Enel and Myst AI: Optimizing energy forecasts
Predicting the energy production of renewable power plants — as well as the expected electricity load by consumers — is crucial to ensuring that utilities can run their day-to-day operations efficiently. Utilities like Enel rely on day-ahead and intraday forecasts for energy production both in order to sell this power into electricity markets and to purchase any additional power needed to meet customers’ needs.
Predicting the future is no simple task, and building forecasts that are highly accurate and reliable can be a time-consuming and challenging endeavor. After seeing the need to provide energy companies with tools to quickly create highly accurate forecasts, Pieter Verhoeven and Titiaan Palazzi founded Myst AI, a Google-backed time series forecasting platform that helps energy companies predict near-term load, renewable production and market prices. This technology enables data science teams to deploy accurate forecasting models in minutes and is specifically focused on the energy industry.
In 2018, Enel Green Power began working with Myst’s data science team to develop forecasts for Italy’s macrozone imbalance, in order to bid on asset production in intraday markets. Enel Energy and Commodity Management followed suit a year later. This type of forecasting is especially important, given that Enel’s Italy team alone is responsible for selling 2.5 GW of solar and wind energy production in wholesale markets.
But, why is forecasting essential to energy production?
As intermittent renewable energies are becoming more ubiquitous, forecasting is becoming increasingly important in ensuring that the grid stays balanced. Solar and wind energy production are affected by weather conditions, so having an accurate prediction of how much power will be produced by these resources at any point in time is key to an optimized, balanced grid. “Myst’s vision is to create a world in which demand and supply are in continuous balance, so that resources are used efficiently and sustainably for a better planet,” says Myst COO Titiaan Palazzi.
As many energy companies look to grow their portfolio of renewable assets, the Myst Platform enables faster model development and deployment workflows that can help data science teams ship new models faster, thereby enabling company growth.
Energy consumption data can reveal trends and predict future patterns in order to reduce risks, decrease carbon emissions and increase profitability. Machine learning models are perfectly suited to this task, as they can learn from past data trends to predict the near future. “We want to support organizations in the transition to a carbon-free electricity system by providing data science teams with the tools to supercharge their forecasting capabilities,” adds Palazzi.