Executive Summary

The Hi-FlowTM Sampler is a portable device developed by GRI to measure emission rates from individual leaking equipment components. It provides very accurate measurements (i.e., to within ±10 to 15 percent) with minimal effort, facilitates the use of simplified screening techniques resulting in reduced leak-detection costs, and perhaps most importantly, it provides the type and quality of information needed to properly evaluate and prioritize individual repair opportunities. The technology has proven to be a useful tool in the control of fugitive emissions at gas transmission facilities in both the U.S. and Canada. The benefits achieved through reduced methane losses have consistently justified the cost of applying the technology. The current study has been undertaken to evaluate the benefits of applying the technology at upstream gas production and processing facilities. The uncertainty is primarily related to potential sectorial differences in average leak rates, frequencies, repair costs, economies of scale, and the overall economics of the available repair opportunities.

To evaluate the merits of conducting leak detection and repair programs enhanced by use of the Hi-FlowTM Sampler at upstream gas facilities, a demonstration program was conducted at two sites: a large gas-gathering compressor station, and a deep-cut liquids extraction plant. Both of these are sweet facilities located in northern Alberta.  The work at the first site was conducted during the period of August 30 to September 1, 1999 by Indaco Air Quality Services, Inc. The second site was surveyed by Clearstone Engineering Ltd during the period of October 3 to 8, 1999.

While the primary focus was on fugitive equipment leaks and the use of the Hi-FlowTM Sampler to help identify cost-effective repair opportunities, an effort was also made to generally identify and evaluate all potential methane control opportunities at the selected sites. This included determining emission rates from vents using direct measurement techniques.  Additionally, flare rates at one site were determined through a combination of measurement and estimation techniques. All cost/price data presented throughout this report are expressed in Canadian dollars.

Total gas losses determined for the two facilities amount to 2 997 x 103 m3/y or $0.318 million dollars annually based on a gas price of $2.80/GJ (i.e., 47 percent by the compressor station and 53 percent by the extraction plant). It is estimated that approximately 90 percent of these emissions would be economical to control, and the avoided losses would have a net present value of 1.41 million dollars. Venting and flaring control opportunities contributed the majority of the economical-to-reduce gas losses at both facilities (i.e., 98 and 99 percent, respectively), with the rest coming from leak control opportunities. The low contribution due to fugitive equipment leaks reflects the fact both facilities were relatively tight; they had average leak rates generally well below the corresponding industry averages.

Typically, the majority of the methane emissions at gas facilities are contributed by only a few significant leakers, not all large leakers are cost-effective to repair, and the probability of finding an economical repair or control opportunity increases with the number of sources surveyed. The results of the current work indicate that sufficient cost-effective control opportunities exist to justify a formal emission survey at compressor stations and larger facilities. The opportunities from fugitive equipment leaks alone, however, did not justify the costs of these surveys. A more comprehensive program that considers fugitive equipment leaks, process venting, flaring and the performance of combustion equipment is recommended. For the portion directed at fugitive equipment leaks, it has been shown that the Hi-FlowTM Sampler is a practical and efficient method of evaluating the identified leak repair opportunities. Moreover, with proper calibrations, the unit can be applied to gas streams of varying compositions, and to non-methane organic vapour streams including LPGs. Procedures for conducting these calibrations are presented herein.

While any economical-to-repair leaking components detected should be fixed, average leaker emission rates based on combined data from both test sites suggest that the most cost effective approach would be to generally focus on the following types of components:

  • block and control valves,
  • pressure relief valves,
  • regulators,
  • flange connections
  • compressor seals, and
  • compressor valve stems and valve caps.

Normally, flange connections would not be a key contributor at gas facilities. Their relative importance here is considered to be an anomaly and is not expected to prevail if more sites are tested.

It should be noted that the accuracy of accounting meters at gas production and processing facilities is typically 0.2 percent of value under perfect conditions, while the amount of natural gas losses as a percent of facility throughput is often less than this value. Statistically, the uncertainty in flow measurements will decrease as the values are aggregated over time due to a cancelling effect of the random noise in the data. Consequently, even though instantaneous flow measurements may not be accurate enough to detect the impact of loss reduction efforts, the long-term impact of such efforts does produce a measurable increase in sales.

The site-specific results are presented below.

Site 1 – Compressor Station

This facility comprised 5 reciprocating compressors, a dehydrator, 2 electric power generators and a produced-water handling system. There were a total of 1657 equipment components (e.g., valves, flanges, compressor seals, relief valves, etc.) in gas service at the site. Of these components, only 2.9 percent were leaking, and of the leakers, only 49 percent were determined to be economical to repair based on a requirement of a positive net present benefit. Other criteria (e.g., a maximum payout period or a minimum required rate of return) may decrease the percentage of leakers deemed to be economical to repair.

The major source of methane emissions at the site proved to be from the use of natural gas as the supply medium for all the pneumatic instruments (49 percent), followed by fugitive equipment leaks (43 percent), and measured venting by the glycol dehydrator (7 percent). Methane losses represent only 32 percent of total CO2-equivalent GHG emissions at the site. Most of the total GHG emissions (68 percent) are due to fuel consumption by the compressor engines and power generators.

A cost-benefit analysis of each identified control opportunity indicates that up to 89 percent of the identified methane losses (i.e., up to 1260 x 103 m3/y valued at $133 500/y) could be eliminated at no net cost and the payout period for many of the key actions would often be less than 1 year. For simplification purposes, this analysis excluded the costs of finding the available control opportunities and the mobilization/demobilization costs for the selected control actions since these would be relatively small on a per-component basis for most reasonable sized-application.

Implementing all cost-effective opportunities would reduce total GHG emissions at the site by 28 percent and yield a net present savings of $635 620.

The primary cost-effective control opportunities identified at the site are as follows:

  • Installation of an air compressor to displace the use of fuel gas as the supply medium for all gas operated devices – this would have a capital cost of $54,700 and result in a net benefit of $38 100 per year (8.991 kt CO2E reduction per year). The estimated payout period is 8 months.
  • Installation of a larger boot or slug catcher on the produced water tank to reduce process gas carry-through – this would have a capital cost of $75,000 and result in a net benefit of $24,860 per year (6.51 kt CO2E reduction per year). The estimated payout period is 1.2 years.

While not always the case (i.e., based on unpublished findings from similar comprehensive surveys conducted at 13 other upstream oil and gas facilities in Canada), focusing on only fugitive equipment leaks at this facility would not have identified sufficient control opportunities to justify the total costs of the program.

Site 2 – Deep-Cut NGL Extraction Plant

This facility comprised 2 inlet reciprocating compressor units, 3 shallow-cut extraction trains complete with molecular sieve dehydration units, 1 deep-cut extraction train, and an LPG and condensate storage facility. There were a total of 17 085 equipment components (e.g., valves, flanges, compressor seals, relief valves, etc) in hydrocarbon gas service at the site. Of these components only 2.0 percent were leaking, and of the leakers, only 47 percent were determined to be economical to repair (i.e., result in a positive net present benefit).  The emission factors developed from the measurement results are all well below the corresponding industry average values published for fugitive equipment leaks, with the exception of the emission factor for compressor seals (CAPP, 1999).  Additionally, the total fugitive emissions determined for the facility are 55 percent lower than the values that would have been predicted using published average emission factors.  In contrast, the total fugitive emissions were 47 percent larger than would have been predicted using the leak rate correlation approach (U.S. EPA, 1995) for all leakers, and no-leak emission factors (CAPP, 1992) for the remaining components.

The major source of hydrocarbon loss at the site is the residual flare rates by the emergency flare systems (i.e., purge gas plus leakage into the flare systems) (73 percent), leakage from the compressor rod packings (7 percent), and leakage from the switching valves on the molecular sieve units (2.2 percent). The rest of the emissions are from miscellaneous fugitive equipment leaks. Compressed air is used as the supply medium for all gas-operated instruments and devices so there were no methane emissions from these items. Overall, hydrocarbon losses represent only 9 percent of total CO2-equivalent GHG emissions at the site. Most of the total GHG emissions (89 percent) are due to fuel consumption at the facility.

A cost-benefit analysis of each identified control opportunity (excluding the costs of finding these opportunities and the mobilization/demobilization costs for the selected control actions) indicate that up to 90 percent of the identified hydrocarbon losses (i.e., up to 1423 x 103 m3/y or $150 800/y) could be eliminated at no net cost. This would reduce total GHG emissions at the site by 8 percent and yield a net present savings of $894 910. The payout period would generally be less that 1 year for most of the key cost-effective control options.

The primary cost-effective control opportunities identified at the site are as follows:

  • Installation of a flare gas recovery system – this would have an estimated capital cost of $99 500 and result in a net benefit of $87 720 per year (2.972 kt CO2E reduction per year). The estimated payout period is 11 months.
  • Repair of the leaking compressor seals – this would have a cost of $6 500 and result in a net benefit of $3,760 per year (1.425 kt CO2E reduction per year). The estimated payout period is 7 months.

As was the case for Site 1, to have conducted an emissions survey focused only on fugitive equipment leaks would not have identified sufficient control opportunities to justify the total costs of identifying and implementing the identified control opportunities.

Final Report