Loaded Question: Is a 50-60 % load for Stability Chambers too? (Or just Incubators?)

We recently received this question based on an article "Troubleshooting Sensor Positioning: Different Sensor Response Times & In Air or In Media."

Hi, In your "Troubleshooting..." article you wrote:

"Generally, an incubator should only be loaded to 50-60% of its actual possible physical volume. Otherwise there will not be enough space to allow for air flow and the resulting cooling/heating processes. (This is based on the French Standard: NFx15-140: Measurement of Air Moisture – Climatic and Thermostatic Chambers – Characterization and Verification)"

Would the same (50-60%) be true for a stability chamber, or does this apply specifically to incubators?
We are validating numerous stability chambers and we are considering stacking the shelves full end-to-end for full-load testing. At the very least, if the 50-60% was used for incubators or anything else, procedural controls would have to be strictly enforced so that the loading does not exceed this when in normal use. What are your thoughts on this?

Thanks,
Validation Engineer J

Paul Daniel answers:

Hi VEJ,
Thanks for emailing us! Yes, I would consider this to hold true for stability chambers, as well as incubators. The guidance we refer to (linked again below) is for "Climatic and Thermostatic Chambers", which includes incubators and stability chambers. The figure 50-60% is an estimate based on the parameters given in the French Standard: NFx15-140: Measurement of Air Moisture – Climatic and Thermostatic Chambers – Characterization and Verification. The exact words are closer to this:

Sensors should be placed at a distance from the walls equal to 10% of the given dimension. So in a 1 meter wide chamber, the sensors would be 10cm from the wall, and our sensors would be 80cm apart.

If we assume a 1 cubic meter chamber that is 1m on each side, then we can calculate the volume inside the sensors to be 0.8m x 0.8m x 0.8m = 0.51 cubic meters. So, the guidance would indicate that in this case the usable space in the chamber
is 51% of its total volume.

There are several other facts to consider:

1. You must map the space you are actually using.
2. "Climatic and Thermostatic Chambers" such as incubators and stability chambers, require air circulation to operate correctly.
3. The Manufacturer manuals (and therefore their recommendations for use) for your incubators will likely have very restrictive guidelines for loading that are similar to that in NFx15:140.
4. NFx140 is a French standard and therefore not enforceable in the US. However, it is one of the few guidance documents available to us, and as such, was incorporated as a reference in to other guidance documents, such as "ISPE Good Practice Guide: Cold Chain Management", which definitely influences the expectations of auditors in the US. I expect the upcoming guidance "ISPE Good Practice Guide: Controlled Temperature Chambers" will also be consistent with NFx15:140.

Your thought to stack the shelves full has some validity. It is very difficult to control loading of chambers by procedural controls, especially when a typical facility doesn't have enough space to anyway. When faced with the choice between violating procedure, and failing to get your work done, most people will "accidentally" overload a chamber. The best way to get around having to do procedural controls is to overload the chamber during mapping.

I think this can a viable strategy, with certain limitations:

1. You are not using the chamber as the manufacturer designed it.
2. This exposes you to risks of premature failure and may void any warranty.
3. You can no longer truly depend on the extensive mapping and testing the manufacturer did when they developed the chamber. Their mapping studies during development are the source of their loading recommendations. Therefore, you should have little expectation that the chamber will perform adequately when fully loaded.
4. You will want to consider using a higher density of mapping points (more sensors) since you can no longer rely on the development testing done by the manufacturer. (It was likely mapped using an array of 27 sensors per DIN 12880).

That's my $0.02. I hope it helps! If all I did was raise more questions, feel free to reach out to me or our European Regulatory Compliance expert Piritta Maunu.

Best regards,
Paul Daniel

Video: The Stuff we do in Life Science Environments
Under 3 minutes + a Catchy tune...

We made a video to describe our systems in life science environments.
If you haven't seen it, check it out! The music is actually pretty good. You can't dance to it though... we tried. ☹