Main Switchboard Blackout Recovery & PMS Load-Sharing: Why the Bus Will Not Stay Up
A blackout that recovers cleanly is a PMS doing its job; a blackout that keeps tripping the bus on restart is a load-sharing, sequencing or protection problem. How to diagnose dead-bus closing, sequential restart and kW/kVAR sharing on a diesel-electric main switchboard before it costs a port call.
What a clean recovery looks like — and where it breaks
A correctly configured Power Management System (PMS) turns a total blackout into a brief inconvenience: the emergency generator picks up the essential bus, the standby diesel generator auto-starts, closes onto a dead main bus, the heavy consumers restart in their configured sequence, and the bus settles at a stable voltage and frequency with the load shared between running sets. When that sequence works, nobody calls us. We are called when the bus will not stay up — it recovers, then a heavy consumer restarts and the whole bus collapses again, or the second generator trips on reverse power the moment it parallels, or the frequency hunts and never settles.
The diagnostic discipline is to treat blackout recovery as four separate functions and test each one independently: dead-bus closing, sequential restart of consumers, kW (real-power) load sharing, and kVAR (reactive-power) load sharing. A failure in any one of the four produces a 'the bus will not recover' complaint, and the four have completely different root causes. Lumping them together as 'a PMS fault' is how vessels end up replacing a PMS controller that was never the problem.
Dead-bus closing — the first generator onto a cold bus
When the bus is fully dead, the first generator to recover does not synchronise — there is nothing to synchronise to. It performs a dead-bus close: the PMS confirms the bus is genuinely de-energised, confirms no other breaker is about to close onto the same bus, and closes the generator breaker directly. The most common dead-bus failure is a race condition where two generators both detect the dead bus and both attempt to close — the PMS interlock that should arbitrate which set closes first has either lost its inter-section communication or has a mismatched priority configuration between the two sections.
Diagnose by reviewing the PMS event log immediately after a failed recovery. The log timestamps every breaker close request and every interlock decision. A dead-bus close that is rejected with a 'bus not dead' flag when the bus is genuinely cold points to a stuck voltage-sensing relay or a feedback contact that is reporting the bus as live. A dead-bus close that two sets attempt simultaneously points to an interlock communication fault between switchboard sections — frequently a single failed comms card or a tie-breaker auxiliary contact that has corroded open.
Sequential restart — why the bus collapses on the third consumer
Once a generator is on the bus, the heavy consumers must restart in a sequence that never exceeds the available generation capacity. A bus that recovers, runs two consumers, then collapses when the third (typically a large pump or a thruster) attempts to start is almost always a sequencing or generation-capacity problem, not a generator fault. The single running set cannot supply the inrush of the third consumer, the frequency dips below the under-frequency trip threshold, and the protection correctly opens the generator breaker to save the set — which blacks out the bus again.
The fix is in the restart sequence and the standby-start logic, not in the generator. Confirm that the PMS starts the second generator before the load reaches the point where a single set cannot accept the next consumer's inrush. Confirm that the heavy-consumer restart table staggers the starts with enough delay for the bus to recover between them. We have closed many 'repeated blackout on restart' calls by correcting a restart sequence that had been re-ordered during a previous yard period and never re-validated under load — the steady-state load was fine, but the transient inrush sequence exceeded the running capacity.
kW load sharing — droop versus isochronous, and the CT on the wrong phase
Two generators on the bus should share real power (kW) within roughly 5% of each other, governed by either matched droop on both governors or a PMS-coordinated isochronous load-sharing line. The classic unequal-kW symptom — one set carrying most of the load while the other idles, or one set tripping on reverse power the instant it parallels — is most often a load-sensing CT problem. A CT on the wrong phase, or reversed, makes the PMS misread one set's contribution, and the set is either over-loaded or driven into reverse power.
Diagnose by capturing the load-share line signal and each set's actual kW with a power analyser during a parallel run. If the PMS thinks the sets are balanced but the analyser shows them unequal, the load-sensing CT or its wiring is lying to the PMS. If the PMS itself shows the imbalance, the load-sharing line, the governor droop setting, or the speed-control interface is the next step. A reverse-power trip the instant a set parallels is the textbook signature of a reversed load-sensing CT, almost always after a maintenance period when the CT was removed and reinstalled.
kVAR load sharing — the AVR side of the same problem
Real-power sharing is the governor's job; reactive-power (kVAR) sharing is the AVR's job, and the two are diagnosed separately. Two sets that share kW correctly but split kVAR badly — one set running at high excitation while the other runs cool — have a kVAR-droop or cross-current compensation problem in the excitation control, not in the engine. The symptom is often a circulating reactive current between the two sets that does no useful work but loads both alternators and trips overcurrent under conditions where the actual bus load is modest.
Confirm both AVRs are set to the same kVAR-droop slope and that the parallel-CT (the cross-current compensation CT) is reading the correct phase on each set. Swapping the suspect AVR or its parallel-CT input between sets isolates whether the fault follows the AVR or stays with the set — the same swap-test discipline that separates a four-figure AVR replacement from a low-cost CT secondary repair. Capture the per-set kVAR with the power analyser before and after any change so the improvement is measured, not assumed.
Proving the recovery before the surveyor asks
SOLAS II-1 requires the installation to recover essential services after a blackout, and the surveyor at the special survey wants the full drill, not a paper statement: main breaker open, bus dead, emergency generator picks up the essential bus, standby diesel generator auto-starts and dead-bus closes, heavy consumers restart in sequence, and the bus settles with the load shared. Running this drill on a quiet alongside watch or at anchor — under controlled conditions, with the engine room manned — proves all four functions at once and produces a class-witnessed log of the recovery times.
We attend blackout-recovery testing and PMS load-sharing verification at all major US ports under the standard wizard flow. The attendance captures the dead-bus close timing, the sequential restart staggering, and the kW/kVAR sharing under a transient load, and produces a single class-acceptable report. Catching a mis-sequenced restart or a reversed CT during a planned test is a day of work; catching it during an unplanned blackout at sea is a different kind of day entirely.
FAQ
- The bus recovers then collapses when a big pump starts — is the generator faulty?
- Usually not. That is a sequencing or generation-capacity problem: the single running set cannot accept the third consumer's inrush, frequency dips below the under-frequency trip, and protection opens the breaker correctly. The fix is in the restart sequence and standby-start logic, not the generator.
- An incoming generator trips on reverse power the moment it parallels — why?
- The textbook cause is a load-sensing CT on the wrong phase or reversed, usually after a maintenance period when the CT was removed and reinstalled. The PMS misreads the set's contribution and drives it into reverse power. Confirm the CT phase and polarity before suspecting the PMS or the generator.
- Can we test blackout recovery without risking an uncontrolled blackout?
- Yes — we run the drill under controlled conditions on a quiet alongside watch or at anchor, with the engine room manned and the chief informed. It proves dead-bus closing, sequential restart and kW/kVAR sharing at once and produces a class-witnessed log of the recovery times.
Book a blackout-recovery and PMS test
/services/switchboard
Source PMS cards and switchboard spares
/supply
Published by Levent Marine — Florida-based, Wyoming LLC — 24/7 worldwide