Benchmarking GHG emissions from cryogenic gas processing
K. Chow, Muse, Stancil & Co., Houston, Texas
Greenhouse gas (GHG) emissions became a focal point in the US gas processing industry in 2009 when the US Environmental Protection Agency (EPA) proposed regulation to require the reporting of GHG emissions in the petroleum and natural gas sectors under its Subpart W rule. The EPA stated that the proposed rule would require sources above certain emissions thresholds to monitor and report emissions, and that the rule would not include the control of GHG emissions.
Even so, the requirements were not straightforward, as industry operators and trade associations raised major concerns regarding the proposed rule, including emissions-monitoring methods, equipment threshold levels that trigger reporting, safety issues related to direct measurement, inaccessible measurement sources, and aggregation of facilities for reporting.
Subsequently, when the EPA signed the final mandatory GHG reporting rule under 40 Code of Federal Regulations Part 98 and published it in the Federal Register on October 30, 2009, Subpart W was excluded from the rule.
In recognizing the scope of the industry’s concerns and the complexity of measuring emissions in the oil and gas sector, the EPA modified its Subpart W rule and did not finalize its final rule, “Mandatory Reporting of Greenhouse Gases from Petroleum and Natural Gas Systems,” until November 30, 2010. The final rule required facility operators to begin measuring GHG emissions in 2011 for the reporting of emissions in calendar year 2011. GHG emissions to be measured and reported are carbon dioxide (CO2), methane (CH4) and nitrogen dioxide (NO2).
Effect on gas processing permitting. While the EPA stressed, during the announcement of Subpart W, that the mandatory reporting rules do not require control of these GHG emissions, the agency made it clear that the data collected under 40 CFR Part 98 would be used to direct policy and regulatory actions. One of these regulatory actions has been the tailoring rule, which affected the permitting process for large gas processing plants.
Under the tailoring rule, any facility that emits greater than 100 Mtpy of carbon dioxide equivalent (CO2e) would be subject to the Prevention of Significant Deterioration (PSD) permitting process. The PSD application requires the applicant to perform a GHG emissions best available control technology (BACT) analysis, where all available control technologies are identified and then assessed for their technical, economic and environmental merits.
Through the EPA’s New Source Performance Standards (NSPS), the agency established additional requirements in 2012 for reducing volatile organic compound emissions from certain equipment and processes at gas processing plants. The NSPS requirements were intended to lower VOC emissions and will also have a similar effect on GHG emissions.
Potential implications for the future. Understanding the GHG emissions from the wellhead to the burner tip is critical, as this data could be used by policymakers to set future regulations that could impact demand for natural gas. The transition of electric utilities and independent power producers to natural gas-fired generation over the past decade is clear evidence of regulations that benefit the natural gas industry.
The EPA could ratchet down the threshold for PSD applications, which would capture a greater percentage of greenfield gas processing plants under the program. One of the risks with PSD is that the BACT evaluation and approval from the EPA administrator is done on a case-by-case basis. This could result in the delay of plant startups, which would, in turn, significantly affect well production and could result in well shut-ins or delays in well completions.
Analysis of EPA GHG reporting results. Against this backdrop of existing regulatory requirements and potential for future policy actions at both the state and federal levels, the GHG emissions results from the 2011 and 2012 reporting years for the EPA GHG reporting program are analyzed to understand the significance of gas processing emissions relative to other parts of the natural gas supply chain.
Fig. 1 illustrates the breakout of reported GHG emissions between the major EPA-defined source categories in the natural gas industry—production, processing and transmission. The GHG emissions are measured in metric tons of CO2e, representing emissions of CO2, CH4 and NO2.
Fig. 1. Reported GHG emissions in the US natural gas sector.
As seen in Fig. 1, the production source category is the largest contributor to industry GHG emissions, with 100 metric MMt of CO2e and 95 metric MMt of CO2e in 2011 and 2012, respectively. The natural gas processing source category is the second-largest contributor, with a reported 62 metric MMt of CO2e in 2011 and 60 metric MMt of CO2e in 2012. By comparison, the GHG emissions downstream of production and gas processing are not as significant.
Note: While upstream GHG emissions are reported through the aggregation of multiple production sites into an EPA-defined “facility,” midstream GHG emissions are reported under the traditional definition of a facility, which is a gas processing plant that meets the emissions-reporting threshold.
When compared to other sectors of the US economy, the natural gas sector is second only to the power sector in reported GHG emissions. Power plants led all industries in GHG emissions with 2,090 metric MMt of CO2e, followed by petroleum and natural gas systems as a distant second, with 217 metric MMt of CO2e. Fig. 2 illustrates the gap in reported GHG emissions between power plants and the rest of the industry sectors.
Fig. 2. Reported GHG emissions in 2012 by sector, metric MMt CO2e.
Gas processing emissions will rise. Gas processing emissions under the EPA reporting program are expected to climb as processing plants expand and as larger plants are built to accommodate the rise in unconventional gas production.
There were 394 facilities reporting GHG emissions for 2012, and the number of reporting facilities is expected to rise with modular cryogenic processing plants (i.e., those with capacities of 100+ MMcfd) becoming the norm for greenfield facilities. Furthermore, plants that were not previously subjected to EPA reporting could potentially qualify as reporting facilities as additional processing trains are added to existing facilities.
Benchmarking GHG emissions from cryogenic plants. This benchmarking study looks at the reported 2012 GHG emissions for a sampling of 10 cryogenic gas processing plants located along the US Gulf Coast, the Midcontinent region (including Oklahoma, North Texas and the Bakken shale formation), the Permian basin and the Rocky Mountain region. Intensity-based GHG emissions metrics were developed using publically available gas plant operating data to enable comparisons between plants with various nameplate processing capacities.
Under the mandatory EPA GHG reporting program, a natural gas processing facility would be required to report its GHG emissions if its annual average throughput was 25 MMcfd or greater. Once that threshold is met, the processing plant is obligated to measure and report GHG emissions from eight specific sources. In Table 1, the reported emissions in 2012 for the 10 plants are summarized using a plant intensity metric based on actual gas plant inlet volumes in 2012, expressed as metric tons of CO2e/MMcf of natural gas processed.
While the GHG emissions intensity ranges from 0.616 metric tons to 3.387 metric tons of CO2e/MMcf processed, it should be noted that the average GHG emissions intensity (1.668 metric tons of CO2e/MMcf processed) and the median GHG emissions intensity (1.546 metric tons of CO2e/MMcf processed) are quite close, indicating that the data distribution is fairly symmetrical.
Delving further into the specific sources of the GHG emissions at each plant, Fig. 3 provides a breakdown of the reported emissions by equipment type. The chief contributor to GHG emissions is stationary combustion. These sources include reciprocating engines and turbines as well as external combustion equipment, such as process heaters, reboilers and thermal oxidizers.
Fig. 3. Breakdown of cryogenic plant GHG emissions sources.
For plants that treat natural gas with high CO2 content, the venting of CO2 emissions from acid gas removal vents is also considered to be a substantial contributor. For the remaining five equipment categories, it appears that their relative significance as GHG emissions sources is dependent on each plant’s unique operational profile.
For example, the quantity of GHG emissions from blowdown vent stacks is a function of the number of planned and unplanned equipment shutdowns that require depressurization of pipeline, compressor or vessel components. Likewise, for centrifugal compressor venting and reciprocating compressor venting, the GHG emissions quantity is determined by the number of compressor vents and the flowrate sampled during the annual measurement.
In the case of Plant E, flare stack GHG emissions exceeded stationary combustion emissions, although that was due to 2012 being the plant’s first year in operation. Gas was likely flared as the plant underwent commissioning and startup procedures.
In Table 2, the emissions results from the 10 plants are summarized by equipment category. The dehydrator vent category has been excluded because only one of the 10 plants sampled reported GHG emissions greater than zero from this source.
Operational insights. Not surprisingly, several equipment categories would require additional data points to produce a normal distribution data set. Even with the limited data set available using a sampling of 10 plants, operational insights can still be derived. For blowdown vent stacks, the emissions quantity is driven by the number of blowdown occurrences, which vary widely from plant to plant.
Blowdown occurrences are caused by planned and unplanned shutdowns, which would imply that a gas processing plant with high blowdown vent stack emissions per MMcf of gas processed is experiencing more operational issues than its peers.
Flare stack emissions represent process gas, vented gas and other hydrocarbon streams that are burned in the flare. While vented gas that is combusted in the flare is regarded as a form of vapor recovery control, process gas and pure hydrocarbon streams sent to the flare typically represent valuable product streams that were directed to the flare due to equipment shutdown. As such, high flare stack emissions reported under the EPA are potential indicators of reduced plant recoveries and residue gas volumes.
In the case of compressor emissions, the reported emissions under the compressor category could be as low as zero if the compressor vents are not open-ended lines and are directed to the flare. The emissions would then be reflected in the flare category.
For reciprocating compressors, if the rod packing case is open to the atmosphere and not connected to an open-ended line or vent, then the quantification of emissions is based on an annual leak detection survey and reliance on flow estimates using calibrated bags or a high-volume sampler.
Emissions from equipment leaks represent the total estimated leaks from specific plant components in gas service. The leaks in valves, connectors, open-ended lines, pressure relief valves, and meters are identified during an annual leak detection survey using approved methods, such as optical gas imaging instrumentation or an infrared laser beam-illuminated instrument. A plant with high GHG emissions under this category should undertake a more active leak-detection program to reduce product loss.
Effect of gas plant use on GHG emissions. Gas processing plants are designed to run optimally under a certain set of process conditions. With equipment at the plant being sized for a specific range of operations, sub-optimal performance could result in plant energy usage inefficiencies and a higher number of upsets. These are reflected in the reported GHG emissions from stationary combustion fuel, flare stacks and blowdowns.
In Fig. 4, the GHG emissions rate for each of the 10 plants is evaluated against its 2012 processing capacity utilization. As discussed earlier, Plant E flared a large amount of gas during startup, commissioning and ramp-up of operations, and this is reflected accordingly in Fig. 4.
Fig. 4. GHG emissions and plant capacity utilization.
At the other end of the spectrum, Plants A and C have lower GHG emissions rates because they do not have any acid gas removal units subject to reporting.
GHG emissions and NGL gallons recovered. NGL recovery is a fundamental component of gas processing economics. While inlet gas composition, flowrate and plant configuration influence plant recovery rates, assessing the GHG emissions per gallon of NGL still provides an indicator of how efficient a plant is at recovering liquids.
Since stationary combustion fuel is the largest source of gas plant GHG emissions, a high GHG emissions rate per gallon of NGL produced would imply that the plant is retaining a disproportionate percentage of gas as plant fuel. Fig. 5 highlights the varying emissions rates of each plant for NGL recoveries.
Fig. 5. GHG emissions and NGL recoveries.
Takeaway. GHG emissions from stationary combustion, flare stacks, and acid gas removal units present the largest sources of emissions from gas processing plants. The amount of GHG emissions per MMcf of natural gas processed can vary widely from plant to plant, based on the type of installed equipment and operational management, which results in a unique emissions profile for each plant.
The NSPS technology requirements for compressors and pneumatic controllers to reduce VOC emissions should reduce reported GHG emissions from individual plants in future reporting years, under the EPA’s Greenhouse Gas Reporting Program.
While CO2 emissions from acid gas removal vents are a significant contributor to GHG emissions, there is, at present, no effective control option to address this issue. For the acid gas removal vents, the emphasis is placed on using the thermal oxidizer to convert any CH4 or VOCs present in the waste gas into less intensive CO2. To address GHG emissions from stationary combustion sources, the selective use of electric motors could reduce the amount of fuel consumed by the plant.
Prudent operations and reductions in plant upsets would lead to fewer maintenance, startup and shutdown events that cause flare and blowdown emissions, with the added benefit of retaining more hydrocarbon products for sale at the plant tailgate. GP
Ken Chow is a principal at global energy consultancy Muse, Stancil & Co., with over 15 years of technical and commercial experience developing assets in the upstream, midstream, LNG and power sectors. He served as a member of the API GHG Upstream Workgroup and the GPA Environmental Subcommittee during the development of the EPA’s mandatory GHG reporting for the oil and gas industry. In addition to his industry background, Mr. Chow has five years of natural gas and refinery consulting experience with Purvin & Gertz. He holds a BEng degree in mechanical engineering from McGill University in Montréal, Canada, and he is a certified Lean Six Sigma practitioner.