Are Nuclear Power Plants Too Safe?

Even though the anniversary of the Japan tsunami is less than a year old, the U.S. is licensing new nuclear power plants for the first time since 1978.  While it would be understandable to react with caution to the Fukushima disaster by temporarily closing plants and suspending or cancelling new or existing plants, the U.S. is moving ahead.  And why shouldn’t they?  Nuclear power generation in the U.S. is very safe. In fact it may be too safe! Although there are a lot of articles warning of the dangers of death and disease caused by nuclear power, most are based on estimates and speculation that are largely unconfirmed. What we do know is that fossil fuel power plants are a much more dangerous place to work than nuclear plants. In the last 15 years there have been no deaths at nuclear plants but over 400 at fossil fuel power plants.

The danger I see for the nuclear power industry comes from overconfidence resulting from the high reliability of processes and equipment in the industry.  Therefore, the real danger may be more from human performance than nuclear radiation. I say that because a lot of what employees do at nuclear power plants is to monitor, inspect and repair equipment. What happens when you monitor something that, because of high reliability, never changes?  The non-scientific description of the result is “complacency.”  The scientific term for complacency is extinction.  It refers to the fact that previously reinforced behavior will eventually stop when it does not produce a reinforcer. Although it is invisible in that nothing is happening, the process does produce tell-tale signs that it is occurring.  Inattention is the one that is most pervasive and dangerous. 

Under extinction there is a slowing response to changes, failure to see small changes, insensitivity to them and uncharacteristic emotional reactions to the behaviors of peers and management. All businesses face a similar problem when many jobs involve monitoring reliable processes. The well-publicized examples of sleeping on the job that occurred in the air control towers last year are examples of what happens when there is no reinforcement for looking at the monitors. When data on monitors rarely change or where no response is required when processes are in control, it is unlikely that employees will be vigilant. The solution that the Transportation Secretary implemented in the control towers (adding another controller) will not fix the problem. The problem is not in the people.  It is in rate of reinforcement built into the job. In most cases it is woefully inadequate. Correcting such problems requires a more in-depth understanding of reinforcement than a pat-on-the-back, an atta-boy or a warm fuzzy. While the nuclear industry pays great attention to processes and behavior surrounding them, the extinction problem may need to be identified and changes made to prevent it. Otherwise, it could be lulled into complacency by being “too safe.”


Read the latest on safety regulations put in place by the NRC in this New York Times article.

Posted by Aubrey Daniels, Ph.D.

Aubrey is a thought leader and expert on management, leadership, safety and workplace issues. For the past 40 years, he has been dedicated to helping people and organizations apply the laws of human behavior to optimize performance.