variable suction conditions

[drain]

Hi
My water supply for a land development project is described below:
Potable water supply from a 18 in. aqueduct
For 2 weeks per year during maintenance, positive suction is 15 - 20 psi
For 50 weeks per year, positive suction is 90 - 120 psi
Discharge head is 160 psi, discharging directly into a potable water reservoir

Initial minimum discharge is 475 USgpm (first 2 years or so), but since I'm only filling a tank once a day, this can be done in a shorter period at higher flow
Ultimate discharge (about year 6) is 1650 USgpm
I was going to use 1 pump initially (with one back-up pump on standby), with a second and third added in years 2 and 4
My pump suppliers tell me that due to the range in suction conditions, variable speed drives is the only way it can be done.
Can anyone think of another way?
Also, re the piping layout, I initially laid it out with suction and discharge headers in the station and the pumps between them, all in one plane. I've since thought more about putting the discharge manifold above the suction manifold to economize space. Are there any drawing resources available showing different station lay outs which I can have a look at?

Thanks.

 

[quark]

How your suction pressure gets varied and what type of pumps you have? As the suction pressure is varying greatly, you can make better use of it if you go for variable speed arrangement.

During high suction pressure period, your pump runs to right of the curve (i.e discharges more and draws up more power) and may overload your motor.

If you don't want to spend more, either you have to control the discharge valve (which results in more frictional losses) or go for a second impeller to suit your requirement. (but this depends upon the pump you selected and may not be practical with such a wide pressure range)

Layout is a trade off between space and pressure loss. Anyhow you should give the details of the pump.

 

[zdas04]

I'd build a manifold arrangement that lets the standby pump act as a booster into the second pump during the maintenance period. Look at the pump curves and see if it can do 15 to 90 psig and design your main pump to always do 90 to 160 psig. As demand grows alternate adding pumps to the standby header and the main header. When maintenance is not being done have them all on the main service. This would require clever valving and piping, but it wouldn't be terribly hard.

David