Yesterday I attended an interesting presentation at Energy Solutions Exhibition in Hamilton – a serious approach of Peak Power Inc to reducing demand during peak hours. Peak Power considers peak chasing a game of smarts, and they lots of them to the table.

Weather your company belongs to Class A of Class B, Global Adjustment is a major energy cost

Peak Power uses big data and machine learning to recommend time to engage on-site batteries to shave peaks. With all the smarts they estimate that their solution captures 90% of possible value of demand reduction. This year PP successfully covered all 5 peaks; it took engaging batteries 21 times over the year.

Predicting peak hours requires serious mathematics

Ducking peak hours has become a serious mathematical task. With growing number of participants, provincial peak demand does not follow weather anymore, it follows what people think the weather will be. We are approaching time when demand will behave like stock market: short term forecast does not follow fundamentals, but rather what on what one player thinks that other player will do.

List of Ontario historical peak demand hours from 2012 to present,  can be downloaded directly from top ten Ontario demand peaks archive  at IESO site.

Chasing demand peaks is a game. This year two peaks happened at 6pm during January weekend. Who knew?!

Companies that expect to succeed in this game by watching Weather Network are destined to pay dearly.