Herramienta para la detección de necesidades formativas de diferentes

5.1 Traceability is characterised by a number of essential elements:

(a) an unbroken chain of comparisons going back to a standard acceptable to the parties, usually a national or international standard;

(b) measurement uncertainty; the measurement uncertainty for each step in the traceability chain must be calculated according to agreed methods and must be stated so that an overall uncertainty for the whole chain may be calculated;

(c) documentation; each step in the chain must be performed according to documented and generally acknowledged procedures; the results must equally be documented;

(d) competence; the laboratories or bodies performing one or more steps in the chain must supply evidence for their technical competence, e.g. by demonstrating that they are accredited;

(e) reference to SI units; the chain of comparisons must end at primary standards for the realization of the SI units;

(f) re-calibrations; calibrations must be repeated at appropriate intervals; the length of these intervals will depend on a number of variables, e.g. uncertainty required, frequency of use, way of use, stability of the equipment.

a basic quality calibration program

5.2 In many fields, reference materials take the position of physical reference standards. It is equally important that such reference materials are traceable to relevant SI units. Certification of reference materials is a method that is often used to demonstrate traceability to SI units.¹

The other document that goes hand-in-hand with this is EA 4/02, Expression of the Uncertainty of Measurement in Calibration. The purpose of this document is to harmonise evaluation of uncertainty of measurement within EA, to set up, in addition to the general requirements of EAL-R1, the specific demands in reporting uncertainty of measurement on calibration certificates issued by accredited laboratories and to assist accreditation bodies with a coherent assignment of best measurement capability to calibration laboratories accredited by them. As the rules laid down in this document are in compliance with the recommendations of the Guide to the Expression of Uncertainty in Measurement, published by seven international organisations concerned with standardisation and metrology, the implementation of EA-4/02 will also foster the global acceptance of European results of measurement.²

By understanding and following both of these documents, a calibration function can easily maintain traceable calibrations for the requirements demanded by their customers and the standard or regulation that their company needs to meet.

To maintain traceability, without using uncertainty budgets or calculations, you must ensure your standards are at least four times (4:1) more accurate than the test equipment being calibrated. Where does this ratio of four to one (4:1) come from? It comes from the American National Standard for Calibration – (ANSI/NCSL Z540.3-2006) which states: “Where calibrations provide for verification that measurement quantities are within specified tolerances…Where it is not practical to estimate this probability, the TUR shall be equal to or greater than 4:1.”

So, if a TUR of equal to or greater than 4:1 is maintained, then traceability is assured. Keep in mind that a TUR of 4:1 somewhere along the chain of calibrations may not have been feasible, and uncertainty calculations were performed and their uncertainty stated on the certificate of calibration. This is correct and acceptable. In most circumstances, where the need to maintain a TUR of 4:1 comes into play, is at the company or shop level, where the customer’s test

40

a basic quality calibration program

equipment is usually used for production or manufacturing purposes only.

So how does calibration and traceability fit into the big picture?

What does the big picture look like? Why do you need a quality calibration program?

You need to establish a quality calibration program to ensure that all operations throughout the metrology department occur in a stable manner. The effective operation of such a system will hopefully result in stable processes and, therefore, in a consistent output from those processes. Once stability and consistency are achieved, it is possible to initiate process improvements. This is applicable in every phase of a production and/or manufacturing program. But especially true in a metrology department.³

Let’s take for example a calibration program that has six calibration technicians on staff. Four of them work in another facility calibrating the same types of equipment as the other two. However, the other two have far more experience and through no fault of their own do not use the calibration procedures that are required by their quality system.

They have calibrated the same items for several years and feel there is nothing new to learn. One of the four calibration technicians (who are always following the calibration procedures) finds there is a fast, more economical way to perform a specific calibration. They submit a change proposal for the calibration procedure and everyone is briefed and trained on the new technique. The four calibration technicians that have been following the calibration procedure improve their production and save the company money. The two ‘old timers’ have a reduction in their production and actually cost the company money. If everyone was using the calibration procedures like they were supposed to, then this would not have happened.

Process improvements cannot take place across the department if everyone is not doing the job the same way each and every time they perform a calibration.

We are not ignorant enough to believe that when calibration technicians have performed a particular calibration hundreds or even thousands of times that they are going to follow calibration procedures word for word. Of course not. But they must have their calibration procedure on hand each time they are performing the calibration. If a change has been made to that procedure, the calibration technician must be trained on the change before they can perform the calibration;

and the appropriate documentation completed to show that training

a basic quality calibration program

was accomplished and signed off. When the proper training is not documented and signed off by the trainer and trainee, then it is the same as if the training never happened.

What is a quality calibration program?

A quality calibration program consists of several broad items referred to in the Quality System Regulation (QSR) from the Food and Drug Administration (FDA). These items are also referred to by other standards (ISO 9000, etc.) and regulations throughout most industries that regulate or monitor production and manufacturing of all types of products. One of the most stringent requirements can be found in the current Good Manufacturing Procedures (GMP).

The basic premise and foundation of a quality calibration program is to “Say what you do, Do what you say, Record what you did, Check the results, and Act on the difference”. Let’s break these down into simple terms.

“Say what you do” means write in detail how to do your job. This includes calibration procedures, work instructions and standard operating procedures (SOPs).

“Do what you say” means follow the documented procedures or instructions every time you calibrate, or perform a function that follows specific written instructions.

“Record what you did” means that you must record the results of your measurements and adjustments, including what your standard(s) read or indicated both before and after any adjustments might be made.

“Check the results” means make certain the test equipment meets the tolerances, accuracies, or upper/lower limits specified in your procedures or instructions.

“Act on the difference” means if the test equipment is out of tolerance, you’re required to inform the user/owner of the equipment because they may have to re-evaluate manufactured goods, change a process, or recall a product.³

“Say what you do” means write in detail how to do your job. This includes calibration procedures, work instructions and SOPs. All of your calibration procedures should be formatted the same as other SOPs within your company. Here is an example of common formatting for SOPs:

42

a basic quality calibration program

1. Procedures 2. Scope

3. Responsibilities 4. Definitions 5. Procedure

6. Related Procedures 7. Forms and Records 8. Document History

After section 4. Definitions, you should have a table listing all of the instruments or systems that would be calibrated by that procedure, along with their range and tolerances. After that you should have a list of the standards to be used to calibrate the items. This table should also include the standard’s range and specifications. Then the actual calibration procedure starts in section 5. Procedures.

Manufacturer’s manuals usually provide an alignment procedure that can be used as a template for writing a calibration procedure. They should show what standards accomplish the calibration of a specific range and/or function. A complete calibration must be performed prior to any adjustment or alignment. An alignment procedure and/

or preventive maintenance inspection (PMI) may be incorporated into your SOP as long as it is separate from the actual calibration procedure.

There are, generally speaking, two types of calibration procedures:

Generic: temperature gages and thermometers, pressure and vacuum gages, pipettes, micrometers, power supplies and water baths.

Specific: spectrophotometers, thermal cyclers, and balances/scales.

Generic SOPs are written to show how to calibrate a large variety of items in a general context. Specific SOPs are written to show step-by-step procedures for each different type of test instrument within a group of items. Possibly, the calibration form is designed to follow specific steps (number wise); and removes doubt by the calibration technician on what data goes into which data field.

“Do what you say” means follow the documented procedures or instructions every time you calibrate, or perform a function that follows specific written instructions. This means following published calibration procedures every time you calibrate a piece of test equipment.

Have the latest edition of the procedure available for use by your calibration technicians. Have a system in place for updating your

a basic quality calibration program

procedures. Train your technicians on the changes made to your procedures every time the procedure is changed or improved – and document the training.

What do you do when you need to make an improvement, or update your calibration procedures and/or forms? A formal, written process must be in place, to include:

• Who can make changes

• Who is the final approval authority

• A revision tracking system

• A process for validating the changes

• An archiving system for old procedures

• Instructions for posting new/removal of old procedures

• A system for training on revisions

• A place to document that training was done

“Record what you did” means that you must record the results of your measurements and adjustments, including what your standard(s) read or indicated both before and after any adjustments are made, and keep your calibration records in a secure location. Certain requirements must be documented in each calibration record. Of course there are many ways to accomplish this, including:

• pen and paper

• “do-it-yourself” databases, e.g. Excel, Access

• calibration module of a computerized maintenance management system (CMMS)

• calibration software specifically designed for that purpose These include the identification of the test instrument with a unique identification number, their part number and range/tolerance. The location of where the test instrument can be found should also be on the record. A history of each calibration and a traceability statement or uncertainty budget must be included. The date of calibration, the last time it was calibrated, and the next time it will be due calibration should be on the form. There should be a place to show what the standard read, as well as the test instrument’s ‘As Found’ and when applicable ‘As Left’ readings.

The ‘As Found’ readings are what the test instrument read the first time that a calibration is performed, prior to alignment, adjustment or repair.

The entire calibration is performed to see any part of the calibration is out of tolerance. If an out-of-tolerance (OOT) condition is found,

44

a basic quality calibration program

record the reading (on the standard and the UUT) and continue with the rest of the calibration to the end of the calibration procedure. If one were to stop at the point where an OOT is found, make an adjustment, then proceed with the calibration, there is a good possibility that the adjustment affected other ranges or parts of the calibration. This is why the entire calibration is performed prior to adjustment or alignment.

There will be times when an instrument has a catastrophic failure.

It just dies and cannot be calibrated. This should be noted in the calibration record. Then, once the problem is found and repaired, an

‘As Found’ calibration is performed. The UUT is treated the same as any OOT unit, but you would not have been able to collect the original

“As Found’ readings.

“As Left” readings are taken after repair, alignment, or adjustment.

Not all UUTs would be considered OOT when “As Left’ readings are taken. In some circumstances, it might be metrology department policy to adjust an item if it is more than ½ beyond its in-tolerance range, while still meeting its specifications. In this type of situation, after the UUT is adjusted to be as close to optimum as possible, a complete calibration is again performed, collecting the ‘As Left’

readings for the final calibration record. Another example would be when preventive maintenance inspection is going to be performed on an item. The calibration is performed, collecting the ‘As Found’ data.

Then the PMI is completed, and an ‘As Left’ set of data is collected.

If the item is found to be out-of-tolerance at that time, there would not be a problem since it was found to be in tolerance during the first calibration. It would be obvious that something happened during the cleaning, alignment or adjustment and that after a final adjustment was completed to bring the unit back into tolerance, a final ‘As Left’

calibration would be performed.

The standard reading, from the working or reference standard you are using to calibrate the UUT, will also be recorded on the calibration form.

Usually, the standard is set at a predetermined output, and the UUT is read to see how much it deviates from the standard. This is a best practice policy that has been in use in the metrology community since calibration started. However, there will be times when this is not possible.

One example when it would not be practical to set the standard and take a reading is during the calibration of water baths. The water bath is set to a predetermined temperature, and the temperature standard is used to record the actual reading. Compare this to the calibration of pressure gages where a pressure standard is set to a standard pressure,

a basic quality calibration program

and the gage(s) under test are then read, and their pressures recorded on the calibration record, and compared to the standard to see if they are in or out of tolerance. In other case, just as the calibration of autoclaves, they are set to complete a sterilization cycle and a temperature device records all of the temperature readings throughout the cycle and the readings are checked to see if the autoclave met its specifications. The same happens when calibrating thermometers. They, along with the standard, are placed in a dry block and a particular temperature is set. The UUT is compared to the reference after equilibration, and a determination is made as to the in or out of tolerance of the UUT. As can be seen by the above examples, it is not always possible to set the standard and take a reading from the UUT.

Also on the calibration form should be an area to identify the standard(s) that were used, along with their next calibration due date(s), plus their specifications and range.

There should also be a place to identify which calibration procedure was used, along with the procedure’s revision number. There must be a statement showing traceability to your NMI, or in the case of most companies in the USA, to NIST, or to any artifact that was used as a standard.

You should include any uncertainty budgets if used, or at least a statement that a TUR of ≥ 4:1 was met.

List environment conditions when appropriate and show if they pass or fail. According to NCSL International Calibration Control Systems for the Biomedical and Pharmaceutical Industry – Recommended Practice RP-6, paragraph 5.11: “The calibration environment need be controlled only to the extent required by the most environmentally sensitive measurement performed in the area.”⁴

According to ANSI/NCSL Z540.3-2006, paragraph 5.3.6 Influence factors and conditions: “All factors and conditions of the calibration area that adversely influence the calibration results shall be defined, monitored, recorded, and mitigated to meet calibration process requirements. Note: Influencing factors and conditions may include temperature, humidity, vibration, electromagnetic interference, dust, etc. Calibration shall be stopped when the adverse effects of the influence factors ad conditions jeopardize the results of the calibration.”⁵

If the conditions within the area that calibrations are being performed require monitoring according to the standard or requirements that must be met, then a formal program must be in place for tracking those conditions and reviewing the data. If this is the case, then there should

46

a basic quality calibration program

either met, were not met, or are not applicable to that calibration.

You should indicate on the form if the calibration passed or failed.

If the UUT had an out-of-tolerance condition, then there should be a place to show what happened to the UUT, with the following possibilities as an example: 

• The user/customer was notified and the UUT was adjusted and meets specifications.

• The user/customer was notified and the UUT was given a ‘limited calibration’ with their written approval.

• The user/customer was notified and the UUT was taken out of service and tagged as unusable.

Notice that in each circumstance that the user/customer must be notified of any and all OOTs. This is called for in all of the standards and regulations. The user/customer, even if internal to the company performing the calibrations, must be informed if their test equipment does not meet their specifications.

There should be an area set aside in the calibration form for making comments or remarks. Enough space should be available for the calibration technician to include information about the calibration, OOT conditions, what was accomplished if an OOT was found, etc.

And finally, the calibration record must be signed and dated by the technician performing the calibration. In some instances, the calibration record requires a ‘second set of eyes’. This means that an individual higher up the chain of command (supervisor, manager, QA inspector, etc.) must review the calibration record and also sign and date that it has been reviewed, audited, or inspected before it is considered a completed record. If this is the case, there should be a place on the form for the final reviewer to sign and date.

What do you do if, after recording your results, you find that you

What do you do if, after recording your results, you find that you