The Puget Sound Clean Air Agency (PSCAA) Supplemental Environmental Impact Statement (SEIS) on Greenhouse Gas (GHG) Emissions Lifecycle Analysis (LCA) has been out since Oct 8th, 2018.  For a review of the LNG project and for general SEIS talking points, please review my LNG page here. On this page, I review some of the numbers behind their calculations and then demonstrate what their model really tells us when we use more updated information.  Each of the paragraphs below will showcase what happens when you modify one variable in their model in terms of the overall GHGs produced by the LNG project versus what is called the No Action Alternative, in other words, if they did nothing and kept using diesel and not LNG.

Review of the Scenarios

Before we dive into the numbers, it is important to remember that there are three scenarios outlined in this SEIS.  They include:

ScenarioDescription
A250,000 gallons of LNG production per day
B500,000 gallons of LNG production per day
No Action AlternativeNo LNG facility

I have only included results comparing Scenario A to the No Action Alternative at this time, but will add in Scenario B versus the No Action Alternative later this week.

Using the Updated Methane 100-year Global Warming Potential (GWP) Value

The first issue arises with the 100-year methane GWP value. The PSCAA model uses a 100-year methane GWP value from 2007, however there are more recent updates to this value as shown in the table below.  The 2007 numbers were used in the SEIS as that is what the Environmental Protection Agency currently uses and it is codified in the Washington Administrative Code, WAC 173-441-040.  I would note that for the updated 2013 values from the UN Intergovernmental Panel on Climate Change (IPCC) Assessment Report 5 (AR5), there are actually two sets of values; one includes climate carbon feedbacks and one does not.  Climate carbon feedbacks are impacts that the methane has on the overall atmospheric chemistry as the methane is oxidized, so it provides a more complete picture of the true impact of methane emissions.

ReportYear100-year Value
UN IPCC AR4 (used in the SEIS)200725
UN IPCC AR5 without climate carbon feedbacks201328
UN IPCC AR5 with climate carbon feedbacks201334
Etminan et al without climate carbon feedbacks201632

The following graph shows the overall emissions when using these 100-year GWP values.  The blue bars represent the LNG Scenario A (blue, think natural gas), and the dark gray represent the No Action Alternative (dark gray, think oil).

Therefore, using the best available scientific data, either from the UN IPCC AR5 report to include the climate carbon feedbacks, or the even more recent 2016 Etminan et al GWP results, one sees that the LNG project is actually dirtier from a GHG perspective than just using the existing solution (1.69% and 0.07% respectively).  For the Etminan et al number, remember, that does not include the climate carbon feedbacks.  I suspect that including that would probably boost the overall GWP to around 38, but that is purely an educated guess.

Using the 20-year Methane GWP Value

The next modification I made was to replace the 100-year methane GWP values with a 20-year methane GWP value.  The project is only slated to last for 40 years, and the recent UN IPCC report indicated that we need to drastically change our energy fuel mix within the next 12 years, or face dire environmental consequences by 2040.  Such declarations warrant using the 20-year GWP value instead of the 100-year value.  The 20-year values are shown in the table below and the impacts to project’s GHG emissions are plotted in the subsequent graph.

ReportYear100-year Value
UN IPCC AR4200772
UN IPCC AR5 without climate carbon feedbacks201384
UN IPCC AR5 with climate carbon feedbacks201386
Etminan et al without climate carbon feedbacks201696

As you can see, using the 20-year methane GWP values has a significant impact upon the results.  Even using the older 2007 values, this shows that the LNG plant would be 32% dirtier over the first 20 years of the project, all the way to 51% dirtier if we used the latest numbers from Etminan et al.

Another way to consider these results are what do they mean in terms of the number of additional cars that would be on the road annually.  The EPA states that the average car emits 4.6 metric cons of CO2 per year.  Given that the estimates above are for the 40-year lifespan of the LNG project, this gives the following number of additional cars per year:

Using a 40-year Methane GWP Value

The SEIS indicates that the lifespan of the LNG facility is estimated to be 40 years.  As such, it would be useful to understand the GWP over 40 years.  This value is not typically provided by the UN or other scientists, so we need to derive it ourselves and then apply it to the SEIS model.  TODO. 

Using an Updated Upstream Methane Leakage Rate

Another value to adjust is the upstream methane leakage rate.  As natural gas is extracted from the ground using hydraulic fracturing, which is the primary extraction technique used in British Columbia which is the source of this gas, a certain percentage of the methane leaks directly into the atmosphere.  This is known as methane leakage and a rate (percentage) is applied that is the percentage of gas that leaks versus how much is extracted.  The SEIS used the methane leakage rate as found in the GHGenius v4.03 software, however a newer version of the software is available with an increased leakage rate, and there have been recent empirical studies indicating that the leakage rate provided by the B.C. Oil and Gas Commission is significantly under reported.  I will consider a variety of justified leakage rates based on these considerations.  These are listed in the table below.  TODO – Complete the values in the table below.

SourceYearLeakage Rate (%)g/mmBTU
GHGenius v4.030.32TODO
GHGenius v5.00.46TODO
Atherton et al2017TODOTODO
Werring2018TODOTODO

TODO – I will provide a graph of the impacts of properly adjusting the upstream methane leakage rates to the overall greenhouse gas emissions once I can confirm the missing data in the table above.

Using an Updated Downstream Vessel Methane Slippage Rate

One of the primary purposes of the LNG facility will be to provide natural gas to TOTE maritime and other potential maritime vessels.  When the vessels burn the natural gas, a certain amount of it leaks into the atmosphere.  This is called methane slippage.  There are some different values for methane slippage that can be used, which are shown in the table below.  The SEIS currently uses, and PSE agrees with this use, the smallest methane slippage shown below, whereas the SINTEF recommendation (SINTEF is the research organization who did the research behind these values) and the manufacturer testbed data values are also provided.  Note that all of these values come from the Stenersen and Thonstad, 2017 SINTEF study, Table 7.2, p. 31.  I would also note that when PSE conducted their own GHG lifecycle analysis that was included as an additional document with the SEIS, they used the 5.3 g/kWh number, yet their peer reviewer indicated that they should include the 7.6 g/kWh manufacturer testbed number as that is considered best practice.

SourceMethane Slippage (g/kWh)
SINTEF Measurement (used in the SEIS)5.3
All Sources, Dual Fuel (SINTEF recommendation)6.9
Manufacturer testbed data7.6

Applying these methane slippage values yields the following graph.

Cumulative Impacts

With all of this background, now we can start looking at the different combinations of these values to understand their cumulative impacts.  For instance, what do the GHG emissions look like when you combine the latest 20-year methane GWP value with the manufacturer testbed methane slippage value?  I show some of the permutations in the graph below.  I will incorporate the updated upstream methane leakage rate information as soon I have confirmed the possible values, as indicated in the upstream methane leakage section above.