Fire Pump Acceptance: What 57 psi at 150% Really Means

The pump fails the 150% point and should not be accepted on this data.

• Churn (0 gpm): 108 psi. The shutoff pressure must not exceed 140% of rated head: 0.140 x 95 = 133 psi. 108 psi passes, and it is a reasonable margin under the 175 psi component rating.
• 100% (750 gpm): 94 psi. Rated head must be achieved within tolerance - the field result should be at least 95% of rated head, i.e., 90.25 psi, and 94 psi passes comfortably. This is the only point the contractor is citing.
• 150% (1,125 gpm): 57 psi. The criterion is not less than 65% of rated head: 0.65 x 95 = 61.75 psi. The measured 57 psi is 4.75 psi low (about 8% of rated head) and fails.
• Speed correction: the pump ran at 1,750 rpm instead of 1,760. Affinity laws scale head with the square of speed: 57 x (1760/1750)^2 = 57.7 psi. Correcting for speed does not close the gap.
• The pattern matters. A pump that is fine at churn and at 100% but falls off at 150% is not 'within 5%' - it is a pump whose curve does not match the submitted submittal. The 150% criterion exists because the pump must be able to protect the system while significantly overdemanding, which is exactly what happens during the early stages of a real fire.

Before rejecting the machine outright, verify the instrumentation and the test method: calibrated gauges, a verified flow measurement (pitot readings at the test header, or a flow meter), correct suction conditions, and that the test header is actually capable of flowing 150% of rated capacity without the header itself becoming the limiting orifice.

Suction is not a credible explanation on these numbers, but it is the right thing to verify before condemning the pump.

• The suction stayed positive. 22 psi static dropping to 12 psi at 1,125 gpm still leaves roughly 12 psi gauge, or about (12 + 14.7) x 2.31 = 61 ft absolute - far above the net positive suction head required by a split-case pump in this range. A pump starved of NPSH would show a much larger, non-linear collapse in head, noise, vibration, and often a gauge fluctuation; a clean 4.75 psi shortfall at the top of the curve is a curve problem, not a suction problem.
• The 10 psi suction drop is normal and expected. It reflects the friction loss through the 6-inch RPZ at 1,125 gpm plus the city main loss. If a designer wants to raise the 150% discharge pressure, reducing that loss (larger RPZ, larger suction piping) would help a little, but you cannot recover 4.75 psi of head by improving suction when the suction is already positive - pump head is measured across the pump, not from atmosphere.
• What to check: gauge calibration and location (suction gauge upstream of the pump flange, discharge gauge downstream, both at the same elevation), pump rotational speed, the submitted versus as-built pump curve, impeller trim, and whether the flow measurement at the header was accurate at 1,125 gpm.
• Disposition: retest with calibrated instruments. If the shortfall repeats, the pump does not meet the specification and the remedy is a replacement impeller/pump or an engineered variance from the AHJ - not acceptance by argument. The contractor cannot simply accept a pump that does not meet the code criterion.

The proposed settings will produce a jockey pump that runs almost continuously and a main pump that starts on trivial fluctuations.

• Start from the churn pressure. The pump produces 108 psi at zero flow. If the jockey stops at 110 psi, it can never reach its stop point - the jockey would run until it overheats or until the relief valve lifts. The jockey stop pressure must be below the pump churn pressure with a margin.
• Keep the sequence in order and spaced. A workable arrangement is: main pump starts on a pressure well above the highest system demand but several psi below jockey start; jockey starts roughly 10 psi above the main pump start; jockey stops roughly 20 psi above the main pump start. For a system whose calculated demand at the base of the riser is, for example, 70 psi, a common arrangement is main pump start at 80 psi, jockey start at 90 psi, jockey stop at 100 psi - all comfortably below the 108 psi churn. The proposed 95/100/110 inverts the required margins and puts the operating band against the churn pressure.
• Check the relief valve. With 108 psi churn and components rated 175 psi, the relief valve must be set to protect the lowest-rated component and to pass the pump's excess flow at churn; it must not be set above the rating of any system component. Do not assume a factory setting.
• No flow meter. Without one, the owner cannot perform routine no-flow runs with any measurement, cannot trend performance, and cannot detect degradation between annual tests. Recommend a listed flow meter in the pump discharge, and confirm the annual flow test and the weekly no-flow run are set up in the NFPA 25 program from day one, with the baseline curve recorded at acceptance. The baseline is the single most valuable document the owner will get from this test day.