Why We Invested: Sunairio

Changing climate = changing weather, and we’re not prepared

A changing climate means changing weather. No deep insights there. Yet the energy sector currently relies on historical weather data when modeling anything beyond 15 days. The industry uses a small sample of historical years to form a “Typical Meteorological Year” (TMY). The problem? There’s no such thing as typical anymore.

Use of backward-looking weather data in energy investment and planning is causing costly mistakes. Winter storm Uri in Texas in 2021 left millions of people without power for days, caused hundreds of deaths, left up to $90B of damage, and saw prices spike to $9,000/MWh (versus market average of $35/MWh). There are a range of interventions that could have provided greater resilience. These include structural solutions that require significant investment and political support – e.g. improved planning processes at ERCOT ISO to account for extremes and climate trends, winterization of existing critical infrastructure, or connection of the Texas grid to surrounding national grids. In the short term, improved weather modeling can help the energy industry be more proactive to prevent the most severe impacts from occurring while more fundamental solutions are sought – e.g. by enabling more accurate real-time and next-day generation scheduling, alerting grid operators to the need for emergency supply procurement, and supporting flexible demand response programs.

Grid variability is increasing as more variable renewable energy sources are coming online. Yet the industry continues to depend on small samples of historical weather data sets that – due to being averaged over long-periods and large areas of land – prevent an accurate understanding of weather extremes, volatility, and trends. In addition, the primary weather variables that affect renewable energy assets – such as irradiation and hub-height wind speed – are highly localized, but the only long-term weather modeling available today (physics-based global climate models) generate regional data that’s too coarse to be of commercial use in asset-level energy models (e.g. relying on data from the nearest airport rather than where the generation assets are located). The lack of high-resolution climate data makes planning difficult for everyone in the energy industry: traders, renewable project developers, grid operators and utilities.

‍Technology Developments

There has been a lot of attention on the need to move away from physics-based models and the associated potential for AI. For example, GenCast (from Google DeepMind) announced its ability to provide better 15-day forecasts than the best physics-based short-term weather forecast (even outperforming its 2023 DeepMind predecessor). Other big tech has also been leaning in – Microsoft has Aurora, NVIDIA has ForeCastNet. And so has venture capital – companies like Jupiter Intelligence and Tomorrow.io have raised hundreds of millions of dollars.

More fundamentally, recent improvements in cloud computing have removed the computational barrier to analyzing enormous amounts (hundreds of TB) of meteorological data and making insights available in real-time.

Buoyant spent the last year meeting with both generalists using AI for weather predictions and vertically integrated solution providers. Today’s crop of new companies are very nimble and are accomplishing a lot without as much overhead as more established players. We are excited to get in the game with our latest investment.

‍Our Investment in Sunairio

‍Sunairio’s models leverage the latest foundational techniques and proprietary technology to deliver actionable probabilistic forecasts of energy demand and supply (at asset and regional levels), incorporating climate change. They provide:

• 15 day - 15 years forecasts: Forecasts beyond 10 days are notoriously unreliable. Sunairio has developed a solution for this unmet need by providing probabilistic forecasts up to 15 years out which are more accurate than any historical data approach and even NOAA’s own flagship seasonal forecast. Traders can use these forecasts to manage their balance-of-month trades, developers can get more accurate production forecasts, and utilities can be more confident in their resource planning.
• High spatial and temporal resolution: Sunairio provides resolution that allows users to model down to the asset-level and relevant times of day. As the image below demonstrates, this level of specificity can yield meaningful differences in data models.

• Commercially-ready solutions: Sunairio’s tools have been developed as an application layer to integrate with customers’ existing workflow. Customers don’t need to change how they make decisions, they just get better information that corresponds to the risks they face. For example, with Sunairio, a utility knows the risk of lower than expected wind speed at its generation assets during peak demand. A similar logic applies across all types of customers including traders and developers.

‍Strong Founder-Market Fit and an Impressive Team

Founder/CEO, Rob Cirincione, felt the pain of inadequate models in his over-a-decade as an energy trader at Constellation Energy and NRG. This deep understanding of the problem has driven Sunairio to deliver products that actually solve problems faced by the industry today. Rob has worked with leading academics at Northwestern and University of Massachusetts to develop the underlying modeling, and with strong engineers and climate scientists who have scaled products at a venture-backed startup before (CTO, Eric Hewitt and Senior Data Scientist, Tim Ivancic). This lean team of 5 is already punching above its weight and we can’t wait to see what the next stage of growth brings.

Welcome to the Buoyant Boat

Buoyant’s founding investment thesis incorporated both climate adaptation and mitigation, and Sunairio supports both. On adaptation, Sunairio allows the energy market to be more resilient by better predicting outlier weather events. On mitigation, Sunairio reduces production risk associated with bringing more renewables onto the grid, thereby supporting the energy transition.

Sunairio also aligns strongly with Buoyant’s digital thesis. Their software tools enhance the impact of existing technology (e.g. solar / wind) while new climate tech continues to scale (e.g. long duration battery storage). 

We are also privileged to be investing alongside Constellation Technology Ventures, MassMutual Venture’s Climate Technology Fund, as well as existing investors including Laconia Capital and Thin Line Capital.

If you’d also like to be a part of the journey, Sunairio is hiring! See open roles here.

See more detail in the press release here and from Axios here.

‍