The panel shows two commanded fire water pump starts, followed seconds later by start commands to all five pumps. Ring-main pressure collapses as the deluge valve opens, every PSLL changes state, and the backup-start logic reacts exactly as configured. The first two pumps need about 15–20 seconds to accelerate, while the valve reaches its open position much sooner.
Start here: prove the event sequence with one synchronized trend. Do not add a blind valve delay or move pressure setpoints before you know which signal creates the five-pump cascade.
Confirm the start-command chain
Trend the following signals against one time base:
-
F&Gdetection and confirmed-fire inputs - Start commands for all five pumps
- Running, speed-ready, and fault feedback available from each driver
- Ring-main pressure at the pressure-switch location
- Each
PSLLstate - Deluge command and valve-position feedback
Read the sequence from left to right. If all five start directly from the detection signal, the problem is in the command logic rather than the pressure transient. If only two receive the initial command and the remaining three start when their PSLL inputs operate, continue with pressure and valve measurements.
| Observed symptom | Probable cause | Next check |
|---|---|---|
| All five start with the detection input | Detection logic commands every pump | Review the cause-and-effect matrix and start outputs |
| Two start first; three follow as pressure falls | Deluge demand outruns accelerating pumps | Measure pressure, valve travel, and pump response |
| One of the first two never accelerates | Starter, driver, permissive, or pump fault | Inspect command-to-running and command-to-speed intervals |
| Pressure rebounds and overshoots after all pumps start | Excess pumping and rapid hydraulic transitions | Record the surge at several ring-main locations |
Measure the hydraulic gap
Record pressure fast enough to resolve valve movement and the 15–20-second pump acceleration interval. Take readings at the switch manifold and at other accessible points on the ring main. A single slow panel value can hide both the minimum pressure and a short surge.
Compare four events: initial valve motion, rapid pressure decline, each PSLL transition, and recovery as pump output rises. Obtain the actual valve-position trace; command status alone does not show travel time.
The pressure collapse occurs while deluge outflow exceeds pump inflow. The temporary volume deficit is:
Volume deficit = integral of (deluge flow - pump flow) over the acceleration interval
Pressure falls as that deficit consumes the stored volume provided by water compressibility and pipe-system compliance. The jockey pump maintains static pressure and supplies small leakage demand. It is not the source for full deluge flow, so increasing its runtime or changing its start band does not correct the transition.
Separate acceleration from setpoint trouble
Check the actual pressure settings and reset behavior of every PSLL. Compare them with the approved cause-and-effect and hydraulic design values. If all switches operate at nearly the same pressure, a single transient can call every pump at once.
Do not lower the switches merely to suppress starts. A lower threshold may hide inadequate pressure and delay legitimate backup operation. Staggered thresholds or sequenced start delays can prevent simultaneous starts in a multiple-pump installation, but the final settings must still start enough pumping capacity before pressure falls below the firewater design requirement.
Next, measure each initially selected pump from command to running feedback and from command to useful hydraulic output. A running contact proves that a starter or driver reached its feedback state; it does not prove rated speed, discharge pressure, or available flow. Pump flow also depends on the operating point where the pump curve meets the system curve. Use speed feedback, discharge pressure, and the applicable pump performance data to establish readiness.
If the pumps respond later than their approved performance requires, repair that problem first. Valve sequencing cannot compensate safely for a slow or unreliable driver.
Choose the resolving branch
Use the measured sequence to choose one branch:
- Incorrect command logic: Restore the approved arrangement in which the required pump pair receives the intended fire signal and additional pumps respond only through their assigned backup criteria.
- Slow pump response: Correct starter, driver, permissive, mechanical, or discharge-path delays. Repeat the acceleration test before changing deluge behavior.
- Valid pump response but excessive transient demand: Evaluate controlled deluge opening with a transient hydraulic model and a witnessed flow test.
- Incorrect or undifferentiated pressure settings: Recalculate the start sequence from the approved pressure requirements, pump curves, static pressure range, and instrument tolerances.
A delayed or slower valve opening is acceptable only when the firewater design authority confirms that the hazard receives its required water application within the required response time. A blind timer tied to an assumed 15–20-second acceleration period is weak protection: a pump can fail while the timer still releases the valve.
If controlled opening is selected, base progression on validated supply conditions and valve feedback. Treat pressure, pump readiness, and timeout as separate conditions. Define what the valve does after missing feedback, power loss, controller failure, or failure of one selected pump. Those responses belong in the approved cause-and-effect, not in an undocumented logic patch.
Implement the approved sequence
- Preserve the direct
F&Gstart request to the pump pair required for the maximum single-zone demand. - Confirm that both start commands reach their controllers and that neither pump has an active fault or missing permissive.
- Define pump readiness using the available speed and discharge response. Do not use running feedback as the only proof of capacity.
- Apply the approved deluge strategy: immediate opening, controlled travel, or a bounded release condition validated by the hydraulic and fire-safety review.
- Retain backup pump initiation at the approved low-pressure conditions. Coordinate thresholds or delays so a real loss of capacity still calls additional pumps.
- Alarm failed valve travel, missing pump readiness, and abnormal pressure decay as distinct conditions so operators can identify the failed branch.
Do not treat valve travel percentage as flow percentage. Valve flow is nonlinear and varies with differential pressure. Confirm the proposed opening profile using valve data and measured flow or a validated hydraulic model.
Verify pressure, flow, and surge behavior
- Test from the normal pressurized condition with the jockey pump controlling the ring main.
- Initiate the same detection sequence used during routine testing and record all synchronized signals.
- Verify that the intended two pumps start, accelerate, and develop the required hydraulic response.
- Confirm that ring pressure remains above the approved minimum or that any backup start occurs exactly at its approved criterion.
- Verify deluge-valve travel, delivered flow, and application timing against the firewater design basis.
- Record the pressure recovery and peak surge after pump acceleration and valve movement.
- Repeat the approved failure cases, including loss of one selected pump, to prove that backup logic still operates.
Pipe rupture in the GRP system is a separate integrity problem. Sequencing may reduce a surge, but it does not raise the pipe rating or correct weak restraints, damaged pipe, or an invalid transient design. Compare measured minimum and peak pressures with the piping design limits and complete the piping correction independently.
Fixes that waste time include enlarging the jockey-pump role, changing PSLL settings without transient data, adding an arbitrary timer, or slowing the valve without checking delivered firewater. Each can conceal the initiating mechanism while weakening protection.
FAQ
Why do all five fire water pumps start during one deluge test?
The first two pumps require about 15–20 seconds to accelerate, while the deluge valve opens much faster. Ring pressure crosses every PSLL threshold, so the pressure-based backup logic calls the other three pumps.
Why does the jockey pump not hold ring-main pressure?
The jockey pump maintains static pressure and supplies small leakage demand. It is not sized or intended to match the full flow from an open deluge zone.
Why does a pump running signal not prove full capacity?
Running feedback only reports a controller or driver state. Confirm useful output with speed, discharge pressure, pump performance data, and the resulting ring-main pressure or flow.
Why not lower the PSLL settings to stop nuisance starts?
Lower settings can delay legitimate backup pumping and conceal an unacceptable pressure minimum. Set and coordinate them from the approved hydraulic requirements, static pressure range, pump curves, and switch tolerances.
When should I stop testing and escalate to official support?
Stop when a proposed sequence delays required firewater delivery, pressure approaches the piping design limit, valve or pump feedback is unreliable, or the transient remains unstable. Escalate through the firewater designer and official pump, driver, deluge-valve, and control-system support channels with synchronized trends, cause-and-effect logic, pump curves, valve data, and hydraulic calculations.