Tuesday, January 4, 2011

Energy Storage & Renewable Energy Sources

Energy Storage & Renewable Energy Sources

Table of contents
Informal abstract
Before reading on, please note:
Introduction to this theme in this discussion.
Aspects of compressed-gas energy

Compressed-gas storage for local power
Compressed-gas energy storage for transport and home
Compressed air for efficient combustion.
Compressed air for residential purposes
Compressed-gas energy for large-scale local use
Submarine Compressed-gas storage for global power
Fluid inversion energy storage.
Submarine tent farms: hazards and objections.
Uses for submarine compressed air.
Sources of energy for storage in compressed air.
In Sum

Informal abstract
It has become an article of faith among denigrators of renewable energy sources such as wind and wave, that they are practically worthless for reducing base load requirements because there is no way to store excess energy against times of low input. This essay describes certain classes of energy storage that should be quite adequate at scales ranging from domestic to city-wide or even country-wide requirements for energy storage during times of scarcity. They involve little or no pollution risk, have long working lives, and in general, unusually desirable attributes from the point of view of the power engineer. For one thing, apart from mitigating problems of intermittent energy supply, the energy storage units would constitute an excellent buffer either to condition diffuse input power for concentrated power output, or to smooth excessive or erratic power input to match the needs of the power grid. They thereby should increase the range of feasibly useful energy resources, including traditional combustion-based or hydropower generation.

This is not a document cast in stone; it is a discussion of non-technical draft proposals. Anyone with doubts, queries, suggestions, objections, elaborations or corrections should please feel welcome to contact me.
If you do so, please let me know whether your correspondence is in confidence, or you would like to be credited with any changes or additions that you might inspire. At the time of writing my email address is jonrichfield@gmail.com
Of course there also would be no hard feelings if you preferred to write an independent document of your own without consulting me.

Wind, wave, solar, and tidal power, together with various other renewable energy sources, may or may not prove to be sufficient in scale to meet the entire needs of world-wide, or even first-world, industrial and lifestyle needs. This essay does not address that question.
I do however accept without reservation that such sources are sufficient in quantity and potentially sufficiently benign in nature to be of considerable practical value, whether or not they on their own are sufficient to meet our global needs. Either they could make a substantial contribution to national and international energy networks, or they could be of value in meeting the full energy requirements of smaller local installations.
Both of these categories of contribution are important, although in effect many critics assert the ludicrous view that no replacement for any part of our fossil fuel consumption can be worth consideration unless it is complete and immediate.
Another point of criticism is that all such sources of energy are intermittent in nature. This largely is true, certainly to the extent that it is a serious problem from every point of view.
However, objections on such grounds are grossly over-generalised. They take two forms, at least by implication. The first objection is as stated above, that any currently imperfect or incomplete solution should be rejected without further consideration. The second is that where there is any intermittency in supply, the base load must be supplied by traditional means and generally on about the same scale as if there had been no intermittent renewable contribution.
I disagree radically with both forms of such assertions. For one thing, as I illustrate below, such sources can be of huge value in partly meeting peak power needs and thereby reducing the base load requirements. Peak loads that represent only a small percentage of power consumption, may none the less determine the cost of traditional infrastructure. By say, halving the peak load (or by increasing the sources of power available to meet peaks) one can drastically reduce the infrastructural costs of the base load as well as improving the flexibility of the total installed infrastructure in meeting emergency power demands.
Anyway, even though that objection is substantially unjustified, I do accept that the cost and technological difficulty of storing energy on a large scale are forbidding. As a technological challenge the field deserves attention.
At the same time I deny that it follows that the problem is intractable. There are options for storage adequate to bridge any plausible or practical break in supply, even on the largest scale. Suitable installations could do so cleanly and economically, using existing technology, even though most options would require some enterprising and exploratory engineering. For some applications it should be feasible to store the energy from wide ranges of sources to achieve brief but drastic power output such as for the economical launching of spacecraft. That is not a form of power demand that lends itself to routine installations or reticulation.
Conversely, energy stored in such ways could buffer either irregularly low or irregularly high levels of power input, permitting smoothly conditioned supply of power to the national grid. This is important in various ways beyond simply lending plausibility to say, wind, wave, and solar power schemes. Apart from permitting more effective and reliable trimming of base load requirements, a large-scale energy-storage medium also serves as a conversion mechanism for all kinds of energy input. Any energy resource that can contribute its input, even a source that otherwise would be regarded as too diffuse, too low-grade, too intermittent, or too undesirably timed, becomes valuable in proportion to how much energy it can yield and what it costs to deliver that energy to the store.
In other words, a successful system would at once increase the range, quality, and quantity of energy available to us. 
This aspect has a wider range of significance than might at first spring to mind. Obviously the intermittent nature of renewable sources is a problem during periods of undersupply. In fact during such periods energy demand is likely to be at its greatest for the very reason that caused the undersupply; for example, solar energy  is at a premium during nights and during periods of cold weather. 
However, during periods of local oversupply, where say, both wind and solar energy are flooding in simultaneously, and the grid is committed to accepting and paying for electricity from private generators, it happens on occasion that the network is locally overloaded by the excess.  At such times any means of energy storage is doubly welcome if it can absorb the temporary excess for consumption during subsequent periods of shortage, thereby converting an embarrassment, even a threat to the functioning of the grid, to an asset.
The main interest in energy storage technology has centred on electric batteries of various sorts, and I do not deny that batteries may indeed some day be sufficiently practical and economical on a scale that renders other media redundant, but at present such prospects seem tenuous, or at best uncertain. I largely ignore batteries for the purposes of this discussion, whether primary batteries such as fuel cells or zinc/air, or secondary, rechargeable cells or capacitors.
Another class of energy storage medium of practical importance that I ignore is chemical storage in forms other than batteries, such as organic fuels, atmospheric oxygen, or electrolytic hydrogen.
Instead I concentrate on what I shall call mechanical storage: storage of potential energy, stress/strain energy, and the like. A good example is clockwork, whether spring driven or weight driven. Clockwork has served as an energy store for centuries, and often very reliably. However, clockwork is not the example that I shall concentrate on. I see little prospect of clockwork making serious inroads into the classes of problem under consideration.
There also are many other interesting and practical storage media that I ignore, such as flywheels, though they too may prove immensely valuable in practice.
Two examples of potentially large-scale modes of energy storage are more promising in context: pumped storage of water or other massive fluids displaced to heights sufficient to yield useful hydroelectric power, and compressed air storage. Pumping water into high-level reservoirs to drive turbines has been of value in certain sites round the world. I know of at least one example in South Africa and one in Britain, and I know there are several more. Such reservoirs work very similarly to the way that the weather raises the water that in the form of rain, fills dams for hydroelectric power schemes. In practice the main differences usually are:
1. The energy to raise the water for pumped up storage schemes is supplied from other usable sources during periods of plentiful supply or low demand.
2. The period of consumption of the pumped storage energy is short, typically hours of peak usage rather than months of routine consumption.
3. The pumped reservoir routinely gets emptied during a cycle, without concern for ecological effects. This differs from say hydroelectric power from a lake, which as a rule one cannot afford to drain periodically.
Typically such pumped-storage schemes are fairly clean and designed on a moderate scale, but not many sites are suitable for the purpose. Pumped-storage dams are expensive to construct, land-hungry, incompatible with other uses of the real estate they occupy, and the facilities tend to be needed in areas where land prices are high.
There is theoretical scope for developing the principle more highly. For instance, water towers could be filled by pumps powered by wind turbines, to be exhausted during peak consumption or when the wind drops, but really, I cannot see such schemes becoming very prominent or very large-scale. I do however discuss some specific applications where raised water could condition the power output of other types of storage.
However it is important to understand that energy storage on such a small scale cannot be of much use for anything beyond dealing with brief peaks in demand, typically local demand. Storing energy on a scale sufficient to bridge days or possibly weeks of calm would require a far larger scale of infrastructure.
In contrast to pumped hydroelectric energy storage, energy storage in the form of compressed air is just starting to be taken seriously in non-trivial applications on a non-trivial scale. Compressed air is in some ways a particularly troublesome storage medium, but it does have attributes that have moved me to devote much of the rest of this discussion to the subject in one form or another. At the time of writing, at least one underground scheme is operational in Huntorf, Germany, and one in Alabama. Both accumulate compressed air in chambers excavated in rock salt.
In another scheme in Britain a chamber has been excavated in a granite mass. Excess off-peak power capacity is used to pump air into the cavity, and during peak hours the compressed air drives turbines to supplement the base capacity.
Notice how the ability to use such stored energy to trim the required peak power generation capacity is so profitable that it can pay for a comparatively expensive and only moderately efficient installation. In the face of such applications, thoughtless claims of the uselessness of unconventional energy sources to reduce base load capacity are not in anyone's rational interest.
At the time of writing, several entertaining and informative discussions of compressed air energy applications are to be found on en.wikipedia.org, under headings such as "Compressed-air vehicle" and "Compressed-air energy storage". They offer a useful background and some interesting history and principles, but they do not deal very directly with the concepts that I propose here. However, the fact remains that there is a current trend towards increased interest in compressed air energy storage and examples are proliferating on the internet.

Aspects of compressed-gas energy
Compressed gas has several attributes that are highly unwelcome, but it also has also several excellent virtues from the point of view of the power engineer. There are various modes in which such power can be used, and some of the modes present almost exactly opposed problems and merits.
The main gases that I discuss in this essay are air and nitrogen, possibly with occasional reference to hydrogen, natural gas (mainly methane), CO2 and noble gases.
The following points are not in any special sequence or carefully structured context:
· Compressed gas safety is variable. Most of the gases under consideration are not particularly toxic, explosive or corrosive, though most are not actually respirable. However, toxicity is seldom a consideration because we are not much discussing enclosed spaces where people might be asphyxiated. Also, except where oxygen, oxygen-enriched air, or fuel gases are considered, there is little risk of flammability.
Still, compressed gas that is explosively released or is kept at pressures near to the safe engineering limit, can be very dangerous in various ways, irrespective of toxicity. Where really high pressures are applied, the gases might dissolve unexpectedly in liquids or even certain solids. This may lead to troublesome chemical and physical effects. In particular it may cause harm or hazards when pressures fluctuate frequently or widely. If liquefied gases are under discussion, as opposed to merely compressed gases, they present yet other hazards. However, in the major applications under discussion the use of pressurised gases presents few serious dangers if well engineered.
· There always are problems in equipment for handling gases under fluctuating pressures. High temperatures, low temperatures, condensation, unexpected chemical reactivity, and similar considerations arise. Such problems are in no way unfamiliar to engineers, and no exotic techniques are contemplated in this discussion.
· Changing pressures cause changes in temperature. The resultant heating or freezing can be a nuisance, but a fundamentally more serious problem is that waste of the heat produced can cause serious inefficiencies. Conservation or constructive application of such heat or cold are ideals that the engineers should bear in mind. Examples of appropriate techniques include heat exchange, heat storage buffers, constructive heat usage, or application of the cold to condense fresh water from moist air in regions where desalination is desirable. In some circumstances actually freezing usable fresh water out of salt water might be worth while as an incidental application.
· For most purposes probably the major inefficiencies in employing compressed gases as energy stores are heat waste during compression, and leakage in the distribution system. To deal with the former, I propose that designs aim at the recapture of such heat, which, though technically demanding, should not be a serious challenge to the ingenuity of appropriately trained and experienced engineers. As for the second inefficiency, either appropriate attention to leak proof transmission systems, or designs that rely on very short range conveyance of the compressed gas to electricity generating units should be routine. Demanding though such requirements might be, the concepts are hardly groundbreaking.
· Storage of gas under high pressure can be an extremely efficient way to contain energy without moving parts, decay, pollution, or unstable compounds, but it demands careful vessel design. Polymers are commonly economical and desirable vessel materials, but many have tendencies to creep more rapidly than metals or ceramics. Such applications require careful choice of material, such as cross-linked oriented polymers, or compound materials, such as reinforced polymers. Polymer engineering is tricky, but it is becoming more routine with increasing experience and with the availability of increasing numbers of polymers with special attributes.
· Given suitable container materials and engineering, it is possible to maintain statically stored energy in impressive quantities and densities at modest cost in the medium of compressed gas. Systems of moderate levels of technological sophistication can rival electric batteries, especially at large energy volumes. Compressed air at millilitre volumes would rarely be worth consideration, but on tonnage scales it could compete very well with rival storage media.
· Compressed air and various other gases, filtered, dried, and heated if necessary, are exceptionally clean media for driving many types of power unit such as turbines or pistons. Such equipment is comparatively easy to maintain indefinitely under conditions of constant use if properly designed. Apart from being valuable simply for driving turbines, compression of fuel gases also can augment the efficiency and power output of appropriately designed internal combustion or dual action engines. Such hybrid systems also can deal very conveniently with the cooling problems of expanding gases by recovering the heat of the exhausts.
· Compressed air stores, like most other stores, commonly present a problem in that they deliver energy at a higher intensity when fully charged than when nearly empty. However, this is tractable to the professional engineer, and some classes of devices deliver nearly constant pressure from when they are nearly full to when they are nearly empty. To withdraw power only within a narrow range of pressures is the crudest approach to the problem, but it is perhaps the commonest in practice. More sophisticated technologies include vessels that contain a reservoir of liquefied gas at a constant temperature, vessels that maintain a gas store under a constant head of liquid, and gases stored under constant external pressure, or under a constant compressive weight.
· From certain points of view certain classes of compressed air stores are poorly scalable. For instance, even under constant pressure, increasing the diameter of a given spherical pressure vessel by a factor of n also increases by a factor of n the stress its skin must withstand. Although the force is spread over n times the section of skin material, the total force is a function of the square of the dimensions, so that the resultant stress remains n2/n. So, roughly speaking, other things being equal, the skin thickness must increase linearly with the dimensional increase. This is not too bad, but for high pressures and large vessels it becomes expensive, and as the vessel is repeatedly discharged the risk of crack growth increases disproportionately in thick skins. There also are related problems with the design of pressure vessels constructed from compound materials such as fibre reinforced plastics or concrete.
· In contrast, from other points of view compressed air stores are highly scalable. Increasing the dimensions of a spherical vessel by a factor of n would increase its capacity by a factor of n3 at the same pressure, so one must beware inappropriate comparisons. After all, increasing the wall thickness and radius of the vessel by a factor of n increases the wall volume by a factor of about n3 too, which is fair enough. Conversely, as will appear in context, in some applications the containers can be enormously scalable at a very modest cost in material.
Note that I have mentioned no fundamentally insuperable obstacles, only problems related to fundamental attributes of gases and structural materials. All technologies depend on engineering to mitigate such problems by design, configuration, and related disciplines. One should be wary of dismissing new technology too glibly. After all, if I were to suggest a means of transport based on devices that must contain repeated internal explosions, or that burn fuel at high temperatures in mechanisms that cannot withstand such high temperatures, informed parties might well cite such considerations as patent supplementary evidence for my perversity or outright stupidity. And yet, a century on we have some piston engines that by the original standards of Benz and Daimler are quite nice. By the standards of A. A. Griffith we also have some quite nice gas turbines, partly by courtesy of F. Whittle.

The main theme of this essay is compressed-gas energy storage on a scale to run national grids, but for purposes of perspective and for interest in passing, smaller scales of application deserve attention. Before proceeding to discussion of the forms of compressed gas storage suited to national and trans-national requirements, note how less ambitious proposals also could be useful in suitable circumstances. Here are a few examples.

Compressed-gas energy storage for transport and home
When there is a need for compressed gas energy storage and there is no infrastructural network of piping to deliver it over long distances, it is possible to build local vessels of modest size from a range of suitable materials, such as fibre-reinforced matrix material or simply from metal. There already is serious interest in cars powered by such compressed gas, either alone or in hybrid fossil fuel engines.
When used for transport, compressed air could be produced either industrially and sold at the pump, or produced locally by the consumer. Whether to charge the vessel by using off-peak power or windmills or any other power sources would depend on local infrastructure or personal resources. The domestic user might use a reticulated power supply, or domestic wind turbines, or local shared wind turbines, or any other source of energy as appropriate. As a matter of practicality, high pressure compressed gas reticulation tends to be challenging and wasteful, but that does not necessarily exclude it as an engineering option for appropriate requirements and resources.

Combination of compressed air, possibly oxygen enriched, with combustible fuel, certainly is an idea with potential. I would like to see the super-charging potential of the gas used more ambitiously than anything I have seen at the time of writing. We have an impending global shortage of fossil fuel, but such fuel as we do have, apart from petroleum gases, tends to be rich in carbon and poor in hydrogen. Accordingly cracking tends to produce an excess of carbon and tarry compounds.
Now, suppose such carbon and other low-grade fuels were converted to suitably structured blocks or possibly pellets. Suppose the compressed gas were slightly enriched in oxygen, say to 25% or 35%. This is an easy and cheap thing to do on an industrial scale, but note that simply supplying the air at a high pressure also promotes efficient combustion. Then by passing the gas through the ignited carbon-rich fuel blocks, one could consume them cleanly. Such consumption would drastically increase the efficiency of use of the gas, efficiently consume the carbon fuel, and hugely increase the range of a car that otherwise must be driven by cold compressed air.
Alternatively compressed oxygen could be used during the ignition phase, with plain compressed air supporting combustion after reaching a working temperature.
An afterburner could ensure that CO production would be far lower than in a petrol car, and the relatively low temperature and non-explosive pressure of combustion would minimise the production of NOx and similar unwanted compounds. The fuel would be far less toxic, less polluting than petrol or even diesel, less of a fire hazard, easier to store, and quicker to install in vehicles at refuelling time. And in emergency circumstances where no fuel blocks were available, the car still could run on pure compressed air, though neither as far nor perhaps as fast.
Such solid fuel also could increase the scale of vehicle that could benefit from compressed air. Heavy duty trucks that could never rely purely on compressed air could use such fuel blocks instead.
Similarly, instead of running domestic turbines purely on compressed air, one could multiply their yield by running them off hot gas plus low-grade fuel. For one thing it would obviate problems of adiabatic cooling of the compressed gas.
All in all, such solid turbine fuel is a subject that deserves a dissertation on its own. It certainly would raise the hackles of persons opposed to the release of carbon dioxide into the atmosphere, but that is a topic I deal with elsewhere than in this essay.
A related combustion principle already is exploited in using compressed air for burning fuel gas in power turbines. This enables one to run the turbines without external air compressors, thereby increasing their efficiency by amazing factors.

For residential purposes simple compressed air containers, with or without combustible fuel, might be charged by pumps attached to renewable energy sources such as wind turbines or solar cells. Such reservoirs could say, supply a house through the dark hours and windless rainy days. All that would be necessary would be for the excess energy from the source to be sufficient to charge an adequate storage tank. Of course, a static installation of a carbon-burning arrangement similar to that mentioned in the previous section could also be useful here.
One practical concern is that high pressure tanks are intrinsically hazardous to install above ground. In an industrial setting, design of compressed-air tanks of reasonable specifications should not present disproportionate problems, but the ingenuity of the laity in creating disasters with even the most innocent domestic devices, legendary though it is known to be, none the less is perennially underrated.
Instead of installation above ground, a tank with a capacity of a few cubic metres could be installed underground in contact with the surrounding earth at a depth of several metres. Though expensive this would have at least four advantages.
  • It would take the most dangerous unit out of reach and out of mind of most meddlers, saboteurs, and many other hazards
  • It would protect the most vulnerable potential victims in the event of a rupture.
  • Storage below ground level also would permit the tank to contain a far higher pressure, as long as the wall were suitably mated to the surrounding soil or rock.
  • The increased working pressure for a given cost of tank construction could offset the increased cost of underground installation.
This might strike one as a toy installation, and not unreasonably so, given the ambitious promises I have been making on behalf of compressed air energy storage, but as routine domestic equipment such facilities could perform a wide range of individually and socially valuable functions all the same.
The cost of underground installation could be reduced if such installations became sufficiently routine to justify the design of specialised equipment for digging and preparing suitable holes. Furthermore, the cost amortisation should be attractive. A well-designed installation should have a long life, many decades at least. A single vessel might well outlast a whole series of houses built above it and utilising it.

Suppose that we increase the scale of design and adapt the form of the store accordingly. Let us consider an approach suited, not to individual dwelling units, but to blocks of flats, small suburbs, large sectional title schemes and so on.
The first step would be to commission some serious geo-engineering investigations. This would be less to determine whether any scheme whatsoever would be feasible, than to see which kind of scheme would be most practical and desirable at any particular site. For example the investigation should include assessment of the ground water situation. For instance, unusably saline ground water, much saltier than seawater, is widespread in Australia and also is common in many other arid regions; such water usually is of no practical value, but we would not want to compromise potable or near-potable water supplies.
We also would need to reckon with the strength and nature of the earth at the site. Granite might be strong, but it is very expensive to excavate into large chambers, and we would need detailed information on the risk of bursts or even cracks. Clay in contrast could not withstand very high pressures, but if moist, it would offer interesting potential for expansion of a hole by high pressure instead of excavation. The hole lining would have to be suited to the surrounding rock. Shale or similarly untrustworthy material might well make certain classes of scheme prohibitive.
Let us suppose for example that we have found a nice, deep, moist layer of clay-impregnated gravel, say 100 metres below the surface, extending to infinity downwards and sideways. We could drill downward into such a deposit, say another couple of hundred metres or so, according to our needs, and excavate a suitable cavity by some competently selected means. On a green-field site we might use brute-force inflation with water or air, or we might dig out material with high-pressure water jets. Once the cavity is complete we might line it with a spray of suitable cement or polymer.
The shape of the eventual cavity would not be spherical because part of the point of digging so deep is to exploit the hydrostatic forces of the soil at that depth. If we assume a soil density of 2, the hydrostatic resistance alone at the top of the cavity 100 metres below the surface would be some 20 atmospheres. If the cavity itself also measured 100 metres from top to bottom, then the hydrostatic resistance at the bottom of the cavity would be about double that at the top. So if we made the top say, circular, 6 metres across, the whole cavity should ideally have a diameter of 12 metres at the bottom. That gives something like 6000 cubic metres, if the back of my envelope is working properly. If the cavity is suitably rigidly lined, it could be emptied in an emergency, but it normally would be maintained at a pressure considerably higher than 20 atmospheres, and tapped as required.
Such installations would suffer from their reduction of effectiveness as the pressure reduces while they empty, but an elaboration on their design could at once increase their energy capacity, level out their pressure delivery, and reduce any tendency to collapse under the pressure of the surrounding earth as their air content reduces. Install a curtain or stocking of flexible polymer membrane, such that when the vessel is full of air, it occupies the space on one side of the membrane. Install a surface or tower reservoir that holds water at least sufficient to fill the underground cavity. As the vessel empties of air, the water fills the space on the other side of the membrane, and as the compressed air is recharged, it displaces the water and forces it back into the reservoir.
The reason for the membrane is that air under pressure dissolves quite freely in water, and this might be very wasteful. Quite a light membrane could prevent that.
A membrane probably would be most practical, but there are alternative approaches. For example a few tens of litres of a fluid that is less dense than water, and does not readily dissolve much air under pressure, could be maintained in the cavity to separate the water from the compressed air. Possibly a floating slurry would cover the water better than a pure fluid.
Yet again, there might be merit to designs that actually encourage solution of the air in water, so as to stabilise the release of the air over a wider range of pressures. There is room for research to establish the most promising technology and design.
In any such installation the air pressure then never drops below about 10 atmospheres, or whatever the pressure of the head of water would amount to. Such a pressure can drive a suitable turbine very adequately. Furthermore, the amount of energy stored in raising say, 6000 tonnes of water 100 metres would be a valuable part of the storage scheme. Depending on the pressures in question, it might dominate the energy capacity of the installation.
This is an example of where a water tower could help condition the output of another energy source.
Air at say ten to thirty atmospheres, on a scale of some thousands of cubic metres, should keep quite a few households or a light industrial installation running for some time. There are however many significant adjustments one could make to the design, and I do not intend to waste too much time on this theme.
There also is no theoretical reason why similar designs on a much larger scale could not be used to store energy for cities. I wonder whether played-out mines, such as those (some 3 km deep or more) around Johannesburg, could not be sealed and used in similar ways without fancy shaping. Johannesburgers are used to earth tremors by now anyway.
However, conical excavations similar to those I have described, but on a scale of a kilometres rather than hundreds of metres, might well be worth constructing where there are suitable energy resources worth storing for a city. The order of magnitude of storage would then be millions of cubic metres rather than thousands. The cost should not be much greater than a major fossil fuel power station, and the pollution would hardly compare with any other power generating medium whatsoever. The useful life of such a resource, if it is suitably planned, should be measured in centuries or millennia rather than decades.
Notice that the problem of adiabatic heat production would be trivial in subterranean cavities in comparison to most pressurised gas containers. Heat doesn't conduct rapidly through more than a couple of metres of soil, so it should be fairly well conserved. In fact, in the deepest mines we even could profit by exploiting the ambient underground heat already present, especially for counteracting the cooling effect of air expanding into the turbine or other target device.
In summary, given enough renewable energy or even off-peak energy, we could do spectacular things with imaginative application of pedestrian engineering.

So far we have been considering local-storage devices, rather than national supplies. They should be very useful and do wonderful things for trimming the scale of base-power installations, but will tend to be expensive and limited in scale. It is hard to imagine any such installation on a scale larger than what would be necessary for running a large city for a day or so.
Note however, that energy storage on a scale adequate to run a large city for a day or so is more than adequate for all normal purposes, not only purposes of peak trimming, but of base load trimming as well.
When we consider marine installations however, we begin to see potential for national or international scales of storage, rather than a scale suited to a city or two. Hardly anything we do at sea is ever really cheap, but some approaches offer attractive options.
Consider the idea of storing air under submarine tents (or capes or balloons if you like) tethered to the sea floor. At first ignore the design details and think of the principles. Air stored under such vessels would be at the pressure of the surrounding water, and that could be anywhere from a usable excess of about one atmosphere to about 400 atmospheres, depending on whether the depth is ten metres or four kilometres.
Assume that the tents are roughly hemispherical and open at the bottom (in practice they probably would have a flexible membrane inside to keep the compressed air from dissolving in the water beneath). The pressure stresses on the empty structures would in most respects be trivial. When the tents are empty their materials would experience only the compressive stresses of the water, which would be negligible for any solid submerged object such as a membrane or a cable. Even stresses from pressure differences within the structures when full of air, would be limited to the stresses resulting from the difference in depth between the tops of the tents and their skirts, in other words the buoyancy forces.
The full tents would exert a buoyant force, which would become a major consideration for a large tent. I have arbitrarily considered 1-million-cubic-meter tents, roughly hemispherical (probably catenoidal in practice), which very roughly should be about 160m in diameter, large enough for a reasonable Ozzie-rules footy field. This might be ambitious, but the optimal size in practice would be for the engineers to determine.
Obviously of course, there would be many a stage of design, redesign, installation and development of infrastructure before anything like the scale I propose here would even be contemplated, but there is no point my wasting time on the initial and intermediate stages; the first prototypes probably would be developed in baths or bathing pools, and none the worse for it.
Be that as it may, given a buoyancy of about 1 million tonnes, such tents could be tethered by ballast sufficient to counter roughly a 1 megatonne lift. Using 1000 tonne working-stress cables, that would require about 1000 cables attached around the skirt at 0.5m intervals. If we use a light-gauge tent membrane, then the tethers could be attached to the skirt of a woven mesh cap that would cover the membrane and take the stress of the buoyancy.
Although it would be possible to attach those tethers to the membrane, it probably would be more practical to use an extremely light membrane, just adequate to confine the compressed air behind the mesh. The membrane would amount to a liner for the mesh cap, which would take the main load. The liner would be much like the relatively weak inner tubes that used to be used in vehicle tyres before tubeless tyres became so nearly universal.
The buoyancy pressures, as opposed to the hydrostatic pressure of water at a depth of some km, would amount to about 80m of (sea)water, or roughly 8 atmospheres. That is a modest working pressure for the lined mesh cap to contain. A membrane of suitable polymer 1 cm thick should be adequate. Probably fibre-reinforced polymer would be better.
Instead of attaching the tethers to the sea floor, I prefer the idea of terminating the tethers in linked chains of weights that, when the tent is empty, would lie flat on the sea floor. Each "link" could have an effective weight of say 10 to 100 tonnes, which would imply 10 to 100 weights per tether. As the tent filled, it would progressively raise more and more of the weights off the bottom. With the tent at its fullest, only the last ring of weights would remain unlifted. Such a design would be independent of ooze on the ocean floor or the strength of the underlying substrate. It also would be ecologically benign, possibly even beneficial. Installation of the untethered chains of weights would be simpler than trying to dig and install concrete attachments. Practically all the abyssal work, whether installation or maintenance, should be undertaken by unmanned robot craft partly under remote control from the surface.
Design of the infrastructure would obviously be an interesting exercise in many respects. Possibly other designs than chains of weights would be superior; possibly the distal weights would need tethering to counter centripetal vectors as the membranes fill with air. I simply describe them as an example that illustrates the practicality and flexibility of the principle.
To assist in clarification of the verbal description, I attach a schematic illustration of a single tent at the end of this section, but the precise form of the tent I leave as an exercise for the engineers to determine. I am uncertain whether a catenary profile or something more sophisticated would be best, but for this discussion I elect to think in terms of a hemisphere as a convenient first approximation. Whatever mathematical form is chosen, for reasons of material efficiency it certainly should not deviate too far from a hemisphere.
The energy storage capacity of a megatonne of water displaced at a depth of 2-4 km is formidable. It dwarfs the already considerable energy capacity of 200-400 million cubic metres of air compressed into 1 million cubic metres. It amounts roughly to lifting 20-40 million tonnes of water 100 metres. That is 20-40 teraJ or 5-10 GWh depending on how many decimal places I have lost.
Now, there is no reason to install just a single tent at any one spot. For the price of extra ballast, tether and membrane, one could stack tents say 100m apart vertically. That would increase the capacity perhaps ten to forty times, enough to run a major city for a few days. Furthermore, most sites suitable for such a stack would be suitable for siting many more stacks in the same vicinity. Siting stacks at intervals of some 300m, one could fit some 9 stacks per square km, or 90000 in a 100 km sq stack farm.
That may sound like a big farm, but if you misplaced an area of 100 km sq even in confined waters like the bay of Biscay or the Mediterranean, it would take a search party to find such a tiny patch, never mind in some of the larger areas off the continental shelf of Africa or the Americas. And note once more that such farms would be ecologically benign in every way. In fact they would amount to submarine conservation areas. This is not just my disordered fancy; it has repeatedly been demonstrated in real life with oil platforms and sundry other engineering works.
The devil is always in the details. One possible concern is the question of how to prevent tackle from tangling, breaking, or sinking when sinking would be disastrous, or floating when floating would be no better. The scheme would hardly have been worth discussing say seven decades ago. Of course, it would be desirable to be able to make suitable members of the structure rigid, others flexible, all of them indifferent to salt or water, and pretty close to neutrally buoyant in sea water. For example, we want the membranes to remain in place inside their mesh caps whether they are empty or full; we want the ballast chains to remain sunk and we don't want cables or struts anywhere to tangle, knot, develop cracks, or snap on becoming fatigued after a few million cycles of charging and discharging.
Some of these attributes are tricky, especially when dealing with depth of kilometres under the sea. For example, deep-sea floats are dangerous toys. An inflated bag is fine near the surface, but a kilometre down...
But at our disposal we now have materials technology that even in my childhood would have been seen as miraculous. We have industrially available polymers ranging from densities well above that of salt water, to just below. Polypropylene and polybutylene PB-1 for example, remain slightly buoyant at all oceanic depths and pressures, and can be fabricated to resist enormous wear and tension. 
The best material for the ballast weights might be lead slugs in dense plastic envelopes; solid, oriented polyester capsules might be especially suitable, strong enough for the tension, non-buoyant, and resistant to water leaching inwards or lead leaching out.
From the ecological point of view, it is interesting to compare such tent farms with oil platforms. Oil platforms originally met with hysterical resistance because of the ecological harm they would do. Eventually the hysteria died down because the oil companies promised to be good and to remove them once they had outlived their usefulness. And when the companies eventually did prepare to live up to their promises, there were hurried negotiations to prevent their doing so. Even during their active lives the platforms had caused little harm if any, and once they stood idle they were a pure ecological boon. 
In this respect the tent farms should differ only in that they would involve no significant ecological harm from scratch and that their useful lives apart from maintenance, should be indefinite.

One alternative to the submarine storage of compressed gas as a means of capturing energy, would for practical purposes avoid the problem of adiabatic temperature changes. It is uncertain however, how practical the expedient would be. It would be a less concentrated means of energy storage, it certainly would be generally more expensive for the same scale of energy storage, and it would be unlikely to be work on anything like so large a scale. Still, it might be worth consideration for smaller scale installations under special circumstances. For one thing it should not require so robust a structure as the compressed air tents.
The scheme would be based on the use of a liquid in a vertically stacked series of tethered, submerged, flexible bags. They would contain a fluid comprising roughly half the total capacity of all the bags in the stack. Depending on the design, the liquid would either be as dense as possible or as buoyant as possible, relative to the density of the external water.
Unfortunately not many dense liquids are at once safe enough, cheap enough, and dense enough to be obviously attractive for such an application. Liquid sulphur dioxide sounds marginally attractive at a density of about 1.4, and it could easily be kept liquid at ambient temperatures in the ocean depths. However, huge volumes would be required, and sulphur dioxide is an unpleasant material to work with. Various concentrated salt solutions might work, but their densities are disappointing, seldom much above 1.1.
Conversely several liquefied gases or other low-density liquids have promisingly low densities. Liquefied petroleum gases have densities in the neighbourhood of 0.6. So does ammonia, but it is not much more attractive to work with than sulphur dioxide.
At all events the principle would be to separate a vertical stack of suitably designed bags with turbines or other mechanisms that could act as power generators when driven by a moving fluid, or as pumps to drive the fluid in the opposite direction. Conceivably the buoyancy of adjacent bag stacks could be neutralised by attaching heavy-fluid bags to light-fluid bags. I admit though, that the proposal strikes me as unpersuasive.
In the charging phase of the cycle the renewable energy source would either power pumps to drive the dense fluid to the top of the structure, or the light fluid to the bottom. In either case the evacuated part of the bags would be permitted to collapse under the external water pressure. In the energy consumption mode, either the heavy liquid would be permitted to flow down through the turbines, or the external water pressure would force the lighter fluid up through the turbines.
Other configurations with multiple fluids of different densities might be of interest, but it is not worth debating their relative merits here. Their advantages over the adiabatic temperature excursions of compressed gases are unlikely to offset the costs of the working fluid, or the potentially almost limitless scale of the submarine compressed air tents. Their main interest here is to demonstrate the conceptual range of approaches to the use of submarine potential energy as a storage medium.

Objections include the risks of tsunamis or underwater landslides. The tsunami risk is laughable because such slender structures on flat sea floor would not couple significantly to tsunami wave motion in the open sea. Underwater landslides should be no problem at reasonably chosen sites. For the most part this means keeping a reasonable distance from volcanic regions, steep shores, and continental shelves. That leaves more than enough sea room for tent farms to serve the entire planet several times over.
Other objections include distance from land and ecological effects. The distance should be nothing to be worried about. Compared to the price of overland electricity transmission, a fixed line to a shore terminal should be a modest overhead. Whether or when it would be advantageous to transmit compressed air instead of electricity, I cannot yet say. Personally I suspect that electricity transmission would in general be preferable.
Ecologically such stack "farms" should be pure profit. The water surface above where the tents were tethered would be where the wave-power, solar and wind-power units would be sited. Such installations would amount to de facto nature reserves or conservation areas, with no fishing allowed, and no shipping routes within tens of km. Their shallow zones should be excellent natural fish nurseries and feeding grounds. Very likely in their operation they could, as a bonus, raise nutritious ooze from the sea bed, encouraging at once the growth of photosynthetic organisms, scavengers, and filter feeders, together with the fish that prey on all those.
The tent stacks themselves would have hardly any biological footprint; they would neither cover, nor as a rule even touch, the sea floor, nor present any hazard to pelagic life. There would be no reason for serious discharge of toxic wastes from any of the industrial activities.
There are adequate illustrative examples of apparently far less ecologically benign marine engineering that in practice turned out to constitute de facto marine reserves. This commonly has happened with drilling platforms, breakwaters, and many others. 
Conversely, there is the question of fouling organisms growing on the structures. Since they are static, this is not as serious a matter as it would be on the hulls of ocean-going vessels.  However, if growths were to become troublesome, it might well be prudent to dedicate purpose-built robots to scavenge and scrape fouling organisms off each other as well as off ropes, pipes and the like on a fixed schedule.  
Although adiabatic effects might be troublesome in the tents and ducts, they could largely be controlled. For one thing, the only place where there should be really gross cooling, would be at the point where the compressed gas gets discharged into the turbines or the like near the surface. Adiabatic heating would also be predominantly near the surface. By intelligent use of insulation, ambient heat, and heat storage buffers, the adiabatic cooling could be nullified or even used as an energy source.
Note that such a store would have special virtues from the point of view of the engineer. The difference between the delivery pressure of a full store and a nearly empty compressed air store would be trivial, typically of the order of a few percent. This not only is a great luxury for the designer, but should make for long-lived power equipment running continuously at near-optimal load. Also, unlike non-hydrostatic compressed-air stores, these stores would experience hardly any stress when being filled or emptied and without penalty they could be nearly completely emptied as often as desired in routine operation.

Not that it should be necessary to repeat this reminder, but anyone about to criticise such schemes because they are not prime sources of energy, should remember that compressed air reservoirs are neither primary energy sources, nor contemplated as such. They would be used as energy stores and conditioners of energy accumulated from intermittent, too-abrupt, or too-diffuse resources such as wind or wave energy. Prime sources of energy are the subject of other essays. These installations also could store energy from off-peak generators,  permitting them to run at optimal load at nearly all times, thereby shaving peak loads and reducing base load infrastructure.
The versatility of such compressed-air energy stores would be formidable as compared to say purely electrical generators. Presumably the stored energy normally would be used for electricity generation, but an example of where it could be used efficiently and directly, could be in the launching of spacecraft.
Given such a source of compressed air, the shuttle launcher would have been unnecessary. Imagine launching a spacecraft in what amounted to a giant airgun, starting from fairly deep down. If we could launch the entire craft at high subsonic vertical velocities, this would replace the least efficient, most energy-hungry phase of the launch. It would remove the requirement for the first stage, or at the least, reduce the size of the first stage. The remaining part of the craft would amount to the second or third stages of the conventional rocket craft, at a fraction of the pollution and cost of the current modes of launching.
One never knows; such an installation might make the likes of space-based solar energy practical.

The sources of energy for accumulation could be anything available. I think it should at the least include floating fields, hundreds of kilometres square, equipped for the combined generation of wave, wind and solar energy. These are technologies that should not interfere with each other (nor indeed with large-scale marine aquaculture or ecology). All three could be harnessed for compression of air into ducts that feed the submarine reservoirs.
At the same time, the sources of energy need not be limited to such devices. I am firmly of the opinion that the idea of putting solar power generation satellites into geosynchronous orbit is completely perverse. The right place for anything of the type would be orbits as low as practical. A string of satellites sharing almost the same low near-polar to tropical orbit, precessing at a rate that keeps the satellite in the sun all year round, could beam down visible light or microwaves or both. Precession of about one degree per day would be more than enough. Precession at such a rate probably would require some active steering, but whether steering would demand say, an ion drive, trailing electric cables, or whether light sails used primarily for energy harvesting should suffice, I couldn't guess.
The satellite energy could be picked up mainly by a chain of floating oceanic stations with steerable antennae, thereby permitting 24-hour power generation world-wide. There would of course be land-based stations as well. Whether land- or water-based, such installations should be combined with installations of other forms of renewable power generation and storage, none of which could rival compressed air for stability, quality and cleanliness.

Probably the greatest single objection and obstacle to the full scale exploitation of the major renewable sources of energy is the problem of energy storage. To this day the storage problem is parroted by bigots and journalists as if it were simple and adequate proof that renewable sources simply could never solve any real-life problems on a national or international scale.
In reality one should not lose sight of the fact that storage is a serious problem in the conventional energy industry as well. If we could solve it, the problem of making allowance for peak demand would practically disappear, meaning that practically any power station could be scaled to the base load rather than to peaks, and often to reduced effective base load at that.
More practical engineers have been researching the prospects for energy sources and making a mockery of such objections by producing several kinds of storage systems, batteries, pumped storage, heat storage, and several others. Some are beginning to look impressive, in spite of tedious canards to the contrary. Not only do such technologies bid fair to solve the existing problems, but to offer new modes of energy usage.
Compressed air offers a particularly promising range of small- to very-large-scale systems that lend themselves to a very wide range of power generation technologies. Some seem promising for transport, some for inland use, but the greatest scale of all emerges from contemplation of the idea of compressed-air storage on a very large scale under deep water.


  1. I agree with your thesis, compressed gas storage has been undervalued and underexploited, especially in conjunction with heat engines'

  2. Thank you John. I suspect though, that you and I are in the minority. I have had very little response on the topic, mainly negative. I find this curious, because of the importance and value of energy storage, even using conventional energy sources.