Hello all! I have some questions on the beginning phase of ISO 17025.

Our calibration lab is under a mature QMS (ISO9001- 10+ years; regulated industry; 1000-4000 employees). For discussion, I will specify the current, comprehensive sytem as the "large" system, and the calibration activities program as the "small" system.

As we sit, my assessment is we have a good program (small and large). There is interest, for various reasons, to fall under the 17025.

For the most part, most of the components specified is 17025 are encompassed within the larger system. We already have procedures in place to handle:

Purchasing
Calibration records
Management/ management review
Nonconforming product
Corrective action program
Document control....
Among others

There does not appear on the surface to be any reason why these would be deficient within the "small" system.

NOTE: There are a few I have identified that need addressing.

Currently (and in the foreseeable future) there will only be internal customers.

So, am I hearing that in the end, it will be better to create a new quality system within the system? So, do I "pretend" that the larger umbrella system does not exist, and create from scratch?

Too, I cannot imagine that the quality professionals responsible for the large system would be completely OK with creating a small system independent of the large. That is my thought; but nothing verified on that one.

Your thoughts, input, and opinions are welcome.:)

For the most part, most of the components specified is 17025 are encompassed within the larger system. We already have procedures in place to handle:

Purchasing
Calibration records
Management/ management review
Nonconforming product
Corrective action program
Document control....
Among others

There does not appear on the surface to be any reason why these would be deficient within the "small" system.


Brad, the devil is halfway in the details (I'll come to the other half later). You really want to do a careful gap analysis of your, for example, Purchasing procedure against the requirements of 17025.

If your procedure dots all the i's and crosses all the t's, you're home free. If not, you have an interesting choice: amend your large-system procedure (and possibly burden the folks who can still operate by the large-system document), or create a supplemental procedure for the small system only.

Beware of the words "policy and procedure." According to A2LA, at least, they literally mean you need both a policy and a procedure. And that brings me to the other half of the devil. A2LA (others may be different) loads heaps and heaps of interpretation on top of the standard. Fortunately for such as you and me, they also provide ample documentation warning you about their interpretations.

NOTE: There are a few I have identified that need addressing.
If by that you mean you have identified procedures that need expansion because of expanded requirements in 17025, then you're OK. Just be sure to do the gap analysis.

Currently (and in the foreseeable future) there will only be internal customers.
That distinction loses a lot of meaning if your lab releases product.


So, am I hearing that in the end, it will be better to create a new quality system within the system? So, do I "pretend" that the larger umbrella system does not exist, and create from scratch?
No. Use the larger system. For example, slip in the one or two minor 17025 points into the management review agenda. Don't have a separate review just for the lab.

But you will probably find yourself creating supplemental documents, such as a separate quality manual supplement (I call it a compliance guide, or something similar) for the lab.

Go to the A2LA website and check out the document finder. There are about ten categories of documents, with three or four documents of interest in most categories. Download 'em and read 'em. If you are going to use another accreditation body, get as much reference documentation from them as you can.

Ken, thank you so much. Let me carry on your thoughts, here.

Brad, the devil is halfway in the details (I'll come to the other half later). You really want to do a careful gap analysis of your, for example, Purchasing procedure against the requirements of 17025.



A gap analysis will definitely be in the works. Those attachments here at the Cove will be very helpful.

If your procedure dots all the i's and crosses all the t's, you're home free. If not, you have an interesting choice: amend your large-system procedure (and possibly burden the folks who can still operate by the large-system document), or create a supplemental procedure for the small system only.



Ahh.. the devil as you suggested. I don't think I can exert influence on the large system. It sounds like I will need to create procedures for the small system, and assure compliance within the large system. That, may be the tricky tightrope to walk.

Brad,

Who is your accreditation body?

Or have you selected one yet?

Ahh.. the devil as you suggested. I don't think I can exert influence on the large system. It sounds like I will need to create procedures for the small system, and assure compliance within the large system. That, may be the tricky tightrope to walk.

Brad, it may not be as bad as all that. Just be careful how you state the scope and purpose. To beat up on Purchasing as a quick example:

This work instruction describes the policy and procedures by which the small quality system conforms to the purchasing requirements of ISO/IEC 17025.

The scope of this work instruction includes only services and supplies for laboratories operating according to ISO/IEC 17025.
Or words to similar effect.

Also, why don't you start a separate thread for this question so it can get the attention it deserves.

Brad,

Who is your accreditation body?

Or have you selected one yet?

No, I have not selected one yet, Cliff. I am still in the information gathering phase on this. My boss will probably want a scope. I'm in Texas, suggestions are welcome. If any of you are skittish about conflict of interest and all, PM me with your suggestions.

Brad, it may not be as bad as all that. Just be careful how you state the scope and purpose. To beat up on Purchasing as a quick example:

This work instruction describes the policy and procedures by which the small quality system conforms to the purchasing requirements of ISO/IEC 17025.

The scope of this work instruction includes only services and supplies for laboratories operating according to ISO/IEC 17025.Or words to similar effect.

Also, why don't you start a separate thread for this question so it can get the attention it deserves.

Ok, will do.

Brad, it may not be as bad as all that. Just be careful how you state the scope and purpose. To beat up on Purchasing as a quick example:

This work instruction describes the policy and procedures by which the small quality system conforms to the purchasing requirements of ISO/IEC 17025.The scope of this work instruction includes only services and supplies for laboratories operating according to ISO/IEC 17025.Or words to similar effect.


Ok, so this is the essence of the gap analysis. Determine where the existing large system is deficient (in the requirements of 17025) and address those in the small system. If I understood Jennifer's (and the others) post on the other thread, there is no need to have formal "this is in accordance with ..... large system reference). We willl need to focus on gaps in the small system documents.

Brad, we went through a similar scenario in 2001. After some discussion with our auditor, we decided to keep our ISO17025 accreditation separate from our TS QMS.

We do reference certain parts of it but thats about it. It has worked well for us and we have had no problems keeping the lab separate. It drives our TS auditor nuts, but he spends very little time in the lab now:agree1:

Greetings Brad,
If I may make a suggestion…..

I know very little about 9000

Do your comparison with your existing documents and the 17025:2005 standard.
Then start cutting and pasting. Add what you need, throw out the rest. KEEP IT SIMPLE. Small entities need less convolutions due to their size which simplifies the communication requirements even right down to streamlining corrective, preventive actions and improvement.
Pay particular attention ( as stated previously)to Purchasing calibrations for equipment standards, consumables for calibration/testing if you use them, and standards themselves. This part is unforgiving.
Training and qualifications are a big issue.
Section 5 should be different than just a management system that you have.
You must also pay attention to the notes as they will be scrutinized. Not to the point of “Shall” but needs to be incorporated “where practical”
Uncertainty can be a bear. You must understand your budgets.:bonk:
Proficiency testing seems like a pain but actually turns into an excellent tool.
You will do just fine.
Remember the all important “what does this mean to us”
9 times out of 10 you will already be doing the requirement as a mater of rote. The trick is to document it in an acceptable way.

OK.....take a deep breath.....you identified some good procedures.....but in the QMS there are differences between 9K and ANS/ISO/IEC 17025, in that 9K is NOT meant specifically for a laboratory, and 17025 is.....

The purchasing is one example, for example, under 17025 you need to identify and control the critical consummables and services.....

Two other and more crucial components are Management Review and internal audit.....Management Review has a prescriptive agenda that must be followed.....now, depending on accrediting bodies (ABs), and I can't speak for A2LA, L-A-B, NVLAP, or ACLASS, only for the one I work for on this.....you can combine or rename the topics, so long as all are addressed identifiably and the minutes reflect that.....internal audit has two parts, the QMS and the technical part, and both must be audited.....

Now, the heart of 17025 is the technical aspects of the laboratory operations.....that is where the rubber meets the road as it were.....

Someone earlier mentioned selecting an AB, and to do yourself justice you should get quotes from ALL the ABs, and talk to the Staff person at the AB before signing with any AB.....the five in the U.S. that are internationally recognized are: IAS, L-A-B, A2LA, NVLAP, and ACLASS (I use the more known name for ACLASS, they may have changed names, not sure about that)

Keep putting the questions up and the discussion going.....

First, a big thank you :agree1: goes out to all who have provided assistance so far; online and offline. It truly is appreciated. For those offering help, beware! I’ll probably take you up on it! :lol:

While I do have knowledge on many areas that I am asking about, I am also way too close to this one. Thus, I need some objectivity on this, so I do appreciate the input from all!

1. Policies/ procedures

Now, what I am envisioning is writing a sub-QMS for the “small” system. I also see it as a three-tiered system. The top procedures would outline the quality system. The middle section would encompass purchasing, management responsibility, and uncertainty calculations. Finally, the third tier would be the work instructions. At some point in the system, I will need to meet with our quality folks in charge of our QMS on this, as I believe they will need to accept the 17025 in lieu of the 9001 QMS.

2. Uncertainty calculations

I would like to see if it is feasible to go with what we currently have for the technical portion. We have established at least 4 to 1 accuracy ratios for our standards. We have a few less than 4 to 1, as the current technology limits the possible standards. This is the fundamental comparison of the tolerance of our standard, compared to the tolerance of the U.U.T.

We have not performed extensive uncertainty calculations. This is due to two components- 1) Assuring proper accuracy ratios has been deemed sufficient; 2) we have somewhere between 75-175 standards.:tg: Of course, these (accuracies) are documented in several places within the system. I do need to get these into one document.

3. AB/ consultant

You state to get an AB sooner rather than later. But, if there is really no known robust system even close to 17025 requirements in-place, what value is gained in the initial stage? It seems to me you would want, say a consultant’s help, to develop the system. Then comes the work of the auditor by identifying initial gaps, efficacy of the system, etc. What value comes from getting the AB involved so early in the game?


Your comments, input, are welcome!:bigwave:

Hello all!

As follow-ups to the above questions, another question (in addition to that question :tg:) came to mind. I realize who you use for your higher-order calibration needs to be scrutized. However, I'm pretty happy with who we currently use. Can you 'grandfather' calibration sources into the program?

It wouldn't fly with an AB auditor to hear "we grandfathered this company in due to their outstanding and consistent performance", you would still need to show evidence of their compliance with ISO 17025, either through audits conducted, third party approvals, or similar methods.

*The laboratory shall maintain a register of all subcontractors that it uses for tests and/or calibrations and a record of the evidence of compliance with this International Standard for the work in question.

*The laboratory shall evaluate suppliers of critical consumables, supplies and services which affect the quality of testing and calibration, and shall maintain records of these evaluations and list those approved.

As to your question regarding MU calculations, I don't believe an ISO 17025 AB will let you approach Measurement Uncertainty in this manner. In the past the 4:1 ratio was often used as a way out from calculating UOM, however, the industry has moved away from allowing this and more stress has been put on calculating realistic values for measurement uncertainty. From reading ISO 17025 there is very little wiggle room when it comes to UOM. From my scope of experience, if you calibrate anything in house you are required to have a procedure for estimating UOM and you must apply that procedure to your calibrations.

Thanks, Benjamin. Let me clarify the vendor question. For example, we have a lot of equipment that has to go to the mfg. for calibration, as we have not found suitable calibration sources for that equipment. Thus, I have to use them.

Too, I have another vendor that is not certified to 17025. However, they do excellent work, shows uncertainty, etc. They have been site-certified, but not under the extra scrutiny of 17025 (if extra scrutiny exists).

I just want to practice what I preach. I believe in improvement, but I don't feel the organization should be strapped for extra expenses and limited calibration vendors solely attributed to having to comply with the requirements. I hope that comes across appropriately..:D

Too, I have another vendor that is not certified to 17025. However, they do excellent work, shows uncertainty, etc. They have been site-certified, but not under the extra scrutiny of 17025 (if extra scrutiny exists).

I just want to practice what I preach. I believe in improvement, but I don't feel the organization should be strapped for extra expenses and limited calibration vendors solely attributed to having to comply with the requirements. I hope that comes across appropriately..:D
Benjamin28 pretty much has it right.

A2LA (can't speak about other AB's) does permit exceptions to the 17025-only requirement, but usually only in the case of unavailability of accredited calibrators.

If you want to save money on calibration, maybe you can stretch the interval. Twelve months is the generally accepted interval, but there's no science behind it. If you can show little or no drift (in comparison to your uncertainty budget) between intervals, you might be able to justify extending the interval.

This would be even more true if you used quality control activities such as repeated runs on a single sample or retesting of retains. Such activities could allow you to detect instrument problems real time and schedule an early calibration if applicable.

To answer your other question, 17025 is a much higher bar than 9001.

Thank you guys.:yes:

I don't think the issue is paying for a good calibration. Quality is not questioned. When I spoke of expense, that would be the additional expense of having to replace existing vendors that are adequately serving our business needs solely based on a stringent (but voluntary) requirement.

We have a lot of equipment, and much of it is not mainstream. I want to improve our operation, and not put ourselves in a corner, so to speak, by limiting who can perform work.

Keep the thoughts coming.

Thank you guys.:yes:

I don't think the issue is paying for a good calibration. Quality is not questioned. When I spoke of expense, that would be the additional expense of having to replace existing vendors that are adequately serving our business needs solely based on a stringent (but voluntary) requirement.

We have a lot of equipment, and much of it is not mainstream. I want to improve our operation, and not put ourselves in a corner, so to speak, by limiting who can perform work.

Keep the thoughts coming.

I see your point. Unfortunately the AB guys seem to be pretty hard-headed about this:(.

Is it a conflict of interest? I think so.

This is the kind of question to use to help select your AB.

Yes, ANS/ISO/IEC 17025 does in fact allow use of non-accredited calibration providers; in fact under the standard accreditation is not a requirement - BUT - your calibration providers need to be able to support your accreditation, so you must be able to show they can do that.....

One way is to audit them yourself, taking along a metrology professional with experience in that discipline, and have the associated costs and time and so forth, or if an accredited provider is avaialable, then that becomes easier and less expensive.....and if there is no other provider for that, then the due diligence and documentation and (depending on the AB) you should be OK.

AB assessors can be a bit hard-headed as one wrote further down, and we have a requirement to be, up to a point.....but where the standard gives some leeway then the assessor needs to remember that also.....

So remember all this when you start shopping for your AB, and please DO ask the tough questions.....

Who will be assigned to shepherd your file?
Can you speak to the Staff who will be your contact and even be at the assessment?

Come up with other questions also and use them.

Three days were spent in training covering IS017025. A few observations:

IAS did a very professional job of covering the standard. Also, several representatives for IAS were there. I think that provides a good strong image that the company generally agrees with philosophy and approach, there is a united front, and they are able to add to the discussion.
As to the specification itself, seems to be a decent-enough document. One of the sections that surprised me regarding internal audit reports:

4.14.3 The area of activity audited, the audit findings and corrective actions that arise from them shall be
recorded.

I thought “Corrective Action” was reserved language, and would not seem appropriate to appear on an auditors report.

I realize the standard is meant to be broad; to say what to do, and not how to do it. Do you feel that the standard should be a little more vociferous on the qualifications of the internal auditors?

4.14.1 Such audits shall be carried out by trained and qualified personnel who are, wherever
resources permit, independent of the activity to be audited.

Contrast that to the near absolute requirement that all calibration vendors must be accredited labs. Are we suggesting that is it much more egregious to- use a vendor that may have a good program but hasn’t “paid” for 17025; than having woefully inept internal auditors?

It was explained to be that most accreditors will cite for using a non-accredited lab, then the customer “appeals” through the Corrective Action for the nonconformance. Why would I want (and why would you want a system) to enter with a known problem? I agree it should be highly preferred, but seems a bit aggressive to mandate it. Your thoughts…

I thought “Corrective Action” was reserved language, and would not seem appropriate to appear on an auditors report.
The audit report includes all audit activity, including follow up, presumably by the auditor. The auditor can explain how he cleared the NC

Contrast that to the near absolute requirement that all calibration vendors must be accredited labs. Are we suggesting that is it much more egregious to- use a vendor that may have a good program but hasn’t “paid” for 17025; than having woefully inept internal auditors?
A2LA, for one, reads audit reports closely and will not hesitate to comment on auditor qualifications. Woefully inept internal auditors will certainly attract unfavorable attention.

It was explained to be that most accreditors will cite for using a non-accredited lab, then the customer “appeals” through the Corrective Action for the nonconformance. Why would I want (and why would you want a system) to enter with a known problem?

It is not always possible to find an accredited calibration lab for all equipment. In that case you have no choice but to go through the silly citation-appeal-appeal granted cycle. Sometimes repeatedly.

Hello all! I have some questions on the beginning phase of ISO 17025.

Our calibration lab is under a mature QMS (ISO9001- 10+ years; regulated industry; 1000-4000 employees). For discussion, I will specify the current, comprehensive sytem as the "large" system, and the calibration activities program as the "small" system.

As we sit, my assessment is we have a good program (small and large). There is interest, for various reasons, to fall under the 17025.

For the most part, most of the components specified is 17025 are encompassed within the larger system. We already have procedures in place to handle:

Purchasing
Calibration records
Management/ management review
Nonconforming product
Corrective action program
Document control....
Among others

There does not appear on the surface to be any reason why these would be deficient within the "small" system.

NOTE: There are a few I have identified that need addressing.

Currently (and in the foreseeable future) there will only be internal customers.

So, am I hearing that in the end, it will be better to create a new quality system within the system? So, do I "pretend" that the larger umbrella system does not exist, and create from scratch?

Too, I cannot imagine that the quality professionals responsible for the large system would be completely OK with creating a small system independent of the large. That is my thought; but nothing verified on that one.

Your thoughts, input, and opinions are welcome.:)

Brad,

First, if you have a copy of ANS/ISO/IEC 17025:2005, great, if not, first thing is to purchase it, likely the best price is from NCSLI http://www.ncsli.org

Now, pick it apart word by word and line by line, write everything down that is a requirement. For example, Management Review has a prescriptive agenda that you must be aware of, including topics that may not be a requirement under ISO 9001.....purchasing, watch criticals; sub-contracting (if any) watch how you qualify organizations; Clause 5.9.1 requires special attention and for it also obtain APLAC PT 001 and PT002, both are free at http://www.aplac.org and similar specific requirements are throughout the Standard.....

Now, go through your already-documented QMS word by word and line by line.....WHERE everything matches up, use the parent system.....where you can't absolutely put a peg in the sand that there is a match, then document.....

GET TRAINING!!!!!!! Training in ISO/IEC 17025 and in Measurement Uncertainty.....there are a number of courses available.....

Do your uncertainty, a requirement for traceability of calibration and a great early warning system for problems in the technical arena.....

It is unlikely that you will seek accreditation, but if you do remember there are FIVE accrediting bodies (ABs) in the U.S., and do NOT choose one till you have quotes from all five and have spoken to all five, and called a good sample of their customers.....

Hope this helps.

Hershal

2. Uncertainty calculations

I would like to see if it is feasible to go with what we currently have for the technical portion. We have established at least 4 to 1 accuracy ratios for our standards. We have a few less than 4 to 1, as the current technology limits the possible standards. This is the fundamental comparison of the tolerance of our standard, compared to the tolerance of the U.U.T.

We have not performed extensive uncertainty calculations. This is due to two components- 1) Assuring proper accuracy ratios has been deemed sufficient; 2) we have somewhere between 75-175 standards.:tg: Of course, these (accuracies) are documented in several places within the system. I do need to get these into one document.


As for the 4:1, first, to be absolutely clear, it is allowed and at times even directed under the now-retired American National Standard ANSI/NCSL Z540-1-1994 and also allowed under the current ANSI/NCSL Z540.3-2006.....key is, you must be able to prove it, so you still have to do uncertainty studies.....

It was explained to be that most accreditors will cite for using a non-accredited lab, then the customer “appeals” through the Corrective Action for the nonconformance. Why would I want (and why would you want a system) to enter with a known problem? I agree it should be highly preferred, but seems a bit aggressive to mandate it. Your thoughts…

According to the Standard, simply because a supporting lab is not accredited, that is not worthy - in and of itself - of a corrective action, because the Standard does in fact provide some leeway.....however, there must then be sufficient documentation to prove the ACTUAL requirements of the Standard which are: traceability, competence, and measurement capability, according to Clause 5.6.2.1.1 which addresses traceability.....

Side note, the IAS policies are available at no cost on their website, as are the policies for A2LA, L-A-B, ACLASS, and NVLAP on their resepctive websites.

Hope this helps.

Thanks, Cliff. Excellent as usual.

The audit report includes all audit activity, including follow up, presumably by the auditor. The auditor can explain how he cleared the NC


But that's my point. Generally an auditory cites the clause and evidence of non-conformity. The correction/corrective/preventive action usually is the domain of the auditee, not the auditor. Some things may be easy fixes while they are there; some will take a while.