What is an Energy Market Forecaster for ESS?

If a battery energy storage system is going to make money in a wholesale energy market, it needs to know something the market hasn’t told it yet – where prices are headed. That’s the job of an energy market forecaster.

How Energy Markets Work

Before getting into forecasting, it helps to understand what a battery system is actually bidding into.

Wholesale energy markets run on offers. Hundreds of generators submit price-and-quantity offers for each time interval. The market ranks those offers from cheapest to most expensive and selects enough to meet total demand – paying everyone at the same lowest price they all would accept.

Here’s a simple example: if four sites bid into a market that needs 100 megawatts, the market takes the cheapest offers first. If the last offer needed to hit 100 MW is priced at $60/MWh, everyone who bid at or below that price gets paid $60 – even the site that bid $1. Sites that bid above $60 get nothing.

This is called price clearing, and it shapes how smart battery operators think about bidding strategy.

What Does This Mean for Battery Systems?

Batteries typically bid low – below where they expect the price to clear. The logic is that if your offer is cheap enough, it’s almost certain to be awarded, and you get paid at whatever the market clears, which is often much higher than your bid.

That strategy depends on one thing: knowing when prices are likely to spike.

A battery that charges when prices are cheap and discharges when prices are high can generate significant revenue. Miss the timing, and you’ve either sold energy at a loss or sat idle during a spike. This is where market forecasting becomes critical.

What Is an Energy Market Forecaster?

An energy market forecaster predicts future electricity prices over a bidding period. It takes available data and produces a price curve the bidding system can optimize against.

The goal isn’t perfect price prediction. What matters most is getting the timing of peaks right. Whether a price spike reaches $500/MWh or $2,000/MWh is less important than knowing it’s coming at 9 PM and not 3 AM. The battery just needs to be charged and ready when it arrives.

Inputs a forecaster typically draws on include:

• Historical market data – energy usage follows weekly cycles, so last week’s prices at the same time are meaningful signal
• Apparent temperature forecasts – heating and cooling demand drives a significant share of grid load
• Solar irradiance and cloud cover – solar generation reduces prices; less sun means less competition and higher prices
• Generator outage data – fewer generators online means the market clears higher

Building a reliable forecaster is genuinely hard. There are whole commercial companies that do nothing else but build these systems.

Building A Simple Forecaster 

A basic but functional approach is to forecast the price for any given hour by averaging the actual prices from the same hour across the previous three weeks. Energy demand follows weekly patterns closely enough that this approach captures daily peaks and valleys with reasonable accuracy.

In Peter Gruenbaum’s course, “Software for In Front of the Meter Battery Systems,” students build exactly this kind of forecaster using real data from California ISO then compare forecasted prices against actual historical prices to evaluate performance.

It won’t catch extreme price spikes like the ones that occasionally hit the Texas market, but it’s a legitimate backup forecaster for when more sophisticated models fail, and it’s a solid foundation for understanding what a production forecaster actually needs to do.

Ready to Build It Yourself?

Peter’s course walks you through building a simple forecaster, a battery simulator, a bid optimizer, and a market simulator. This is the course to take if you are looking for a hands-on way to learn how to create software for front of the meter battery systems. 

Enroll in Software for In Front of the Meter Battery Systems today!