How Do You Value Grid Connected Energy Storage?

This is almost a part II to my earlier post, but I was reminded again of the problems we face when it comes to defining, and then valuing, Grid Connected Energy Storage (GCES).

The recent New York Times article, "Companies Race to Develop Utility-Scale Power Storage" pointed up the problems and potential for confusion. "Power storage" technologies listed included the Beacon flywheel, the NGK molton sodium-sulfur battery, the A123 lithim ion battery, and compressed air (again!). Quoting a report by GTM Research, this article made a very insightful distinction between "power oriented" technologies, used mainly to regulate short-term changes to grid frequency, and "energy oriented" storage -- in which energy use is shifted to other times of the day. However, the author could have done a better job applying this distinction and pointing out the difference in cost.

For example, the article discussed the $69 million Beacon project in New York, where they will install, "...hundreds of "flywheels" to store 20 megawatts of electricity, enough to power 200 homes for a day." In reality, the flywheel is designed to store only 15 minutes of power and falls into the "power oriented" category above. Its total energy storage will only be 5 MW hrs, about enough to power 40 homes for a day, although it will never be used for that purpose.

Also, the article reported on the $25 million requested by Southern California Edison for an A123 "32-megawatt-hour battery" - but is it really 32 MW hrs? I pose the question because the system will be designed as an 8 MW battery with 4 hours of storage (32 MW hr), but the application is at a wind farm, where multiple cycling is needed to firm wind - a "power oriented" application. Lithium ion batteries are good for about 500 - 600 complete charge and discharge cycles. If it is used in an "energy oriented" application, shifting wind power at night to the day, then it will only last about 2 years. However, in a "power" application, where the battery is barely discharged, it will last for many thousands of cycles. In fact, this is how it is currently applied. In this case it would be operated like an 8 MW flywheel, with usable energy storage of only about 2 MW hrs.

So how do you value these installations? If we value the flywheel and the li-ion systems by the MW hr, then their cost is $13.8 million and $12.5 million respectively. However, if all we care about is their power capacity, then the cost is $3.45 million and $3.125 per MW. (The NGK battery is the only energy oriented technology mentioned in the article, but no cost information was provided.)

By contrast, a VRB-ESS (vanadium redox flow battery - energy storage system) will provide both energy and power, with nearly unlimited cycles, full or partial, for about the same cost per MW of the flywheel or li-ion battery. However, the VRB-ESS will also include 4 - 8 hours of storage, dropping the cost per MW hr to a fraction of the cost for a "power oriented" system.

For example, a 5 MW system with 6 hours of storage would cost about $18 million, with all costs included - a complete turn-key system. That would provide 30 MW hrs of energy at a cost of about $600 thousand per MW hr. The cost of power is only $3.6 million per MW.

Although not directly relevant to the discussion, it's good to know that the VRB-ESS will last 10 years before needing refurbishment. This consists of replacing the PEM (proton exchange membrane) at a cost of about $3 million. The system is then good for another 10 years!

Bottom-line? - It's important to understand the application, whether energy, power or both, and then determine the cost per energy (MW) and/or the cost for power (MW hr), when evaluating the technology.