Rapaire, there *is* fusion research going on, funded internationally (including by the US) - JET in Switzerland has been going for years, and there's a new one in progress (can't remember the name offhand) which is intended to be the last step before commercial fusion. It can and will be done on Earth, almost certainly in my lifetime.

I suspect you mean fast-breeder reactors, and on that you're right. BNFL got majorly screwed over that. Their plan was that Britain became the international centre for fuel reprocessing, they built up a stash of plutonium (which was a "waste product" for uranium-based reactors) from reprocessing, and then used that plutonium as essentially "free" fuel for fast-breeder reactors. This would use up the plutonium (plutonium is *not* forever if you use it as fuel!) and convert it to shorter-halflife elements which solves an awful lot of Charlie's "hundreds of thousands of years" problem. But anti-nuclear protests prevented any significant building of fast-breeder reactors in the UK, so BNFL has been shafted with storing it basically ad infinitum.

Charlie, my dad worked as the environmental officer at BNFL's Salwick plant, so I've unavoidably picked up a fair amount of info. This doesn't mean I'm pro-nuclear, but it does mean I know a fair amount of the facts. I'm basically nuclear-agnostic, and fossil-fuel-agnostic and mega-dam-agnostic too - I'm not actively opposed to any of them on principle, but if we can find safer generation methods then I think we should.

On that front, I have to take issue with the "massive low-level radioactive contamination" from fuel processing. Is that "massive" as in high-level, or "massive" as in covering a wide area? And is this "contamination" significant compared to background radiation? or compared to radioactive particles from fossil-fuels? My dad had personal experience with this particular subject, when FoE produced a study showing the potential for radioactive particle concentration in the river Ribble. At some expense, BNFL got a study of the *actual* river behaviour, and found that not only was there no such concentration, the river behaviour was fundamentally different to what the initial study showed, indicating that the initial study's river model was so fatally flawed as to be worthless. FoE hadn't done any such checks.

Let's not dance around it though. Nuclear fission reactions *are* potentially dangerous, and so are the processing methods for producing fuel rods. What stops them being *actually* dangerous are the design of the reactor and the design of the processing plant, reliable automated control systems (including interlocks to prevent human errors), and human monitoring to ensure that everything's working properly.   Where accidents have occurred, it's always been a failure in one of those three elements: bad design, unreliable control systems or inadequate monitoring.

In that, they're no different to coal/oil/gas power, or dams, or chemical factories for that matter. Coal mines and oil/gas rigs are hugely dangerous, and those power stations can give off huge amounts of hazardous chemicals - check out Russia's industrial wastelands for example. You think Chernobyl was bad, then check out China's record on dams - a major failure in 1975 drowned 26,000 people, with another 145,000 dying from disease and famine in the aftermath. And the potential problems with chemical factories should need no explanation, given that just about everything in there is toxic, explosive or both - Bhopal is a good example here, or the many places in the US contaminated with dioxins.

To give an example from somewhere else, consider planes. The DC-10 was designed with its control cables running along a single slot which was vulnerable to damage. When this failed, the planes weren't controllable and crashed. At least one Airbus is known to have crashed because of failure of its electronics (most likely due to software). The number of plane crashes due to faulty radar, faulty landing-assist systems and simple pilot error is simply huge. We'd have to be crazy to fly - wouldn't we...? Well actually no, because the FAA and other organisations try to learn from what's gone wrong and ensure it doesn't happen again, and airplane manufacturers also try to learn from what's gone wrong before because their company is at stake.

But does this happen with nuclear? That's my main worry with nuclear is with the second and third points - control systems and monitoring. Reactor designs are very mature, and all nuclear plants in the West are designed to fail safe. But control systems are open to interference from management saying "we've got a deadline and it's good enough, so just do it" - Les Hatton quotes examples of this in his book on software reliability. They may also be open to sabotage or other active stupidity, such as with Chernobyl (which was not a failsafe design but relied on automated controls to keep it safe). And there have been several cases of monitoring being circumvented - one case involved X-rays of welds being falsified to save money.

This isn't unique to nuclear plants though - they just happen to be the highest-profile example of this general human problem. The Buncefield depot fire in Hemel Hempstead is one example. Or the many coal-mining disasters, most of which have turned out to be avoidable. Or zillions of chemical spillages, or homeowners being told to stay indoors because some dangerous gas has been released from their local ICI plant because of some screwup there. In all these cases, designs are mostly good because the potential problems with a chunk of metal or concrete are visible and civil engineering is a very mature area, but control systems are less visible, and human factors rely on individual people being consciencious *and* systems being in place to make sure they are. Basically, I'm trying to say that *all* heavy industry is a potential danger, and I think nuclear is no more or less a danger than other heavy industry.

Graham.