Refinery Sector Rule Focus on Flare Combustion Control Parameters

Paper from the AFRC 2016 conference titled Refinery Sector Rule Focus on Flare Combustion Control Parameters

Recent United States Environmental Protection Agency (USEPA) regulations included in the; Refinery Sector Rule (RSR) modifications concerning flares have focused on improving; combustion efficiency at refinery flares. As part of these new regulations, the USEPA has; required facilities to comply with a Net Heating Value in the combustion zone (NHVcz) for each; affected flare. This compliance point requires active control of a flare that has not been required; previously. The RSR provided two methods for showing compliance with the NHVcz, the direct; method and the feed forward method. Each method has details that need to be considered and; the flare monitoring technologies chosen will likely drive the decision to which option a facility; utilizes. The presentation will provide a brief overview of the new flare requirements of the RSR; and will provide an in depth discussion of the flare control options.

AFRC 2016 Industrial Combustion Symposium Kauai, HI September 2016 Title - Refinery Sector Rule Focus on Flare Combustion Control Parameters Author: Herman Holm Client Guardian Sage ATC Environmental Consulting, LLC Co-Author: Kevin Lingard Project Technical Sage ATC Environmental Consulting, LLC Summary Recent United States Environmental Protection Agency (USEPA) regulations included in the Refinery Sector Rule (RSR) modifications concerning flares have focused on improving combustion efficiency at refinery flares. As part of these new regulations, the USEPA has required facilities to maintain a minimum Net Heating Value in the combustion zone (NHVcz) for each affected flare. This compliance point requires active control of a flare that has not been required previously. The RSR provided two methods for showing compliance with the NHVcz, the direct method and the feed-forward method. Each method has details that need to be considered and the flare monitoring technologies chosen will likely drive the decision to which option a facility utilizes. The presentation will provide a brief overview of the new flare requirements of the RSR and will provide an in depth discussion of the flare control options. Regulatory Overview Regulatory requirements for refinery flares could be comprised of Federal and State requirements. The Federal requirements were found in the New Source Performance Standards (NSPS) Subpart A at 40 CFR §60.18 and in the National Emission Standards for Hazardous Air Pollutants (NESHAPs) Subpart A at 40 CFR §63.11. These regulations contained similar requirements for proper flare operation and were utilized by refineries as a means to show that a 98% destruction removal efficiency (DRE) was achieved. Flares were expected to:     Be designed and operated with no visible emissions, except for a total of five minutes during any two consecutive hours; Be operated with a flame present at all times; Meet the heat content requirements in the vent gas (multiple options based on the type of flares); and, Operate below the exit velocity requirements (multiple options based on the type of flares). These operating requirements were determined to be sufficient and were utilized from the mid1980s to present, with some minor revisions through the years. However, over the past 10 years, new information has been found that indicates that the previous parameters are not sufficient. This insufficiency has stemmed from the lack of inclusion of the effects of assist gas (steam or air) on combustion efficiency. Through testing completed by the Texas Commission on Environmental Quality (TCEQ) and the University of Texas (UT) in 2011 and testing performed at industrial facilities through the Consent Decree process, etc., parameters have been found that can be used to better determine the DRE of refinery flares that take into account the vent gas and assist gas. On June 30, 2014, the USEPA published the proposed Petroleum Refinery Sector Rule (RSR) which included changes to 40 CFR 63 Subpart CC (Refinery MACT 1) and Subpart UUU (Refinery MACT 2). Changes to the regulations included the requirement for affected units that utilized flares as control devices must meet certain operating criteria in order to demonstrate 98% DRE. The changes proposed were modified and finalized in the rule published on December 1, 2015. The requirements that a flare must meet were included in 40 CFR 63 Subpart CC (MACT CC) in sections §63.670 and §63.671. Similar to the requirements found in §60.18 and §63.11, MACT CC includes the following requirements:      Be designed and operated with no visible emissions, except for a total of five minutes during any two consecutive hours when the flare is operating below its smokeless capacity and regulated material is sent to the flare; Be operated with a pilot flame present at all times when regulated material is sent to the flare; Meet the combustion zone operating limits when regulated material is sent to the flare for at least 15 minutes; Meet the dilution operating limits for flares with perimeter assist air and, Operate below the exit velocity when regulated material is sent to the flare. The pilot flame presence, combustion zone operating limits, and exit velocity are determined on a 15-minute block that runs with the clock (i.e., 12:00 am to 12:15 am, 12:15 am to 12:30 am, etc.). If a refinery flare is meeting the flare requirements as provided in MACT CC, the flare does not need to meet the requirements in §60.18 or §63.11. However, if the flare operates in a State with its own specific flare operating requirements, the flare will be required to continue to comply with those requirements as well. In §63.670(l), MACT CC prescribes the calculation methods for determining flare vent gas net heating value. In §63.670(l)(5) (i) and (ii), the regulation details the feed-forward and direct calculation methods, respectively. The principle behind the feed-forward method is to utilize the Net Heating Value of the vent gas (NHVvg) obtained in one 15-minute period and apply it to the subsequent 15-minute period. Additionally, the feed-forward method allows the amount of supplemental natural gas added to the flare in the previous 15-minute block and the current 15minute block to be factored into determining compliance. The equation to determine the NHVcz utilizing the feed-forward method from §63.670(m)(2) is shown in Equation 1: Eq. (1) Where: NHVcz = Net heating value of combustion zone gas, British thermal unit (Btu)/ standard cubic feet (scf). NHVvg = Net heating value of flare vent gas for the 15-minute block period, Btu/scf. Qvg = Cumulative volumetric flow of flare vent gas during the 15-minute block period, scf. QNG2 = Cumulative volumetric flow of supplemental natural gas to the flare during the 15-minute block period, scf. QNG1 = Cumulative volumetric flow of supplemental natural gas to the flare during the previous 15minute block period, scf. For the first 15-minute block period of an event, use the volumetric flow value for the current 15-minute block period, i.e., QNG1=QNG2. NHVNG = Net heating value of supplemental natural gas to the flare for the 15-minute block period determined according to the requirements in paragraph (j)(5) of this section, Btu/scf. Qs = Cumulative volumetric flow of total steam during the 15-minute block period, scf. Qa,premix = Cumulative volumetric flow of premix assist air during the 15-minute block period, scf. For the direct calculation method, the average NHVvg of the current 15-minute period is utilized to show compliance for that 15-minute period and values are not carried forward to adjacent 15minute periods. The equation to determine the NHVcz utilizing the direct calculation method from §63.670(m)(1) is shown in Equation 2: Eq. 2 Where: NHVcz = Net heating value of combustion zone gas, Btu/scf. NHVvg = Net heating value of flare vent gas for the 15-minute block period, Btu/scf. Qvg = Cumulative volumetric flow of flare vent gas during the 15-minute block period, scf. Qs = Cumulative volumetric flow of total steam during the 15-minute block period, scf. Qa,premix = Cumulative volumetric flow of premix assist air during the 15-minute block period, scf. The refinery must notify the USEPA of which method it will use to show compliance. If the refinery wishes to change the method of calculation, the USEPA must be notified 30 days prior to the change. Calculation methodologies are also included in the regulation to determine the dilution parameters for flares that receive perimeter assist air. These calculations are similar to the equations presented above with the exception that the cumulative flow of perimeter assist air is also included in the denominator of each equation. This paper will only focus on flares that do not receive perimeter assist air. Two items are important to note at this point. The first is that the flow rates provided in the equations are cumulative flow rates, which means that compliance during the 15-minute block cannot be determined until the end of the time period. The second is that these equations only prescribe how compliance is calculated and MACT CC does not detail how control of the flare system is to be accomplished. How the control strategy functions is left to the refinery to determine. Control Strategy Implications As compliance with the combustion zone operating parameters are required to be met on 15minute block averages, the speed at which sample data from the vent gas are obtained is anticipated to impact the ability to comply and impact which calculation method is chosen. There are a limited number of analyzers that have been utilized in flare service to determine the NHVvg. The analyzers fall into two general categories: analyzers that speciate the compounds in the vent gas and analyzers that provide the NHVvg only. Based on data obtained from the Flare Management Plans (FMPs) submitted as part of NSPS Ja compliance, two typical analyzers used to speciate the vent gas are gas chromatographs (GCs) and mass spectrometers (MS). Btu analyzers or calorimeters have also been installed on flares in order to only provide the NHVvg. These technologies have pros and cons; however, one of the biggest differentiators when it comes to data is the speed at which the analysis is performed. To perform the analysis for the required hydrocarbons and inert compounds, typical GC analysis times range from six to 15 minutes and MS analysis times are typically one to two minutes. A calorimeter can provide Btu data on an approximately minutely basis. In order to determine how analysis time would impact the two calculation methods, scenarios were developed to determine at what time lag would compliance become problematic for each method. For these scenarios, we chose a flare with the natural gas addition point, equipped with a dedicated flow meter, upstream of the analyzer and vent gas flow meter. The flare was setup as a typical 24 to 30-inch tip that would receive 500 lb/hr of minimum cooling steam (178 standard cubic feet per minute [scfm]), and a typical steam to waste gas ratio of 0.5 lbsteam/lbwg. Waste gas is the vent gas stream without any supplemental fuel included. Flare operations will be highly dynamic; however, it is anticipated that the need to add supplemental gas will be limited in scenarios as the use of nitrogen and steam to the flare are typically limited to startup, shutdown, and maintenance activities. Therefore, the scenario was also simplified to approximate the release of a low Btu gas to the flare similar to depressuring a vessel, with a rapid rise in flow that trails down in magnitude as the pressure would decrease in the vessel that corresponds to a swift decrease in NHVvg that trails upward as the gas decreases in flow. The flare is a continuous flare that does not have a water seal. The scenario, which we will designate as the Rapid Scenario, is represented graphically in Figure 1. 800 600 2000 400 1000 200 0 0 0:00 0:15 0:30 0:45 1:00 1:15 1:30 1:45 2:00 NHVwg (Btu/scf) Qwg (scfm) 3000 Time Qwg (scfm) NHVwg (Btu/scf) Fig 1: Rapid Scenario for Waste Gas Flow and NHV As the flare is a continuous flare, the scenario begins with the flare in compliance with NHVcz without the need to add supplemental fuel. Over the second 15-minute block, the waste gas flow rate is increased from 250 scfm to 2,000 scfm with a corresponding drop in NHV of the waste gas (NHVwg, NHV of the vent gas minus any contribution of the supplemental fuel). The waste gas due to the vessel pressuring was estimated to be 48,000 scf of 50 Btu/scf material. The time interval between analytical samples were then simulated to determine whether the direct method would be able to remain in compliance. Figures 2, 3, and 4 illustrate a 1-, 7- and 15-minute sample time. 700 600 2000 500 1500 400 1000 300 200 500 100 0 0 0:00 0:15 0:30 0:45 1:00 1:15 1:30 1:45 2:00 Time Qwg (scfm) Qng (scfm) NHVcz (Btu/scf) NHVcz = (270 Btu/scf) NHVwg (Btu/scf) Fig 2: Direct Calculation Method - Rapid Scenario - 1-Minute Analytical Time NHV (Btu/scf) Flow Rate (scfm) 2500 700 600 2000 500 1500 400 1000 300 200 500 NHV (Btu/scf) Flow Rate (scfm) 2500 100 0 0 0:00 0:15 0:30 0:45 1:00 1:15 1:30 1:45 2:00 Time Qwg (scfm) Qng (scfm) NHVwg (Btu/scf) NHVcz (Btu/scf) NHVcz = 270 (Btu/scf) Fig 3: Direct Calculation Method - Rapid Scenario - 7-Minute Analytical Time 700 600 2000 500 1500 400 1000 300 200 500 NHV (Btu/scf) Flow Rate (scfm) 2500 100 0 0 0:00 0:15 0:30 0:45 1:00 1:15 1:30 1:45 2:00 Time Qwg (scfm) Qng (scfm) NHVcz (Btu/scf) NHVcz = (270 Btu/scf) NHVwg (Btu/scf) Fig 4: Direct Calculation Method - Rapid Scenario - 15-Minute Analytical Time Figure 2 shows that the direct method was able to add supplemental fuel in sufficient quantity and quickly enough to remain in compliance for the 15-minute block periods. When the analytical time is increased to a 7-minute run time, the method is not able to stay in compliance for one of the 15-minute blocks utilizing a set point of 270 Btu/scf as shown in Figure 3. However, if the set point for control of the NHVcz was increased to 432 Btu/scf, the method would show compliance for all the 15-minute blocks associated with the scenario. The analytical time was further increased to 15 minutes as shown in Figure 4. Figure 4 shows that the direct method was not able to add sufficient natural gas quickly enough and was out of compliance for two 15-minute blocks. The set point would need to be increased to 523 Btu/scf to remain in compliance. For the Rapid scenario created, the elevated set point for the 7- and 15-minute analysis times would result in supplemental natural gas being continually added to the flare in order to be able to respond to a flaring event of this magnitude. The scenario in Figure 4 had the sample result being reported in the 12th minute of each block. A fourth scenario was run to see if having the result reported in the beginning of the 15-minute block would impact the ability to remain in compliance. When the sample time was changed to the 2nd minute of each block, the direct method was not able to remain in compliance for one block. The set point would need to be increased to 334 Btu/scf in order to remain in compliance. Interestingly, this new elevated set point does not require the supplemental natural gas to be continuously supplied. This comparison indicates that the timing of the event, with the timing of the analytical results, can have large impacts to the system. 1000 800 600 500 400 200 0 0 0:00 0:15 0:30 0:45 1:00 1:15 1:30 1:45 2:00 NHVwg (Btu/scf) Qwg (scfm) The system was then tested to see if the waste gas was not released as quickly to the flare, would the direct method be able to respond better to the event and remain in compliance. The total volume of flared gas was held constant at 48,000 scf of 50 Btu/scf material. The scenario is shown in Figure 5 and is referred to as the Slow Scenario. Time Qwg (scfm) NHVwg (Btu/scf) Fig 5: Slow Scenario for Waste Gas Flow and NHV Again the system was tested at a 1-minute analysis time as shown in Figure 6. 700 600 500 400 300 200 NHV (Btu/scf) Flow Rate (scfm) 1000 900 800 700 600 500 400 300 200 100 0 100 0 0:00 0:15 0:30 0:45 1:00 1:15 1:30 1:45 2:00 Time Qwg (scfm) Qng (scfm) NHVwg (Btu/scf) NHVcz (Btu/scf) NHVcz = (270 Btu/scf) Fig 6: Direct Calculation Method - Slow Scenario - 1-Minute Analytical Time 1000 900 800 700 600 500 400 300 200 100 0 700 600 500 400 300 200 NHV (Btu/scf) Flow Rate (scfm) Figure 6 shows that the direct method was able to add supplemental fuel in sufficient quantity and quickly enough to remain in compliance for the 15-minute block periods. When the analytical time is increased to a 7-minute run time, the method is not able to stay in compliance for two of the 15-minute blocks utilizing a set point of 270 Btu/scf as shown in Figure 7. 100 0 0:00 0:15 0:30 0:45 1:00 1:15 1:30 1:45 2:00 Time Qwg (scfm) Qng (scfm) NHVcz (Btu/scf) NHVcz = 270 (Btu/scf) NHVwg (Btu/scf) Fig 7: Direct Calculation Method - Slow Scenario - 7-Minute Analytical Time However, if the set point for control of the NHVcz was increased to 295 Btu/scf, the method would show compliance for all the 15-minute blocks associated with the scenario. The analytical time was further increased to 15 minutes as shown in Figure 8. 700 600 500 400 300 200 NHV (Btu/scf) Flow Rate (scfm) 1000 900 800 700 600 500 400 300 200 100 0 100 0 0:00 0:15 0:30 0:45 1:00 1:15 1:30 1:45 2:00 Time Qwg (scfm) Qng (scfm) NHVcz (Btu/scf) NHVcz = (270 Btu/scf) NHVwg (Btu/scf) Fig 8: Direct Calculation Method - Slow Scenario - 15-Minute Analytical Time Figure 8 shows that the direct method was not able to add sufficient natural gas quickly enough and was out of compliance for five 15-minute blocks. The set point would need to be increased to 461 Btu/scf to remain in compliance. For the scenario created, this elevated set point would result in supplemental natural gas being continually added to the flare in order to be able to respond to a flaring event of this magnitude. The scenario in Figure 8 had the sample result being reported in the 12th minute of each block. A fourth slow scenario was run to see if having the result reported in the beginning of the 15-minute block would impact the ability to remain in compliance. When the sample time was changed to the 2nd minute of each block, the direct method was not able to remain in compliance for four blocks. The set point would need to be increased to 321 Btu/scf in order to remain in compliance. Similarly, this new elevated set point does not require the supplemental natural gas to be continuously supplied. This comparison indicates that the timing of the event, with the timing of the analytical results, can have large impacts to the system. The various scenarios under the direct method are summarized in Table 1. Table 1 provides, for the Rapid Event and the Slow Event, for each analytical time, the number of 15-minute blocks the system was out of compliance with a set point of 270 Btu/scf and the volume of total supplemental fuel added. The table also provides the set point in order to show compliance with all the 15-minute blocks in the scenario with the additional and total amounts of supplemental fuel required. Finally, a 5-year historical cost of $0.00459/scf for natural gas was utilized to provide a cost for the supplemental fuel. Calculation Method Table 1: Comparison of Direct Calculation Method for Various Analytical Timeframes Scenario Direct Method Rapid Event Slow Event 1 Supplemental Fuel Required Set point to be in compliance for all 15minute blocks Additional Supplemental Fuel Required (scf) ($)1 (Btu/scf) (scf) ($)1 (scf) ($)1 1-Min 0 25,120 $115.30 270 0 $0.00 25,120 $115.30 2 1 18,614 $85.44 432 35,208 $161.60 53,822 $247.04 2 10,294 $47.25 523 59,642 $273.76 69,936 $321.01 1 23,729 $108.92 334 13,134 $60.29 36,863 $169.20 0 3 19,384 18,310 $88.97 $84.04 270 295 0 4,853 $0.00 $22.28 19,384 23,163 $88.97 $106.32 5 15,381 $70.60 461 47,064 $216.02 62,445 $286.62 4 19,392 $89.01 321 10,658 $48.92 30,050 $137.93 15-Min Late3 15-Min Early 1-Min 7-Min 15-Min Late4 15-Min Early Natural gas cost estimate is the past 5 year estimate. 0.00459 Total Supplemental Fuel Required for Compliance (No.) 7-Min Notes: No. 15minute blocks out of compliance with 270 Btu/scf set point $/scf at 920 Btu/scf 2 This scenario would also result in approximately 72 scfm of supplemental gas being added at all times. 3 This scenario would also result in approximately 220 scfm of supplemental gas being added at all times. 4 This scenario would also result in approximately 299 scfm of supplemental gas being added at all times. The feed-forward calculation was then utilized with the same Rapid scenario to see how the system reacts and to determine the amount of natural gas that is needed in order to remain in compliance. Figures 9, 10, and 11 show the 4-Minute, 7-Minute, and 15-Minute Rapid Scenarios. 2500 700 Flow (scfm) 500 1500 400 1000 300 200 500 NHV (Btu/scf) 600 2000 100 0 0 0:00 0:15 0:30 0:45 1:00 1:15 1:30 1:45 2:00 2:15 Time Qwg (scfm) Qng (scfm) NHVwg (Btu/scf) NHVcz (Btu/scf) NHVcz = 270 (Btu/scf) Fig 9: Feed-Forward Calculation Method - Rapid Scenario - 4-Min. Analytical Time 2500 700 Flow (scfm) 500 1500 400 1000 300 200 500 NHV (Btu/scf) 600 2000 100 0 0 0:00 0:15 0:30 0:45 1:00 1:15 1:30 1:45 2:00 2:15 Time Qwg (scfm) Qng (scfm) NHVcz (Btu/scf) NHVcz = 270 (Btu/scf) NHVwg (Btu/scf) Fig 10: Feed-Forward Calculation Method - Rapid Scenario - 7-Min. Analytical Time 700 600 500 400 300 200 100 0 -100 -200 -300 Flow (scfm) 2000 1500 1000 500 0 0:00 0:15 0:30 0:45 1:00 1:15 1:30 1:45 2:00 2:15 NHV (Btu/scf) 2500 2:30 Time Qwg (scfm) Qng (scfm) NHVwg (Btu/scf) NHVcz (Btu/scf) NHVcz = 270 (Btu/scf) Fig 11: Feed-Forward Calculation Method - Rapid Scenario - 15-Min. Analytical Time 1000 900 800 700 600 500 400 300 200 100 0 700 600 500 400 300 200 NHV (Btu/scf) Flow (scfm) Based on the results of the three timeframes for the Rapid Scenario, with a set point of 270 Btu/scf, the 4-, 7-, and 15-minute analysis times did not show compliance for two, two, and three 15-minute blocks, respectively. In order to show compliance for all the 15-minute blocks, the set point would need to be increased to 541, 537, and 663 Btu/scf for the 4-, 7-, and 15-minute analysis times, respectively. The feed-forward calculation was then utilized with the same Slow scenario to see how the system reacts and to determine the amount of natural gas that is needed in order to remain in compliance. Figures 12, 13, and 14 show the 4-Minute, 7-Minute, and 15Minute Slow Scenarios. 100 0 0:00 0:15 0:30 0:45 1:00 1:15 1:30 1:45 2:00 2:15 2:30 Time Qwg (scfm) Qng (scfm) NHVcz (Btu/scf) NHVcz = 270 (Btu/scf) NHVwg (Btu/scf) Fig 12: Feed-Forward Calculation Method - Slow Scenario - 4-Min. Analytical Time 700 600 500 400 300 200 NHV (Btu/scf) Flow (scfm) 1000 900 800 700 600 500 400 300 200 100 0 100 0 0:00 0:15 0:30 0:45 1:00 1:15 1:30 1:45 2:00 2:15 2:30 Time Qwg (scfm) Qng (scfm) NHVwg (Btu/scf) NHVcz (Btu/scf) NHVcz = 270 (Btu/scf) 1000 900 800 700 600 500 400 300 200 100 0 700 600 500 400 300 200 NHV (Btu/scf) Flow (scfm) Fig 13: Feed-Forward Calculation Method - Slow Scenario - 7-Min. Analytical Time 100 0 0:00 0:15 0:30 0:45 1:00 1:15 1:30 1:45 2:00 2:15 2:30 2:45 3:00 Time Qwg (scfm) Qng (scfm) NHVcz (Btu/scf) NHVcz = 270 (Btu/scf) NHVwg (Btu/scf) Fig 14: Feed-Forward Calculation Method - Slow Scenario - 15-Min. Analytical Time Based on the results of the three timeframes for the Slow scenario, with a set point of 270 Btu/scf, the 4, 7, and 15-minute analysis times did not show compliance for three, five, and six 15-minute blocks, respectively. In order to show compliance for all the 15-minute blocks, the set point would need to be increased to 594, 558, and 511 Btu/scf for the 4-, 7-, and 15-minute analysis times, respectively. As part of the direct calculation method, the time at which the result was reported was varied to determine the impact on the system. The data shown in Figures 11 and 14 have the data reported in minute 2 of the 15-minute block. Having the data report at a later timeframe showed a similar trend to the direct method and is not presented. The various scenarios under the feed-forward method are summarized in Table 2. Table 2 provides, for the Rapid Event and the Slow Event, for each analytical time, the number of 15minute blocks the system was out of compliance with a set point of 270 Btu/scf and the volume of total supplemental fuel added. The table also provides the set point in order to show compliance with all the 15-minute blocks in the scenario with the additional and total amounts of supplemental fuel required. The set points determined to ensure compliance during the 15minute blocks would require supplemental fuel to be constantly added to the flare. These constant rates have also been provided. Finally, a 5-year historical cost of $0.00459/scf for natural gas was utilized to provide a cost for the supplemental fuel. Feed-Forward Method Calculation Method Table 2: Comparison of Feed-Forward Calculation Method for Various Analytical Timeframes Notes : Scenario Rapid Event Slow Event 1 No. 15minute blocks out of compliance with 270 Btu/scf set point Supplemental Fuel Required Constant Supplemental Fuel Addition for Compliance Additional Supplemental Fuel Required (scf) ($)1 (Btu/scf) (scfm) (scf) 4-Min 7-Min 2 2 23,932 18,614 $109.85 $85.44 541 537 215 208 98,666 86,064 $452.88 122,598 $562.72 $395.03 104,678 $480.47 15-Min 3 14,618 $67.10 663 520 199768 $916.94 214,386 $984.03 4-Min 7-Min 3 5 18,740 18,310 $86.02 $84.04 594 558 320 245 141,135 $647.81 159,875 $733.83 111,909 $513.66 130,219 $597.71 15-Min 6 17,943 $82.36 511 168 84,763 $/scf at 920 Btu/scf ($)1 Total Supplemental Fuel Required for Compliance (No.) Natural gas cost estimate is the past 5 year estimate. 0.0045 Set point to be in compliance for all 15minute blocks (scf) ($)1 $389.06 102,706 $471.42 The previous cases were for a continuous flare. Many flares are equipped with water seals and flare gas recovery (FGR) systems that enable a flare to operate without regulated material being sent to the flare when the water seal is intact. For these types of flares, when the water seal is intact, the flare is expected to operate with the minimum purge gas (natural gas) and the minimum steam addition. At these minimums, the flare is expected to not meet the required 270 Btu/scf in the combustion zone, but it is not required to as regulated material is not being sent to the flare. Once the water seal is broken and at least 15-minutes of flaring occurs, the system will need to be in compliance with the combustion zone limits. The system was evaluated for the direct method with 1- and 7-minute analysis times and the feed-forward method for the 7-minute analysis time. The system was also evaluated to see how the start time of the event related to the 15-minute block affected the results. Figures 15, 16, and 17 illustrate the direct method for the 1-minute direct, 7-minute direct, and 7-minute feed-forward methods, respectively. 2500 1000 900 2000 800 700 1500 600 500 1000 400 300 500 200 100 0 0 0:00 0:15 0:30 0:45 1:00 1:15 Qwg (scfm) Qng (scfm) NHVcz (Btu/scf) NHVcz = 270 (Btu/scf) 1:30 1:45 2:00 2:15 NHVwg (Btu/scf) Fig 15: Direct Calculation Method - Water Seal Scenario - 1-Minute Analytical Time 1000 900 800 700 600 500 400 300 200 100 0 2000 1500 1000 500 0 0:00 0:15 0:30 0:45 1:00 1:15 1:30 1:45 2:00 NHV (Btu/scf) Flow Rate (scfm) 2500 2:15 Time Qwg (scfm) Qng (scfm) NHVwg (Btu/scf) NHVcz (Btu/scf) NHVcz = 270 (Btu/scf) 2500 1000 2000 800 600 1500 400 1000 200 500 NHV (Btu/scf) Flow (scfm) Fig 16: Direct Calculation Method - Water Seal Scenario - 7-Minute Analytical Time 0 0 -200 0:00 0:15 0:30 0:45 1:00 1:15 1:30 1:45 2:00 2:15 Time Qwg (scfm) Qng (scfm) NHVcz (Btu/scf) NHVcz = 270 (Btu/scf) NHVwg (Btu/scf) Fig 17: Feed-Forward Calculation Method - Water Seal Scenario - 7-Min. Analytical Time When the flaring event starts at the beginning of a 15-minute block, the direct method is able to remain in compliance for the 15-minute blocks utilizing a 1-minute analytical time. At a 7minute analytical time, two 15-minute blocks are not in compliance and would require a set point of 300 Btu/scf in order to show compliance. The feed-forward method with a 7-minute analysis time does not show compliance for four 15-minute blocks and would require a set point of 537 Btu/scf in order to show compliance. The system was also tested with the event beginning at minute 8 of a block. Figures 18, 19, and 20 illustrate the direct method for the 1-minute direct, 7-minute direct, and 7-minute feed-forward methods, respectively. 2500 1000 900 2000 800 700 1500 600 500 1000 400 300 500 200 100 0 0 0:00 0:15 0:30 0:45 1:00 1:15 1:30 Qwg (scfm) Qng (scfm) NHVcz (Btu/scf) NHVcz = 270 (Btu/scf) 1:45 2:00 2:15 NHVwg (Btu/scf) Flow Rate (scfm) 2500 1000 900 800 700 600 500 400 300 200 100 0 2000 1500 1000 500 0 0:00 0:15 0:30 0:45 1:00 1:15 1:30 1:45 2:00 NHV (Btu/scf) Fig 18: Direct Calculation Method - Water Seal Scenario - 1-Minute Analytical Time 2:15 Time Qwg (scfm) Qng (scfm) NHVcz (Btu/scf) NHVcz = 270 (Btu/scf) NHVwg (Btu/scf) Fig 19: Direct Calculation Method - Water Seal Scenario - 7-Minute Analytical Time 2500 1000 Flow (scfm) 600 1500 400 1000 200 0 500 NHV (Btu/scf) 800 2000 -200 0 -400 0:00 0:15 0:30 0:45 1:00 1:15 1:30 1:45 2:00 2:15 Time Qwg (scfm) Qng (scfm) NHVcz (Btu/scf) NHVcz = 270 (Btu/scf) NHVwg (Btu/scf) Fig 20: Feed-Forward Calculation Method - Water Seal Scenario - 7-Min. Analytical Time When the flaring event starts at minute 8 of a 15-minute block, the direct method is not able to remain in compliance for one 15-minute block utilizing a 1-minute analytical time and two 15minute blocks utilizing a 7-minute analytical time. The set points would need to be increased to 299 and 300 for the 1- and 7-minute analysis times, respectively. The feed-forward method with a 7-minute analysis time does not show compliance for four 15-minute blocks and would require a set point of 555 Btu/scf in order to show compliance. The various scenarios under the water seal scenarios are summarized in Table 3. Table 3 provides, for the two different event start times, for each analytical time, the number of 15minute blocks the system was out of compliance with a set point of 270 Btu/scf and the volume of total supplemental fuel added. The table also provides the set point in order to show compliance with all the 15-minute blocks in the scenario with the additional and total amounts of supplemental fuel required. The set points determined to ensure compliance during the 15minute blocks would require supplemental fuel to be constantly added to the flare. These constant rates have also been provided. Finally, a 5-year historical cost of $0.00459/scf for natural gas was utilized to provide a cost for the supplemental fuel. Water Seal Breach Calculation Method Table3: Comparison of the Water Seal Scenario for Various Analytical Timeframes and Methods Notes: Scenario Early Event Late Event 1 Direct 1-min Direct 7-min Feed Forward 7min Direct 1-min Direct 7-min Feed Forward 7min No. 15minute blocks out of compliance with 270 Btu/scf set point $/scf at 920 Btu/scf Additional Supplemental Fuel Required Total Supplemental Fuel Required for Compliance (No.) (scf) ($)1 (Btu/scf) (scf) ($)1 (scf) ($)1 0 2 34,221 36,786 $157.07 $168.85 270 300 0 6,694 $0.00 $30.73 0 43,480 $157.07 $199.57 4 36,786 $168.85 537 96440 $442.66 133,226 $611.51 1 2 33,942 36,198 $155.79 $166.15 299 300 6,264 6,601 $28.75 $30.30 4 36,198 $166.15 555 Natural gas cost estimate is the past 5 year estimate. 0.00459 Supplemental Fuel Required Set point to be in compliance for all 15minute blocks 40,206 42,799 $184.55 $196.45 106,509 $488.88 142,707 $655.03 Conclusions MACT CC has provided the direct and the feed-forward calculation methods in order to determine compliance with each 15-minute block that regulated material is sent to a flare. One key question that refiners have is what type of technology is needed in order to ensure compliance and what are some of the issues associated with each method. As some of the largest differentiators between technologies is analysis time, the impact of this was utilized to see how a system would react under a few different scenarios. Using the direct calculation method, a rapid and slow scenario, and 1-, 7-, and 15-minute analysis times, a few conclusions can be drawn. 1. The direct method was able to remain in compliance for the scenarios used for all the 15minute blocks in the scenario. 2. As the analysis time increases, the number of 15-minute blocks that are out of compliance increases with a set point at the required NHVcz of 270 Btu/scf. The amount of supplemental fuel the system attempts to add decreases as the interval increases, leading to a conclusion that there is not enough data in order to add sufficient supplemental fuel. One possible solution to overcome these blocks of non-compliance is to increase the set point to add a "cushion" for the system to absorb the changes in flow and NHV. Again, as the analysis time increases, the set point needs to be increased in order to stay in compliance. 3. Depending on the magnitude of the event (both in magnitude of flow and NHV of the waste gas) and the NHVcz that the continuous flare operates at under routine conditions, elevating the set point will cause supplemental fuel to be added at all times once that normal operation NHVcz is exceeded. This will cause the annual cost of supplemental fuel to greatly increase in importance. Using the feed-forward calculation method, a rapid and slow scenario, and 4-, 7-, and 15-minute analysis times, a few conclusions can be drawn. 1. Based on the scenarios utilized, using a set point for the NHVcz of 270 Btu/scf resulted in 15-minute blocks that were out of compliance. 2. As the analysis time increases, the number of 15-minute blocks that are out of compliance increases with a set point at the required NHVcz of 270 Btu/scf. The amount of supplemental fuel the system attempts to add decreases as the interval increases, leading to a conclusion that there is not enough data in order to add sufficient supplemental fuel. One possible solution to overcome these blocks of non-compliance is to increase the set point to add a "cushion" for the system to absorb the changes in flow and NHV. For the rapid event, as the analysis time increases, the set point needs to be increased in order to stay in compliance. Conversely, for the slow event, the set point did not increase as the analysis time increased. 3. For the scenarios simulated, for both the rapid and slow event, using the feed-forward method requires supplemental fuel to be added at all times. Using the direct and feed-forward calculation methods, a water seal scenario, and 1- and 7analysis times, a few conclusions can be drawn. 1. Based on the scenarios utilized, using a set point for the NHVcz of 270 Btu/scf resulted in 15-minute blocks that were out of compliance, with the exception of the direct calculation method with a 1-minute analysis time. 2. As the analysis time increases, the number of 15-minute blocks that are out of compliance increases with a set point at the required NHVcz of 270 Btu/scf. The amount of supplemental fuel the system attempts to add decreases as the interval increases, leading to a conclusion that there is not enough data in order to add sufficient supplemental fuel. The feed-forward method resulted in a greater number of blocks out of compliance and required a larger amount of supplemental fuel to show compliance under the same scenario and with the same sample time as the direct method. While these conclusions have given some clarity to issues that will be observed trying to comply with the combustion zone parameters for MACT CC, the scenarios experienced by a facility will be much more complex and can have other implications for the type of analyzer chosen and the sizing of supplemental gas lines.