As the world transitions from fossil fuel generated electricity to renewable electricity, the intermittency of the clean energy sources with the highest potential (wind energy, solar energy and marine energy) is a significant drawback. Matching supply with demand is extremely challenging and, in truth, an effective solution at a grid level has yet to emerge.
One promising approach is to employ energy storage. Like many things, processing the raw product to improve the quality and/or make it more convenient to use increases value. In this case, excess electricity generation is stored in some form; when there is later insufficient production to meet demand, the deficit is drawn from storage. Superficially this works well, and it is frequently claimed that this approach solves the problem of intermittency. However, there are subtle problems that only emerge in the detail but are still a serious barrier. It is only by building plausible models that these issues become clear.
It is largely accepted that batteries are most suitable for short term storage, smoothing fluctuations over a period of days. The efficiency is high and batteries are extremely reliable. Longer term storage is also required (perhaps to carry energy across seasons) and in this case batteries are unsuitable because of mass, volume and cost. Though the efficiency is lower, hydrogen storage is a better option for bulk storage because the capacity can be increased simply by increasing either the gas storage volume or pressure without necessarily scaling up the capacity of the conversion components.
In the model above, hourly wind data from a location in Scotland are used. Hour 1 of the model is midnight on 8th August 2019, and, including the 2020 leap day, that is 8784 hours in total. The wind data is from a standard height of 10 m, but is adjusted to a wind turbine hub height of 100 m – a generic wind turbine power curve is then applied to find the hourly power. A 4.5 MW wind turbine is chosen. The solar data is the energy that would be produced by a 1 kW rated solar PV panel flat on the ground. The figure is multiplied by 2,500 to represent the hourly power output from a 2.5 MW PV array. The hourly consumption data of a small town is used, but it can be scaled up to a city if required. The default battery capacity is 10 MWh and we do not concern ourselves with the charge or discharge rates. The hydrogen store can be pressured up to 700 bar and has a capacity of 50 cubic metres at this pressure. Most parameters can be altered, but some are fixed (such as 95% battery efficiency for charge and discharge and the equivalent of 65% for the hydrogen system) where there is no learning benefit in altering these. The stores are assumed to be initially half full.
The model is run with the ‘Start’ button and the display shows the energy being generated by the renewable energy system and the town/city demand. There are three energy losses. The battery losses are the consequence of 95% charge and discharge efficiency (hence 5% loss on each transfer) and there is an equivalent loss associated with hydrogen storage. The curtailed electricity is electricity that could be produced (and is therefore included in the renewable energy generation total), but it not because there is nowhere for it to go – there is insufficient demand and both stores are full. In such circumstances the generators are normally disengaged. The grid electricity figure is electricity that must come from the grid at times the stores are empty and instantaneous production does not meet demand.
Running the default model, we see that of the annual consumption of 25,160 MWh, only 2,462 MWh comes from the grid. You might think that’s pretty good; you only need one tenth of the electricity from fossil fuels you previously used. However, it is not that simple - there are consequences. The grid system still needs to be maintained but the grid operator now has much less revenue than before, and the need to supply missing electricity at random times requires conventional power stations on permanent and costly standby. You have largely offloaded a difficult balancing problem to someone else.
What about trying to supply all the energy without grid intervention? This would be possible if the complete local energy system included a reliable clean source such as a biomass generator (though that is not without problems), or perhaps an aerobic digester to produce methane, or even a waste incinerator. Alternatively, you might try controlling the way people use electricity through appropriate pricing measures implemented with smart meters, or you might take a smart grid approach where appliances in the home are externally and remotely controlled. The use of heat pumps might also be appropriate. These are very complex solutions and perhaps the mixing of many sources simply hides rather than solves the underlying problem.
You can see zoom in on the problem by experimenting with the wind turbine, solar PV, battery and hydrogen storage capacities in the model above. By increasing generation capacity, less electricity is drawn from the grid, but the amount never quite reaches zero unless very large numbers are chosen (there is clear evidence of diminishing returns). The penalty is high capital and maintenance cost and a vast increase in wasted energy. Perhaps something can be done with this energy, but the battery and hydrogen waste heat are difficult to recover and the curtailed electricity spirals and is difficult to use for any new industrial process because it is highly variable, even more variable than the original energy (and consequently of even less value). The temptation would be to use it to heat homes to displace some demand from fossil fuels, but the effectiveness of this approach would need to be tested by constructing a separate model.
Another way of reducing grid reliance is to increase storage, but this comes at an enormous cost and there is the problem of requiring very expensive equipment utilised for only a short time. We will look in detail at this elsewhere, but some explanation here might help. Imagine you have storage equipment costing €1,000. This allows you to store 1 kWh of electricity at a point in the day when it is not needed for use later in the day when otherwise you would have paid €0.20 for a unit from the grid. In one year, you have saved €72 and the equipment pays for itself in 14 years. Now imagine you have an expensive system that is carrying energy across many months. The gain each year from multiple daily accumulations is then not possible and the economics are less favourable.
The bottom line is that bulk storage needs to become very efficient, reliable and cheap for an intermittent renewable energy and storage combo to completely displace reliable fossil fuels. In the meantime, a possible solution is accepting very expensive grid electricity (when needed), perhaps 2-5 times the current price, to cover the operator's additional costs; a bigger electricity bill overall, but much less carbon dioxide released into the atmosphere.
Investigations