Energy Innovation Brief
Issue 30 | April 27, 2023

In Western Canada and around the world, the energy sector is rapidly transforming to one that promises to be cleaner, greener and more efficient. Each month, the Canada West Foundation’s Energy Innovation Brief brings you stories about technology innovations happening across the industry – in oil and gas, renewables, energy storage and transmission. If you have an idea for a story, email us at:

Exploring AI in energy

This month’s EIB was written by CWF’s Johnson Shoyama Policy Analyst, Tyler Robinson. This special issue explores the ways that Artificial Intelligence (AI) is being used by the energy sector to improve safety, reduce costs and improve environmental outcomes. Increasingly, AI is being tasked with jobs that are best carried out by an always-on, incredibly fast, non-human mind: monitoring, image recognition and predictive modelling of supply and demand, for example. But should people in the energy sector be worried that AI will make them obsolete? To answer that question we went directly to the source. We asked ChatGPT, an AI program that generates responses to text inputs, “Should people in the energy sector be worried about losing their jobs to AI? Please answer in the style of the Muppets’ Swedish Chef.” Here’s what it had to say:

“Bork bork bork! Vell, let me tell ya someting, my dearie. In da energy sector, da use of AI is on da rise, but dat doesn’t mean people should be worried about losing their jobs. In fact, AI can be a helpful tool to improve efficiency and accuracy in da industry. Plus, it can create new job opportunities, like AI technicians and data analysts. So, no need to be a-fretting like a chicken with da head cut off! Embrace the technology and let it help you cook up a storm in the energy sector! Bork bork bork!”

Not bad for a robot.

01|  Process optimization
02|  Predictive modelling 
03|  Balancing supply and demand on electrical grids
04|  Monitoring and fault prevention 

Process Optimization

Process optimization harnesses the data processing power of AI to find efficiencies and allow organizations to get more bang for their buck—whether that means more throughput in a pipeline, a more efficient power plant, or simply using less energy to do the same amount of work. It’s an area where AI excels as it requires sorting through and learning from massive volumes of data—something machine learning is incredibly good at.

  • Calgary-based CruxOCM replicates the way that auto-pilots assist—but don’t replace—human pilots, but for control room operators in the oil and gas transmission and storage industry. The software uses machine learning to automate those parts of control room operations that are labour-intensive, prone to error, and subject to changing conditions. It allows operators to run multiple systems with the press of a single button and has helped to expand pipeline throughput capacity by up to 10 per cent, with similar reductions achieved in power usage.
  • QuantumBlack AI is a neural-network model, a form of machine learning modelled after the human brain that can learn to identify patterns in data. The technology is owned by McKinsey and has been deployed across multiple industries, including the energy sector. In Texas, power generator Vistra Corp. used QuantumBlack to balance thermal efficiency and fuel requirements at its Martin Lake power plant. By parsing through two years’ worth of data, QuantumBlack more accurately tuned the plant’s fuel efficiency, resulting in an annual reduction of 340,000 tonnes of CO2 emissions and $4.5M a year in savings.
  • EZ Ops is an oil and gas operations software platform that independently prioritizes activities for field teams, based on urgency, cost efficiency and location in order to reduce truck rolls and save on operating costs, while maintaining safety. EZ Ops estimates that with their optimizations they can reduce emissions and lower operating expenditures by 15 per cent or more.

Predictive Modelling

Predictive modelling uses statistical analysis based on historical data to predict future outcomes. By analyzing the patterns in massive amounts of data, AI can make predictions that help companies take proactive action to address issues rather than waiting for problems to arise.

  • Brainbox AI reduces energy waste by adding AI to HVAC systems. This “predictive brain” learns how to use less energy while optimizing comfort within the building. By predicting future demand, Brainbox can make adjustments in advance, rather than being reactive like traditional HVAC systems, in turn lowering the emissions impact and energy bills all at once.
  • Using technology based on IBM’s Watson, the same supercomputer recently featured on Jeopardy, researchers are eliminating some of the uncertainty involved in solar forecasting. Funded by the U.S. Department of Energy SunShot Initiative, the endearingly titled “Watt-Sun” uses weather modelling, solar radiation measurements and cloud tracking to improve solar forecasting by as much as 30 per cent. This kind of certainty can help ensure that renewable energy is fully integrated into the grid and minimize the problems of intermittency.

Balancing Supply and Demand in Electrical Grids

Balancing supply and demand on the electrical grid is a complex process, especially since it must be done in real time to avoid temporary blackouts. AI can help by collating data from weather forecasts, facility outputs and trends in consumer behaviours to inform grid operators what adjustments must be made—or even make the changes itself. By doing so, AI increases reliability and helps utilities maximize the benefits of storage, renewable and distributed resources.

  • Ottawa-based BluWave-ai is being used to predict load on the grid across all of Ontario, after initial successes in India and the U.S. The software was trained using five years worth of models and data, and now provides day-ahead and intra-day forecasts to the province’s Independent Electricity System Operator (IESO).
  • Stem’s Athena AI software helps companies manage their onsite power generation and storage assets by automating demand-side management activities. The software runs thousands of scenarios using data such as historical power use, electricity prices and weather to determine the best times for companies to use onsite solar panels, charge battery storage or draw power from the local electrical grid. By charging batteries when prices are low and exporting power when prices are high, the software can maximize the profitability of a company’s energy system.

Monitoring and Fault Prevention

While humans can be flexible and adapt to many situations, we do have a few limitations. We can only be in one place at a time, we’re heavy and cumbersome to transport, our bodies don’t do well in toxic environments, and even the best of us can struggle to pay close attention for prolonged periods. Using AI and sensors for remote monitoring mitigates many of these problems and allows energy companies to keep tabs on assets, engage in preventative monitoring, maintain site security—and even just delay human travel until it is required.

  • Osperity’s system gathers visual data from cameras and analyzes the data using AI-based “computer vision.” It learns to recognize objects and activities so that its industrial users are alerted when specific incidents occur. For oil and gas, this could include detecting failing equipment or leaks. For electric utilities, it could be sensing unusual conditions at a remote substation. For EV charging stations, it can be used to detect changes in cleanliness at the station.
  • Qube has built a system that is part of the “industrial Internet of Things”, combining small, solar-powered continuous monitoring stations with back-end software to continuously monitor and measure emissions of carbon dioxide, methane, and other gases. AI and machine learning are used to better quantify, locate and classify emissions so they can be reduced. Qube states that its technology is not only 80 per cent cheaper than average optical gas imaging (OGI) surveys, but also up to 50 per cent more effective in reducing emissions than tri-annual OGI surveys. The technology is currently being used in Alberta at Enhance Energy’s sites.

The Energy Innovation Brief is compiled by Brendan Cooke and Marla Orenstein. This month’s edition features contributions by Tyler Robinson. If you like what you see, subscribe to our mailing list and share with a friend. If you have any interesting stories for future editions, please send them to .