Is it worth extending the duration of merchant batteries?

The short answer is no as long-duration merchant batteries are not yet cost-effective, and the associated margins (i.e., time spreads) are not large enough. For example, if we use the ERCOT West hub, the average values captured in day-ahead and real-time energy price arbitrage under normal conditions — since 2018 — are $15.80/kWh-year, $10.30/kWh-year and $6.05/kWh-year for batteries of 2-h, 4-h, and 8-h duration respectively. The capital cost of storing energy with chemical batteries today is at least $200/kWh, making it too expensive. Merchant batteries are most valuable today for providing flexible capacity than capturing time spreads.


ERCOT (i.e., Texas) has about 25% of its installed capacity coming from intermittent renewable generation, which produces about 20% of the energy. It has minimal access to hydropower and therefore has to rely solely on thermal generating assets and demand to mitigate energy intermittency. For this reason, ERCOT has an effective capacity day-ahead market for regulation and contingency reserves to guarantee sufficient capacity to meet the residual load (i.e., load less renewable energy). The West hub has the highest share of wind and solar generation in ERCOT, combined with the lowest load. These features make this hub an ideal location for short-term merchant storage.

What is the intrinsic value of merchant batteries in ERCOT?

The intrinsic value of a merchant battery in ERCOT consists of optimizing the battery in relation to the day-ahead (energy and capacity) and real-time markets. Every day we have to decide for the next day how much power to buy and sell, and what ancillary services to sell and how much. Moreover, every five or fifteen minutes, re-optimizing the battery based on real-time energy prices and expected future prices.

We divide the associated revenues into three streams, namely time spreads, capacity, and price spikes. Time spreads are the margins made by buying energy at a lower price than selling in the day-ahead and real-time markets. The capacity is associated with the sale of ancillary services (regulation and contingency reserves) on the day-ahead market. And, price spikes are the margins when the supply and demand balance of ERCOT is tight, resulting in excessive prices (energy and capacity) for a few hours. We make this distinction to avoid contaminating the value of energy with that related to capacity. When the system is tight, any generating asset is very profitable. It is also the most vulnerable value when we build new generating assets.

We have developed a mathematical program to assess the intrinsic value of electrical energy storage assets and thermal power plants based on historical day-ahead and real-time prices. This value is quite close to what our dispatching and scheduling algorithms achieve in practice — we use a combination of approximate dynamic programming and deep learning to implement our algorithms.

The following graphic shows the results of our assessment effort based on historical prices in ERCOT West hub between January 2018 and August 2020, where the break-even margin is the minimum resource price required by equity investors in the absence of other revenues.

We can easily observe that the most significant sources of intrinsic value for merchant batteries are the capacity market and price spikes. Indeed, batteries do not need to burn fuel to provide regulation-down/up or responsive reserves, making them quite competitive for offering these services. When it comes to batteries, the cost of fuel is replaced by an opportunity cost. Without price spikes, the 4-h battery would not be profitable, and the 8-h battery is just too prohibitive now.

We provide more details on how merchant batteries get optimally dispatched into ERCOT in the following dashboard, which we update weekly.

Our levelized cost analysis

Those are our main assumptions behind the break-even margins (or break-even resource prices).




CEO, Pyxidr

Love podcasts or audiobooks? Learn on the go with our new app.

Recommended from Medium

A deep dive into Starbucks Dataset

Determining Significant Features in a House Sale

Usage of Big Data Analytics in Organizations

Neo4j data processing with springboot

Another Wave, the Same Mistakes

3 Lines of Python Code to Write A Web Server

Broadcasting in Numpy: A Powerful Technique You Should Know

A New Take on Analytical Workflow Management

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store
Robin Duquette

Robin Duquette

CEO, Pyxidr

More from Medium

Bridging the Dive Industry and Ocean Conservation Sectors

“CommBank Is Funding Coal Projects…” Advertisements

CANCER THE DEFEATED FOE Cancer, which is known as the second deadliest disease in developed…

The To-Go Cocktail Debate