ISO 17025:2017 clause 7.6.2 - Performing calibration of its own equipment shall evaluate the measurement uncertainty

[Khairul_jaafar]

Hi everyone,

I'm quite new with ISO 17025 standard and still need a lot to learn and understand.

My current company is accredited in ISO 17025 as a Testing Laboratory. However my company also performing calibration on equipment especially our own equipment which currently I found that they do not have the measurement uncertainty. While some of the equipment, my company will send to accredited Lab to perform calibration where they will provide a complete certificate with the measurement uncertainty.

My question is since we are not accredited in calibration do my company still need to provide calibration with evaluation of measurement uncertainty to satisfy clause 7.6.2?

 

[lawcch]

Hi everyone,

I'm quite new with ISO 17025 standard and still need a lot to learn and understand.

My current company is accredited in ISO 17025 as a Testing Laboratory. However my company also performing calibration on equipment especially our own equipment which currently I found that they do not have the measurement uncertainty. While some of the equipment, my company will send to accredited Lab to perform calibration where they will provide a complete certificate with the measurement uncertainty.

My question is since we are not accredited in calibration do my company still need to provide calibration with evaluation of measurement uncertainty to satisfy clause 7.6.2?

definitely. YES.. either your lab accredited or not, you still have to ensure that every calibration work must calculate and include all TOTAL measurement uncertainty (MU) base on GUM documents or technical standards. It is a a very difficult technical subject to explain here. I advise you to attend professional training and you must be very good in statistical techniques and data analysis. Also, you must identify sources contributed to your total MU base on the number of processes or steps involve in the calibration processes.
Not only you need to attend technical training but you must have working experience in calibration works in order to become a competent technicians to do calibration or testing jobs. It is all mentioned in ISO/IEC 17025 on competence of personnel and calibration of your own measuring equipment in the lab.

 

[Khairul_jaafar]

definitely. YES.. either your lab accredited or not, you still have to ensure that every calibration work must calculate and include all TOTAL measurement uncertainty (MU) base on GUM documents or technical standards. It is a a very difficult technical subject to explain here. I advise you to attend professional training and you must be very good in statistical techniques and data analysis. Also, you must identify sources contributed to your total MU base on the number of processes or steps involve in the calibration processes.
Not only you need to attend technical training but you must have working experience in calibration works in order to become a competent technicians to do calibration or testing jobs. It is all mentioned in ISO/IEC 17025 on competence of personnel and calibration of your own measuring equipment in the lab.

Thanks for the explanation

 

[lawcch]

My question is since we are not accredited in calibration do my company still need to provide calibration with evaluation of measurement uncertainty to satisfy clause 7.6.2?
I think you must specify what type of accuracy you want to achieve in your own verification of measuring equipment. If you want to perform your own calibration even though you do testing work , your reference measuring equipment must has traceable to national or international standard .and you must specify what level of accuracy you want to achieve in your calibration. Therefore, I would NOT advise you to do your own calibration work for yur measuring equipment which use in your testing lab. It is an expensive affair to do a lot calibration for your measuring glassware, thermometers, relative humidity , dewpoint meter, oven or furnaces , etc.
If your management is serious about lab accreditation , you must spend a lot of money to do calibration works on all your measuring equipment so that you can calculate your own MU in doing testing works.

 

[Khairul_jaafar]

My question is since we are not accredited in calibration do my company still need to provide calibration with evaluation of measurement uncertainty to satisfy clause 7.6.2?
I think you must specify what type of accuracy you want to achieve in your own verification of measuring equipment. If you want to perform your own calibration even though you do testing work , your reference measuring equipment must has traceable to national or international standard .and you must specify what level of accuracy you want to achieve in your calibration. Therefore, I would NOT advise you to do your own calibration work for yur measuring equipment which use in your testing lab. It is an expensive affair to do a lot calibration for your measuring glassware, thermometers, relative humidity , dewpoint meter, oven or furnaces , etc.
If your management is serious about lab accreditation , you must spend a lot of money to do calibration works on all your measuring equipment so that you can calculate your own MU in doing testing works.

Yes, we have Master or reference equipment which is used to either calibrate or verify the equipment and It is traceable to national or international standard. The issue that concern me is no measurement uncertainty but only data been recorded after the calibration been performed by the responsible technician. I would take into account your suggestion and advice, Thank you.

 

[Gus]

Yes, we have Master or reference equipment which is used to either calibrate or verify the equipment and It is traceable to national or international standard. The issue that concern me is no measurement uncertainty but only data been recorded after the calibration been performed by the responsible technician. I would take into account your suggestion and advice, Thank you.

I'd like to keep this short and useful so here's the basics that i think are at fault (i am the supervisor of a 17025 accredited Test Lab) and not to confuse you with the "why its important", "GUM" etc.. @lawcch covered this pretty thoroughly


For one thing, you as a Test Lab are required by 7.6.1 to identify and evaluate Uncertainty on your own methods, this by itself requires that you factor into this "math" the uncertainty of your equipment, therefore forcing you to have uncertainty reported on every single Calibration certificate from your equipment that has a impact on the results of your test.

i second the opinion that its not ideal for you to calibrate your own equipment, having said that, there are cases where you might have to do so, (for example in our case we use an externally calibrated metrology well and PRT to calibrate our thermocouples, we use loads of them, and it would be ridiculously costly and time ineffective to send each one of them for external calibration) but you absolutely have to try in keep the traceability chain as short as possible on your side to avoid headaches when calculating additional uncertainty, you have to make sure you get some training on uncertainty measurement, i recommend you hire an expert and work on your particular case and document it well,you might even make it so you measure uncertainty on the entire process and you are home free!, IMHO the GUM is quite a handful to try and do it alone.

Another thing to watch out for when doing this is auditor bias... calibrating your own equipment may be viewed as a risk to impartiallity or some other kind of risk

 

[Khairul_jaafar]

I'd like to keep this short and useful so here's the basics that i think are at fault (i am the supervisor of a 17025 accredited Test Lab) and not to confuse you with the "why its important", "GUM" etc.. @lawcch covered this pretty thoroughly


For one thing, you as a Test Lab are required by 7.6.1 to identify and evaluate Uncertainty on your own methods, this by itself requires that you factor into this "math" the uncertainty of your equipment, therefore forcing you to have uncertainty reported on every single Calibration certificate from your equipment that has a impact on the results of your test.

i second the opinion that its not ideal for you to calibrate your own equipment, having said that, there are cases where you might have to do so, (for example in our case we use an externally calibrated metrology well and PRT to calibrate our thermocouples, we use loads of them, and it would be ridiculously costly and time ineffective to send each one of them for external calibration) but you absolutely have to try in keep the traceability chain as short as possible on your side to avoid headaches when calculating additional uncertainty, you have to make sure you get some training on uncertainty measurement, i recommend you hire an expert and work on your particular case and document it well,you might even make it so you measure uncertainty on the entire process and you are home free!, IMHO the GUM is quite a handful to try and do it alone.

Another thing to watch out for when doing this is auditor bias... calibrating your own equipment may be viewed as a risk to impartiallity or some other kind of risk

Thank you for your explanation. Both of you are really helpful.