Large IPP uses Bidding Benchmarks to support analytics stack

How a developer/owner/operator of large renewable generation in California and Texas uses Enertel's Bidding Benchmarks to reduce evaluation time, improve their P&L, and support their trading operations.

Bidding Benchmarks shows load, solar and wind profiles how different trading strategies might select different markets in the day-ahead market.

Intro + Company Snapshot

A leading player in the clean-energy sector boasts a substantial portfolio of over 2 GW of operating solar PV and 2.4 GWh of battery storage, with an additional 6 GW+ of projects slated for groundbreaking in the next few years.

The company's advanced analytics team collaborates with its trading arm to optimize its expanding asset portfolio. Key inputs for their daily trading decisions include comprehensive price forecasts for each asset, derived from both internal models and external sources.

Business Challenge

Since 2022, they have used Enertel forecasts to inform their market assessments and anticipate risk across various market segments for their assets. During this period, Enertel consistently introduced new machine learning models, presenting a significant challenge for the company: effectively translating these continuous model enhancements into quantifiable profit and loss improvements, given their team's capacity.

"We knew that the Enertel models were picking up patterns that complemented our own analysis, but there was a bottleneck to figuring out how to integrate those patterns into new trading strategies, and ultimately an improved P&L"

Senior Director, Trading at Large IPP

Historical backcast data provided to clients includes hourly data with 12 separate percentiles. This means a 6-month backtest package for a single node can include over 500,000 data points. For a rapidly expanding enterprise focused on growth, the time needed to integrate new models into an existing workflow proved to be too long.

The company required a solution that could drastically reduce backtesting time from weeks to hours, while supporting the extensive experience of their trading team. Although metrics like RMSE, CRPS, and calibration rates were valuable, the ultimate measure of success for new models needed to be their direct contribution to enhancing trader profitability and mitigating risk.

Section Recap

• The analytics team at a fast-growing IPP had many competing priorities to support the trading team

• New modelling techniques from Enertel took too long to effectively integrate into trading strategies

• Metrics were easy to calculate, but didn't always translate to trader profitability

Solution

Enertel's bidding functionality empowers analytics team members and traders to evaluate transparent bidding strategies directly trained using Enertel's forecasts. These strategies are drawn from a diverse library of risk profiles, enabling traders to select appropriate benchmarks. Historically, the tool offers transparent insights into the core metrics the company prioritizes: cumulative profit, volatility in returns, and the ability to conduct granular analysis of specific days and hours.

"The bidding functionality from Enertel removes most of the backtest analysis that used to take weeks and provides actionable bidding recommendations alongside our own, delivered daily to our analytics and trading teams."

Senior Director, Trading at Large IPP

Enertel's solution eliminates the need for a data scientist to bridge the gap between predictions and decisions, allowing traders to evaluate forecasting models in terms they inherently understand: profit, value-at-risk, and Sharpe ratios. It also ensures transparency, letting traders double-click into any given day and review the forecasts that informed the bidding suggestions.

"The team at this renewable developer is technically savvy and highly detailed. They needed to save time on boilerplate backtesting, which made them a great candidate for the new bidding product."

David Murray, cofounder of Enertel.

Section Recap

• The analytics team can backtest new modelling techniques directly in the Enertel application, with no work required

• Calculated metrics matter for their business: profit, value-at-risk, and sharpe ratios

• Different trading strategies can be tailored to their view of the market, meaning their in-house expertise can do more analytics with fewer menial tasks

Conclusion

This forward-thinking renewable owner, a recognized leader in its field, has demonstrated exceptional capacity in scaling its contributions to the U.S. power grid. A rapidly expanding portfolio of assets means their dedicated team of analysts and traders remain focused on critical performance indicators, rather than encumbered by the complexities of historical data analysis. In the broader landscape of energy transition and technological advancement, optimizing even minor backtesting tasks yields substantial, compounding advantages over time.

Enertel AI provides short-term energy price forecasts and bidding strategies for utilities, independent power producers (IPPs), and asset developers. Take a look at our Bidding Benchmarks product in an interactive demo.