Joint oil sands monitoring program emissions inventory report

April 2016

Executive Summary

The Joint Canada-Alberta Implementation Plan for Oil Sands Monitoring (JOSM) identified air emissions inventories as critical to obtaining accurate information on what is being emitted to the atmosphere from all activities related to oil sands developments. Such information is necessary for understanding the impacts of oil sands operations on the surrounding environment. Current and comprehensive air emissions inventories are also key inputs for air quality modelling.

As part of JOSM, Environment and Climate Change Canada’s Global Environmental Multi-scale - Modelling Air quality and CHemistry (GEM-MACH) model is being run to provide high-resolution forecasts of air quality (AQ) in the oil sands region. It is also being used to improve understanding of the underlying science related to oil sands emissions and their fate in the environment by helping to build a coherent picture of the processes underlying air quality in the region, connecting emissions to atmospheric concentrations and then to deposition in the environment.

Over the past several years, the JOSM Air Emissions Inventory Working Group has endeavoured to gather, review and identify the most accurate existing oil sands region emissions data, and improve the accuracy of related information required for modelling. These activities were carried out in order to support the GEM-MACH modelling and to help guide specific monitoring activities attempting to address potential scientific knowledge gaps related to oil sands emissions. This work has resulted in the development of a comprehensive oil sands emissions inventory, called the Joint Oil Sands Monitoring Program Emissions Inventory (JOSM Inventory). The JOSM Inventory is based on a synthesis of the best available information from several existing emissions inventories and related sources, up to October 31, 2014.

This report details the development and compilation of the JOSM Inventory, including identifying the data sources, compilation procedures, gap analysis, and how priority data gaps were addressed. This report also provides summary information on the JOSM Inventory and accompanying database, details the associated emissions processing that was carried out to generate gridded input emissions files for modelling, and outlines the setup of the GEM-MACH modelling system.

The JOSM Inventory was developed by first reviewing the existing 2006 Environment and Climate Change Canada National Pollutant Release and Air Pollutant Emissions Inventories, and ancillary data for the spatial and temporal allocations, against the stated objectives and requirements of the GEM-MACH modelling project for the oil sands region. This review revealed that the existing 2006 modelling inventory was not sufficiently detailed or up-to-date for the oil sands study area, which is roughly 100 km by 100 km in size and is located in the part of Athabasca oil sands region just north of Fort McMurray (see figure below). Numerous specific data gaps and areas for improvement were identified to meet modelling needs. These data gaps were then prioritized and those that were identified as “high” were further investigated.

JOSM Inventory Oil Sands Study Area (black outline region)

Various available inventory datasets were then reviewed to attempt to fill these data gaps, and this included reviewing and utilizing data from Environment and Climate Change Canada, the Alberta Energy Regulator, Alberta Environment and Parks, as well as relevant Environment Impact Assessments for facilities in the region. The main datasets used in the inventory were: stack-specific hourly emissions measured by Continuous Emissions Monitoring Systems, the Cumulative Environmental Management Association Inventory, the 2010 Air Pollutant Emissions Inventory and the preliminary 2013 National Pollutant Release Inventory. These inventories represented the most specific and recent information available for the primary study area in the Athabasca oil sands region.

Although by necessity the JOSM Inventory was compiled by blending inventory datasets that cover several different calendar years, it is intended to represent average annual air emissions for the year 2013, and improvements were specifically focused to represent August and September 2013. These improvements were done in order to use the inventory for air quality modelling in support of JOSM summer 2013 intensive air measurement campaign that took place during that period, and afterwards, for comparison with estimates of emissions and process-based analyses from the summer campaign.

The JOSM Inventory focuses on oil sands sources within the study area, but it also includes oil sands sources outside of the study area, non-oil sands sources within and outside of the study area, and industrial and non-industrial sources for the rest of the province of Alberta. The JOSM Inventory also includes sources across North American, since the GEM-MACH modelling system also includes most of North America within its outermost modelling domain. The JOSM Inventory includes emissions data for carbon monoxide (CO), oxides of nitrogen (NO_X), sulphur dioxide (SO₂), particulate matter less than 10 micrometres in diameter (PM₁₀), particulate matter less than 2.5 micrometres in diameter (PM_2.5), total and speciated volatile organic compounds (VOC) and ammonia (NH₃).

The JOSM Inventory represents the best information available to the JOSM Air Emissions Inventory Working Group for the study area. However, the review was focused on the study area, and other areas of the oil sands development regions may require further review and improvement. The table below presents the JOSM Inventory industrial emissions for the study area and the three oil sands development regions, as well as total Alberta emissions (excluding natural sources). By species the study area represented between 1% and 20% of provincial anthropogenic criteria air contaminant emissions. The three combined oil sands regions accounted for between 1% and 24% of provincial emissions, which highlights the significance of the large oil sands operations in the study area.

Note: Emissions data are for the year 2013 for the facilities in the Athabasca oil sands, and the year 2010 for all other sources, see Chapter 2.

The emissions data included within the JOSM Inventory consists of several dozen emissions processor and modelling files. A Microsoft Access database, called the JOSM Emissions Inventory Database, was also developed to help make portions of the JOSM Inventory useable for stakeholders not familiar with emissions processor and modelling files. The Database includes all industrial sources in the three oil sands development regions that were included in the JOSM Inventory. However, it does not include all of the data and information included in the entire JOSM Inventory. The Database does not include industrial sources outside of the oil sands development regions, non-industrial area or mobile sources inside or outside of the regions. It also does not include temporal profiles, spatial surrogates, GIS shapefiles, or other related supporting input information needed for emissions processing. These additional sources of information are available as part of the inventory, but are not part of the Access database itself.

Air quality models such as GEM-MACH cannot directly use emissions inventories as inputs. Instead, the inventories must first be converted to hourly emissions on a model grid using additional data, and software packages called emissions processing systems. Models require emissions inputs which are resolved on the temporal and spatial scales needed for modelling (the input emissions are required hourly and placed appropriately on the model grid). These data, as well as the inventory, are used by an emissions processing system to create model-ready hourly gridded emissions files.

The SMOKE (Sparse Matrix Operator Kernel Emissions) emissions processing system was used on the JOSM Inventory to generate the required modelling input files. This involved temporal allocation, chemical speciation and spatial allocation of emissions. Temporal allocation was performed to convert the inventory annual emission rates into hourly values. Chemical speciation was carried out to break down certain criteria air contaminants into individual model species, particularly for volatile organic compounds and particulate matter. Spatial allocation was performed to distribute non-point source emissions to appropriate grid cells.

Several methodological improvements were made to enhance the accuracy and representativeness of emissions processing for the largest emitting facilities in the study area. To account for the large spatial extent of the six Athabasca oil sands mining facilities, a set of 18 facility-specific spatial surrogates was generated for allocating emissions within each of these mining facilities. Facility-specific monthly temporal profiles and VOC speciation profiles were also developed for these six facilities. After emissions processing was completed, the resulting gridded, speciated, hourly emissions were then used by the GEM-MACH model.

Three levels of spatial nesting were used for the GEM-MACH simulations. An outer grid with 10-km horizontal grid spacing was used to cover most of North America. A middle 10-km grid with a different orientation was used to cover most of north-western Canada and the north-western United States. Lastly, a grid with 2.5-km grid spacing was used to cover the provinces of Alberta and Saskatchewan, including the study area. The nested combination of these three successive grids was used to produce the best possible meteorological and air quality forecast from the final 2.5-km nest of the model.

Two types of simulations have been performed with GEM-MACH for the study area. The first type includes near-real-time forecasts of air pollution on the 2.5-km domain. These forecasts are being run continuously as an experimental forecast by Environment and Climate Change Canada, and were also used during the August to September 2013 timeframe to guide instrumented aircraft flights for the JOSM summer 2013 intensive air measurement campaign. The second type includes retrospective simulations of the measurement campaign time period, to evaluate the model using observations, and use the model to help elucidate the processes which gave rise to the observations. Both sets of simulations include outputs that allow the assessment of human health impacts (e.g., Air Quality Health Index) and ecosystem impacts (e.g., deposition of sulphur and nitrogen to ecosystems within the 2.5-km domain). The model outputs are also being used to develop methodologies to evaluate the suitability and effectiveness of the spatial placement of monitoring site locations.

Although the JOSM Inventory represents the best information available to the JOSM Air Emissions Inventory Working Group at the time it was compiled, there were several data gaps identified that are still being addressed by ongoing inventory development at Environment and Climate Change Canada. This ongoing work includes further examination of the mine-fleet emissions, the investigation of the coverage of small non-thermal bitumen extraction sites, the use of additional Continuous Emissions Monitoring System (CEMS) measurements and associated release parameters, improvements to biogenic emission rates, and the use of aircraft observations to infer emissions magnitudes and chemical speciation. In addition to these ongoing improvements, as discussed near the end of this report, there are further areas where the JOSM Inventory could potentially be improved.

Lastly, air emissions from the oil sands are changing constantly, as oil sands production increases together with the introduction of new technologies that are intended to improve environmental performance through reduced emissions. In order to remain useful and representative for use in ongoing GEM-MACH modelling and oil sands environmental monitoring work, the JOSM Inventory will need to be maintained by updating it on a regular basis to reflect changes in sources and emission rates in the region. Remaining priority data gaps will also need to be addressed through enhanced or additional monitoring and by continued dialogue with industry.

For information on the report and the emissions inventory data, click here:
Full Report
JOSM Portal