Document ojE0EVOaY2VrKVmRq3zGMjrr

DATE: SUBJECT: FROM: THRU: TO: UNITED STATES ENVIRONMENTAL PROTECTION AGENCY REGION 5 77 WEST JACKSON HOULEVARD CHICAGO, ILLINOIS 60604 OCT t 52018 CLEAN AIR ACT INSPECTION REPORT Kensington Processing Plant Gregory Gehrig, Environmental Engineer AECAB (Ml/WI) Sarah Marshall, Section Chief AECAB (Ml/\Vl) File BASIC lNFORMATlON Facility Name: Kensington Processing Plant Facility Location; 11543 State Route 644, Kensington, OI-! Date of Inspection: September 20, 2018 EPA Inspectors: 1. Gregory Gehrig, Environmental Engineer 2. Kenneth Ruffatto, Environmental Engineer 3. Cary Secrest, Environmental Protection Specialist Other Attendees: 1. Yousheng Zeng, Providence Photonics 2. Sri Motyala, Providence Photonics 3. Jon Morris, Providence Photonics 4. Christopher Bender, Environmental Manager, Momentum Midstream, LLC 5. Mark Hay, Plant Manager, Momentum Midstream, LLC 6. Dave Pratt, Momentum Midstream, LLC Contact Email Address: Christopher Bender <C.Bender@m3midstream.com> Purpose of Inspection: CAA Compliance Facility Type: Midstream oil and gas processing facility Facility Name: Kensington Processing Plant Facility Location: 11543 State Route 644, Kensington, OH Date of Inspection: September 20, 2018 Regulations Central to Inspection: SIP, NSPS and MACT Arrival Time: 1240 hrs Departure Time: 1605 hrs Inspection Type: ~ Unannounced Inspection D Announced Inspection OPENING CONFERENCE ~ Credentials Presented ~ CBI warning to facility provided The following information was obtained verbally from facility personnel. Company Ownership: The facility is owned by Utica East Ohio and operated by Momentum Midstream, LLC (Momentnm). Tbe facility has had the same ownership since it began operating in 2013. Process Description: The Kensington Processing Plant (Kensington) is a midstream oil and gas processing facility. Momentum processes oil and gas extracted at wells via two pipe lines. These two pipelines are the high pressure and the medium pressure pipelines. Liquids are removed by a 'slug catcher', stabilized and then sent via pipeline to a fractionation facility. Gas is cooled in cryogenic process and separated into four product streams: methane, ethane, propane and butane. There are the four cryogenic trains at this facility. Four flares are used at this facility, althongh only three were operating at the time of the inspection. Observations of the flares are summarized below. Various storage vessels are also used to hold liquids, including condensate, methanol, and natural gas liquids. TOUR INF0RMAT!ON EPA toured the facility: Yes Data Collected and Observations: EPA contracted Providence Photonics (Providence) of Baton Rouge, LA to conduct flare observations at Kensington using a Video Imaging SpectroRadiography (VISR) imaging device, or VISR. VISR uses infrared sensing and imaging technology to detennine the combustion efficiency (CE) of hydrocarbons combusted by a flare with a visible flame. Providence used the VISR to observe the flaring along with a dedicated standard video camera. EPA also utilized a FUR camera to observe the flares and collect images. Page 2 of 4 Facility Name: Kensington Processing Plant Facility Location: 11543 State Route 644, Kensington, OH Date of Inspection: September 20, 2018 At the time of inspection, the facility was operating normally. Observations of the flares were: High Pressure (HP) flare - Control room data showed a rather steady vent gas (VG) flow rate of approx. 275 lb/hr to this unassisted flare. VlSR observations indicated CE of approx. 99% on average. Tank flare (associated with T00l ) - Very little flaring was observed from this unassisted flare during the inspection. Several short events of venting lasting a few seconds each were seen. One of these short events had noticeable visible emissions, although they all burned bright orange. VISR data showed the CE for these short events was bet\veen 90 and 99%. Emergency flare (associated with HP stabilizer) - Only the pilot lights were visible with the FLIR, and therefore only CE readings for the pilots were collected with the VISR. This flare is air-assisted and the headers leading to it are swept with natural gas from the Kensington natural gas product delivery line. Loadout flare - This unassisted flare was not operating at the time of inspection. A report from Providence on the VISR data and images is provided in Appendix A. EPA and Providence conducted a limited tour of the facility for the purposes of observing flare operations. EPA observed the flare operating parameters in the control room and took readouts detailing conditions at the time of inspection. Field Measurements: were taken during this inspection. RECORDS REVIEW Five control room flare operating SCADA information printouts (taken). CLOSING CONFERENCE Requested documents: None at the time of inspection. / 4 .. SIGNATURES /, ~ Lead Inspector /~~~~ Date Section Chief: Date: / O-Jl - /1 rD/ ( SJr8 Page 3 of 4 Facility Name: Kensington Processing Plant Facility Location: 11543 State Route 644, Kensington, OH . Date of Inspection: September 20, 2018 DIGITAL IMAGE LOG* 1. Inspector Name: Gregory Gehrig 3. Company/Facility Name: Kensington Processing Plant 5. Number of Images: 9 2. Date(s) of Inspection: September 20, 2018 4. Street Address, City, State: 11543 State Route 644, Kensington, OH 6. Archival Record Location: CD-R labeled Kensington Images 9.20.2018 Time Image name (EDT) Description IMG 0147 1442 Emergency flare at distance IMO 0148 1448 Emergency flare closer IMO 0149 1515 HP flare MOV 302 1411 FLIR video of Tank flare MOV 303 1413 FUR video of Tank flare MOY 305 1427 FUR video of Tank flare associated v.rith 99.5% CE MOY 306 1455 FUR video of Emergency flare (imaged bv G. Gehrig) MOV 307 1517 FUR video of HP flare MOY 309 1529 FLIR video of HP flare *Note additional images of :flares were collected by Providence Photonics as detailed in the report in Appendix A. Page 4 of 4 Facility Name: Kensington Processing Plant Facility Location: 11543 State Route 644, Kensington, OH Date oflnspection: September 20, 2018 APPENDIX A: FJELD NOTES AND FIELD MEASUREMENT DATA Video Imaging Spectro-Radiography Report provided by Providence Photonics Appendices Page 1 of 1 Flare VISR Test Report Facility: Kensington Process Plant September 2018 Prepared for US Environmental Protection Agency Region 5 77 West Jackson Boulevard Chicago, IL 60604 PROVIDENCE PHOTONICS PROJECT NO. 1198-001 PREPARED BY Providence Photonics, LLC I1201 Main Street, Baton Rouge, LA 70802 . . PROVIDENCE WPHOTONICS Proj ect No. 1198-001 US EPA Regio n 5 Table of Contents INTRODUCTION ...................................................................................................................................... 2 SUMMARY OF VISR TEST RESULTS AND OBSERVATIONS ........................................................................ 2 Test Results and Observati ons - Tank Flare ... .. ....... .. ... .. ..... .... ...... ...... ... ............................ .......... ....... .... ... ...... .. 4 Test Results and Observat ions - Eme rgen cy Flare ...... .... ..... ... .... ...................... ........ .... .... ..... ... ......................... 8 Test Results and Obse rvation s - Hig h Pressure (HP) Flare...... .. .. ... ....... ...... ............ ........ .......... .. ....... ....... ... .... . 10 ELECTRONIC DATA GENERATED FROM THE TEST .................................................................................. 13 REFERENCES ......................................................................................................................................... 13 Revised: 10/11/2018 5 :43:00 PM Page 1 !13 Project No. 1198-001 US EPA Regio n 5 V ISR Flare Testing Introduction EPA Region 5 retained Providence Photonics, LLC (Providence) to assist the inspection of Kensington Processing Plant (Kensington Plant) located at 11543 State Route 644, Kensington, Ohio 44427. Providence measured the combustion efficiency (CE) and other performance metrics of the three flares at the Kensington Plant using the Video Imaging Spectral Radiometry (VISR) method. The VISR method utilizes a mu lti-spectral mid-wave infrared imager to remotely and continuous ly measure relative concentrations of combustion gases. Based on these relative concentrations, flare CE is calculated. The VISR method provides continuous measurement of the following five flare performance metrics at 1-second data rate. Combustion Efficiency (CE): This is the primary metric that VISR measures . CEis the concentration of unburned hydrocarbons relative to the concentration of carbon dioxide (CO2, the end product of combustion) in the post-combustion gases. CE is expressed in percentage. If combustion is complete and there are no unburned hydrocarbons in the post-combustion gases, CE wi ll be 100%. Smoke Index (SI): SI is used to indicate the level of smoke in the fla re flame. It has a range of 0 to 10. When flare has no smoke, the SI is very low (generally below 1 to 2) . As smoke level increases, the SI monotonically (but not linearly) increases. Flame Stability (FS): FS is used to measure the degree of flare radiance fluctuation within 1 second, which indicates stability of the flare. It has a range of 0-100. When the FS value is in about 80 or higher, it generally indicates a very stable flare. Flame Footprint (FF): FF represents the flame cross section area (in square feet) of the flare and associated hot post-combustion gases perpendicular to the VISR instrument. The footprint is based on the Infrared radiance, and it is generally larger than the size of the f lame observed in the visible spectrum by the naked eye. Heat Release (HR). HR measures the amount of heat released by the flare in regions of the midwave Infrared (MWIR) spectrum monitored by the VISR instrument. HR is reported as Btu/min. Because the HR only represents the energy released by the flare in the MWIR, the total heat released by the flare is larger than the HR metric reported by VISR. More information about the VISR method and its validation is provided at https://www.providencephotonics.com/flare-monitoring. Summary of VISR Test Results and Observations Date: September 20, 2018 Personnel: Gregory Gehrig (EPA Region 5) Kenneth Ruffatto (EPA Region 5) Cary Secrest (EPA OECA) Yousheng Zeng (Providence) Jon Morris (Providence) Sri Mutyala (Providence) Revised: 10/11/2018 5:43:00 PM Page 2 I13 Facility: Kensington Processing Plant 11543 State Route 644, Kensington, Ohio 44427 Location of Flares and VISR Instrument: see Figures 1-3. Project No. 1198-001 US EPA Region 5 VISR Flare Test ing Figure 1. Tank Flare. Left: Google map with a red circle indicating the location of the flare and a red number "1" indicating the location of the VISR instrument. Distance: 150 ft. Right: A photo of the flare ta ken during the test. Figure 2. Emergency Flare. Left: Google map with a red circle indicating the location of the flare and a red number "2" indicating the location of the VISR instrument. Distance: 213 ft. Right: A photo of the fla re taken during the test. Rev ised: 10/n/2018 5=43:00 PM Page 3 I13 Project No. 1198-001 US EPA Regio n 5 VISR Flare Testi ng Figure 3. High Pressure (HP) Flare. Left: Google map with a red circle indicating the location of the flare and a red number "3" indicating the location of the VISR instrument. Distance: 369 ft . Right: A photo of the flare taken during the test. Test Results and Observations - Tank Flare Test results for the Tank Fla re are summarized in Table 1. Variability (or trending) of the flare during the test can be seen in Figures 4-6. According to facility personnel, this flare is in idle condition most oftime with little vent gas being sent to the flare (only pilots are lit). Figure 7 provides a snapshot of the spectral image and a visible image of the flare when it was in the idle condition. When the flare was idling, only the pilot flame was visible from the measurement position and some of these frames did not meet data quality criteria set for this test. Therefore, some 1-second CE data points were missing in Figure 4. When tank pressure is built up to a set value, excess gas is vented to the flare (see Figure 8) and the flaring lasts less than 1-2 minutes. During the inspection, the interval between the flaring events was observed to be about 20-30 minutes. There was one flaring event which occurred while the VISR instrument was being set up. We conducted measurement for nearly 25 minutes. Towards the end of this 25-minute period, there was another flaring event - see the spikes near the right end of the FF and HR chart in Figure 5. Figure 8 includes images that correspond to the spikes (flaring) in Figure 5. Revised: 10/11/2018 5:43:00 PM Page 4 [ 13 Proj ect No . n 98-001 US EPA Regi on 5 VI SR Fl are Testin g Table l. VISR test results for the Tank Flare Average for the Duration (2 :05 pm - 2:30 pm) 1-sec (Avg.-SD, Avg.+SD) Combustion Efficiency(%) 99.4 (97 .0, 100.0)* Smoke Index (unitless) 1.12 (0.02, 2.22) Flame Stability (unitless) 92 .2 (85 .7, 98.8) Flame Footprint (ft2) 2.39 (O, 8.25)* * Heat Release (Btu/min) 31 (0, 144)* * *: When (Avg.+SD) >100, the value of 100 is assigned to t he upper bound. * * : When (Avg.-SD) <0, the value of Ois assigned to the lower bound . 1-min (Avg.-SD, Avg.+SD) (99.1, 100.0)* (0.41, 1.89) (88.6, 95 .2) (0, 8.00) * * (O, 147)** 100.0 90.0 80.0 Ill 70.0 LL. --"C 60.0 C: ra 50.0 ";:R w 40.0 u 30.0 20.0 10.0 0.0 N 0 0 -0 :i: Kensington - Tan k Flare (1-sec results) fT _ ---r -"--'1LJnI-VllAII1" 1 I N N N N N 0 ~ r-:, N w u, 0 u, 0 u, 0 -0 -0 -0 -0 -0 ~ N-0 s s s $ s s Local Time CE (1-s) - FS(l-s) - SI (1-s) Figure 4. Tank Flare. Time series plot of 1-second results fo r CE, FS, and SI. 7 .00 - 6 .00 - 5.00 4.00 vi 3 .00 2.00 1.00 0.00 N w u, -0 s Revised: 10/11/2018 5:43:00 PM Page 5 [ 13 Project No. 11.98-001 US EPA Region 5 VISR Flare Testing 45.00 40.00 35.00 - 30.00 ,:I= . 25.00 C" ~ 20.00 u. u. 15.00 10.00 5.00 0.00 N 0 0 "Cl s Kensington - Tank Flare (1-second results} _________,.., s s s s s sN N N N N N a I-" I-" N N "Tl VJ "Cl "Cl "Cl "Cl "Cl V, 0 V, 0 V, iii" "Cl 0 :::! . .:::J (IQ Local Time - FF(l-s) - HR (1-s) Figure 5. Tank Flare. Time series plot of 1-second results for FF and HR. 700 600 500 -C: .. 400 ..E..... .:...l. 300 co -ex: ::J: 200 100 0 N VJ u, "Cl s 100.0 99.0 98.0 97.0 96.0 ~ 95.0 w u 94.0 93.0 92.0 91.0 90.0 N 0 0 "Cl s Flare CE in Different Averaging Periods Kensington - Tank Flare N N N N 0 I-" I-" N u, 0 u, 0 N N N w V, 0 "Cl "Cl -0 "O -0 "Cl s s s s s s Local Time - CE (1-s) - CE (1-m) - cE {15-m) - CE (duration) Figure 6. Tank Flare. Time series plot of CE with different averaging periods. Revised: 10/11/2018 5:43:00 PM N w Ln "Cl s Page 6 [ 13 Project No. 1198-001 US EPA Region 5 VISR Fla re Testing Figure 7. Tank Flare (idle condition). Left: A representative composite spectral image generated by VISR instrument. Red color represents CO2; green color represents hydrocarbon; and blue or bluish white color represents aerosol (e.g., soot) or hot solid object (e.g., the shield for the pilots). Right: A representative image in visible spectrum. Figure 8. Tank Flare (flaring condition) . Left: A representative composite spectral image generated by VISR instrument. Red colo r represents CO2; green color represents hydrocarbon; and blue or bluish white color represents aerosol (e.g ., soot) or hot solid object (e.g., the shield for the pilots) . Right: A representative image in visible spectrum . Revised : 10/11/2018 5:43:00 Pfvl Page 7 [ 13 Proj ect No. 1198 -001 US EPA Region 5 VISR Flare Testing .Test Results and Observations - Emergency Flare Test results for the Emergency Flare (North Flare) are summarized in Table 2. Vari ability (or trending) of the flare during the test can be seen in Figures 9-11. Fig ure 12 provides a snapshot of the spectral image and a visible image ofthe flare. During the test only the flare pilots were detected (see Figure 12). Table 2. VISR test results for t he Emergency Flare Average for the Duration (2 :56 pm - 3:02 pm) Combust ion Efficiency(%) 99.3 Smoke Index (unitless) 1.46 Flame Stabi lity (unitless) 86.9 Flame Footprint (ft2) 1.86 Heat Release (Btu/ min) 8 1-sec Avg. SD (98.6, 99.9) (1.07, 1.85) (77.6, 96.2) (1.23, 2.50) (5, 11) 1-min Avg. SD {98.8, 99.8) (1.17, 1.74) (83.8, 89.8) (1.37, 2.32) (6, 10) 100.0 90.0 80.0 V, 70.0 LL. l"C 60.0 C: (ti 50.0 w 40.0 u 30.0 20.0 10.0 0.0 N u-, u-, "O ~ Kensington - Emergency Flare (1 -sec results} --- N N N N (JJ r.,.J r.,.J u-, u-, V, Ln 0 en -..J 00 \0 0 0 0 I-' N "O "O "O "O "O "O "O ~ ~ ~ ~ ~ ~ ~ Local Time CE (1-s) - FS (1-s) - SI (1-s) Figure 9. Emergency Flare. Time series plot of 1-second results for CE, FS, and SI. 2.50 2.00 1.50 1.00 - 0.50 0.00 Revised : 10/11/2018 5:43:00 PM Page8 l 13 Project No . 1198-001 US EPA Region 5 VISR Flare Testing 3.50 3.00 2.50 E 2.00 ci-;":;':- 1.50 u. 1.00 0.50 0.00 N u, u, "O s Kensington - Emergency Flare (1-second results} N u, 00 "O s Local Time - FF (1-s) - HR (1-s) Figure 10. Emergency Flare. Time series plot of 1-second results for FF and HR. 16 14 12 - 10 C: E -8 .:.:.I re 6 0:: ::i::: 4 2 0 Flare CE in Different Averaging Periods Kensington - Emergency Flare 100.0 99.0 98.0 97.0 -~ 96.0 95.0 LU u 94.0 f 92.0 93.0 I 91.0 90.0 N N N N N u., u., u., u, u, u, u, V, 0 u, (J) -..J 00 \D 0 0 0 ~ N "O "O "O "O '"CJ "O "O "O s s s s s s s s Local Time - CE (1-s) - CE (1-m) - CE (duration) Figure 11. Emergency Flare. Time series plot of CE with different averaging periods. Revised: 10/1.1/2018 5:43:00 PM Page9!13 Project No. 1198-001 US EPA Region 5 VISR Flare Testing Figure 12. Emergency Flare. Left: A representative composite spectral image generated by VISR instrument. Red color represents CO2; green color represents hydrocarbon; and blue or bluish white color represents aerosol (e.g., soot) or hot solid object (e.g., the shield for the pilots). Right: A representative image in visible spectrum. Test Results and Observations - High Pressure (HP) Flare Test results fo r the High Pressure (HP) Flare are summarized in Table 3. Variability (or trending) of the flare during the test can be seen in Figures 13-15. Figure 16 provides a snapshot of the spectral image and a visible image of the flare. Table 3. VISR test results for the HP Flare Average for the Duration (3:20 pm - 3:49 pm) Combustion Efficiency{%) 98.9 Smoke Index (unitless) 1.62 Flame Stability (unitless) 94.6 Flame Footprint (ft2) 76.91 Heat Release (Btu/min) 719 1-sec Avg. SD (98.1, 99.7) (0.90, 2.34) (89.4, 99.7) {69.55, 84.27) (634,805) 1-min Avg. SD (98.4, 99.3) (1.22, 2.03) {93.2, 95.8) (71.22, 82.15) (653, 780) Re vised: 10/1.J./2018 5:43: 00 PM Page 10 [ 13 Project No. 1198-0 01 US EPA Region 5 VISR Fl are Testing 100.0 90.0 80.0 V') 70.0 u.. -e"'C 60.0 C: !'Cl 50.0 40.0 LU u 30.0 20.0 10.0 0.0 (.>.) i-'.>, u, -0 s Kensington - HP Flare (1-sec results) w w (.>.) (.>.) (.>.) (.>.) (.>.) (J,,) (J,,) N N 0 V, 0 :i:;. :i:;. u, V, 0 V, 0 -0 -0 "'CJ "'CJ "'CJ "'CJ "'CJ s s s s s s s Local Time CE (1-s) - FS (1-s} - S I (1 -s) Figu re 13. HP Flare. Time series plot of 1-second resu lts for CE, FS, and SI. 4 .50 4 .00 3.50 3 .00 2 .50 vi 2.00 1.50 1 .00 - 0.50 0.00 (J,,) u, V, "'CJ s 120.00 100.00 - 80.00 ~ O" 60.00 ~ u.. u.. 40.00 20.00 0.00 (J,,) I-" V, "'CJ s Kensington - HP Flare (1-second results} w w (J,,) (.>.) (J,,) (J,,) (J,,) (J,,) (J,,) N i0 0 V, 0 :i;. :i:;. V, V, 0 V, 0 "'CJ "'CJ "'CJ "'CJ -0 -0 "'CJ s s s s s s s Local Time - FF (1-s) - HR (1-s) Figure 14. HP Flare. Time series plot of 1-second results for FF and HR. Revised: 10/11/2018 5 :43:00 PM 1,200 1,000 -800 "2 E 600 :;:l +- c:o 0::: 400 ~ 200 0 f,)J V, V, "'CJ s Page 11 \ 13 Project No. 1198-001 US EPA Region 5 VISR Fl are Testing Flare CE in Different Averaging Periods Kensington - HP Flare 100.0 99.0 98.0 - 97.0 e 96.0 9 5.0 LU u 94.0 93.0 92.0 91.0 90.0 .... w w (.>J (.>J (.>J w (.v w (.>J w w r--:i r--:i ln 0 V, 0 :i:, :i:, V, V, V, 0 1./1 0 1./1 -0 -0 -0 -0 -0 ,::, ,::, -0 -0 s: $ $ s s: s $ s s Local Time - CE(1-s) - CE (1-m) ..,_CE (15-m) - CE (duration) Figure 15. HP Flare. Time series plot of CE with different averaging periods. Figure 12. HP Flare. Left: A representative composite spectral image generated by VISR instrument. Red color represents CO2; green color represents hydrocarbon; and blue or bluish white color represents aerosol (e.g., soot) or hot so lid object (e.g., the shie ld for the pilots). Right: A representative image in visible spectrum. Revised: 10/n/2018 5:43 :00 PM Page 12 I13 Project No. 1198-001 US EPA Region 5 VISR Flare Testing Electronic Data Generated from the Test The following electronic data are submitted to the US EPA Region 5 as part of this report: A Microsoft Excel file containing second-by-second results from the VISR instrument, derived 1minute average results, and 15-minute average results when applicable. Video recording of the flare during the test, taken by a regular video camera in visible spectrum. Spectral video generated by the VISR instrument. References 1. Va lidation of a New Method for Measuring and Continuously Monitoring the Efficiency of Industrial Flares, Y. Zeng, J. Morris, M . Dombrowski, J. Air & Waste Manage. Assoc. 66:76-86 (2016). 2. Performance Review of Remote and Direct Flare Combustion Efficiency Monitor, Y. Zeng and J. Morris, presented at American Fuel & Petrochemical Manufacturers {AFPM) 2017 Environmental Conference, Denver, Colorado, October 15-17, 2017 3. Video Imaging Spectral Radiometer {VISR) Based Flare Monitor, Y. Zeng, presented at Air Quality Measurement Methods and Technologies, Long Beach, California, November 7-9, 2017 More information about the VISR method and its validation can be found at https://www.providencephotonics.com/flare-mon itoring. Revise d: 10/11/2018 5:43:00 PM Page 13 I13