Nuclear power plants must be designed to be fail-safe, if not fail-proof. To be fail-safe means that if an accident does take place, the system goes to a safe state. A simple example of this is an electrical fuse. If too much current tries to flow across the fuse, it melts and stops the flow of electricity. Neither Chernobyl nor Fukushima were fail-safe designs. Future nuclear plants must be designed in a way that provides the public with an absolute degree of confidence that the designs are fail-safe.
Public expectation may be that nuclear designs need to be fail-proof, but there are many reasons why that metric will never be achieved. The most fundamental reason is that we simply can’t guard against every possible outcome. Thus, we try to mitigate possible consequences, and implement fail-safe designs.
If the fail-safe metric can be met, it is still going to be a tough sell to a skeptical public. The reality is that nuclear power is a very low risk way to produce power. But the potential consequences of an accident can be extremely high, as we saw when entire cities were permanently evacuated in the aftermath of Chernobyl and Fukushima.
Those images will be hard to overcome, but if the nuclear power industry is to regain momentum the public will have to be convinced that there is zero chance of another major release from a nuclear power plant.