11 replies to this topic

[Murae]

Hi All,

 

We have new machinery installed that will produce a product that is for pet food.

 

in the process of production the mixed product is formed then passes through a cooling system to harden the product, this product is packed and naturally softens in the warehouse in ambient temperatures.

 

We are BRC accredited and this new machine will be in the scope of our next audit my query is that they operators currently record temperature of the cooling system every hour and document. the limit of the cooler has a small tolerance of 6°c.

The problem is that the cooler temperatures changed regularly so that the end product comes out hardened, leaving the temperature at a set point may work one day/per production run but then not the next leaving a soft end product.

 

The recording is quite different to the tolerance and I have been asked weather we need to have a tolerance to aim for at all seeing as the final product being hardened is the sole aim of the cooler. 

 

The thoughts I have are if there is no target then recording seems to be futile.

If the tolerance is set to incorporate all that different temperatures that it has worked at then it will be a large margin 15°c

 

Any advice on this would be appreciated.

 

to note: the analytics of the end product are the same weather it is softer or hardened the sole purpose is to have a hardened product which makes the product not fall to pieces when being packed.

 

Murae

 

[Charles.C]

Hi Murae,

 

afaik, BRC includes some quality requirements in addition to safety.

 

Do you mean that the finished product specification / customer has no interest in whether the item is in pieces form or not ?

 

if otherwise and the result is "pieces" for some units, there would seem to be a quality problem ?

[Murae]

Hi Charles,

 

The customer(s) do not want the product in pieces which is why the product is passed through a cooler to ensure this does not happen.

 

The temperature of the cooler will be constantly adjusted to ensure the product is as intended (a whole item)

 

we record the temperature but the variance is a lot because the final product as a whole item is what the specification is.

 

what it means is as long as the end product is to spec then it doesn't matter what the temperature of the cooler is as, each time it is run to get the correct end product yelds a different temperature.

 

Murae

[Charles.C]

Hi Murae,

 

I deduce there is no known consistent relationship between cooler setting and occurrence of hard/soft output ?

 

In that case i do not see how a tolerance (in the statistical sense) can exist.

[Murae]

Hi Charles,

 

That is correct, there is no known relationship for temperature and output.

 

the tolerance was put in as was belief at the time that the temperatures would be within.

 

from a audit point of view how would an auditor look at it if there are temperatures recorded but with no outcomes

 

is it worth recording the temperature hourly even though there will be no outcomes based on this, the only temperature changes happen when the final product does not meet specification when it comes out of the cooler.

 

Murae

[Charles.C]

Hi Murae,

 

Bit difficult to recommend without knowing the method but It sounds to me that yr SOP for the cooler is basically a  ‘trial-and-error” technique.

 

However I anticipate that the operator has some kind of  “rule-of-thumb” for homing in on the appropriate setting on a lot-by-lot basis, eg he starts with some initial level based on experience and then adjusts up or down (somehow). Hopefully the method is not too hit-or-miss.

 

Perhaps you can write a “smoothed” Procedure on that basis to justify the recorded temperatures.

 

One difficulty may be if there is a high failure rate at output. Hopefully not.

 

Conceptually, this method is not so different to a long-famous  procedure of Chemical Process Optimization called EVOP.

[jkoratich712]

Is there a documented check of the product after the cooler on whether it meets the spec for hard/soft? I'm thinking that is what you document and then the temperature adjustment is made to ensure that the spec is made.

 

We do something similar with oven temperature - we document the temp at the beginning and end of a run. It is only used as a reference for the next time we run that product. The temperature can then be adjusted during the run based on the finished color of the product.

[Murae]

Hi Jkoratich,

 

We perform penetrometer tests on the product at kg/cm² to ensure they are at required target level.

As this product was something that we used to purchase, we carried out trial tests on these that gave us our target level, when we started out own trials with the product that we made we tested against the same level which gave similar results.

Trials performed show that the target level ensures correct product specification, if the target is not met then the temperature is changed and the process starts again.

 

Murae

[CMHeywood]

Can you state it as:  Target +/- 6C.  Target temperature is affected by environmental, ambient condtions.  Target temperature ranges from xC to yC  +/- 6.  Target temperature depends on the correct hardness of the product.

[Murae]

Hi CM Haywood,

 

What I was thinking of is still recording the temperature each hour, but not have a target statement.

 

I am still unsure if the temperature needs recording each hour if there is no know accurate temp to finished product match up.

 

would a start up and finish temperature be ok?

[BrummyJim]

Hi Murae

 

Is this akin to a pasteurisation process where time and temperature will cause the level of solidification you require? If so, it is important that you measure the relevant temperature.

 

What exactly are you measuring here? Is it the air temperature at a point in the cooling system, the surface temperature of the product, the core temperature of the product? Is dwell time an issue?

 

Jim

[Murae]

Hi BrummyJim

 

The recording is taken from the display of the cooler every hour currently, however this fluctuates and is changed accordingly so the end product is at the correct specification when tested using a penetrometer.

 

as stated previously initially there was a target range set in as this was what the temperature would be to produce the end product correctly. but carrying out trials on the product shows in 1 run the end product will be whole based on temperature x but the next run the end product could be in pieces/broken. therefore the temperature is changed so that the end product I whole/solid again.

 

the customers specification is only that the end products are whole, there is no requirement how hard they need to be etc. as they soften naturally, we only want them hard initially so that when going through the packing process they do not break.

 

Murae