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

 

[Gilles12]

Hi EveryBody,

Thanks @LittleInch,

You are right! see below information asked

- Design Pressure 104 bars/ Test pressure 127 bars
- wall thickness: 12.7 mm
- Pipeline grade API 5L grade B
- MAOP: 100 bars

&Concerning Pumps there are centrifugal and identical and parallel layouts, there are designed to deliver 90 bars at 100 m3/h. you'll find attached the pump's curve and station PID.

the pipeline is 258 km long with a 194 m altitude between the inlet and delivery stations. the highest point is located 249.7km from the inlet station and is 243 m high.

Yes, we are discussing the creation of a boosting station 151 km from the inlet station.

 

[Gilles12]

here PID

 

[georgeverghese]

And what is max suction pressure to these pumps ? If it from a tank, what is max elevation diff between high level and pump centre line?
Dead head dp = 95bar at density of 860kg/m3.
And what is actual density of this "DDO" at min operating temp?

 

[pierreick]

Gilles,
C'est quoi le veritable probleme/crainte? rupture de canalisation ou de partie(s) d'equipements ou autre?
The pump data sheet was issued in 2008 , meaning that the line has been in operation for 15 years. Tell us what the real issue is?
Very flat curve (Delta P vs Flow rate), wheels already machined, 2950 RPM. Did you get other proposals, from other vendors?
Can you share the specification sheet (scope of this project), where we can see the variability of your process? (Min, Max, Normal values).
Can you share the altimetry, both suction and discharge lines to start to build a model? What about the pressure at the suction of the pump and discharge of the line?

Difficult to read the PID, missing the source. Can we get a PDF?
More info better answer.
Pierre

 

[LittleInch]

Also arrival pressure, trip pressure and what flow rate are you operating at with two pumps?

I think it might be about 220 to 230 m3/hr?

Not sure how you manage to start with one pump without going off the end of the curve? How does it start up?

When you shut down how do you do it? What is the settle out pressure at the pump end?

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

 

[LittleInch]

Also, as said above, your pump curves are really quite flat and when running pumps in parallel you can easily find one pump doing >60% of the flow. With three pumps it will be even worse.

But my previous comment applies - basically your third pump is going to add about 10m3/hr to the total if that.

You need about 21 bar to get any flow past the high point.

Do you have a flow assurance dept or team to get them to develop a system curve for you. Or decide where to the best location is for your booster station?

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

 

[Gilles12]

georgeverghese

the max suction pressure is about 1 bar, and tanks connected to pumps are at atmospheric pressure
The max elevation is about 13 m
The actual density range of transported products is 720- 870 kg/m3

Perreick,

I am thinking about how to start the third pump because of modification of the process (boosting station, VFD stater chosen design by the project team) and I was wondering about the influence of motor actual speed on pump minimal admissible flow (pump security). in the literature, it said a pump might be operated between 80% and 120% of the Best Efficiency Point which is about 100 m3/h. I don't worry about the line because we set up a security level of up to 97 bars.
Pumps data are very close to actual data.

Suction line: 1 m

Discharge line: 2.5 m but the pipeline is buried after a few meters ( see pid)

as I told you, in normal operating conditions suction pressure varies from 0.2 to 1 bar and discharge pressure from 88 to 90 bars

LittleInch,

Yes the flowrate is about 220 m3/h and goes up to 250 m3/hr ( when pumping gasoline)

the arrival pressure is 7 bars.

Start-up process :
- the control valve (PCV10 see PID) set to 20% or 25% opening
- Starting the first pump
- we set the control valve to 30% opening (to avoid the first pump being overloaded. At this stage, motor power reaches 495 Kw with a nominal of 450 kW) then we start the second pump and open gradually but quickly the valve so that we limit pressure surge on the pumps discharge side. so when reaching permanent operation conditions discharge pressure is 90 bars and the flow is 100 m3/hr. note that the discharge valve of each pump is automatically opened at the pump starting up ( 60 s to be fully opened).

Finally yes we have a project team, but a former team already chose the location I indicated and began the project on this basis. it is now very difficult to change plan.

 

[LittleInch]

Gilles,

Your method of starting is not good. IMHO.

The first pump is going way off the end of the curve if you're hitting 495kW - you must be doing about 150m3/hr...

You might need to start with the control valve at 10% and control on flow for one pump to get the line moving then start the second unit. As you're so close to the MAOP limit you might benefit from a soft start, but the pump speed needs to get up to the first one within about 30 seconds otherwise it will just stall as it's not producing enough head.

For your booster station I would start all three first then start the booster station.

SO yes, a soft starter might be needed or fit a proper control valve to the discharge of each pump so that you can control the initial surge of pressure until they all even out. As said before though, with that flat a pump curve it will be a miracle if all three units operate within 5% of the same power draw.

Your concerns about pump minimum speed and operating within 20% of the BEP are different things and I don't think you really understand what the issues are here. They are fixable, but might need a different mode of operation.

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

 

[Gilles12]

Thanks LittleInch,

 

[georgeverghese]

So max dead head pressure would be 1+(95*.87/.86) = 96.1barg, which is some 95% of MOP of 100barg - looks ok to me.
The process sketch attached shows no min flow recycle control loop for each of these pumps - this should help with startup woes. At this pump power, a dedicated safety low flow trip for each of these pumps would be good.

 

[pierreick]

HI,
If surge is the issue, you may consider this piece of equipment.
Good luck

Pierre

 

[LittleInch]

George,

Gilles has said "security level of up to 97 bars". Now I think that's lost in translation, but suspect it means trip or at least alarm.

Also I missed the information that the arrival pressure is 7 bar.

I hate min flow recycle loops, this only needs a control valve on each pump and you can play tunes on that.

gilles - Thanks for the response but a little more feedback would be appreciated. Did it answer all your questions?
Are you going to do anything differently?

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

 

[pierreick]

Hi,
Some hydraulic calculation to confirm that 3 pumps in // will not improve much the performance of the process.
2 pumps in //, qv = 220 m3/h @ 1000 mCl
3 pumps in //, qv = 230 m3/h @ 1050-1100 mCl
Good luck .

Pierre

 

[georgeverghese]

There is an FS 11 / 12 / 13 on these pumps - dont know what these are.
So what is the nameplate rated power on each of these motors ? - it isnt on the pump datasheet.

You will need a flow recycle loop during flow turndown, at least one loop on the common discharge header. The VFD on each of these pumps will probably get a speed reset from the pipeline PIC.

 

[pierreick]

Hi,
They are flow switches.
Detailed information available on the right side of the PID. Complete set in French, well documented.
Note: Big money if you intend to purchase VFD for each pump.

Pierre