Can EPANET calculate a network with multiple pumps?

i have a network with 3 bore hole pumps that connect to a main manifold and from that manifold one pipe fill a tank.
can epanet simulate such a network?

Its possible, I try to model a system having 12 boreholes and successfully run ,
Just assign neg demand at each node which represents a boreholes. Then run


how to model booster pump?

If you want to model the borehole pump,
you can model each borehole as a sequence of reservoir and pump
The reservoir head will represent the water level in the aquifer
The pump is characterized by a pump curve

1 Like

thanks gal.
I’m actually know how to model a single borehole pump (I’m doing it exactly as you wrote).
the question is can epanet simulate multiple pumps (pump’s curve, it doesn’t necessarily need to be a bore hole pumps) that feed one manifold.
I’ve attached a simple drawing to show what i mean.

those pumps are individual and connect parallelly, not in a row.
those pumps connect to a collecting manifold, and from that one a single pipe fill a reservoir.

@aridzv Have a look at this network here with three pumps in parallel:
The INP file is here:

thanks elad.
i have 2 more quotations, if I may:

  1. i didn’t see any P.S.V’s after the pumps. how do this model set/fix the pump’s flow?
  2. can i use Cramer’s rule to develop the head-flow equation from a given pump graph?


@aridzv, regarding your questions:

  1. The pump flow is determined by the intersection of the pump curve and the system (resistance) curve. If you aim to set a constant flow you might consider using a variable speed pump.
  2. I am not familiar with Cramer’s rule in the context of pump curves. EPANET allows you to insert the curve as a set of discrete points and then it will build the continuous graph on its own



  1. my question was referring, in my case, to a pump station design, or as it calls in the irrigation world - system head or head control.
    when I do the design I will use a reservoir at a set level and then I will get the head loss.
    the flow (Or demand in EPANET language…) is usually known. the required pressure, or head, in the outlet of the pumping station is also known, so in this way and with a proper use of PSV and FCV I get head loss between the pump and the end of the pumping station, add it to the required pressure at the end of the pumping station and in that way I get the pump working point, and then I can select a pump.
    the use of a PSV AFTER the pump is selected is to set the pump’s flow, and it is the way it is done in the real life, not only as a design method.
    and this lead’s us to next issue:

  2. EPANET simple way of creating a pump curve is to enter points of head and flow. this is basic.
    I wonder if EPANET allows you to enter a pump equation instead.

  3. Cramer’s rule allow you to use a set of known values to produce an equation.
    for example, 3 point will get you a Y=ax^2 + bx + c.
    my question\dilemma is, if this equation is accurate enough?
    I saw in some publication and software’s that use it to get a pump curve equation, but most pumps
    equations provided by the manufactures are not straight forward exponential equations.

i think it is better to add only one(1) valve to the collecting manifold, economically and for maintenance point of view.