Pump Application

I have a utility related question, in regards to pump control...Sounds like it's kinda up Tom's alley... I don't have a lot of experience with pump control, but I've been asked to do a fair amount of it lately.

I guess I can do it moderately well, most of the applications are simple and relatively low HP, usually under 4 pumps.

When running a series of pumps in a lead/lag fashion, should you alternate the lead pump? Some people recommend to me that you want even pump/motor wear. In my opinion, doesn't this just make them all wear out around the same time?

I know you want to start/stop a motor as infrequently as possible, but I think I'm missing something here.....Some people really seem to be STRONLY opinionated to having even run time on the motors/pumps on a particular bank, so much that they ask for Hobbs meters! Is there something fundamentally that I'm missing....other than someone's personal preference?

TIA for any input,

Greg

While waiting for our resident expert to weigh in, I'll through out my 1.25 cents worth. Probably 99 % of duplex and triplex pump panels I've seen do not alternate the lead pump.

I think personal preference does play a lot of it.

With a good plc and program, you can track hours on each pump AND alternate the starting pump daily, weekly, or with each cycle.

While a duplex or triplex pump panel can be made easily with out a plc, a small AD, GE, or entertron unit could easily be added and other additional features.

regards.....casey

PS RE: Hobbs Meters - I once had to hook three meters up and run them until they all read the same, even though each one was supposed to have a different duty cycle.

Now this i can help on. i am not an expert on plc's but i did have to care for roughly 1600 pumps on my previous job at a paper mill. ranging from 1/2 horse up to 1250 hp.
from 120 volt to 2300 volt.

the reason that they like to have even times on pumps where i was is because of consistincy, downtime tracking and failure analisys.but consistincy was extremeley critical to the product.

as for your question should you alternate. its a matter of personal opinon. we did not get any more run time out of the ones that we swapped around v/s the ones that ran constant. i have never found any real reason to go thru the trouble it takes to maintain a system like that(except where it is critical that the flow and volume and pressures be consistant) because in what we wasted on controls and monitoring we could have bought enough new pumps to last a lifeime. hope this helps. good luck, david

Lets look at this mechanically, the motors and pumps will have bearings which should have pre-determined life expectancy. Lets say its 10000 hours (just a number not an actual figure) for each pump/motor. Using this as criteria then you would always want to "attempt" equal hours of runtime per pump/motor to maintain the longevity of the total system. NOTE I said "attempt" because it depends on the Preventative Maintenance system for details....ie should pump/motor1 maintain a 500 hour lead and it be taken offline at 10000 hours for rebuild. They may have the pump(s) monitored and rebuilt on a schedule, thats why they insist on equal run time.

There are bearings and moving parts, which do you think would wear out first? A pump/motor that runs constant or a pump/motor that just runs on demand.

Ok, my 2 cents...

I'm gonna say it depends a lot on the application. So on hydraulic elevators the pump or pumps run only in the up direction, (the elevator comes down by gravity). So there are a lot of short cycles wich is harder on the equipment. So on large units with multiaple pumps of the same size we alternate the lead starting pump. On units with mutli pumps that are not the same size, we always start the smallest pumps 1st.

The reality is if all the pumps are started for each cycle, the differance in wear between pumps/motors is nominal. On applications that may call for only one pump, or not all the pumps on every cycle, I would then recomend alternating pump starts.

Thanks for the responses...Sounds like either

A.) It depends.
B.) Doesn't matter.
C.) Largely plays on preference.


I'm glad to hear with your mill experience, and all of those pumps/motors, you don't have any concrete eveidence to support longer life due to lead pump rotation.

Tom, if you're out there, I guess what I'm looking for is something concrete that supports the favor of rotating the lead pump...


TIA again,

Greg

Greg,

Let's look at it this way:

One thought for rotating your lead pump is to make certain that all pumps are working. If your application is considered to be critical if it fails, then you can't afford to have one pump not operate. By rotating lead pumps, you increase the certainy that all pumps are operational.

I would think the last thing your customer would want to happen, is find out that the last pump isn't operational when it is the only thing standing between you and an overflow. The cost to rotate pumps is much less than damage costs.

The majority of our customers that have pump applications, rotate their primary pumps. Some for preference...some to distribute total usage...some for the reason stated above.

As stated previously, rotation can be done in a number of ways. If your PLC has a real time clock, you can do it based on a calendar. If not, then the suggestion of some timers and counters would be more than sufficient.

Hope this helps.

God Bless,

There is one factor that hasn't been addressed above - a pump that goes for long periods of time without operating will generally fail soon after it begins operation. This counter-intuitive fact is due to the damage that can occur sitting idle. Without occasional operation the lubricant isn't distributed in the bearings and corrosion can result. Packings get stiff and set, and won't seal right. And vibrations tranmitted through floors cause the bearings to Brinnell - develop fatigue, wear and flat spots from the vibration causing minute movement of the same ball or roller continuously on the same spot on the race. It is better for the equipment to be run periodically.

Pumps aren't like the Deacon's One Hoss Shay ( http://eldritchpress.org/owh/shay.html ) and with equal hours won't all come apart on the same day. Experience has shown that even with alternators and "equal" hours there is time to repair the first pump to go before it's companion bites the dust.

It is true that motors get the most thermal stress during starting, but on small pumps it doesn't much matter and on larger ones it is only necessary to make the interval for alternation long enough, typically weekly alternation, and the impact is negligible.

So, I do recommend alternating equipment on a regular basis, but there isn't a need to be fanatical about insisting on absolutely equal hours on all of them.

With automaticaly controlled equipment it isn't tough to program alternators. For systems with intermittent alternation a first on/first off alternator is fine. For continuously operating equipment it isn't hard to program an automatic alternator to switch lead equipment onr a set time and day of the week. However, I don't generally recommend automatic alternators in my systems, which have continuous operation of at least one unit. Instead I make it easy for the operator to walk up to the operator interface and simply set a new lead unit once a week. The system does the start-up and shut down. That forces the operators to at least look at the system once in a while, and avoids problems with sequencing when one unit is down for service or whatever.

Great answer Tom!

I second just about everything said in this thread.

Although it's just more trivia:

If you have true motor drives (not reduced voltage or multistage starters) on the pumps, then the starts per hour/starts per day does not apply at all. A true drive NEVER exceeds FLA and it starts excitation very near to zero Hz, which means the motors don't get the starting stresses they get when started at 6 times the FLA across the line at 60 Hz excitation (or 50 Hz for the Euro's).

Another opinion is that it's easier to implement the swapping than it is to argue its validity. I'd say, "If you want to alternate, then I'll make them alternate. If you don't, then I won't." It's just more code.

Thanks for the input Tom, and everyone else. Instead of the requested Hobbs meters, I will just have runtime displaying on the VFD's. Less holes in the panel, in an already wet environment.


Greg

Tom has really nailed the answer! We switch day pumps monthly to keep hours roughly close and to keep control over the iron that the wells produce. We have to switch dose rates for individual wells and it is much simpler to manually do it with the pump change by the operator. This keeps us on our toes checking the system.