Lightning and Surge Protection Coordination for Generator, ATS and Control Panels


A standby generator is often described as the last line of defence for a building. In a lightning-rich climate it is also a long conductor waiting to be hit. The surge that damages a generator installation rarely stops at the service entrance: it divides between the earthing system, the transfer switch, the control harness and the alternator winding according to impedance, and the smallest share can still destroy a controller. Coordination is the discipline of deciding where each share is diverted, and of proving that those paths exist in every operating mode.

A lightning surge is a current event before it is a voltage event

Lightning protection is usually discussed in terms of voltage, because voltage is what breaks insulation. The physical event is a current source. A cloud-to-ground flash delivers a short, high-amplitude pulse, and a nearby flash induces a smaller pulse into every conductor that runs through its magnetic field. Both are currents looking for a return path, and they divide among all the paths that are bonded together at the moment of arrival.

The waveform matters for selection. A direct strike is represented by a long-duration impulse, while an induced surge is represented by a shorter, steeper waveform. Devices are rated for one or both, and those ratings come from the manufacturer’s data. What the designer has to decide is where each share is diverted and which equipment must withstand what remains.

A standby plant adds paths that an ordinary building does not have. The generator frame, the fuel tank, the transfer switch enclosure and the control panel are all metal, all bonded, and often spread across a yard. A surge that enters at the utility service can leave through the generator earthing conductor, and on the way it passes the transfer switch control harness. That is why a generator installation needs a coordination plan rather than a single device at the main board.

Where a surge can enter a generator installation, and what is expected to deal with it first. Device ratings and test classes are set by the manufacturer’s data.
Source Coupling path Where it appears First response
Direct strike to the structure or its air termination Conductive current into the earthing system, then into bonded metalwork Structural steel, cable trays, generator frame Bonded down conductors and a Type 1 device at the entry boundary
Nearby strike Magnetic and electric field coupling into loops and cables Overhead utility line, control cable, sensor loop Type 2 device at the distribution board; shielded and earthed cable
Utility switching Conducted transient on the supply Main board, transfer switch utility terminals Type 2 device coordinated with the upstream device
Transfer switch operation Arc at the contacts; inductive kick from coils and small motors Switch power contacts, control circuits Contact gap, coil suppression, Type 3 device at the control panel
Static during fuel transfer Charge accumulation and discharge Tank, fill point, pipework Bonding and earthing, not a surge device
Capacitor or motor switching on site Oscillatory transient on the local busbar Busbar, panel, variable-speed drives Type 2 device near the load, as specified by the manufacturer

Give each boundary a device, not each cable

The usual way to organise this is the lightning protection zone concept. Zone 0 is exposed to the direct flash. Zone 1 is inside the structure or behind a bonded screen, where the current is partly diverted. Zone 2 is inside an enclosure or a cabinet, where the field is further reduced. Every time a cable crosses a boundary, the surge travelling on it has to be reduced at that boundary, or it arrives at the next zone with most of its energy intact.

The device type follows the boundary, not the room. A Type 1 device at the service entrance deals with the partial lightning current that the structure’s bonding lets through. A Type 2 device at a distribution board deals with the residual and with switching transients. A Type 3 device at the terminals of the controller deals with what is left. The type is defined by its test class, which the manufacturer states. The designer’s job is to make sure the energy rating decreases from upstream to downstream and that the voltage protection level at each stage is below what the downstream equipment can withstand.

Coordination between stages relies on impedance. The cable between two devices provides some decoupling, and the manufacturer’s instructions state the minimum length or the decoupling element to use when the distance is short. Two devices of different sizes fed from the same bar are not coordinated; they share the surge in a ratio nobody designed.

Devices by zone boundary. The same cabinet can contain more than one boundary, which is why the layout matters as much as the bill of materials.
Boundary Device role What it protects Coordination check
Service entrance, zone 0 to 1 Type 1, partial lightning current Main switchgear and the utility side of the transfer switch Bonding to the earthing system; backup protection per manufacturer
Distribution board, zone 1 to 2 Type 2, residual surge Submains, transfer switch load side, panel feeders Upstream and downstream energy ratio; cable length or decoupling element
Control panel incoming, zone 2 to 3 Type 3, terminal-level protection Controller, battery charger, relay modules Voltage protection level against equipment withstand
Sensor and communication loop Signal device with matched impedance 4-20 mA loop, bus transceiver, remote annunciator Bandwidth, capacitance and shield bonding arrangement
Generator output terminals Type 2 or Type 3 as specified Alternator winding, excitation circuit, output cables Coordination with the transfer switch and the utility-side device

The transfer switch is a surge path as well as a switching device

The automatic transfer switch is usually placed where the utility supply enters the building, which makes it one of the first pieces of equipment a surge meets. It also decides whether the generator is connected at all. A surge that damages the mechanism or its controller can leave a healthy generator unable to take load.

The transfer contacts are a gap in the circuit. During transfer that gap opens and closes, and a surge arriving at that moment can arc across it. The contact gap, the transfer speed and the suppression on the operating coils are set by the switch manufacturer.

The neutral arrangement deserves separate attention. A three-pole switch leaves the neutral solidly connected in both modes; a four-pole switch breaks it. When the neutral is switched, the surge return path changes at the instant of transfer, and equipment referenced to the neutral sees a different earth potential. The single-line diagram should show the neutral and earth paths in both modes, and the protection should be arranged so that neither mode leaves a device without a return path.

A bypass-isolation switch adds another mode: the load can be fed while the switch itself is isolated for maintenance. In that mode the surge path may bypass a device mounted on the switch. The protection layout has to be drawn for every mode, not just the normal one.

How the transfer switch arrangement changes the surge path. The neutral and earth paths should be drawn for each mode before the devices are positioned.
Arrangement Surge path in normal mode What changes during transfer or bypass Design response
Three-pole switch, solid neutral Neutral and earth remain common to both sources Neutral is never broken; the earth reference stays stable Check that the generator neutral is not bonded twice
Four-pole switch, switched neutral Neutral path depends on the source in service Neutral is broken and re-made; the return path changes Show both neutral paths; check devices in each mode
Bypass-isolation switch Normal path through the switch and its device Load can be fed while the switch is isolated Provide protection on the bypass path, not only on the switch
Closed-transition switch Both sources briefly connected Circulating current and a shared neutral path Confirm the protection scheme remains coordinated in parallel
Open-transition switch Load is interrupted between sources Control and sensing circuits lose supply briefly Check ride-through of the controller and the UPS bridge

Control panels and the DC system are the fragile end

The generator controller, the engine control module, the battery charger and the instrumentation are low-voltage electronics with a much lower withstand than the alternator. They are also connected to long cables: a sensor on the engine, a remote annunciator in the guardhouse, a modem in the office. Each of those cables is both an antenna and a conductor.

The practical rules are unglamorous. Keep control and power cables in separate trays or separated by the distance the manufacturer requires. Bond the shield at the panel end, and at the field end only where the instruction says so, because bonding both ends in a lightning environment can circulate current through the shield. Fit a signal device at the panel boundary for every loop that leaves the enclosure. Protect the charger’s AC input, because the charger is often fed from a small local board that has no surge device of its own.

The excitation circuit is part of the same problem. The automatic voltage regulator and the rotating rectifier are exposed to transients on the output cables, and the suppression around the field circuit is specified by the alternator maker. A generator that trips on over-excitation after a storm has not necessarily lost its regulator; it may have been hit through a path nobody had drawn.

Interfaces at the control end of the installation. Each one is a boundary that needs a decision, not a default.
Interface Typical exposure Protection measure Record
Battery charger AC input Local board with no upstream device Type 3 device at the charger, or a Type 2 at the local board Device rating and upstream coordination
24 V DC distribution Long runs to solenoids and relays Suppression at coils; separation from power cables Coil suppression fitted and checked
Engine sensor harness Long cable close to the alternator and starter Shielded cable, single-end shield bond, signal device Shield bonding method and device type
Remote annunciator or modem Cable leaving the building or crossing a yard Signal device at each end as specified; fibre where practical Loop schedule and device schedule
AVR and field circuit Transients on the output and excitation cables Suppression per the alternator maker’s data Maker’s instruction and fitted parts
Battery monitoring Small signals on long leads Signal device or short, protected route Route drawing and device fitted

Bonding and earthing decide whether the device can work

A surge device works by diverting current, and diverted current has to reach the earthing system through a path of low impedance. If the connection is long, looped or shared with a noisy load, the voltage across the device during the surge rises and the protected equipment sees more, not less. The classic failure is a device mounted neatly at the panel with a two-metre connection to a remote earth bar. The same device mounted with a short, wide connection to a local bonding bar behaves completely differently.

Every zone boundary should have a bonding bar, and every metal service that crosses the boundary should be bonded to it: the generator frame, the enclosure, the fuel tank, the transfer switch, the cable tray and the panel chassis. Bonding is not the same as earthing. Bonding makes the metalwork sit at one potential during the surge; earthing gives the current a route into the soil. Both are needed, and the electrode system has to be verified rather than assumed.

For a generator yard, the electrode arrangement and the earth-fault loop are normally checked at commissioning. The test method and the acceptance value come from the project specification; the measurement is a text record, not an impression. A scheme with a good device and a poor earth does not prevent damage. It moves the damage to a different place.

Bonding checks that make the difference between a protected installation and a drawing that looks protected.
Item What to verify How Record
Device earth conductor Short, wide and dedicated to the local bar Measure length and conductor size against the manufacturer’s instruction Length and size on the as-built
Panel to bonding bar Low-resistance connection, no shared noisy return Continuity measurement at the installed condition Measured value and instrument
Generator frame Bonded to the installation earthing system Visual and continuity check at the frame lug Lug location and continuity result
Fuel tank and pipework Bonded across joints and flexible sections Continuity across each joint Joint schedule with results
Transfer switch enclosure Bonded at the cable entry and at the panel Continuity check at both ends Both readings
Cable tray and armouring Continuous and bonded at boundaries Continuity along the route Route drawing and readings
Electrode and earth-fault loop Within the value set by the project specification Commissioning test with a recorded method Test method, value and date

Let the manufacturer’s data and the specification set the numbers

Selecting a device is a set of comparisons, not a preference. The maximum continuous operating voltage has to be above the highest sustained voltage the system can present, including regulation and tolerance. The voltage protection level has to be below the withstand of the equipment being protected, with a margin set by the equipment maker and the project specification. The nominal discharge current and the impulse current have to match the exposure and the position in the zone scheme. The short-circuit rating has to be compatible with the backup protection.

Ambient conditions change the answer. The reference ambient for equipment selection on this site is 40 °C, and a device inside a metal enclosure in direct sun runs hotter than the air temperature. The manufacturer’s derating data for temperature, altitude and enclosure applies. A device selected only for its headline current may be outside its thermal envelope for much of the year.

Over-specification is also a trap. A device with a very high discharge rating but a high protection level may protect itself and not the controller. The right question is not which device is the largest, but which device gives the lowest protection level that can survive the expected current. That answer comes from the manufacturer’s data and the project specification, and it should be written down with the calculation so the next person can check it.

Inputs that decide the device selection. None of these values should be copied from a catalogue headline without checking the system it sits in.
Input Why it matters Source
Maximum continuous operating voltage Must stay above the highest sustained system voltage, including tolerance System design and manufacturer’s data
Voltage protection level Must sit below the withstand of the protected equipment, with margin Equipment maker and project specification
Nominal and impulse discharge current Matches the exposure and the position in the zone scheme Manufacturer’s data and site exposure assessment
Short-circuit rating and backup protection Keeps the device safe if it fails at end of life Manufacturer’s data and upstream device rating
Temporary overvoltage tolerance Prevents a device from operating during a healthy voltage rise System study and manufacturer’s data
Ambient and enclosure temperature Sets the real thermal margin at 40 °C and above Manufacturer’s derating data
Cable length between stages Provides decoupling, or requires a decoupling element Manufacturer’s installation instruction
End-of-life indication Allows the annual check to see that the device has operated Manufacturer’s data and maintenance plan

Commissioning, annual checks and the load test

Commissioning is where the scheme is proved or quietly lost. The checks are simple but they have to be recorded: the device type and rating against the schedule, the conductor size and length, the torque on the terminals, the bonding continuity between the panel, the transfer switch and the earthing bar, the status indication on each module, and labels that show what is upstream and downstream.

After commissioning, the scheme needs an annual visit. The status window on a modern device shows whether it has absorbed an event, and some devices keep a counter or use replaceable cartridges. The reading should be recorded, not glanced at. The same visit should check that nothing has been added to the panel without a protection review, because a new modem or a new sensor loop is a new path.

That visit can be combined with the annual load test at 50 % to 100 % of rating. The load test is the one that proves the machine. A set that is only ever exercised at 10 % to 20 % of rating is not being proved, and the low-load band has its own risks. Combining the two checks does not reduce either one, and it avoids a second mobilisation. For a 500 kVA set, two hours at full load consumes about 200 litres at the published 101 litres an hour; fuel is a real cost, which is another reason to plan the visit once and do both jobs properly.

Larger plants follow the same logic. A 1000 kVA set running at about 200 litres an hour uses about 4800 litres in 24 hours, so a long outage is an expensive event that the plant is expected to survive. In a data centre the UPS covers the seconds between mains loss and generator load, and a PUE of 1.4 shows how much of the load is critical. The transfer switch and its control circuits are what turn those seconds into a stable supply. A surge that damages the controller does not appear as a fuel problem; it appears as the outage the generator was installed to prevent.

Records should include the device schedule, the commissioning results, the annual status readings, any replaced cartridges, and the as-built single-line diagram showing the surge path in each mode. If the generator is later paralleled or a panel is added, that record is the only way to know whether coordination still holds.

Does a Type 2 device at the main board protect the generator controller?

Not on its own. The controller is usually fed from a local board or a charger circuit and is connected to long sensor cables. The residual voltage at the main board plus the cable length means a second stage is needed at the panel boundary. The manufacturer’s coordination data states the required upstream and downstream combination.

Can surge protection be installed only on the generator output?

That protects the alternator and the load side, but it leaves the transfer switch, the charger and the controller exposed from the utility side. The scheme should cover each zone boundary, including the utility entry, the transfer switch, the control panel and the loops that leave the enclosure.

What changes when the transfer switch breaks the neutral?

The return path changes at transfer. Equipment referenced to neutral sees a different potential, and a device that relied on the old path may have no path at all in the new mode. Draw the neutral and earth paths for each mode and check the device arrangement against those drawings.

How can I tell whether a surge device has operated?

Most devices have a status window, a flag or a counter, and some use replaceable cartridges. The indication is set by the manufacturer. Record it at each annual visit and replace the module when the manufacturer’s instruction says its duty has been exceeded.

Is a separate earth electrode needed for the generator?

Not necessarily, and a separate electrode that is not bonded to the main earthing system can create a dangerous potential difference. The earthing arrangement is set by the project specification and the system design. What matters for surge protection is that all earths are bonded to one reference at the installation.

Does surge protection remove the need for a lightning protection system?

No. A lightning protection system deals with a direct strike to the structure; surge devices deal with the current let through and with induced surges. They are complementary, and coordinating them is part of the design.

Send the single-line diagram and the site exposure

Tell us the generator rating, the transfer switch arrangement, the control architecture and how the site is earthed. The reply sets out the zone boundaries, the device classes to coordinate and the checks to record at commissioning, in writing.

Request a quote WhatsApp the sales desk

Written and maintained by the KVA Grid sales engineering desk. The figures on this page are the ones published elsewhere on this site: neutral bonding from generator earthing and the neutral; electrode and loop verification from generator earth electrodes; the annual load test at 50 % to 100 % and the 10 % to 20 % low-load band from the maintenance schedule; the fuel figures of about 0.24 to 0.29 litres per delivered kWh, and 101, 80 and 60 litres an hour for a 500 kVA set at full, 75 % and 50 % load, from fuel consumption; the 1000 kVA example at about 200 litres an hour and about 4800 litres for 24 hours, the UPS bridge between mains loss and generator load, and the PUE 1.4 basis from data center backup power; and the 40 °C reference from hot and humid climates. Device types, discharge ratings, protection levels, decoupling distances, derating and replacement criteria are set by the manufacturer’s data and the project specification. Reviewed 5 November 2026.

Leave a Comment

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

Scroll to Top
WhatsApp