3 replies to this topic

[QM-OS]

Hi, don't know if this is the right part of the forum, but I''ll give it a try.

 

Would like some input about frequency and amout of quality controls during production.

 

We produce large batches at a time, around 100-400 tonnes (typically 200 000 kg/week).

Samples are taken once per hour to check the quality, so that produced goods are in accordance with specification.

 

If you count on a batch of about 200 tonnes, that means about 46 samples. At the moment there is 300 grams per sample = 14 kg in total. Is that really enough? 14 kg out of 200 000 kg?

The number of samples, 1 per hour sounds reasonable, but the amount of quantity I think borders on too little.  

The "300 grams per sample-rule" is something that's been always been in force at this company, ever since production started, so no one has really questioned it before now.

 

What are your thoughts?

[BrummyJim]

It depends on what you use your samples for.

• Accelerated shelf life
• Micro testing
• References
• Authenticity
• CoA analyses

Micro typically requires 10g/test (plus an amount if re-testing is required) so 100g here

Shelf life would be another 50g

Reference samples to compare against customer complaints might be another 100g, possibly more depending on granularity

Authenticity/CoA analysis might be another 50 - 100g

 

If all the above, you might be a little short.

[QM-OS]

It depends on what you use your samples for.

• Accelerated shelf life
• Micro testing
• References
• Authenticity
• CoA analyses

Micro typically requires 10g/test (plus an amount if re-testing is required) so 100g here

Shelf life would be another 50g

Reference samples to compare against customer complaints might be another 100g, possibly more depending on granularity

Authenticity/CoA analysis might be another 50 - 100g

 

If all the above, you might be a little short.

 

We use external laboratories for micro, heavy metals, pesticides etc. So these quantities are not part of previous mentioned 300 gr samples.

(We take almost 1000 gr for analysis of each batch)

Samples for shelf life and reference are not included either. We save between 500-1000 gr of each batch for that.

 

 

I felt that out of 200 tons we "only" check about 26 kg against the specification tolerance (% of other plant materials, % of damaged/unripe/over riped pieces etc)

To me that sounds little but maybe it isn't?

[BrummyJim]

If your 300g is enough for your assesment of the product against the specification, then it is a sufficient sample. The next challenge is to determine if an hourly check is enough. A risk assessment is a good place to start. What is the variability within the product over time? For example, I used to work with ice cream which was made in 7 tonne batches and stored in tanks. A day's run was up to 4 tanks. The tank was continuously agitated which kept it homogeneous, so we were testing each tank for homogeneity as well as comparing each batch.

 

Look at the process and your experience.