Management more important than location. California’s energy‑storage systems are divided by a financial gap. What does this mean for Poland?

Published: Estimated reading time: 3 minutes
Source: Shutterstock

Building a large‑scale energy‑storage system alone does not guarantee financial success. A report by Gridmatic reveals enormous differences in the profitability of batteries operating on California’s CAISO market. It turns out that the key to profits is not geographical location, but a flexible trading strategy based on algorithms. These conclusions are fundamental for the developing storage market, including in Poland.

A financial gap in revenues

California is a powerhouse in energy storage – over the past six years, its battery capacity has grown to 17.4 GW. However, the market is highly opaque because the grid operator anonymizes public trading data. To break this barrier, Gridmatic published the CAISO Storage Report 2024–2025, in which it reconstructed and analyzed the real performance of 30 large storage systems with a combined capacity of 2.3 GW.

The market analysis produced surprising results:

  • The median market revenue for the analyzed fleet was 2.67 USD per kW‑month (a metric that allows comparing storage efficiency regardless of size).
  • The weakest units generated below 1 USD per kW‑month, while market leaders exceeded 6 USD per kW‑month.
  • As many as 19 out of 30 analyzed storage systems earned less than 4 USD per kW‑month.

These massive differences in earnings occurred even between facilities located in the same price zone, which theoretically had identical market conditions.

The trap of rigid bidding and equipment failures

Why do some storage systems earn a fortune while others generate losses? The Gridmatic report points to two main reasons:

  1. Overly passive energy trading – most operators use rigid, static bid curves submitted to the market only once per day. As many as 12 out of 30 installations used completely unchanging bid templates 100% of the time. As a result, these batteries could not react to sudden, minute‑scale price spikes on the balancing market.
  2. Low technical availability – the average availability of the analyzed batteries was only 79.7%. Downtime was caused by equipment failures and fires. For example, the Gateway storage system was offline for 24% of the time, and units at Moss Landing have been idle since a fire in January 2025. Beyond losing direct revenue, such long outages expose owners to high financial penalties from the grid operator.

At the opposite end of the spectrum was the Caballero storage system (100 MW / 400 MWh), owned by Alpha Omega Power. Thanks to automation and machine‑learning algorithms that continuously adjusted pricing strategies to market fluctuations, the facility generated 2.45 times higher revenue than the market median. If the remaining 29 storage systems had operated as efficiently as Caballero, the entire fleet would have earned an additional 98 million USD over two years.

The capacity market is not enough

The conclusions from California should be required reading for investors in Poland. Our large‑scale energy‑storage systems are only beginning to emerge, and their business plans rely heavily on stable revenue from the capacity market. However, the long‑term profitability of Polish batteries will depend on energy trading on the wholesale market and providing system services for PSE.

California’s experience shows that Polish investors cannot treat energy‑storage systems like maintenance‑free solar farms. To profit from price arbitrage (charging batteries during low‑price periods and discharging during peaks), traditional trading models will have to be abandoned. Success in the Polish market – characterized by increasing price volatility due to growing wind and solar generation – will require advanced AI algorithms and automated real‑time trading. Otherwise, Polish storage systems may end up at the bottom of California’s performance statistics.

Change consents