MELISSA BLOCK, Host:
NPR's Joe Palca has been talking with engineers about that question.
JOE PALCA: When humans first started building things, determining whether they would work as planned was mostly a matter of trial and error. Build a bridge, drive a chariot across it and see if the bridge collapsed.
HENRY PETROSKI: And it really wasn't until the Renaissance, until Galileo, that there was what we would today call a rational approach to design through calculation.
PALCA: Henry Petroski is a professor of engineering at Duke University. He says now, computers and mathematical models can predict how a bridge will behave under different conditions. But Petroski says if you ask engineers to build a safe bridge or a safe nuclear plant, they ask you what you mean by that.
PETROSKI: The definition of safe is not strictly an engineering term. It's a societal term. Does it mean absolutely no loss of life? Does it mean absolutely no contamination by radiation? What exactly does safe mean?
PALCA: Michael Corradini is a professor of nuclear engineering at the University of Wisconsin. He says first, you design for normal operating conditions.
MICHAEL CORRADINI: And then with any engineering design, anything at all, you then say, okay, well, that's fine. That's what it operates on. But what if something happens that it doesn't - it experience some unusual conditions?
PALCA: So, for example, let's say you're designing a bridge for a place where gale force winds and heavy snows are virtually unheard of.
CORRADINI: But you still design for it so that the bridge safely performs its function, or if it fails, it fails gently, so there's no catastrophic effects.
PALCA: Corradini says nuclear power plant designers try to think of everything.
CORRADINI: They have a listing of all the things that they think can go wrong, and how they can go wrong, and some sort of ordered list of which is more likely than others.
PALCA: And don't think I'm joking about the possibility of a meteorite falling on the plant.
CORRADINI: It's there somewhere; it's just a very low probability.
PALCA: But sometimes, when you set out to build a nuclear power plant, you don't know exactly what the likelihood of a particular calamity might be.
YOTARO HATAMURA: When we plan something, always we make assumption.
PALCA: He says accidents happen when engineers don't constantly test and revise the assumptions they've made about all the things that a nuclear plant needs to operate.
HATAMURA: We can avoid these accidents by thinking way - backward thinking.
PALCA: And backward thinking is to say, look at the previous steps...
PALCA: ...see if that worked properly.
PALCA: And if it didn't, change it so that you go forward with a better path.
PALCA: Hatamura says it's not always easy to look back and admit to problems that you should have thought of in the first place.
HATAMURA: (Foreign language spoken)
PALCA: Joe Palca, NPR News, Tokyo.
NPR transcripts are created on a rush deadline by Verb8tm, Inc., an NPR contractor, and produced using a proprietary transcription process developed with NPR. This text may not be in its final form and may be updated or revised in the future. Accuracy and availability may vary. The authoritative record of NPR’s programming is the audio record.