A response to the report titled “Investigation of Fine Particulate Matter Characteristics and Sources in Edmonton, Alberta”
By Dr. Joe Vipond, CCFP (EM), MD, BSc (hon)
In November, 2015, Drs. Kindzierski and Bari, of the University of Alberta’s School of Public Health, released a report focused on whether coal fired power plants lying to the west of Edmonton had an impact on the city’s air quality. This non-peer reviewed [AR1] [JV2] [JV3] report was released solely on the website of the company Transalta, and has not had the opportunity to go through the regular scientific process of discussion and rebuttal. There are a number of assumptions and omissions within the report that deserve further examination, and could call into question the conclusions of the study.
The report states that wind conditions were evaluated because “…meteorology may play a major role in influencing ground level concentrations of air pollutants across the area”.
The wind roses provided on figures 1.1 and 1.2, were derived from data at the Edmonton MacIntyre station, taken from the years 2006-14. It has been provided “… to understand the strength and directions of the prevailing winds contributing to Edmonton air quality.” The graphic on figure one suggests that the overwhelmingly predominant wind direction is from the South, however the text states that over this period “Prevailing winds blew blew from the W-NW direction 40% of the time and from the S 23% of the time…”
As weather is highly variable, it is not unremarkable that various locations will produce different windroses[AR4] . Often multiple sites will be evaluated within a regironregion to get a good picture of what is happening. Indeed, the reports methods section states “Wind rose plots were generated to show historical 17 year and 9 year average prevailing wind directions at existing air monitoring stations in Edmonton…[AR5] ”, however, only the MacIntyre station result areis provided in the report. Further review of Other windroses at other stations provide a better understanding of the substantial variability of wind directions even within city limits (See figure 4, 2014 Capital AIrshed Report). . Using wind directions from a single station does not provide a reasonable perspective of the actual variability present.
Furthermore , Looking at the windroses over a long term period doesn’t help resolve what the cause of the exceedances were, which occurred in the winter of 2010. A Government of Alberta study (need reference) appropriately looked at the windrose over a shorter time span when the 2010 exceedances occurred. What this study was able to ascertain was that wind speeds were low during these exceedances and temperature inversions were present causing the trapping of air pollution near ground level. This tells us that the overall pollution loading in the capital airshed is too high to maintain good air quality during all weather conditions.corroborate that the predominant wind is from the West/WNW (attach figures or hyperlinks).
It is important to note that the three main coal plants lie directly west of the city (see figure S2.2 in the report). It is also important to note that the average wind direction is not particularly important, as the wind direction is continually shifting and therefore the wind will blow pollution from a single site across the landscape over time. A nice example of this can be found in the video model produced by Environment Canada showing the modelled wind eaffect on the transportation of SO2 from point sources in the Alberta/Saskatchewan region for the month of August 2013. (insert hyperlink).
Forest fires or equipment changes [AR6] caused the PM2.5 exceedance in 2010
It is true that CAPE media report released in April 2015 did not remove forest fire events, and that this would have caused the numbers to be higher than had they been removed for 2010. This was clearly stated in the presentation of these results. However, CASA Gguidance documents for the achievement of Canadian ambient air quality standards (CAAQS) states that “…events demonstrated to be caused by forest fires should be removed from the calculation of achievement demonstration”. Indeed, the Government of Alberta’s data analysis which precipitated the trigger exceedance under CAAQS did have theseis forest fires removed (see http://aep.alberta.ca/about-us/cumulative-effects/regional-planning/capital-region/documents/CapitalRegion2013AirQualityReport-Mar2015.pdf[AR7] .)
The devices used to measure PM 2.5 did change in 2009, and evidence indicates that this would have caused more accurate, and slightly higher, PM2.5 measurements, especially in the winter. This is an important consideration when performing longer-term analysis’ of trends in pollution, however this has little impact to the recorded exceedances from these air monitoring stations. The However, this data from the new devices isare still the approved measurement devices that are to be used that against whichto report against the CAAQS [AR8] (or the new CWS) is measured, therefore the exceedances still stand as true. Our ability to more accurately measure the exposure to fine particulate matter is not a reason to disregard the need to reduce this pollution back to safe levels[AR9] .
Graphical and data analysis
Figure S1.2 Shows PM2.5 historical trends for four monitoring stations. While referencing the decrease in PM2.5 over time at MacIntyre, the authors ignore the obvious rising trends at the other three stations.
Figure S1.4 Indicates that for the pollutants PM 2.5, NOx and SO2 (all of which are acknowledged as emissions and secondary derivatives from coal combustion) , but are unscientifically ascribed to other industrial sources rather than coal[AR10] ), there is an obvious rise in levels during the years 2012-14 that is ignored by looking at a 10-year trend. This is not acknowledged within the report making it not useful in the investigation of the cause of specific events, or short term trends. these events. Again the issue we are identifying is the recent increase in total air pollution and the associated measured events that remain unattributed to a specific source.
Long term trends can be useful to ascertain an overall population’s likely long-term exposure to air pollution, but in no way is this helpful in determining the specific cause of a specific event.
Sources of pollutants in the Capital Airshed
The 2015 Kindzierski report states that coal-fired power emissions minimally contribute to the PM 2.5 issue in the Capital Airshed. This stands in stark contrast to a study he co-authored in 2001 looking at the sources of PM10 (coarser grained pollutants, of which PM2.5 is a component) in two months in the Devon area, located just SW of the city of Edmonton. That study concluded “The PM10 collected during the two sampling seasons originated from four main source categories that combined to account for 86% of total amount of particulate matter collected. The largest contribution came from crustal material sources (30%) followed by coal emissions (22%), combustion sources (18%) and regional sulphate (15%).” It is further worth noting that both the combustion sources, and regional sulphate, were also partially composed of coal emissions. Indeed, this report stated: “This background regional sulphate was likely due to the high amount of oil and gas activity in Alberta (that often results in gas flaring and other industrial processes), emissions from coal-fired power plants, and emissions from other primary industrial sources that cannot be singled-out.”
Another paper co-authored by Kindzierski on PM1 in early 2015 also stated: “Since some elements (e.g., Pb, Cd) are reported to be transported over long distances by atmospheric flow and deposited far from their emission sources,61,66 the presence of one coal-fired cement factory in the northwest, three major coal-fired power generating units 60 km west of Edmonton (in Wabamun Lake) and one coal-fired power plant in 200 km southeast (in Battle River) are inferred to be associated with this factor (SI Figures S8, S11−13). Therefore, (one component of PM1) was interpreted as coal combustion and it accounted for 2.7% (0.010 μg/m3) and 3.6% (0.018 μg/m3) of total measured elemental mass in PM1 for indoors and outdoors, respectively…(Another component of PM1) was characterized by a large mass fraction of S, which explained more than 60% of the variation indoors and outdoors. This factor was assigned to secondary sulfate and it was the largest source in Edmonton, accounting for 44% (0.16 μg/m3) and 43% (0.21 μg/m3) of the total measured elemental mass in PM1 for indoors and outdoors, respectively. This is interpreted to be related to Alberta’s background regional sulfate that is found in high abundance due to oil and gas extraction and production activities (SI Figures S8 and S9), other industrial processes like coal- and gas-fired industrial boilers and power plants and other nonspecific industrial sources.”
Furthermore, a paper frequently referred to by the Transalta study is Jeong, 2011. It contains the following statement, referring to Edmonton: “Interestingly, the maximum concentrations of the sulfate factor, the second largest source (19%), were observed in winter due to primary emissions from winter–coal combustion.” And “Overall, more than half (60% – 82%) of the PM2.5 mass in Edmonton appeared to come from local sources.” The final conclusion stated “In Edmonton many VOCs were strongly correlated with the nitrate, sulfate, and biomass burning factors, indicating the importance of local contributions… Thus, PM2.5 can be reduced at all the sites through local controls, particularly Edmonton.”
Finally, in the Government of Alberta’s own report, the Capital Region Fine Particulate Matter Science Report, figure 16 shows that during the concerning coal season events (where PM2.5 exceeds the CAAQS), the primary components are secondary PM2.5 with Ammonium Sulfate and Ammonium contributes over 60% of the PM2.5. The report further states that the SO2 and NOx are likely from regional, and not local sources (page 17,20). Finally, figure 21 states that in the region, 44% of the NOx, and 69% of the SO2, is from electrical power generation (aka coal fired electrical generation).
This is in sharp contrast to the statements in the Transalta paper, such as “The observed apparent higher levels in winter months are presumably due to… SO2 from local sources and regional transport of oil and gas activities…” (note no mention of coal). “Only a strong contribution to secondary sulfate from… where coal combustion sources are located can be observed from (our models)…” “… other important regional precursor sources of secondary nitrate are located in Alberta South of Edmonton, NW BC and Southern SK…” The assertion that areas in Southern Saskatchewan contribute substantially to NOx in Edmonton, over a massive source 50 km away, seems disingenuous.
The report further suggests that “…the use of natural gas for residential/commercial space heating in Edmonton is more common and also contributes to elevated secondary sulfate contributions during winter and spring months.” This is incorrect, as the combustion of natural gas is not associated with SO2 emissions, the precursor for secondary sulfate.
A abstract from B. Williams (2014) states (Andrew, Can you help with a better SOA/coal reference?): “It has been recently discovered that along with known emissions from coal combustion and coal-biomass cofiring (e.g., greenhouse gases, refractory particles), organic aerosol can also be directly emitted as primary organic aerosol or produced in the atmosphere as secondary organic aerosol from aging of volatile organic compounds emitted from inefficient combustion/pyrolysis processes.” Compare this with the Kindzierski report: “The potential for Secondary Organic Aerosals formation in coal combustion plumes is considered to be small or unimportant based on our review of published literature”. Further, in regards to SOA, Kindzierski states “…the presence of a preferred trajectory path aligned with the Yellowhead transportation corridor west of Edmonton and and area southwest of Edmonton during winter…that is associated with elevated levels of PM2.5 for the SOA factor” There is no mention that this is the same location of the 12 coal generators in the Wabuman area. A similar claim is made regarding NOx and the Yellowhead transportation corridor.
One suggestion would be that a more accurate representation of the coal-fired power plants impact on the Edmonton airshed would be to model the emissions from these point sources and how they impact the surrounding areas, like Environment Canada has done with the SO2 emissions. Another solution would be to have adequate monitoring in the areas between the plants and Edmonton, such as Stony Plain, Spruce Grove, and West Edmonton, areas which are currently lacking in monitoring stations.
Funding and timing of report
Through personal communication with the author and persons at Transalta, it was ascertained that the funding for the report came from an unrestricted grant from Transalta. When the report was released, on the Transalta website, this fact was stated on the news release associated with the report. Since then, this reference has been removed from the website, for reasons unknown. The author has stated that funds were received and placed into an fund at the university, but were never spent. When the university was contacted to discuss the details of the grant and the account, they were not able to do so, citing privacy concerns. A FOIP request is pending to further elucidate the details of this relationship.
This report was presented to Alberta Environment in early September, 2015, and was mentioned by Transalta CEO Dawn Farrell during the Pembina Climate Summit in September 2015. The report was released to the public on November 20th, two days prior to the province’s Climate Leadership announcement.
With the Governments investigation not pointing to a single cause of the exceedances, this indicates that overall pollution loading from all sources must be addressed. Coal-electricity, with viable replaceable technology with much lower to no air contribution, must also be addressed and we feel should be the focus of the strategy to reduce air pollution in the Capital region and across the province.
[AR1]Not sure if there is a standard of peer review, but two doctors as authors might be considered peer-reviewed. I’m not certain though.
[AR4]Great visual of this from Alberta Capital Airshed annual report http://capitalairshed.ca//cmsFiles/documents/document55ff227abe602.pdf
[AR5]Additional point that could be added. Looking at the windroses over a long term period doesn’t help resolve what the cause of the exceedances were, which occurred in the winter of 2010. The Government study appropriately looked at the windrose over a shorter time span when the exceedances occurred. This provides the necessary insight to understand what wind conditions were present when the exceedances occurred, however this does not provide any definitive evidence of the source of the pollution causing the exceedances.
What the GOA study was able to ascertain, somewhat unsurprisingly, was that wind speeds were low during these exceedances and temperature inversions were present causing the trapping of air pollution near ground level. This tells us that the overall pollution loading in the capital airshed is too high to maintain good air quality during all weather conditions. This is why air pollution throughout the capital region must be reduced.
[AR6]Might be helpful to separate these two under separate headings.
Warren’s study doesn’t appear to have removed forest fire events, making it more difficult to infer anything from the study.
This point should be highlighted. CAPE didn’t remove and neither did Warren’s. The best analysis is the GOA’s for evaluation of CAAQS.
Warren has consistently included forest fire events in his statistical work, which isn’t inappropriate to determine an overall populations exposure to air pollution, but isn’t appropriate in ascertaining the source of pollution. Can show his study on long term trends in Fort Mac as evidence of this, and his disregard for the issue of spikes in air pollution which is the problem we’re trying to address.
[AR8]For semantics, the CAAQS are the new CWS’s. The CWS’s are the old standards and CAAQS are the new standards.
[AR9]Another point to consider: In addition, the elevated measurement in 2010 could be ascribed to this new equipment if the measurements remained high following the measurement. This is not the case, with the years following 2010 still showing elevated pollution measurements but not as extreme as the 2010 event.
[AR10]We might want to explore this a little more, and review the language he uses. Ultimately he disregards NOx and SO2 because it’s also produced by O&G if I recall correctly.