How do you determine the % tolerance allowed in the mass balance of your traceability of packaging materials?

Hi all, how do you determine the % tolerance allowed in the mass balance of your traceability of packaging materials?

I feel like it should be 5%, but I don't have any sources to back that up.

Thanks!

let me first understand the question , you  want to develop a specification for packaging material with some sort of tolerance , where the tolerance can account for all the uncertainties . 

This is required when doing reverse traceability from product to RM  ? 

My approach would be like this ,

First develop a sampling plan for the container / bag weight . If you get a 500 bags and you sample ten , you should measure the uncertainty with it . Then figure out the standard deviation of your result and divide by the square root of the sample size . This gives you the error , now add the uncertainty to the error. In  theory this should solve all your problems . But , we don't live in a perfect world as well . 

So here is what is to be done , collect the result and plot a moving I chart and note down the SD for it , generally any value under 1 SD is good , but above 1 SD it warrants an investigative procedure . 

So document it as a variable or revise the specification once in a month or so until you have a stable system .  now take the data overtime and plot a regression curve , this will tell you how  it will change over  time , ideally the r value should be near to zero. even though it is zero your sum of residuals should be random , if you see a pattern , you need to understand there is a error which is happening . 

Do an internal verification with 1 SD ( standard deviation ) and check the result , if it is good , document it . If not, you need to collect more data over a period of time . This can be communicated to  your packing material  supplier as well for better control over material. 

Hope this helps .

Hi all, how do you determine the % tolerance allowed in the mass balance of your traceability of packaging materials?

 

I feel like it should be 5%, but I don't have any sources to back that up.

 

Thanks!

Hi brynecox,

Is this for Food Production (as Forum posted) or Packaging ?

I'm not a SQF user but I could not find any specific reference to "mass balance" in either of SQF Food or Packaging Standards although a few SQF Food posts here claim that mass balance  is  expected to be done within "Product Traceability "?

?

I also see no mention of "mass balance" in the SQF Food/Packaging Guideline material ?

Regardless, continuing -

There are many, mostly BRC, "mass balance" related food threads  on the Forum (eg see links below) but I only found one for Packaging (5th link below).

As illustrated in the large range of numbers for mass balance results in the links below, I anticipate that the "tolerance" conclusions for BRC, SQF, etc for both Food and Packaging will be similar, eg "it all depends". :smile:

https://www.ifsqn.co...all/#entry35984

https://www.ifsqn.co...drawal-for-brc/

https://www.ifsqn.co...nce/#entry29566

https://www.ifsqn.co...uring-an-audit/

https://www.ifsqn.co...e-mass-balance/

By "mass balance" I assume you refer to the type of calculation exampled in this BRC  (food) Guideline -

BRC - F054.pdf   5.89MB   69 downloads

Note the concluding caveat -

It is unlikely that the mass balance check will account for all materials with 100% accuracy. However, packed final product should never exceed the quantities of the raw material batches. The company shall justify any discrepancies and understand the nature of the variance. This may be inherent in the product characteristics, for example, the dehydration of fresh ingredients or portion variances.

(However, the above "never" is not aligned with an NSF document indicating an acceptable recovery range of 99.5 - 105% in  Post 3 of 2nd link above)

Here is another, BRC8 FAQ, Food  quote -

CLAUSE 3.9.3: IS 100% TRACEABILITY RECOVERY NO LONGER REQUIRED? (THE WORD ‘FULL’ WAS REMOVED FROM CLAUSE)?

 

The word ‘full’ was removed from this clause as it was a common cause of confusion during Issue 7.  However, the meaning and intent of  the clause remains unchanged from Issue 7 to Issue 8. It is unlikely that the mass balance check will be able to account for all materials to an accuracy of 100%; however, the company needs to be able to justify any discrepancies and demonstrate that it understands the nature of the variance (e.g. through processing factors such as dehydration of fresh ingredients, typical wastage on equipment, or portion variances).

The Standard requires the site to test traceability ‘across the range of products’ each year. Where the traceability for all products manufactured by the site is the same or similar, then a minimum of one traceability test a year must be completed. However, if there are significant differences or specific traceability challenges relating to one product or a group of products, this may necessitate additional tests specifically related to that product or group of products.

In addition, where the site makes a product claim, the requirements of clause 5.4.4 apply (i.e. traceability tests should occur at a frequency to meet any particular scheme requirements or at least every 6 months in the absence of a scheme-specific requirement)

 

and from BRC I.Guide -

The principle [of mass balance] is to ensure that the traceability system is effective. Mass balance is a key measure of its workability and highlights areas for improvement.

So, offhand, I suggest that for SQF no specific %tolerance exists. Welcome to be corrected of course. :smile:

it is for food production. The costco addendum makes us account for the food contact packaging used. 

We are currently trying to figure out what is a reasonable recovery rate for the lids during traceability exercises.

it is for food production. The costco addendum makes us account for the food contact packaging used. 

 

We are currently trying to figure out what is a reasonable recovery rate for the lids during traceability exercises.

Hi brynecox,

Basically, as from previous post(s,) it should ultimately be more a question of having the data to  justify  yr net result.

Nonetheless, I would  have thought that wastage in actual packaging material would be "minimal" ?