In short: The UK energy industry has bought the lone worker safety stack, man-down wearables, connected gas detectors, check-in apps, but most of it assumes a mobile signal that tank farms, terminals, compressor stations and substations frequently do not have, especially inside steel structures and at remote perimeters. Lone working is a coverage problem before it is a device problem, and a private 5G network turns a best-effort duty of care into an engineered one: guaranteed signal, real-time location, and mustering that works where GPS and public networks fail.
Key Takeaways
- The safety stack assumes coverage the estate doesn't have — man-down alarms, connected gas detectors and check-in apps all fail silently in the dead zones of tank farms, terminal buildings and remote compounds, which is exactly where lone workers go.
- Lone working is growing as the estate ages — decommissioning, late-life assets and leaner operations mean more technicians visiting more remote sites alone, while HSE guidance makes clear the employer must provide a reliable means of raising the alarm.
- Private 5G makes the duty of care engineerable — one licensed network across a site delivers guaranteed coverage, sub-metre positioning inside steel structures where GPS fails, and real-time mustering, with ATEX-rated devices available for classified zones.
In a nutshell
Why is lone worker safety suddenly on the agenda?
Because the shape of the UK energy workforce has changed underneath the safety systems that were designed for a fuller estate. The British Safety Council's working estimate is that up to eight million people in the UK, around a fifth of the workforce, do some form of lone working, and the energy sector's share is growing for structural reasons: late-life assets run with leaner crews, the North Sea's long decommissioning tail sends small teams to quiet installations, and onshore the pattern is one technician in a van visiting compressor stations, block valve sites, tank farms, substations and wind sites that used to be permanently staffed.
The law has not changed, but its application has sharpened. The Health and Safety at Work Act and the Management Regulations require employers to assess and control the risks of lone working, and the HSE's guidance (INDG73, "Protecting lone workers") is specific on the point that matters here: a lone worker must have a reliable means of raising the alarm, and the employer must have a reliable means of knowing something has gone wrong. "Reliable" is doing a lot of work in that sentence. A man-down alarm that cannot get a signal out of the compressor house is not a control measure; it is the appearance of one, and it is the sort of gap that looks very bad indeed in an incident investigation.
Most operators know this, which is why the sector has invested heavily in the device layer. The uncomfortable truth is that the device layer is the solved half of the problem.
Where does the current lone worker stack fail?
Walk the estate with a coverage map and the pattern is consistent. Public mobile networks are built for population, not for industry: the terminal at the end of a peninsula, the tank farm behind earthen bunds, the wellsite compound in an agricultural valley and the cable landing station on a cliff top all sit at the ragged edge of commercial coverage, because there is no consumer revenue there. Ofcom's coverage obligations have improved the picture on A-roads and in villages; they were never designed to reach the inside of a gas terminal.
Then the structures themselves take over. Energy infrastructure is a radio-hostile environment by construction: steel tanks, dense pipework, concrete bunds, plated buildings and switchrooms behave as shields. A worker who is fine at the gate loses signal three rows into the tank farm and is entirely dark inside the pump house, and GPS, which every lone worker app leans on for location, degrades or fails in the same places, under steel, inside buildings, between tall structures. The result is a stack of well-intentioned kit with a hole in the middle: the wearable that alarms locally but cannot phone home, the check-in app that shows a worker's last position as the car park where they parked two hours ago, the connected gas detector, of the kind operators are deploying for continuous methane monitoring, that buffers its readings until it wanders back into coverage.
Some operators respond with process: buddy systems, timed check-in calls, radios. These work, and we would not argue anyone out of them, but each carries a cost. Buddying doubles the labour for routine tasks; timed check-ins interrupt work and generate false escalations; analogue radio raises an alarm but tells the control room nothing about where or what. The trade-off operators are actually making, often without stating it, is between paying that recurring process cost forever and fixing the coverage once.
How does private 5G change the safety case?
A private 5G network inverts the starting assumption: instead of hoping the public network reaches the site, the operator builds guaranteed coverage across it, in licensed shared access spectrum, engineered to include the bunds, the pump houses and the perimeter rather than stopping at the office. On that foundation, the lone worker stack starts doing what its brochures promised.
Firstly, the alarm path becomes reliable, and provably so: a man-down or panic activation reaches the control room from anywhere on site, every time, and the network's coverage can be surveyed, documented and presented to the regulator as an engineered control rather than an assumption. Secondly, location works where GPS does not. Cellular positioning within a private network, augmented by the density of small cells, delivers useful accuracy inside steel structures and between tanks, which transforms both routine assurance ("where is the technician who has not moved in forty minutes?") and emergency response, real-time mustering during an incident instead of a clipboard at the assembly point. Thirdly, the same network carries the rest of the connected worker load: wearable gas detection streaming live rather than buffering, video from a helmet camera when a remote expert needs eyes on a valve, and the site's fixed sensors besides. ATEX-certified 5G devices and routers now exist for classified zones, a genuine change from a few years ago when the certified-device catalogue was the honest blocker.
The candid limits: a private network covers the site, not the drive between sites, so the van on the A9 still rides on public networks and a sensible deployment uses both, private on site, public in transit, with devices that move seamlessly between them. And for a single small compound visited monthly, a satellite-messenger and a disciplined check-in procedure may remain the proportionate answer. The private network case is strongest where sites are large, structurally complex, hazardous, and visited often, which describes refineries and terminals exactly.
What does this look like across the UK energy estate?
The pattern generalises across the sector's site types. Gas terminals such as those at St Fergus, Bacton and Easington combine scale, steel and classified zones, the perfect storm for both public coverage and GPS. Tank farms and fuel terminals add the bund problem: earthworks designed to contain a spill also contain radio. Compressor and block valve stations along the transmission network are small but numerous and remote, the classic one-technician visit. Offshore, wind farm operations have the same lone-and-remote profile with a boat ride attached. And the new build, the hydrogen and carbon capture clusters, has the opportunity the old estate never had: to design the safety network in from day one rather than retrofitting it, a point we have made about the net-zero clusters before.
It is also worth noting what the network does for the safety case beyond lone working: the same coverage carries permit-to-work verification at the job front, drone inspection video that keeps people off ladders, and the process telemetry that the wider energy sector is already moving to wireless. As with most private 5G deployments, no single application carries the network; the safety case plus two operational ones usually does, and the safety case has a way of settling boardroom arguments that pure efficiency cases do not.
Where should an operator start?
Start by testing the assumption your current controls rest on: commission a coverage survey of your highest-risk sites, all four public networks, outdoors and inside the structures where people actually work, and lay it over your lone working risk assessments. Where the two disagree, and they will, you have a documented gap between the duty of care and the infrastructure it depends on. That document tends to move budgets.
We design, build and manage private 5G networks for energy and industrial sites, remote, steel-heavy, hazardous, and sized for the UK's mid-tier operators rather than the supermajors. If your lone worker devices are only as good as the signal in your worst pump house, talk to us about engineering the coverage they assume.