New AI-based software to optimise hydropower operations

Upstream Tech, an innovator of AI-based climate software solutions, has announced the release of HydroForecast ST-3, the latest generation of its water forecasts.

HydroForecast ST-3, a solution for hydropower operators and water resource managers, uses AI model breakthroughs to boost the efficacy of the software.

The AI breakthroughs employed in ST-3 result in an accurate distributed model and additionally facilitates integration of specialised regional weather forecasts, such as NOAA's High Resolution Rapid Refresh.

Furthermore, HydroForecast ST-3 is designed to withstand individual input failures, produces forecasts more frequently and provides detailed per-weather source forecast traces.

Have you read?
Who is your energy transition digital superhero?
Generative AI has ‘turbocharged the power of the human’ says Octopus Energy founder Greg Jackson

Upstream Tech explains in a release that with increased frequency of extreme weather events, climatic unpredictability is causing significant challenges for sectors dependent on water resources. This is especially so for hydropower, which is subject to reduced power generation under drought conditions and, conversely, damage to hydropower infrastructure and downstream areas owing to storm surges.

Consequently, these sectors have become reliant on short- and medium-term water forecasts to make decisions, where accuracy, reliability and interpretability are paramount.

However, water forecasts are only as reliable as their inputs, like meteorological forecasts, which historically face disruptions from power outages, technical issues, and other unexpected events. In the context of extreme weather, these kinds of disruptions jeopardise the safe and efficient operation of infrastructure.

By leveraging a distributed machine learning model, HydroForecast ST-3 mitigates missing or delayed inputs, ensuring a forecast is always available during crucial operational events. For hydropower operators, this results in safer operations and base load resilience, according to Upstream Tech.

Marshall Moutenot, CEO of Upstream Tech, commented in a release: "We are trying to change the way organizations make water decisions by arming them with the best possible forecasts. ST-3 is the culmination of years of R&D conducted by our team and represents a step change in the kind of accuracy and resiliency we can offer to our customers.

"We are passionate about the energy transition and the role hydropower is playing in increasingly renewable energy grids. Arming our hydropower partners with information that will lead to safer operations in the face of extreme weather, flexible operations as our grid evolves, and optimal generation to improve revenue - those are huge wins that my team is exceedingly proud of."