The quick answer is A LOT!
Depending on the output of the factory or the complexity of the product, CIPs could be initiated up to 20 times per day, every day. Not only is that a lot of data, but it’s a lot of resources being consumed and valuable production time being spent not producing. While it’s great that there is so much information available, it’s also overwhelming as it can be difficult or impossible to harness it, understand it, and make good use of it.
At a fundamental level, we want to be confident that the production lines are clean or cleaned properly at a specific time in the past, and ready for the next production run. Food safety and brand protection relies on good hygiene and is why our industry is regulated and invests heavily in GMPs, HAACPs, and validation. Technological advances in automation have also played a key role in improving food safety and efficiency: the more automated the process the lower the chance there is for mistakes to be made.
You would expect that if your CIP processes are automated, then each cleaning cycle would be the same every time the CIP program is activated - the cleaning parameters are programmed in the system, the CIP runs, and then production (re)commences. However, this is not the case. Once you are able to visualize a CIP dataset over time, it can tell an unexpected story. Process variance, over and under-consumption and completely random outliers will typically be uncovered and may all be positively or negatively impacting on quality and sustainability.
Uncovering Process Variance to Improve Repeatability
Below is a snapshot of 30 instances of the alkaline step on the same object (a tank).
The target temperature parameter in the alkaline step is between 15 and 20 minutes at 65°C. Looking at the 30 events below, the alkaline step ran well above the cleaning protocol. On one occasion (first column on the left-hand side), it lasted around 23 minutes; on four occasions (third column from the left ), it ran for approximately 31 minutes; and on 10 occasions (the five columns accumulative from the right-hand side), it ran for over 40 minutes.
Plotting data on the IntelliCIP statistical analysis graph makes it possible to uncover process deviations and start asking “why?”. Why has an automated cleaning step not followed the program? Why is the automation set up with such large allowances? Why is more water, caustic, thermal energy, and time being used than needed? Does this explain why costs were higher than forecasted last month?
IntelliCIP provides clarity and structure to CIP data to go beyond “what was cleaned when?” and uncover “what is happening and how can we improve?”. Decisions can be made to reduce variance in processes to make them more efficient, and improvements can be continuously tracked and measured over time.
Why was the alkaline step running out of scope?
Identifying some strange behaviour in the CIP process led to troubleshooting to get to the bottom of what was happening. After a short investigation, it was discovered that, when the OEM installed the equipment, they programmed the cleaning step with very generous setpoints to meet their commissioning goals. The setpoints were reprogrammed to the validated, expected targets.