Single Point Calibration

If I have a freezer at a fixed -20C. Can I calibrate it at that point.......The setpoint is never changed during normal use.

Sounds reasonable then (especially if you have linearity records or tables for the existing temperature probe response).

Don't forget you can also easily calibrate at room temperature (to give you a second calibration point) by turning off the freezer, opening up the door(s) and letting it stabilize at the new temperature.

You should try to find a second point to check the gain. One "unit" could be 0.1 or 10 degrees.

I don't see anything wrong with that, but I would suggest attaching an envelope with the calibration info on a piece of paper inside and date....

Calibrated by:

on date:

Accurate at -10 C:

Calibration methodology:

Then write "calibration info" on envelope....

You need to verify that -20 is -20 and not -15. They have product that has to be kept at that temperature. You can't just rely that the indication on the front of the unit is correct.

" You can't just rely that the indication on the front of the unit is correct."

It sounds like that is exactly what you have been doing.

Besides, if -20 is good then -30 would be better.(safer)

I might refer to this as verification instead of calibration, and if you are relying on a visual indication, unless there is someone sitting in front of the indicator you might not know what the temperature is.

You need an alarm and/or a back-up compressor.

I offer this:

Calibration of temperature control and monitoring devices

Dealt with the state and DOH requirements with respect to research laboratories, the requirement is to keep and maintain a log of temperature readings of every freezer(s)... The temperature reading samples taken daily or hourly depending on the requirement as dictated by a specific laboratory applications...

The accuracy of the freezer's built-in temperature meter is the one that may need to be validated /calibrated at least twice every year...sometimes monthly or quarterly based on application criticality.....

Validation is done by comparing the built-in temperature readings using a certified industry standard temperature meter...This reference meter in turn has passed an annual testing and calibration. with an NIST traceable certification issued by a certified independent testing laboratory....

<...You need to...> Nope. Untrue. There is no obligation here to do anything in particular.

<...They have...> Then the <...They...> will have the standard against which the calibration is to be carried out. Otherwise the <...They...> would not be able to assure product quality and the installation and its documentation would not survive more than casual scrutiny..

<...has to...> By what obligation?

<...at that temperature...> Plus or minus what?

<...kept...>

At what location in the installation:

• by the entrance door or
• deep within the product stack,

for example?

Perhaps some freezers used to calibrate other freezers.--_{in pluto}

Or some tweezers could be used for calibrating other brain tweezers (not teasers).

Perhaps some freezers just need verification as most all of us in here agreed to.

Is this what you are referring to ? :

Thanks, but I do not see any reference to a single point calibration in there.

Its not called calibration at a single point its called routine check.

That is part of validation. The validation checks to see if the temperature can be maintained over a period of time. Calibration verifies that the reading at a single point in time is accurate.

How is it done at present, then?

Or is the facility shipping uncontrolled product at present? If so, then how long can it get away with this state of affairs?

First, a unit must be "calibrateable" to be calibrated. It sounds as if this is not one of those.

Most critical equipment requires a calibration verification by a certified organization or individual on a yearly basis. This is simply a "certified" measuring device placed in the unit and verifying that it is indeed within the stated specification.

This is usually coupled with an in house weekly, daily, hourly, per run, verification of the units function. Simply placing a testing device such as a recording thermoprobe in the freezer for a specified amount of time and recording the results. If in spec. Ya! if not, corrective action is needed per your "out of Spec" procedures.

An actual calibration of a piece of equipment is different than the calibration of a measuring device. Equipment is something more involved and usually includes two set points to establish the acceptable range. This term "acceptable range" is your stated, validated, controlled and documented, requirement per SOP, GMP, GLP, etc.

In the case of stand alone, laboratory freezers, they will come with this validation information and if you are buying at the Pharma level ($$$) a complete, NIST documented calibration. The company will usually offer calibration services as an extra cost item.

Calibration. This is the actual, physical act of changing the sensors, logic, or control range of a device to keep it within the previously mentioned, documented, specification. This is required by FDA, USDA, you name it, regulation.

You can "establish" any range you wish for a piece of equipment as long as you can document that the range is achievable, achievable consistently, repeatable by any testing organization, documentable, and meets your industries standards for the purpose or process. if there are no industry standards then it is up to your QC/QA dept. to establish this standard.

So, step one, establish the range that is acceptable for the unit. (Caution, do not set this too tight or you will be sorry)

Step two, remove all product from unit and place several certified thermoprobes with a valid calibration certificates in various locations in the unit and allow it to run undisturbed for 24 hours. Check the recorded temperature for the test period and see if the unit is capable of holding the temperature within your expected range.

Step Three, document, document, document, As long as your unit is routinely checked and within this set range it is considered good. That is all you can do if your unit is a "factory set unit".

There is a whole other level that I won't get into here if you are working with temperature critical materials. Then you need to determine Temperature rise rates, Open condition times, recovery periods, entries per hour/shift/day. A whole world of details that in a pharma product environment are required.

To calibrate it, you normally would have a master gauge that is certified, and use that for calibration.

Of course if there air movement within the freezer itself, one can 'assume?' that is the temperature through out the freezer,... If the required temperature is -20 with no air movement and if there isn't air movement, I would calibrate it by the thermostat.

With no air movement depending how tall your freezer is, and product movement in and out of the freezer, the temperature may be stratified.

And if your thermostat does not match your master gauge, calibration is required... how, I can't answer that.

You can have constant verification with a calibrated chart recorder. Cheap to install and operate. Charts can be hourly/daily/weekly/etc. There are electronic versions that can maintain a log for a very long time without needing to change charts. Both styles can be set up with alarms when the temperature varies from the set point.

good idea, plus it would give a history of any issues.

I have seen single point cals done on freezers and refrigerators used for materials in pharmaceuticals. I'm not sure single point cals would pass an audit.

The screen shot from the FDA's Office of Regulatory Affairs (ORA) shows that single point cals are acceptable for specific instruments: working and reference hydrometers, but freezers and refrigerators have a different requirement.

The entire document is at
https://www.fda.gov/downloads/ScienceResearch/FieldScience/LaboratoryManual/UCM092154.pdf

Thanks for the input guys. Let me give you a little more history. I have been a metrologist for over 30 years. I know how to calibrate. As for any external devices, that will not work. They have the unit hooked up to the building maintenance system. That gives them 24 hour coverage and alarms in case it goes out of specifications. The probes are not removable without some serious work. The whole question here is simple single point calibration and is it acceptable.

Single point verification of freezer system:

(1) toss in at least two temperature loggers and record a minimum of 24-hours for two separated locations inside the freezers.

(2) introduce a temperature transmitter, that you have already calibrated accurately to a NIST standard thermometer over the range of perhaps -25 to -15 °C. The error of your secondary standard transmitter must be within the limits acceptable to your freezer calibration, i.e., if freezer must hold ±2 °C, then 0.5 °C accuracy of your testing equipment is deemed suitable to 4 standard deviations (right?).

If the logged data range and other statistics show the range is being met by the freezer, then all you need look further to is any offset between freezer display and the secondary standard transmitter. I would locate the secondary standard transmitter probe as near as practical to the internal temperature probe of the freezer.

Anything is acceptable if you write the procedure correctly. The old adage of write down what you do and do what you write down carries a long way with auditors.

Write your own procedure for your equipment. Be as specific as possible. Devise a checklist for your calibration (I would call it a verification) that documents each step. Carefully file them away.

In the event an auditor does not like what you are doing it is up to them to cite the spec to which you are not adherent. Then you have something to work with.

The short version of this plan is Baffle them with BS. ;)

...or dazzle them with brilliance!!!

I don't see why not. This is a storage freezer so I wouldn't expect the temperature to be cycling all over the place in normal operation with the door closed, nor would I expect the near constant stabilized temperature of the freezer to be warmer unless there was a problem with the freezer seal, etc.

If the freezer stabilizes at -20 deg C and you need it to be under -15 deg C and your measuring and calibrated at -20 deg C then you should be fine. If the freezer fails in some way (eg-door seal, compressor failure, etc) then the temperature should gradually rise triggering an alarm since this is monitored. A 1 deg C (for example) measurement error is unimportant in this application.

If this is a very critical application you may need to consider adding an additional removable temperature probe suitably calibrated, either permanently fixed in position or used to perform the calibration on the existing temperature probe.

be very careful if and when making a new perforation for the new temperature sensor well. You need to at least be able to prove to yourself, the additional BTU leakage in will not perturb control out of normal.

That is what the little "disposable" temperature loggers are for.

I wasn't thinking of going to the trouble of a well for a freezer application. Obviously sealing when passing through the walls is of importance.

We are already monitoring with a second probe that is attached to the building maintenance system. The question is not how to calibrate. I know how to do that. The question is if there are any regulations that allow it at a single point. I am trying to rewrite our procedure, but our QA will not allow me to go to a single point unless I can prove it is acceptable to calibrate at a single point.

Ask them to prove it is not acceptable . . .

Actually, change the language. As has been said several times, you are not calibrating this system, you are merely verifying that the instrument reading is correct. You can only do that at a single point . . .

Verification testing can be as elaborate or not, but one it is, it is always one thing.

depends on device type, if it is resistance type device, and not voltage producing, IMHO, the characteristics of the device will be homogeneous for the device type, with offset as the only necessary calibration - single point.

• <...We...>
• <...I...me...>
• <...our...>

It sounds as though there is a lot of defining to do, including the <...acceptable...> criteria.

For this situation, single-point calibration probably should be considered adequate. But that doesn't mean that everyone will consider it adequate. For instance, overactive customers or bureaucrats might require more, whether or not reasonable.

If the sensor is expected to be accurate over a wide temperature range, then zero and span would be required. But for a narrow range, no normal sensor will drift much, if single-point calibrated at the principal temperature of interest.

<...official document...> What does that mean? What is an un<...official document...>? Aren't all documents <...official...>? Which office, in particular?

This thread currently scores very highly on the <...official...> gobbledeygook scale.

When you write a love letter and try to sneak it to your intended it is an unofficial document

When the teacher intercepts it and reads it in front of class, then it becomes an official document.

It is still considered bad form to get caught.

This would be a calibration, not a verification. Although in this case there is little to no difference other than the fact that the units need to be calibrated on a regular basis according to internal documents. As for the word "official" that means from an entity that is official. They will not accept a procedure from another company, but would accept something from a USP or IEC or similar.

And the TLAs mean what, please?

<...They...>?

<...an entity that is official...> What [TF] does that mean?

This thread is still in the realm of la-la land...

Seems to me that you are trying to verify that the temperature displayed is close to the actual temperature AT THE PROBE. So to verify that it can ONLY be a single point measurement AT THE PROBE with a calibrated device. All official requirements would, i expect, relate to the freezer performance of which accuracy of the display may be only one aspect.

Jim