Home
If you can not measure it, you can not improve it.
Lord Kelvin (William Thomson)
ISO 9001 certified organisations have to make decisions regarding where to send their measuring instruments for calibration. Many calibration laboratories are accredited to ISO 17025 but some are not. Many accredited laboratories are not accredited for all the services they offer. The use of non-accredited calibration labs, or non accredited services of partially accredited labs, may reduce operating costs in the short term, but could turn out to be costly in the long term. Examination of ISO 9001 (2000) and ISO 17025 suggests that ISO 9001 certified organisations should select their calibration labs carefully and make sure that the labs they use are properly accredited for the services they provide.
Organisations certified to ISO 9001 are required to calibrate all measurement equipment used to verify or control quality, and all such calibrations are required to be traceable to national or international standards (ISO 9001 1994 section 4.11, ISO 9001 2000 section 7.6). Records of calibrations are required to be kept and corrective action taken when measurement equipment is found to be out of specification.
In this topic we discuss some of the implications of calibration and traceability requirements for ISO 9001 certified organisations and for calibration and test laboratories. We investigate the meaning and components of the term `traceable' and show that ISO 9001 certified organisations should use laboratories accredited to ISO 17025 for the calibration of all test and measurement equipment used to verify or control quality.
Many calibration laboratories claim accreditation to ISO 17025[2]. In Australia NATA[1] is the accrediting body, and accredited Australian labs are entitled to use the NATA logo on their documents and web pages. Accrediting bodies in some other countries are listed below[7]. ISO 17025 is an international standard that specifies quality and technical competence requirements for testing and calibration laboratories. ISO 17025 replaced ISO Guide 25 in 1999.
In Australia the ISO 17025 accreditation process includes an initial on-site visit by NATA officials and NATA-appointed technical experts who assess the calibration environment and the qualifications and competence of the technical staff. The assessment includes discussions with technical laboratory staff to allow them to demonstrate their knowledge and expertise. A measurement audit may also be carried out in which the lab is asked to calibrate a well characterised instrument or artifact. These activities are designed to bring to light any deficiencies in the technicians' understanding of the calibration processes. Subsequent to the initial assessment, the laboratory's performance is regularly assessed through proficiency testing[6]. Proficiency testing involves regular round-robin calibration or testing of pre-prepared artifacts, instruments or samples, and comparison of individual lab results with either the group result or a reference lab result.
An ISO 17025 accredited lab is required to perform internal audits (section 4.13) to verify that its operations continue to comply with the requirements of its quality system. Labs are also encouraged to maintain in-house quality checks on their calibration standards so that departure from specified performance is detected early (ISO 17025 section 5.9). For example, a voltage calibration lab might calibrate a stable in-house voltmeter regularly and statistically evaluate deviations from mean values to warn of possible problems with their voltage calibrator.
Services from ISO 17025 accredited labs are usually a little more costly than apparently identical services from non-accredited labs. Calibration labs that are not accredited to ISO 17025 have reduced costs and hence can deliver lower cost calibration services. In an era when share prices and quarterly returns rule, it is very tempting to shop around and simply use the cheapest service. This article attempts to show that this approach may not be optimal in the long term.
Hiring and keeping competent technical staff, internal audits, maintenance of in-house quality checks and participation in proficiency testing programs improves the likelihood of an error-free service but can never guarantee complete absence of calibration or other errors. However, customers of reputable ISO 17025 accredited labs can expect to be informed promptly and fully of errors when they are discovered, and of the particular consequences related to the calibration of their equipment (sections 4.9, 4.10).
The ISO 9001 requirement for traceable calibration of test and measurement equipment raises questions concerning the term `traceable'. We examine definitions and components of traceability extracted from ISO 9001, ISO 17025 and other documents, and discuss the implications.
VIM[3] defines traceability as the property of the result of a measurement or the value of a standard whereby it can be related to stated references, usually national or international standards, through an unbroken chain of comparisons all having stated uncertainties .... The unbroken chain of comparisons is called a `traceability chain'.
An unbroken chain of comparisons is a logical and easily understood component of traceability. In its simplest form a traceability chain can be thought of as a pedigree or list of makes, models and serial numbers of instruments or artifacts in the chain. The manager of a non-accredited lab might claim that his/her calibrations are traceable because he/she is able to trace the calibration pedigree of the references and standards he/she uses. However, there is more to traceability than a simple list of hardware.
We discuss this aspect by example. Assume we keep a set of weights which we use to check balances in a chemical laboratory. If we can show that our weights are calibrated against weights which have a traceability chain that leads to the standard kilogram in Paris can we claim that our weights are traceably calibrated? Consider briefly the process of using a balance to compare our weights with a set of calibrated weights. The balance should have a resolution and repeatability necessary for the uncertainty required in the final result. It must be properly serviced and maintained, mounted on an appropriately rigid and vibration-free bench in a temperature controlled environment, and not abused in any way. Air movement around the balance may need to be restricted. If the weights to be compared are of different densities compensation for buoyancy might be necessary. Buoyancy compensation might require measurements of air temperature, humidity and barometric pressure. If the lab provides other calibration services then the presence of other equipment nearby may alter the environment in the vicinity of the balance, e.g. a temperature calibration oven might alter the mean radiant temperature in the vicinity of the balance.
If we appreciate the potential complexity of the calibration process then we should require that the lab calibrating our weights employ a technician with sufficient competence and training to appreciate all the potential sources of error in the calibration. He/she should be capable of setting up the equipment properly and deciding which errors are significant and which can be ignored for a particular calibration.
Competence as a component of traceability is addressed in ISO 17025 section 5.6. Section 5.6.2.1.1 states that ... traceability of measurement shall be assured by the use of calibration services from laboratories that can demonstrate competence, measurement capability and traceability. The use of the word `shall' in a standard usually means that there is no other way to achieve compliance. ISO 17025 further notes that:
Hence traceability as defined by VIM[3] and ISO 17025[2] contains a recursive element that requires ISO 17025 accreditation at each step.
No measurement is ever true. There is always a difference between the true value of a measurand and the output of an instrument. Measurement uncertainty is an quantitative statistical estimate of the limits of that difference.
VIM[3] defines measurement uncertainty as a parameter associated with the results of a measurement, that characterises the dispersion of the values that could reasonably be attributed to the measurand.
There are a number of reasons for the inclusion of uncertainty estimates as essential components of traceability. We discuss two below.
More than one calibration lab has commented to us that many of their customers do not appear to be interested in uncertainties associated with the calibration of their instruments. The customer should view uncertainty estimates as confirmation that his/her instrument was calibrated against a reference of adequate performance and that all potential sources of error were under control during the calibration process.
The authors have seen evidence that instruments from ISO 9001 certified top-100 Australian companies are being calibrated in laboratories that are not NATA accredited. These organisations might be making small savings in the short term by using non-accredited labs. If, however, inadequately calibrated instruments are used to verify or control quality, then those organisations may find themselves in an embarrassing situation if their products are subsequently found to be out of specification. In extreme cases it may be necessary to recall all products manufactured since the last time the instrument was traceably calibrated. Organisations using non-accredited calibration labs do not conform to ISO 9001 and should not claim conformance.
Some calibration laboratories offer a wide range of calibration services but are accredited for only a subset of those services. In some cases labs claim `ISO 17025 accreditation' but are vague about exactly which services are accredited and which are not. ISO 9001 organisations should be careful to select calibration labs that are explicitly accredited for the services they are using. In Australia NATA[1] keeps an up-to-date publicly available list of accredited labs with details of the calibration services for which they are accredited and their least uncertainties of measurement.
In a manufacturing environment it is often the case that more than one measurement system is used to monitor or control the quality of the product, and inevitably some measurements contribute more than others to uncertainty in product quality. ISO 9001 does not require all measurement systems to be calibrated - only those that contribute significantly to the control or verification of the quality of the product. One approach to this problem might be to perform uncertainty analyses on quality-related measurements using techniques similar to those outlined in the ISO GUM[4] to determine which measurement systems require calibration and the maximum associated uncertainties.
ISO 9001 certified organisations should analyse the measurement systems they use to verify or control quality, make informed decisions on which instruments require calibration, and have these instruments calibrated by selected ISO 17025 accredited labs.
©2002 Martin Turner B.Sc. (Eng) Ph.D.
Engineering and Measurement Consultant
12 Goodman Place, Cherrybrook, NSW 2126, Australia
Tel: 0403-007 305 (International: +61-403-007 305)
Email: mjturner at biccard.com
Disclaimer The views expressed and information provided in
these documents are the opinions of the authors and do not
represent specific advice on any topic.
First published: 8 Sept 03 Last modified: 23 Feb 03
Home