Varying head rise in Centrifugal pumps

[kbka]

Hi PF!

I have a question regarding the use of centrifugal pumps.

I have a system with a constant flowrate of water at 2m^3/s pumping water from a lower to a higher reservoir. The inlet condition is a head of 10m. The Height of the upper reservoir will vary from 83m in the beginning to 287m at the end.

My questions are:
Should I use more than one pump?
If so
Any litterature that covers the need of multiple pumps? (My fluid mechanics books don't)
And
how does the varying head effect efficiency of the pump/pumps?

 

[Jobrag]

You need to get the head flow curves from the pump manufacturer.

 

[kbka]

I tried. With no succes. It seems that at this high flow rate and pressure, very few options exists. I've only been able to find performance curves etc. for smaller pumps. (Head and flow)

And again i can't really seem to find anything regarding how the varying head effects efficiency of the pump?

 

[kbka]

Is the solution to just use a pump that can deliver 287 m of head. And let it pump at that setting throughout the whole process?

 

[brewnog]

The manufacturer of your pump will certainly be able to provide a flow curve. Have you asked them for one? What's the pump?

 

[kbka]

I've asked some pump manufacturers about this and I never got a reply. Any suggestions to a manufacturer making centrifugal pumps at this scale?

 

[Q_Goest]

Hi kbka,
This sounds like another school project created by a clueless prof. The pump you're talking about is very large and the pressure is relatively high. The pipe this machine would be going into would be roughly 1 meter in diameter and would have to be relatively thick wall. Power is on the order of 10,000 hp (7500 kW). Any industry project that would install a pump like this would be budgeted for millions of dollars and have at least a handful of engineers working on it, possibly many more depending on the civil work required which no doubt, would be very significant.

The way these projects are done is generally something gained from considerable experience. The engineering team has similar projects to fall back on and compare to and a number of suppliers for machinery this large. That engineering firm would be familiar with all the manufacturers for machinery like this and experience with the issues involved. This isn't a project that someone picks a pump off a shelf or from a page on the internet, so you won't likely find a flow curve. And whether or not a single machine is used or multiple machines would be part of a trade study done to estimate the overall cost of installation and operation. The cost of operating this system (energy) would be a prime consideration in a system this large. So the type of pump, how many would be used, the elevations they would need to be placed at, etc... would all be considerations looked at by the trade study.

Pumps for this would be one-off machines made for the application with manufacturers drawing on a set of pump frame sizes and tailering the machine to fit the application. It could be a multistage machine and you may find that you need a way of varying pump speed due to the fairly broad range in output pressure. Another way to accommodate the outlet pressure variation might be to install multiple pumps at varying levels, but all those considerations are part of a trade study that would be done to determine the most economical solution. Hope that helps...

 

[Q_Goest]

I've asked some pump manufacturers about this and I never got a reply. Any suggestions to a manufacturer making centrifugal pumps at this scale?

I'm afraid, that's because they don't want to spend time educating college students (no offense intended). You might want to consider 'creating' data for your class project based on much smaller machines that are available off the shelf.

 

[kbka]

Q_goest,

Thank you for your thorough reply!

This project is a setup that would (if ever realized) cost many millions of dollars. It would be a high capacity centralized Energy storage system. Its a new concept (Which I unfortunately have to keep secret) where I'm estimating the round trip efficiency of the system.
I'm not necessarily looking for a certain make and model, I just need to estimate the energy consumption of the pump during operation.

I think I'll just take a "regular" head flow curve and make it fit my system.

 

[Q_Goest]

Thanks for the clarification. I take it you're doing the preliminary work for a trade study of sorts then? I might suggest looking through the literature for similar installations around the world. I'm sure there must be peer reviewed, published articles on similar systems, various measured efficiencies, and many other aspects of the project. They should give you a good starting point and some documentation to fall back on to give your study legitamacy. I'm sure there would be much to learn by researching similar systems.


(just thinking out loud here...) I wonder if these systems have multiple pumps for redundancy, for example, two pumps either one of which might be capable of most if not all the flow required. Three pumps, two of which would be capable of producing the flow required, etc... By doing it this way, the pump size is reduced and reliability/redunancy is increased.

 

[russ_watters]

I'd browse the aurora or b&g catalogues and see what the biggest pumps are, then start looking at how efficiency varies along the curve. You could spend data playing around with different configurations and at this level that's probably sophisticated enough.