The development of the coal seam gas to liquefied natural gas industry is currently a contentious issue in Queensland. Election promises of moratoriums and regional bans for the industry reflect the degree of public concern over the apparent lack of scientific understanding of the industry’s environmental impacts. In a bid to restore public confidence, government and industry have funded the establishment of several independent research bodies to undertake further scientific research into the industry’s environmental impacts.
The industry position on science and evidence-based public policy
On face value the Australian Petroleum Production and Exploration Association, the industry voice for the CSG-LNG industry, is supportive of these initiatives, citing the importance of “science-based” public policy and a “science-based approach to matters regarding the gas industry’s expansion”. However, upon scrutinising its representation of the CSG-LNG industry as an important player in the fight to reduce global greenhouse gas emissions, this commitment to a ‘science-based approach’ does not appear to translate into practice.
APPEA promotes gas as “clean” energy as it produces “50-70% less greenhouse gas emissions” than coal when combusted to produce electricity. To give credence to these statements, APPEA commissioned the global engineering consultancy Worley Parsons to undertake an independent analysis to compare the lifecycle GHG emissions of CSG-LNG against coal when combusted in Chinese power plants.
Upon the report’s release in November 2011, APPEA’s chief operating officer Rick Wilkinson stated that the report’s findings were evidence that gas “is [a] significantly cleaner energy source than coal” and that “the Greens, and other anti-fossil fuel activists, have in recent times become blind to the science on gas’s greenhouse benefit”.
Science and engineering: a difference that matters
In this statement the inference that the analysis was undertaken in the true spirit of science — data collection, rigorous analysis of evidence and objective interpretation of results — is strong. However, what APPEA commissioned was not a scientific investigation, but an engineering report, and when it comes to accuracy and conclusive results, there is a significant difference between the two.
Whereas science involves the systematic investigation and testing of hypotheses in order to develop objective knowledge about the world and how it works, engineering is the application of scientific knowledge in order to problem solve, design and create. In doing so, engineers will apply scientific principles to predict or investigate how a system performs under different operating conditions.
This process is known as modelling, and the accuracy of a model (that is, how well its predictions line up with reality) comes down to the adequacy of the mathematical equations and parameters selected to characterise the model. In turn, the adequacy of a model’s parameters comes down to the currency and comprehensiveness of the data from which it is derived. Engineers develop models that are fit for purpose.
If you need a model that delivers highly accurate and conclusive results, they’ll commit the time and resources to collecting adequate data and develop a robust model. Similarly, if you need some numbers to assist in focussing your decision making process, they’ll make some quick assumptions and do some back of envelope calculations for you.
Analysing the APPEA report
So what purpose, then, is the APPEA report fit for? What is clear is that it is not fit for supporting the claim that CSG-LNG is conclusively less GHG intensive than coal.
The report does support the conclusion that CSG-LNG is potentially up to 70% less GHG intensive given certain assumptions, however this is just one of 54 possible conclusions that could be reached by comparing the six CSG-LNG export scenarios and nine coal export scenarios modelled in the report. Across this range, CSG-LNG goes from being 70% less GHG intensive to being 45% more GHG intensive than coal.
Which, then, of these 54 comparisons is the most meaningful? Given that the report states that gas is “likely to add capacity rather than compete against coal”, that the capacity-adding base load coal power plants currently under construction in China are efficient (supercritical and ultra-supercritical plants, and that base load gas power plants are generally open cycle gas turbines (OCGT), it makes sense that the most meaningful comparison is between the OCGT and supercritical/ultra-supercritical power plants scenarios.
Across these comparisons, CGS-LNG becomes 49% less GHG intensive to 45% more intensive than coal. If we look solely at the “base case” scenarios, the margin narrows to 5-10% less GHG intensive than coal.
What sort of data are we talking about?
In a similar vein, the concept of “best available data” should not be confused with sufficient or adequate data — data that is current, comprehensive and representative — when evaluating the report’s findings. Take the report’s estimate of fugitive emissions for example. In the case of estimating leaks from the LNG plant, gas gathering pipelines and mains transmission pipelines, this “best available data” appears to be that collected in two US EPA lead research projects from the early to mid-1990s.Why? Because these studies were the basis for the emission factors contained in the American Petroleum Institute (API) Compendium, which in turn is the default for the National Greenhouse Accounts (NGA) Factors. The NGA Factors are the Australian government standard for estimating industry GHG emissions, but this does not mean that the data it is based upon is current and representative it is of the Australian CSG industry. It is simply the ‘best available data’.
The venting of gas during well maintenance (well work over) was estimated in the report by using an ill explained figure from a CSG company environmental impact statement (EIS). The EIS stated that in the absence of an NGA emission factor, venting calculations were “based on an operational understanding of existing CSG facilities.” Now, the NGA Factors may not contain an emission factor for well maintenance venting, but the API Compendium does.
In 2010, the US EPA issued a background technical support document that made corrections to several emission factors that are derived from the studies upon which the API Compendium is based. One of them was the well work over emission factor, which was upgraded from 0.05 m3 CH4/well-year to 177 m3 CH4/well-year. If applied to the ‘base case’ scenario and adjusted for the fact that in Australia most of those emissions would most likely be flared, the footprint of the base case CSG-LNG scenarios increase by 10%.
Increase the leakage and venting rate for wells and processing equipment from 0.04% (as per the report’s findings) of total gas processed to 2% — a conservative estimate based upon recent research findings from the University of Cornell and the National Oceanic and Atmospheric Administration – and the CSG footprint increases by an additional 10%.
In two simple adjustments we’ve managed to strip the GHG ‘advantage’ that CSG-LNG had over coal in the base case scenarios and put coal out front. All this without even factoring in the impact of a revised methane global warming potential (GWP) factor.
According to recent research, methane’s GWP factor is more likely to be 33 (that is, methane’s ability trap heat in the atmosphere is 33 times greater than that of carbon dioxide) than 25 as currently recognised by the Intergovernmental Panel on Climate Change. The APPEA report used a methane GWP factor of 21.
Not only is it clear that the report is not a source of conclusive findings on the comparative GHG footprint of CSG-LNG and coal, but it is clear that there is a general lack of data needed to confidently quantify the GHG footprint of the CSGLNG industry. However, is it necessary to postpone making a call on the industry’s GHG performance until adequate data has been collected and comprehensive modelling undertaken?
Maybe in the isolated case of comparing the export of a limited quantity of CSG and coal from Australia to China, but not so much if trying to predict the contribution of a global uptake of gas to slowing global warming; the Worley report itself admits that the “…absolute impact [of CSG-LNG uptake] on world GHG is beyond the scope of this report.”
What does the science suggest for policy?
Over the last twelve months three studies have been released, the first by the International Energy Agency, the second by a senior researcher at the US National Centre for Atmospheric Research, and the third by researchers at Stanford University that suggest a rapid and global uptake gas would not sufficiently reduce GHG emissions to slow the increase in global average temperature over the 21st century.
As neatly summarised by the IEA executive director Nobuo Tanaka, “an expansion of gas use alone is no panacea for climate change” as natural gas, regardless of its GHG footprint, “is still a fossil fuel. Its increased use could muscle out low-carbon fuels, such as renewables and nuclear.”
In terms of answer the question about CSG-LNG’s role in limiting global warming, these studies, though doubtlessly also confronted with data adequacy issues, are far more ‘fit for purpose’ than the APPEA report.
It is clear that APPEA can not rightly claim with scientific certainty that CSG-LNG is less GHG intensive than coal. At best it can suggest that its GHG performance is better than coal under certain scenarios. If it truly supported a ‘science-based’ public policy, not only would APPEA also suggest that under alternative scenarios CSG-LNG is more GHG intensive than coal, it would recommend that more adequate data and detailed modelling be carried out before a science-based public policy position can be reached on the issue.
But it does not. So until APPEA starts to communicate transparently and objectively around the issue, scientists, engineers and greenies alike will continue to “be blind” to what APPEA considers to be “science”.
*Rebecca McNicholl is a Brisbane-based consulting environmental engineer. A fully referenced version of this article is available on request from FAQ Research.