Edition Date Standard by the American Petroleum Institute, 10/01/2011 Standard by the American Petroleum Institute, 02/01/2014 Pass Criteria Equipment Packing adjustment Media Temperature Pressure 500 ppmv maximum after one adjustment Specified fixture simulating a valve One allowed1 Methane 97% minimum purity 500F (260C) 600 psig (41.4bar-g) 100 ppmv maximum

Valve being qualified None allowed Methane 97% minimum purity 500F (260C) 600 psig (41.4bar-g) Number of valve stem actuations 1510 310 Number of thermal cycles 5 3 Leak measurement method Method 21 Method 21 Leak measurement details Done with stem in static state Done with stem in static and dynamic states Leak measurement frequency Every 50 actuations of the stem Every 50 actuations of the stem

Prepared by and used with permission of Garlock 6 API 622 Test Data 7 PPM Readings in API 622 and 624 Probe split in two to simultaneously check both stem and packing Tin foil used as a partial shroud to try to collect leakage from any point around the stem and packing Record ppmv data over a minute and report the average and maximum with no background correction Some report both static and dynamic (stem moving) readings This paper uses the maximum readings recorded in either static or dynamic modes This should be conservatively higher than traditional M21 readings 8 Mass Emissions from ppm For petroleum industry: Default Zero EF (reading zero) = 0.0000078 kg/hr/source Correlation equation = 2.29*10^-6*SV^0.746 kg/hr/source For chemical industry: Default Zero EF (reading zero) = 0.00000066 kg/hr/source Correlation equation = 1.87*10^-6*SV^0.873 kg/hr/source

No need for Pegged EF in Low E testing! These emission estimates are for the instant the measurement occurs Still need to estimate emissions over time 9 Accelerated Wear Time API 622 includes 1510 mechanical wear cycles and 5 thermal stress cycles API 624 includes 310 mechanical wear cycles and 3 thermal stress cycles, but also requires packing to have passed API 622 before 624 testing The tests take place over 3 to 6 days, but represent a much longer period because of the accelerated use The equivalent process life may vary with: Continuous vs. batch processes Process application of the valve (manual isolation, drain, sample, motor operated, control, etc. valve) Other factors too process specific for consideration here 10 Continuous Process Cycles to Time Accelerated Wear Cycles to Operating Time Continuous Processes Frequency per year % of Total Weighted Frequency Valve Application Low End

High End Valves Low End High End Manually operated block isolation valves 1 10 78% 0.78 7.8 Block valves isolating pumps 5 24 2% 0.1 0.48 Drain valves 12

120 5% 0.6 6 Sample valves 52 795 2% 1.04 15.9 Motor operated valves 150 1000 2% 3 20 Process control valves 500

5000 11% 55 550 Totals 720 6949 100% 60.52 600.18 Average Annual Operation Cycles 330.35 Years of Operation for 1510 Operating Cycles 4.57 11 Batch Process Cycles to Time Accelerated Wear Cycles to Operating Time Batch Processes Frequency per year

% of Total Weighted Frequency Valve Application Low End High End Valves Low End High End Manually operated block isolation valves 12 365 75% 9 273.75 Block valves isolating pumps 24 365 2%

0.48 7.3 Drain valves 12 365 5% 0.6 18.25 Sample valves 52 795 2% 1.04 15.9 Motor operated valves 150 1000 5%

7.5 50 Process control valves 500 5000 11% 55 550 Totals 750 7890 100% 73.62 915.2 Average Annual Operation Cycles 494.41 Years of Operation for 1510 Operating Cycles 3.05

12 Continuous Process Thermal Cycles to Time Thermal Cycles to Operating Time Continuous Processes: Scheduled Unsheduled Total Run Lengths Between T/A, yrs Shutdowns/yr Shutdowns/yr Shutdowns/yr 5 0.20 0.5 0.70 4 0.25 0.4 0.65

3 0.33 0.3 0.63 2 0.50 0.2 0.70 1 1.00 0.1 1.10 Average 0.76 Years Operation for 5 thermal cycles 6.61 13 Batch Process Thermal Cycles to Time Thermal Cycles to Operating Time

Batch Processes: Scheduled Unsheduled Total Campaign Run Lengths, days Shutdowns/yr Shutdowns/yr Shutdowns/yr 30 12.17 0.05 12.22 20 18.25 0.04 18.29 10 36.50

0.03 36.53 5 73.00 0.02 73.02 1 365.00 0.01 365.01 Average Years Operation for 5 thermal cycles 101.01 0.05 14 Summary Accelerated Wear Time 1510 mechanical wear cycles represent 3.1 to 4.6 years of operating time for batch/continuous process units 5 thermal cycles represent 0.1 to 6.6 years of operating time for batch/continuous process units EPA defines Low E as <100 ppm operations for 5 years, and accepts API 622 data as satisfying that definition The thermal cycles in API 622 falls far short of the potential

thermal cycles in 5 years for a batch process, but API 622 accelerated wear cycles average to around 5 years equivalent operation for continuous operating units and mechanical wear for batch units This paper, therefore, assumes that the API 622/624 readings are spread evenly over a 5 year period 15 API 622 Emissions Over Time Example 1510 cycles with a reading every 50 cycles (including beginning and end readings) results in 42 readings 42 readings spread evenly over 5 years would be one reading every 45.6 days (roughly every 6 weeks) Default zero and correlation equations for petroleum and SOCMI are used to calculate instantaneous emission estimates Emissions from the previous to current readings are averaged over the 45.6 days between readings Emissions are summed over the 42 total readings and divided by the time in years to create an average emission factor in kg/yr/source Emission Factor for static and dynamic readings are averaged 16 API 624 Emissions Over Time Example 310 cycles with a reading every 50 cycles (including beginning and end readings) results in 14 readings 14 readings spread evenly over 5 years would be one reading every 140.4 days (roughly semi-annual) Default zero and correlation equations for petroleum and SOCMI are used to calculate instantaneous emission estimates Emissions from the previous to current readings are

averaged over the 140.4 days between readings Emissions are summed over the 14 total readings and divided by the time in years to create an average emission factor in kg/yr/source Emission Factor for static and dynamic readings are averaged 17 Emission Factors Summary Summary of API 622 and 624 Testing-Based Emission Factors Petroleum Chemical Calculated Calculated Static Dynamic Emission Emission Average Maximum Average MaximumFactor, kg/hr Factor, kg/hr 4 20 6 23 6.9203E-06 7.0746E-06 5.1 37 5.7 12 8.8547E-06 1.141E-05 1 5 1 8 7.3361E-06 1.8432E-06 1 4 1

4 4.5681E-06 2.228E-06 16 33 16 34 1.8124E-05 2.1304E-05 10 38 13 46 1.4043E-05 1.5093E-05 10.1 49 NA NA 1.1793E-05 1.1769E-05 3 15 7 20 9.8515E-06 1.0118E-05 Methane, ppmv Test Packing Chevron Teadit 2236 624 Teadit 2236 622 Garlock 1303 FEP 622 Garlock 212 ULE 622 Chesterton 1622

622 Chesterton 1622 622 Nippon Pillar EDP 15 624 Nippon Pillar EDP 15 Valve Velan 4 300 Ladish 4 600 Velan 4 300 Stuffing Box Stuffing Box Velan 4 300 Stuffing Box Velan 4 300 API 622 and 624 tests reports were provided by and thanks go to: Ron Walters of Teadit North America Todd Haberkost of Ladish Valves Jim Drago of Garlock Scott Boyson and Rodney Roth of A.W. Chesterton Company Josh Erd of Nippon Pillar Corporation of America 18 Low E Emission Factor Considerations All of these emission factors calculated from low emission packing and valves are near the default zero emission factor for valves One approach to encourage voluntary use of Low E would be to allow use of the default zero factor to predict emissions for permitting Another approach would be to allow an average emission factor (or control efficiency) for Low E valves and packing to date to be used to predict emissions for permitting Another approach would be to require applicants to include

calculations of an emission factor for the specific Low E equipment they will use 19 Control Efficiency Approach In lieu of developing emission factors for Low E, it would also be possible to develop a control efficiency for application of Low E that could be applied to the normal average emission factor The following table shows control efficiency estimates calculated as: CE%=(1-(Average EF/Low E EF))X100 Each Average EF above is for valves in a specific service category in a specific industry Each Low E EF above is for valves in a specific industry, but all testing is done on Gas/Vapor service (methane, 600 psi) 20 Low E Control Efficiency Low E Control Efficiency Estimates Valve Avg. EF Low E EF Control Industry Service kg/hr/source kg/hr/source Efficiency Refining Gas/Vapor 0.0268 1.01864E-05 97.32% Refining LL

0.0109 1.01864E-05 98.91% Refining HL 0.00023 1.01864E-05 99.98% Refining Average All 98.74% Chemical Gas/Vapor 0.00597 1.01049E-05 99.40% Chemical LL 0.00403 1.01049E-05 99.60% Chemical HL 0.00023 1.01049E-05 99.98% Chemical Average All 99.66% Low E Emission Factors are the average of all the API 622/624 tests using either petroleum or chemical correlations/default zero emission factors. 21 Conclusions The measurements done in API 622/624 testing give

methane ppmv data that can be used to estimate emissions API 622/624 are accelerated wear tests that appear to represent a five year period for the average valve (with the exception of thermal cycles on a batch process) Emission factors (EFs) developed for API 622/624 data range in value from 4.6E-6 up to 1.8E-5, falling about one order of magnitude higher than the default zero EFs (6.6E-7 up to 7.8E-6) Control efficiency numbers for the Low E EFs compared to EPA Protocol Average EFs range from 97.3 to 99.9% 22 Conclusions These Low E EFs and control efficiencies could be used for permitting, where actual components that can be monitored do not exist yet EPA should also consider allowing use of Low E EFs for non-monitored valves (such as HL service, UTM, etc.) Neither EPA nor the states currently accept these Low E EFs and control efficiencies for permitting We ask that EPA review these calculations and/or replicate their own approach to similar calculations EPA-sanctioned Low E EFs and/or control efficiencies would add another incentive for every new facility or modification to be done with Low E technology 23

Recently Viewed Presentations

  • Red Scare in America, 1919 - Somerset Independent Schools

    Red Scare in America, 1919 - Somerset Independent Schools

    , or citizens of other countries living in the U.S., just belonging to certain groups considered radical could lead to . deportation, or being sent back to one's own country. In late 1919 Palmer's forces arrested thousands and deported hundreds....
  • The Emergence of Modern Canada - Ms Kubiak&#x27;s School Site

    The Emergence of Modern Canada - Ms Kubiak's School Site

    The Emergence of Modern Canada. 1896-1914. Introduction. ... and Donald Mann to extend their lines through the Yellowhead Pass, and down the Thompson River to Kamloops, and through to Vancouver. ... Officials and police officers tried boarding the ship, but...
  • P, PI &amp; PID Controllers

    P, PI & PID Controllers

    AP, EECE, Dept. DIT University, Dehradun (c) Nafees Ahamad, EECE, DIT University, Dehradun. Controllers . A device or system or program for improving transient and steady state response of a control system. (c) Nafees Ahamad, EECE, DIT University, Dehradun. Proportional...
  • the linguistic situation of the Caribbean

    the linguistic situation of the Caribbean

    the linguistic situation of the Caribbean Silvia Kouwenberg overview & background L32A 2006/07 Week 2 Language in the Caribbean: overview and background Topics: (lack of) geographical unity pre-Columbian population & languages of the Guianas, the Antilles the special case of...
  • Scrapbooks &amp; Stall Decorations

    Scrapbooks & Stall Decorations

    I really wanted a calf that maintenance wise would be a little easier to take care of and that maybe I could decide at a moment's notice to load her up and take her to a show without a lot...
  • Eisenstein&#x27;s Ivan the Terrible Part II: The Politics

    Eisenstein's Ivan the Terrible Part II: The Politics

    Ivan the Terrible Part I. Deals with the idea of absolute power granted by God in order to "gather lands" and restore Russia's grandeur. Events: Ivan's early years, coronation, marriage and death of the wife, Kurbski's treason, victory over the...
  • Preview of New 2016 Boy Scout Rank Requirements

    Preview of New 2016 Boy Scout Rank Requirements

    Repeat from memory the Scout Oath, Scout Law, Scout motto, and Scout slogan. In your own words, explain their meaning. Understand and agree to live by the Scout Oath or Promise, Scout Law, motto, and slogan, and the Outdoor Code....
  • 48x72 poster template - University of California, San Diego

    48x72 poster template - University of California, San Diego

    Primary productivity measurements are taken once a day. For these experiments, the watercolumn is sampled in the late morning at six depths cooresponding to the isolumes (97, 40, 10, 4, 1, 0.3%). Incubation bottles are inoculated with 14C-bicarbonate at local...