Asking the Right "whys" in a 5-Why Investigation

Phil La Duke — Fri, 17 Jul 2009 14:24:25 GMT

For years I have been frustrated by the poor use of the 5 whys in root cause analysis. It's a powerful methodology, but in general people ask the wrong questions and spend a lot of time needlessly pursuing red herrings. When it comes to incident investigations and/or continuous improvement events I think instead of asking a simple "why?" we should be asking "how?" as in how does this contribute to an unsafe workplace? instead of why?

Any ideas?

Phil La Duke
www.safety-impact.com

Hazard Stacking

Phil La Duke — Fri, 17 Jul 2009 10:34:05 GMT

I've run into a lot of companies that are smugly pleased with themselves because despite a workplace filled with hazards, a hazard correction process that doesn't work, and serious other safety issues they have a low incidence rate and hurt relatively few workers. There are companies out there whose safety rates are true representations of the relative safety of the workplace, but far more that are time bombs that don't realize that there's a problem until someone is killed.

I'd like to think that most of us aren't working in death traps out there, but we are working in places where process variation creates a kind of a risk stack. I thought of the concept of Risk Stacking as I was thinking about my days as a Geometric Design and Tolerancing (GD&T) training. One of the most interesting concepts in GD&T is Tolerance stacking; basically it's the problem of variance building in a design to the point where parts (while in tolerance in and of themselves) no longer fit together because some are on the high end of the variation limits while others are on the low end of the variation. There are complex formulas for calculating tolerance stack but I don't advocate applying them to hazard stacking.

EXAMPLE: Think of a plastic bottle and bottle cap manufacturer. Neither the bottle nor the cap are ever the exact same dimensions, but as long as they are withing+ or - a certain number they are acceptable. But what if the bottle is at the top of the + tolerance and the cap is at the bottom of the - tolerance. (the bottle is too big but acceptable and the cap is too small but acceptable) While the cap may still fit it will likely be difficult to install and may cause damage to the bottle threads. It's rare that two components would create serious problems, but when you consider a product with many parts that must work simultaneously or in concert with one another the results can be catastrophic.

Same can be true of hazards, but to a much more serious degree. If one element of the process has variation there might be a remote chance that an injury could occur, but as hazards stack (many hazards in the same area) the risk of injury and the likelihood that a severe injury will occur is almost certain. I understand that JSAs seek to find all potential hazards, but that is really more of a FMEA approach and tends to deal with what could happen from an "as designed" perspective (what happens on paper) and the concept of tolerance stack I think is appropriate to devising a comparative risk tool for organizations. Does anyone have any suggestions?

SafetyIMPACT Forums / SafetyIMPACT Forums / SafetyIMPACT! Workshops

Asking the Right "whys" in a 5-Why Investigation

Hazard Stacking