A dive into grid-scale Energy Storage options
It's been a while since my last post, as I've been rather busy working on other projects and research into CO2 emissions (which has yet to see the light of day). Today, I wanted to dive into energy storage, particularly grid-scale energy storage - something that will be of increasing importance as our energy grids move to renewable sources.
Why does it matter?
Traditionally, energy grids have consisted of generating capacity, consumers, and connections between them. Grids generally have base-load power - constantly on power sources - and peak-load - capacity that can be spun up rapidly to meet peaks in demand. Balancing a grid then consisted of looking at the long-term demand, and making sure a reasonable balance of peak- and base-load power capacity was in place to meet demand, ideally without wasting large amounts of capacity but with headroom for exceptional demand and power failures.
However, in recent years, with the rise of renewables, power generation has changed. Whereas it previously consisted of fuel-based systems which can be turned on at our whims (albeit some with a significant ramp-up time), we now have a number of competitive renewable options which are dependent on external factors - the sun and the wind. This means we have power sources that generate energy at times decoupled from demand, which means we need a means of storing energy. For example, solar panels produce no electricity in the dark, but if we can store energy through the day, we can then release that energy through the nocturnal down-time and produce consistent power output - a drop-in replacement for existing power generation methods. Energy storage also gives us better ability to deal with unexpected surges in demand or supply, as well as power failures.
Many types of energy storage have been around for a long time, but large scale usage of storage hasn't really been required due to the convenient nature of coal, gas and nuclear power. With the growing threat of climate change, it's time to revisit energy storage and see what options are available, both old and new. I'm going to focus on technologies that will work at grid-scale, to limit the scope slightly. Some of the technologies I'll touch on are mature and in use, some are very much under development, and not all are cost-effective.
Chemical Energy Storage
The first chemical storage mechanism to consider is Power-to-gas. This involves taking electricity and using it to create gas, usually either hydrogen (through electrolysis of water) or methane (through electrolysis and the Sabatier reaction, consuming water and CO2). Hydrogen has the advantage that burning it produces water as a product (so eliminating pollution), whereas Methane will produce greenhouse gases when combusted. Neither gas offers great round-trip efficiency (~40%), but using cogeneration plants (that produce both heat and electricity), that can be raised to around 60%.
A similar suggestion is Power-to-ammonia. Ammonia can be produced via the Haber-Bosch process (compressing Nitrogen and Hydrogen under high pressure), and has about 1/3 the energy density of diesel. It's also produced in large quantities already for use in fertiliser, and can be burnt cleanly to produce energy. However, it is highly corrosive in its concentrated form, and requires a large amount of energy to produce.
Mechanical Energy Storage
Mechanical energy storage encompasses two broad areas of energy storage - moving things, and compressing things. In the latter camp lie technologies such as CAES - Compressed Air Energy Storage. This is what it sounds like - compressing air using electricity, and storing it in some fashion (large installations have used underground caverns for this). The compressed air is then either used to spin turbines directly, or more frequently combusted in a gas turbine, which lowers the fuel usage of the turbine considerably. There are two large installations, in Germany and in the USA at the 100+MW scale.
The other main compressive storage technique is the hydraulic accumulator, which pressurises water in a vessel, for either direct mechanical usage, or for converting back to electricity. Hydraulic accumulators were used a lot at the turn of the 20th century, for driving boat locks, lifting bridges and other mechanical mechanisms, but are used much less frequently in the modern era, and not for generating electricity. A modern variant of this is QuidNet, who are planning to pump pressurised water into old oil wells to store energy.
Onto moving things. These mechanisms split into gravitational methods - that lift things up using energy, then reclaim the energy by lowering the things again - and kinetic methods - that store energy by moving things about, and slow them down to regain the energy.
The largest and most mature gravitational energy storage mechanism is pumped storage, accounting for 99% of all bulk energy storage globally (as of 2012). This involves two water reservoirs at different heights. Off-peak energy is used to pump water up to the upper reservoir, and when power is needed, the pumps are reversed and become generators, producing energy. There are several GW-scale pumped storage facilities in the USA and China, and many more at MW-scale across the world, but pumped storage is very dependent upon having the right geography. This naturally limits where it can be implemented, so it cannot placed at will, but it can scale to very large sizes indeed.
Another gravitational technique is much more simple to grasp - simply storing energy as gravitational potential energy by lifting heavy weights. There are a couple of variants of this, including using mine-shafts to hold weights that are raised and lowered vertically, using cranes to stack heavy blocks, and using electric rail lines to lift heavy cars up an incline. Most of these techniques are relatively simple, but are still in development, but have potential to scale to large capacities.
The last technique is kinetic storage, i.e. flywheels. At their most basic, flywheels are really rather simple - spin a mass really fast to store energy, then slow it down to regain the energy. They've been used in everything from buses to racecars, but for energy storage purposes have been refined to a fine art in order to minimise energy losses. This includes using magnetic levitation and magnetic bearings, making the flywheel out of carbon fibre, and running the entire thing in a vacuum to remove drag from air.
Electrical Energy Storage
Electrical energy storage mostly consists of capacitors - electronic components capable of holding energy in an electrical field. They're used extensively in electronics, but also in grid applications for power factor correction. One variant is the supercapacitor - capacitors with much higher capacitance values (albeit lower voltage limits). These can store 10-100 times the energy per unit mass versus conventional capacitors, and can deliver and store charge much faster than batteries (though they are less energy dense). They're used as buffers for grid power (to reduce fluctuations from heavy loads such as electric vehicle charging), as well as to stabilise the output of variable energy sources such as wind power. They've also seen extensive usage in various kinds of vehicles, for applications like regenerative braking and short-term energy storage. They're still fairly expensive compared to batteries, so have seen little adoption as a grid-scale energy storage mechanism.
Another potential electrical energy storage mechanism is SMES - Superconducting Magnetic Energy Storage. This involves storing energy in the magnetic field produced by a current flowing through a coil of wire. Superconductors are conducting materials that, under certain conditions, have no current losses, allowing current to flow in a coil indefinitely. This is a fascinating technology, but is currently limited by the limitations of the superconducting materials currently known about. All the currently known superconductors require cryogenic cooling to become superconducting, meaning any SMES system would need a refrigeration system to cool it down potentially as low as 4K (-269°C). Superconductors also tend to be brittle ceramics, making creation of wires for large coils tricky. They also have a critical magnetic field - if the magnetic field they are in exceeds a certain level, they stop superconducting, losing their benefits. All of these various disadvantages have meant no grid-scale applications have been developed as yet, although they are used for power quality control in certain manufacturing applications.
Electrochemical Energy Storage
Electrochemical energy storage can take many forms, but the most common is batteries. There are countless types of battery technology, but there are only a few that really scale to grid-scale applications. Lithium Ion is one of the best known battery technologies, popularised through the rise in smartphones and laptops, and more recently through electric vehicles. These have plummeted in price over the past decade, and have been installed in a number of MW-scale installations across the world.
The the other principle form of battery technology used in grid-scale storage is flow batteries. They consist of two liquid chemicals separated by a membrane, which allows ion exchange between them. They have a lower energy density than Li-Ion, but can be scaled up to an almost unlimited degree simply by scaling the electrolyte tanks. Like most forms of battery, these come in a variety of chemistries, but vanadium flow batteries are one of the most mature forms, and have been rolled out to several energy storage facilities in China, Japan and the USA.
Molten state batteries are also under investigation as a grid-scale technology. They consist of two molten metal alloys separated by an electrolyte, and are relatively simple to manufacture, although dealing with the high temperatures of molten metal adds some difficulties. There are a number of different technologies under this umbrella, including ZEBRA and Sodium-sulphur batteries. Most of the efforts in this area are research-stage, but there are some deployments of the technology in Japan and USA.
Solid-state batteries are another up-and-coming technology which has received significant attention in the past few years. They use solid electrodes and a solid electrolyte, and are believed to be capable of 2.5x the energy density of existing batteries, due to their ability to tolerate higher temperatures. However, at the moment they are incredibly expensive, and their manufacture is believed to be resistant to economies of scale. There are also unresolved issues with the formation of dendrites (essentially one of the electrodes grows towards the other, compromising the battery more and more).
A related technology worth touching on is the use of electric vehicles for grid scale storage, sometimes referred to as Vehicle-to-grid. Electric vehicles require significant batteries for power, and as electric vehicle adoption grows, the energy capacity contained in these vehicles will become significant. A large amount of research has been conducted into using them as grid backup. There are also concerns that this would damage the vehicles' batteries. Vehicle batteries that are beyond end-of-life may also be usable as stationary storage, extending their usable life further (stationary storage isn't impacted by lower energy density as much as vehicular applications).
Thermal Energy Storage
Thermal systems are usually used to store heat directly (for heating/cooling) rather than electricity, but are worth touching on, as heating and cooling are major power draws, and some systems can also be used for electricity storage. Broadly speaking, they tend to divide into systems that store cold and those that store warmth.
Cryogenic energy storage (CES) is in the former category, and uses electricity to liquify either nitrogen or just air. It was originally envisaged as a way to power vehicles, but there is a pilot in the UK using it for electricity storage. The efficiency is low, around 15%, but might be increased if the cold lost from evaporating the liquid gas is captured somehow and reused. Along similar lines is ice storage air conditioning. This is generally used on a per-building level, and involves using off-peak electricity to create ice, which is then used for peak time cooling. Whilst it's not particularly suited for electricity storage, it can help reduce the sharp peaks in demand caused by cooling systems in hot areas/times of year.
Moving onto the hot stuff, we have molten salt storage. This involves storing energy as, you guessed it, molten salt. It can be used to store electricity in isolation, but is often used in Concentrated Solar Power plants, where the solar energy is used to heat giant tanks of molten salt, which can then act as an energy store through the night, allowing such plants to continuously produce electricity. The Gemasolar Thermosolar Plant in southern Spain produced electricity continuously for 36 days in this manner - a record for the technology.
Another similar technology is molten silicon, which replaces the salt with silicon. Silicon melts at vastly higher temperatures - around 1400°C - and so has a huge energy density per unit of volume (although containing substances at that temperature brings its own challenges). A company in Australia is currently attempting to commercialise the technology for grid-scale storage, yielding both electricity and heat.
Steam Accumulators are also used for storing heat, though this time as steam. They have been used for district heating systems and industrial applications in the past, although some solar power plants are using steam as the storage mechanism.
Most of these storage mechanisms are relatively short term, but there is also a form of thermal storage designed to work on a yearly cycle - seasonal thermal energy storage. This takes a whole host of forms, from storing energy in aquifers (used to heat/cool the Reichstag in Germany), using boreholes, energy pilings (heat exchangers built into the pilings of buildings during construction), and more. They all take advantage of seasonal differences in temperature to stockpile heat from the warmer seasons in some form of large store, then reclaim that heat during colder months for heating.
Here endeth the tour
So there you have it - a whistle-stop tour of potential and actual grid-scale storage mechanisms. There are many different approaches to energy storage, and whilst a number are still at a development stage, many have no major impediments to wide-scale usage. However, deploying energy storage at grid scale is expensive, which makes it a hard industry to enter into, and without cooperation of major industrial companies, it will be hard to transform the grid in the way that it needs to meet climate targets.
This is an area I'm researching more, as there are huge opportunities in the industry. As always, if you have any feedback, questions, or comments, please do let me know, via email or on Twitter.