# Machine safety: Does ISO 13849-1: 2006 weight severity, frequency, and probability equally?

## New quantitative requirements for designing safety-related parts of the control system (SRP/CS) have created many discussions. Even with new requirements from ISO 13849-1, this updated standard begins with the same old qualitative approach to determine the “goal” (Performance Level required- PLr) for any safety function, asking about severity, frequency, and probability.

05/20/2013

New quantitative requirements for designing safety-related parts of the control system (SRP/CS) have created many related discussions about machine safety.  Yet, even with these new requirements from ISO 13849-1, this updated standard begins with the same old qualitative approach to determine the “goal” (Performance Level required- PLr) for any safety function. The same three questions are still asked; Severity, Frequency and Probability.

EN954-1 came out in 1996 with an amazing way to put more teeth into determining a hazard level and related mitigation solution for any recognized hazard. In so doing we had to analyze each hazard by evaluating the related potential injury by severity, frequency, and probability according to the graph below.

Ten years later in 2006 ISO 13849-1 was updated and released introducing Performance Levels and the requirement to develop the PLr, which I call the goal. To develop the PLr, we again use the qualitative approach by evaluating the related potential injury by severity, frequency, and probability also shown in the graph. There’s a whole lot more we could get into here but let’s keep it focused at the three questions.

What was the criteria for approaching these three questions in their order of severity, frequency and probability? Is severity weighted the most because it’s the first question? Such as; S = 50%, F = 30% and P = 20%? Or is probability asked last because of its greater impact? Such as; S = 25%, F = 35% and P = 40%? Or, does it matter at all? Can all three questions be equally interchanged?

Can anybody provide some insight and background? Has this presented you with any new perspectives? Add your comments or thoughts to the discussion by submitting your ideas, experiences, and challenges in the comments section below.

Related articles:

Contact: http://www.jbtitus.com for “Solutions for Machine Safety”.

Anonymous , 05/20/13 12:41 PM:

The way I see it the starting point in the 13849-1 diagram is 50%. We go up or down from there. If you treat the severity as 25% and the two other decisions as 12.5% each, you end up with the five levels representing 0, 25%, 50%, 75% and 100%. Assuming that's what was intended, the F and P decisions can be treated as interchangeable, but S comes first. I don't know much about the rationale behind the decision process. The standard doesn't do a good job of explaining how the decision process relates to the risk reduction it shows in Figure 2.