Waste Oil Pump Control

[wdreynolds1]

Lancie, et al...

Let me say again how much I appreciate the input.

I've attached my emulation program for 5 pumps... I made some slight modifications to your suggestions and have debugged thoroughly.

I will be adding alarms later.

Any suggestions, comments or feedback would be appreciated.

Regards,

W

 

[Lancie1]

Any suggestions, comments or feedback would be appreciated.

It looks like it will do the job. Congratulations! And they said it couldn't be done with a FIFO!

You may find that splitting it into subroutines will make it easier. 17 sets of the same logic for each pump is a lot to scroll through when you need to find a particular pump.

 

[wdreynolds1]

Thanks again for all the help!

W

 

[robertmee]

Couple of concerns I have with your program:

Never wire Stop PB's as NO. They should be NC. Otherwise what happens if you lose power or a wire to your Stop PB. You won't be able to stop a pump. Dangerous. So you should have XIC not XIO's on the stop Inputs after you change the wiring.

Similarly, your high level switch should be the same way. What happens if you lose power to your switch or the wire breaks. You're going to overrun the tank because you'll never detect high level.

Next, granted I didn't simulate or spend alot of time, but I don't see a mechanism for stopping a pump that's in the queue to start without waiting for it to actually start. Suppose someone presses the Start on Pump 10 and it's in the queue several pumps back. You should be able to press the Stop on Pump 10 and it leave the queue. Otherwise another dangerous condition where someone has pressed stop on a pump and it starts later because of the FIFO unload. A way to resolve this is to set a flag for each pump that it's been loaded into the queue. Reset this flag on exit of the queue and starting of the pump OR anytime a stop PB is pressed. A condition to start is that this flag is still set when the pump exits the queue.

 

[wdreynolds1]

Robert,
Thank you very much for the feedback... I hadn't considered the need to stop a pump that is waiting in the queue. That really helps me improve the code. I will use a latch/unlatch on the start/stop to inhibit the start of a pump in the queue that has been requested to stop. It will unload as normal but have no effect on the output.

The stop PBs in the this program (I'm assuming you're looking at the all-stop) are wired NO just for testing purposes in the emulation... they will, of course, be NC when I write the actual program.

Thank you again for the feedback... its been invaluable!

W

 

[Lancie1]

EDIT: Never mind, your stops in the PLC program should appear as open contacts.

 

[wdreynolds1]

Lancie,
If there is no power on the stop, flow or pressure switches in the field then we'll have an all-stop command in the program. Always NC in the field and NO or NC in the program... depending on what we're looking at.

Thanks for the clarification,

W

 

[Lancie1]

I see you added a timer for when 5 pumps are running and there is a high pressure event. If one pump stops on high pressure (as required by the original specification) then the pressure will drop. Unless the pressure switch dead band (hysterisis) is set correctly, then when the pressure drops, another pump will immediately be unloaded from the FIFO and started up. Actually the specs call for the SAME pump that just stopped, now be restarted (not after some time period) but as soon as the pressure drops below 50. I think to make this work, the pressure switch must have a large dead band, and it must be set so that it must not close again until another pump has finished pumping. Otherwise there will be a vicious cycle of high pressure-pump stop-low pressure-same pump restarts-high pressure again, then repeat.

Many pressure switches have such a dead-band setting, for instance well pump switches. These can be set so that the pump fills the tank up to say 60 PSI, then the switch pressure switch opens, but does not reclose until the pressure drops to the minimum pressure required by the home water system, say 20 or 30 PSI. That is the type of pressure switch you need here. Then no timer would be necessary. Actually the timer does not eliminate the repeating cycle, only delays the time until the repeat.

 

[wdreynolds1]

Lancie,
Good observation. Actually, we were able to talk the customer into an analog pressure switch. I'll be coding the dead band in the program... we don't permit another pump to start when pressure is above 40psi, we shut down the last pump at 50psi, and we delay the restart for 15 seconds (user adjustable) once pressure has dropped below 50psi. I'll tune this control once I'm in the field and we run tests to see what our real world pressures are going to be.

Thanks again,

W

 

[Lancie1]

...we shut down the last pump at 50psi, and we delay the restart for 15 seconds (user adjustable) once pressure has dropped below 50psi.

Think about this some more. Each pump is probably the same size and is going to increase the header pressure by some X amount (not the same value but always about the same pressure increase for each pump position. Pump 1 will go from 0 to V PSI, the second pump will increase PSI by W PSI, and so on up to 17. So if you kill the 5th pump, then PSI will drop by Z amount and after 15 seconds probably Pump 5 will restart and the pressure will increase by Z amount, and Pump 5 will again stop and repeat, over and over. I think you really need the pressure to drop by Y + Z amount to indicate that another pump has finished, before Pump 5 restarts. Then the 5th pump will not stop as soon as it restarts. So your analog pressure input should be used like this: If pressure above 50, stop the last pump started, and don't allow any more to restart UNTIL pressure is less than 40 INCLUDING the last pump started.

About the stop function. If a guy at some station presses Start, walks off and says "Gee, I really should not have pressed that last button", then will he remember WHICH of the 17 stations that he accidently requested to Start? I really do not think undoing a mistaken start request is a big deal in this case. It is not like it will cause any danger. So that pump starts up when its oil tank is empty. No problem, just press the stop button!

Really, all you need is one pump with 17 control stations for this system. 17 pumps is going to be expensive for not much gain.

 

[wdreynolds1]

Lancie,
We are in total agreement... however, the customer has requested individual pumps at each station and would not come off that request. Oil viscosity is not a constant in this case... some stations may be pumping antifreeze while another is pumps gear oil... pressure will stabilize after start up but I won't really know the RW reaction to a start/stop until I'm in the field.

W

 

[Lancie1]

I see. Antifreeze would definitely be thinner than most oils. I am surprised that the antifreeze is added to the mix. I suppose it all separates again at the refinery. With a mixture of different weights of fluids, setting the Cut On and Cut Off pressure is going to be a challenge. Also, what happens when a station pumps all the available oil but the operator is not close by to shut off the pump? Can these pumps run dry without harm?