The placement of generators at the base of the containment structures was already a mitigation to another risk: that ground movements common to the area might damage or destroy the generators.
A complex environment is one in which multiple risks constrain options and operations. Generators can be placed neither high nor low without assuming risks or requiring additional mitigations.
If a solution space is a search through a complex multidimensional topology, constraints place bounds on that space.
Come on, there's a lot of know-how to building a platform that can resist an earthquake.
An alternative is build a waterproof wall around it. Or enclose it in a waterproof enclosure. (We do know how to build submarines.)
With some thought, you can think of many more arrangements much more resilient than what they had, and would not have been particularly expensive.
Such as bulldoze a mound and put the generators on top of that.
Other problems with Fukushima had similar simple solutions. Instead of venting hydrogen gas into the enclosed building, where it built up and exploded, vent it to the outside. Cost: a vent pipe.
Not an expert in any of the related fields, but Japan is also a heavy seismic activity zone. Piling up loose dirt is just asking for liquefaction.
My armchair spectator opinion is that I want each reactor 'unit' built in an engineered safe solution, far up, supported by a seismic isolation system. Up in the hills, where people would be sent to evacuate after the earthquake due to the tsunami risk? Yeah, put that thing there.
Also the plant engineers need to be personally authorized and held responsible to do whatever it takes to keep things "safe" in the event of a disaster. They're on site, they're seeing the details, let them have the authority to ensure the greatest possible safety.
> Piling up loose dirt is just asking for liquefaction.
Having investigated this myself when I bought property for a house, this is not expensive to deal with. Heck, my neighbor did. You bore a hole with a big auger, fill it with concrete, and set the structure on it. You can even build skyscrapers on fill with such (see San Francisco).
The only reason these problems were not addressed when Fukushima was built is because people didn't think of it in their failure analysis, not because it is expensive, impractical, or impossible.
There is nothing fundamentally wrong with the Fukushima design, just details.
That limits the cooling water supply. Almost all large reactors are near bodies of water. Palo Verde in Arizona is one of the very few exceptions. They operate a sewerage disposal plant to get water.
The fixes to that design would be straightforward. For example, putting the backup generators on a platform so they wouldn't get flooded.