
What You’ll Learn
- The three risk assessment methods Annex F.6 names — brainstorming, checklists, and the risk assessment matrix — and what each one is for
- How a risk assessment matrix is built: severity down one side, likelihood across the other, values assigned to every row and column
- Why five categories is the practical ceiling for a useful matrix — and why too few categories fails just as surely
- How a risk register scores a task by adding severity, exposure, event likelihood, and avoidance values — and what the committee’s “higher than 10” threshold does with the total
- Why a switchboard breaker operated twice a year ends up requiring PPE while a lighting breaker operated twice a day does not
From Judgment to a Number You Can Defend
Every risk assessment ends with a decision: work as planned, add controls, or stop. Annex F.6 addresses the step before that decision — turning the assessment into something structured enough that two people looking at the same task reach the same answer. There are many risk assessment methods, and F.6 says the method, or combination of methods, should be chosen based on the application, the desired result, and the skill level of the persons performing the assessment. The annex names three methods directly, and its enhanced commentary adds a fourth family — qualitative scoring systems such as risk ranking and risk registers. This article works through all of them, then follows the annex’s own example end to end: two circuit breakers, scored two different ways, with a result that surprises most readers the first time through.
Method One: Brainstorming
Brainstorming is the simplest method on the list: an open group discussion regarding hazards, the associated risk, and risk control methods. Annex F places it inside the work crews already do — it can be used as part of pre-job planning and during a job briefing session. Its strength is coverage: the people who will actually perform the task surface hazards and controls that a form filled out alone in an office would miss. Its weakness is the mirror image — with no structure, the output is only as good as the discussion, and nothing is scored or compared.
Method Two: Checklists
A checklist is a list of common hazards and possible control methods, used as a tool for pre-job planning and for job briefings. Where brainstorming depends on what the group happens to think of, a checklist guarantees the known items are at least asked about, every time, in the same order. Annex F points to Annex I for an example of a job briefing and planning checklist. Checklists and brainstorming are not rivals: the checklist sets the floor, and the discussion explores what the list does not cover for this particular job.
Method Three: The Risk Assessment Matrix
A risk assessment matrix is commonly used to quantify levels of risk. A matrix is an ordered presentation of data in rows and columns that defines the cells of the array; a risk assessment matrix is the simple table version of that idea, grouping risk based on the two factors every evaluation must consider — the likelihood of the occurrence of harm and the severity of the injury that could result. The rows and columns are assigned values, such as low and high, and each cell of the grid then carries a risk level, often color-coded, that can be used to assess the need for remedial action — such as the use of PPE for a given task — and to prioritize safety issues against each other.
Building one has a built-in tension. A matrix can become cumbersome if too many columns and rows are used, and it can fail to impart the necessary information if too few are used. Practically speaking, the maximum number of categories for a useful matrix is five. The annex’s own Figure F.6 stays at the simple end: a qualitative two-by-two matrix with two categories for the rows (improbable and possible) and two for the columns (first energy level and second energy level). The matrices in the annex and its commentary are illustrations of aids for prioritizing and determining remedial actions — other published matrices, such as those in ANSI/AIHA Z10, offer different interpretations of the same idea.
The Risk Register: Four Values and a Threshold
The commentary’s worked example uses a risk register — a scoring system that breaks likelihood into the same three factors the F.1.1 risk analysis names, and adds them to severity. Each factor is read from a committee-selected table:
- Severity of injury — driven here by incident energy: a non-arc-rated switchboard at 31 cal/cm² scores high; a panelboard at 5.8 cal/cm², where the enclosure is assumed to hold at 8 cal/cm² or less, scores low.
- Frequency and duration of exposure — how often the task happens. Twice a year scores low; twice a day scores high.
- Likelihood of a hazardous event occurring — for properly applied, properly maintained breakers, negligible in both scenarios.
- Likelihood of avoiding or limiting injury — whether the enclosure can be counted on to contain an event. A switchboard that might be breached scores badly; a panelboard assumed to stay intact scores well.
The four values are totaled into a risk score, and the score is compared against a threshold the committee sets in advance: in the example, a risk score higher than 10 requires additional safety controls. The register’s virtue is that it forces the exposure factor and the severity factor to be weighed separately, instead of letting one dominate the discussion.
Worked Scenario
An electrical safety committee must analyze two tasks to determine whether they require additional safety controls. Both breakers are operated (turned on and off) while energized, in dead-front configuration, under normal operating conditions, within rated life, with no damp environment — and of the possible electrical hazards, only the arc flash hazard is assessed.
Task 1 is a 1,600-ampere breaker in the main 480Y/277-volt, non-arc-rated switchboard of a building with 35,000 amperes of available short-circuit current. The switchboard’s incident energy level is 31 cal/cm². The breaker is operated twice a year, when facility maintenance is scheduled.
Task 2 is a 20-ampere high-intensity discharge (HID) lighting breaker in a 480Y/277-volt lighting panelboard with 25,000 amperes available. The panelboard’s incident energy is 5.8 cal/cm². This breaker is operated twice a day, to turn the lighting on and off.
Scored on the risk register, Task 1 draws a severity value of 6 (31 cal/cm² into a switchboard that is not arc rated), an exposure value of 3 (twice a year), an event-likelihood value of 1 (negligible, because the breaker and switchboard are properly applied), and an avoidance value of 3 (it is not anticipated that the enclosure will always contain an arc flash event, so the likelihood of avoiding or limiting injury is rare). Total: 13 — above the committee’s threshold of 10.
Task 2 draws a severity value of 1 (at 5.8 cal/cm², the panelboard’s integrity is assumed not to be breached), an exposure value of 5 (twice a day — the highest-frequency task of the two), an event-likelihood value of 1 (negligible, for the same proper-application reason), and an avoidance value of 1 (the enclosure is assumed to remain intact, so avoiding or limiting injury is probable). Total: 8 — below the threshold.
The committee’s conclusion follows the scores. For these scenarios and this installed equipment, only administrative controls and PPE from the hierarchy of controls are in play — and based on the analysis, PPE is required when operating the 1,600-ampere switchboard breaker, and is not required when operating the 20-ampere HID lighting breaker. The task performed twice a day came out safer than the task performed twice a year, because the register weighs what happens if the event occurs — 31 cal/cm² against an enclosure that may not hold it — more heavily than how often hands are on the handle.
The committee then ran the same two tasks through a risk assessment matrix built on incident energy, using their assumption that incident energy levels of 8 cal/cm² or less will not breach the integrity of the enclosure, and rating the likelihood of an arc flash event as unlikely because the equipment has been properly selected, installed, and maintained. Task 1 — incident energy greater than 8 cal/cm² and less than 40 cal/cm², likelihood unlikely — lands in a cell color-coded yellow, a moderate risk, and the committee recommends that management consider the additional controls the matrix’s action table indicates. Task 2, at 5.8 cal/cm², falls in the matrix’s low-energy band on the same unlikely likelihood. Two methods, two different mechanics, the same ordering: the big, rarely touched breaker is the one that demands action.
One caution from the annex itself: these examples should not be relied upon for any particular installation. They demonstrate the thinking process a risk assessment requires — they are not a scoring table to copy onto different equipment.
Common Mistakes and Safety Notes
- Judging risk by frequency alone. The twice-a-day task scored lower than the twice-a-year task. Exposure is one factor of four; severity and the ability to avoid injury can outweigh it.
- Building a matrix with a category for everything. Beyond five rows or columns a matrix becomes cumbersome and stops communicating; too few categories fails the other way, hiding real differences.
- Treating the example’s numbers as your numbers. The annex states its examples are not to be relied on for any particular installation — the severity, exposure, and avoidance values came from tables the example’s committee selected for its own equipment.
- Scoring the breaker instead of the task. The register scores a task on specific equipment under stated conditions — normal operation, rated life, dead-front, dry environment. Change the conditions and the values change.
- Assuming a low incident energy settles the question. Task 2’s low score also depended on the enclosure-integrity assumption and a negligible event likelihood from proper application and maintenance — an unmaintained breaker would not earn the same values.
Summary
Annex F.6 offers a toolbox rather than a single formula: brainstorming for open group discussion in pre-job planning and job briefings, checklists to guarantee the common hazards and controls are always asked about, and risk assessment matrices that group risk by likelihood and severity in a grid of assigned values — five categories being the practical maximum before the grid stops being useful. The commentary’s risk register scores a task on four values drawn from committee tables — severity, frequency and duration of exposure, likelihood of the event, and likelihood of avoiding or limiting injury — and compares the total against a preset action threshold, higher than 10 in the example. Run through both tools, the annex’s two breakers reverse the intuitive answer: the 1,600-ampere switchboard breaker at 31 cal/cm², operated only twice a year, scores 13 and requires PPE, while the 20-ampere lighting breaker at 5.8 cal/cm², operated twice a day, scores 8 and does not — because severity and enclosure behavior, not frequency, carried the decision.
Quick-Check Questions
- What three things should drive the choice of a risk assessment method under F.6?
Answer: The application, the desired result, and the skill level of the persons performing the assessment. - Why is five the practical ceiling for the number of categories in a risk assessment matrix?
Answer: A matrix becomes cumbersome if too many rows and columns are used, while too few categories fails to impart the necessary information — five is the practical maximum for a useful matrix. - In the worked example, what risk score triggers additional safety controls, and which task exceeded it?
Answer: A score higher than 10. The 1,600-ampere switchboard breaker task scored 13 (severity 6, exposure 3, event 1, avoidance 3) and exceeded it; the 20-ampere HID lighting breaker task scored 8. - Why did the twice-a-day lighting breaker end up not requiring PPE while the twice-a-year switchboard breaker did?
Answer: Because the register weighs more than frequency: the lighting task’s low severity (5.8 cal/cm² into an enclosure assumed to stay intact) and probable avoidance of injury outweighed its high exposure value, while the switchboard task combined 31 cal/cm², a non-arc-rated enclosure, and a rare chance of avoiding injury. - How did the matrix analysis characterize the switchboard breaker task?
Answer: With incident energy greater than 8 cal/cm² and less than 40 cal/cm² and the likelihood of an arc flash event rated unlikely, the task fell in a yellow, moderate-risk cell, and the committee recommended management consider the additional controls indicated.
Based on NFPA 70E (2024), Informative Annex F, paraphrased for study.
Video lesson for this module is in production — coming soon.
Quiz for this module is being prepared.
Flashcards for this module are being prepared.



