Pumps in series with transfer pump using VFD.

[chethboy]

Hello,

I am working on a crude transfer pump application where in we need to add a booster pump as there is no sufficient NPSHa. The booster pumps shall be a centrifugal inline type ( 53m3/hr@50 m head)that feeds a high pressure centrifugal transfer pump (53m3/hr @ 690m head). The high pressure transfer pump need to operate with variable speed drive to vary the flow and head as required by the process downstream. I need your help to know if we can operate the booster pump at constant flow and head using a DOL starter? Do we need to use an automatic flow control valve between the booster and transfer pump ? Or will a simple bypass line with orifice to meet booster pump MCSF will suffice ?
In terms of starting sequence can both the pumps start together or do we need to have few seconds of delayed start between booster and transfer pump?
Anything else to take care within these pump loop ?

Appreciate your inputs and thank you in advance.

 

[Artisi]

First question is, if there is insufficient NPSHa for the high pressure pump,is there sufficient for the booster pump?

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)

 

[georgeverghese]

In most cases, it should be possible to use the min flow controls provided for the HP transfer pumps since LP and HP pumps are in series, assuming you can set the min flow to be suitable for both pumps. Presume the min flow recyle you have now is in dual service for min flow and level recycle.

If the above is possible, and if the LP pump is less than 100hp, then an FIA-L may be adequate on the LP pump discharge to account for accidental startup of the LP pump with discharge valve closed. If hp > 100hp, an FZLL is typically required.

Post a sketch of your current perception of pump line up, process controls both within and outside this pump set so we get an overall view of capacity controls, including the source vessel level / pressure controls.

 

[chethboy]

Thanks Artisi,
I completely agree with your statement, we are looking for a cost effective solution and adding a booster with out proper analysis does not come cheap. We do not have sufficient NPSHa to run the HP pumps directly through out the pump curve. There is sufficient NPSHa to meet the LP pump NPSHr. Attached P&Id provides the details.

Hello George,
Thank you for the feedback. I have attached the prelim P&Id that I could think of for your reference. We need a cost effective solution hence was thinking if we can live with restriction orifice to maintain the pumps MCSF rather than the Automatic re-circulation valve. My major concern is to make sure that the DOL LP pump does not suffer while we vary the flow on HP using VFD. Also I need to be sure if I am placing the NRV's and PRV at correct locations, any feedback with really help me finalise these.
Any input on starting of these units,do we need to have a time delay.
Appreciate your time and efforts in providing these inputs.

 

[Artisi]

Simple answers to simple questions.
You could start the feed (booster) pump DOL, this should be prior to the main pump start-up.
Depending on the frequency of minimum flow rate of the feed pump will or could dictate the best method of by-passing flow, this should be a design consideration. Simple bypass with orifice could be sufficient, however I cannot make an experienced comment on an automatic flow control valve.

Question: why such high head requirement (50m) for the feed pump, you should keep this to the minimum head to satisfy the main pump hydraulic requirements.

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)

 

[LittleInch]

For some reason I can't open your attachment, can you check it has a proper file type name and try again (maybe shortern the file name?)

I'm not a fan of min flow bypasses, but occasionally have to use them. The key is what is the operation? Long term flow with a varying flow or a stop / start operation.

I'm having difficulty believing that the main HP pump can vary flow that much that it would affect the booster pump.

Centrifugal fixed speed pumps operate effectively as constant pressure sources ( within 20%) and vary flow depending on the downstream system.

A relatively small pump like the booster pump should be able to operate continuously at about 30% of rated flow without damage to anything.

If your operation is relatively continuous and is one booster feeding one pump then I wouldn't bother with any flow diversion between booster to HP pump, start the booster first, wait either a few seconds or use the HP inlet pressure to tell you when to start the mainline HP unit and go from there. Just have a delay on main hp pump contactor so if the hp pump isn't running then it trips the booster



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

 

[georgeverghese]

File attached doesnt load up. Agreed, previous experience was also for simultaneous startup of both LP and HP pumps, hence the min flow recycle on the HP pump discharge should do. This is because the HP pump will be started with recycle valve forced fully open at min speed - no load operation. A timer delayed PZLL on HP pump suction or loss of run contact at MCC can be used to trip out the HP pump.