Heat engines require cooling, to turn heat energy into mechanical energy (and then, via a turbine-connected generator, to electrical energy). This is an unavoidable physical principle, and is typically exploited via the Carnot cycle. Usually, this cooling requirement uses water.
Why do I raise this point? Because it seems to be a source of much confusion (innocent and deliberate) amongst the energy illiterate, especially when mounted as an argument against nuclear energy generation (and, implicitly, as a reason for adopting renewable energy). For instance, Friends of the Earth have decried:
Nuclear power plants consume large amounts of water –35-65 million litres daily. Indeed nuclear power is the thirstiest of all energy sources. A December 2006 report by the Commonwealth Department of Parliamentary Services states: “Per megawatt existing nuclear power stations use and consume more water than power stations using other fuel sources. Depending on the cooling technology utilised, the water requirements for a nuclear power station can vary between 20 to 83 per cent more than for other power stations.” Global warming and water shortages are likely to exacerbate problems experienced by the nuclear power industry during heatwaves in recent years. Nuclear power plants in several countries, including France and the US, have had to operate at reduced capacity, or to shut down temporarily, because of reduced water supply or to avoid breaching regulations limiting the heat of expelled water.
So what’s the story? Are water limitations and discharge regulations destined to be a major limiting factor for nuclear power, especially for places that are experiences increasing water shortages, such as Australia? The short answer is no — this is classic FUD. For the longer answer, read on.
All thermal power plants, by definition, make use of heat engines with heat exchangers, and so require cooling (although this need can be reduced in various ways, as explained below). This includes coal-fired, nuclear fission, oil-fired, conventional gas-fired, solar thermal and geothermal power stations. The renewable energy sources that don’t have this cooling requirement are hydropower, wind, wave, tidal and solar photovoltaic power.
Water is used in two ways in thermal power plants: (a) Internal steam cycle: to create steam via the energy source (fossil fuel combustion, fission chain reaction, heat exchange with deep rocks [hot dry rock geothermal] or a heat transfer fluid [concentrating solar power]) and convey it to an electricity-generating turbine, and (b) Cooling cycle: to cool and condense the after-turbine steam (this condensation dramatically decreases the volume of the expanded steam,creating a suction vacuum which draws it through the turbine blades), and then to discharge surplus heat to the environment.
In the internal energy transfer water circuit, only a comparatively tiny amount of water is lost (this is the case for nuclear, coal, conventional [rankine cycle] gas, etc). It’s effectively a water –> steam –> water –> steam –> water etc. closed-loop system. In a pressurised water reactor (see diagram), there are two steam loops, only one which is exposed to the nuclear core. In a boiling water reactor, there is a single internal water circuit. Clearly, this cannot be what worries people, as the water consumption for an internal steam cycle is essentially a once-off affair.
So let’s look at (2), the cooling cycle. The amount of heat discharged to the environment depends principally on the plant’s thermal efficiency. High efficiency is achieved by having a large temperature differential, whether it comes from high internal heat or a low temperature external environment, or both. The thermal efficiency of today’s nuclear power plants are around 35%, whereas hotter coal burners can reach 40% or more. Advanced high temperature gas- or molten-salt-cooled nuclear reactors (AHTR) can reach efficiencies as high as 70%. Bottom line: higher thermal efficiency = lower water usage. In this context, hot dry rock geothermal, which operates at fairly low outlet temperatures of 150 — 250 °C (compared to 300 — 550°C for nuclear and 850°C for AHTRs), doesn’t look that great.
The options for wet cooling are once-through (direct), or recirculating. Once-through uses water from a large body — the ocean, a big lake, or a high-flow river — to bring in water and then reject roughly the same amount after cooling, which is a few degrees warmer; there is little net loss. To use the recirculating method, water is drawn from some available source (e.g. a river) and then water is sprayed down hyperboloid-shaped cooling towers, which exploit the physics of evaporation, in a natural chimney draft, to cool the water. Using the recirculating method, roughly 2.5 litres of water are lost to evaporation per kWh of electricity generated. A typical 1 GWe plant operating uses about 75 megalitres per day (25 Olympic-sized swimming pools).
Some countries only use once-through cooling using only sea water (which is in infinite supply) — UK, Sweden, Finland, Japan, Korea, China, etc. Canada uses water from the Great Lakes. In the US and France, coastal plants use sea water and a large number of inland reactors use cooling towers or once-through river/lake sources.
For Australia, this raises an important point. Our coal-fired power stations are clustered in regions such as the Latrobe and Hunter Valleys. Why? Because that way, they’re located right on top of the coal seams. When you have to feed 4 million tonnes of this black rock into a 1 GWe plant each year, it makes a lot of sense to avoid te need to move hundreds of thousands of loaded rail cars across the country, and instead to put your electricity generation plants where the fuel is (if you can — in most of the US, they can’t). The caveat is that you must use the water that can be piped to these locations — fresh water — for recirculating cooling. However, if Australia replaced all of its coal plants with nuclear reactors, it could save all that valuable fresh water. Why? Because their geographical location is unconstrained by the fuel supply, since an utterly trivial 25 tonnes of fuel must be supplied to a light water reactor, or just 1 tonne for a fast spectrum or thorium reactor. As such, all of the nuclear power plants could be built along the coastline and cooled by sea water.
An alternative, for any thermal plant, is dry cooling, whereby heat is transferred directly to the air via high-flow forced drafts (using industrial-sized fans, finned radiator pipes etc.). This is a less efficient method than wet cooling, because the cooling fans consume considerable power and the temperature differential that’s established is necessarily smaller. Yet, it may end up being the only feasible option for large-scale desert-based solar thermal power. The sunny and dry desert is one place where water scarcity really bites. For instance, a German solar developer, Solar Millennium, has reluctantly decided to use a dry-cooling method for their two 250 MWe (peak) CSP plants, after the nearby residents feared their aquifers would be sucked dry by the use of 1.3 billion gallons of cooling water per annum. The inefficiencies created by air cooling will increase the size of the mirror fields required to yield a given amount of power. For some interesting further reading on this proposed solar thermal project (with an ecological impacts slant), see here.
Naturally, if water is a real limitation in a given area that requires electricity, then what’s good for the solar thermal goose is good for the nuclear gander — nuclear power can use air cooling too, if necessary. Or, in most cases, you take the win-win option of saving inland fresh water by closing down coal plants and building your nuclear plants by the sea. I guess Friends of the Earth didn’t think of these points — or, perhaps, they just chose not to mention them. But at least now, having read this TCASE post, you’ll not be tricked by this anti-intellectual sleight-of-hand.
For further information, I can thoroughly recommend that you read this from the WNA: Is the Cooling of Power Plants a Constraint on the Future of Nuclear Power?