To Calibrate or Not to Calibrate various Weight Scales

[Golfman25]

I have a client that has weigh scales in both their receiving department and mixing department. They use the scales for overall verification of weights when materials come in and when materials are mixed together in a vat. Although they verify that the weight measurements they are getting from the scales are in the ballpark by checking with objects of known weight, they are not concerned with the level of precision for either application. The mixed materials undergo specific visual inspections later on in the process and they have never had an issue where the mixed product had to be rejected. They have assessed the risks of mixing with a weigh scale used for reference only and have decided that the cost of verification/calibration is much more than the cost due to risk of product batch failure. Therefore, since product inspection related to proper mix of materials is carried out later on in the process, they have decided to remove the weigh stations from a calibration program and label them "not calibrated - use for reference only".

Personally, I agree with their reasoning, and their right to make this decision. As long as their product quality is protected through the product inspection later in the process, the organization should be able to weigh the risks/costs of calibration. Thier ISO 9001 Registrar may not agree.

What are your views?

Under the cirumstances you describe, I see absolutley no reason why these scales need to be calibrated. Weight of incomming product -- all you really need a reference measurement. Enough to make sure the vendor isn't off by more than a few lbs. (or kg, whatevery you use up north). Mixing ratios -- same thing assuming thier process doesn't require more precision (which sound like it doesn't).

Here is a perfect opportunity to keep your client's system simple -- don't put the scales in and see what happens. Why complicate things? However, you are right to be concerned about your Registrar. They tried to nick us -- we never really concerned ourselves with scales. They aren't that important in our process. We had them verified once per year by the original vendor. Guess what -- they where not a ISO certified "lab." Can you say potential nightmare? These are the areas in which the registration process and standard quickly can get out of hand. I wish your client luck.

 

[Jim Wynne]

Bluntly put, if the calibration of the item does not matter, then REMOVE the item from the process, but if the item actually matters to the process, then calibrate the item.

This is an answer to a question that wasn't asked. The OP says

Although they verify that the weight measurements they are getting from the scales are in the ballpark by checking with objects of known weight, they are not concerned with the level of precision for either application.

My emphasis. They're doing all of the calibration that they feel is necessary. No violation of the standard (or common sense) is in evidence.

 

[howste]

Re: To Calibrate or Not to Calibrate

I was having Deja Vu reading this thread, so I went and found the thread I remembered with almost this same discussion: Weighing (Weight) Scales Verification - Blank weight of material

It seems this discussion is always pretty polarized. Some people are adamant that the calibration must take place. Others don't believe they should be. From a requirements standpoint, I don't believe that all of the scales "shall" be calibrated. From a business standpoint, the answer may be different.

 

[bobdoering]

Personally, I agree with their reasoning, and their right to make this decision. As long as their product quality is protected through the product inspection later in the process, the organization should be able to weigh the risks/costs of calibration.

If the process never fails, and the mix has broad side of a barn tolerances, I would take 3 known weights (to track some linearity) and on some occasion (annually?) verify the scale is still reading "correct enough" (I doubt at this point they have any idea what the tolerance is, so this term is a real toss up). As long as those weights were verified at some time on a traceable scale, you have about the lowest form of calibration - and it is pretty darn reasonable in cost.

I would buy "used for reference" if they were actually using a volumetric approach for control, and questioned if the density was a problem. Or, if the mix was in bag quantity, and somebody was curious if the bag weight was correct. But, if you can not take the device out of the process (yes - that is a very nice test) then it really isn't "reference only".

 

[michaelkhart]

Yes, the issue is certainly contensious. In the end the client decided to continue with verification/calibration but revise the frequency of from one year to two, based on the history of past calibration results, and the added safety of product inspection further down the stream. The registrar auditor, during a maintenance audit today, agreed that this would not consitute a future nonconformance to ISO 9001.

Thank you all for your replies and opinions!