Evaluation of Emissions in the Upstream Oil and Natural Gas Industry

David Picard, Clearstone Engineering Ltd.
GL 09-9183-50

Alberta Environment is currently re-evaluating the ambient air quality objectives (AAQOs) for the province. This initiative includes reviewing the values of currently targeted substance and potentially adding additional substances to the list. A Stakeholder Advisory Committee (SAC) has been established to participate in this review and evaluation process. The first two substances to be reviewed are Acrolein and Arsenic.

The purpose of this study is to provide upstream oil and natural gas (UOG) industry with facility related data that can facilitate defensible science-based decisions in this matter, and allow potential risks to the UOG industry to be assessed.

Accordingly, a series of source emission tests were conducted at facilities in various
operational regions of Alberta covering a range of different UOG industry subsectors.
The emission tests were conducted on a wide range of sources types. Operational
regions included the south east, central, north east and north. Industry subsectors
included sweet gas, sour gas, conventional oil, cold heavy oil and thermal heavy oil
production facilities. In total, thirteen combustion source samples were collected from
six different combustion source types located at seven different facilities across

Source sampling was completed using two types of sampling trains. Organic
substances and inert gases were sampled using a canister sampling train and
inorganics (metals) were sampled using a two stage impinger train. Canisters contents
and impinger solutions were analyzed by Alberta Innovates. The organics protocol
quantified the target substance Acrolein, and approximately 187 other substances.
The inorganics protocol quantified Arsenic, and approximately 35 other substances.
In addition, the fuel gas associated with each combustion source was sampled for
organic substances and inert gases using the same organics analytical protocol.

A rigorous material balance, considering all fuel and flue gas substances identified,
was conducted for each combustion source to determine the actual air to fuel ratio,
combustion efficiency and flue gas to fuel gas ratio. Based on the results of this
material balance, emission factors for all compounds expressed in terms of mass
emission per unit of energy input were determined.

Acrolein was observed in only one combustion source emission test and the
determined emission factor was 9.09E-5 kg/GJ. This value is well above the detection
limit of 7.91E-10 kg/GJ. The emission was associated with a compressor engine that was operating with a combustion efficiency of 95.6 , well below the average
efficiency of 99.5 for all of the combustion devices samples. The emission factor
determined for Acrolein was within the published range for 4-cycle Rich Burn,
Natural Gas Internal Combustion Engines (1.13E-3 – 2.65 E-6 kg/GJ).

Based on these tests, it is appears that acrolein formation and emission from
combustion sources may be associated with poor combustion efficiency and a follow
up sampling program is recommended to potentially confirm this hypothesis. For
combustion sources operating within the design specifications of good to excellent
combustion efficiency, acrolein formation and emission does not appear to be an

Arsenic was observed in all combustion source emission tests and the average
emission factor was determined to be 2.40E-11 kg/GJ, approximately two orders of
magnitude above the lower detection limit of 1.46E-13 kg/GJ. Based on evaluations
of combustion air and fuel gas as the potential source of arsenic in the flue gases, it
was concluded that fuel gas was the most likely source. Calculated theoretical
concentrations of arsenic in the fuel gas were within the range noted by others for
arsenic in natural gas.

A sampling program to quantify arsenic and potentially other metal substances in raw
natural gas, process fuel gas and sales gas in various Canadian regions is

In addition to acrolein and arsenic, numerous other organic and inorganic substances
were identified and quantified. Each source was characterized by its own suite of
substances with some substances being associated with several sources. Emission
factors were determined for all substances for each source sampled. These results
serve as a database for future assessments of trace comtaminant emissions.