A recent twitter conversation started by the incisive Suzy Waldman highlighted the hope many energy-conscious folks have that some form of economical large scale storage technology will soon be deployed to start enabling variable renewable energy to effectively meet demand – much like dispatchable capacity already does.
If this were possible, with consideration of scale of materials and energy required to construct all this storage, then it would be great to see, especially if the life cycle analysis indicated nice low emissions per kilowatt hour (kWh) of supplied electricity.
However there are a few undeniable considerations against it being adopted at a significant scale, at least in the form envisaged by optimistic variable renewable proponents.
The limitations on net energy return on investment have been patiently fleshed out. The actual prohibitive scales of land area or physical amounts of refined metals and chemicals have also been estimated. Another factor which I suspect hasn’t occurred to many, if not most, proponents is a simple matter of economics.
As explained by FERC:
Pumped storage projects move water between two reservoirs located at different elevations (i.e., an upper and lower reservoir) to store energy and generate electricity. Generally, when electricity demand is low (e.g., at night), excess electric generation capacity is used to pump water from the lower reservoir to the upper reservoir. When electricity demand is high, the stored water is released from the upper reservoir to the lower reservoir through a turbine to generate electricity.
Electricity can be thought of as an homogeneous good, and storing excess when demand is low – and thus its cost is low – so that it can be sold later when demand – and the price – is high is the economical way to operate a pumped hydro storage facility. Further, these daily periods of low and high demand are effectively set in stone, which is exactly how an operator of an industrial facility would prefer them.
The FERC précis goes on to say:
Pumped storage projects are also capable of providing a range of ancillary services to support the integration of renewable resources and the reliable and efficient functioning of the electric grid.
Which seems at odds with the previous logic, which of course it is, while also being technically true. An energy storage installation could buy oversupplied summer sunlight in the hours around lunchtime and sell it back around dinner time. Or it could buy oversupplied wind generation… whenever that happens to be, with a view to selling it at the next peak demand time. It is “also capable of providing a range of ancillary services to support the integration of renewable resources” but doing so rapidly erodes the economics of operating the storage.
So when economical storage is proposed as the large-scale answer to the shortcomings of variable renewable energy, what autocratic mechanism do proponents support for ensuring that dispatchable sources such as coal, CCGT (combined cycle gas turbine) and nuclear power are excluded from supplying it more economically?
Remember, too, that it’s the excess summer sunshine which is destined for storage. In this envisioned storage-balanced renewables grid, substantial overbuild is unavoidable.
Rokkasho-Futamata 51 MW Wind Farm, which is backed up with storage that can supply a maximum of 34 MW for 7 hours.
To make it work, wind farms such as Rokkasho-Futamata feature co-sited battery installations for load levelling and reserve capacity. As the case study states:
The Japan Wind Development Company’s 51 MW Rokkasho-Futamata Wind Farm project was integrally developed with energy storage capacity of 34 MW, using NGK Insulators’ sodium-sulfur (NaS) batteries for load leveling, enabling the storage of low cost off-peak wind power for sale/ distribution during peak demand times.
A similar capability is seen for some pre-commercial concentrating solar thermal projects, where molten salt stores solar energy as heat. The cost of doing this is already apparent in such proposals as Alinta‘s pilot project.
It would be interesting indeed to see the re-evaluated economics of variable renewable energy projects if such load levelling integrated storage became standard.
The Actinide Age 
Yes… as per several tweeters, hard/impossible to see why cost effective storage would not work as well or better in partnership with the production of electricity from a nuclear plant. It would re-jig the economics of a planned system… fewer NPPs running mostly full-time, storing and releasing energy in response to demand.
First ground to be broken will, I think, be small home storage to partner with PV and respond to peak. I think that would be an excellent thing, albeit EROI considerations need to be included. EROI is why I really think smaller levels of home storage with PV while remaining grid-connected for the rest is likely the most efficient solution overall in terms of energy, materials and resources.
I can’t find the tweet now but someone pointed out that incumbent fossil generators would probably be able to supply the cheap power at low demand times for storage most economically, i.e. if grid storage was deployed rapidly from this point, it would probably perpetuate fossil fuel use rather than enable variable renewables – unless market-distorting restrictions were enforced (which is, of course, as likely as seeing grid-scale batteries in any such time-frame).
Excellent post. Thank you! Since this issue pops up again and again I also had thoughts along similar lines some time ago. See here for my “random thoughts”. I would still be interested in a more thorough comparison. A comparison that actually optimizes for the amount of storage and the way it is used as a function of storage cost and then compares variable RES with reliable baseload plants. I am not sure I would know how to do that. Seems way harder than “random thoughts” for a blog post 🙂
I’m worried that we’ll run out of Next Big Things. Around the turn of the century it was carbon capture and a few years later it was dry geothermal. Now it’s energy storage. Somewhere in a BNC article it said pumped hydro was about 5 Gwh a day in Australia yet we use 249,000/365 = 682 Gwh a day. Just one Gwh would require 2 million lead acid batteries of half a kwh capacity. Untroubled by such numbers we’re told energy storage will bring about a ‘utility death spiral’. Not any time soon.