It is likely that at some point in the future, all energy will come from nuclear sources. How far way that is, I wouldn't like to say. However, I would hope that the type of nuclear energy we shall be using at that time will be something other than the uranium fission reactor which predominates today.
Almost all current reactor designs are based on research done in the 1940's Manhattan Project. The aim of that project was to create a nuclear weapon, and the processing of uranium ore was seen as the best approach, yielding as it does Uranium 235 and Plutonium 239, both suitable materials for weapons. When it was realised that nuclear processes could also be used commercially for electricity generation, the uranium route was followed almost as if it were a foregone conclusion that this also was the best approach for powerplants. Even as early as the 1950's some scientists were questioning that decision, stating that other approaches might be more suitable, but the financial backers decided to go with the 'safe bet' of uranium rather than risking their capital on other largely untested ideas.
The design of powerplant reactors then evolved through several phases, to settle on the 'big two' the PWR and the BWR. A few other more radical designs such as the UK-designed AGR are also in service, but by far the greatest number of powerstations employ uranium fuel with some form of pressurised water cooling.
As a brief rundown on the tech, pressurised water reactors and boiling water reactors operate on a similar principle, in which enriched uranium is reacted in a water-filled vessel, the water serving as both a neutron moderator and coolant. The difference is that in the BWR the water in the reactor itself is boiled to produce steam which drives the powerplant. In the PWR, the reactor water is under greater pressure and thus cannot boil. A heat exchanger transfers the heat from this into a secondary water circuit where boiling occurs, producing steam to drive equipment. From a safety point of view the two designs are not greatly different.
Some older designs such as the Chernobyl reactors rely on blocks of graphite surrounding the fuel rods as the neutron moderator, but still employ water as coolant. The advantage of that design is that graphite being a more effective moderator, the fuel need not be so highly enriched. The disadvantage, put simply, is that graphite has some rather worrying modes of unstable behaviour when in a nuclear environment, and these instabilities demand greater care in ensuring correct operating procedures. Thus, graphite moderators have largely been phased-out from modern designs.
As we all know, there have been a number of relatively serious accidents with uranium reactors. At least two have resulted in major releases of pollution to the environment. The two most serious such incidents have proven wrong the experts' predictions that the worst which could happen to an out-of-control reactor is a meltdown. The worst which can happen is, in fact, a substantial explosion. What is less often realised is all but one or two of these accidents were primarily down-to chemical causes rather than nuclear ones. That may seem an astounding statement, so to see why let's take a look at the issues involved:
Uranium is a metal which, like aluminium, is quite reactive in air but on which an impermeable oxide coating tends to accumulate, passivating the metal against further corrosion. It does not ignite easily, but will continue to burn of its own accord once ignited. This flammability was the key factor in the Windscale reactor fire, which would have resulted in serious pollution had the chimneys not been fitted with particulate filters.
The answer to this one is to use uranium oxide as reactor fuel. The oxide, being noninflammable, does not present this risk. Almost all modern reactors use oxide fuel. So, basically, solved.
A disadvantage of uranium oxide is that it is not a particularly strong material. Thus, unlike uranium metal which has a similar strength to steel, it requires a container of some sort to give it structural strength.
Fuel rods or pins, acting as containers for uranium oxide pellets, have to be constructed of a material with good strength and corrosion resistance, but which does not hinder the passage of neutrons into and out of the fuel. Only a few materials are able to meet all of these requirements. One that does is an alloy of the metal zirconium.
Zirconium was the metal 'wool' in old-fashioned camera flashbulbs, of the one-use type that were the standard before electronic flash came along. Its properties are similar to those of magnesium. Like magnesium it is not easy to ignite, but once ignited it burns fiercely with a bright white flame. Moreover, once ignited it is extremely hard to extinguish. Pouring water on a zirconium fire will only result in the zirconium continuing to burn underwater! This happens because the burning metal is so keen to 'grab' oxygen from any available source that it will strip the oxygen component out of water, leaving the hydrogen to escape as gas.
Thus, a zirconium fire is not only hard to extinguish, but if water or steam is in contact with the burning metal it will result in a buildup of hydrogen gas. Hydrogen gas is of course itself inflammable, and not only that, it forms an explosive mixture with air over a very wide range of mixing ratios. Thus, the consequence of a fuel-cladding fire in a confined space such as a reactor containment vessel will likely be an eventual hydrogen explosion. Putting water on the fire will only hasten that event. The force of such an explosion is easily sufficient to rupture the strongest primary containment vessel.
There is currently no off-the-shelf answer to this zirconium fire risk where PWR/BWR reactors are concerned. The only proper solution would seem to be to investigate other reactor designs.
In open air, water boils at too low a temperature to allow efficient operation of a heat engine such as a turbine generating set. To reach higher temperatures without steam voids forming inside the reactor, the cooling water must be held under very substantial pressure. Any loss of pressure will result in an energetic steam escape which is of course a hazard in itself, but more importantly will leave part of the reactor core with no cooling.
A secondary concern here is that the cooling water must be constantly circulated to keep temperatures under control. A failure of the cooling pumps will lead to water boiling off, steam voids forming, and a zirconium fire. What we are dealing with here, is basically just car radiator technology on a grand scale. The difference is that if your car radiator boils over you can turn the engine off and wait for things to cool. Even when a uranium reactor is fully shut down it will still take hours or days for residual heat output to cease, therefore a sudden and total loss of cooling is a very serious contingency.
Not strictly a safety issue, but one of the key limitations of current uranium reactor designs is that only a few percent of the available energy in the fuel is extracted. The fuel then becomes too inactive to maintain the reaction, and has to be replaced. This has two consequences. World supplies of uranium are finite, and with so dreadfully inefficient an extraction of energy those supplies may well run out even before supplies of fossil fuels such as coal are depleted. A more efficient reactor which extracted most of the available energy would extend the life of our reserves to many thousands of years, but at present they may only last a century if we're lucky. This is the basis of the statement by Greenpeace that nuclear (uranium) energy 'is not sustainable energy' because its lifetime is too limited. Basically, current reactor designs are 'gas guzzler' technology that show no concern whatsoever for efficient or clean use of the fuel.
This partial combustion of fuel also has the consequence of creating large amounts of high-level nuclear waste, for which no completely safe disposal route currently exists. More efficient combustion of the fuel would generate far less such waste, or possibly none at all.
When we look at what actually happened at Chernobyl and Fukushima, we see that although the reactor designs differed, as did also the circumstances which led to the accidents, what followed was of a very similar nature. At Chernobyl a surge in reactor output due to mis-operation caused severe damage. At Fukushima, flooding knocked-out the coolant pumps. In both cases, cooling failure and ensuing zirconium-in-steam fires released hydrogen gas which blew the reactor buildings apart and scattered radioactive pollution around the site.
The lesson to be learned here, which the nuclear industry is understandably reluctant to acknowledge, is that not the nuclear fuel itself, but the presence of chemically inflammable materials in the reactor is the Achilles' heel of nuclear safety. I think it's likely that this chemical fire/explosion risk factor was uppermost in German politicians' minds when they decided to close down some of their nuclear stations following Fukushima.
Talking of consequences, you will see some nuclear proponents trying to play-down the seriousness of these two incidents. Don't believe them! If you take the trouble to look-up some of the Russian sites covering the actual events at Chernobyl and the environmental aftermath, you will see that not only was it extremely serious, but that it was only through the swift action and dedication of the disaster-control teams that it was not far more serious. As a worst-case scenario it was reckoned that a further hydrogen explosion might have evacuated the entire core, in which case a substantial part of Europe would have suffered hazardous levels of fallout.
So, the key lesson to be learned from these incidents is that any reactor with chemically inflammable materials in its core is a pollution catastrophe just waiting to happen. Unfortunately, this includes almost all of the 400-or-so nuclear powerplants in the world today.
In spite of these issues I do believe that nuclear power can be safe, but to achieve that we need to completely rethink reactor design. Paradoxically It isn't the nuclear fuel that's the primary risk of catastrophic failure, that is down-to the use of zirconium fuel cladding and pressurised water coolant. Eliminate those and the main remaining risk is a meltdown - White-hot radioactive lava flowing out of a failed reactor, to collect in a sump below the site. The end of the reactor for sure, but with proper design even this need not be an environmental catastrophe. Though, any environmentally-friendly reactor will need to be a lot more efficient in its use of fuel, both to conserve stocks and to reduce waste.
It's worth reflecting on the fact that one of the reasons for the extreme cost and long construction time of EDF's Hinkley Point proposal is the sheer extent of the proposed safety features. Safety features necessitated by the use of a zirconium cladding / water coolant combination. For me, this whole approach doesn't inspire confidence anyway. Better to address the actual hazards at source than to take the added safety-feature approach.
So. Having detailed the thinking which says that the current crop of uranium-fuelled nuclear power stations are a singularly bad idea, and why we shouldn't build any more of that kind, in the following sections we'll look at what the alternatives are.