In a meltdown, when the core hits water, and radioactive steam causes an explosion, what do the citizens do? How far downwind will people be affected? Who will tell us when to leave, and where to go? Will our home insurance cover this type of disaster? A partial meltdown happened in Russia, and it was difficult to contain. Can we contain a total meltdown? How? How long will the area be uninhabitable? Will a meltdown cause and EMP? Will computer chips function?
If there's only a one in a thousand chance of this happening, I want the answers. The sooner the better.
In an LWR, the core is already underwater, so the core would not "hit" the water. The ECCS is designed to flood the core and keep it covered. There are also spray systems to remove residual heat in the event of long-term ECCS failure (e.g., "run-dry"). By then, residual heat is decayed to a level such that evaporative cooling can handle the load. Disabling the EECS and in-vessel systems is even more difficult than penetrating the containment structure, which is no mean trick (a strtegic nuke could probably do it, but external impacts from large objects, even those carrying fuel load, is part of the design basis accident envelope).
How far downwind will people be affected?
Depends on the release term. For example, with the TMI-II release, the safest thing to do, statistically speaking, was to stay put. The millirem-range exposures you might have got if you camped at the plant boundary for the duration of the accident (which no one did), presented a risk smaller than that for driving on the local highways any reasonable distance.
Who will tell us when to leave, and where to go?
Power plant operators are required to have in place an emergency plan that includes evacuation for persons with the EPZ (look that one up). Those plans are reviewed and approved by federal and local agencies. It is rehearsed and practiced on an annual basis.
Will our home insurance cover this type of disaster?
Not homeowners, but the Price-Anderson liability pool. Plant operators pay a premium every year to participate in a liability pool managed and administered by private insurers. The Price-Anderson structure was established by the federal government as the model for the structuring of the liability pool, but the operators pay their way on the premiums, and claims are settled by the private underwriters.
A partial meltdown happened in Russia, and it was difficult to contain. Can we contain a total meltdown?
You are probably thinking of the Chornobil (Ukranian spelling for it, please don't "correct" me) accident and that is irrelevant to consideration of credible accidents involving the LWR technology we use in the West. Meltdown is totally contained in all credible accident scenarios (including external impacts) because the containment integrity is maintained.
How? How long will the area be uninhabitable?
The containment structure allows the release to be contained. The surrounding area may have to be temporarily evacuated if there is some residual release of short-lived noble gases (the most likely effluent), but those disperse (they are chemically inactive) and are short-lived from a radiological viewpoint. The downwind area will likely be repopulated within a few days. For example, the somewhat overly-cautious evacuation of the Middletown, PA area did not result in any long-term abandonment of the territory.
Will a meltdown cause and EMP?
EMP is a phenomenon related to nuclear explosions, which the physics of LWR systems preclude from happening (fuel enrichment too low, insufficient in-core flux, et al.). Don't let Tom Clancy or Whitley Streiber fiction trip you up.
Will computer chips function?
If they're not running Windows ME, my guess is that they will.