VFD lead lag water pumps

I have two pumps that are on VFD's, supplying water. Each runs with identical speed setpoints up to 50%. Above 50% the lead pump will take on the extra load.

should I run these two pumps off of one PID loop and clamp the lag pump to 50% or run off of two PID loops with the lag PID loop clamped at 50%...

as far as the lead lag alternation...should I do that on time or alternate on exceeding 50% on the lead and then returning (which would then switch the lead pump to lag)

With 20+ years in the potable water arena, I strongly recommend against equalizing the wear on your pumps, especially if the application is at all critical. Keep one unit with lower hours so when the most used pump needs work, the other still is solid to go. Speed control method really is a matter of preference, in my opinion.

MasterBlaster said:

With 20+ years in the potable water arena, I strongly recommend against equalizing the wear on your pumps, especially if the application is at all critical. Keep one unit with lower hours so when the most used pump needs work, the other still is solid to go. Speed control method really is a matter of preference, in my opinion.

Click to expand...

Thanks for that input MasterBlaster. question....


The control narrative has them "alternating lead/lag" so do you recommend something like 70/30 split lead lag?

That’s a reasonable split, I think. In a three pump scenario, you could 50/50 two of them and keep one in standby (run only enough to keep it lubricated and such) until a failure occurs. But try to plan for trouble, cuz it will find you . . .

MasterBlaster said:

That’s a reasonable split, I think. In a three pump scenario, you could 50/50 two of them and keep one in standby (run only enough to keep it lubricated and such) until a failure occurs. But try to plan for trouble, cuz it will find you . . .

Click to expand...

Thanks

MasterBlaster said:

That’s a reasonable split, I think. In a three pump scenario, you could 50/50 two of them and keep one in standby (run only enough to keep it lubricated and such) until a failure occurs. But try to plan for trouble, cuz it will find you . . .

Click to expand...

Wouldn't this fall under preventative maintenance plan?

Even something super simple, it started this many times, it ran for this many hours. Rebuild/replace.

Newer sensors can have vibration monitoring and I think some have HART feedback that lets you know when its on its last leg.

(This isn't my area though, I've just been around some PM management projects.)

VAN said:

Wouldn't this fall under preventative maintenance plan?

Even something super simple, it started this many times, it ran for this many hours. Rebuild/replace.

Newer sensors can have vibration monitoring and I think some have HART feedback that lets you know when its on its last leg.

(This isn't my area though, I've just been around some PM management projects.)

Click to expand...

Very true, in theory...

But in reality, I've never seen any PM on a pump of any type. They are forgotten as long as they work, and 'a piece of $#!+' when it fails, regardless how long it ran without any attention!

I have a lot of experience with sequencing water pumps. You should always run two pumps in parallel at the same speed because they must produce the same head. Also, running two pumps in parallel does not result in twice the flow because of pipe friction. You should run one pump up to 100% and alternate. You could compare energy usage with one pump at 100% vs. two pumps at lower speeds in parallel. After all the purpose of a VFD is to save energy. This can be as simple as monitoring kW manually at different speeds and using that output as a switching point to go to parallel operation.

Gene Bond said:

Very true, in theory...

But in reality, I've never seen any PM on a pump of any type. They are forgotten as long as they work, and 'a piece of $#!+' when it fails, regardless how long it ran without any attention!

Click to expand...

ive worked with some water districts who will purposely dont do PM on large pumps/motors, because the board has shown they respond quicker to emergencies over large PM bills.

We used to have the standby pump way of doing things too. 2 alternating and one just sitting there. We found that when it came time to run that standby pump for a longer period of time than the monthly rotation tests, the pumps or the related support devices showed failures in short order from the longer duration run.

We now alternate all the same sized pumps equally and the oversized fire pumps that are sometimes present, every 20th cycle or so.

kittydog42 said:

I have a lot of experience with sequencing water pumps. You should always run two pumps in parallel at the same speed because they must produce the same head. Also, running two pumps in parallel does not result in twice the flow because of pipe friction. You should run one pump up to 100% and alternate. You could compare energy usage with one pump at 100% vs. two pumps at lower speeds in parallel. After all the purpose of a VFD is to save energy. This can be as simple as monitoring kW manually at different speeds and using that output as a switching point to go to parallel operation.

Click to expand...

I have not done tandem water pump for pressure control previously.



-- one requirement is to be able to change the speed of each pump when the PID is in manual.



based on that requirement I will be using two PID loops with the PV being the combined Press Output. Thinking this through, I am not sure how to ensure load balancing between the pumps to the target 50% output. The PIDs could fight each other leaving one pump with almost all the load. Since you have done this before what is the preferred method for this control?



[one thought I had was to clamp the lead pump to 50% and let the lag come on line as needed up all the way to 100% then switch lead lag on hours]

two PID issues

the more I think about this the more I am inclined to use ONE PID controller. because for example the Operator could have 2 different setpoints and then totally jack with my lead lag fail-over.



I think One PID controller that moves the output into 4 separate Floats. F[1] F[2] F[3] F[4]. when NOT in manual F[1] F[2] Floats get moved into the AOs and F[3] F[4] are for bump-less transfer.



when the PID is taken to manual the MOVs above are disabled (F[3] F[4] hold the last output sent) and the MAN rung uses two MOVs to move F[3] F[4] into the AOs. F[3] and F[4] can now be adjusted by the operator.

Your last post is what I have done. If both pumps feed into the same header, you should keep the speeds matched when in auto (one PID) for best results. You can use a oneshot from each PID mode change to get a bumpless transfer to and from automatic if only one pump is run in manual. The oneshot to manual mode can copy the PID control output to the manual setting for all the pumps, but then the operator could be allowed to run more than one pump and adjust their speeds independently. That might rarely if ever be useful, but I tend to lean toward giving operators the most flexibility as long as it doesn't adversely affect the equipment.

As for avoiding equivalent run times, I understand in theory why you would not want to wear out all the pumps at the same time. In reality that is unlikely to happen for several reasons, but I like the let the operators decide when to allow the pumps to alternate: a) not at all (pick a lead pump) b) each call c) daily. I recently had a system where the operator wanted to alternate pumps weekly, so I let him choose the day of the week to do it, but his HMI still has a control to let him change that to one of the other 3 more commonly used choices.

kittydog42 said:

You should always run two pumps in parallel at the same speed because they must produce the same head.

Click to expand...

OkiePC said:

If both pumps feed into the same header, you should keep the speeds matched when in auto (one PID) for best results.

Click to expand...

Good Morning, Various booster pump stations we work on, the pumps do not run at the same speed. They all pump into the same header, but if the first pump gets to 100% and is not able to meet the pressure demand, the 2nd pump kicks in on the PID loop and controls the speed to meet the PID setpoint. So, for example, the first might be running at 100% and the 2nd at 30% to get the correct pressure.

Would it be better if they would both be running at 70%?

cjjeeper said:

based on that requirement I will be using two PID loops with the PV being the combined Press Output.

Click to expand...

cjjeeper said:

Thinking this through, I am not sure how to ensure load balancing between the pumps to the target 50% output. The PIDs could fight each other leaving one pump with almost all the load.


[...]

the more I think about this the more I am inclined to use ONE PID controller. because for example the Operator could have 2 different setpoints and then totally jack with my lead lag fail-over.

Click to expand...

Your theory about two PIDs chasing the same PV has proven true in my experience. Two PIDs is a bad call IMO. I would use one PID.