The question of whether SWMM5 can accurately simulate the flow conditions in pump station force mains in a wastewater collection system was raised and discussed in the SWMM Users Group in 2008. Force mains by definition always flow full. Rises in pressure occur in the force main following pumps turning on, and drops in pressure occur following pumps turning off. These changes are called “water hammer” and can result in pressure and flow rate oscillations in the force main.
SWMM5 has a new force main feature that did not exist in earlier versions of the SWMM program. Conduits may be specified as circular force mains that always flow full and have friction loss calculated using the Hazen-Williams or Darcy-Weisbach formulas.
In SWMM, the junctions are always considered to be vented and the drop in pressure is limited to the invert of the conduit. It is a common practice to install air and vacuum valves on the high points of a discharge header/force main. These valves release trapped pockets of air when pressure rises and allow air to enter the force main when pressure drops occur. The entrance of air prevents the pressure from dropping below atmospheric which thereby prevents vapor cavities from forming. After vapor cavities form, they later collapse when pressure rises thereby damaging the conduit.
The capacity and location of the air and vacuum valves in pump station discharge headers/force mains is evaluated using software programs such as Water Hammer and Mass Oscillation (WHAMO) by the U. S. Army Corps of Engineers. The issue raised in the SWMM Users Group is not related to the design of air and vacuum valve systems at pump stations and in force mains, but whether SWMM5 can reasonably simulate the flow out of the pump station force mains as pumps turn on/off.
A SWMM5 model was set up with a pump station that lifts flow from a low reservoir into a force main that runs to a high reservoir. This SWMM5 model was run using the dynamic wave solution technique with a range of computational time steps. The results were compared to those from simulations using the Transient Analysis Program (TAP), WHAMO and Infoworks CS.