was successfully added to your cart.


Understanding average weights…

Average weights is a complicated subject and perhaps this is why many sites procedures do not comply with the legislation. In this post I’ll break down the rules, so that you can check your systems to make sure your procedures and records comply.

average weights

Minimum weight or average weight?

If at all possible I would always advise sites to run their weight systems by minimum weight rather than average weight, as it makes things so much simpler!  However, to run to minimum weight you need to check your packaging first (or specification if you produce product for further processing). If your pack (or spec) has the e mark, then this indicates average weights, so you would have follow the average weight rules that I’ll go through below.

If there is no e mark then you can run to minimum weight, which means that you need to carry out weight checks, but the frequency and sample rate is up to you – as long as the pack weight is at least the minimum weight, that’s all you need to prove.

What does it mean to run average weight?

Where your product has an e mark, it is a legal requirement that you meet the average weight rules and be able to prove (through records) that you have done so. The average weight rules are split into what they call the ‘3 packers rules’.  If trading standards were to visit you, they would want to see evidence of your compliance all 3 packers rules.  Generally, when I visit sites they usually meet the average weight rule, but forget to comply with the other two rules.

The 3 packers rules are:
  1. The average of the batch must be at least the nominal weight
  2. There must be no more than 2.5% T1s
  3. There must be no T2s in the batch

It sounds complicated and it is a little, but we can break it down to try and explain it.  Before we explain the 3 packers rules, there are 2 really important points to understand first:

  • You must define what you mean by a ‘batch’
  • You have to work out what sample size to apply

It’s important that we understand what ‘a batch’ is, as there are a few rules around what can be defined as a batch.

A batch must be:

  • The same product (recipe and weight)
  • Made on the same day
  • Made in the same place (line)
  • From one production run
  • Be no more than 10,000 units
  • When checks are carried out at the end of the line, the maximum number for a batch is the hourly throughput

Once we’ve established our batch, we must work out how many samples are required to be assessed per batch.

The legislation states that the sampling rate must be at least 50 packs per 10,000 packs produced, the minimum time to produce 10,000 packs is 1 hour and the maximum allowed is 1 day. Here’s a guide:

  • If you make less than 10,000 units per day, then you must sample 50 units across your run.
  • If you make more than 10,000 units per day, you need sample 50 packs from each 10,000 units made.
  • So if your throughput was 10,000 units per hour you would need to sample 50 packs per hour.
  • If your throughput is 30,000 units per hour, then you’d need to sample 50 packs every 20 minutes (1 hour divided by 3).

3 Packers Rules:

Now you know what your batch is and how many samples you need to check in that batch per hour, then you need to confirm for each batch that you’ve met the 3 packers rules. I’ll go through each one individually.


For all the samples that you’ve taken you need to confirm that the average weight is the same or greater than the weight declared on pack (known as the nominal weight). So; add up all the weights of the samples taken and divide the total weight by the number of samples, this will give you the average weight for the samples taken.

RULE 2 – The batch must not have more than 2.4% -T1’s

So, now we need to understand what is meant by ‘-T1’.  The legislation defines ‘T’ as the tolerable negative error, but just try to think of it as a set quantity that is taken off and added onto the weight declared on pack. First you need to find out what ‘T’ should be for your pack size. The legislation provides this table:

Tolerable negative error

Nominal quantity in grams and millilitres

As a % of nominal quantity

g or ml

5 to 50


from 50 to 100


from 100 to 200


from 200 to 300


from 300 to 500


from 500 to 1,000


from 1,000 to 10,000


from 10,000 to 15,000


above 15,000


In order to work out what ‘T’ would be for your pack, you would need to look up your pack size in the left hand column and then look across to see whether a % or g/ml is provided.

WORKED EXAMPLE 1: – Your pack size: 75g

Let’s go with the g or ml column first as that’s easier.  If your pack size was 75g for example, it would sit between 50 and 100g, reading across the columns on the table above – you’ll see the tolerable negative error for that pack size is 4.5g.

So T = 4.5g.  So, T1 (which is T x 1) is 4.5g x 1 = 4.5g.

Therefore to work out –T1 would be 75g (for your pack weight) – T1; 75g – 4.5g = 70.5g.

So for this example, to comply with rule 2, that no more than 2.5% of your samples must be –T1, means that – no more than 2.5% of your sample must weigh less than 70.5g.

So to do this, you would need to work out what –T1 is for your pack size and then, highlight how many of your samples weighed from your batch weigh less than your –T1 figure.  Then work out what % this is per batch.

WORKED EXAMPLE 2: – Your pack size: 450g

Going back to the table above, you’ll see that if your pack size is between 300 and 500g and read across the table, you’ll see that you’re not provided with a g/ml number but a %.  For a pack of 300 – 500g it’s 3%.  This means you need to work out what your T is by working out what 3% of your pack size is.  So for a pack of 450g it’s 13.5g, so T would be 13.5g.  –T1 would be 450g – 13.5g = 436.5g.

To help you with working out your –T1 and also –T2 (which we’ll come on to below), I’ve created an excel spreadsheet, which you can just pop your figures into and it’ll do the rest for you!  To get your free average weight calculator – fill out out form below and we will send you one by email. We will then add you to our Quantity Control list to keep you updated with advice and info for this area of compliance in BRC Issue 8.

RULE 3 – None of the packs must be -T2

So, we’ve worked out what the –T1 is above, now we need to work out what T2 is.  Easy – it’s just your T figure x 2.

So, -T2 is your T2 figure taken away from your declared weight on pack.

To comply with this rule, you must make sure that none of your packs weigh –T2 or less.  Any of your samples that you weigh that are –T2 must be disposed of, or re-worked.  If you get –T2’s in your batch you should also apply corrective action to make sure you increase your sampling, in case there are other –T2’s in your batch.  In the worst case, you would need to check all of the packs produced to ensure that you don’t send out any –T2 packs, as Trading Standards would see these as ‘illegal’ packs.

Signing off the batch

For each batch produced you must positively confirm that you’ve adhered to the 3 packers rules. This means you need to work out the average weight of the batch and record it. Then sign it off to say it conforms (if it does).

  • Calculate the % of -T1’s in the batch and again sign if off to say it conforms if it’s less than 2.5%.
  • Confirm that there have been no -T2’s for the batch and sign it off.

Where any of the batches do not pass these rules, corrective action must be applied.  The aim of the corrective action must be to ensure that the batch is adjusted to make sure that it meets the 3 packers rules before it is dispatched.  You would need to ensure that all the –T2’s are removed, that there are no more than 2.5% -T1’s and that the average weight of the batch is at least that on pack.  To do this, you can go through more samples (depending on the scale of the problem) and record the results, or go through all the packs and record the results.

Working with volume rather than weight

If you declare your product in volume rather than weight (ml or litres), you will probably check the weight (and apply the 3 packers rules) in weight.  It is important if you do this that you know the density of your product, as you’ll need to convert your weight to volume to be able to determine if the weight is correct.

I hope that’s made sense, I think you’ll really find the average weight calculator useful!  If you have any questions please get in touch as always. We have created a Quantity Control Documentation Pack compliant to Issue 8 of the BRC Standard too, so if you would like to order this, click our link below to find out more…


  • Athanasios Blanas says:

    Thank you very much. Easy and comprehensive.

  • Charles B says:

    Very informative. thanks for the easy break

  • Alex says:

    Hi Kassy,
    say your units “per run” is only 1000 – do you sample proportionately or do you still have to do 50? Also, what if you have a multi-head filler – does this factor into it? We have a range of fillers from 1 filling head to 48 filling heads?

    • Kassy Marsh says:

      Hi Alex,
      I’m afraid you’d still need to do 50. If you have 48 filling heads, in theory you should be able to show that you’ve sampled them all – because each head could be filling differently. Perhaps one rotation of the filler – gives you 48 samples per run. Producing just a 1000 samples on a 48 head filler must have it’s problems, so I can appreciate this probably isn’t the answer you’d like 🙂

  • Clem Griffiths says:

    Hello Kassy when you say ” unit” do mean retail unit as in a can of beans, a package of bacon or a bottle of pop? . My application is bread , frozen dough to be exact and I am running a slow line packing 60 pieces into each case. Probably doing about 100 cases per hour. My strategy was to take a case (60 pieces) at start , middle and end of run and do a verification check prior to shipping. This means N=180.
    My nominal or declare weight is 90 grams, what do you think?

    • Kassy Marsh says:

      Hi Clem,
      A unit for average weight, is the unit where the weight is listed. For example, a retail pack would have the nominal weight listed on the pack, so one unit is one product. If you are supplying a box of frozen dough, where is the nominal weight listed? Is it on the box or on the specification for the individual frozen doughs. It’s the nominal weight that you have to be able to prove. Are you sure you are working to average weight on your frozen dough? Typically, product that goes for further processing would not need average weight, and so, you’d work to minimum weight.
      The number of samples you need to take depends on how big the production run is. 60 pieces would be a minimum for a run under 10,000 units. How long would it take you to produce 10,000 units? This will obviously depend on whether a unit is a box, or whether there are 60 units in a box…
      Let me know and I’ll try to help further.

  • Clem Griffiths says:

    Thanks, I am declaring 90 grams but as you correctly suspected it is the retail customer that is taking that case of 60 pieces and repackaging it as they deem appropriate lets say into 3-packs. What I want to do is validate that the process delivers pieces that after final in store bake off will be 90 grams or more as per my specification. It is the Retailer who will have to deal with weights and measures.
    Thanks again

  • adam says:

    Easily understandable and well written. It’s amazing what the brain remembers.

  • karlie litchfield says:

    Hi Kassy,

    Is the weight of the packaging taken into consideration when calculating the T1/T2? Or is it solely on the contents of the packaging?
    Thank you in advance for your help

  • John says:

    Hi Kassy

    How would you recommend keeping records of compliance to packer rule 2 and 3 when using manual checks? If I weight 10 packs per hour manually, 1 pack between T1 and T2 would straight away give 10% when the limit is 2.5%.

    Best regards

    • Kassy Marsh says:

      Hi John
      Keeping records of compliance to each of the packers rules is a must, so that you can prove that the product that you have produced in the past was compliant. You would need to prove this to Trading Standards on an inspection. Remember, you 2.5% is across your batch. So, if you weigh 10 packs an hour, yes you are right 1 pack would be 10%, but if you did a run of 4 hours on the same batch, then 1 pack over the 4 hours would be 2.5%.
      Thanks, Kassy

  • Renata says:

    Hi Kassy
    Is there any legislation / guidance on the give away weight?
    Thank you


    • Kassy Marsh says:

      Hi Renata
      If you follow the letter of the law, you shouldn’t ‘give away’ more than 2.5% T1 or any T2’s. But in practice, Trading Standards are not going to worry if you are giving the customer too much. Unless, that is, you are producing a product with a nutritional claim, where the excess product could mean you are not meeting the claim.
      Thanks, Kassy

  • Alex says:

    Your definition of what a batch is is very helpful – is this an official recognised list or not?

  • Alex says:


    From my previous comment I meant the below list in particular – is this more of a guideline or a recognised definition by governments?


    It’s important that we understand what ‘a batch’ is, as there are a few rules around what can be defined as a batch.

    A batch must be:
    •The same product (recipe and weight)

    etc. etc.

  • Keith Barrasford says:

    Hi Kassy
    Thanks for the above – very good explanation.
    One question I have is:
    We are planning to pack 1.5 litres of product into saleable units and use the minimum weight as packaging does not have the average logo.
    We will weigh every unit and ensure that these are all above our calculated minimum of 1536 ( product and package ) or will be rejected.
    These will then be packed in cases of 6. If we record the weight of 1 individual unit every 15 minutes along with 1 case of 6 is that sufficient
    to meet the requirement – potentially we will produce 5000 individual units every 24 hours.

    • Kassy Marsh says:

      Hi Keith,
      You’ll be pleased to know – that if you’re not declaring your product as being packed to average weight (by adding the e Mark) then you don’t need to comply to the average weight regs. So you can define your own sampling plan, which means what you are suggesting would be fine.

  • Rick Nadal says:

    Good explanation.
    Thank you!

  • Laura says:

    Hi Kassi

    Is there a need for T1’s to be destroyed? I have procedures from a previous business that states T1s should be destroyed as well as T2. Was the technical team there being overly cautious?

    • Kassy Marsh says:

      Hi Laura,
      You can’t have more than 2.5% T1’s in the batch. Therefore, if your monitoring shows (during production of the same batch) that 2.5% or more has been achieved, then you could in theory increase the dough weight at the divider and/or increase the batch size (i.e keep producing) to ensure that the perecentage of T1’s is reduced across the whole batch.
      If this isn’t possible, then the right number of T1’s would need to be wasted to ensure that there were no more than 2.5% in the batch.
      Does that make sense?

  • Matthew Richards says:

    Hi. I used to run a line making ready meals. On a minimum weight system,the average weight is the number of kilograms produced divided by the number of meals produced and the more meals produced let’s say 5000 meals and you altered the weight slightly then it would take time for the average weight to lower as compared to less meals produced.
    Now on nominal weight program, it doesn’t follow that rule. …Lets say i produced 5000 meals and the number of kgs produced was 2,040,000 then the average weight would be 408 grams and should take a while for it to lower if you altered the weights less but it doesn’t ….the average weight changes every few minutes. ..just wondered why. Hope you get in touch
    Kind regards

    • Kassy Marsh says:

      Hi Matthew,
      I’d be happy to help, but I’m not sure I understand what you’re asking. Are you saying that if you reduce the weight of the product, the average weight on your system should come down slowly and it doesn’t because it goes up and down every few minutes? If that’s the case, it would either mean that the weights of the product being checked varied massively (a wide enough weight range to change the average weight quickly), the batch size in the machine isn’t set up right or there’s something wrong with it!
      Let me know if I’ve misunderstood 🙂

      • Richard says:

        You are correct, the best way of determining whether “the filling Line”, is suitable for average weight packages is to measure the standard deviation of the sample and as in this case it’s greater than -T1 on the nominal then he needs to correct it. The packer has a legal duty to ensure the packs he produces meet the 3 packers rules, and T2’s are a failure of the packing process and all packages in that BATCH should be recalled and every package re weighed. Each packer in the UK is duty and must have their system checked by local trading standards annually and the FILLING SYSTEM tested to ensure it PASSES a “reference test”, otherwise he should pack to MINIMUM weight regulations.

  • Larry says:

    Hi Kassy,
    Thanks for this information
    We run a spirits bottling line @ 200 bottles per minute both with/ without the e-mark. ( 192,000 bottles a day)
    What quantity of bottles should you check for fill level & how often should you do it?
    Thanks for the help

  • Andria says:

    Hi Kassy,
    We are a small microdistillery. Our production run is an average of around 1344 bottles at 350ml. Production takes around one day as the process is mostly manual. We do not have the ‘e’ mark on our bottles. We have 4 pneumatic filling heads. Currently we do 12 samples in all, 3 each for each filling head at the beginning of each run. Are we currently sampling enough? What would be an ideal sample plan for a production run of this size and what would be the best way to record our sampling?

    Thank you.

    P.S: I have signed up to the news letter and not sure where I can download the calculator mentioned above.

    • Kassy Marsh says:

      I think Mel’s helped you out with the calculator, so hopefully you’ll have that now. If you do not have the e mark on your bottles, you don’t have to run to average weight and therefore, you can pick the sample size that you think works for you best.

  • Richard Ross says:

    The above post is very interesting, and minimum weight is simpler , but you have failed to put across is hat since 2006, all weighing equipment used for minimum / average weight MUST be trade approved, and that the packer MUST ensure their company is compliant with the regulations. The problem in UK is both sets of regulations are in force.
    The 2006 average weight regulations allows smaller packers to pack to average weight without the need of recording provided ALL packages in if the BATCH size is less than 100 items provided all packages are passed over a suitable approved scale thereby providing smaller producers an opportunity to export to EC .
    The Local Trading Standards in UK are responsible for applying the law and sampling processes must be agreed with them to ensure compliance with the 2006 regulations for average weight, and I would suggest you packers contact them, as some have now become Primary Authorities – link https://primary-authority.beis.gov.uk/about
    Hope this helps your clients

  • Patricia Ababio says:

    Thanks so much. Very easy to understand.

  • Scott says:

    What’s the view on products that change weight after packing. For example, if at point of pack, product batch complies with average weight legislation, but during distribution/ shelf life, net product weight changes, due to for example moisture loss, oil migration etc, and may lead to products on sale that fall outside the legislation. Does this have to be compensated for?. UK Legislation states at point of pack, WELMEC guidance (which TSO’s generally use) ‘recommends’ adherence throughout distribution.

    • Kassy Marsh says:

      My view would be that the pack should weigh the nominal weight declared on pack, at the point at which the customer buys it. However, legally you would have a defence if challenged, if you could prove that the packs complied with the 3 packers rules…

  • Fiona says:

    Hi Kassy,
    Great article really helpful,

    One question when I was looking at the EC Directive the table that works out the ‘tolerable negative error’ is divided into Class A and Class B? I work in a bakery and just wanted to check I’m going by the right category?


  • Jerome Cowan says:

    The 75g – 4.5g example is incorrect. The answer should be 70.5g.

  • Emily says:

    Do any other manufacturers create their own T2 test packs to ensure their machinery is working correctly?
    If so, how do you go about this?

    • Kassy Marsh says:

      Yes and I think your online checkweigher company may provide these too. It doesn’t need to be actual product, it just needs to be the right shape, size and – weight of course! I would recommend you speak to the manufacturer first, as they may have the easiest solution for you.

  • Rebecca says:

    First off great article – thanks.

    Does anyone do small runs and average weights and is it worth it or even do-able? For the area we can change to average weights, some batches are <50 units, most are 50-300 and the biggest 500. We packs lots of different 100g packs of produce currently to minimum weights. We use the Marco weighing system for context.

    • Kassy Marsh says:

      Hi, if you’re doing minimum weight and you can stick to that, I would highly recommend doing that. I’d only do average weight really if you have to – either for your customers or if you’re exporting product. Thanks, Kassy

  • Rebecca says:

    HI Kassy, I’ve done a bit of extra research since my original post. For batches <100 you have to have a sampling rate of 100% so this makes it not worth it.
    For actual calculations and sticking to the statistics etc the Marco (and I assume similar systems) do all of this for you and along with the sampling scales make compliance easy. It was just the cost/benefit of small runs vs savings in using the average weight rather than minimum that I was querying.

Leave a Reply