Need to do Expected value calculations on Nuke plants
 

It seems to me that we now have enough occurances and the cost of these nuclear accidents to start factoring in the probability of an incident and its cost into the total financial picture when making decisions on Nuke power plants.

Way back they would just say a plant will withstand the impact of a small airliner or it will withstand a level 7.7 quake , etc. Well now we have 3 mile Island , Chernobyl and Japan incidents. We can now generate probability of an accident per reactor per year , etc. Example , lets say based on the 3 incidents mentioned and their years of operation and the total nukes in the world that we can predict a 1% incident rate over the lifetime of a reactor. Lets say an incident costs 100 billion dollars. Then we need to include .01 X 100 billion = 1 billion into the captalization of a new plant. How does an extra billion up front impact the decision of Nuke vs oil vs coal vs wind power , etc.

Anyway , back in the 70's we had no real data to support prdeictions but I think now we do have the data to start putting a number into the financial calculations used to justify a nuke.

Now don't get me wrong , I have always been pro nuke in the past. I am just starting to wonder if its worth it or not , now that they have a concrete record of some level of failure rate.

If you're really interested, compare the environmental impact, loss of life, energy costs and generated energy per acre of things like oil, coal, NG and you'll see that nuclear is still a winner by far.

Don't get me wrong, there is definitely a very long-term effect when there are incidents but 3-4 incidents out of the hundreds of plants over 30 years is far from a convincing argument - especially considering what current designs require as safety measures. Taking the environmental aspects out - coal, oil and NG won't be here forever.

And consider that TMI and Chernobyl were major human errors whereas Japan's problem was a quake 100 times greater than thought possible there FOLLOWED by a Tsunami 30% higher (and far more powerful) than expected. Two items deisnged for but independently of the other (also consider that the Japanese Nuke engineers didn't have lights in the control room until 2 days ago but I digress).

Modern systems are supposed to be much better and designed to automatically shut themselves down in blackout conditions where there is no power to run the systems - not the same as these designed 40 years ago (TMI and Fuk). Chernobyl design is supposed to be the worst plant design you could ever do, even by Soviet standards. A nuclear dirty bomb WITHOUT any kind of containment.

Fuk's problem was from what I read a cascade of worst case scenarios taking place one after another;

• Quake hits, Reactors shutdown by design (takes months to cool all the way down but a long as cooled they cannot go critical). They would have failed over to regional power supplied by one of several local non nuke plants nearby but those were damaged so much as to be offline. Redundant system 1 failed.
• The system then failed over as designed to backup diesel generators (BDGs). Redundant system 2.
• Tsunami wipes out BDGs 30 minutes later, system fails over to Redundant system 3 battery backup with 8 hour plus capacities even with extensive facility damage (look at the before / after pics).
• Battery backups fail after a day. Redundant system 3 failed.
• BDGs restarted (Redundant system 2) but fuel is fouled by Tsunami damage (our BDG fuel storage is supposedly very difficult to damage and also has multiple ways to route to system - redundancy to redundancy). Redundant system 3 failed.
• Explosions of venting Hydrogen from reactor vessels into unventilated area (supposedly our plants of this design were back fit many years ago to allow venting to occur and not have explosive gases contained in outer building).
• Additional BDGs flown in but either incompatible by design AND/OR connections not accessible within buildings damaged by explosions. Well beyond redundant systems now and trying to MacGiver 8 reactor plants.

It is amazing they have not had worse failures to date since the quake.

Also from what I have read, the recent designs are set to naturally cool in blackout conditions like what hit the Fuk plants multiple times. So while they will have Redundant system on Redundant system, the design of the plant will from the get go cool itself even without power.

My only point is that we now have a concrete track record that can be used to predict the frequency with which actual accidents happen. Back in 1970 it was all speculative. Just as they use 100 year or 200 year flood plains to predict problems and adjust the cost of insurances , etc, they now have concrete numbers to use when looking at Nukes. Yes all the things at the Japan site were unique and to piggy back one unique event on another it improbable , we now know for sure it has some finite probability and that can be used in thinking about future builds.

People can say new designs or unlikely natural events will or will not hurt or help but we no longer have to totally speculate. We have solid historical evidence that should be used in future calculations. The money side is always used to justify or unjustify certain technology applications such as the high initial investment of solar. I just think we need to do total financial analysis on all the various options going forward.

One other point to make is the environmental impact brought up of oil , solar , nuke coal , wind , etc. Yes they all have an environmental price and they all have a potential cost in lives lost or other significant health issues. Many of the deaths from the nukes will be 20 years in the making before they surface. The same is true for coal and oil fired plants. These long term effects are often discounted while the chance of a plane hitting a windmill tower are often over stated because their results are immediate. I think that while the recent Nuke prolems in Japan may not totally damn future Nuke reactors , if people fairly include these now measurable finite probabilities of disasters at these plants , the case for energy ideas like wind and solar cannot help to be looked at as more inviting options than they were before the recent nuke issues. Not more inviting than nukes specifically , just more inviting overall in light of the recent disasters that show nuke power plants do have finite probabilities of big problems from time to time.

If they didn't before, they will definitely start factoring in "compounding disasters" in the construction of nuclear facilities.

I doubt anyone thought that they would get a 1-2 punch when designing the reactor. Heck, unless you have had prioir experience with similar "domino effect" scenarios, you don't plan for them.
Here in New England, particularly along the coast, we deal with :compounded situations" every year: We get a severe winter storm and high tides to give us heavy snow and coastal flooding.
Although that can't compare to an earthquake and sunami, it gives you an idea of how confounding and complicated Japan's scenaio really is.

I think a lot of the arguments against nuclear are more sensationalistic than factual. I don't mean to nitpick, but I would say 3 incidents over 30 years with hundreds of nuclear power plants in 40-plus countries is "solid historical evidence" for how *reliable* and safe nuclear power is compared to other alternatives.

I used this analogy the other day:
When a plane crashes, everyone on the plane dies. However, it's still statistically safer to board a plane and fly across the country than it is to get in your car to go buy milk.

If the wind in japan starts blowing south , 20 million people in Tokyo could be in serious danger. A plane crashes and 400 die. 400 hundred is not good by any means but compared to the possible 20 million plus , the nuke single incident has to be considered "big".

A nuclear incident is absolutely considered big, but like I said, it's still statistically safer than all other options that generate a similar amount of power.

By statisically you are using 3 incidents over 30 years in x number of facilities?

I'm basing it off of a number of articles I've read comparing the different methods of generating electricity. There have been a number of articles in the last few weeks detailing the safety, environmental and cost differences between nuclear and oil/coal/NG. If you're interested, I'll try to find them again.

Not statistically valid, given the different deigns of the plants. Its like calculating a death rate in auto accidents for sub-compact cars and then applying it to 18 wheelers.

If they can put a nuclear plant on a ship and power a small city of America's best then I think we have the ablility build safe nuclear land plants