My point was that a nuke at idle costs about the same amount to operate as a nuke producing at capacity. It is all about spreading fixed costs over units of production. No production -- infinitely high unit costs -- production at capacity essentially the same total costs spread out over design capacity of the plant.

As to why liability relief is needed, as a libertarian, my basic belief is that that Government shouldn't be bestowing relief on one class of producers. However, the counter to that is the fact that there is simply no way to commercially insure the risk involved with a serious accident at a nuke on the open market. No responsible party will write these policies as the theory of insurance requires some predictability of occurence [or at least a basis for handicapping the odds] and the more critically the financial capability on the part of the issuer to pay off. Both are lacking in the case of nukes.

Without effective risk management to place limits on liability, a company operating nukes will be seen as [sorry -- pun intended] radioactive.

In in commercial sense, a major nuke accident represents the unthinkable. In a broader sense, the Russians had the unthinkable occur and for the most part life goes on.

It all comes down to the responsibility of the board of directors to manage risk and avoid whenever possible optional, but potentially fatal risks to the corporation. The board's duty is to the shareholders and not to the greater good -- whatever that might be.

IMO, [and from what I surmise to be Alan's perspective] the debate really is over whether we want to burn more coal; compel behavior changes --- probably at the point of a gun; simply let the SHTF as the grid collapses from overloads; go to progressively longer rolling black outs -- same end result as the other grid collapse scenario although initially the process of break down of society might be less dramatic; or build nukes.

BTW, I am positive on solar [and maybe wind] coupled with better storage technologies over the long haul, but nukes appear to be the best approach to getting through what I expect will be a nasty transition.

I don't know whether you consider the above to be an adequate explanation, but that is most of the story in few paragraphs. An even more abbreviated version might be: Generate or degenerate.

There are quite a few inches of procedures to be followed to stop & restart a nuke, but it is NOT unsafe.

Reactionary is correct, economics (CHEAP fuel, expensive capital) make 100% load the preferred mode.  Occasionally, when the power source with the cheapest fuel (hydro) has a surplus of fuel and no place to store the excess water, they will slow down WHOOPS 2 (not enough distribution lines to take all available hydro + nuke out of area).  The French also throttle back a handful of their nukes as night (AFAIK).

To restart a steam NG plant from cols takes the equilavent of 30 minutes fuel at 100% load just to get it hot.  Coal takes even more.

Given a choice between burning coal at 3 AM and idling an NG plant, or burning NG at night and turnign of a coal plant, idle the one with the more expensive fuel.

Base load plants have thicker turbine blades at the roots (last longer at constant load), whilst peakers have thinner blades (heat & cool easier, last longer when cycled).  When the City of Astin was looking at swicthing one of their NG steam plants from base load (load with cycle between ~33% & 100%) to peaker, they planned to remachine the turbine blades at considerable expense.

Karahnjukar hydro turbines were designed for a steady load.  Very different turbines would have likely been built if it were a peaking hydro plant.

So it is not the fuel, but a number of engineering decisions that are made in design.  What works best for cycling use (on/part load/off) does not work best for steady 100% load use.  Cycling a base load plant, regardless of fuel, is just NOT a good idea.  And running a peaker at full load for 500 days straight is also not a great idea.