SHARE Knowledge Hub

SFARP: So Far As Reasonably Practicable

So far as reasonably practicable (SFARP) is a qualified duty in Great Britain and Northern Ireland health and safety law. A dutyholder considers the risk and the further control available, then implements it unless the sacrifice in money, time or trouble is grossly disproportionate to the risk reduction. The judgement starts with the hazard, recognised good practice and credible control options, not a preferred budget.

Updated 2026-10-22UK focus

Understanding SFARP: So Far As Reasonably Practicable

So far as is reasonably practicable is a qualification used in important Great Britain health-and-safety duties. It requires a dutyholder to reduce risk unless the sacrifice of a further preventive measure, in money, time or trouble, would be grossly disproportionate to the risk reduction achieved. It is a demanding, context-specific test, not a promise of zero risk or a licence to choose the cheapest option.

Why this matters at work

Hazard identification is not confined to observed harm. Long-latency respiratory disease, exposure to noise or vibration and low-frequency major accidents may have few warning injuries. If the work could plausibly produce serious harm, it needs a prevention decision even without a recent event.

Identify employees, agency staff, contractors, cleaners, maintenance workers, visitors, service users, neighbours and other members of the public where relevant. State the harm pathway for each group. A vague entry such as all staff can conceal who lacks training, who crosses the traffic route or who inhales a released substance.

How organisations manage this in practice

Hazard identification should use observation, worker knowledge, manufacturer information, near misses, ill-health records and non-routine-task review. Risk assessment considers severity, likelihood, exposure, existing controls and uncertainty. Control selection asks first whether the hazard can be removed or redesigned, then how access or exposure is prevented or reduced.

HSE presents five connected steps: identify hazards, assess risks, control risks, record findings and review controls. The cycle begins with seeing actual work and asking workers, then considers who could be harmed and how. It must end in action and verification; a completed assessment with no control change is not a finished risk decision.

Review the decision when technology, good practice, workforce, plant, process, occupancy, incident evidence or law changes. A new feasible control may alter what is reasonably practicable. Set triggers and a named owner; do not wait for a scheduled three-year review if a serious near miss or weak signal appears now.

Review after an incident, near miss, health concern, failed control, new evidence or worker report, as well as at a sensible routine interval. Determine whether the change affects other sites, shifts or similar tasks. If the control is no longer adequate, take interim protective action before waiting for the next formal review date.

Record the further measure and the evidence for the decision

So far as reasonably practicable requires a reasoned assessment of risk and further precautions. A cost comparison, a low score or a completed form is not an automatic permission to leave risk uncontrolled.

Safety professional reviewing guidance and documents in a UK workplace office. Photorealistic illustration.
Illustrative UK workplace scene. Read the practical checks below for responsibilities, evidence and limits. Download image · Download smaller image
  1. Credible harm

    Describe the actual conveyor intervention, energy path and people affected. Assess the credible seriousness and likelihood using available evidence and explicit uncertainty.

  2. Applicable good practice

    Establish the relevant legal requirements and recognised good practice first. Do not use a local calculation to justify falling below an applicable minimum requirement.

  3. Further precautions

    Consider avoiding entry, improving the feed arrangement and appropriate guarding/isolation before residual administrative or PPE measures. Assess the full method and any new risks.

  4. Risk reduction and sacrifice

    Record what the measure would achieve and relevant feasibility, time, trouble and cost evidence. Apply the legal concept with competent advice; a simple equal-cost benefit balance is insufficient.

  5. Unresolved evidence

    Obtain competent technical or legal advice where needed and keep affected work on hold when essential safety evidence is missing. Uncertainty cannot create a green approval.

  6. Decision and review

    Identify the competent decision owner, reasons, implemented precautions and review triggers. Review when the work, technology, knowledge or conditions change.

Read the complete controls: ERICPD planning prompts

Discipline verifies and sustains all the controls; it is not a separate substitute for them.

SFARP: So Far As Reasonably Practicable — ERICPD planning prompts
Prompt or levelComplete task-specific example
EliminateChange the clearing method or remove the blockage source.
ReduceAssess an alternative feed design.
IsolateProvide suitable fixed or interlocked nip-point guarding.
ControlUse verified isolation and a written clearing method.
PPEGloves cannot make contact with a moving nip safe.
DisciplineTest guards and stop for damage, bypass or failed isolation.
Read the complete controls: ESEAP effectiveness order

Compare stronger feasible controls while preserving the actual task, checks and release requirements.

SFARP: So Far As Reasonably Practicable — ESEAP effectiveness order
Prompt or levelComplete task-specific example
EliminationChange the clearing method or remove the blockage source.
SubstitutionAssess an alternative feed design.
Engineering controlsProvide suitable fixed or interlocked nip-point guarding.
Administrative controlsUse verified isolation and a written clearing method.
PPEGloves cannot make contact with a moving nip safe.

No cost threshold, risk score or generic drawing establishes that a live task meets the applicable SFARP duty.

What does SFARP mean?

SFARP requires a risk-and-sacrifice judgement about further controls, with serious harm and recognised good practice considered before cost.

The decision starts with the credible harm and a control that could prevent it. For a machinery nip, the dutyholder should test an interlocked guard against recognised practice and record how it will be installed, checked and maintained before weighing any sacrifice. HSE machinery safeguarding.

The phrase appears simple, but the judgement is demanding. A dutyholder must first understand the risk and what more could be done to reduce it. Only then can they consider whether the sacrifice in money, time or trouble would be grossly disproportionate to the risk reduction achieved.

For serious risks, the balance is deliberately weighted toward safety. A control that prevents fatality, life-changing injury, major occupational disease or a major accident will usually require a much stronger reason before it can be rejected.

SFARP is also time-specific. A decision that was reasonably practicable with yesterday's knowledge, technology and standards may not remain acceptable after an incident, new guidance, better equipment, changed work or evidence that controls are failing.

ConceptMeaningProfessional caution
Reasonably practicableA balance between risk and the sacrifice needed to reduce it.Not the same as cheapest, easiest or industry habit.
RiskLikelihood, severity, exposure, vulnerability and uncertainty around harm.Do not rely only on a matrix score.
SacrificeMoney, time, trouble, technical difficulty and operational disruption.Must be weighed against the risk reduction benefit.
Gross disproportionWhere the duty is qualified by reasonable practicability, rejection of a further control requires sacrifice grossly disproportionate to risk reduction; more serious risk strengthens the required justification.The balance is weighted toward safety.
Good practiceRecognised guidance, standards and accepted controls.Departing from good practice needs strong justification.
Hierarchy of controlHigher-order controls should be considered before PPE or behaviour.Weak controls are harder to defend where stronger controls are available.
Residual riskRisk remaining after controls.Needs ownership, communication, monitoring and review.

How can the hierarchy of controls guide risk reduction so far as reasonably practicable?

Reasonably practicable judgement tests available stronger controls before accepting weaker safeguards.

At a conveyor nip, first eliminate access by changing the clearing method or removing the blockage source. If that is impossible, consider an alternative feed design, then a fixed or interlocked guard that physically separates hands from moving parts. The engineer specifies and tests the guard; maintenance records its inspection; the supervisor verifies it is present before start. Isolation, a written clearing method and training address residual intervention, while gloves cannot make contact with a moving nip safe. In ERICPD (Eliminate, Reduce, Isolate, Control, PPE, Discipline), a teaching framework used here, Discipline means continuing supervision and checks that guards and isolation remain effective. Under the HSE five-level hierarchy, the guard is engineering control and the procedure is administrative control. A damaged guard or bypass is a stop point, not an occasion to claim the risk is already ALARP. HSE HSE hierarchy of controls HSE.

How should SFARP decisions be made?

A defensible SFARP decision follows a structured chain from risk understanding to options, comparison, decision, evidence and review.

The first step is not cost. The first step is risk understanding. What could happen, to whom, how severe could it be, how often is exposure created, how reliable are existing controls and what uncertainty exists? Without that analysis, a cost argument is premature. HSE.

The second step is control options. A dutyholder should consider elimination, substitution, design, engineering, isolation, safer systems, competence, supervision, maintenance, monitoring and PPE. The final decision should explain why the selected control package is adequate and why further controls are or are not reasonably practicable.

StepQuestionEvidence
Define the hazard scenarioWhat could credibly cause harm?Risk assessment, incident history and technical evidence.
Estimate risk and uncertaintyHow severe, likely and uncertain is the risk?Exposure data, consequence analysis, worker input and monitoring.
Identify available controlsWhat further controls could reduce the risk?Good practice, standards, supplier options and specialist advice.
Assess control effectivenessHow much would each option reduce the risk?Control reliability, hierarchy position and residual risk.
Assess sacrificeWhat money, time, trouble or operational impact is involved?Cost/time estimates and feasibility review.
Apply gross-disproportion testIs sacrifice grossly disproportionate to the risk reduction?Recorded judgement with competent approval.
Implement and verifyAre selected controls installed, used and maintained?Action closure, inspections, maintenance and assurance.
ReviewHas knowledge, technology, guidance or work changed?Review trigger and management record.

How would SFARP apply to a machinery guarding decision?

A worked example shows why cost alone is not enough to reject a stronger control for serious machinery risk.

HSE’s work-equipment safeguarding guidance explains fixed and interlocked guarding; the exact machine and work location still determine the duty and suitable design.

Decide controls for a conveyor nip point

Task material and setting: A production conveyor jams at a reachable nip where blockage clearing could amputate fingers; the outcome is a reliable feed and safe clearing method.

People and prerequisites: The engineer, operator, maintenance team and accountable dutyholder identify foreseeable access, machine instructions, recognised guarding practice and isolation needs before restart.

Before: Test removal of the blockage source or a redesigned feed. Where access remains foreseeable, specify and commission an interlocked guard and an isolated clearing method, then train operators on stop conditions.

During: The operator checks the guard and isolation before clearing and reports faults; maintenance tests the interlock and prevents bypass. The supervisor observes real use, not only a document.

Hold or stop: Any bypassed or defective guard, unexpected access or failed isolation stops the line. Signs and gloves cannot replace the physical barrier.

Completion and hand back: Engineering verifies guard function after maintenance, reinstates isolation and hands the line back to the supervisor; the dutyholder reviews any unresolved further control against risk and sacrifice.

Evidence: Keep the harm scenario, good-practice comparison, options and gross-disproportion reasoning, commissioning test, defect log, training observation and signed hand-back.

Assume a production conveyor has a nip point exposed during blockage clearing, with a credible amputation outcome. Before restart, the engineer tests whether the blockage source can be removed or the feed redesigned; if access remains foreseeable, an interlocked guard is specified and tested under fault and cleaning conditions. The manager cannot reject it merely because installation interrupts a shift. During work, operators use the isolated clearing method and report any bypass or fault; a bypass stops the line. After commissioning, the engineering authority records the guard test, residual tasks, maintenance checks and hand-back. Only a documented comparison of risk reduction and grossly disproportionate sacrifice could support rejection of an additional measure; the existing sign and gloves cannot substitute for the guard.

Decision elementWeak reasoningStronger SFARP reasoning
Risk descriptionOperator could be injured if careless.Operator could suffer amputation when clearing a foreseeable blockage at an exposed nip point.
Existing controlsTraining and warning sign.Training relies on perfect behaviour during pressure and does not prevent access to danger.
Further controlGuarding too expensive.Fixed/interlocked guard would remove access during operation and substantially reduce serious injury risk.
SacrificeProduction downtime and engineering cost.Quantified cost and downtime considered against life-changing harm potential.
DecisionLeave as is until next year.Install guarding, set interim controls and verify after modification.
EvidenceManager says operators are experienced.Risk assessment, PUWER review, engineering quote, action plan, training and inspection record.

How does SFARP apply beyond machinery guarding?

High-hazard work usually needs stronger evidence because the potential harm is serious and the control decision may depend on competence, design and verification.

Machinery guarding is an easy example because the harm pathway is visible. The same SFARP logic applies just as strongly to temporary work at height, DSEAR ignition control, process isolation, confined-space entry and other high-hazard decisions where a weak control can lead to fatality or major loss.

The professional test is whether the dutyholder has considered recognised good practice and stronger controls before accepting residual risk. A short duration, awkward programme, familiar job or expensive control does not automatically make a weaker approach defensible.

Consider a ten-minute roof inspection in Great Britain. The planner first asks whether the inspection can be completed from the ground; if not, they select a safe existing place or collective fall prevention before considering measures that only limit a fall. Duration changes exposure, but does not erase the fatal-fall consequence. The owner records why the chosen access method is suitable, verifies it before use and postpones the inspection if the planned protection is unavailable. This follows the HSE work-at-height decision sequence.

For solvent transfer at a Great Britain process site, the decision begins with the substance and credible release, then whether the quantity or release can be reduced and ignition kept out of any hazardous atmosphere. A past absence of fires is weak evidence that the release and ignition controls work. The technical owner should verify the selected containment, ventilation and equipment arrangement against the actual process before restart; an unverified loss of ventilation is a stop point. The HSE DSEAR control order supports this Great Britain example. For a Northern Ireland site, confirm the corresponding NI duties and HSENI guidance before applying a legal conclusion.

High-hazard decisionWeak reasoningStronger SFARP reasoning
Temporary work at heightThe job only takes ten minutes, so a ladder and care are enough.Fatal fall potential remains even for short work; consider avoiding height, using fixed access, scaffold, MEWP, edge protection, rescue arrangements and competent supervision before relying on behaviour.
DSEAR ignition controlWe have never had a fire and people know not to smoke.Assess release scenarios, vapour/dust accumulation, zoning, ventilation, equipment selection, bonding/earthing, hot-work control, gas testing and proof that ignition sources are controlled.
Process isolationThe operator will close the valve and check the line.Serious stored-energy or loss-of-containment risk may require positive isolation, lockout, drain/vent, verification, permit control, line-breaking precautions and handback evidence.
Confined-space entryIt is a quick inspection and the space looks clean.First establish whether the purpose can be achieved without entry. If entry is unavoidable, the applicable GB/NI confined-space instrument requires a safe system and emergency arrangements before work, including competent assessment of isolation, atmosphere, ventilation, rescue and any permit controls. A permit alone does not justify entry.
Live electrical workSwitching off would delay production.In GB, all regulation 14 conditions must be met: making the conductor dead is unreasonable in all the circumstances; working on or near it live is reasonable in all the circumstances; and suitable precautions prevent injury. Production delay alone does not establish this. Dead working uses secure isolation and proving dead; that is separate from the exceptional live-work decision. Use the separate NI Electricity at Work Regulations 1991, regulation 14, for NI.

For a recurring roof inspection, compare relocating the inspection point or using remote access with the remaining need to work at height. Where access remains necessary, the competent planner assesses collective fall prevention before personal protection, including safe access and rescue. Record assumptions, access frequency and how the chosen arrangement will be checked. Reopen the decision when roof condition, task or available equipment changes. This illustrates the decision chain; it does not approve a ladder, platform or harness for an unspecified site. HSE height-control sequence.

The live-work gateway is a specific legal test, not a general budget comparison. See HSR25 regulation 14 and its supporting explanation; retain the distinct NI instrument. Confined-space planning starts by avoiding entry rather than selecting precautions after entry has already been assumed.HSE HSR25: regulation 14 live-work conditions HSE confined spaces: avoid entry and plan safe work

How do SFARP and ALARP relate?

SFARP and ALARP are usually treated as expressing the same core reasonably practicable test, but the language is used in different contexts.

SFARP is shorthand for a qualified duty; the exact statutory phrase must be read in the applicable instrument. ALARP is commonly used in risk management, engineering and major hazard contexts to describe the level to which risk should be reduced. HSE.

For practical purposes, both require more than a risk score. They require a structured judgement about risk, further measures, effectiveness, sacrifice and evidence. In high-hazard sectors, ALARP demonstrations may be formal and technical; in lower-risk workplaces the same principle may be recorded in a simpler decision note.

TermTypical usePractical meaning
SFARPLegal duty wording.Do what is reasonably practicable to reduce risk.
SFAIRPAlternative spelling/abbreviation for so far as is reasonably practicable.Same practical idea as SFARP.
ALARPRisk management and engineering shorthand.Risk reduced as low as reasonably practicable.
Gross disproportionDecision test where risk reduction is compared with sacrifice.For serious risk, sacrifice must be very high before control is rejected.
Good practice benchmarkGuidance, standards or recognised controls.Following good practice is usually expected unless there is a good reason not to.
ALARP demonstrationFormal evidence package in higher-hazard contexts.Shows options, reasoning, controls and residual risk.

The offshore well-examination guide, page 16, explains the underlying SFAIRP/ALARP and gross-disproportion concept. Its well-examination requirements are sector-specific and do not establish machinery or NI duties. The actual statute and topic-specific safeguarding guidance determine those duties. HSE conceptual explanation, offshore source.

How does SFARP compare with other legal duty wording?

Qualified duties require judgement; absolute duties do not allow the same risk-cost balance.

safety qualification learners often lose marks because they treat all legal phrases as the same. They are not. 'So far as reasonably practicable' allows a risk-sacrifice balance. 'Practicable' is closer to what is technically possible. 'Absolute' duties require compliance where the duty applies.

Terms such as suitable and sufficient, adequate and effective also matter because they describe the quality of assessment or control. A control can be cheap and present but still not adequate.

The particular duty governs the decision. HSE PUWER guidance requires effective measures against access to dangerous machinery parts and describes the safeguarding hierarchy; a generic SFARP comparison does not replace that requirement. First identify the applicable safeguard duty and recognised practice, then assess any further option. Do not infer that a small employer or a small machine automatically needs less guarding. HSE specific safeguarding duty.

What does gross disproportion mean in practice?

Gross disproportion means the sacrifice must be clearly out of proportion to the risk reduction before further control can be rejected.

The gross-disproportion idea prevents dutyholders from rejecting life-saving controls because they are merely inconvenient or expensive. The higher the risk, the greater the sacrifice expected before a control can be dismissed. HSE.

This is why a control that might be excessive for a low-risk office task could be entirely reasonable for work at height, confined spaces, process safety, live electrical work, hazardous substances or machinery entanglement.

The starting presumption is to adopt the risk-reducing measure. To reject a further guard or independent shutdown, the dutyholder needs a credible estimate of the harm and how much the measure would reduce it, plus evidence of the money, time and trouble involved. The comparison is weighted toward safety, especially for severe outcomes. An unsupported statement that the company cannot afford it, a low matrix score or a quiet incident history does not establish gross disproportion.

Risk situationLikely judgementWhy
Low-risk office paper cut hazardExtensive engineering control unlikely to be required.Risk reduction benefit is tiny.
Warehouse vehicle-pedestrian crossingPhysical segregation likely to be strongly expected if feasible.Fatality potential is high and barriers can be effective.
Machinery nip pointGuarding/interlocking likely to be reasonably practicable.Severe injury risk and higher-order control available.
Silica dust cuttingOn-tool extraction and RPE are expected controls.Occupational disease risk is serious and controls are recognised.
Tank overfill major accidentIndependent alarms/trips and operating discipline need strong evidence.Major accident consequences require robust barriers.
Temporary contractor roof accessProper access/edge protection expected.Short duration does not remove fatal fall potential.

What evidence makes a SFARP decision defensible?

Defensible decisions show what was known, what options existed, what was chosen, what was rejected and why.

After an incident, vague statements such as 'we considered it too expensive' are weak. A stronger record shows the risk scenario, good practice reviewed, options considered, cost and feasibility, control effectiveness, decision owner, interim controls and review date.

The evidence should be created at the time of the decision, not reverse-engineered after harm. It should be clear enough that an auditor, regulator, insurer or court can understand the reasoning.

EvidenceWhat it should showQuality question
Risk assessmentCredible harm scenario, people affected and existing controls.Does it explain how harm could occur?
Good-practice reviewRelevant HSE guidance, ACOPs, standards or industry practice.Was recognised good practice considered?
Options registerFurther controls identified and compared.Were higher-order controls considered first?
Effectiveness estimateHow each option reduces likelihood, severity or exposure.Is the control benefit credible?
Sacrifice estimateMoney, time, trouble and feasibility.Is the sacrifice evidenced, not asserted?
Decision recordChosen control, rejected options, rationale and authority.Who accepted residual risk and why?
Interim controlTemporary measures until permanent control is complete.Is temporary risk managed and time-limited?
VerificationEvidence that controls were installed and working.Was the decision implemented in reality?

What is the difference between a poor and defensible decision?

A defensible decision explains control options and proportionality before accepting residual risk.

A poor decision usually starts with the desired answer: 'we cannot afford it' or 'we have always done it this way'. A defensible decision starts with the risk and works through the controls.

The most useful management habit is to record the point at which further risk reduction was considered and why a chosen control package was adequate.

IssuePoor decisionDefensible decision
CostToo expensive.Cost compared with serious-harm prevention and alternatives.
TimeWe need to finish today.Temporary controls and stop/go decision made by authorised person.
Good practiceOther sites do it this way.Guidance and standards checked; deviation justified if needed.
Worker behaviourThey should be more careful.Task design, supervision and recovery controls reviewed.
Residual riskAccepted by manager verbally.Recorded with owner, review date and evidence.
ReviewLook again next year.Review after change, incident, new guidance or control defect.

How does SFARP apply to major hazard and process safety decisions?

Major hazard decisions need stronger evidence because the potential consequences are severe and controls are often technical.

In process safety, SFARP/ALARP thinking is applied to loss of containment, fire, explosion, toxic release, overpressure, runaway reaction, emergency shutdown and safety-critical elements. The decision may involve HAZOP, LOPA, SIL assessment, relief design, inspection regimes and emergency planning. HSE.

A site with no recent injuries may still fail SFARP if safety-critical barriers are overdue, alarm bypasses are normalised or MOC actions remain open. Personal safety performance does not prove major accident risk is controlled.

For a tank overfill, the operator must consider how a loss of containment could start, which independent barriers prevent or limit it, how those barriers are tested and what happens if one is unavailable. Recognised design and operating practice should be established before a cost argument about additional protection. A technical authority reviews consequence, likelihood and uncertainty; the dutyholder funds and verifies necessary measures. An impaired level alarm or isolation valve is a hold or restricted-operation decision under the site's safety arrangements, not a reason to recalculate risk after the fact.

ScenarioSFARP questionEvidence
Relief valve overdueIs continued operation tolerable while the protective device is overdue?Relief register, inspection plan, temporary risk review and authorisation.
Trip bypassWho can authorise bypass and what compensating controls exist?Bypass register, time limit, risk assessment and handover.
Tank overfillAre independent high-level controls reliable?Proof test, alarm response, operating envelope and training records.
Hot work in process areaCan ignition risk be eliminated or controlled before permit?Isolation, gas testing, DSEAR controls and fire watch.
MOC open actionsCan start-up proceed with unresolved actions?Pre-start-up safety review, action categorisation and senior approval (distinct from the Pressure Systems Safety Regulations).
Ageing plantIs inspection sufficient for degradation risk?Integrity plan, inspection evidence and defect management.

When should SFARP decisions be escalated?

Escalate when risk is serious, good practice is not being followed, controls are impaired, specialist judgement is needed or cost is being used to reject control.

SFARP decisions should sit with people who have the competence and authority to make them. A supervisor can stop work or apply known controls, but should not be expected to accept unresolved fatality risk, technical uncertainty or legal departure from recognised good practice. HSE.

Escalation should include options, risk reduction benefit, sacrifice, interim controls and decision required. This turns a concern into a management decision rather than a local argument.

An operator can identify a failed barrier and stop the task, but cannot make the organisation's final risk-sacrifice judgement. A supervisor should record the defect, keep people clear and call the engineer and accountable dutyholder. The decision record should state the applicable GB or NI duty, good practice, options, assumptions, people exposed, reviewer and date for reassessment. Restart depends on demonstrated control, not merely a signed decision sheet.

TriggerWhy it mattersEscalation route
Fatal or major-harm potentialThe balance is weighted strongly toward further control.Senior risk owner and competent adviser.
Departure from good practiceMay be hard to defend without strong evidence.HSE lead, legal/technical authority.
Higher-order control rejectedPPE/procedure may be weak if engineering control is feasible.Manager with budget/design authority.
Technical uncertaintyRisk or control effectiveness not understood.Specialist assessment.
Interim controls becoming normalTemporary risk acceptance is drifting.Management review and time-limited authorisation.
Cost used as main argumentCost must be weighed against risk reduction, not asserted.Documented SFARP review.

Where do SFARP decisions commonly fail?

SFARP fails when organisations reverse-engineer decisions, understate risk, ignore good practice or treat cost as a trump card.

The classic weak decision is to choose the preferred business option first and then search for safety arguments afterwards. That is not SFARP reasoning. A defensible judgement starts with risk and control options.

Another common failure is over-reliance on past success. 'We have never had an accident' is not evidence that further controls are not reasonably practicable. It may only show that exposure has not yet resulted in harm.

Failure or weak practiceHow it developsWhat good looks like
Cost-first reasoningControl rejected before risk is understood.Risk and options assessed before sacrifice.
Generic risk scoreMatrix number used as evidence.Scenario and control effectiveness explained.
Good practice ignoredKnown guidance not checked.Relevant HSE/industry standards reviewed.
Behaviour-only controlTraining and signs used where engineering control exists.Hierarchy of control considered.
No ownerResidual risk accepted informally.Decision owner and authority recorded.
No interim controlsAction deferred with no temporary protection.Time-limited interim controls and review.
No verificationDecision made but control not checked.Field verification and assurance evidence.

What evidence shows SFARP has been achieved?

Evidence should show risk understanding, control options, reasonably practicable judgement, implementation and review.

There is rarely a single document that proves SFARP. Evidence comes from the assessment, control design, good-practice review, option comparison, action tracking, verification and management review.

For high-risk decisions, a clear decision record is essential. It should be understandable to someone who was not in the meeting and should show why the decision was reasonable at the time it was made.

Evidence itemWhat it should showAssurance question
Risk assessment or work-impact assessmentScope, affected people, risks, controls, actions, owner and review date.Does it match the work as done?
Action trackerActions, owners, dates, priority, interim controls and close-out evidence.Are important actions verified rather than marked complete?
Training and briefing recordsWho was briefed, on what, when and how understanding was checked.Can workers and managers explain the control?
Inspection or monitoring recordsPlanned checks, defects, findings, trends and corrective action.Are checks focused on real controls and weak signals?
Worker feedbackConcerns, suggestions, barriers, confidential routes and follow-up.Do workers trust the system enough to speak up?
Management reviewTrends, resources, assurance confidence, exceptions and decisions.Are leaders seeing useful risk and wellbeing information?
Good practice reviewHSE guidance, standards, ACOPs, industry practice and technical advice checked.Has deviation been justified?
Options analysisRisk-reduction measures considered with hierarchy of control.Were stronger controls assessed before weaker controls?
Sacrifice assessmentMoney, time, trouble and feasibility compared with risk reduction.Is sacrifice evidenced?
Decision authorityWho accepted residual risk and why.Was the decision made at the right level?
Review triggersChange, incident, guidance, technology or control failure.Will the decision be reopened when facts change?

How should SFARP decisions be assured?

Assurance should test whether risk reduction decisions were made with evidence, implemented in practice and reviewed after change.

First-line checks should verify that chosen controls are actually in use. Second-line assurance should sample SFARP decisions and action quality. Third-line review should test high-risk, technical or legally sensitive decisions independently.

The assurance question is direct: if a serious incident occurred tomorrow, could the organisation explain why further controls were or were not reasonably practicable?

Assurance layerWhat to checkExample test
First lineControls selected by assessment are present and used.Observe task and compare to assessment.
Second lineOption analysis, good practice and action closure.Sample rejected controls and challenge reasoning.
Third lineHigh-risk legal/technical decisions.Independent review of major hazard, guarding, work at height or COSHH decisions.
Management reviewOverdue risk-reduction actions and residual risk acceptance.Check whether leaders approve or resource controls.
Incident learningWhether incidents reveal previously rejected controls.Reopen SFARP judgement after near miss or failure.
Worker voiceWhether workers can challenge impractical controls.Ask if current controls are usable and reliable.

How should management check this work?

Use a defined plan, carry out the controls, check performance and correct what does not work.

For sfarp, assign a competent owner to set the work controls, check their use and review them after change, failure or new evidence. HSE managing health and safety

A management-system label or certificate is not approval of a site limit, procedure or individual work decision. Apply GB and NI duties to the correct location.

What should each role learn and demonstrate?

Competence is evidenced by a person making and explaining the right decision for a realistic task, then responding to change.

A record should show the assessed scenario, observer, action taken and the trigger for reassessment; attendance alone is not proof of performance.

RoleLearningDemonstrated competenceAssessor and recordReassessment trigger
Operators and maintainersKnow the guard, isolation and stop points for the assigned task.Identify a bypassed interlock and refuse start until engineering release.Supervisor practical observationGuard defect or changed method
SupervisorsDistinguish a routine control check from a high-consequence SFARP decision.Escalate an unguarded nip and document interim protection without using an accident rate as proof.Engineering/H&S lead scenario reviewDefect, near miss or disputed control
Engineers and safety advisersEvaluate credible harm, good practice, options, benefit and sacrifice.Produce a peer-reviewed option record with commissioning criteria and uncertainty.Competent technical authorityDesign change or new guidance
Senior dutyholdersChallenge cost-based rejection and verify critical barrier funding.Approve an evidenced decision and sample the installed barrier after commissioning.Board assurance or independent reviewMajor-hazard change or barrier failure

What do the acronyms and key terms mean?

These terms explain the decisions on this page.

Acronym or termFull expansionMeaning here
SFARPSo far as reasonably practicableQualified duty to reduce risk unless further control is grossly disproportionate.
SFAIRPSo far as is reasonably practicableAlternative abbreviation closer to statutory wording.
ALARPAs low as reasonably practicableRisk-management expression for the same underlying test.
COMAHControl of Major Accident HazardsGB major-hazard regime with separate NI arrangements.
HSEHealth and Safety ExecutiveGB regulator and guidance publisher.
HSENIHealth and Safety Executive for Northern IrelandNI regulator.
AIArtificial intelligenceAutomated response tool; use the verified source and task facts.
COSHHControl of Substances Hazardous to HealthSubstance-control regime; GB and NI instruments differ.
DSEARDangerous Substances and Explosive Atmospheres RegulationsGB fire and explosion risk framework; NI has a separate instrument.
GBGreat BritainEngland, Scotland and Wales for the stated legal route.
HAZOPHazard and operability studyStructured process deviation review.
L111HSE publication L111Guidance code for the GB COMAH regulations.
LEVLocal exhaust ventilationCaptures airborne contamination near its source.
LOPALayers of protection analysisMethod for examining independent protective layers.
MEWPMobile elevating work platformPowered temporary access equipment.
MOCManagement of changeReview of risk before a planned change.
NINorthern IrelandSeparate legal and service route in this article.
PPEPersonal protective equipmentLast-line protection for remaining exposure.
PSSRPressure Systems Safety RegulationsGB pressure-systems duties; check NI instrument separately.
PUWERProvision and Use of Work Equipment RegulationsGB work-equipment regime; NI has a separate instrument.
RPERespiratory protective equipmentBreathing protection selected for residual exposure.
SILSafety integrity levelReliability designation for a safety instrumented function.
UKUnited KingdomGeographical umbrella; GB and NI law can differ.
ERICPDEliminate, Reduce, Isolate, Control, PPE, DisciplineTeaching framework for control selection and sustained supervision; not a statutory rule.