SITELIVE 
A duty without a limit
There is no workplace exposure standard for heat in Australia. The duty is the general one — eliminate or minimise the risk so far as is reasonably practicable — with Safe Work Australia's guide for managing the risks of working in heat setting out the process, and WorkSafe Queensland stating the position bluntly: heat stress indices are guides, not safe/unsafe limits. That is not regulatory reluctance so much as physics. ISO 7243:2017, the international wet bulb globe temperature standard, describes itself as a screening method for establishing the presence or absence of heat stress across a working day of up to eight hours, and directs users to the analytical method in ISO 7933 once the screening values are exceeded. WBGT characterises the environment. It does not characterise the worker standing in it, their work rate, their clothing or their acclimatisation state.
The United States is attempting to put numbers on the process instead of the physiology. OSHA's proposed Heat Injury and Illness Prevention standard, published in the Federal Register on 30 August 2024, would require a written prevention plan and defines two thresholds: an initial heat trigger at a heat index of 80°F (26.7°C) or a WBGT equal to the NIOSH Recommended Alert Limit, and a high heat trigger at 90°F (32.2°C) or a WBGT equal to the NIOSH Recommended Exposure Limit. At the initial trigger it would require suitably cool drinking water in sufficient quantity for one quart — roughly one litre — per employee per hour, break areas with shade or cooling, paid rest breaks where needed, and acclimatisation protocols for new and returning employees. At the high trigger, paid 15-minute rest breaks at least every two hours become mandatory, alongside a hazard alert and symptom observation via a buddy system or a supervisor watching no more than twenty people. The informal public hearing concluded on 2 July 2025 and post-hearing comments closed on 30 October 2025; it remains a proposal, not a rule, and Australian sites have no equivalent numeric obligation.
The finding that should change your induction
The CDC's MMWR review of OSHA general-duty-clause heat enforcement cases from 2012–2013 examined 20 cases, of which 13 involved a worker's death. Nine of those deaths occurred within the first three days of working on the job, and four occurred on the worker's first day. None of the employers had a complete heat illness prevention program: twelve had no program at all, seven managed water inadequately, 13 provided no shaded rest area, and exactly one used scheduled work-rest cycles. The illnesses and deaths occurred across a heat index range of 84.0–105.7°F (29.0–41.0°C), which is an ordinary Australian summer rather than an extreme event. Acclimatisation is the control that addresses this pattern directly and costs nothing but programme: NIOSH's schedule limits a new worker to no more than 20% of the usual duration in the heat on day one, increasing by no more than 20% each day, and returning workers to 50/60/80/100% across the first four days, with close supervision for 14 days. Acclimatisation also decays quickly, which is why the first shift back after a fortnight of leave, or the first hot week of the season, behaves like a first day.
What wearables can and cannot tell you
The obvious answer to individual variation is to measure the individual, and the market now sells wrist and chest devices reporting an estimated core body temperature. The underlying science is real and admirably honest about itself: the Kalman-filter model that estimates core temperature from sequential heart rate observations was validated across laboratory and field studies totalling more than 52,000 observations and showed an overall bias of −0.03 °C with a standard deviation of 0.32 °C, with 95% of estimates falling within ±0.63 °C. The authors' own framing is the one to quote back to a vendor — accurate enough to give a practical indication of thermal work strain, and not a replacement for direct measurement of core temperature, particularly given that the clinical reference methods themselves, oesophageal versus rectal, average limits of agreement of about ±0.58 °C. The practical consequence is that a wearable reading should trigger a human check, not an automated pass/fail gate that crews learn to ignore or work around. The same discipline applies to environmental numbers: a heat index taken from the nearest weather station is not the heat index in a trench, a plant room, a roof space or under vapour-barrier coveralls, and workers in direct sunlight can experience a heat index up to 15°F (8.3°C) above the reported figure.
What it means on site
Write the trigger table before summer and tie it to work rate and clothing, not the forecast alone. Ramp anyone new, returning or reassigned into hot work, and treat the ramp as a scheduling constraint the programme has to absorb — this is the most evidence-backed control available and it fails because it is inconvenient, not because it is unproven. Make rest genuine: shade with air movement, cool water within walking distance, and breaks paid so nobody trades a cool-down for output. Give supervisors a symptom script and a buddy pairing rather than a poster. Then log it — conditions at the workface, the trigger reached, the breaks taken, the acclimatisation status of every new starter and every returning one. In a jurisdiction with no numeric limit, what was reasonably practicable is decided afterwards, from whatever record you kept.