Hello All,
This is going to be a bit long winded, but I'll try to give as much information as possible!! I'm a humble electrician and I've have been placed in a plant and have been asked to look at making some process modifications to more efficiently control our pumps. The place has been poorly engineered from conception and we have no engineers on site to run these things by, so I'd like to get some opinions and input from you all on a few ideas we have thrown about.
Our batching plant has 26 underground tanks and 26 4kw induction motors directly coupled to the pumps. Each pump sends its product through its own 2" pipe into the plant and there is a tee-off point at each mix tank and then the pipes are dead ended. This is our main problem - no return line to the feed tanks (and no we can't even contemplate puttting them in due many restriants). Each line has a non-return near the pump, and a differential pressure switch downstream from the non returns.
The bulk of the product is pumped into 10000 litre large mix tanks, however (and this is the awkward part arises), we also have hand filling stations from which we take 20-1000 litres at a time from. Currently the product pipes are run at a pressure of around 400kPa during filling - not real nice when you are trying to get 20 litres out into a drum! The hand filling is done via 1" petrol pump bowser type handles/hoses. The only good part about the bowser set up is the fact there is a switch at each that triggers the pumps to start when a a hose is taken out of its cradle. This switch could be useful down the track.
When not filling the pumps lie dormant and the only way the pumps can start and stop is via differential pressure switches in each line or via the cradle switches at the bowsers. The pipes are getting up to pressure REAL quick and the pumps are cutting on and off via the pressure switches a ridiculous number of times per hour. In short this is destroying them. Our targets are -
AIM 1 = Achieve maximum flow, whilst maintaining safety for the operators.
AIM 2 = To maintain a fill pressure of say 400kPa to the large mix tanks.
AIM 3 = To drop the pressure to say 50 kPA for safety when ONLY filling by hand.
AIM 4 = When filling to the large mix tanks AND by hand, throttle pressure back if possible (Hard one this one, any solutions welcome!!)
AIM 5 = When filling ends , throttle the pump back to a safe point above motor stall speed for ten minutes, then turn off thus allowing only 6 motor starts per hour (evenly spaced to save the motors).
AIM 6 = Turn off where possible to conserve energy (some pumps are not used for days on end).
My thoughts were to replace the differential pressure switches with pressure transmitters in the pipes and to have the pressure 4-20mA signal feedback into the Allen Bradley SLC500 plc. From there have software PID's controlling the output signal to a VSD, trying to maintain an appropriate pressure for each types of filling.
I am far from being an expert when it comes to PID loops, but hey this is a pilot project on one fill line only for now and i'm keen to learn more!! My concern is that to cover our different fill pressure requirements, i would need 3 PID's with different pressure SP's controlling the ONE COMMON analog output. I have not been in this situation before and I don't know if this is possible. I would imagine it's not a good idea to cut PID's in and out like this!! Can this be done??
Logic wise, each PID would have to be turned on only if the pressure transmitters feedback sits within a certain band. These LIM (limit) logic instructions then switch off the PID's not in use via an XIC instruction.
Situation #1 - Filling to large mix tanks.
The first LIM instruction covers a PV band of 150->450kPa and controls PID#1 (large tank fill occuring thus maintain SP=400kPa). When filling ends and the 2" fill valve is closed, i take it we should have a pressure rise due to no flow, to say 450->500kPa when dead headed. The non return valve at the pump should trap this 450->500kPa in the line. If the pressure returns below 450kPa, keep going as was.
A second LIM instruction now takes over as it covers a PV band of 450->500kPa (pressure is above normal fill pressure thus no fill occuring). If the pressure remains between 450->500kPa for more the than 1 minute, this should turn off PID #1 as its safe to say filling's over.
From here we want to slow the pump down to a safe point above the motor's stall speed. Our intention is to install a return line (from between the discharge side of the pump and the non return) back to the underground tank and to cycle the product back. After a further 9 minutes if the trapped pressure is still 450->500kPa turn the pump off entirely to conserve energy. This would keep us to our possible 6 starts an hour limit on the motors.
Situation #2 - Hand filling to small drums.
When filling by hand only (from a bowser), utilise the switch at the cradle to enable PID#2 with an SP of say 50kPa to keep the pressure low. As hand filling
begins this should let the trapped pressure dissipate from 450->500kPA down to 50kPa then to try to maintain it. A restriction orifice placed in the hand filling hose should
dampen the initial high trapped pressure.
Situation #3 - Filling to large mix tanks whilst hand filling small drums.
Would be the same control arrangement as Situation #1, but when the bowser switch changes state throttle back by cutting to situation #2.
If you managed to bother to read this far, thanks for your time and any suggestions, simplifications, alternatives, solutions, criticism or logic examples are
welcomed. If things aren't totally clear let me know and i'll do my best to resolve the situation.
All the best, Regards - Chris.
This is going to be a bit long winded, but I'll try to give as much information as possible!! I'm a humble electrician and I've have been placed in a plant and have been asked to look at making some process modifications to more efficiently control our pumps. The place has been poorly engineered from conception and we have no engineers on site to run these things by, so I'd like to get some opinions and input from you all on a few ideas we have thrown about.
Our batching plant has 26 underground tanks and 26 4kw induction motors directly coupled to the pumps. Each pump sends its product through its own 2" pipe into the plant and there is a tee-off point at each mix tank and then the pipes are dead ended. This is our main problem - no return line to the feed tanks (and no we can't even contemplate puttting them in due many restriants). Each line has a non-return near the pump, and a differential pressure switch downstream from the non returns.
The bulk of the product is pumped into 10000 litre large mix tanks, however (and this is the awkward part arises), we also have hand filling stations from which we take 20-1000 litres at a time from. Currently the product pipes are run at a pressure of around 400kPa during filling - not real nice when you are trying to get 20 litres out into a drum! The hand filling is done via 1" petrol pump bowser type handles/hoses. The only good part about the bowser set up is the fact there is a switch at each that triggers the pumps to start when a a hose is taken out of its cradle. This switch could be useful down the track.
When not filling the pumps lie dormant and the only way the pumps can start and stop is via differential pressure switches in each line or via the cradle switches at the bowsers. The pipes are getting up to pressure REAL quick and the pumps are cutting on and off via the pressure switches a ridiculous number of times per hour. In short this is destroying them. Our targets are -
AIM 1 = Achieve maximum flow, whilst maintaining safety for the operators.
AIM 2 = To maintain a fill pressure of say 400kPa to the large mix tanks.
AIM 3 = To drop the pressure to say 50 kPA for safety when ONLY filling by hand.
AIM 4 = When filling to the large mix tanks AND by hand, throttle pressure back if possible (Hard one this one, any solutions welcome!!)
AIM 5 = When filling ends , throttle the pump back to a safe point above motor stall speed for ten minutes, then turn off thus allowing only 6 motor starts per hour (evenly spaced to save the motors).
AIM 6 = Turn off where possible to conserve energy (some pumps are not used for days on end).
My thoughts were to replace the differential pressure switches with pressure transmitters in the pipes and to have the pressure 4-20mA signal feedback into the Allen Bradley SLC500 plc. From there have software PID's controlling the output signal to a VSD, trying to maintain an appropriate pressure for each types of filling.
I am far from being an expert when it comes to PID loops, but hey this is a pilot project on one fill line only for now and i'm keen to learn more!! My concern is that to cover our different fill pressure requirements, i would need 3 PID's with different pressure SP's controlling the ONE COMMON analog output. I have not been in this situation before and I don't know if this is possible. I would imagine it's not a good idea to cut PID's in and out like this!! Can this be done??
Logic wise, each PID would have to be turned on only if the pressure transmitters feedback sits within a certain band. These LIM (limit) logic instructions then switch off the PID's not in use via an XIC instruction.
Situation #1 - Filling to large mix tanks.
The first LIM instruction covers a PV band of 150->450kPa and controls PID#1 (large tank fill occuring thus maintain SP=400kPa). When filling ends and the 2" fill valve is closed, i take it we should have a pressure rise due to no flow, to say 450->500kPa when dead headed. The non return valve at the pump should trap this 450->500kPa in the line. If the pressure returns below 450kPa, keep going as was.
A second LIM instruction now takes over as it covers a PV band of 450->500kPa (pressure is above normal fill pressure thus no fill occuring). If the pressure remains between 450->500kPa for more the than 1 minute, this should turn off PID #1 as its safe to say filling's over.
From here we want to slow the pump down to a safe point above the motor's stall speed. Our intention is to install a return line (from between the discharge side of the pump and the non return) back to the underground tank and to cycle the product back. After a further 9 minutes if the trapped pressure is still 450->500kPa turn the pump off entirely to conserve energy. This would keep us to our possible 6 starts an hour limit on the motors.
Situation #2 - Hand filling to small drums.
When filling by hand only (from a bowser), utilise the switch at the cradle to enable PID#2 with an SP of say 50kPa to keep the pressure low. As hand filling
begins this should let the trapped pressure dissipate from 450->500kPA down to 50kPa then to try to maintain it. A restriction orifice placed in the hand filling hose should
dampen the initial high trapped pressure.
Situation #3 - Filling to large mix tanks whilst hand filling small drums.
Would be the same control arrangement as Situation #1, but when the bowser switch changes state throttle back by cutting to situation #2.
If you managed to bother to read this far, thanks for your time and any suggestions, simplifications, alternatives, solutions, criticism or logic examples are
welcomed. If things aren't totally clear let me know and i'll do my best to resolve the situation.
All the best, Regards - Chris.
