AI can power green growth

The technology is allowing firms to unlock the vital asset data to inform investment decisions and validate green acquisitions, says Gareth Brown, chief executive of Clir.

Gareth Brown

It’s increasingly clear to many leading fund managers that renewable energy is a smart investment decision that will continue to weather wider market challenges. But many new and prospective investors worry about the consistency and long-term performance of wind and solar assets owing to resource risk – a concern exacerbated by a lack of clarity over how technical performance translates to asset financing and returns. Without a full understanding of asset performance, investors cannot properly assess the risk profile of an asset, resulting in poor valuation before an acquisition or unnecessary costs after.

Many argue this uncertainty can be solved by collecting more data, but modern renewable energy generators already collect more than enough data to answer questions around long-term energy yields and risk. The missing step that turns asset performance data into a real understanding of asset performance and performance potential is putting this data into the assets’ real-world context.

Analysing data streams from the asset alongside data streams from its environmental context, such as resource levels and the location of surrounding structures, provides investors with a full picture of renewable energy assets’ performance and highlights whether any lower-than-expected output is truly the result of a fluctuation in resources or of underlying technical defects or operational missteps.

The full picture

Traditionally, generating these contextualised, in-depth insights prior to acquisition can take consultancies weeks, if not months. Short cuts are therefore taken, which can lead to reduced value.

Instead, investors can use domain-specific data labelling, combined with machine learning and artificial intelligence, to generate a full picture of the asset in hours, thus avoiding any drag on the deal timeline and cutting unnecessary costs.

In-context data analysis mediated by domain-specific AI allows costly issues, such as imminent gearbox failure, to be recognised much earlier than standard methods via unusual patterns of temperature and vibration. A prospective investor can use these findings to push back on a project’s valuation based on the cost of downtime required to repair the turbine and the value of replacement parts.

“Investors with a contextual, data-based understanding of an asset can secure a great deal for an easily optimised project”

Likewise, opportunities for operational optimisation are often overlooked, resulting in a project being valued below its potential. Investors with a contextual, data-based understanding of an asset can therefore secure a great deal for an easily optimised project. Asset owners that apply this level of understanding to their existing fleet can benchmark any prospective investments against their own, high-performing assets. By leveraging the potential acquisition against a fleet’s-worth of data, the investor can work out the true opportunity and risk.

A full picture of asset performance enables investors to maximise output from their newly purchased assets, and quickly. This increases revenue and provides certainty around returns and production risk. Investors that take a data-driven approach are able to confidently assess and value potential acquisitions. This allows them to commit to a greater share of renewables in their portfolios.