Is there an 'standard' for the length of calibration record retention

[kidvegas19]

Is there an "standard" for the length of record retention with respect to calibration? I have recently taken over the position of calibration coordinator for this manufacturing company. The company has been around for a long while. The current database has around 500 active tools/gages that require calibration and about 3 times that in tools that do not require cal or have been lost/stolen/sold. For most of these devices, we have on file the calibration certs going back to the beginning of time. Our record retention policy with regards to certs states 'indeterminate' which is then defined as a case by case situation with record retention decided by the departments affected. Apparently, the previous record keepers confused indeterminate with indefinate.

I was thinking of keeping on file a history of the current cal cert plus the previous 2. The electronic database will reference the current cert number as well as the last and next cal due dates. Is this acceptable? I can not find information on what is the norm.

What about tools that no longer exist due to lost/stolen/sold/disposed? How long are the records for that tool expected to kept? For example, I pulled a random file from the file cabinet (which, believe me, took actual physical effort because every drawer is packed so tightly with files) to discover this tool had been bought, calibrated, used, and sold in 1993!

Any information will be greatly appreciated.

KV

 

[gpainter]

We saved ours for as long as we had the Measuring Instrument.

 

[Kevin H]

re: Length of Record Retention

Congratulations on a first post. I ran an Guide 25 accredited lab in a steel mill, typical retention for calibration was set up for 5 years. If an instrument was damaged/repaired, etc. that information was maintained on those records, and only disposed of when that 5 years were up. After that they were disposed of (destroyed). Test results were kept locally for 1 year. After that time the computers were downloaded and test results were stored on magnetic tape. The magnetic tapes had a retention time of 25 years.

Part of determining record retention time revolves around customers - at a different customer some of our product ended up being used to prodcue the raw material for aircraft turbine blades. Because of that, our customers required us to keep records, including calibration ones basically forever. What do your customers demand/need for records retention.

 

[Claes Gefvenberg]

Welcome to the Cove KV

/Claes

 

[qualitytrec]

It really does depend on what and who you are working with? I would ask your customers to define this information in writing. Then determine if you should for legal reasons keep them longer.
JMO
Mark

 

[Bev D]

Agreed. For 'tough' customers like the aerospace or if you are regulated bya governmental agency such as the FDA, the record retention period for Quality Records (Which calibration records are) can range from 3 years to 30 years to the life of the program. I've seen them all. This period should be in your customer's specifications for you. If you are in a more relaxed industry, your Quality Manager should know what your company's stated retention period is...
Note that typically the RECORD must be kept even if the Gage is destroyed, lost or obsoleted. The intent is to maintain traceability of what happened to your product...

 

[Graeme]

Is there an "standard" for the length of record retention with respect to calibration?

Greetings and welcome!

As others have mentioned, unless you are in a regulated industry there isn't a "standard" time - it is whatever your company decides it will be.

I have found, though, that it is best to treat calibration records in two different ways, depending on the use of the instrument or tool.

If the calibrated item is a working tool, then "normal" record retention can apply. Better practice is to keep the record for as long as it can apply to a product that was touched by that calibrated tool - the calibration interval plus the product warranty period, for example. A regulated industry may have other rules - the FAA, for instance, requires calibration records of aircraft maintenance tools to be kept for at least two years.

If the item is a calibration standard, however, there are other considerations. A calibration standard is used to calibrate oter tools and is a link in the traceability chain to the SI units of measure. The historical calibration data of that item is its pedigree and is as important as the current results. Analysing and using the history of the item gives you the ability to determine the current bias and uncertainty, and predict the future performance of the standard. That is important for producing quality measurement results. It can also enable using a standard to a higher degreee of accuracy than relying only on the published specifications.

An example: I have a nominal 1000 Ohm standard resistor with calibration history going back 15 years. The published specifications give me a stability of 10 pm per year and an uncertainty of +/-50 ppm. BUT, from my control charts of actual calibration data I know that the actual performance is much better than that - maybe a stability of 1 ppm per year and uncertainty of +/-10 ppm. That knowledge recently saved me a lot of money, because I could use that existing resistor for a +/-50 ppm calibration instead of spending several thousand dollars for a new one at that performance level.
​

The lab I am working with has set a record retention policy for calibration standards of the life of the standard plus enough time to exceed the calibration recall of everything it touched.

In a way, you can say that calibration standards get better with age - but only if you have the data to prove it.

 

[kidvegas19]

Thanks For the Replies All

Luckily, the markets we build our products for are not as stringent as aerospace and the like. Likewise, the end users of our the products do not risk life and limb with each use. Using the suggestions and input from this thread, I do believe I have a good handle on our needs for retention. I am thinking the computer database will have the complete history of a tool therefore the hardcopy records could contain the current cert and the two previous. This will provide physical evidence for auditing and for the most part, cover the warranty time of the products we build. The last thing I need to do is consult with management to see what they feel would be required for legal reasons.

Thanks again all for the welcome and the replies.

kv