Generator control panel and protections: what the settings actually do


A genset control panel is a start and stop device with a list of thresholds attached, and the settings on it decide whether the set stops on a real fault or rides through a normal disturbance such as a motor start. Every protection is a measurement, a setpoint and a delay, and the engineering is in those three numbers rather than in the display. The panel hardware only carries out what the settings ask of it.

What a genset control panel is, and what it is not

The panel on a diesel generating set is usually one integrated controller in a steel enclosure on the set, with a display, a keypad, indicator lamps and terminals for everything it has to talk to. Those terminals carry the engine sensors, the alternator voltage sensing, the current transformers, the circuit breaker or contactor control circuit, the mains sensing on an automatic mains failure set, and the remote inputs and outputs. A controller in the DSE7320 class, one of the widely used automatic mains failure units, does all of it in a single box, and the parameter list that comes with it runs to several pages.

What the panel is not is just as important, because a panel is often asked to do a job that belongs somewhere else.

  • It is not the site protection study. The panel protects the set. Whether the set’s own breaker trips before the incoming breaker, or the other way round, is decided by a discrimination study of the whole distribution, not by the controller.
  • It is not the alternator’s differential protection. A large machine has winding protection of its own, in separate relays with separate current transformers, and the panel does not replace it.
  • It is not medium voltage protection. Above about 1 kV the protection moves into dedicated relays and switchgear, and the controller may only send a start command and read a status.
  • It is not the grid connection protection. A set that runs in parallel with the mains needs loss of mains and vector shift or rate of change of frequency protection, to the network operator’s rules. That is a separate scheme, a separate approval and a separate paper trail.

This page is about low voltage sets up to about 1 kV, which is what is supplied here, and about the settings rather than the wiring. The figures are commissioning starting points; they are not a substitute for the engine and alternator datasheets or for the site protection study, and on a machine with an unusual alternator or an awkward load the datasheet wins every time.

What the controller measures, and what it does with it

A controller watches three groups of signals at once. The engine side reads speed, oil pressure, coolant temperature, coolant level, fuel level and battery voltage, and whether the alternator is charging. The alternator side reads three phase voltage, current through the current transformers and frequency, and from those it calculates kW, kVA and power factor. The mains side reads voltage and frequency, which is what tells an automatic mains failure panel that the supply has gone.

Every one of those readings can produce one of four responses, and the four are not the same thing.

The four responses a controller can give to a measurement outside its window, and what each one does to the set. The grade given to a signal is a commissioning decision, not a property of the sensor.
Response What the controller does Typical signals
Indication Shows the value and logs it, with no output Load, fuel level, running hours, battery voltage
Alarm or pre-alarm Closes an alarm output and shows a message while the set keeps running High coolant temperature below the shutdown level, low fuel, low coolant level
Electrical trip Opens the generator breaker and leaves the engine running for a cool-down Overcurrent, earth fault, over and under voltage, over and under frequency
Shutdown Stops the engine and latches, so the set needs a deliberate reset Low oil pressure, overspeed, high coolant temperature, emergency stop, fail to start

The difference between the last two matters more than it looks. A shutdown on low oil pressure is right, because the engine is damaging itself. An electrical trip on an alternator fault can be right as well, because the engine is not at fault, stopping it in one step removes the airflow and oil flow that a hot turbocharger still needs, and the set may still be carrying part of the load on the other side of the breaker. Most controllers let the grade be changed for each signal, and it should be a deliberate choice rather than a factory default that nobody read.

Every controller also keeps an event log, and after a shutdown a snapshot of the last few seconds of run data. That record is what turns a message on a screen into a diagnosis, and it is worth reading before anything is restarted. How to read those messages is set out in diesel generator fault finding.

Remote monitoring is the same measurements leaving the panel. A dry contact is a relay output that is either open or closed, and it carries a simple signal to a building management system: running, on load, common alarm, mains failure. A serial link carries the whole register map, so the building management system or a remote terminal can read every value, and on some controllers change a setpoint. The usual arrangement on the sets supplied here is dry contacts for the status signals and a Modbus link over RS485 for the values, with an ethernet or GSM module where the site has no fixed network.

The protection settings, and the figures behind them

Almost every setting on a generator controller is a threshold, a delay and a grade, and the same short list appears on every panel. The table below gives the commissioning windows used most often on a low voltage 400 V 50 Hz set. They are starting points rather than settings: the final value depends on the alternator’s own curves, the engine datasheet, the load, and the discrimination with the devices downstream.

Typical settings for a low voltage 400 V 50 Hz diesel set. The window is the range seen in practice, and the final value comes from the alternator data, the engine datasheet and the site protection study.
Protection Measured against Typical window Typical delay Usual grade
Overcurrent, long time Alternator rated current 105 to 125 % 5 to 60 s, inverse time Electrical trip
Overcurrent, short time Alternator rated current 150 to 300 % 0.2 to 1 s Electrical trip
Overcurrent, instantaneous Alternator rated current 250 to 400 % Under 0.1 s Electrical trip
Earth fault Alternator rated current 10 to 20 % 0.5 to 5 s Electrical trip, or alarm only
Over voltage Nominal voltage 105 to 110 % 1 to 3 s Electrical trip
Under voltage Nominal voltage 80 to 90 % 1 to 10 s Electrical trip
Over frequency Nominal frequency 105 to 108 % 0.5 to 5 s Electrical trip
Under frequency Nominal frequency 90 to 95 % 0.5 to 10 s Electrical trip, or first stage of load shedding
Overspeed Rated speed 108 to 115 % Under 1 s Shutdown
Low oil pressure Engine datasheet 1.0 to 2.0 bar 0.5 to 3 s, bypassed during the start Shutdown
High coolant temperature Engine datasheet 95 to 105 °C 1 to 5 s Shutdown
Low coolant level Level switch Switch setting 1 to 10 s Alarm, or shutdown
Low battery voltage Battery nominal 11 to 12 V on a 12 V system 10 to 30 s Alarm, or shutdown while cranking
Fail to start Crank attempts 3 attempts Crank and rest as set Alarm and lockout

Two of those rows do most of the work. The overcurrent window is not a number an engineer likes but a number the alternator forces, and it is treated on its own below. The temperature and pressure rows are the engine manufacturer’s figures, and the engine datasheet is the only correct source for them; the windows here are what those datasheets usually contain, not a reason to skip reading one.

The voltage and frequency windows look generous until the load is considered. A 400 V set is allowed to sit between about 360 V and 440 V before the panel calls it a fault, which is wider than a data centre or a hospital will accept at its own terminals. That is the point: the panel protects the machine, and the site’s own equipment protects itself. Where the load is sensitive the thresholds are tightened and the transfer scheme is designed around them, which is the subject of data centre backup power and how to size a generator for a hospital. The volt and hertz relationships behind those windows are set out in generator voltage and phases, and the power a set can actually hold for hours is a step below its nameplate, as prime and standby power explains.

Why almost every protection has a delay

A protection without a delay is a protection that trips on everything. The delay exists to let a normal event pass and to let a real fault last long enough to be certain, and its size is the difference between a set that carries its load and a set that is always stopping.

The events a delay has to ride through are known, and they are all short. A motor starting draws six to eight times its running current for a second or two, and on a loaded board the voltage dip reaches 15 % or more, which is inside the undervoltage window of the set itself. A load step makes the frequency dip and the voltage dip while the governor and the AVR catch up. A transfer from the mains to the set has a dead period and a re-energising transient. Each of those lands inside a protection window for a moment, and each of them is normal.

What each delay is really waiting for. A delay shorter than the event it is meant to ride through turns a normal operation into a shutdown.
Protection The normal event it must ride through What goes wrong if the delay is too short
Overcurrent, long time A motor starting, and the seconds of overload that follow a large load step The set trips on every start of the largest motor
Under voltage The dip caused by a motor starting on the same board The breaker opens on a start that would otherwise have completed
Under frequency The frequency dip while the governor picks up a load step The set trips on load acceptance instead of shedding load first
Over frequency Very little on a healthy set, so the delay is short An unnecessary shutdown on a small transient
Earth fault Charging current from long cable runs and harmonic current in the neutral Nuisance tripping on a healthy installation with a long feed
Low oil pressure The first seconds of running, before the pump has built pressure A healthy engine is shut down on every start, which is why the start bypass exists
Overspeed and emergency stop Nothing; these are the last line of defence An engine destroyed to save a fraction of a second
Mains confirmation timer A mains dip or a reclose that is not a real failure The set starts and transfers for a disturbance the mains itself rode through

The last row belongs to the transfer scheme as much as to the controller, and the sequence of timers that runs a mains failure is set out in ATS and AMF compared. The same logic runs through every setting on this page: a delay is not slack in the schedule, it is a definition of what counts as an event.

The link between the delays and the starting current of a motor is worth following one step further, because it comes up at almost every commissioning. A set that has been sized on the running load alone will see a start it cannot ride through, whatever the delay is set to, and the two ways out are a different starting method or a larger set. That comparison is worked through in generator starting current and motor starting.

Overcurrent is three protections, not one

An overcurrent setting on a controller is really three elements, each with its own threshold and its own delay, and together they cover the range from a mild overload to a bolted short circuit. The names vary between manufacturers; the shape of the current and time curve does not.

The three overcurrent elements plus earth fault, with what each one is there to clear. On an alternator the instantaneous element cannot be set as high as it would be on a transformer.
Element Threshold, against rated current Delay What it clears
Long time, inverse 105 to 125 % Seconds to a minute, shorter as the current rises Cable and alternator overload, and a slowly rising load
Short time, definite 150 to 300 % 0.2 to 1 s A fault on a feeder that the downstream device has not cleared
Instantaneous 250 to 400 % Under 0.1 s A short circuit close to the alternator terminals
Earth fault 10 to 20 % 0.5 to 5 s A fault to earth, which may be a small current no phase element sees

The three elements are set against the same measured current, so they are only useful if the measurement itself is right. That depends on the current transformers, their ratio and their polarity, and a reversed or wrongly rated CT makes a correct panel lie about both the load and the fault. It is the first thing to check before any setting is changed.

What the alternator can supply into a fault

The instantaneous element is the one that catches people out on a generator, and the reason is the alternator itself. Its fault current is limited by its own reactance, so on most machines the current into a short circuit is only about three times the rated current, sustained for around ten seconds before the excitation collapses; the same rating of transformer delivers fifteen to twenty times. That comparison, and the way it sets the breaker rating, is in generator cable sizing and voltage drop.

Put numbers to it. A 500 kVA set at 400 V has a rated current of about 722 A, and its fault current is roughly 2 200 A. An instantaneous element set at 400 % would need 2 888 A to operate, and the alternator will never deliver that, so the setting would sit there for the life of the set and never trip. On an alternator the instantaneous element is therefore set low, often between two and three times rated, or it is replaced by a definite time element with a short delay. The element that is the first line of defence on a transformer is, on a generator, a backup that may never see its own threshold.

The two sets used here as the worked example are the 500 kVA diesel generator and the 800 kVA diesel generator, and the same three times rated figure applies to both unless the alternator datasheet says otherwise. That is the number to write against every current setting on the commissioning sheet.

Grading, so that only one device trips

Protection is not a single device, it is a chain, and the chain is only as good as the grading between its links. The rule is simple: the device closest to the fault operates first, and every device upstream of it waits. On a generator supply the chain is short, it has an alternator at one end that cannot push much current, and it has a changeover device in the middle, and all of that has to be worked out at the drawing stage rather than in the field.

A typical low voltage chain from the alternator to the load, and what each device is set to do. The generator controller is a backup to the switchgear, not a replacement for it.
Device Where it sits What it must clear How it is set
Final circuit device At the load, a moulded case breaker or a motor starter A fault in one final circuit Tightest and fastest in the chain
Distribution board device Feeding a group of final circuits A fault the final device did not clear Slower than the device downstream of it
Incoming or main breaker At the changeover or the main board A busbar or cable fault on the site side Slower again, and rated for the site fault level
Generator circuit breaker On the alternator, inside or beside the panel A fault between the alternator and the main board Matched to the alternator short time withstand
Controller overcurrent In the generator controller, sensing the same current A fault the generator breaker has not cleared Backup, set just above the breaker curve

Two of those rows are generator specific. The generator circuit breaker has to be rated for a fault current only a few times the running current, which makes it look oversized on paper and correctly sized against the alternator data, and the controller’s own overcurrent is a backup to it rather than the primary protection. Where the site fault level is high and the set is small, the two curves can cross, and the honest answer is that a discrimination study is needed rather than another attempt at the settings. The site’s earthing and neutral arrangement decides what an earth fault current actually is in the first place, and that is a design decision rather than a setpoint.

Settings that have to agree with the ATS and with paralleling

A controller does not work alone. On an automatic mains failure set it shares its job with the changeover device, and on a parallel installation it shares it with the other sets and the paralleling panel. A setting that is correct in isolation can be wrong in either arrangement, and the ones that cause trouble are always the same handful.

Settings that cross the boundary between the controller, the changeover device and the paralleling panel. Each one belongs to more than one piece of equipment.
Setting What it must agree with on an ATS What it must agree with when sets parallel
Engine start delay The mains confirmation timer, so a dip does not start the set The number of sets asked to start, and the order they come on
Transfer delay The time the alternator needs to reach voltage and frequency before the load lands The synchronising window before a breaker is allowed to close
Return or mains stability timer Long enough that the mains has really returned, not just flickered The transfer back to the mains, and the order the sets come off
Cool-down timer Runs the set off load before it stops, which the engine needs Runs after the breakers open, before the set is taken off the bus
Under voltage Set slower than the transfer transient, or the set trips as it takes the load Set against the bus, so one set’s dip does not trip the others
Overcurrent and earth fault Discriminated against the site incoming device Discriminated against every other set’s contribution to the same fault
Reverse power Not used on a single set A few per cent of rated kW with a delay, so a set that has stopped is taken off the bus
Load shedding Not used on a single set Under frequency or under voltage opens the non-essential loads before the set trips

The first four rows are timers, and the sequence they run in is the transfer sequence described in ATS and AMF compared. The last three rows exist only because there is more than one source of power, and the protections a parallel bus adds are set out in generator paralleling. A parallel plant also needs a decision about the neutral and about the fault current, because several sets contribute to the same fault at once, so the settings are written against the total and not against one machine.

There is also a boundary the settings cannot cross on their own. The controller measures at the alternator and cannot see a fault beyond the changeover device; everything downstream of that point is the site protection, and the two have to be graded by whoever is responsible for the whole distribution.

Mistakes we see in setpoints

  • Commissioning on the factory defaults. A default setpoint is a safe average for a generic machine, not a value for this alternator, this engine and this load. The numbers on this page are a starting point; the commissioning sheet is what closes the job.
  • Setting instantaneous overcurrent above the alternator’s fault current. On a machine that delivers about three times rated into a short circuit, an element set at four times rated is a decoration. It is the most common single error on a generator panel.
  • A delay shorter than the normal event. Undervoltage set faster than the dip from the site’s own largest motor turns every start into a shutdown, and the fault is then blamed on the set.
  • Trusting a reading without checking the current transformers. Every current based protection and every kW reading depends on the CT ratio and polarity, and a reversed or wrongly rated CT makes a correct panel lie.
  • Never testing the shutdowns. An overspeed or low oil pressure trip that has never been proved is an assumption rather than a protection, and it has to be proved with the engine hot and running. What those tests look like across a year is set out in the maintenance schedule.
  • Expecting the panel to protect the site. The controller protects the set. Whether the rest of the distribution is graded is a study of the whole installation, and no setpoint on the controller can stand in for it.
  • Widening a window instead of finding the cause. If a set trips on a start it should ride through, the fix is the starting method or the load, not a relaxed protection setting that will now miss a real fault. The symptoms and the checks are separated in diesel generator fault finding.

Questions we are asked

What is a genset control panel?

It is the integrated controller on a diesel generating set: a start and stop device with a display, a keypad and a list of protections, wired to the engine sensors, the alternator voltage sensing, the current transformers and the breaker or contactor. On an automatic mains failure set it also senses the mains and drives the changeover. A controller in the DSE7320 class, one of the common units, does all of that in one box, and the settings on it are what decide how the set behaves.

What does a DSE7320 controller actually do?

It is an automatic mains failure controller, so it starts the set on a mains failure, transfers the load, and returns it when the mains is stable, and it protects the engine and the alternator while it runs. It measures speed, oil pressure, coolant temperature and battery voltage on the engine side, and three phase voltage, current, frequency, kW and kVA on the alternator side, and it offers configurable timers, alarm outputs, a dry contact interface and a serial link for a building management system. The exact parameter list and the limit of each setting belong to the manufacturer’s manual for the unit, and the values that matter for a particular set are the ones on its commissioning sheet.

Why do protection settings have delays?

Because almost every normal event looks like a fault for a moment. A motor starting draws six to eight times its running current and dips the voltage by 15 % or more, a load step dips the frequency while the governor catches up, and a mains transfer has a transient of its own. A protection without a delay trips on all of them. The delay is what separates a normal disturbance from a real fault, and a delay shorter than the event it is meant to ride through is the usual reason a set trips when nothing is wrong.

What overcurrent settings should a generator have?

Three elements, not one: a long time inverse element at 105 to 125 % of rated current with a delay of seconds to a minute, a short time element at 150 to 300 % with a delay of 0.2 to 1 s, and an instantaneous element at under 0.1 s. On an alternator the instantaneous element has to be set low, because the fault current is only about three times rated and lasts around ten seconds, so an element set at four times rated would never operate. The final values come from the alternator curves and the discrimination study.

Why is a generator short circuit current so low?

Because the alternator’s own reactance limits it. The fault current is a property of the machine, and on most alternators it is about three times the rated current, falling away over roughly ten seconds as the excitation collapses. A transformer of the same rating delivers fifteen to twenty times, because its impedance is far lower. The consequence for protection is direct: the settings are written for a fault current of three times rated, not fifteen, and the instantaneous element is set inside that figure rather than above it.

Should an earth fault trip the set, or just alarm?

It depends on the installation. On a site with a long cable feed, the charging current alone can be a few per cent of rated current, and a tight earth fault setting with a short delay will trip on it with nothing wrong. Where the installation is small and the earth path is close, an earth fault trip with a delay of 0.5 to 5 s is the right answer. Where the feed is long, the first stage is an alarm and the trip setting is raised until it sits above the standing current, and that decision belongs in the protection study.

Do I need different settings if the set runs in parallel?

Yes. Everything a single set needs is still needed on each set, and the parallel bus adds protection that exists only because the machines can drive each other: reverse power, circulating current or reactive imbalance, busbar supervision, and short circuit withstand for the total fault current of all the sets at once. The timers also become a shared decision, because the transfer, return and cool-down functions are coordinated between the sets and the paralleling panel rather than by one controller alone.

What do you need from me to check a set of protection settings?

The engine and alternator datasheets, or at least the model and the alternator transient data; the set rating and the supply voltage and frequency; the load schedule with the largest motor and its starting method; the single line diagram with the device ratings and settings at each level; and the neutral and earthing arrangement. With those we can put the settings on paper with delays that suit this load and check them against the grading through the rest of the distribution.

Send the single line diagram, not only the set rating

Tell us the set and the alternator, the load schedule with the largest motor and how it starts, and the device ratings at each level of the distribution. We will come back with a commissioning sheet, the delay that suits each protection, and the grading that keeps one device tripping instead of several.

Request a quote WhatsApp the sales desk

Written and maintained by the KVA Grid sales engineering desk. The settings on this page are commissioning starting points for low voltage installations, and the final values come from the engine and alternator datasheets, the controller manufacturer’s manual and the site protection study. The fault current figure of about three times rated current, sustained for around ten seconds, is a planning value for an alternator used on the same basis as our cable sizing page, and the transformer comparison of fifteen to twenty times is the usual figure for a machine of the same rating. Protection settings and device grades affect the safety of an installation and are set by a competent person against the actual data. The position of a single set relative to a parallel installation is in generator paralleling, and the transfer sequence in ATS and AMF compared. This page describes general practice and is not a quotation or advice on any specific project. Reviewed 8 October 2026.

Leave a Comment

Your email address will not be published. Required fields are marked *

Scroll to Top
WhatsApp