Clean-in-Place System Prototype Details
Background:
Working for a company designing a robotic saucing machine, the sauce supply lines were not removeable. Per NSF guidelines, this required a Clean-in-Place device. The original device consisted of a simple centrifugal pump, manual ball valves and two water holding tanks. Each sauce line would be flushed with detergent, fresh water and then sanitizer. The question remains, is this sufficient.
Step One,
Identify you have a Problem
The original Clean-In-Place system had an output pressure of 15PSI and a flow rate of about 0.75gal per minute, per sauce tube. After disassembly of a sauce pump, it was found it failed to clean the tight corners within the pump.
Step Two:
Quantify the Problem
With the use of Hygiena’s ATP surface testers, we could quantify how bad the problem was, start changing things and compare their effects.
Step Three:
Crazy Idea #1 and 2
First tried just cranking up the flowrate on the CIP system, but this had a small improvement and after talking with the operators, they do NOT like the idea of using more cleaning chemicals, because they are expensive.
Thus, I tried agitation with some ultrasonic transducers. We mounted them to the base of the pump, got them dirty and tried to clean with ultrasonic action. This worked better but again not perfect and nobody liked (even me) the high voltage and water combo.
Step Three:
Crazy Idea #3
I know a source of pressure is needed, looking at a piston type, positive displacement this time. This pump has a greater agitation which will improve cleaning around seals and fittings.
Selected a simple two piston pump but not sure how fast the pump can be operated safely attached to the saucer system. So, I selected a 3ph motor and attached it to a VFD for speed control. The VFD can connect to a microcontroller for activation.
Step Four:
Making it Better
The first run of the device showed more agitation is needed but one cannot increase the flowrate beyond 1gpm. This is a problem, but ideas keep coming…
Talked with the operators cleaning the saucer system, they currently need to haul 5 gal buckets of water to the centrifugal pump style CIP system; they would prefer to simply hook up a 30ft hose to the faucet. This unlocks new possibilities.
Took apart the positive displacement pump and removed the discharge check valve, allowing the pumped water to flow backwards. Then connected the inlet of the pump to the faucet with a flow control regulator. Now, one can independently control flowrate with the regulator and the pulsation action with the VFD, perfect.
Step Five:
Add more…
With the change to direct connect to the faucet for water, the ability to add cleaning chemicals needs revision. The old version, just add then to the 5 gal bucket, easy. Now, I need some way to inject chemicals directly into the water stream.
Two positive displacement FMI pumps were used to inject the chemicals. The pumps powered by a stepper motor and lead-shine controller. The pumps had a Hall-effect switch to locate top-dead-center on its rotation. This is needed due to chemicals only dispensing for the first 180 deg pump rotation, the remaining rotation is the suction-stroke. This required a microcontroller to measure the incoming water flow rate, rotate the FMI pump at the correct speed during the discharge stroke to achieve the desired mixing rate, then rapidly rotate motor during the suction stroke.
With converting to a micro-controller, this allowed the entire cleaning process to now be fully automated, just connect and press the start button (instead of manually timed and valves manually operated).
Next, I needed a flow-chart to aid in programming the microcontroller.
Step Six:
Bring to Life
Generate some code, perform some machining on the Mill, a few McMaster part orders later, add some electrical wiring and a noise filtering capacitor, you got yourself a fully automatic, pulsating CIP rig with chemical injection
Step Seven:
Test its cleaning ability.
Over a period of two weeks, numerous sauce pumps were made dirty by pumping sauce in the lab, then cleaned with this CIP system. The sauce pumps were then taken apart and inspected with ATP testers. The results were outstanding. Visually, there was no presence of sauce in any void or connection joint AND the ATP tests came back below 1/10 of industry recommended levels.
The CIP system was tested even more thoroughly by allowing the sauce to stay within the pump for 24hrs at room temperature, then cleaned. The results were the same, success.