Prepared by Mahkah Wu for Hair on Fire Oregon
2305 Ashland St #412, Ashland, OR 97520
The Pacific Connector Gas Pipeline Project is a proposed 232-mile, 36-inch diameter pipeline designed to transport 1.07 up to 1.55 billion cubic feet of natural gas per day (Bcf/d) from interconnects near Malin, Ore., west and north to the Jordan Cove LNG terminal in Coos Bay, Ore., where the natural gas will be liquefied for transport to international markets.
The Pacific Connector/Jordan Cove LNG Export Project, if built, would contribute the CO2 Equivalent of 18 Coal Fire Power plants initially (each averaging 3.5 million tons of CO2/year) up to 27 Coal power plants if operated at its full capacity (within 4 years).
The planned Pacific Connector Pipeline would move 1.07 Bcf/d of natural gas from the Ruby and Gas Transmission Northwest Pipelines to the planned Jordan Cove natural gas liquefaction facility in Coos Bay. However, the liquefaction facility plans to add a fifth and sixth set of trains after its first four years of operations, which will necessitate delivery of 1.55 Bcf/d of natural gas through the pacific connector pipeline. Although the pipeline plans to operate at 900 psig at the supply interconnections near Malin and at 850 psig at the Jordan Cove LNG terminal,[i] the pipeline’s maximum allowable operating pressure (MAOP) of 1480 psig allows for significantly higher pressure drops and thus greater delivery volumes.[ii]
The difference in liquefaction capacities has important implications for the chronic methane (the chemical name for natural gas along with CH4) loss that occurs at several points of the natural gas supply chain, most notably during extraction through fracked wells. In addition, the possibility of accidents and ruptures in the pipeline necessitates consideration of the impact of concentrated short term release of methane.
Chronic Methane Loss
Even operating without incident, the natural gas supply chain emits methane at several points, most notably during fracking extraction. In addition to fracking effects, natural gas infrastructure includes both intentional leaks, usually for safety purposes, and unintentional leaks, such as faulty valves and cracks in pipelines.[iii] Although proponents of natural gas are correct that it releases 50% as much carbon dioxide upon combustion relative to coal,[iv] methane gas has 86 times the warming potential of carbon dioxide over a twenty year period.[v] Thus, the improvements in carbon dioxide emissions must be considered against methane emissions when determining if natural gas is a viable ‘bridge’ fuel. A number of studies have considered this question, and while there is some disagreement on exact quantitates, there is considerable consensus that the amount of natural gas released is significantly underreported.
A study, “Methane Leakage from North American Natural Gas Systems,” published in the February 2014 issue of Science, synthesized findings from more than 200 studies. It found that nationwide rates of methane emission from natural infrastructure are 25-75% higher than EPA estimates.[vi] This, however, pales in comparison to misestimates made of the methane released during fracking extraction. Another study, “Toward a better understanding and quantification of methane emissions from shale gas development,” published in the March 2014 issue of Proceedings of the National Academy of Science (PNAS), directly measured atmospheric methane concentrations over the Marcellus formation in southwestern Pennsylvania. It found “large emissions averaging 34 g CH4/s per well from seven well pads in the drilling phase” which is between 100 and 1000 times greater than what the Environmental Protection Agency (EPA) estimates![vii] In response to these criticisms, the EPA proposed a new rule in January 2015 to require better accounting of methane released from fracking operations and natural gas compressor stations and pipelines.[viii]
Estimates of the total methane released by the extraction and transport of methane vary. On the low end, the previously mentioned Science study found natural gas leak rates of 5.4%.[ix] On the high end, a study, “Remote sensing of fugitive methane emissions from oil and gas production in North American tight geologic formations,” published in the October 2014 issue of Earth’s Future, used satellites to measure methane released in the Bakken (Montana and North Dakota) and Eagle Ford (Texas) formations, two of the fastest growing fracking regions in the country. From 2006-2008 it found leakages of 10.1% ± 7.3%, and from 2009-2011 leakages of 9.1% ± 6.2%.[x] Another study published in PNAS, “Greater focus needed on methane leakage from natural gas infrastructure,” assessed the benefits of natural gas usage over coal as a function of methane emission. It found that at a 5.4% leakage rates, natural gas is worse than coal for climate change for the first 50 years and at a leakage rate of 7.6% worse for a century.[xi]
Applying these numbers to the Pacific Connector Pipeline, the lower bound estimate of 5.4% leakage corresponds to initial methane emission of 21.1 Bcf/year, ramping up 30.6 Bcf/year once additional liquefaction trains are completed. The upper estimate puts initial emissions at 39.4 Bcf/year with the potential for 57.1 Bcf/year of emissions once Jordan Cove is fully operational. These values range from 1.5% to 4.2% increases in the United States’ total annual methane emission![xii]
Accidental Methane Loss
Although the amount of methane emitted in even a worst case scenario accident pales in comparison to the amount emitted in normal day to day operations of the oil and gas industry, the localized release of a large amount of methane can have catastrophic effects. A number of events might result in breakages in the pipeline, including earthquakes and subsequent tsunamis, landslides, and human induced accidents. In addition, incidents like the Deep Water Horizon oil spill in the Gulf of Mexico demonstrate that unforeseen issues can lead to catastrophic accidents.
The proximity of the pipeline to both the Cascadia Subduction Zone (CSZ) and the seismic hotspot near Klamath Falls is troubling—492 earthquakes have been recorded within 100 miles of the pipeline. Although many of these were of negligible magnitude, 24 were magnitude 5.0 higher, the threshold for which earthquakes have engineering significance. The two most recent of these significant earthquakes occurred in 1993 about 15 miles from Klamath Falls—the first was a magnitude 5.9 event, followed two hours later by a second magnitude 6.0 event. Most troublingly, a major Cascadian earthquake of magnitude 9.0 is thought to have occurred off the coast in 1700. These events occurred multiple times in pre-history and reoccur irregularly on ranges of 100-1000 years. A similar event would generate substantial vertical grounds shifts, potentially shearing the pipeline at multiple points, with lessening hazard in the eastward direction of the pipeline.[xiii]
In the event of a less serious accident that only breached the pipeline at a single point, gas flow would be turned off at the mainline block valves (MLVs). Table 1 shows the gas that would be contained between each MLV at both the initial operating pressure and at MAOP.
In the event of catastrophic leakage, the most immediate danger is, surprisingly, not from its combustibility—the natural gas would not have sufficient oxygen to catch fire and rapid depressurization would cool the gas. While a lack of oxygen would prevent fire, it would be a serious problem for any breathing organism caught in the rapidly expanding cloud of natural gas, including those that breathe oxygen through gills. This problem is not without precedent. Naturally sequestered carbon dioxide was disturbed by a seismic event in 1986 and killed around 1,700 people and 3,500 livestock by asphyxiation.[xv] Although this event was the result of 2.8 Bcf of gas, a significantly smaller emission killed 38 people in 1984,[xvi] so the comparison to the 86.2 MMcf potentially contained in longer sections of the pipe is warranted. It is also conceivable that a small break in the pipeline could occur without detection, sequestering much larger amounts of gas underground until it is released by aftershocks or other seismic activity.
However, the gas at the surface level would quickly dissipate. The mixing of the colder natural gas with warmer air would generate high winds, especially if the accident occurred on a summer day. As the gas dissipated, many highly flammable interfaces where air and natural gas mix would form, creating many possible ignition points. At its most flammable mixture with air, the maximum amount of natural gas between MLV 3 and MLV 4 would have a volume of 195.2 MMcf, more than the volume of eight hours of the Rogue River’s discharge. Given the scale of forest fires triggered by a lightning strike, unattended campfire, or even a cigarette butt, it is difficult to conceive of the fire that a cloud of gas a fraction of this size could generate.
These estimations are by no means the worst case scenario. They are assumptive of the operator’s ability to quickly shut down mainline block valves and of isolated breaks in the pipeline. An event like the last major Cascadian subduction earthquake could rupture the pipeline at multiple points and hamper the operation of MLVs, multiplying these impacts.
[i] The pressure drop is actually approximately 48 psi, rather than 50 psi, due to changes in atmospheric pressure between Malin and Coos Bay.
[ii] Draft Environmental Impact Statement, pg. 2-30: https://www.ferc.gov/industries/gas/enviro/eis/2014/11-07-14-eis.asp
[iii] America’s natural gas system is leaky and in need of a fix, new study finds: http://news.stanford.edu/news/2014/february/methane-leaky-gas-021314.html
[iv] Is Natural Gas ‘Clean’?: http://opinionator.blogs.nytimes.com/2013/09/24/is-natural-gas-clean/
[v] Anthropogenic and Natural Radiative Forcing, pg. 714: http://www.climatechange2013.org/images/report/WG1AR5_Chapter08_FINAL.pdf
[vi] Methane Leakage from North American Natural Gas Systems: http://www.sciencemag.org/content/343/6172/733.summary
[vii] Toward a better understanding and quantification of methane emissions from shale gas development: http://www.pnas.org/content/111/17/6237.abstract
[viii] EPA Moves to Count Methane Emissions from Fracking: http://www.scientificamerican.com/article/epa-moves-to-count-methane-emissions-from-fracking/
[ix] Methane Leakage from North American Natural Gas Systems: http://www.sciencemag.org/content/343/6172/733.summary
[x] Remote sensing of fugitive methane emissions from oil and gas production in North American tight geologic formations: http://onlinelibrary.wiley.com/doi/10.1002/2014EF000265/full
[xi] Greater focus needed on methane leakage from natural gas infrastructure http://www.pnas.org/content/109/17/6435
[xii] EPA: http://www.epa.gov/climatechange/ghgemissions/gases/ch4.html
[xiii] Draft Environmental Impact Statement, pg. 4-259 to 4-261: https://www.ferc.gov/industries/gas/enviro/eis/2014/11-07-14-eis.asp
[xiv] These values were computed based on a 36 in diameter pipe using a density estimated assuming natural gas in the following molar ratio: 96.5% methane, 6% carbon dioxide, 3% nitrogen, 1.8% ethane, 0.45% propane, 0.1% iso-butante, 0.1% n-butane, 0.05% iso-pentane, 0.03% n-pentane, 0.07% n-hexane (See ISO 12213-2 (2006) Natural gas – Calculation of compression factor – Part 2: Calculation using molar-composition analysis). Pipeline data was retrieved from the Draft Environmental Impact Statement, pg. 2-31: https://www.ferc.gov/industries/gas/enviro/eis/2014/11-07-14-eis.asp
[xv] The 1986 Lake Nyos Gas Disaster in Cameroon, West Africa: http://www.sciencemag.org/content/236/4798/169
[xvi] Origin of the lethal gas burst from Lake Monoun, Cameroun: http://www.sciencedirect.com/science/article/pii/0377027387900023