HRN with ISO 12100: a tool to support the process, not a “shortcut to compliance”
In day-to-day risk assessment work under ISO 12100, the HRN method is sometimes treated as a “ready-made risk formula”. It is convenient, but it leads to a fundamental mistake: ISO 12100 does not define risk as a single number produced by a simple calculation. The standard expects risk to be estimated for specific scenarios (hazardous situation → hazardous event → harm), and it treats the “probability of occurrence of harm” as a function of exposure, the occurrence of the hazardous event, and the possibility of avoiding or limiting the harm. HRN can be used correctly, but only when it is applied as a tool that supports this logic—rather than replacing it.
This distinction directly affects what we are actually “calculating” in HRN. ISO 12100 is not limited to asking what hazards exist on a machine. It drives you towards scenarios: when a person is within the hazard zone, during what task, in what machine state, and for what reason. Only then does the risk become meaningful and real.
That is why HRN should not be applied “to the hazard” in isolation, but to a described use case. In practice, an HRN score only makes sense when it is tied to a specific accident scenario, not to a machine part as such.
Probability of occurrence of harm: the ISO 12100 element HRN cannot ignore
The biggest gap between the informal way HRN is often understood and the logic of ISO 12100 is the concept of “probability”. In ISO 12100, probability of occurrence of harm is not a single judgement such as “low / medium / high”. The standard presents it as a function of three components:
That third component is, in practice, the one most often lost when teams use HRN—and yet it is frequently one of the most decisive factors in real incidents. Two situations can have the same exposure and a similar chance of a hazardous event occurring, but still represent very different risk levels because in one case harm can realistically be avoided, while in the other case it is effectively unavoidable.
If HRN is to be consistent with ISO 12100, this aspect must be assessed deliberately and must influence the outcome—even if “pure HRN” does not include a dedicated A (avoidance) parameter.
“Pure HRN” in practice: relating LO, FE, NP and DPH back to ISO 12100
Classic HRN uses four factors (LO, FE, NP, DPH) and combines them into a single result (most commonly as a multiplication). The formula itself is not the issue. The issue is how teams define what each factor is supposed to represent.
If you want to use HRN without modifying the method, the key is consistent mapping to the risk elements in ISO 12100:
In practice, “pure HRN” only aligns with ISO 12100 when LO is not selected “in a vacuum”, but only after working through the three core considerations reflected in the ISO 12100 risk elements: who is exposed, what the hazardous event is, and whether harm can be avoided.
ISO 12100: qualitative and quantitative methods—where HRN really sits
ISO 12100 indicates that risk reduction decisions should be supported by a qualitative method, and—where appropriate—a quantitative one. At the same time, the standard explicitly recognises the limits of quantitative approaches: they only make sense when the necessary data are available, which in practice is often difficult.
That explains exactly why HRN is so popular. HRN produces a number, but it is most often not a quantitative method in the probabilistic sense. It is a scoring method—semi-quantitative—that structures the assessment, while still relying on expert judgement.
A “more quantitative” approach with HRN only starts when an organisation deliberately builds data foundations and anchors for the scale. In practice this means:
Without that, HRN can still be useful, but it functions as a structuring and comparative tool—not as a probability calculation.
How to use HRN deliberately: rules that keep the method aligned with ISO 12100
In day-to-day risk assessment, the aim is not to “calculate HRN perfectly”. The aim is for the HRN to drive sound design decisions and to remain consistent with the ISO 12100 logic. To achieve that, what matters most are working rules that reduce subjectivity and prevent the common shortcuts that distort results.
The most important ones are:
Within an ISO 12100 approach, HRN delivers the most value when it is used iteratively: once before a protective measure is applied and again afterwards. The number is not a “proof of safety”; it is an indicator showing whether the risk elements that should have changed have actually changed.
Case study: a “clean HRN” for a jam-clearing scenario (and what that means in ISO 12100)
In practice, one of the best tests of HRN is a situation where the risk does not come from “normal operation”, but from a typical ancillary task.
Machine and task context
Consider a section of a packaging line with a conveyor and a pair of infeed rollers (a draw-in / nip point). During operation, film or product jams occur and the operator clears them manually.
In an ISO 12100 assessment we describe hazardous situations, not a generic “hazard in the machine”:
Only at that level does it make sense to move on to HRN.
Assumed HRN scale (example)
To keep the case study clear, assume a simple internal rating scheme (what matters is consistency, not “perfect numbers”):
Note: this is not the “only correct HRN scale”. It is simply a consistent set that makes the decision mechanism transparent.
HRN rating – two variants of the same task
Variant A: clearing a jam in “JOG” mode with reduced speed and hold-to-run control
Practical assumptions:
Parameter selection:
HRN = 3 × 3 × 2 × 1 = 18
Design conclusion: the risk still exists, but it is “kept under control” by conditions that increase the operator’s ability to avoid harm (this is exactly the ISO 12100 aspect that HRN must intentionally reflect in the LO factor).
Variant B: the same task, but under “production must keep running” conditions (restart possible, higher speed, time pressure)
Practical assumptions:
Parameter selection:
HRN = 3 × 3 × 4 × 1 = 36
Design conclusion: what changed is not the “type of hazard”, but the human–machine relationship in a specific operating state. HRN shows the difference—but only if LO genuinely covers the ability to avoid harm.
NP: why the “number of persons multiplier” can be a trap (and how to handle it sensibly)
In the classic HRN method, NP is a multiplier. That works mathematically, but in practice it is often misread: the result starts to imply that the risk to an individual operator “increases” simply because a second person is nearby.
In a more up-to-date way of using HRN (also reflected in practical tools discussed in ISO/TR 14121-2), the number of persons more often appears as part of the context:
it increases the significance of the issue,
it raises the priority for action,
it influences how the area and access are organised,
but it does not have to be a multiplier of “individual risk”.
How to apply this in practice without undermining ISO 12100
The clearest approach is to separate two perspectives:
Individual risk (unit risk) – consistent with ISO 12100 logic
You calculate the HRN for the “most exposed person” and, in the vast majority of cases, you assume NP = 1. This aligns with how, in practical design risk assessment, harm to an individual is evaluated for a specific interaction.
Organisational priority / “scope” – a management decision, not the definition of risk
If the scenario could involve more than one person (e.g. an open area, operator + maintenance technician, bystanders), state explicitly in the report:
“possible simultaneous exposure of multiple persons: YES/NO”
“maximum number of persons in the area during the task: …”
and use this to increase the priority of actions (e.g. within a retrofit plan, implementation schedule).
This approach addresses a common HRN pitfall: you avoid artificially inflating the “operator risk” while still recognising that the same scenario may affect more than one person.
When NP makes sense as a multiplier (and doesn’t distort the result)
Using NP as a multiplier is generally only justified when dealing with scenarios such as:
- energy release / ejection of parts,
- fire, explosion,
- situations where a single failure can affect several people at the same time.
Even then, it is worth asking a basic methodological question:
should “multiple casualties” be reflected in the severity of harm (extent of harm) rather than through a multiplier?
As with any risk assessment, what matters is the interpretation of the outcome rather than the numeric value itself, and the final risk evaluation decision ultimately sits with the engineer.