One of Helen Caldicott’s favourite quotes is that “in essence, a nuclear reactor is just a very sophisticated and dangerous way to boil water”. And - leaving aside the question of danger for a moment - she’s right. Every nuclear power station, and virtually every coal-fired power station, currently in operation does just that. In the process, huge amounts of water are used, and much of it goes straight up the cooling tower. But it’s not often appreciated why this is so. So, in the hope of explaining some of the science and engineering behind turning heat into electricity, I bring you a quick primer on the heat engine.
The Aeolipile. Source: Wikipedia.
Since the 19th century, electricity has been produced by spinning mechanical devices called generators. They’re marvels of efficiency. Roughly 95% of the mechanical energy you put into a generator is turned into electricity. But what do you use to turn your generator? If you’ve got some falling water nearby, you can use it. High, well-watered mountains quickly became a source of electricity that remains amongst the world’s cheapest. But if you don’t have that, the next most convenient energy sources (at least until recently) were fossil fuels, whose chemical energy is released as heat. As we all know well, Australia has vast amounts of conveniently located coal, just waiting to be dug up. But how do you get heat to spin your generator?
The first mechanical device to convert heat into energy was the Aeolipile, apparently invented by Hero of Alexandria back in the first century AD. It’s a pretty simple device, really. You get steam into a chamber with some narrow holes either side. As the steam escapes, it causes the chamber to rotate. A perfect demonstration of Newton’s laws of motion, except that we had to wait another millennia and a half for Newton to invent them. And the principles haven’t changed much to this day.
It wasn’t until the 17th century that heat engines became widespread, though it was a rather different design than the Aeolipile. A British gentleman called Thomas Newcomen invented a steam engine that could be used to pump water out of coal mines. As you can see by the natty animation (from Wikipedia), hot steam pushed up a piston. The cylinder holding the piston was then cooled down through the release of cool water, which created a vacuum pushing the piston back down again. The action of the piston moving up and down provided the useful work - in this case, pumping water out of the mine!
Newcomen’s engine worked pretty well for this purpose, except for one thing; it required an enormous amount of coal to do its job. That’s fine if you’re operating a coal mine, but not if you need to bring the coal from elsewhere. That’s where James Watt came in. He figured out that the Newcomen engine wasted a lot of steam, through the process of cooling the cylinder down. Watt made a separate condensing chamber that was permanently water-cooled, connected by a pipe to the main cylinder. When the time came, a valve was opened to connect the two; the coolness of the condensing chamber sucked the hot steam out of the main cylinder, without wasting energy cooling the cylinder itself. The Watt engine required two-thirds less steam than the Newcomen engine to do the same work. In modern terminology, its thermal efficiency was improved by a factor of three. This led to its wide adoption across England, and ultimately to the steam locomotive and the Second Industrial Revolution (and to the economic observation known as the Jevons Paradox, but there’s another story).
Meanwhile, a French military engineer, Nicolas Carnot, examined the theoretical basis of steam engines, and made a fascinating observation. He analysed a theoretical heat engine, known as a Carnot engine, and showed that no heat engine, whether it used steam or some other working fluid, could outperform his theoretical design. Any practical heat engine would be, in fact, less efficient.
His efficiency equation looked like this:
The key numbers in there are TC and TH, which correspond to the absolute temperature of your source of heat, and your source of cooling, and they’re expressed in degrees Kelvin (degrees Celsius above absolute zero - so add 273.15 degrees to whatever the temperature is in Celsius). You don’t generally have much control over TC - it’s the ambient air temperature, or water temperature if you’re using water cooling. But the implication of Carnot’s calculation is that if TH can be increased, your heat engine will be more efficient - that is, more energy will be turned into useful work, and less will be wasted.
There was a big gap between Carnot’s theoretical limits and the Watt engine; without even being aware of it, many steam engine inventors made improvements that closed the gap. In the late 19th century, the reciprocating steam engine was replaced with steam turbines, and innumerable detail refinements have been made since to extract every last bit of energy from the steam available. But, while we haven’t reached the Carnot maximum, the gap between the best turbines and the Carnot maximum are due to things we can’t easily change. So, we’re back to Carnot’s insights - if you want to improve the efficiency of your heat engine, increase the temperature at which it operates.
That fact explains much of what’s going on in the world of engine design, big and small. Diesel engines emit less CO2 than petrol engines because they’re more thermally efficient, in large part because they have a bigger effective temperature difference between the hot and cold reservoirs. The latest generation of coal-fired power stations don’t use steam at all - they use “supercritical water” that is so hot, and under so much pressure, that there isn’t any abrupt difference between liquid and gas. The extra temperature increases the thermal efficiency substantially. The difference in emissions intensity between coal and natural gas isn’t just because gas is a cleaner fuel; it’s because natural gas turbines run at a higher temperature than (older) steam turbines, and are therefore more thermally efficient. Combined-cycle plants use an extra trick to get better thermal efficiency; the water used to cool the gas turbines is then used to run a steam turbine, the extra step getting a bit closer to the Carnot maximum efficiency.
And there’s one other common characteristic of heat engines. No matter what the design - a petrol or diesel engine, a gas turbine, a steam turbine, or even a piston steam engine - more heat is lost to conduction, and through friction, in small engines than in large ones, all other things being equal. You can also run big engines a bit hotter than small ones. So big really is better, if you’re trying to convert an expensive, environmentally damaging resource into electricity, and you don’t have anything to do with the surplus heat.
Much of what you have said here also applies to concentrating solar thermal power stations (CSP), and geothermal for that matter) as these facilities work by heating oil with solar radiation, then using that heated oil (400c in one system and 275 in another) to boil water to run steam turbines. These running temperatures are well below the systems that you are refering too, but as the energy is relatively free and no emissions are made that is not as important.
The other thing to add to your post is that it is not necessary to lose water in any of these systems. The water loss comes from the cooling of the cooling water if water evaporation is used to improve the effectiveness of cooling towers. As you have pointed out in the past just air movement is sufficient if there is enough of it (larger cooling towers for the same generation capacity and more electricity to run the larger fans). I am told that the temperature of the steam in the final condensation stage of a steam turbine power plant is around 68 deg C. So as you say the higher the starting temperature the better it works.
Interesting post Robert; care to hazard a guess as to how the greater efficiencies of new generation coal fired power stations would compare to the microgen model descibed on catalyst a few weeks back (where small power stations - inside an office block say - used the otherwise wasted heat by capturing and piping it as heating or use it to power airconditioning)?
In other words, where should the money be invested?
It is probably worth making a comment about boilers as well. Burning fuel makes very hot gas. The energy in the gas is available to transfer to another medium to do work. Boiling water takes water at ambient temperature then raises it to 100deg C or much higher when the space over the water is contained to make a pressure vessel. It takes time to do this energy transfer, but hot gasses do not like to stick around. They need to be on the move (residency time and surface area). So boilers are generally arranged so that the initially combusted gasses come in contact with the hottest of the generated steam, then circulate to the coldest end of the boiling system as the combustion gasses cool keeping a uniform temperature gradient between the two mediums throughout. Simply put boiling water in a pot with a gass flame is very inefficient as most of the hot gasses flow past the pot surface with out transfering their heat because the gas residncy time is too short and the temperature gradient is too high. The best way to boil water with gas would be to start the flame on low then progressively increase the flame size (heat) as the water temperature increases. I believe.
RF: Theoretically, microgen can be very, very efficient - if you’ve got a good use for the waste heat, as I alluded to in my last sentence.
However, most of the work on microgeneration seems to be coming out of the UK, where you need heating for much more of the year - and even so, as I understand things it’s turning out to be substantially less efficient than hoped.
Even so, the life cycle carbon emissions of using natural gas in combination with microgeneration is still way worse than centralized nukes or renewables.
Furthermore, cost-effectiveness and thermal efficiency are not one and the same. Capital costs, and maintenance costs, have to come into the picture, and maintaining 1 big engine tends to be cheaper than maintaining 1000 small ones.
[cough] what about stirling engines [cough]
Liam: they’re heat engines too. Same deal applies; the greater the difference between your heat source and your cool source, the higher the efficiency.
But one thing you may be able to do with a nuclear power station is have a working fluid that gets really hot, using a Brayton (constant volume) rather than a Rankine cycle. Molten sodium metal used to be talked about, but of course handling large quantities of a super-hot, extremely chemically reactive, heavily irradiated fluid raised some interesting safety problems.
And you’re right about bigness=efficiency, all else equal. But that’s only in generation - distribution is opposite.
Interesting how the fundamental problems of long term Economic sustainability comes down to engineering solutions for the generation of power isn’t it? The Green party could use some engineers. Probably got a few layin’ about.
(Will what ‘er ye doin’ laddie. Git o’ yer arse and replace these ruhdikulass energee powlassies they g’t goin. Thar squanderin’ ‘n opportuniteh! Aye! Th’arrrgh. Jimmeh. )
Engineers are notorious for political conservatism however. It’s not necessary for them to be politically active to solve environmental problems. In fact it’d probably get in the way.
Enjoy your posts Robert.
Robert - thanks for the really interesting post.
Even if you don’t need to heat the offices, surely hot water would
be useful - maybe you’d need to distribute to nearby appartment buildings as well.
As an aside, I wonder how many office building there are out there in Australia where in winter they heat the rooms where the workers are and at the same time separately cool the rooms that the computers are in….
Robert
I think the size=efficiency argument is obvious on engineering grounds, but how do the inefficiencies in the distribution network affect the total outcome?
ie. let’s say the whole country has a single generator operating at greater total efficiency than could be achieved by a whole of of smaller distributed plants closer to their customers.
What are the trade offs in the losses in the distribution network in those two alternatives (sorry for asking, I don’t have a good handle on current figures but I remember that years ago transmission losses and inefficiencies were pretty large)
Good point on the distribution losses. There are 2 main types of microgenerators out there. The Capstone gas turbine microturbine is a 100 Kw generator set which produces a lot of heat and runs on gas. Now if you were going to produce that heat by burning gas then the turbine makes sense as it gives flexibility and the sale of power to the grid helps to offset capital cost. The other type is the Whispertech (NZ) sterling cycle power generator which is a 1Kw gas powered unit. This is suitable for domestic installation and is efficient for the same reasons under the same conditions. These units are popular in Northern UK and Europe in remote regions and for boats. They cost $10,000 plus, have a 40,000 hour running life and no maintenance.
Another type of microgenerator is a system used in skilodges where they install a piston engine generator set which provides heat and power. These get replaced every few years. Noisey and smelly.
None of these solutions do anything good for global warming.
JM: it’s not a simple relationship between size and distance, I know that much.
If you really need to send power a long way, you use high-voltage DC transmission. According to Wikipedia, typical line losses are under 3% per 1000 km. That gets you Perth-Adelaide with less than 10% line losses.
So are the whispertech bods telling porkies:
A definite porkie. Even for a thermodynamically ideal engine that would need a working temperature of about 3000 degrees C.
Aidan,
What Robert said is if transmitting power over long distances high voltage DC “can” be used. That is not what is necessarily used in pracitce for shorter distances, and it certainly is not what is used in the near distribution network as the power gets closer to the end user(accumulating losses with each stepdown from the 400,000volts to the 240volts for the end user).
What the Whispertech guys are saying, I think, is that their units produce electricity, possibly hot water (I haven’t looked) and space heating, all from the gas supply with the 10% loss being in the heat of the exhaust CO2 flue gas.
Well they are presumably counting the useful water heating they generate as well. But is the 40% number accurate as well?
40% is reasonable for single-cycle gas. Combined cycle is somewhere between 50-60%.
Robert,
Putting your thinking cap on, if you have a pipe containing a substance at, say, 120 deg c, this will be releasing infrared radiation at a certain rate. As infrared radiation can be focussed, shouldn’t it be possible to focus this infra red rediation on a smaller surface to produce a heated surface at a higher temperature than the original 120deg c. Is there some obvious flaw in this thinking?
Well, in principle you can concentrate infra-red radiation; that’s what concentrating solar power (both thermal and PV) does.
I doubt you could ever do so practically with low-intensity, artificial sources of heat. Too much would be lost through other sources of heat transfer, and the radiating source wouldn’t be directional enough to focus.
So the photons would be coming off the hot surface in angles in a spherical aura, so the trick would be to be able straighten them up to a common direction. Are you familiar with the science of the nano sized holes in a metal foil? I can’t remember what the feature was but I recall that this arrangement was able to capture photons and then release them from the foil on the other side in a special way.
So, BilB, when you said
were you correct? The whispertech guys are saying they are 90% efficient, and noone here seems to be gainsaying that. Sounds better than 60% (combined cycle power plant) - transmission efficiencies.
That supposes that you have something useful to do with the hot water, but that doesn’t sound too far fetched to me. All those Europeans with gas-fired hydronic heating can replace their conventional gas boiler with a whispertech and generate grid electricity when they are warming their house up. They’d probably use more gas than otherwise, but on the upside they’d offset their electricity use and overall make more efficient use of the CO2 producing fuel.
Isn’t this a good thing?
What I was saying is that the whispertech still uses CO2 emitting gas. And the extended advantages of the Whispertech are only relevent in colder climates where the overflow heat can be used for space heating (to get the full effect). The reason why I am especially interested in the Whispertec’s unit is that it is a sterling cycle engine. What that means is that this is a mechanism that does not consume fuel internally. It works on the flow of heat. ie it has a hot spot and a cold spot. There is a gas burner which projects a flame at the hot spot, internally the heat causes hydrogen to expand pushing a piston and then cool/contract in another part of the engine attached to the cold spot. The temperature gradient between the 2 spots governs it effectiveness and the power it converts to mechanical energy released in the form of electricity. So in principle this engine will work on any source of heat as long as the energy can be delivered to the hot spot effectively. This engine can be run on direct concentrated solar radiation, or it could be arranged to work from oil heated by the sun which is then delivered to the hot spot, or it could also work with heat from a log fire or burning biomass.
What is also special about this engine is that it is a sealed unit similar to your refrigerator motor. This means that it cannot accumulate contaminates thereby extending its operating life without maintenance.
Combined cycle(Gas turbine and steam turbine)can get over 70% Carnot efficiency in ideal conditions. The really good thing about gas is that it is a high hydrogen content fuel (Methane mostly)The CO2 emissions from methane drainage from coal beds are less than 50% those of coal burnt in boilers. Raises the question: why don’t the coal fired generation companies just close the mines and install methane drainage on the cola beds ? These are vast by the way and already piss raw methane into the atmosphere day and night - this is at least 10 times worse than CO2 as a GG.
Carnot also rules in fuel cells but you use entropy instead of temperature to calculate efficiency (simplistically. So the program for OZ should be to convert our entire coal generation to gas. Forget nuclear it is so last century.
The problem with fuel cells BTW is that they are an impedance limited energy source with really bad voltage compliance and totally run out of volts with even a small overload. So when you put your foot down to pass that semi-trailer you will take about an hour to do it. Splat.
Huggy
I used to drive an old and grossly underpowered Land-Rover, so I’m kind of used to that. Tractors and electric milk floats were about my fighting weight.
On a serious note, an excellent post. If more people were conversant with the findings of Monsieur Carnot and the Laws of Thermodynamics, the market for fraudulent miracle engines would shrink dramatically.
On the subject of microgeneration, I doubt if there is a one size fits all universal solution.
For instance, in a situation where a generator has access to plentiful, cheap, flammable plant waste matter, the use of an inefficient but relatively simple and rugged system involving either a relatively crude boiler and a reciprocating steam engine (where water supply is not restricted) or a Stirling engine, as mooted by Liam (where water is scarce and/or expensive) may represent a viable and responsible option. Elsewhere it might be ridiculous to even consider such an idea.
Intellectual Bogan
If you want to scare yourself google arctic methyl hydrate (essentially natural gas).
A great post, Robert.
I can’t help but think, every time I hear a mention of free energy machines or water powered cars or what ever, that we’d benefit greatly from a greater appreciation in the community of the laws of thermodynamics.
Some things which are consequences of the above, that I’d like to elaborate on, for everybody’s benefit:
The idea that nuclear energy is no good because it uses too much water or consumes lots of water is complete baloney.
The laws of thermodynamics are not prejudiced in any way in favour of or against nuclear fission as a heat source.
For any power plant - nuclear, gas, coal, geothermal or solar thermal, whatever - that is operating at a given heat source temperature and discharging energy into the environment with the heatsink at a given temperature, the rate of energy flow out of the heatsink is a fixed multiple of the plant’s electrical power output. A power plant requires the same cooling tower capacity or cooling water flow rate, for the same electricity output, irrespective of whether it’s nuclear, coal, geothermal or whatever, unless the heat source temperature of the plant can be increased.
Also - it’s worth mentioning cogeneration. Cogeneration basically just means that you’ve captured some of the waste heat from, say, a gas turbine power plant, and extracted more energy using a Rankine cycle (steam cycle) power plant, which means you’ve increased the overall efficiency of the process.
You might also be using the waste heat to heat water for heating a building, or something like that.
However, contrary to how it’s implied, say, in the writings of Amory Lovins, cogeneration is not a means of clean energy generation. It’s just a means of getting a bit of an efficiency improvement from a given source of energy. Most commonly, the energy source associated with cogeneration is a natural gas burning turbine - and burning methane in a gas turbine is in no way “clean, green” energy, even if you can increase the efficiency through cogeneration.
Of course, the idea of cogeneration can be applied equally to a range of thermal energy sources - coal, gas, or nuclear energy or a range of other things - especially, for example, if a high temperature nuclear power reactor driving a gas turbine is considered.