Skip navigation
An abstract illustration of the functioning of artificial intelligence

AI And Machine Learning for Better Energy Demand Response

Ontario start-up improves accuracy of peak prediction, helps businesses save electric costs.

The fast-growing energy management industry has witnessed strides in innovation and, while exciting, many of these new ideas never make it to market. However, after running a successful pilot last year, the Ontario-based technology company, EnPowered, has proven that it has the grit to succeed. The start-up challenged the demand response market with its market predictions tools that use artificial intelligence (AI) computing and machine learning technologies to predict energy markets. EnPowered helps its customers better manage demand response efforts to control energy costs, minimize curtailment or cogeneration events, and generate significant savings.

EnPowered piloted its technology last year with 20 Ontario-based businesses, helping them collectively garner US$40 million in energy savings. Founder and CEO, Tomas van Stee, states, “We are excited to officially launch and help more businesses save on their electricity costs. Our clients need to focus on their businesses, so we make things easy and provide clear, explicit forecasts and inform them when to curtail or alter usage. Every EnPowered customer increases savings and reduces electricity costs, with minimal operational impact.”

Businesses look to EnPowered’s forecasts to eliminate stress, reduce oversight and avoid unnecessary shutdowns. The company’s precision is also why its pilot customers, like Ensyn of Renfrew, Ontario, could amass such considerable savings. Mike Ackman, Ensyn’s general manager, says, “One of the challenges of the [DR] program was trying to do it on our own. By signing on with EnPowered, the difference was night and day, you didn’t have to worry. You could go about your business and let EnPowered worry about letting you know when to curtail. The main benefit was the dollar savings. We are looking at roughly a 40-50% reduction in energy costs.”

Introduced to control electricity grid constraints, demand response programs incentivize businesses to scale back energy usage during peak usage periods. Predicting these trends is valuable, as these fluctuations influence electricity prices and present opportunities for businesses to save millions, or generate revenue. For businesses in markets with 5 Coincident Peaks (5CP) programs, such as the Industrial Conservation Initiative (ICI) in Ontario, Canada, the problem is knowing in advance when peaks will occur and which ones are tied to expensive electricity rates.

Unfortunately, determining energy demand in advance is difficult and is becoming increasingly complex. While reliable trends exist relating to energy spikes during extreme weather events or people returning home from work each day, there are a growing number of factors at play, such as expected weather, consumer behavior, or demand response efforts of other customers in the market. Exacerbating the problem further is that these factors interact with one another, and therefore the complexity of these relationships increases exponentially. This means that even if one were to account for every factor that affects energy demand in a given area, they still need to consider how one factor influences another. For example, how does air temperature change human behavior. As a result, publicly available predictions are becoming increasingly inaccurate and businesses are responding more frequently to false peak events, thereby eating away potential savings and unnecessarily impeding production.

The market predictions platform notifies its customers of impending price sensitive peaks. Its predictions use advanced computational models that account for historical trends and current data, along with a multitude of interrelated variables, such as air temperature, wind speed, dew point, humidity, probability of precipitation, human behavior and energy response fluctuations. With 100% accuracy and 70% fewer disruptions than the market, its forecasts have shot to the top of the industry. Last year, in attempts to identify Ontario’s top five peaks, the public operator sent out 33 alerts. EnPowered’s technology narrowed this down considerably, sending out only 10.

The company pivoted from its early days as a group buying service for homeowners. EnPowered remains committed to its initial vision of helping customers save money on energy. Focused on this, the company is looking forward to its expansion across the province of Ontario and helping additional industrial businesses manage their energy usage and participation in demand response programs. Later in 2019, EnPowered is set to begin expanding across other North America markets.

EnPowered is pushing hard to support businesses in controlling energy costs through demand response and ICI management. EnPowered delivers the peak usage predictions which are crucial to customers in the demand response market. The technologies developed by EnPowered are fueled by advanced AI and machine learning technologies. Demand response programs offer significant savings to business, but they have become very difficult to manage. As a result, businesses are looking to find ways to eliminate stress, reduce oversight, and avoid unnecessary shutdowns — letting the tech users focus on their business. This is why it’s so important for predictions to be structured in a way that helps customers to better manage demand response efforts to control energy costs, minimize curtailment or cogeneration events, and generate significant savings.

Founded in 2015, EnPowered is headquartered in Waterloo, the key technology city within the Province of Ontario in Canada.

Hide comments

Comments

  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.
Publish