What were the two most unnoticed energy storage trends in 2014?
Throughout 2014 there were several articles on significant energy storage events or announcements, as well as future predictions for the coming years. However, there were two transformational “side effects” that were skipped over and barely noticed. The first is the return to multi-hour (high-energy) storage applications, and the second is emergence of the “distributed bulk” storage to challenge traditional larger central storage systems.
The return to multi-hour storage applications and devices
Japanese utilities started using multi-hour Sodium Sulfur (NaS) and flow batteries for peak shaving almost two decades ago. Outside Japan, in the US, American Electric Power (AEP) started using NaS batteries in early 2000’s for its grid improvements, which was soon mirrored by some other utilities. In the absence of adequate utility interest, a few companies in the US—including Beacon Power and AES Battery Storage—successfully pioneered the movement to use storage for short-duration applications (i.e., frequency regulation) which shifted the market focus to storage devices with discharges much shorter than an hour. Part of the success was due to the fact that such devices are less expensive than their multi-hour counterparts on the $/kW basis. Their success, at least in the US, took the attention off the former high-energy (multi-hour) applications and focused it on high power applications.
There is no doubt that Southern California Edison’s (SCE) procurement of 260 MW of storage for its local capacity requirement for the West Los Angeles Basin region is significant in stimulating other utilities and showcasing a variety of different storage technologies. But the real significance of this 2014 event is that it is swinging the pendulum back to the multi-hour storage applications and devices for capacity and other high-energy applications that have a much larger market than the short-duration applications like frequency regulation. Multi-hour storage devices, as dominant in the SCE procurement, can serve multitude of grid and customer needs such as capacity, investment deferral, peak shaving, renewables integration, backup power, demand reduction, energy cost management, frequency regulation, and many more. SCE is taking the storage community back to the much larger market.
The emergence of the distributed bulk storage
Central versus distributed storage has been an ongoing debate and each has pros and cons. There are instances where an advantage of one over the other is obvious, such as central bulk storage at geological sites where caverns or water reservoirs with different elevations are almost readily available. But the real question is: which one would the market or investors prefer when no geological advantages are readily available? Two key advantages have been keeping our hopes high for central storage: lower cost per kWh and availability in large GWh scale. Both of these key advantages were challenged, if not shattered, in the last year.
Tesla Motor’s $5 billion Gigafactory is being built to deliver li-ion batteries, 15 GWh/year of which is dedicated for stationary applications. There is no doubt that this and similar investments will accelerate the already declining battery cost to be competitive with bulk central storage—but there is another significance in this announcement. Apparently, 96% of the pumped hydro plants in the world have a generating capacity less than 15 GWh. Finding accurate information on pumped hydro facilities around the world is a dire task and the DOE’s Global Energy Storage Database only lists 342—and much of the information is still missing due to the fact that it is a voluntary reporting system. Going with what we have available, we notice that only a handful (about 4%) of all pumped hydro plants in the world have a generating capacity over 15 GWh. Figure 1 compares GigaFactory’s planned annual production for stationary applications with the generation capacity of the world’s largest pumped hydro plants.
The real significance is not in the size but the fact that it takes almost a decade to build a pumped hydro (or other large central storage plants), which has a storage capacity that is less than what GigaFactory can deliver within a single year with no need for permits or special geological formations. GigaFactory alone can produce the equivalent of a few average pumped hydro power plants each single year, and that is without considering its much larger production for the electric vehicles.
Tesla’s GigaFactory, along with other companies that are already in place to aggregate distributed storage into effectively larger storage plants, is all it takes to realize what I call “distributed bulk” storage, which effectively functions as a pumped hydro or CAES would work for massive storage and time-shift of energy on the grid. This should clear distributed storage from being disregarded as too small to fit national-scale storage needs. When aggregated properly to benefit the energy grid performance, especially if the storage units are on the utility side of the meters or controlled by the grid operators, distributed storage will effectively serve as much needed bulk storage.
2014 had many significant news announcements on storage, but I believe there were only two noteworthy trend-setting events to mention: moving to distributed storage in bulk scale and addressing multi-hour applications. Both trends have, and will continue to, opened up the storage market of utilities and behind the meter applications.