Chernobyl Reactor Explosion

by ThinkReliability Staff

On April 26, 1986, reactor #4 at the Chernobyl Power Plant exploded, spreading radioactive contamination.  There is much debate about the effects, the magnitude of the effects, and the causes, but we can put together a summary of the root cause analysis here.

It is estimated that thousands (perhaps tens of thousands) of people will die from the aftereffects of Chernobyl.  More than 4,000 children have contracted thryoid cancer.  Additionally, between 50 and 250 million Curies of radioactivity were released, more than 350,000 residents have been resettled, a large area remains contaminated, and over 20 countries received radioactive fallout.

The radioactivity, which had built up in the reactor, was released by an explosion and a fire that occurred due to an uncontrolled power surge.  Inadequate containment resulted in the radioactivity spreading beyond the plant.  The power surge resulted from several actions that increased power and disabled safety systems, and from an unsafe reactor design.  (The reactor was designed so that increased steam production leads to an increase in power. US reactor designs are the opposite.)

The after-effects of Chernobyl continue.  The applications of lessons learned from root cause analysis have been applied in many areas – nuclear power, evacuation planning, radiation health treatments, and food supply.  The only remaining reactors of this type are being shut down.  Hopefully this will not only ensure that another Chernobyl never occurs, but will also improve the safety of many other industries.

Finding Solutions

By ThinkReliability Staff

Once you’ve finished your root cause analysis, determined what the causes of a given incident are and built the Cause Map, now comes the really important part: how do you make sure it never happens again?  To keep an incident from happening again, an organization needs to implement solutions. The first step to implementing solutions is to find possible solutions.  We do this by brainstorming.  The brainstorming process is made easier by the root cause analysis, because instead of finding a solution for “person falls down stairs” we brainstorm solutions for very specific causes, such as “stairs were wet” and “handrail doesn’t extend far enough”.  There are many different methods for brainstorming, but the important point is: don’t discount any suggestions.  Write them down, and move on.  We’ll sort through them later.  Attach the solutions to the causes they control (for example, a solution to “stairs were wet” is “cover stairs from exposure to rain”).  Some causes won’t have any solutions, and some solutions will appear on more than one cause.

Have a wide variety of personnel available for brainstorming.  Sometimes it’s easier for someone farther from the work to see potential solutions, and sometimes the people who do the work every day will have great suggestions they’ve been waiting to bring up.  The more suggestions, the better!  Sometimes a seemingly crazy suggestion will lead to a very practical solution.  Allow people to add on to others’ suggestions.  This can result in a synergistic solution better than the original suggestion.

Once the brainstorming is complete, you’ll have a list of possible solutions.  There are as many ways to select solutions as there are to brainstorm, but I suggest something like the following.  First, make a list of the solutions.  Rate the effectiveness of each solution at preventing similar types of incidents (from 1 to 10, 1 being not very effective, 10 being very effective).   Then rate the ease of implementing the solution (from 1 to 10, 1 being not very easy to implement, 10 being very easy to implement).   Multiply the two together for each solution’s score.  Then, rank the solutions.  The solutions at the top will give you the most “bang for your buck”, or are the most easily-implemented, effective solutions.

Grounding of the Empress of the North

Download PDFby Kim Smiley

On May 14 2007, the 300 foot cruise ship, Empress of the North, grounded out on rocks while rounding Rocky Island during a trip through Alaska’s Inland Passage.  There was significant damage to the hull and the two starboard propellers needed to be replaced.  Costs of repairs totaled more than $4.8 million.  Luckily no one was injured, but over two hundred passengers had to be evacuated from the ship.

This is a common route for cruise ships and the rocks were a well-known hazard clearly marked on navigation charts.  So what happened?

A root cause analysis shows that there were many causes that contributed to the accident.  One of causes is that there were no lookouts at the time of the accident.  The crew members who would have acted as lookouts were performing security rounds.  This was in violation of regulations requiring lookouts at all times and appears to have been a common practice for the crew.

When determining causes it’s important to ask, what is different?  In this case, this was the first watch as Deck Officer for the officer in charge.  He had recently graduated, was newly licensed and inexperienced.  He was not familiar with the deck procedures and the equipment. There was a lot of confusion about watch team roles and he didn’t attempt to take charge of the ship’s navigation until seconds before the grounding occurred.  The National Transportation Safety Board (NTSB) found that the actions, or inaction as the case may be, of the Deck Officer were one of the major factors contributing to the accident.

It’s tempting to stop at this point, but the analysis needs to go farther than just identifying the actions of the Deck Officer as a cause to do a thorough investigation.  Why was he standing watch if he wasn’t fully qualified?  Why wasn’t he prepared adequately prior to being given the responsibility?

The crew member originally assigned the watch was ill.  There are a limited number of possible replacements on a ship this size.  The Master of the ship believed the watch would be a good training watch because it was an easy watch with minimal course corrections needed.  It was also not the practice of the crew to have specific night orders for the overnight watches so the newly arrived junior third officer found himself standing the midnight to 4 am watch with minimal guidance.

Many investigations lead back to human error, but it’s important to ask questions beyond that point.  Changing how people are trained, improving the environment, and providing specific writing inspections can help prevent human errors in many cases.

(The photo above is an official Coast Guard photo.)

Preventing Dog Attacks

Download PDFby ThinkReliability Staff

The occurrence of dog attacks is a significant ongoing problem.  An estimated 4.5 million people are attacked each year, of whom 800,000 seek medical care.  These statistics only include attacks that were significant enough to be reported, so the actual incidence is no doubt larger.  One action that has been taken to reduce the incidence of dog attacks is banning specific dog breeds associated with aggressive tendencies (mostly large breed dogs like Pit Bull Terriers, Boxer Dogs, and German Shepherd Dogs), known as Breed Specific Legislation (BSL).

Although BSL is gaining popularity, it does not address all the causes of dog attacks.  A root cause analysis of dog attacks identifies factors related to the dog (inherent temperament, socialization, protective tendencies, location and level of restraint), the owner (treatment and control of the dog) and the victim (behavior, location, age and experience with dogs).  The etiology of a dog attack is multifactorial and as such, should be dealt with in a broad and diverse approach.

Some suggested alternatives to BSL that take into account the complex nature of dog attacks and are targeted at preventing all dog attacks follow:

– Education about proper behavior around dogs would greatly decrease the potential for dog attacks.  Approximately 80% of attacks are by a known dog and more than half of attacks are against children under 12, suggesting that human behavior around a dog is an important trigger since children are more likely to engage in activities that may be perceived as threatening (such as loud noises, running, improper touching).

– Proper enforcement of existing legislation is a readily available method of reducing dog attacks, as many municipalities have restraint laws that are poorly enforced.  An attack cannot occur without the interaction of a dog and person.  Proper restraint on and off private property would reduce the potential for attacks.

– Stricter regulations and more frequent inspections of breeding operations could play a role in reducing improper treatment of young dogs.  Early socialization plays a large role in that puppies that have little interaction or negative interaction with humans are more likely to develop aggressive tendencies.  In most cases this early interaction occurs within breeding operations.

– Encouragement of voluntary spaying and neutering takes advantage of a widely available procedure to reduce the potential for dog attacks.  One of the most significant predictors of attack is a sexually intact dog.  Outside of a breeding operation there is little reason for not spaying or neutering, and the procedure can have additional benefits for the health of the animal, help control the dog population, and reduce unwanted dogs.

To view the PDF file including the root cause analysis of a dog attack, please click “Download PDF” above.