InnerWeb
Skip to main content

Risk scale with thresholds 20, 70, 200 and 400: recommendations for the OHS manager

The risk level in the editor is the product of probability, exposure and severity, compared with the thresholds 20, 70, 200 and 400. An OHS manager who loads the catalogue for the first time needs a few rules so that assessments from different halls can be compared. Below are seven recommendations with their rationale.

4
risk level thresholds: 20, 70, 200 and 400
Source: feature card

Who these recommendations are for

We address these recommendations to the OHS manager and to the person loading the risk catalogue when the system is started up at a production plant. We assume the plant already has occupational risk assessments in its documents and transfers them to the catalogue, in which every risk has an assessment before and after the measures are applied.

The risk level in the editor is calculated as probability times exposure times severity and is compared with the thresholds 20, 70, 200 and 400. The editor shows it live as a level marker, so the person entering the data sees the consequence of every change immediately. The assessment is prepared and approved by a person, and the system does not assess risk automatically.

Recommendations 1 and 2: first define what the values mean

Recommendation 1: before entering data, write a one-page table that states what each value of probability, exposure and severity means at your plant. The product is only as comparable as its components. If one hall rates severity by injury and another by downtime, the thresholds 20, 70, 200 and 400 will mean different things.

Recommendation 2: enter a reference example for each threshold and describe it in words. An example from below 20 and an example above 400 help the people doing the assessment understand the limits of the scale. Keep the reference examples next to the instructions for loading the catalogue, so that a new person on the team has something to refer to.

Example: product before and after preventive measures

Example: product before and after preventive measuresBar chart: risk level 126 before measures and 42 after measures, with thresholds 20, 70, 200 and 400.050100150product pointsBefore measures: 126 product points126Before measuresAfter measures: 42 product points42After measures
Source: Sample calculation: 3 x 6 x 7 before and 1 x 6 x 7 after the measures; thresholds 20, 70, 200, 400 from the feature card

Recommendations 3 and 4: the assessment after measures is not a second assessment from zero

Recommendation 3: assess the risk before the measures and after them separately, but remember what usually changes. After the measures are applied, exposure and severity usually stay the same, and probability drops. That is why the editor by default copies exposure and severity from the assessment before the measures and stops doing so once the user changes a field by hand.

Recommendation 4: change exposure or severity after the measures only when the measure actually changes them, for example when a physical barrier limits the time a person spends in the zone. A manual change switches off the automatic copying, so when the values before the measures are corrected later, you must make sure the values after the measures stay consistent.

Recommendations 5, 6 and 7: thresholds, exceptions and review

Recommendation 5: do not treat the thresholds as certainties, only as signals. A level calculated from the product shows that a risk needs attention, but the decision on a measure belongs to the person responsible. Recommendation 6: for risks where the severity is irreversible, set an internal rule that, regardless of the product, they require a separate decision by the manager. The numerical scale does not replace this.

Recommendation 7: after the first quarter, review the distribution of levels in the catalogue. If almost all risks fall into one range, the scale is probably badly calibrated. In that case, improve the descriptions of the values, not the thresholds.

  • Write down the meaning of the probability, exposure and severity values in one table.
  • Describe a reference example for each of the four thresholds.
  • Assess before and after the measures separately, changing mainly the probability after the measures.
  • Change exposure or severity after the measures by hand only when there is a real change.
  • Treat the thresholds as a signal, not an automatic decision.
  • For irreversible severity, set a separate decision rule.
  • After a quarter, check the distribution of levels in the catalogue.

Calculation example and its interpretation

Illustrative example: probability 3, exposure 6, severity 7 gives a product of 126. This result sits above the threshold of 70 and below 200, so it requires decisive attention. After preventive measures are introduced, probability drops to 1, exposure and severity remain, and the product is 42, which is below the threshold of 70 and above 20.

The values in the example do not come from a specific plant and serve only to show the reasoning. Your own values must be set on the basis of the definitions adopted at your plant.

Example calculation

Example of calculating the risk level

Assumptions
ParameterValue
Probability before measures3 pts
Exposure6 pts
Severity7 pts
Probability after measures1 pts

Result: Before measures 3 x 6 x 7 = 126; after measures 1 x 6 x 7 = 42.

Note: Sample calculation with illustrative values, not taken from a specific plant.

Video

Frequently asked questions

How is the risk level calculated?

As probability times exposure times severity. The result is compared with the thresholds 20, 70, 200 and 400, and the editor shows it live as a level marker.

Why are exposure and severity the same after the measures?

Usually measures lower the probability and do not change exposure or severity, which is why the editor copies them by default. It stops doing so once the user changes a field by hand.

Does the system assess the risk on its own?

No. The documentation does not describe any automatic assessment by artificial intelligence. The assessment is prepared and approved by a person.

Let's go through the risk scale on your data

Book an editor demo: together we will enter several risks and show how the level marker reacts to a change in the values.

Let's talk