We often look at technology as either a solution to problems that we already have, or as offering us some extra enjoyment from direct interaction. When it comes to Energy Storage however we need to look at it as something else entirely. Energy storage devices have the most potential if they are used as a pre-emptive measure. Not only will their presence on the national grid allow for more efficient use of energy we currently have, but it will encourage the growth of renewable energy generation and forestall problems of the unpredictability inflicted on the grid by this growth. This is not technology solving problems we don’t want, this is technology pre-empting problems from other technologies which we do want.
Currently electricity is supplied by three categories of generators: baseload, intermediate, and peaking generation. Baseload generation is stable, static, and cheap, being comprised of either coal or nuclear. It is difficult, time consuming and costly to switch these facilities on/off or to ramp them up/down, thus their output remains more or less flat. In order to deal with changes in demand throughout the day therefore, more flexible generation is required.
Formed partially from oil and hydroelectric, but consisting predominantly of gas, the intermediate and peeking generation deals with the general changes in load throughout the grid. Gas and hydro also form ancillary services like regulation of short term fluctuations in load as well as emergency operations like black starts from grid failure. All of this adds cost and requires more carbon emissions.
Renewable variable generation (RVG), on the other hand, does not act like any of these. It is usually used to replace some intermediate or peaking generation, but unfortunately renewable generation has the bad habit of supplying electricity when we don’t really want it. Consequentially, if RVG increases like we want it to then it will be increasingly providing surplus energy. This electricity will be wasted and the wind or solar farms will be temporarily disconnected from the grid in order to maintain its integrity.
There are two real ways of dealing with this. The first is to reduce the baseload generation by shutting down selected coal plants and relying on RVG to pick up the slack. Of course, given their unreliability this will require more gas plants to take over when the wind stops blowing and the sun goes down. This in turn will increase the cost to the consumer as gas is more expensive than coal. On the other hand it will reduce carbon emissions as gas is also cleaner than coal; we are not going to reduce carbon emissions without having to pay for it.
There is a better way however. The addition of electricity storage devices to the grid would allow for a more secure supply and a much more efficient use of RVG and consequentially encourage its growth. It would also either allow for the complete replacement of baseload generation, with the consequential reduction in carbon emissions, or take the part of intermediate and peaking generation with the consequential reduction of cost. But the most financially advantageous use of storage is in the replacement of ancillary services.
Storage devices do have considerable problems however, a combination of the lack of regulation, safety concerns, and highly variable cost (especially in comparison to the fluctuations in gas prices) means that industry uptake has been stunted at best. Perhaps the most intriguing cause of slow growth in storage use is the wide variety of choice, both from increased efficiency of old technologies like batteries, and from new technologies like thermochemical storage. Such a wide selection should be a good thing except for the fact that participants (and governments) don’t know which choice is best at the moment, let alone which will be best in ten years.
When choosing a storage device two main things need to be considered: the geology of the area, and the power output required for the service the storage device will be supplying.
It turns out that not every storage device is good at everything. If you wish to smooth out the short term changes in load then you need a storage device that is capable of responding on the scale of seconds or less. On the other hand if you wish to participate in wholescale arbitrage (storing electricity at times of surplus and selling it at times of deficit) then you need a device of relatively low power and response time, but one that can store much more energy for a longer time. As I am arguing here for the use of energy storage as a stimulus and insurance device for RVG, then I will focus here mainly on the high energy arbitrage solutions.
The current state of energy storage is such that 95% of global capacity is made up of Pumped Storage Hydroelectric (PSH). This essentially involves using excess/cheap electricity to pump water from a low reservoir into a higher reservoir, and then at times of high electricity prices allowing the water to flow back down and selling that electricity back into the grid. It is very efficient, low power, and high capacity. It also requires a very specific geographic arrangement, not only does it require the ability to dam a river, but it requires the ability to do so again within a short distance in such a way that the second dam will also create a paired reservoir with the first. There have been solutions where an underground reservoir is created, but this also requires a rather specific geology and is much more expensive.
Additionally there are problems with adding more PSH even where the geography would allow them. Increased regulation regarding building dams and the flooding of large areas means that building PSH is harder than ever.
Compressed Air Energy Storage (CAES) is only just less specific in its requirements. In order to run in a cost effective manner it requires an underground cavern into which air is pumped until it reaches high pressure. When electricity is desired, the air is allowed to escape through an expansion turbine. In new generations of CAES natural gas is combusted to create heat during the expansion process allowing for a much more efficient return on energy.
If you don’t have an underground cavern or a conveniently arranged mountain range to hand however, there are a few things you can look into. In terms of the scale of single units there is nothing to match the Giga tonne level of CAES or PSH, however the combination or dispersion of smaller yet less geographically dependant solutions can achieve the same end.
Batteries have been referred to in the past as rather dirty ways to store energy. And in the days of lead batteries this was accurate. However, with the development of modern batteries like Lithium Ion batteries there is very little environmental damage and recycling or safe disposal is relatively easy. Lithium-ion batteries were, until very recently, prohibitively expensive, however that has been changing and they are set to become the dominant player in the battery market.
My personal favourite emerging technology for medium scale electricity storage however is called Pumped Heat Energy Storage (PHES, see Isentropic thought there are methods and proposals differing in the specifics). This works by pumping argon gas from one large silo filled with uniform small rocks into a similarly filled silo. The pumps are run at times of cheap electricity and rapidly compress the gas from one silo, consequentially heating it to 500oC and releasing it into the other silo. As the hot gas moves through the rocks it forces gas at ambient temperature out the other end. This gas is then rapidly expanded by similar pumps cooling it to -160oC. This process sets up both a pressure and temperature differential between the silos which can be maintained without any extra energy input for long time scales. When electricity is desired, the process is allowed to run in reverse (with hot gas at high pressure displacing the cold gas at low pressure), this forces the pumps to move which runs a turbine.
Though there are other technologies that are highly flexible in their ability to be deployed, such as thermochemical energy storage, these are the two that seem in the best current position to mimic the function of PSH anywhere in the world. Though they are smaller in capacity per unit with PHES going up to 5MW, they can be deployed in groups or even placed strategically throughout the grid in order to increase efficiency and reaction speed.
However it is done, it is an unavoidable fact that if we want to increase our capacity for renewable generation we are going to need a similar increase in our capacity to store the variable energy produced. Unlike how we have become used to treating technological progression however, as an iterative process in solving a continual and connected series of problems; electricity storage needs to be encouraged and implemented as a driver of change, not a response to it.