(When Others Thought We’d Miss)
Authored by Son Phan, Research Scientist, REsurety
Research Scientist
REsurety
At REsurety, our research and power markets team is dedicated to modeling energy prices, project generation, and emissions impact. The team is interdisciplinary: we have experts in power markets, statistics & analytics, meteorology, and grid operations.
Why do we need such a deep bench of experts? Because forecasting electricity prices is notoriously complex. These values determine what projects get built, sold, and invested in – months before we’ll know whether the wind will blow or the sun will shine. It’s high stakes, and our goal at REsurety is accuracy.
If you’re an energy buyer, we know better forecasts mean better budgets – helping you to engage internal stakeholders, build confidence in your portfolio strategy, and plan for future investments.
If you’re an investor, better forecasts help reduce uncertainty around whether projects will hit their rate of return or lead to stable, long-term yields.
And if you’re a trader, credible data means hedge providers can more accurately price offtake agreements. Improved pricing means better risk management with an eye towards the bottom line.
With so much in play, we take this job very seriously, and invest resources internally to focus on providing the analytics our clean energy customers need to lead. Often, it pays off: in July, for example, REsurety predicted ERCOT prices in the mid-$30s per MWh – and we were only off by one dollar. That doesn’t seem all that impressive unless you compare to market forwards: this spring, when we made this prediction, the market forecast predicted prices around $60/MWh – double actuals. Without REsurety data, buyers may have assumed twice what they received in revenue.
So, how were our forecasts that accurate? It boils down to our unique approach and our commitment to a weather sensitive, fundamentals-based model.
How We Optimize Our Models: Customer-First Thinking
Let’s take a step back to talk about our customers for a second. Our customers are looking to buy, sell, or trade clean energy. Take this example: a buyer is looking to enter into a PPA from a solar farm in their region, with a specific budget and generation capacity in mind for the year. When that customer uses REsurety price forecasts, they are not only looking for good data to value the solar power for sale right now against other solar assets, but also looking to compare against a baseline price provided by their grid operator – i.e., if they just pulled power from the grid, at whatever energy mix the grid operates at, and didn’t make the clean power investment at all.
Whether the purchase of solar power – from that specific solar asset – will “pencil out” is based on what its electricity is expected to sell for in the future. This price determines their clean energy buyer business case, and determines key financial metrics like payback period and cashflow volatility. For some corporations, if the renewable energy isn’t at price parity with traditional sources offered by the local grid operator, they won’t be able to buy due to budget constraints. Other companies have more flexibility, but often, our clients are working hard to justify their clean energy purchasing year to year, just like every other department at a company.
So with that in mind, our primary goal is to provide an accurate distribution of clean energy prices with actionable insights for our clients, recognizing their goals. We are not incentivized to inflate our prices for one side or another; we’re incentivized to help our customers keep making, using, and buying clean energy.
Our Fundamentals-Based Approach: The REsurety Difference
Forecasting electricity prices is tricky. In a grid like ERCOT, with significant intermittent energy generation sources from solar and wind, the task is even trickier. Weather, a significant and inherently difficult variable to account for, is a main driver of price forecasts.
Many traditional methods, like time-series and statistical models, often fall short. They struggle to meaningfully incorporate extreme weather events which are expected but only happen very rarely. This weather variability issue is exacerbated as the volatility and fundamentals of the market change as it grows. For a dynamic market like ERCOT, which is prone to short-term price spikes and has cases of extreme weather events such as 2021’s Winter Storm Uri, statistical models are particularly ill-suited.
Instead of relying on finding a causal signal in historical data, we employ a sophisticated fundamentals model. This model simulates how a market operator would dispatch supply to meet demand, providing a realistic view of market dynamics. While our point estimates don’t aim to perfectly predict every short-term spike, we acknowledge their possibility through our Weather-Smart distribution of prices. This distribution allows us to illustrate a range of potential outcomes based on various weather scenarios.
Our Advantage: Calibration and Pragmatism
Our forecasts are built on a foundation of robust data and expert analysis including publicly available market data, such as ERCOT’s own forecasts for load and generation evolution, our own market research & expertise, and comprehensive weather data.
This means that we don’t blindly accept forecasted load and capacity growth. Our model is continuously refined as the grid evolves, with improvements focused on areas like transmission constraint accuracy, granularity of generation additions, and impacts of policy changes.
Our advantage lies in our pragmatic approach to calibration:
- Supply Must Meet Demand: We operate under the crucial assumption that grid capacity will be able to meet demand the vast majority of the time. This means that our expert team interprets various data points – such as the 300 GW of data centers in load interconnection queues across the country – and, rather than accepting them at face value, judiciously calibrates them to ensure that the result is a functioning power system.
- Supply will be Built Out in a Realistic Way: We don’t assume that the supply chain can work to instantly build generation to meet unrealistic demand forecasts. Instead, we consider the practical realities of construction and project approval rates. If an ISO, like ERCOT, forecasts an unprecedented demand increase, and historical data shows such rapid build-outs are improbable, we factor that into our projections. We analyze interconnection queues, understanding that only a fraction of proposed projects actually come online.
- Generation Source Behavior Matters: We meticulously examine the commitment and dispatch of specific generation types under different weather scenarios. For example, if wind saturation increases, we assess whether this realistically leads to a reduction in gas peaker commitment and, consequently, lower prices.
How REsurety’s Accuracy Drives Your Strategy
We incorporate weather variability to capture the short- and medium-term volatility in times of stress for the grid. This pragmatic, data-driven approach empowers you to:
- Make Informed Decisions: Improve your understanding of when to buy and when to sell energy.
- Improve Budget Forecasting: Achieve more accurate and reliable energy cost projections, empowering your financial planning.
- Navigate the Clean Energy Market: Our forecasts help you understand the evolving dynamics of clean energy trading, even if you’re not an “energy nerd.”
At REsurety, we provide our customers with forecasts based on complex energy simulations and power markets expertise, giving them the insights they need to make informed investment decisions and navigate the energy market with confidence.
Want to learn more? Check out more resources from REsurety here.
