More Reasons BWR Power Plants Are Unsafe

By Béla Lipták, PE, Columnist

I received a number of letters pointing out that I did not give a complete list of the reasons why the old BWR (boiling water reactor) power plants are unsafe (Preventing Nuclear Accidents by Automation -- Part 2).

Readers  noted:

• 40 years old plants are still operating, and some are applying for extension of their permits by another 20 years.
• The Diablo Canyon Power Plant in California is almost directly over an earthquake fault, yet it has also applied for a 20- year extension of its operation.
• Several plants can be flooded or are in hurricane regions.
• Many BWR plants still store their spent fuel rods in water pools instead of dry casks that need no cooling and are easier to protect.
• In case of many plants, the "planning zone" for evacuation in the event of a serious accident is only 10 miles instead of 25 or 50.
• Some designs assume that the grid and the backup diesel generators can’t be knocked out simultaneously.
• Many designs provide battery backup for only  four or eight hours instead of days.
• In many BWR plants, the loss of electricity results in loss of cooling, because no “uninterruptible” coolant source (in elevated open or in pressurized closed tanks) has been provided.
• Excess steam pressure is often relieved into quench tanks, which provide only a limited heat sink, instead of being released directly to the atmosphere.
• I did not mention that in August the Fukushima reactors were still out of control, the contaminated water was still not removed, and the building of the new cooling system will not be completed until 2012.

My reason for not discussing the topics on this list is that my articles deal only with automation errors only, not with general design problems. The purpose of this series is to describe only the automation errors that exist in the old American BWR plants and show how they must be corrected to protect against the repetition of the Fukushima accident. Also, each of my articles deal with only one automation or control error and, therefore, the fact that I did not discuss a measurement or control error yet, does not mean that I will not do that later.