Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations GregLocock on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

head calculation in parallel pumping

Status
Not open for further replies.

moideen

Mechanical
May 9, 2006
360
See an example, two cooling tower, two identical pump parallel, 1350 gpm each required for one CT to run one chiller. During the head calculation, I consider one pump throughout 1350 gpm? One veteran told me that put the 2700 gpm from header. Is it true, because I want to calculate the head for only one pump; throughout year, one pump will only work out.

ENERGY_STUDY-1_BABY_CHILLER_piwivt.jpg
 
Replies continue below

Recommended for you

Moideen,

I assume this is a carry over of your cooling tower post...

Your problem is first to define your system.

Is this only to ever run at 1350 pgm? if so all the flow in the system is simple - it's all 1350 pgm - straightforward.

However if you want two pumps to run and two chillers and two CTs then you're in a different space.

what you will find is that it is impossible to do this with two identical 1350 gpm pumps. The common bits of the system will experience an increase in friction drop of a factor of 4 for a doubling in flow. your pump cant increase head at a fixed flow if it is a fixed speed pump.

So your options are either:
Use a PD pump which will pump a certain flow and vary the head and hence power according to the frictional losses seen at that flow (lower for 1350 than for a total of 2700)
Use a variable speed pump with low speed calculated for 1350 GPM and higher speed to cater for 2700 gpm

How is your cooling tower doing by the way?

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.
 
IMO, besides the pipe sizing due to the friction loss, the symmetrical configuration may be considered for the pump and CT piping system to ensure the equal flow in a parallel operation system.
 
LittleInch: system design flow is 1350 gpm to be met by one pump. In my calculation with 1350 gpm , found the head 84 fthead, 2700 gpm from header to the common line head will be 90ft/hd. Now I will go with 1350gpm 84 ft/hd. I recall a discussion in this forum about the parallel pump operation that run two identical parallel pumps to meet the design flow, is the energy efficient operation. Is it better here to apply that two pump with flow rate 1350 with 84ft/hd.
 
Each system is unique so what works best for one application won't work for another.

In your case optimize the pump you buy for 1350 gpm and that will be more efficient than two pumps designed for 1350 gpm but actually operating at ~700 each.

Get the most efficient pump possible for a duty of 1350. At this sort of size, the efficiencies between different vendors can be quite big and hence a pump which may be a little more expensive to buy, but more efficient, will pay for itself in less electricity costs in a couple of years. But get the data and do the comparison.

Just use one flowrate for the entire system if this is what you are going to run it as - i.e. 1350 gpm.

If I remember right, your current pumps are just too small and hence you're not getting enough flow so it sounds like you're replacing the pumps??

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

LittleInch :
Yes, you are correct; the current pump is undersized, 55 ft/hd only. What you mean 700 each that select two pump each 700gpm with 84 ft/hd?
I have another benefit that if I consider parallel pumping that then can be evaded the electrical modification. The existing cable size 16mm2 and motor input is 18.5 kw (32A). If I go with 1350 gpm single pump with 84 ft/hd then the modification to be required for electrical side and motor base civil work. However, if I selected the 675 gpm/90 ft, I can use the same cable size and its contactors.
 
Don't you want the 2nd pump as 100% spare similar to the 100% spare cooling tower arrangement?If the 2nd pump is truly only spare why do you have to upgrade power? Make controls such that only one pump & one cooling tower can run.
 
Moideeen

That wasn't quite what I had meant, but yes you could do that.

However you need to note that getting two even supposedly identical pumps to do 50% each is quite hard. You will need to either choose a pump where the pump curve at your duty point is quite steep or have some means, e.g. orifice plates to balance up the pumps once installed. Otherwise one pump can be doing 60% flow to 40% the other one.

So doing that could solve your electrical cable issue, but as pointed out above, means you won't have a spare pump.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.
 
LittleInch
Thank you for your advice, the existing system is a 3 pump system, so i will replace with 3 each 700 gpm.one will be spare.did you mean the orifice plate is circuit setter? if all are identical pump, is then required circuit setter?
 
Basically yes. It's very difficult to get truly identical pumps or piping set ups and with low head pumps like these it doesn't take much for one to be out of sync. Maybe you won't need it but allow for it.

Install them them monitor current load is the easiest way to see if they are all pumping more or less the same (+ - 5%).

Be sure to let us know how it works out.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.
 
That's a pretty flat pump curve so a small change in actual head required or imbalance between units might result in some big changes in flow.

You will need to make sure your differential head of 92 ft is as accurate as you can make it without any margins.

As said before, just make sure you have some means of adding a few feet of head loss to one or both pumps to make sure they run at similar flowrates and at the intended flow of 700 GPM and not 800+ (only 5 feet less head loss).

You will though need to run both together at the same time always as one on its own will just run off the end of the curve and trip on excess amps / temperature unless you can really throttle it down or slow it down.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.
 
Is there only (1) 450 ton chiller and no plan to add future for spare or expansion? If so make sure to provide valved taps on the chilled water piping and valved bypass on the chiller to allow connection to truck mounted rental chiller in case the chiller breaks down.
 

LittleInch :
I have accounted 10% safety margin, so there is chance in head variation, so you are correct suppose 5 feet system head loss in operation, flow will increase to 800 gpm. The existing system already have DRV(FLOW REGULATING VALVE)or Better go with steep curve, please see the link grundfoss.
 
lilliput1 :there is no future expansion program, this is residential multy story building. same capacity standby chiller aslo installed.
 
It's easier to go with a steeper curve so that you don't need to go around adjusting flow valves, but up to you. Just make sure turn them both on together otherwise the first one will trip on excess amps

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.
 
LittleInch: the grundfos sales said that they dont have steep curve in end cuction model, steep curve availble in the type multistage system, will use multi stage in cooling tower operation? i also have contacted bell and gosset.
 
That's up to you to look at costs, performance, motor size etc.

My point is that if you have uncertainty about what actual head you need to get the flow and you only have fixed speed motors, then you could get some big changes in flow for only a few feet head difference.

You can install the units you listed, but will probably need to throttle them down a bit so as to give you your 700 gpm and not 800+ or to balance flow between the two pumps properly.

Once you have them set up then they should run fairly smoothly, but look at your options from other vendors.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.
 
LittleInch: May be I would go with the selection of grundfos, I got ksb pump selection, but it’s NPSHR is high(24ft). now I am thinking why I am replacing it’s motor. Replace only the pump. Because required motor specification is matching to the existing motor. Rpm, motor input is the same. Are there any other precautions required? thank you
 
Well I would send the motor spec and data sheet to the pump vendor to make sure the torque curves match the pump and then it's more a matter of physical compatibility in terms of the pump weight, location of mounting feet, whether inlet and outlet flanges don't hit anything on the pump skid, height of the shaft, diameter of the shaft and making sure your coupling connects to the pump shaft correctly.



Remember - More details = better answers
Also: If you get a response it's polite to respond to it.
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor