Submersible Pumps Fed from VFD's 1

[powersoff]

Hello,

I wandered over here from the Electrical side. I am looking for help with understanding the control of submersible pumps, in order to possibly modify plc code or recommend something else. The system I am working with has 5 pumps all supplied by vfd's. The pumps are trying to maintain system pressure. The problem I am seeing is excessive pump starts(+100 a day)and low flow(which results in pump shutting off). The VFD speed reference is coming from PID's running in the plc code. The PID is trying to maintain system PSI(setpoint) and ramps the vfd speed accordingly.
Any insight would be appreciated

 

[dougarthur42]

It seems to me that you need to evaluate your system. 5 pumps all on vfd may be to much to control, you may want to run a couple continuously and swing the others on vfd. You may also want to have a minimum flow recirc, which will recirc when flows are low instead of shuting down pumps. You also need to remember about the pump curves, running 3 or 4 pumps at 8-14 % each may be to low on the pump curves, and it may be better to run 2 pumps at 40 %. There are alot of possibilities here, good luck

 

[itsmoked]

dougarthur42; Has great points to consider.

The typical method is to run the pumps on fully with one modulating. As more is needed another is turned on while the modulating pump is reduced back again. Trying to run them all at reduced speed can cause numerous headaches.

I think there are some fellows who have done this successfully in forum237.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[TenPenny]

Step one is to evaluate your system demand. Track it over the course of a day, for several days. If you are having over 100 starts per day with 5 pumps, someone didn't understand the system fully.


You probably need to measure flow as well, and by using the curves, determine that if your flow is between x and y, you can control your pressure by running, say, 3 pumps on vfd, but if the flow demand increases to z, you'll need to bring an additional pump on line...

 

[BigInch]

Rediculous. 5 pumps and they are on VFD? Give me a break.

**********************
"Pumping accounts for 20% of the world’s energy used by electric motors and 25-50% of the total electrical energy usage in certain industrial facilities."-DOE statistic (Note: Make that 99% for pipeline companies)

http://virtualpipeline.spaces.live.com/

 

[powersoff]

Thanks for the replies. Why is it ridiculous?

 

[BigInch]

I did assume all 5 pumps go to the same header. If they pump to different destinations, forget I mentioned it.

For flowrates between 10% and 110% of total capacity, you can get within 10% of any flowrate you target without even using any vfd at all, just by turning on 1,2,3,4 or 5 pumps.

You probably shouldn't be running at less than 20% of total capacity anyway; just a guess, anyway certainly not 10%. Below 50% motor load, efficiency for motors and vfds drop off very fast, so there's not much benefit, if any at all, for having vfds at those flow ranges. If you're running submersibles, maybe you have a high static head and vfds might not generate enough static head at those low rpms anyway.

If you run X pumps on rated speed and try to run X+1 on vfd, you have to match the head (discharge pressure) of the other units to properly share load. If you produce less head (pressure) with the unit on vfd then you are producing with the other units, they try to backflow into your vfd controlled unit until your vfd unit brings up its pressure to equalize. So then you wind up running X+1 at rated speed too. That's no good.

So, do you really need flow accuracy to within better than 10% of any flowrate you want to target? If not you're home free without VFDs. If you do, I'd try putting a spill back control valve on a line going from discharge header to suction header that could modulate 10% of the total flow capacity and quickly forget about all that electronic cable and power conditioning and bearing current headaches, etc. etc. etc.




**********************
"Pumping accounts for 20% of the world’s energy used by electric motors and 25-50% of the total electrical energy usage in certain industrial facilities."-DOE statistic (Note: Make that 99% for pipeline companies)

http://virtualpipeline.spaces.live.com/

 

[Zapster]

BigInch, where would one find a suction header on for multiple submersible pumps?

 

[BigInch]

Good question! Change to spill back to tank or whatever.

**********************
"Pumping accounts for 20% of the world’s energy used by electric motors and 25-50% of the total electrical energy usage in certain industrial facilities."-DOE statistic (Note: Make that 99% for pipeline companies)

http://virtualpipeline.spaces.live.com/

 

[BigInch]

You don't disagree with anything else do you?

**********************
"Pumping accounts for 20% of the world’s energy used by electric motors and 25-50% of the total electrical energy usage in certain industrial facilities."-DOE statistic (Note: Make that 99% for pipeline companies)

http://virtualpipeline.spaces.live.com/

 

[Gorcan]

You may have resolved this, if so feel free to ignore me. But I would table the question what is the fluid source? Are the pumps in an oilwell, riser or flowline? Before I throw in my tuppenceworth.

 

[powersoff]

The fluid source is an aquifer. We are in the process of having drawdown/pumping test done.

 

[Artisi]

A good description of the project might go a long way in helping everyone to understand what it is all about. Are you pumping from a well system? - and starting from basics like establishing inflow etc sounds like a good move to me.

 

[Gorcan]

Hi Guys

Yeah it sounds like the Aquifer cant supply the fluid required to maintain the pressure, so it’s not really an electrical problem. If you are testing drawdown, and inflow that is the best route. The system may have to be rethought around what your aquifer can supply, and indeed will supply, with VFD's you can lower the frequency which will to reduce drawdown, or else leave, some offline. Or alternatively batch produce and allow the source to recharge. Hope it all goes well rgds.

 

[BigInch]

Without having vfdtrying to provide high head at low rpms working against fixed static head you can always turn the pump off and on again. A rather novel idea, no?

**********************
"Pumping accounts for 20% of the world’s energy used by electric motors and 25-50% of the total electrical energy usage in certain industrial facilities."-DOE statistic (Note: Make that 99% for pipeline companies)

http://virtualpipeline.spaces.live.com/

 

[itsmoked]

29 May 09 7:54
Without having vfdtrying to provide high head at low rpms working against fixed static head you can always turn the pump off and on again. A rather novel idea, no?

Yes. But nowhere near as informative as finding some point where the pumping rate exactly matches the infill rate.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[Artisi]

that's if the inflow can meet the required flow at the drawdown.

More detail needed, at the moment we are second guessing.

 

[BigInch]

Why is that so important. Its not the same rush you get when you land a plane in a 40 kn crosswind by tracking it exactly over the rabbit, land on the white line and still make the first turnoff. Fact is that on/off works just as well if not better.

**********************
"Pumping accounts for 20% of the world’s energy used by electric motors and 25-50% of the total electrical energy usage in certain industrial facilities."-DOE statistic (Note: Make that 99% for pipeline companies)

http://virtualpipeline.spaces.live.com/

 

[wimple]

Hi,

Just some thoughts...
If the number of starts is so high, could means that the pump is delivering higher capacity then the capacity pump has been selected/quoted.
One probable reason is that the system curve is very flat and system losses are lower then expected.


I deem that lowering pump speed (all pumps at same speed) could allow to reduce/mitigate the number of starts...

So it looks like :
- The system resistance downstream has changed during pumps lifecycle or there is a strong variations : min/max scenario of the system curve.

- Selecting 5 pumps in parallel is quite unusual. In any case, It seems the pumps have been quoted with the OP shifted on far left (what is impeller type : is it closed/open impeller, screw like for handling fiber, sand, etc? ) fitted for high losses scenario

Just trying some brainstorming ... one could try two control scenario :
(1) max system losses (5 pumps together with OP at left let say 20%)
(2) min system losses (2 or 3 pumps and cyclical swaping to reduce total capacity and then reduce starts)

 

[thewellguy]

I guess I don't fully understand whats going on with this system, however it sounded like he is using submersible well pumps. Most of then require a min speed of 50% or higher depending on the tdh due to the thrust bearing and min flow requirements. Sound like someone need to get a handle on the hows and whys of the system.