Webb Confidence Index

by Charlie A. Webb CPP, CMC
Staff Writer

Over the last nearly two decades I have spoke with thousands of packaging engineers and a reoccurring question is often presented to me. How often should I perform a thermal calibration on my packaging equipment? There are so many factors one must consider when developing calibration timelines with critical packaging equipment. The first thing we need to consider is by what process is the machine sensing temperature at the platen? Is it through a solid temperature sensor, a thin film temperature sensor, or is it a RTD system using the active platen for thermal feedback? Because some of these systems may be considered sturdier then others, their calibration may stay accurate longer.

What I’ve developed in my own lab is what I call the “Webb Confidence Index”. I developed this basic index that I consider a solid metrology verification plan. When first commissioning in a new piece of equipment, I confirm if possible, in my lab, that the device is in fact calibrated. If this piece of equipment requires special machinery to perform the calibration, I often purchase this equipment from the vendor in order to have the same ability to manage the metrology of the equipment as does the maker or vendor for this critical equipment. I do this even if I will not be performing the official annual calibration just so I can commission the device back into my lab after being sent out for calibration. Also this allows me to use the same metrics as the manufacture so our processes are harmonized.

After verifying that the device is calibrated I grant the device a 10 on my confidence index, showing that it was calibrated, and is still in an expected state of performance on arrival. Typically after 30 days I will check the calibration, I will perform this task despite the fact that the vendor tells me it need only be calibrated annually. I invoke the trust-but-verify system, just to be certain that nothing has changed. After performing my own internal verification, I call this ”verification” because I am not doing this in the spirit of a true full calibration. I am simply checking, and this checking process is designed to correspond with my preventative maintenance plan for my total quality requirement. If the device is still in calibration, I grant the device a 10 position. If the device can be checked in terms of varying degrees of accuracy from my initial point of calibration, then I may diminish that value from 10 to say an 8. But under my index, anything over 5 is an acceptable verification. I will check it again after 30 days and if the device is shown to still fall within that 5 to 10 accuracy area, I will continue to verify monthly. If however, during my first few months of verifying, the device has shown that it moved outside of calibration, that device is immediately given a grade of 1 due to a diminution in my confidence with this equipment.

Even if the device is at the very edge of calibration, it may be granted the under 5 number and that will require that it have a complete new calibration and be checked weekly. Or even daily on some highly critical equipment. After the daily checks, if it continues to show that it’s performing accurately for a five day period, I will then run back to the one-month check. There are a variety of different formulas and patterns you can use, but the important takeaway is that annual calibration, even if it’s the manufacturers recommendation is risky business. At the very least you should be checking or verifying the machine’s performance at least quarterly, if not monthly. We have received several machines into our lab that were poorly maintained and as a result of a marginal PM and verification plan the machine arrived out of calibration. Some of these company’s had shipped tens of thousands of parts into the market and ultimately into the sterile field, this left their quality department with a huge question mark hovering over their head as to how many possible compromised packages are now in the marketplace.

Remember any device can change over time and when you are talking about a system such as impulse sealing equipment that heats and cools thousands of times per day, this change could happen quickly. Also systems that incorporate consumable parts may have a performance change after servicing the equipment, so after a major service always check calibration. I believe companies need to train maintenance staff on calibration and not just leave the task of checking machine performance with their metrology departments. Packaging machine performance is centered on a tight preventive maintenance plan and calibration so keep a close eye on these issues to thwart the pains of a packaging recall.