As anybody who looks at current* wind output figures will know, we are presently blessed with less than 0.2 Gigglewatts of wind power from the total UK wind fleet, the rated capacity of which is close to 8 Gigawatts. For the last 10 days, output has been under 1 Gigglewatt and this means that the actual wind power is probably negative because each machine requires around 200 kilowatts of power just for its life support systems.
It is often claimed that wind and solar will be valuable if only they can have effective storage systems. This set me thinking and I append below a summary of my current thoughts. I would be very pleased to have any comments which can make this case stronger......
With such a low penetration it is not to be expected that Japan will encounter the problems in other countries, such as UK and Europe, where high penetrations from wind and solar are causing very significant problems for distribution and in increasing costs.
The recent report by the Adam Smith Institute – “The limits of Wind Power”, shows that any amount of wind penetration beyond 20% is prohibitively expensive and that the ‘sweet spot’ is between 10% and 15%. Beyond that point, the cost of having to have standby facilities on line and ready to carry full load becomes very high.
The problem with wind power and many other renewables is that they are inherently unstable and largely unpredictable and are thus quite unsuitable for any form of base load energy supply. Wind, in particular, is very variable and can change from high output to almost zero output and back again on a very short time scale, often on a basis of minutes. If we are to avoid the very serious consequences of such variability then we must have either a constantly available back up from conventional power sources, or some form of energy storage that would provide a constant and smooth output from the original wind power generation.
In order to overcome the inherent generation instability of wind and some other renewables (such as solar) it is necessary to have the capacity to store energy on large scale for protracted supply times. This, so far, has proved to be either very difficult or very expensive.
There are many possible methods of energy storage, all of which require a change of state from, say, wind to electrical to another form of energy and then a return to electrical energy. Each change of state involves an unavoidable loss of efficiency in that it is impossible to get out all the energy that was originally developed. This is a basic fact of physics that we cannot overcome. All that we can do is to try to minimise losses, often at considerable expense for meagre gains.
In one sense, we all rely totally on energy storage. All our food is actually solar energy that is converted into chemical states in plants, which are then converted again by chemical changes into the energy that keeps us alive. Fossil fuels, biomass and wood are simply ancient solar energy that has been stored as coal and oil and from which the energy is again released chemically into other forms of energy.
However, the immediate problem is to find ways in which we can store electrical energy from renewables in such a way that it can later be released in a controlled manner that is convenient to us. Thereby hangs the problem, for which there are currently few solutions that are operable economically on the large scale that we need.
There are many types of energy storage available to us, of which the main ones are as follows: -
Pumped Hydro is in practical use in many countries. It involves the use of cheap electrical power during off peak times to pump water from a low to a high level. The water can then be released as required to meet sudden peak demands and can respond very quickly. The higher you can raise the water, the less water you will need for a given power output. Therefore, countries such as Norway, which are very mountainous, can install such a system fairly easily. In UK, we have limited ability to do this and have used most of the readily available sites already. Low lying countries have very little opportunity to do so because the system would require huge land areas to accommodate all the water.
The biggest pumped hydro installation in UK is Dinorwig, in Wales. However, the total installed pumped capacity is equal, to only 1.2 GigaWatt hours of electricity and can deliver approximately 500 Megawatts for 13 to 15 hours until it is exhausted. The total installed capacity of pumped hydro in UK would produce at this level for not more than 22 hours. This means that it is just not capable of covering the capacity shortfall when our UK wind fleet can be producing almost zero power for several days at a time.
We can also look at this system from the point of view of energy losses. Let us ignore any inefficiency from production of power from wind factories and just assume that our electricity is from conventional sources.
When we pump up water for energy storage we have electrical losses to drive the pumps, then there are pumping losses and to this we must add the pipeline energy losses. The end result is that the stored energy loss costs us about 20% to 25% of the input electricity.
When we release the water to generate power we have pipeline losses, water turbine losses and further electrical losses. These may easily be as much as 20% to 25% in total and possibly more at peak powers due to pipeline losses.
Overall, therefore, we would be fortunate to get back as much as 60% of the input power, and would probably not see more than 50%. This is OK as long as we use very cheap, off peak electrical power, but if it is to be supplied by wind turbines we would not have cheap power because of the various incentives that are applied to wind power generation.
One can conclude, therefore, that the use of pumped hydro is only useful in very specific instances for peak power coverage and that it is not suitable for the longer term smoothing that is needed for wind power. Furthermore, any significant extension of pumped hydro installations can only be done at the expense of damming and flooding high level mountain valleys. This may be a problem because people tend to live in valleys rather than mountaintops and there are few available unoccupied mountainous valleys.
Pumped air. This is a very common method of power storage and is widely used for driving pneumatic tools. It simply involves the use of a motor to drive a compressor that supplies compressed air to a reservoir. The compressed air can then be released to drive a suitable machine that may be used to drive a generator to produce electricity.
It is all known technology for which most of the sums have been done and experience gained. The problem is that it has many efficiency losses and is currently used only on small-scale applications where the advantages outweigh the disadvantages. There are very few larger scale systems in operation and these are only experimental at present. In order to operate in the huge scale needed to support renewable energy variability, we need to go very big indeed.
The basic problem of compressing the air is relatively easily solved and could well involve such means as serial axial flow compressors such as are used for pumping on gas pipelines. However, we need to have very big facilities to store the compressed air and to deal with the heat exchange problems when compressing the air and when expending it for power generation. Of these, the storage is the most demanding.
One solution that has been proposed is the use of what are basically very big inflatable balloons that would be moored offshore in very deep water. The compressed air would be supplied to them and then sent back as power is required. There are many problems here, not least of which is the idea of having very large numbers of these devices moored in deep water, together with connecting pipe work and subjected to tidal flows etc. Condensation would be a problem also. For the GigaWatt scales that are needed, this just does not seem to be a sensible solution.
In order to obtain the huge volumes that are needed for air storage we need to think of underground storage in old mine workings, disused salt mines, oil wells etc. This requires that there are sufficient huge underground storage facilities that are easily accessible and reasonably close to the point of use of the power.
Even if we can find suitable storage, we still have the problems of inefficiency in the process. Compressing air is far less efficient than increasing water pressure and the same applies to its expansion to produce power. Even if we ignore possible losses of air due to leakage, it is very doubtful if we could expect more than a 40% overall efficiency.
Chemical conversion. As has been previously said, we rely on chemical conversion for almost all of our energy. However, in this context, we are looking at using renewably generated electricity to cause a chemical change of state to store energy so that it can later be released.
First off are storage batteries, as used in cars, for example. There is a whole range of batteries now available, including some exotics such as LI-on types. All of them rely on a chemical change caused by the incoming electricity so that a reversal of the change will produce electricity. The amount of storage capacity is a function of its construction and size and construction influences the discharge rate and hence the output capacity. Batteries use all sorts of special and possibly toxic materials and many of these materials cause great environmental problems during extraction. Battery malfunctions are not unknown (such as those currently affecting the LI-on batteries in the Boeing Dreamliner aircraft) and can cause serious fire and chemical risks. There is also the problem of limited life, as we all know from our cars.
There is, as yet, no battery system that can cope with long-term charge and discharge rates that are needed for the huge electrical loads that are required for back up to renewable generation. In any case, there are still the inefficiencies involved in taking a high voltage supply from the grid, reducing it to a lower DC voltage for the batteries and then reversing the process to give a mains output. Whilst this is common on small scales, it has yet to be shown to be viable on very large scales.
Another scheme that is being considered is to use surplus electricity to produce hydrogen by electrolysis. Quite easy, actually, and was a common experiment in my school days. Take water and a pair of electrical contacts in the water and, hey Presto, you get hydrogen and oxygen emitted. Collect the hydrogen and you have a good clean fuel ready to be stored for future use, either in cars or as a fuel for generators to resupply electricity. If the hydrogen is combined with CO2 we can get synthetic methane, another good fuel gas.
The big problems are of storage and efficiency. To be useful, hydrogen storage must be very large capacity, sufficient to run a generator for several days during lack of wind and/or solar power. That is a very big ask when we are dealing in Gigawatts and it has not been achieved so far. As for efficiency, we have to face the age-old problem that, whenever you do something, there is an energy loss. Each stage of producing hydrogen, compressing it, storing it and then releasing it for combustion will involve an energy loss so the end output will be considerably less than the energy input. The system would only be economical if the original input electricity is very cheap and even then, the output power will only be as clean as the source of the energy input.
There are several other possible chemical energy storage systems, but they all suffer from the same problem of storage capacity and process losses.
Mechanical storage. This simply means using various mechanisms to store energy for later release. It is actually quite common and in every day use.
For example, we can use a spring to store energy, as in a clock. Or we can use a weight, as in pendulum clocks. Very easy to use and understand, but quite incapable of storing large amounts of energy.
Another method could be to use a flywheel, which can absorb energy for later release. However, it is very unlikely that we can see any form of flywheel that can absorb the energy needed for compensation of power outages over days. Anybody who has seen an old internal combustion or steam engine running will have noted the huge flywheels that they need to keep a constant speed during power fluctuations for each stroke. These machines, big as they are physically, run only at kilowatt power levels. It us easy to see that a flywheel system to operate at GigaWatt levels for hours or days would have to be absolutely enormous. It is simply not feasible.
Thermal storage. This is a system that uses heat from a power source or direct from solar energy to heat a material so that the heat can be stored. The heat is then used to heat water to provide steam, which will then drive turbines to produce electricity.
The most famous of these systems is the Gemsolar Array in Andalucia, in Spain. This has an enormous array of steerable mirrors that focus solar energy on to a tower. The tower contains molten salts, which are heated and circulated to insulated storage vessels. The hot salts are used, via a heat exchanger, to produce steam, which then drives turbines that produce electrical power. The system has been operational and can produce up to 19.9 megawatts of electrical power. Because there is a large storage capacity of thermal salts, the system can continue operation even during the night, thus overcoming the most difficult problem of using solar energy.
It is theoretically possible to use wind-powered electricity to heat a salt in a similar manner and is not a huge technical problem (think of immersion heaters in hot water cylinders and kettles). However, the actual problems are very big indeed. The Gemsolar array can carry sufficient heat capacity to provide about 18 hours of electrical power before it literally runs out of steam. For any gigawatts scale system the heat storage would have to be enormous and would almost certainly involve substantial underground storage facilities.
Even if such storage were available, we would still have the ever-present losses to accommodate. Just consider this sequence of using a wind turbine to power a system using thermal storage.
Turbine > electricity > electrical converter > heat exchanger > thermal storage > pipelines > heat exchanger > steam generator > steam turbine > electrical generator > electrical grid.
Each (>) represents a stage at which energy will be lost through inefficiencies. If we assume no other losses and that each stage operates at something like 90% to 95% efficiency, which is high, it is easy to see that overall losses will be around 50% at best. This is hardly the basis for an efficient energy storage system and it could only be viable if the initial energy were to be very cheap, which is not the case with wind turbines in the present economic environment.
From the above it can be seen that there is currently no viable energy storage system that can allow us to use variable renewable energy sources to simulate base load electricity systems with controllable, economic, deliverable power over long periods of time. The only possible exception is pumped power storage, as at Dinorwig, but this is limited in availability and would require huge extensions of land usage in order for it to be useful. It also requires that the initial supply of energy should be at a low, economic cost.
Absent any new developments of efficient and cheap energy storage, it seems to be impossible for us to have renewable and variable power sources as part of our energy grid at levels beyond, at maximum, 20% penetration. The idea, therefore, of having any country with 100% of its energy supplied from renewable sources, is not tenable.
| Wind -
| Wind -