What a Surge Protective Device Actually Protects

Why this matters

A site with a properly listed device at the service entrance loses a control board in a storm, and everybody concludes the protection failed. Usually it did not. A surge protective device limits the voltage difference between the specific conductors it is connected across, measured at its own terminals and made worse by the length of its own leads. It is protection for a path, not for a piece of equipment. Once you can list every conductive path into a machine, you can say in advance which ones are covered and which one the next surge will use, and that list is a deliverable a customer can act on.

Before anything is landed or opened

Installing or inspecting a device at a panel is work inside energized equipment until it is not. De-energize first: 29 CFR 1910.333(a)(1) allows energized work only where the employer can demonstrate that de-energizing introduces additional or increased hazards or is infeasible due to equipment design or operational limitations, and "the customer would rather not lose the freezer" is neither. Lock and tag under 29 CFR 1910.333(b)(2), which also requires stored electric energy that might endanger personnel to be released. The general lockout standard at 29 CFR 1910.147 does not govern here; it excludes exposure to electrical hazards from work on conductors and equipment in electric utilization installations at (a)(1)(ii)(C), and 29 CFR 1926.417 is the construction counterpart. Prove dead with an instrument checked live-dead-live per NFPA 70E-2021, 120.5, and select PPE from the risk assessments at 130.5 and 130.7, both in the edition your employer's electrical safety program has adopted. A device landed on the line side of the service disconnect cannot be de-energized by anything in the building, which is the single most common reason this work gets handed to the utility or to a contractor with a different scope.

What the device does, in one paragraph

A surge protective device is a voltage-dependent shunt. Above a threshold it conducts, diverting surge current away from the conductors it is bridging and clamping the voltage across itself to a level the connected equipment can survive. It does not absorb the surge in any meaningful sense and it does not stop a surge from entering the building. It equalizes two points, and everything it protects sits between those two points.

The numbers on the label, and what they are measured under

• Voltage protection rating (VPR) is the let-through voltage measured under the listing standard's combination wave at 3 kA, per UL 1449 in the edition the device is listed to. It is a comparison number between devices, not a promise at every surge current: at higher surge current the device clamps higher, and the actual let-through at a given current is on the manufacturer's curve.
• Maximum continuous operating voltage (MCOV) is the steady-state voltage the device tolerates without conducting. It has to sit above the highest normal voltage on that system, or the device runs warm and degrades on ordinary service.
• Nominal discharge current (In) is the surge current the device is rated to survive repeatedly at the stated waveform.
• Type describes where it may be connected: Type 1 on the line side of the service disconnect, Type 2 on the load side, Type 3 at the point of use with a minimum length of conductor between it and the panel. The type is a listing property and the connection rules for it live in NEC Article 242, in the edition your authority having jurisdiction has adopted.

The reference waveform behind most of these is the 8/20 microsecond combination wave, which rises to its peak in about 8 microseconds. That rise time is what makes the leads matter.

Why the leads are part of the device

A conductor carrying a fast-rising current develops a voltage across itself equal to its inductance times the rate of change of current, and that voltage adds to whatever the device is clamping at. For a round conductor in the usual single-conductor geometry, away from its return path, inductance runs on the order of 0.4 microhenry per metre and is nearly independent of wire gauge. That figure is a working approximation and not a property of the wire alone: inductance is a loop property, so what actually sets it is the area enclosed between the outgoing and returning conductors, which is why running the two leads together and keeping them straight matters more than upsizing them.

Take a 10 kA surge with the 8 microsecond rise above. The rate of change is 10,000 amperes divided by 8 microseconds, which is 1.25 x 10^{9 amperes per second. Half a metre of total lead at 0.4 microhenry per metre is 0.2 microhenry, so the leads add 0.2 x 10}-6 times 1.25 x 10^9, which is 250 volts on top of the clamping voltage. Run the same device on 1.5 metres of lead and the inductance is 0.6 microhenry, so the adder is 750 volts. Three times the lead, three times the added volts, same device, same surge.

That number sits on top of a clamping voltage which is itself above the labelled VPR at 10 kA, because VPR is a 3 kA figure. Two devices with identical VPRs, one landed with short straight leads and one landed across the panel, are not equivalent installations, and the manufacturer's installation instructions covering lead length are enforceable through NEC 110.3(B) in your adopted edition.

The path map, filled in for one machine

This is the artifact. One row per conductive path that enters the equipment, and the surge uses whichever row is empty.

Site: a packaged rooftop unit with a remote controller located in a separate building on the same property.

Two of the four rows are covered and two are empty, and the empty two are the ones that leave the building.

Now read what that means physically. Both power rows terminate at devices referenced to the same building's grounding electrode system. The control pair and the network drop terminate at equipment referenced to a different building's electrode system. When a nearby strike or a fault raises one building's reference relative to the other, the difference appears across whatever bridges them, and what bridges them is the input circuitry of the controller and the unit's control board. The power-side devices see nothing to clamp, because the difference is not between line and ground at either panel; it is between two grounds. Every device in the table did its job and the board still died.

      power path signal path
    building panel controller in a
    with surge device different building
          |                                |
          +-------- rooftop unit ----------+
                        |
              both paths land inside
              the same board, which
              now bridges two different
              ground references

That is the whole failure, and it is invisible on a walk-around because there is nothing to see at either panel.

What the empty rows need

The fix is a device on each unprotected path, listed for that path's signal type and rated for its operating voltage and, for a data path, its bandwidth. A power surge device is not a signal surge device and putting one on a low-voltage control pair either does nothing or breaks the signal. The other half of the fix, which is the half more often skipped, is that the protective devices on the two paths must be referenced to the same point at the equipment end. Two protectors bonded to two different places leave the equipment between them still bridging a difference, which is the same failure with more hardware in it. The bonding and electrode requirements themselves are their own subject and this library covers them separately; what belongs here is that a path map with two protectors referenced to two points has not been closed.

Where a run genuinely leaves one building and enters another, the design question stops being a device selection and becomes a system question about how the two structures are bonded, and that is an engineering and code question for the adopted NEC edition and the authority having jurisdiction, not a field call.

What a map like this cannot settle

Two things stay outside a technician's judgment, and saying so on the ticket protects the shop.

Whether a modification voids a listing is not decided on site. Adding, extending or reworking the leads on a listed assembly, or mounting a device in an enclosure it was not evaluated in, is a question for the listing agency and the authority having jurisdiction. The safe field position is that the equipment is installed per its instructions or it is not, and a deviation gets written up rather than absorbed.

Where a machine sits in a classified location, none of the above is sufficient on its own. What a classified location requires of an enclosure, a device and a wiring method is settled by the area classification documents and the authority having jurisdiction, and a surge device chosen only for its electrical ratings can be the wrong device there for reasons that have nothing to do with surges.

Who owns which number

• Let-through at a real surge current belongs to the manufacturer's curve, not to the VPR on the label, which is a 3 kA listing figure.
• Where a given Type may be connected, and what overcurrent protection it needs, belongs to NEC Article 242 in the edition your authority having jurisdiction has adopted.
• Lead length limits and mounting belong to the manufacturer's instructions, which NEC 110.3(B) makes enforceable in that same edition.
• Whether the site needs coordinated protection across structures belongs to a designer with the site's grounding and bonding arrangement in front of them.

References

• 29 CFR 1910.333(a)(1) and (b)(2) for the de-energized-work gate and electrical lockout; 29 CFR 1910.147(a)(1)(ii)(C) exclusion; 29 CFR 1926.417 construction counterpart
• NFPA 70E-2021, 120.5, 130.5 and 130.7, in the edition adopted by your employer's electrical safety program
• NEC Article 242 (surge protective devices) and 110.3(B), in the edition adopted by your authority having jurisdiction
• UL 1449, the product safety standard under which surge protective devices are listed and their VPR determined, in the edition the device is listed to
• See related: Surge Protection Reference; Why Equipment Fails After a Power Event Rather Than During It