FMEA Severity Index - Leading contributor to the RPN value

Kevin Mader

One of THE Original Covers!
Leader
Admin
Chad,

Severity may be the leading contributor to the RPN value. Severity measures the impact or outcome of an event. I have heard that Severity of 9 and 10 require corrective action, but nothing in regard a 7 rating. Either way, the difference may be meaningless as the outcome of an event may not be avoidable. It could be me, but I think someone is playing a numbers game. Perhaps only to see if you are prepared to defend your position.

In the interest of producing better products for consumers, organizations need to practice good "business ethics". A severity rating that is a 9 or 10 should challenge an organization to determine if their are ways they can reduce impact. For example, building secondary or tertiary safety features into a design or process (provided they exist) to prevent injury or death may be required. These additions would be perceived as "value added" as they protect the customer. In the nature of Continuous Improvement, I suppose an organization should explore reducing the severity of an event, even for lower ratings (8>). But where does an organization draw the line? Resources may be limited and reacting to every 7 rating may not be possible. Nor is it always necessary. Reducing Severity should be weighed against the value that is gained. Is the customer willing to pay the additional costs of secondary and tertiary safety/convenience features? Probably not. Often these additional items create waste in a product or process. The Customer does not want to pay for waste! Explore the lower cost improvements, those that can increase 'delight' within the customer base and are value added.

So what does the Customer consider waste? Ask them. Otherwise you run the risk of introducing waste to the system. Everyone loses under that premise.

Regards,

Kevin
 
C

Chad Eilers

Kevin,

Thanks for your thoughts. I agree with you, totally.

We are a Magnet Wire manufacturer (insulated, copper wire) and the highest severity index we have is an 8 (non-hazordous but could result in 100% scrap). Therefore, we concentrated on improving our occurences and detectability. Most RPN's for the high severities are < 32. I felt like we were doing the right thing, and i truely beleive we are. Your comments seem to be very close to my idea of what the standard was asking.

Thank for your time.
 

Kevin Mader

One of THE Original Covers!
Leader
Admin
Chad,

I am glad to help. Keep in mind that sometimes Registrars, or auditors for that matter, challenge your decision making. This isn't necessarily to try to catch you doing "something wrong", but sometimes just to determine how you, or your organization, arrived at a conclusion. As Marc often points out here in the Cove, you must be prepared to defend yourself and, occasionally, fight the fight. I believe that they are mostly looking for how well you make decisions (are you considering things?).

Personally I find the FMEA process to be a very good Risk Analysis process. Considering the subjectivity of things in general, the FMEA guidelines to a fair job of creating continuity in the process. What you need to consider as an organization is that it won't necessarily fit every application as spelled out in the reference manual. There is a single line (often overlooked) where it instructs the reader to create similar criteria more meaningful for the organization. In theory, you may not denote a severity rating of 9 as a risk of injury to the user. But organizations should be warned that abrupt deviations from established "norms" may meet with heavy resistance and, perhaps, make it not worth pursuing.

Regards,

Kevin
 
C

Chad Eilers

FMEA Severity Index

During our QS-9000 certification, it was brought to my attention that if the severity level is equal to or above 7 we must do corrective actions for those items. However, the only way to lower severity is to perform a design change. We do not do design, but the auditor said that this "AIAG Interpritation" is for both the Process and Design FMEA.

Has anyone come across this issue in the past, and how should I handle this?

Thanks for your insight!
 

Wes Bucey

Prophet of Profit
My understanding is TS16949 adopted the AIAG manuals (FMEA, PPAP, etc.) in their entirety, but "suggests" applicability is only based on customer requirement. (i.e. have the core tools ready, but only use them if the customer says so.)
 

The Taz!

Quite Involved in Discussions
A severity score can only be lowered by a change in design . . design the potential failure out . . . Occurrence is dictated by past experience and is a product of the Detection. Put in better Detection, and you will see a lowering of the occurrence over time.

In all cases I have seen (Fuel systems), any environmental impact failures (Gas or oil leakage for example), have also rated a 10 for severity.

The severity is dictated by the customer in most cases using guidelines and they are determined and stated during design reviews. You have little control over it unless you have full design responsibility, and have the ability to dictate your own scores.

Ever wonder what the severity score for delamination of tire layers was?. . . and now is? who determined it?

I guess I'm saying, don't cheat yourself and expose your organization to undue risk. If the score should be a 9 or 10 . . . then so be it and do your due diligence and detect the heck out of it. Detection is the only score you have direct and immediate control over.
 
B

Bill Ryan - 2007

The Taz! said:
Occurrence is dictated by past experience and is a product of the Detection. Put in better Detection, and you will see a lowering of the occurrence over time...
Gotta disagree Taz. The Detection rating has NO effect on the Occurence. The only way to affect Occurence is with a prevention control of some type.

Bill
 
R

Rob Nix

This is a methodology that doesn't lack for variety of application and interpretation!

So this is MY take, FWIW IMHO etc etc:

Once a failure mode and its EFFECT is identified, its SEVERITY rating will NEVER CHANGE. A certain effect has a certain severity - period; you may be able to eliminate the effect, and therefore drop it from the analysis, but you cannot lower or raise its severity.

So, the ONLY way to lower an RPN is by lowering its probability of occurence (read, frequency) or by improving the ability to detect (or better yet, prevent) the failure.

As far a choosing a certain number (e.g. 7) as a cutoff point where action is REQUIRED, that is for the customer to decide; although I think it is nonsense to use subjective numbers or calculated numbers (i.e. RPN) AT ALL when performing FMEAs. As I've said in other threads, the value of the FMEA is in the talent in the room, and the exercise itself. The ratings serve no value at all. The group decides: shall we do something to improve the design for this particular failure mode or not? yes or no; don't rate it first.

There may or may not be a correlation between the "occurence" and "detection" factors. It is rather a moot point.
 
B

Bill Ryan - 2007

Rob Nix said:
There may or may not be a correlation between the "occurence" and "detection" factors. It is rather a moot point.
As with your "other post", I agree with you regarding the value of the exercise - however - automotive SQEs have recently begun actually reviewing supplier PFMEAs. Both OEM and Tier 1 have issued their own specific "ranking guidelines". It is imperative that, as an automotive supplier, I understand that the Detection rating does not, and cannot, have an effect on the Occurence rating. The Detection Control is, very simply, my organization's ability to catch the Failure Mode, after it has been produced , and not ship it to our customer.

I realize we may be "splitting hairs" here, but understanding this can "save many lives" on the Supplier/Customer "battlefield".

As always, JMHO.

Have a great weekend everyone (and Good Luck on the office pools for the first round!) :bigwave:

Bill
 
Top Bottom