handle situation of pumps defined as "constant power"

[jjstickel]

EPANET input files sometimes specifies pumps as constant power. This is nonphysical and will result in constant-valued objective functions (at least with respect to that pump). We can assume the specified power is the power at the BEP, but more information is needed to get flow and head at the BEP. It seems the way out is to run a few flow-feasibility simulations of the network and examine the flows through the pump.

[jjstickel]

The "Kentucky" test networks largely define pumps as constant power (POWER designation in the inp file). Kentucky network 3 (ky3.inp) seems to be the smallest.