Surge protective device (SPD) types and NEC requirements field guide

What a surge protective device is

A surge protective device, the SPD, is the part that limits a voltage transient before it reaches the equipment. When a fast overvoltage rides in on the conductors, the SPD turns on in nanoseconds, diverts the surge current to the grounded and neutral conductors, and holds the voltage across the protected circuit down to a level the equipment can survive. When the transient passes, it switches back off and the circuit runs normally.

Most SPDs work off metal-oxide varistors, the MOVs. An MOV is a high resistance at normal voltage and a near-short above its clamping point, so it sits idle until the surge arrives and then dumps the energy to ground. That on-off behavior is the whole job. The device does nothing the other 99.9 percent of the time, and then earns its keep in a few microseconds.

An SPD is not a circuit breaker and not a grounding electrode. The breaker handles overcurrent that lasts; the SPD handles overvoltage that does not. The grounding electrode system gives the diverted current somewhere to go. The three work together, and the SPD is useless without the other two behind it. That tie back to the grounding electrode system is covered in the grounding and bonding guide.

Why surge protection earns its place

Transients come from three places, and the dramatic one is the least common. A direct or nearby lightning strike couples a huge fast surge onto the lines, and it gets the headlines. But the utility switching its own gear, capacitor banks coming on and off, and reclosers operating, throws transients at the service all day. And the largest share is generated inside the building, by motors, elevators, HVAC compressors, and any inductive load that switches off and collapses its field into a spike.

The damage from the small repeated hits is what people miss. A single big surge can punch a hole through a power supply and the failure is obvious. The steady drip of internal switching transients degrades electronics quietly, knocking years off the life of variable frequency drives, building controls, LED drivers, and anything with a circuit board. The equipment does not die at the strike. It dies early, months later, and the surge that started it is long gone.

That cumulative degradation is the real argument for surge protection, more than the lightning story. A facility full of drives and controls runs cleaner and longer with the transients clipped at the service and again near the gear. The protection is cheap against the cost of a control board that takes a line down. The lightning protection system guide covers the strike itself; the SPD handles the transient that gets past it.

What are the SPD Types 1, 2, 3, and 4?

SPDs are classified by Type under UL 1449, and the Type tells you where the device is allowed to go in the system, not how good it is. Type 1 connects on the line side or the load side of the service disconnect and is listed to be installed without an external overcurrent device. Type 2 connects on the load side of the service disconnect overcurrent device, typically at the panel, and it is the workhorse of the trade.

Type 3 is point of use, the receptacle, cord-connected, or plug-in unit, and it has to sit a minimum conductor distance downstream of the service, commonly given as 10 m, about 30 ft, of conductor. Type 4 is a component assembly, a recognized component built into other listed equipment, not a device you field-install on its own.

Get the Type matched to the location and the rest of the selection gets easier. Put the wrong Type in the wrong place and you have either an unlisted installation or a device that cannot do its job. The most common version of that error is a plug-in Type 3 strip treated as if it were protecting the whole building, when there is nothing at the service holding back the big transients ahead of it.

What is the difference between a Type 1 and a Type 2 SPD?

The difference is where each can connect and whether it needs its own overcurrent device. A Type 1 SPD is listed for the line side of the service disconnect, ahead of the main, where there is no overcurrent device in front of it. To be allowed there it must survive a fault on its own, so it is listed without requiring an external overcurrent protective device. A Type 1 can also be installed on the load side, which makes it the more flexible of the two.

A Type 2 SPD is load side only. It goes after the service disconnect overcurrent device, anywhere on the load side, and the manufacturer's instructions almost always call for it to land on a dedicated branch breaker. It cannot go ahead of the main. That single line, line side versus load side, is the practical fork: if you want protection ahead of the service disconnect, you need a Type 1.

For a typical panel install either one satisfies the dwelling requirement at the service, and many SPDs are dual-listed as Type 1 and Type 2. Pick based on where you are landing it. Ahead of the main, Type 1. At or after the main on a breaker, Type 2 is fine and usually cheaper.

Does the NEC require surge protection?

Yes, for the occupancies the code now names, the NEC requires it, and that is a genuine change from older editions. The 2020 NEC added 230.67, requiring a Type 1 or Type 2 SPD at the service for dwelling units, and requiring one to be added when an existing dwelling service is replaced. The SPD has to be an integral part of the service equipment or located immediately adjacent to it.

The 2023 edition kept 230.67 and expanded the surge requirement to other occupancies beyond the dwelling units already covered in 2020, reaching dormitories, hotel and motel guest rooms and suites, and patient sleeping rooms in nursing homes and limited-care facilities. The direction of the code is clear: surge protection is moving from optional to required, occupancy by occupancy, edition by edition.

Treat the exact requirement as edition-dependent and verify it against the code the jurisdiction has actually adopted, with any local amendments. Many areas are still on an older cycle where 230.67 does not exist or reads differently. The safe statement on a jobsite is that recent NEC editions require a service SPD for dwellings and a growing list of other occupancies, and you confirm the adopted edition before you quote the requirement to an inspector or an owner.

Where SPDs are required and where they are smart

Required and recommended are two different lists, and a good install reads both. Under recent NEC editions, the clearest mandate is the service SPD for dwelling units at 230.67, with the 2023 edition reaching the dormitory, hotel, and care-facility occupancies named above. Other parts of the code carry their own surge requirements for specific systems, and the section numbers move between cycles, so confirm them against the adopted edition.

Beyond what is written, surge protection is the smart call anywhere the loads are sensitive or the downtime is expensive. Emergency systems, fire pumps, and the critical operations power systems that have to ride through a storm are exactly the places a transient cannot be allowed to take the controls down. Industrial machinery built to NFPA 79 carries its own surge expectations for the control circuits.

Data centers, healthcare imaging, building automation, and anything full of variable frequency drives all belong on the recommended list whether or not a code line forces it. The rule of thumb: if losing the equipment costs more than the SPD and a couple hours of labor, protect it. That covers most of what is worth protecting.

The specs that matter: VPR, MCOV, SCCR, and In

Four ratings decide whether an SPD fits the system, and reading them in the right order keeps you out of trouble. VPR, the voltage protection rating, is the let-through voltage the equipment downstream actually sees during a standard surge; lower is tighter protection. MCOV, the maximum continuous operating voltage, is the steady voltage the device tolerates without degrading, and it has to sit above the system's normal voltage. SCCR, the short-circuit current rating, has to be at or above the available fault current at the point of installation. In and Imax describe the surge current the device can take.

Match all four to the system, not just the one on the front of the box. A low VPR is worth nothing if the MCOV is too low and the device cooks itself on normal line swings, and it is dangerous if the SCCR sits below the fault current available where you mounted it. The fault rating in particular is a code-relevant limit: you cannot install an SPD where the available fault current exceeds its rating.

For the surge ratings, the nominal discharge current, In, is the number that describes durability, the repeated 8 by 20 microsecond surge the device is rated to take over and over. Imax is a single-shot maximum it survives once. A higher In matters more than a headline Imax, because the real world is many moderate hits, not one record event. Recent NEC editions tie a minimum In to the service requirement, commonly 10 kA, so check the In against the adopted edition.

What is VPR on a surge protector?

VPR, the voltage protection rating, is the voltage that gets through the SPD to the equipment during a standard test surge. It is the let-through, the number that says how much overvoltage the protected gear still has to ride out after the device has done its work. A lower VPR clamps tighter and leaves less for the equipment to survive, so on sensitive loads, lower is what you want.

VPR is measured per mode, and the modes are not equal. A device has a VPR for line to neutral, line to ground, neutral to ground, and line to line, and a spec sheet that only quotes the best one is hiding the others. The mode that protects the equipment in your installation, usually line to neutral for connected electronics, is the one to compare across products.

VPR is not the same as the suppressed voltage ratings on older labels, and it is the value UL 1449 standardized on. Read the VPR for the relevant mode, compare it against what the downstream equipment is built to withstand, and remember that the installed let-through is worse than the rated VPR once you add the lead length. The next section is why.

MCOV: keeping the device above the system voltage

MCOV is the highest voltage the SPD can sit on continuously without degrading, and it has to be above the system's nominal voltage with margin for normal line swings. Set it too low and the device runs hot in normal service, the MOVs degrade early, and you get nuisance end-of-life or outright failure with no surge ever involved. The line normally drifts above nominal, and the MCOV has to stay clear of that drift.

Match the MCOV to the actual system. A 120/240 V split-phase service wants roughly 150 V line to neutral and 300 V line to line at the device. A 277/480 V system needs a different MCOV again. Grabbing a 120 V-rated device for a 277 V circuit is a fast way to a failed SPD and a tripped breaker, and it is a common stocking mistake when crews carry one part for everything.

The trade-off runs against VPR. Push the MCOV higher for safety margin and the let-through voltage tends to rise, so you do not want it arbitrarily high either. Set it above the system with sensible headroom for the real line variation, and no higher.

Why do SPD leads need to be short?

Short leads matter more than almost any other thing about an SPD install, and it is the part crews get wrong most often. The reason is the speed of a surge. A transient rises in microseconds, and at that speed even a short length of wire develops a real voltage across its own inductance. Every inch of conductor between the SPD and the bus adds let-through voltage on top of the device's rated VPR, and the rule of thumb in the field is on the order of tens of volts per inch.

That means the installed protection is the rated VPR plus the lead drop, and long leads can add hundreds of volts to what the equipment actually sees. A device with a clean low VPR mounted on two feet of looped wire performs like a far worse device. You can buy the best clamping number on the market and throw it away at the lugs.

Keep the leads short and straight, the shorter the better, and many manufacturers want them under a foot. No sharp bends, no service loops, no coiling up the slack. Mount the SPD as close to the breaker or the bus as the enclosure allows, take the most direct path to the grounded and neutral conductors, and where the leads run together, twist them so the surge currents cancel some of the inductance. If the install instructions give a maximum lead length, it is not a suggestion.

The breaker ahead of the SPD

Whether the SPD needs its own overcurrent device depends on the Type and the listing. A Type 2 connects on the load side and the manufacturer's instructions normally call for a dedicated breaker sized per those instructions, both to disconnect the device for service and to clear it if the MOVs fail short. Follow the instruction sheet for the breaker size; do not guess it.

A Type 1 is listed without requiring an external overcurrent device, which is exactly what lets it sit on the line side ahead of the main where there is no breaker to give it. That does not mean a breaker is forbidden on a Type 1 used on the load side, only that the device does not depend on one. The listing and the instructions tell you what is allowed.

Keep the breaker close. The point of the dedicated breaker is partly to give a short, clean connection, and a long run out to a distant breaker fights the short-lead rule from the last section. Mount the SPD next to its breaker, not across the panel from it.

Layered protection, service to equipment

One SPD at the service is the start, not the finish. Surge protection works best in stages, each layer knocking the transient down further before it reaches the gear. A Type 1 or Type 2 at the service entrance takes the big incoming hits and the worst of the utility and lightning energy. A Type 2 at each downstream subpanel catches what gets past the first stage and the transients generated inside that part of the building. A Type 3 at the sensitive equipment trims the last of the let-through close to where it matters.

The reason for the cascade is that the service device cannot give a low enough let-through at a distant receptacle on its own. By the time the surge has traveled to the far panel and the equipment, and after the lead inductance at the service device adds to its VPR, the voltage at the load can still be more than sensitive electronics want. The downstream layers exist to clamp it again, locally, on short leads.

Coordinate the layers so the upstream device takes the energy and the downstream device trims the voltage. The 10 m, 30 ft minimum distance for a Type 3 downstream of the service is part of that coordination, letting the conductor impedance share the work between the stages. On a critical facility the layered approach is the difference between protecting the panel and protecting the equipment.

The SPD is only as good as its ground

An SPD diverts surge current to the grounded and neutral conductors, which means it needs a low-impedance path to ground to do anything. If that path has resistance or inductance in it, the diverted current develops a voltage across the path, and that voltage rides right back onto the system the SPD was supposed to protect. A high-impedance ground turns a good SPD into a poor one.

This is the same low-impedance ground that the grounding and bonding guide covers in full, and it is why the two topics cannot be separated. Bond the SPD's grounding terminal to the equipment grounding system on short, direct conductors, into a grounding electrode system that is actually built and bonded the way it should be. A loose, corroded, or undersized ground connection at the SPD is as damaging as a long lead.

On the line-side and service installs in particular, the SPD's effectiveness depends on the quality of the service grounding ahead of it. Check the ground bond when you commission the device, not just the line connections. A green status light on a device tied to a bad ground is telling you the MOVs are fine, not that the protection is working.

SPD versus lightning protection system

These are different things that get confused constantly, and the distinction matters when someone asks whether they are covered. A structural lightning protection system, the air terminals, down conductors, and grounding built to NFPA 780, handles the strike itself, giving the lightning a path around the building to ground. It does almost nothing for the electronics inside.

The SPD handles the transient, the fast overvoltage that couples onto the electrical and signal conductors, whether from that strike, from a nearby strike, or from the utility and the building's own loads. It does nothing to intercept or carry a direct strike. The lightning protection guide covers the structural side in full; this guide covers the transient side.

On a building that has both, they are complementary, not redundant. The lightning protection system keeps the strike from blowing the structure apart, and the SPDs keep the transient that rides in alongside it from taking out the controls and the electronics. A facility with a lightning protection system and no surge protection has protected the building and left the equipment exposed, and that is a common and expensive gap.

Protect the data and signal lines, not just the power

A surge rides more than the power conductors. It couples onto every conductive path into the equipment, the coax, the Ethernet, the antenna lead, the control and instrumentation cabling, the dialer and alarm lines. Protect the power and leave the data lines open, and the transient walks in the back door and takes the equipment anyway, through a port that had nothing in front of it.

Signal-line SPDs exist for exactly this, sized for the voltage and the data rate of the line so they clamp the transient without killing the signal. Coax surge protectors for antenna and RF feeds, network SPDs for the data runs, and dedicated units for control loops all belong on the list when the equipment is worth protecting. The protection has to cover every path that reaches the box.

The single most common surge failure on protected equipment is a transient that came in on an unprotected data line while the power side was covered. Inventory the cables into the critical gear and ask which ones have a surge device. If any path is open, that is the one the next surge will use.

Status indicators and monitoring

An SPD is sacrificial, and the status indicator is how it tells you whether it is still alive. Most devices carry a green light or a per-mode indicator that shows the MOVs are intact and protecting. When a mode degrades to its end of life, the indicator changes, usually a green light going dark or a red light or flag coming up, and on many devices an audible alarm or a dry contact for remote monitoring.

The indicator only reports the device, not the install. A green light means the MOVs have not failed; it does not promise the ground is good or the leads are short. Read it for what it is, a sign the suppression element is intact, and verify the rest of the install separately.

On critical systems, wire the remote-monitoring contact into the building management or alarm system so the failure shows up where someone sees it. An SPD that quietly reaches end of life behind a closed panel door protects nothing and nobody knows, until a surge gets through to the equipment and the question becomes why the device was never replaced.

End of life and replacement

MOVs wear out, and that is by design. Each surge the device absorbs degrades the metal-oxide elements a little, and a single large hit can spend a device in one event. Over time, or after one big surge, the suppression element reaches the end of its life and the device stops protecting, even though it still passes power through normally. The power keeps flowing, which is exactly why the failure hides.

Replacement is part of owning surge protection, not a defect. After a known large surge or a nearby lightning event, check the indicators and replace any device that has tripped its end-of-life signal. On a maintenance walk, the SPD status is a thing to look at, the same as you would check anything else in the panel. A device showing end of life is doing nothing and should come out.

Build the replacement into the plan and the budget. Treat the SPD as a consumable that gets checked on the maintenance schedule and replaced when the indicator says so, the way you would treat any protective element that takes the hit so the equipment does not. The cheap mistake is installing it once and never looking at it again, then wondering why the gear it was supposed to protect took a surge.

Commissioning the SPD

Commissioning is where the install gets verified against the system, and it is short. Confirm the VPR, MCOV, and SCCR on the installed device actually suit the system voltage and the available fault current at that point, because the part that got delivered is not always the part that was specified. The fault rating in particular is the one that has to be confirmed, not assumed.

Check the install, not just the device. Are the leads short and straight, with no loops or sharp bends, within the manufacturer's maximum length? Is the grounding connection short, tight, and bonded into a real grounding electrode system? Is the dedicated breaker, where the device needs one, sized per the instructions and close to the device?

Then read the status. Every mode indicator should show normal, every light green or whatever the device uses for healthy, before you call it done. Record what you verified, because the commissioning record is what tells the next person the device was right when it went in, and gives them a baseline to check against when they come back to it.

Layered protection on a data center or critical facility

On a data center or any facility where downtime is the whole problem, surge protection is a layered design, not a single device, and it tracks the power chain. A service-entrance SPD takes the incoming utility and lightning energy. SPDs at the distribution and PDU level catch what gets through and the internal switching transients from the mechanical plant. Equipment-level protection trims the last of the let-through close to the racks, and the data and signal lines get their own protection on every path in.

Pair that with the grounding and bonding the facility already needs and the structural lightning protection where the building has it, and the surge devices become the layer that keeps the transient off the IT and control gear. The whole point of the staged design is that no single device has to do everything, and the equipment at the end sees a clamped, coordinated result rather than whatever the service device alone could manage.

What to document

A surge install that nobody recorded is one nobody can verify or maintain. The record is what tells the next person what is protecting the system, where it is, and whether it was right for the location, and it is what turns a green-light walk into a real check.

Capture each SPD by Type and location, the system voltage and phase it serves, its VPR for the relevant mode, its MCOV, its SCCR against the available fault current at that point, its In rating, the breaker that feeds it where it has one, the status at commissioning, and the date it went in. When a device is replaced after a surge, log the replacement and the reason, so the history of what the system has taken is on paper.

Field checklist

Common mistakes

• Long or looped SPD leads that add hundreds of volts of let-through on top of the rated VPR.
• The wrong Type for the location, such as a plug-in Type 3 treated as service-level protection with nothing at the service.
• MCOV set too low for the system, so the device degrades on normal line swings and fails with no surge involved.
• SCCR below the available fault current at the point of installation, which is not a code-legal install.
• Ignoring the end-of-life status indicator and leaving a spent, non-protecting device in service.
• Protecting the power conductors and leaving the data, coax, and control lines open for the transient to walk in.
• Bonding the SPD to a high-impedance or loose ground, so the diverted surge rides back onto the system.

Standards and references

The NEC, NFPA 70, is where the installation requirements live. Recent editions require a Type 1 or Type 2 SPD at the service for dwelling units under 230.67, including when an existing dwelling service is replaced, and the 2023 edition expanded the surge requirement across other occupancies. Article 242 is the consolidated overvoltage-protection article in recent editions, covering SPDs on premises wiring of 1000 V or less and surge arresters above that, after older editions split the same material across Articles 280 and 285.

The exact section numbers and the scope of the requirement shift between code cycles, and many jurisdictions are still on an older edition. Confirm 230.67, Article 242, and any In or location requirement against the edition the jurisdiction has actually adopted and its local amendments before you cite them on a submittal or to an inspector.

UL 1449 is the product standard that lists SPDs, defines the Types by location, and standardizes the VPR. NEMA, through the Surge Protection Institute, publishes application guidance on Type selection and ratings. Industrial machinery built to NFPA 79 carries its own surge expectations for the control circuits. Above all of it, the manufacturer's instructions govern the breaker, the lead length, and the rating selection, and where the listing or the instructions are stricter than the rule of thumb, they win.

Units and terms

Surge protection carries a stack of acronyms, and the same device gets described with different ones across a spec sheet, a submittal, and a code section.

An SPD is a surge protective device, the term that replaced the older surge arrester and TVSS labels for premises devices. Ratings read in volts and kiloamps: VPR and MCOV in volts, In and Imax in kiloamps, SCCR in amps or kiloamps of available fault current. Surge current is tested on the 8 by 20 microsecond waveform, the standard shape for the repeated discharge rating. Modes describe the path the device protects, line to neutral, line to ground, neutral to ground, and line to line.

SPD

Surge protective device, the unit that clamps voltage transients and diverts the surge to ground

VPR

Voltage protection rating, the let-through voltage during a standard surge; lower clamps tighter

MCOV

Maximum continuous operating voltage the device tolerates without degrading; set above the system voltage

SCCR

Short-circuit current rating; must be at or above the available fault current at the install point

In / Imax

Nominal discharge current (repeated, durability) and maximum single-event surge current

MOV

Metal-oxide varistor, the sacrificial element in most SPDs that degrades with each surge

Related tools

Calculators and readiness checks for this work

Compare your options

FAQ

People also ask

• Does a lightning protection system stop lightning?
• What is the difference between grounding and bonding?
• What is the difference between a Type 1 and Type 2 SPD?
• What is an electrical disconnect?
• What is cathodic protection?