These are engine driven fire pumps, its required by NFPA to run them once a week (diesel for 30 minutes).

What they did not account for is the diesel are two cycles (94 still made) and don't take kindly to low load (and they are hot rodded for the application as they will run less than 500 hours in 50 years.

Anyway, also required is an initial acceptance test to prove they do what spec says as well as yearly to confirm they still put out.

All that means flowing 4500 gpm in this case (maximum). Only way to do that is a test header so they T the discharge and isolate it with a valve. Want to test open valve (and the play pipes outside the building for various flow rates up to 4500 gpm.

Never meant to do more than one at a time, not intended as a remedy, but the pumps don't like churn (damage inside volutes) and the engine don't like low load so we have been testing at 100% flow, helps the engines but the pumps still get a weird form of cavitation even though we are flowing at 1800 gpm.

Flow loop should stop that, trying to figure out how much flow, good idea to test two at the same time and cut the test time down, but that created a recirculation loop through the non running pump (would stop if you turned it on) ergo, turning backward and doing an ugly sort of combustions via residual oil. All fun