Verification of checkweighers
We are currently verifying our checkweighers at 30 minute intervals using test packs. The test packs are the same size as product and are periodically replaced. The test packs are made up of a pack without the product contents but with a 'filler' to the weight that is required.
This is in addition to start up checks where a calibrated weight is used.
The checkweighers reject at T1 and we have never had an issue with underweights
We feel that what we are doing is somewhat over the top but as the original staff who set the system up some years ago are no longer with the company there is no one to confirm why it was decided to verify at 30 min intervals
Can anyone advise what the practice is for verifying the checkweighers on their sites and whether or not you use test packs.
Thanks
Hi Classic,
I would suggest you try a little searching for checkweighers.
There are 2 possible aspects -
(1) whether the procedure must conform to the details of the UK weights and measures regulations. There were extremely detailed discussions here over this general topic about 4-5 years ago. I can recall that the requirements were complex and the relevant documents were all uploaded here.
(2) If the query is independent of (1) and confined to generically verifying a checkweigher, a table of typical calibration intervals (possibly vs weight range involved) as recommended by a manufacturer was also uploaded here a few years back. I recall it followed the usual practice of initially starting at an elevated frequency whch was then gradually decreased if satisfactory check results were obtained.
Offhand, assuming the calibration is stable, it sounds over the top to me also. But perhaps depending on the value of the Goods. :smile:
addendum to previous -
The post i was thinking of is probably -
http://www.ifsqn.com...ion/#entry36561
welmec document is, i think, including checkweigher requirements as related to weights and measures regs.
Later sartorius link is a basic checkweigher calibration schedule from memory. Sadly the link is long-gone, will post back if in my archives.
(added - found but not relevant)
Assuming aspect (2) in previous post, some support for a daily check here -
UKAS,LAB14, Calibration weighing machines.pdf 108.37KB 253 downloads
(eg sections 3.12, 6)
A brief overview here -
http://www.scsukltd.com/weightlaw
------------------------------------------------------------------------------
This is a recognized method (amongst several) for revising calibration intervals -
Method 1: Automatic adjustment or “staircase” (calendar-time)
Each time an instrument is calibrated on a routine basis, the subsequent interval is extended if it is found to be within e.g. 80 % of the maximum permissible error that is required for measurement, or reduced if it is found to be outside this maximum permissible error. This “staircase” response may produce a rapid adjustment of intervals and is easily carried out without clerical effort. When records are maintained and used, possible trouble with a group of instruments indicating the need for a technical modification, or preventive maintenance, will be known.
COP Calibration Industrial weighing systems.pdf 1.28MB 182 downloads
Guidelines determination Calibration Intervals.pdf 66.95KB 176 downloads
Hey Classic,
Yes, weight verification is a pain, especially in the UK. It is good news that your customer complaints on this issue are low. By the sounds of it, your company are doing a good job on this front.
If you are looking for an addition to your system, throw in a monthly or weekly test whereby you take 40 consecutive packs and weight them. It's just a statistical analysis, but it backs up both what you're lads on the floor are saying in addition to your customer's (lack of) feedback. Are all random packs within T1? If so, happy days. The regularity will depend on your throughput, so I'll leave that up to you!
Last point, you said you check at start up and every 30 minutes thereafter. Like metal detection, don't forget the finish. If your production lads carry out a test at 23:00 and finish at 23:20, who is to say your systems have been in control for the last 20 minutes?
So go on - tell us the gorey details of the ridiculous non-conformance!