Lalmohan wrote: No normal scenario analysis based safety design would have saved fukushima - what happened was indeed abnormal. under such conditions, clearly a fail safe design would have been better - but is no guarantor of absolute safety.
Yes. The LalBrof and LalChix have it right. Drumrolls pliss !!

.
Indeed that is the point I have been saying in my earlier posts on this thread.
However no one answered my quiz on VikAd and Vikrant on the surface and Arihant and others underwater in one of my replies to Bade Mian's post

.
And no one wants to buy my India Tsunami Protection Options either!
Yes, the current engineering (including Gen III +) of the Nuke Plants are incremental in nature ,based on a 1950s /1st gen philosophy and based on investment protection and marginal investment business case of the established nuclear establishment. They make more money when they can tweak existing designs minimally, build large complex nuke plants and then make a continuous cash flow by offering "services" and selling custom fuel for the daily operation of the plant. Sort of like a HP Printer and Cartridge example (though HP practically gives the printer away because it is a highly competitive market, the existing guys have made the system so complex that it costs a bum to set up the plant).
The industry needs fundamentally new disruptive designs that are fail safe. The only one of that was the PBMR kind of stuff , but even then more conventional things can be done. Stuff like LOCA has to be catered to , assuming that it WILL happen , and the plant doesn't go Paki when it does . The current effort (like the Gen1 designs of the 50s is to build redundancy and hope that it doesn't happen), but however a station blackout can scupper it.. a single point of failure.
Now, pliss to answer the kwij on Arihant below the surface and you can get hints on the kind of reactors that can become possible. Minaturized, small sized ones , completely underwater in the seas, no LOCA possible, no pakiness, possibly situated in the bosom of a concrete gravity stabilized offshore rig kind of structure with large concrete "boorqas" with massive reserve of water ityadi and immediate heat transfer in a shutdown state to the sink kind of jazz and even entombing it in a concrete boorqa in a dire situation is just easy by pumping in concrete.
In fact, from the common pics we see of research reactors, completely under the pool, and certified to run for a couple of weeks with no need for any watch whatsoever is the kind of model to go for.
Yes, you need fundamentally new engineering and make it failsafe. Building Tsunami walls, raising height is secondary.
Okay, lets take Sanku and Channakya's case of history based planning (thought he 38m at Fukushima definitely NEVER happened, there would be records of it), you build a 38m capable wall and cater to the current event or even put in a 10% margin, how do you know that the next event is NOT going to be a greater than your design wave with 100% certainty ? You cant. It is statistically impossible. You cant control that. You can only control what you can. That is your design.
What the GOI should do is to reject all current GenIII kind of designs offered and spur the DAE and the indian establishment and issue a challenge/RFP/whatever to the world at large for a reactor design that is failsafe against all possible reactor failure modes and only when proven ,go for a big construction spree.
The Fukushima incident is a timely reminder that all in all, the nuke industry (possibly because it got nearly killed and was starved of investment), has really not progressed much engineering wise beyond the Gen 1 types and any improvement has been rather marginal and not fundamental/radical/step change.