Image of a calendar focussing in on the 19th

What happened on July 19th, 2019?

Missed a Peak on July 19th, 2019? We didn’t. pTrack™ was able to accurately predict the peak by taking into consideration reaction to IESO data. Consumers who relied solely on IESO estimates would have missed the peak day since the peak hour was inaccurately projected by the IESO peak tracker.
Did your Global Adjustment, Peak Demand Factor (PDF) Suffer From a Missed Peak? It Didn’t Have To.

It was a warm summer Friday, with temperatures brushing 30 degrees. Although there was a slight chance of rain, people preferred to stay indoors to avoid the heat. For participants in the ICI program looking for Global Adjustment (GA) savings, it was a perfect day to anticipate a peak.

The IESO Global Adjustment (GA) cost for Class A consumers participating in the Industrial Conservation Initiative (ICI) is calculated based on the consumer’s average demand contribution to the top five peak hours over a year-long base period, May 11th to April 30th. This means that consumers who track these Coincident Peaks and curtail their energy use during these hours of high demand will pay substantially smaller GA charges on their monthly electricity bills. Many ICI program participants rely on peak prediction providers to minimize the number of times they need to activate their curtailment action plan. Thus, accurate peak management is crucial. On hot summer days (like on July 19th, 2019), the likelihood of a peak increases as individuals crank up their air-conditioners.

pTrack™ Prediction

At 10am on July 19th, 2019, Edgecom Energy’s ICI Peak Prediction Service, pTrack™, predicted a 93% chance of a peak occurring between 12pm and 4pm in the afternoon. The pTrack™ Model’s initial projection was 21,570MW at 12pm-1pm and a small chance of a rebound peak of 21,994MW at 5pm-6pm.

The small chance of a rebound peak was due to two reasons:

  1. Participants who curtailed earlier during the day would end their curtailment at 4pm or 5pm, resulting in higher consumption immediately afterwards.
  2. Re-increasing temperatures after the rain passed over the GTA would lead to increased energy consumption.

Due to this, the pTrack™ model predicted a slight chance of a rebound peak, and to accommodate that, the end of the curtailment window was extended from 4pm to 5pm so customers had a chance to continue to curtail to protect against rising PDF and GA charges in the event of a rebound peak.

IESO Predictions

The Independent Electricity System Operator (IESO) predicted a peak of 22,200MW at 5pm-6pm, expecting the rebound peak to be more significant. As the cool rain passed over the GTA, the IESO projection for 5pm and beyond was significantly reduced due to continued curtailment activity anticipating a rebound peak.

However, risk analysis in the pTrack™ model revealed a higher chance of peak earlier in the day as IESO’s prediction would lead to consumers curtailing significantly during the later hours. As the pTrack™ model was updated with this new information and the chance of a rebound peak was lowered, a Resume Regular Operations notice was sent at 3:54pm. The curtailment window for pTrack™ was 12pm-4pm, which deviated from the IESO’s projection of 5pm-6pm.

The Result

In the end, the Coincident Peak occurred at the hour between 12pm-1pm, which fell right under the pTrack™ curtailment window of 12pm-4pm. Overall, approximately 200MW of curtailment occurred during 12pm – 4pm, and a much higher 800MW of curtailment occurred during 5pm – 8pm. Curtailment, combined with continued cooler temperatures and sporadic rain, significantly reduced the rebound peak, and the earlier peak remained as the number 4 peak of the base period.

Taking into consideration the reaction to IESO data, pTrack™ was able to predict that energy demand would be substantially greater earlier in the day, when most users were not curtailing their consumption. Consumers who relied solely on IESO estimates would have missed the peak day since the peak hour was inaccurately projected by the IESO peak tracker. Using advanced machine-learning and AI, pTrack™ was able to accurately predict the peak hours, saving customers thousands of dollars on energy bills.

Actual Demand, IESO Projections, and pTrack Projections on July 19th, 2019
Share the Post:

Related Posts

Download Our
Free eBook

The Guide to Industrial Energy Management will provide you with the foundational knowledge for understanding IoT and AI in the energy industry.

The Guide to Industrial Energy Management