Pump duty rotation

Join Date
Aug 2002
Location
Manchester
Posts
527
Hello
I have up to 8 pumps that, depending on demand, I'll run however many I see fit. Pressure is my setpoint and a pid and 8 inverters deal with that nicely.
My remit is to equally share the pumps' work.
My initial thinking was to turn on/off the pumps based on their run time, but I'm thinking number of starts might be a better metric, or even a combination of both.
Maybe even start the one with the lowest number of starts, and the stop is based on run time.
I suppose my success will be determined by all 8 pumps failing on the same day šŸ˜Š
Is there a preferred methodology?
Thanks, pp
 
Thank you. Still interested if it's number of starts, running time, or a combination of both that is the ideal governor for who starts/stops next.
The logic isn't a problem.
 
Do you have a maintenance plan in place? How often (runtime or starts) will a pump be removed for rebuild and how long does the turn-around take? I would think these would be important factors that would be part of the utilization schedule.
 
Number of starts and overall run time both seem like reasonable data points for determining wear on the pumps/motors. The relevant question would be how much weight you would assign to starts vs. runtime.

That said, it occurs to me that if you started with 8 brand new pumps and did a perfect job of balancing utilization, then you could expect all of the pumps to eventually fail at pretty much the same time.

Have you considered setting aside a group of pumps for lower than normal utilization?
 
I've done this type of setup before but only using 3 pumps. I based all of my decisions upon the "Total hours on each motor" so which ever motor had the least amount of total run time accumulated would be my next motor to start, and I did not switch motors unit that run time then equaled the next least total accumulated run time.
 
One issue you have to consider is what happens when a pump is replaced. If you were to reset the run time to zero then that one would need to run until it catches up with the others.
What I have done in the past was to trigger the pump switch based on a time clock and add some extra run time to the lowest hours pump. It takes a while for the pumps to equalize, but they eventually do.
 
Thank you. Still interested if it's number of starts, running time, or a combination of both that is the ideal governor for who starts/stops next.
The logic isn't a problem.

I have had a problem with a 2 pump method where the run-times were kept close.

Pump 1 failed first. We didn't have a spare in inventory since there was a 'hot spare'. The new pump was ordered. The second pump failed before the new pump was received. But that was a 6 week delivery, over the Christmas holiday.

We now keep the a rebuilt kit on hand for 'hot spare' pumps, and the switching of the pumps keeps them within about 400 hours instead of 50 hours, if I remember correctly. There is also a condition that each pump must be started once a month.

In your case, with 8 pumps, if you need all 8 to run at the same time once in a while I'd keep one or two at 2/3 of the hours of the others. If you never need all 8, the extra effort is likely not worth it. As long as the delivery for pump/motor parts is not 6 months ...
 
I used to try to balance motor run times until I read some of the threads on here and realized that, for the most part, it's a bad idea. What I do now is rotate who is first on a set time based on application. I'm currently doing one that has 5 pumps and it will rotate the 1st position every 24 clock hours regardless of runtime. This application will have, at most, 3 pumps running at a time and sometimes will have 0 pumps running. I think this will stagger the run times enough that it shouldn't sync failures.
 
Hello
I have up to 8 pumps that, depending on demand, I'll run however many I see fit. Pressure is my setpoint and a pid and 8 inverters deal with that nicely.
My remit is to equally share the pumps' work.
My initial thinking was to turn on/off the pumps based on their run time, but I'm thinking number of starts might be a better metric, or even a combination of both.
Maybe even start the one with the lowest number of starts, and the stop is based on run time.
I suppose my success will be determined by all 8 pumps failing on the same day ļæ½ļæ½
Is there a preferred methodology?
Thanks, pp


I think for all reasons of redundancy and maintenance the best strategy would be avoid starting and stopping pumps to balance their overall runtime.

If pumps are run until failure they will fail at different times anyway (and for different reasons). So there is no point in trying to balance their runtimes. It's better to balance their usage instead and make sure everyone pump is used but not needlessly started and stopped.

I suggest keeping track of

  • number of starts,
  • total runtime,
  • current runtime (if running)
  • current stoptime (if not running)
When you need to start a pump you start the one that has the longest current stoptime.

When you need to stop a pump you stop the one that has the longest current runtime.

Number of starts and total runtime is just used to inform operators and maintenance.

With this strategy all pumps will be used about equally.
A pump that has been serviced or replaced will still get about the same use as the others. It will not get extended amount of use just because it's new and this is a good thing. For redundancy it's better to have a spread of new ones and old ones.

This also works if there a pumps that are taken out of service. While they're out of service the other pumps will be balanced in usage and starts and stops. When the pumps are put back into service they will be the ones with the longest stoptimes and the first in line to be started up again.
 
Last edited:
I think Pete.S's reply is pretty good one, but perhaps you'd still want to consider a maximum running time before switching over? I'm assuming here that there will always be a pump or more in use.

Another option, although I don't like it as it's usually left unattended, is to have assignments for the operator (read maintenance) to have some control of the pumps running. So whatever logic you create will only work on the pumps selected by maintenance.
 
I have done it the way Pete described with the addition of an hour counter that will ensure that a pump gets started at least once every x hours...in my case it was every 72 hours.

In my application, the pumps ran to control a sump level with a PID in charge, so it was common for them to run at a steady sump level without changing the number of pumps called for very long periods...but if a pump did not run for 3 days, sludge could harden inside the pump, so we added the extra "idle hours" counter feature to prevent this.

I also like to have a minimum stop time and minimum run time for most pumping applications. With VFDs this is not always necessary, but with motor starter operated pumps, I want to prevent short cycling.
 
The preferred methodology would depend on the process and what its requirement is. If your process is filling a tank and the process will require a pump to be on for a few hours then off for a few hours, etc., then a simple alternator would be best. If you have a process that requires a Lead pump and periodically uses a Lag pump, then you could use a Lead alternator to determine the Lead pump and a separate Lag alternator that will cycle through the other pumps when a lag is requested. If you have a process that uses a pump 24/7, then you could alternate an alternator based on the current runtime.

Thatā€™s three different process requirements with three different options. An effective way to alternate the pumps depends on the process requirements. What will your process be asking of the pumps? Without knowing how the process is using the pumps, thereā€™s really no way of giving a good answer on how to alternate them.
 

Similar Topics

Dear all, I am doing a PLC program for 2 duty and 1 standby pump rotation control by using Mitsubishi Q series PLC. The control sequence of the...
Replies
11
Views
10,071
Dear all Can any one share me the logic for 2 Pump Duty and 2 Pump Standby as per Running hours selection in RS Logix 5000 Rev 20?
Replies
3
Views
2,026
i hav attached the power and control diagram. can you guys please help me out to make a plc program using schneider zeliosoft 2 software. any plc...
Replies
1
Views
2,915
heyy guys. can you guys please help me out. plc programming in schneider plc zelio soft 2. can you guys help me out as soon as possible.??? and...
Replies
5
Views
2,620
<P>can anybody please give me a sample ladder program/idea for 3 pump duty,1 standby operation with minimum number of timers</P>
Replies
4
Views
3,273
Back
Top Bottom