April 8, 2016 at euanmearns.com
Many countries have committed to filling large percentages of their future electricity demand with intermittent renewable energy, and to do so they will need long-term energy storage in the terawatt-hours range. But the modules they are now installing store only megawatt-hours of energy. Why are they doing this? This post concludes that they are either conveniently ignoring the long-term energy storage problem or are unaware of its magnitude and the near-impossibility of solving it.
The graphic below compares some recent Energy Matters estimates of the storage capacity needed to convert intermittent wind and solar generation into usable dispatchable generation over different lengths of time in different places. The details of the scenarios aren’t important; the key point is the enormous differences between the red bars, which show estimated future storage requirements, and the blue bars, which show existing global storage capacity (data from Wikipedia). It’s probably not an exaggeration to say that the amount of energy storage capacity needed to support a 100% renewable world exceeds installed energy storage capacity by a factor of many thousands. Another way of looking at it is that installed world battery + CAES + flywheel + thermal + other storage capacity amounts to only about 12 GWh, enough to fill global electricity demand for all of fifteen seconds. Total global storage capacity with pumped hydro added works out to about 500 only GWh, enough to fill global electricity demand for all of ten minutes.
Yet microscopic additions to installed capacity are apparently considered a cause for rejoicing. Greentechmedia recently waxed lyrical about the progress made by energy storage projects in 2015 . “Last year will likely be remembered as the year that energy storage got serious …. projects of all sizes were installed in record numbers ….” But when it goes on to list “the Biggest Energy Storage Projects Built Around the World in the Last Year” we find they’re all 98-pound weaklings:
Also notice that while megawatts are specified MWh usually aren’t. There are two possible explanations for this. First the facilities aren’t designed to store energy. They are primarily for frequency control, load following etc. The MW are important but the h aren’t, or at least not very. Second, the policymakers who mandate these facilities don’t see any difference between a MW and a MWh.
And I say “mandate” because that is what the state of California recently did. California recognized that it would have to solve some grid stability problems before it could expect to meet its 50% renewable energy by 2030 target, so in 2013 it passed a “Huge Grid Energy Storage Mandate” that required the state’s big three investor-owned utilities to add 1.3 gigawatts of energy storage to their grids by 2020. Three points are worthy of note here:
- Relative to California’s 50GW peak load 1.3GW can hardly be described as “huge”.
- The mandate again doesn’t say how long the storage should last, i.e. how many gigawatt-hours are needed.
- The proposal specifically excludes pumped hydro storage projects of 50 megawatts or more.
And the rationale for excluding pumped storage projects over 50 MW deserves a paragraph all to itself:
The California Public Utility Commission concluded that although large-scale pumped storage hydro meets the statute’s definition of an energy storage system, it must limit the size of eligible pumped storage systems in order to encourage the development and deployment of a broad range of energy storage technologies. In the CPUC’s view, the goal of creating a new market for a range of storage technologies would be undermined if the IOUs could meet their targets by acquiring a pumped storage facility: The majority of pumped storage projects are 500 MW and over, which means a single project could be used to reach each target within a utility territory.
What is this broad range of storage technologies that pumped hydro threatens to undermine? Based on proposals received to date they include bi-directional EV charging stations, molten sulfur batteries, zinc hybrid cathode batteries, lithium-ion batteries, thermal energy stored in ice, in used EV batteries and in rechargeable electrolytes. In short, California will consider any type of energy storage system provided it isn’t pumped hydro, the only large-scale energy storage technology that can be guaranteed to work.
Which brings up the question of which of the technologies don’t work. In the recent ARES post Greg Kaan made the following comment:
This thread is turning into complete nonsense, not due to the commentators here (thanks Greg) but simply through the “solutions” being presented to try and cope with intermittent power production.
And Greg is quite correct. The solutions being presented to cope with intermittent power production range from green dreaming to downright bonkers. Here’s a selection, courtesy of Wikipedia:
Compressed air
Liquid air
Batteries
Electric vehicles
Flywheels
Underground hydrogen storage
Power to gas
Hydro and pumped hydro
Superconducting magnets
Thermal storage.
To which I will add:
ARES rail storage, which we recently looked at.
The 500m-diameter underground granite cylinder that moves up and down without ever cracking, leaking or getting stuck
Flat Land Energy Storage, which was reviewed here.
Anyone who can see a way of commercializing any of the unproven technologies on the list is encouraged to provide details. (Although two of them are in fact capable of providing meaningful amounts of storage. The first is power-to-gas, which was dismissed here as being far too complicated, inefficient and uneconomic. The second is very-large-scale pumped hydro, which was discussed here. The project delivered 6.8TW of storage but involved turning a large chunk of the Scottish Highlands into an inland sea.
So here we have an impossible situation, with green pipe-dreamers and utilities whom one suspects should know better trying to solve an unsolvable problem with technologies that have no chance of solving it. So what happens next? Well, at some point something obviously has to give, but what, where and when is the question.