NT fatigue code: safe staffing and recovery are more than a roster preference
Fatigue in a custodial workplace is not simply a worker feeling tired. It can affect judgement, communication, emotional regulation and the ability to respond safely. A roster may comply with an employment instrument while still requiring a separate assessment of work health and safety risks.
The Northern Territory has approved its Managing the Risk of Fatigue at Work Code of Practice. The Attorney-General’s instrument, dated 12 January 2026, appears in Government Gazette G5 of 12 March 2026, page 4, paragraph (a)(iii). The NT WorkSafe code page also expressly identifies it as approved under section 274 of the Work Health and Safety (National Uniform Legislation) Act 2011. This is local approval, not a national model waiting to be adopted.
Two risks must be addressed
The code separates preventing fatigue from preventing harm when someone is already fatigued. Both matter. An employer should assess what makes a shift exhausting and what could happen if attention, coordination or decision-making deteriorates.
In corrections, practical questions could include whether successive shifts allow real recovery, whether staffing makes breaks achievable, whether transport duties follow a demanding shift, and whether checks protect against consequential errors. These are applications of the code, not allegations about any particular facility. Sections 3.1–3.3, pages 14–21 of the full official code, cover hours, demands, environment, reporting and records.
Look beyond scheduled hours
Actual work can differ sharply from a published roster. Overtime, late handovers, call-outs, work questions during recovery periods and shift swaps need to be visible in the assessment. A quiet period also is not necessarily restorative if it requires sustained vigilance or passive monitoring.
The code recommends comparing planned and actual hours and examining vacancies, incidents, complaints and leave patterns. It warns against relying on a single source or a walk-through inspection. Workers and elected health and safety representatives need a genuine opportunity to explain what the figures mean. An empty fatigue-reporting register can reflect barriers to reporting rather than absence of risk. See sections 2.2 and 3.3, pages 12–13 and 18–21.
A worker’s agreement does not remove the employer’s duty
The code expressly rejects treating willingness to work extra hours, preferred shifts or additional pay as substitutes for risk management. Employment contracts, awards and enterprise agreements do not displace WHS duties. Workers also retain reasonable-care duties, including following reasonable safety instructions and notified policies. Those obligations operate together; responsibility cannot simply be shifted onto one side. See sections 5.2 and 5.4, pages 25 and 31–32.
The code offers numerical roster guidance, but it does not establish a universal statutory maximum shift. For example, page 26 recommends a 30-minute break every five hours at minimum and discusses recovery arrangements including at least 12 hours between shifts. More protection may be needed when demands are high. Any applicable award or agreement must be checked separately. Neither compliance with a suggested figure nor worker consent proves a roster is safe.
Change the system before relying on endurance
A useful response might remove non-essential work, improve relief coverage, reduce unnecessary task switching, introduce predictable rosters or provide suitable recovery facilities. Controls must fit the work. Night duties that cannot be eliminated still require risk reduction, including attention to the tasks performed, staffing, supervision and recovery.
Detection technology is not itself a control. It may help identify deteriorating alertness, but someone must act on the result. Likewise, caffeine cannot replace adequate sleep. The code discusses napping as a possible supplementary measure, with suitable facilities and time to recover from post-sleep grogginess, not as the sole routine answer to unsafe schedules. See sections 5.2–5.3, pages 24–31.
Respond fairly to reports and errors
When a fatigued worker makes an error, investigate contributing conditions rather than assuming carelessness. The code’s illustrative HR case study examines staffing, overtime and recovery before deciding how to respond. It is an example, not a rule that fatigue excuses misconduct or proves causation in every case.
Keep the reported event, alleged conduct, possible fatigue contribution, actual impact and proposed response separate. Check chronology and independent evidence. Protective action can be necessary before fault is resolved, but it should not be described as a completed finding. Monitoring data also needs careful handling: workers should understand its purpose, and fatigue information should not quietly become a different form of performance surveillance. See page 54 and sections 3.3 and 7, pages 20–21 and 37.
The next practical step
Request a review using actual hours, break access, work demands, recovery opportunities and relevant incident information. Record the controls, the person responsible, the completion date and the test of effectiveness. An approved code informs what is known about risks and controls; it does not automatically determine liability or compensation. The immediate question is whether the next roster can be worked safely in practice.
*General information, not advice on a particular dispute. Original commentary; no source tables, diagrams or substantial text reproduced.*

Comments