An industrial site, an ERP, or a data center is not just a roof and walls. It is a set of assets, people, processes, and often insurance requirements. Faced with lightning, you must therefore choose a lightning protection framework that holds up both technically and documentarily.
The real pitfall is not “installing a lightning rod”. The pitfall is installing something correct, then finding yourself without a solid file to attest to your prevention of occupational hazards during an audit, inspection, or after an incident. Here, we compare NF C 17-102 and IEC 62305, then clarify what to document, and why.
What NF C 17-102 and IEC 62305 really seek to prove

You can sum up the difference as two ways to “tell” lightning protection. NF C 17-102 speaks mainly of an external protection system based on lightning rods with early streamer emission devices (ESE). It provides a method of design, installation, and verification, with logic centered on the equipment.
For its part, IEC 62305 (presented in France as NF EN 62305) treats lightning protection as a lightning protection system. It starts from risk, then unfolds external protection (LPS, often including mesh cages), internal protection (equipotential bonding, shielding, routing), and overvoltage (SPD). Since the 2024 edition, certain concepts have evolved, notably around how to qualify risk entries and certain parameters of the analysis (according to available synthesis guides).
To frame the comparison between these international and national standards, we keep a simple rule: the first justifies a solution based on early streamer emission devices, the second justifies a level of overall protection, in line with expected consequences.
We often rely on reference resources to align design, maintenance, and QHSE teams, especially regarding compliance with mandatory standards. For example, the LPS France summary on the difference between NF C 17-102 and IEC 62305 helps lay the groundwork, especially when you need to explain the choice to an insurer or client.
A good normative framework is a file that “holds up” even when the person you’re talking to wasn’t there at the time of the work.
Before deciding, let’s look at the logical gaps, because these dictate the evidence to produce.
Here is a quick read, useful in the preliminary phase:
| Subject | NF C 17-102 (ESE/PDA) | IEC 62305 (LPS based on risk) |
|---|---|---|
| Starting point | ESE equipment and installation rules | Risk analysis, risk calculation, protection levels (LPL) |
| Scope | External protection centered on ESE | External + internal + overvoltage |
| Expected proof | Compliance with standard, controls, tests | Risk approach, layered consistency |
The table does not say “the best standard”. It says what each framework allows you to defend, backed by evidence.
How to choose the right framework according to the site, insurer, and international context

Choice is rarely made “by preference”. You choose because a context demands it. First, there is geography and customer standards. Then there is business criticality. Finally, there is the ability to demonstrate risk control, not just the existence of equipment.
In an ERP, you manage people, so safety and continuity requirements carry more weight. On an industrial site, the consequences of a shutdown or fire change the analysis. In a data center, external protection without an overvoltage protection strategy is like a reinforced door with an open window.
In practice, here is the framework for your site in three common scenarios:
- Project with strong international exposure: IEC 62305 (or NFPA 780) is more easily retained, because it is “read” everywhere, and because it structures insurance discourse.
- Site in France with ESE already in place: you can stay with NF C 17-102 for external protection, provided you have a clean verification file, and address the rest (grounding, equipotentiality, overvoltage) with a compatible framework.
- Critical site (ERP, sensitive process, data center): IEC 62305 is favored to drive risk analysis and internal protection, even if integrating an ESE solution if the specifications allow it.
On critical sites, integrating storm detectors is a way to improve overall protection.
Here, a point is often underestimated: IEC 62305 is not just a “calculation standard”. It is a justification method. The risk assessment, assumptions, and choices of protection levels become “insurance” documents.
To follow recent developments, you can also rely on the LPS France guide on IEC 62305:2024 changes (NSG, TWS), useful when you need to update a risk analysis or explain a difference between versions.
In parallel, when you need public access to NF C 17-102 to review a requirement (while waiting for the official AFNOR version), some use the PDF distributed by technical sites, for example the PDF of NF C 17-102 (2011) in English. However, we maintain discipline: we cite the exact version and trace the source, because the audit does not forgive imprecision.
What to document for audits, inspections and insurers (and how to avoid the “incomplete” file)
You can install a perfect system and lose the argument for lack of documents. Conversely, a clear file can avoid endless discussion, especially when maintenance changes teams. The objective is continuous traceability between design, execution, and verifications.
Documentation is generally structured in 4 blocks.
1) Justification of choice and protection level
For an IEC 62305 approach, you keep the risk analysis (assumptions, input data, reduction objectives). You also keep the protection level decision and trade-offs (for example, why this LPL). For NF C 17-102, you document the logic of ESE placement, the protected zone, and site constraints.
2) Design file and plans
You archive installation plans, descent routings, grounding points, and equipotential bonding connections. You include overvoltage protection coordination when it exists, with consistency of Type 1, Type 2, Type 3 overvoltage protectors according to architecture.
3) Product evidence and compliance
You keep technical data sheets, test reports, certificates compliant with UNE-EN IEC 62561, and component references. For durable connections such as exothermic welding in grounding, the UNE-EN IEC 62561 standard guarantees quality and electromagnetic compatibility of internal protection. To clarify certain debates (ESE vs conventional), a substantive resource like the article accessible via ResearchGate on ESE vs conventional systems can help, at least to frame internal discussion. However, avoid making it a substitute for the standard or specifications.
4) Verifications, maintenance, and field traceability
You establish a simple register: inspection dates, grounding measurements, continuity, state of bonding, dated photos, non-conformities, and corrective actions. One point helps a lot: link each issue to a location, a responsible party, and a closure date.
The day the insurer asks “prove maintenance”, you do not look for emails, you pull out a register.
This is where a platform like LPS Manager makes full sense for managing the lightning protection system. You centralize sites, keep reports, track verifications, and avoid scattered files. To align teams, you also rely on LPS France content, for example lightning protection standards in France and worldwide, practical when a multi-site group juggles between country requirements, clients, and insurers.
Conclusion: deciding also means preparing the evidence
Choosing between NF C 17-102 and IEC 62305 is not a schools debate. You seek the framework that fits the site, the stakes, and the way you will need to demonstrate comprehensive lightning safety. In many projects, you gain peace of mind by combining well-defined external protection with IEC 62305 logic for risk, internal protection, and overvoltage.
At the next audit, you want a file that reads quickly, with clear assumptions and tracked maintenance that provides reliable preventive protection. If you had to retain one simple rule, it would be this: a lightning protection system is worth as much by its installation of protection systems as by the evidence you know how to produce and retrieve, years later.