Mimum Flowrate centrifugal pumps

[Gilles12]

Hi everybody,

We use three centrifugal pumps to transfer refined product through a 12" and 258 km pipeline. For a softer start-up, we want to install a VSD driver.
I wonder if there is an empiric formula to determine minimal flowrate in the function of pump rotational speed (like the law of similarities). I didn't hear about that.

Thanks

 

[RVAmeche]

Don't take on that liability. Get the minimum continuous stable flow (MCSF) from the vendor. If they just call it "minimum flow" but don't specify what kind, push back on them. I've gotten burned in the past being told a certain minimum flow was continuous but really it was thermal minimum flow. Whoops...

 

[EdStainless]

What is the reason for the softer start?
Power side or piping side?
Are you starting against closed valves that you gradually open as the pumps reach speed?
Have you looked into other soft start options?
Full VFD just for soft start is overkill.
And whenever you are running you will have power conversion losses, higher motor losses, and maybe reduced motor life.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[LittleInch]

Softer start up of what?

Flow? Motor Amps?

I don't understand why you want to do this or what the question actually is.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[1503-44]

Vfd controlled pumps do not really have minimum flow requirements. Power consumption at low flow is so low, no significant amount of heat is generated.

--Einstein gave the same test to students every year. When asked why he would do something like that, "Because the answers had changed."

 

[LittleInch]

So much depends on the profile in the line.

If you've got a high spot then until you generate that head you will have no flow...

Not enough detail to work out your issue.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[1503-44]

Right, in which case vfd does nothing for you.

--Einstein gave the same test to students every year. When asked why he would do something like that, "Because the answers had changed."

 

[Gilles12]

Hi everybody,

Thanks for your answers.

RVAmeche, thanks. I will contact vendor

Edstainless, yes we start pumps one after another using a closed valve being open gradually in a common discharge line. Each pump starts in a stars-triangle mode. This usually cause very high pressure because of the line design. The Line was designed for two pumps and now because of the need to increase pipeline capacity, we are running three pumps. But i wonder if each pumps will have the same minimum flow running on differents speed?


Littleinch, the soft starter is for motors. For more understanding, read the answer to Edstainless.

Littleinch and 1503-44 the problem is only
about the rapid increase in pressure in the discharge line.

 

[LittleInch]

Sorry but still makes no sense.

Can you describe a bit more your current design, operation and what your problem is with pressure, flows etc.

I really don't think there is an issue about minimum flow if you're trying to pump more flow. But your third pump if its in parallel might not increase flow very much if at all.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[georgeverghese]

Why do you see the need to calculate this, pump vendor ought to provide.
Pump startup current concern may not be the only reason for VSD, power savings at part load should be the main reason for VSD.
If startup current concern is the only reason, there are other ways of enabling slow ramp up on speed - speak to your electrical engineer.

 

[1503-44]

Gilles, High discharge pressure required. Exactly. Why have a vfd start against a closed valve. It is constant speed pumps that must start against closed valves. When starting against a closed valve, a vfd does nothing for you. It would only increase the time to reach discharge pressure and start flow into the system.

You have electric motor, so no warmup/speedup cycle for a diesel engine is needed.
Why would you otherwise want to delay your pump from reaching max pressure as soon as possible.
Electric motors have high output torque on startup. Do you have a very small, underpowered motor?

What is the new pipeline pressure?

I think you would be creating a need for vfd by installing a vfd.

As others have said, a soft starter might be what you need, and only to minimise inrush current draw.

--Einstein gave the same test to students every year. When asked why he would do something like that, "Because the answers had changed."

 

[pierreick]

Hi,
Probably good to review your start up procedure! Valve being closed @ discharge you are reaching the shut off pressure. You may want to start one pump with discharge closed or partially closed then open the discharge and start the other pumps one by one. I guess you have Check valve on each line to prevent back flow, before the discharge valve.
As suggested supply a PID of your unit to better visualize your system and some info about pump (head, flow rate and pump curve).
I don't see the need for a VFD.
You may want to review the operation of pumps running //.

https://www.pumpsandsystems.com/pumps/what-you-need-know-about-parallel-pump-operation

https://blog.craneengineering.net/operating-centrifugal-pumps-in-series-or-parallel

My view.
Pierre

 

[FMJalink]

Hi,
If I understand your question correctly, I think there is not a simple emperic function. Your main problem is to accellerate the 258 km of liquid in the pipeline without getting too high pressure for too long time during start up with a filled pipeline. You can get an impression of the situation by simulating the situation. The total mass in the pipeline has to accellerate, which can be seen as a differential pressure. The pipeline also has system curve forming function of pressure drop caused by flowrate.

Those 2 pressure drops together with a possible static pressure drop in your system has to be delivered by the pumps.

The pump curve is determined by the the rotational speed of the pump. The pressure of the parallel pumps at the pipeline entrance is equal. If the pump discharge valve is not (yet) fully opened, this gives additional resistance (less pump pressure).

This combined should allow making curves in a spreadsheet showing situation and determining a safe start up acceleration of the pump rotational speed.

I hope this helps.

 

[Gilles12]

Hi,

Thanks to all,

FMJalink I got it and will do as you said and come back.

Pierrick, your links are very helpful

 

[LittleInch]

"This usually cause very high pressure because of the line design."

If you explain the line design and how the units start with pressure / flows etc then we might be able to help more, but you're not giving us any details.

I really can't see how starting a main line pump on a pipeline is going to give you very high pressure. That's a long pipeline for only one set of pumps and there will be a need to maintain the pipeline in a pressurised condition when you stop it otherwise it will be very difficult to start without generating large surge pressures. Is this what is happening?

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[Gilles12]

Hi everybody,

For more détail please note that:

- pumps are 100 bars/ 100 m3
- pipeline spec : Api 5L Max P : 100 bars
- pumps are started-up directly with a
control valve closed being open
gradually. The valve is operated
manually.

- with two pumps running with the valve fully opened the pressure in the Line is about 90 bars. So when starting-up the third pumps without soft starter there will be a surge of pressure.

 

[georgeverghese]

What is the max source pressure for these pumps ? And max density of the pumped fluid.
Can we also see a Q-h curve for these pumps at the speed these pumps run at the moment.

 

[1503-44]

What reason is there to start with a closed control valve?
Starting with vfd and closed control valve makes no sense. It just takes longer to get to the point when the control valve can open.
Probably better to start in parallel with a check valve in parallel with your other pumps.
API 5L is not a design specification/code. It is a pipe spec only.


--Einstein gave the same test to students every year. When asked why he would do something like that, "Because the answers had changed."

 

[pierreick]

Gilles12,
Please provide the documents requested to get meaningful answers (PID, Pump curve, Operating point, start-up procedure). Every day we are discovering something new!
Please tell us about your role within your organization for better support.
Regards
Pierre

 

[LittleInch]

Gilles,

With respect, the answers you are giving tell me you really don't have a lot of experience in this field and you seem to be expecting miracles. What is your design code, what is the DESIGN PRESSURE of the pipeline, wall thickness, pipe grade etc (API 5L comes in many grades)

If your current two pumps are running at 90 bar inlet and your design pressure? is 100 bar, there is virtually no increase in flow available to you.

You need in practice to have a bit of a gap between max operating pressure and design pressure to allow for alarms, trips and relief valve settings. Typically this is 8-10% below design pressure.

So basically you are already operating flat out.

Pumps - I'm assuming these are centrifugal pumps - do not come in units of flow so 3 x "100m3/hr pump" does NOT mean you can flow at 300m3/hr. It depends on what the system curve of your pipeline looks like. Centrifugal pumps are essentially a constant pressure machine with variable flow.

What you propose - VFD - will not work in this instance as until the speed increases to the point where the head of the pump (proportional to speed^2) is equal to or a bit higher than the other two running pumps there will be no flow through this third pump.

All that will happen is that your three pumps will more or less share the flow between them, but this flow in total will not be more than 5% higher than what you are getting with two pumps. So there is no point in starting a third pump. However if you really really want to to prove the point then simply start it against a closed control valve (globe valve or similar) and then SLOWLY open it. Starting against a gate valve or a ball valve won't give you the control you need. Even then your three pumps might exceed you max operating pressure / pump high pressure trip.

Bottom line - give us ALL the information we've asked and then we can tell you for certain. But you can't re-invent the laws of fluid flow.

If you want more flow then you need to build a booster station at the hydraulic middle point. This might not be the middle of the pipeline physically. Depend son the topography / profile of the pipeline.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.