NB: this challenge description, including Figures is available here as a PDF

There are roughly 200 northern and remote communities across Canada that rely on imported diesel fuel for nearly all of their heating and electricity needs (see figure 1) [1]. The total amount of diesel fuel used annually by these communities to meet their electricity demand alone is upwards of 380 million litres, which, when burned, produce over one million tonnes of GHG emissions. This is equivalent to the GHG emissions generated annually from 312,000 passenger vehicles [2].

Many remote communities are pursuing measures to reduce their reliance on diesel fuel, through a combination of energy efficiency improvements and renewable energy projects. One example is Fort Chipewyan’s Indigenous-owned remote solar farm, the largest of its kind in Canada (see Figure 2). In combination with a battery storage system, the project is expected to supply about 25 percent of the community’s energy needs and reduce emissions by about 1,743 tonnes per year [3]. In addition to reducing GHG emissions, these projects have many additional benefits including improved local air quality, greater energy self-sufficiency and resiliency, and reduced risk of fuel spills.

The federal government has supported remote community energy projects through various programs such as the Clean Energy for Rural and Remote Communities (CERRC) program [4]. The Government recently announced an additional investment of $300 million over five years to advance its commitment to ensure that rural, remote and Indigenous communities that currently rely on diesel have the opportunity to be powered by clean, reliable energy by 2030 [5].

Proposed challenge:

Natural Resources Canada (NRCan) is currently supporting remote communities’ energy transition by modeling pathways for communities to reduce dependence on diesel fuel in a reliable, cost-effective manner. There are many options for remote communities to reduce diesel fuel consumption including energy efficiency measures such as lighting and insulation upgrades, new renewable generation such as wind turbines and solar panels, and energy storage such as lithium-ion batteries.

A significant portion of total community energy use, as high as 70% for some communities, goes toward heating buildings. In order to build accurate models to assess the available options for reducing heating fuel use, granular data on existing energy use is required (e.g. hourly thermal load data), along with detailed information on existing buildings in the community.

Canada currently lacks a comprehensive database of remote community building stock that includes residential buildings, commercial/institutional buildings such as offices, schools, and arenas, and industrial facilities such as wastewater treatment plants. Some information is available from Census data, such as the number, type and occupancy of residential dwellings (see data source 2). Data available from other sources, such as Statistics Canada’s Open Database of Buildings (see Figure 3, data source 3) or Microsoft’s Canadian building footprint data (see Figure 4, data source 4), is currently missing or incomplete for most remote communities.

The challenge is to develop a feasible approach using AI/Machine Learning techniques to generate a complete characterization of building stock, for every remote community in Canada, comprising the following:

Identification of all buildings for a given radius surrounding the centre of the community Calculation of building footprint in m2 for each building in the community Classification of buildings into archetypes (e.g. single-detached house, multi-unit residential building (MURB), commercial, industrial etc.) For MURBS, determining the number of units in each, perhaps through a count of entranceways

A visual method to distinguish between houses and other types of buildings would be a valuable contribution. For commercial buildings, distinguishing between building types with different thermal profiles (e.g. arena, warehouse/garage, municipal building) is desirable.

Another valuable outcome of this work would be to identify which buildings would be suitable candidates for rooftop solar (e.g. south facing, free of large obstructions etc.). Several methods have been developed for automated detection of rooftop solar potential [6], [7], however these have yet to be applied to remote communities in Canada.

Data sources

Some helpful public data sources are listed below. Participants are encouraged to use other publicly available datasets as required.

NRCan Remote Community Energy Database A list of remote communities is provided in the information package including the community name(s), population, latitude/longitude and other information. This is the same list of communities in the Remote Communities Energy Database available at the link below, filtered to only include communities whose primary energy source is diesel. https://atlas.gc.ca/rced-bdece/en/index.html

Statistics Canada Census Data The Census Profile provides a statistical overview of various geographic areas based on a number of detailed variables including household and dwelling characteristics. https://www12.statcan.gc.ca/census-recensement/2016/dp-pd/index-eng.cfm

Statistics Canada’s Open Database of Buildings The Open Database of Buildings (ODB) is a collection of open data on buildings, primarily building footprints, and is made available under the Open Government License – Canada. https://www.statcan.gc.ca/eng/lode/databases/odb

Microsoft’s Canadian building footprint data Contains computer generated building footprints in all Canadian provinces and territories and is licensed by Microsoft under the Open Data Commons Open Database License (ODbL). https://github.com/microsoft/CanadianBuildingFootprints

Global Visualization Viewer (GloVis) Glovis hosts publicly available remote sensing data from multiple sources. https://glovis.usgs.gov/app Community Profile - Grise Fiord, Nunavut A partial community profile is provided as an example for the community of Grise Fiord, Nunavut,including photographs of various building types.

References:

[1] Government of Canada, Remote Community Energy Database. https://open.canada.ca/data/en/dataset/0e76433c-7aeb-46dc-a019-11db10ee28dd [2] Natural Resources Canada, Greenhouse Gas Equivalencies Calculator. https://oee.nrcan.gc.ca/corporate/statistics/neud/dpa/calculator/ghg-calculator.cfm

[3] Government of Canada, News Release. https://www.canada.ca/en/natural-resources- canada/news/2019/08/solar-energy-moves-indigenous-communities-toward-a-renewable- future.html

[4] Government of Canada, Clean Energy for Rural and Remote Communities: BioHeat, Demonstration & Deployment Program Streams https://www.nrcan.gc.ca/reducingdiesel [5] Government of Canada, Backgrounder: A Healthy Environment and a Healthy Economy.

https://www.canada.ca/en/environment-climate-change/news/2020/12/a-healthy-environment- and-a-healthy-economy.html

[6] Google, Project Sunroof. https://www.google.com/get/sunroof [7] PV Magazine, Using machine learning and cheap satellite data to design rooftop solar power.

https://pv-magazine-usa.com/2019/08/12/using-machine-learning-and-cheap-satellite-data-to- design-rooftop-solar-power/

Built With

  • nrcan
Share this project:

Updates