In short: A gas detection system on a UK refinery, fuel terminal or gas terminal is only as useful as the path from each detector to someone who can act on it, and on most sites that path is hard-wired for the original process units and missing almost everywhere else. Wireless links, including a site-owned private 4G/5G network, can bring tank farms, new plant and turnaround area monitors into the control room, provided they are kept for detection and alarm and the executive trips stay on the hard-wired safety system.
Key Takeaways
- Cable stops at the old plant boundary — Fixed gas detectors on original process units are hard-wired to the control room, while tank farms, jetties, new skids and outbuildings are often left with local beacons or none at all.
- Turnaround monitors alarm locally — Transportable area monitors placed around hot work and confined spaces during a turnaround usually sound on the spot or in a contractor cabin, and the central control room never sees the reading.
- Wireless for alarm, wire for trips — A private 4G/5G network can carry every detector reading and alarm to the control room, but executive actions such as an emergency shutdown should stay on the hard-wired safety instrumented system.
In a nutshell
What does a gas detection system on a UK hazardous site have to do?
A gas detection system on a major hazard site has to do three things in sequence: sense a flammable or toxic gas at a defined point, raise an alarm that people nearby will hear and see, and tell the control room which detector tripped, at what concentration and when. The first two happen at the detector. The third depends entirely on a communications path, and it is the third that decides whether the operator on shift can isolate a line, stop a transfer or hold a permit before a small release becomes a large one.
The legal framework is familiar to anyone running a COMAH or DSEAR site. The Dangerous Substances and Explosive Atmospheres Regulations 2002 require the risk assessment that tells you where flammable atmospheres can form, and the hazardous area classification that follows from it. BS EN 60079-29-2 covers the selection, installation, use and maintenance of flammable gas detectors, and BS EN 45544 does similar work for toxic gases such as hydrogen sulphide, which matters on refinery sour water units and at gas terminals handling sour gas. None of these standards insists the reading reaches a control room by cable. They insist that the detector is in the right place, works, and that the alarm leads to a response.
Why does gas detection coverage stop where the cable stops?
On many UK sites, fixed gas detection follows the history of the plant rather than the risk. The process units built decades ago at Stanlow, Fawley or Pembroke were designed with detectors cabled back to marshalling cabinets and then to the control room, and that wiring has been maintained and extended ever since. The ground outside those units tells a different story: tank farms covering tens of hectares, jetty heads, road and rail loading gantries, pump houses on the far side of a bund, vapour recovery units bolted on in the 2000s, and the newer skids for biofuel blending, hydrogen or carbon capture.
Cabling a single new detector into a Zone 1 or Zone 2 area means armoured cable, certified glands, tray or trench, a spare input on a controller and an inspection regime under BS EN 60079-17. Running that cable 400 metres across a tank farm, under a road and through a bund wall is a civils job, and it is usually the cable route rather than the detector that kills the business case. The result is a common place pattern: a detector with a local beacon and sounder that alarms perfectly well, and a control room that learns about it when someone radios in.
There is a second-order problem. Detectors on the periphery tend to be the ones nobody looks at between calibrations, so a fault, a poisoned catalytic bead or a drained battery can go unnoticed for weeks. A detector that reports its health continuously to a central system is a materially safer detector than one that reports only when it alarms.
What happens to gas detection during a turnaround?
A turnaround changes the gas picture overnight. Units are opened, purged and entered; hot work happens next to lines that were live a week earlier; and the site population can rise from a few hundred to a few thousand contractors. To cover this, sites bring in transportable area monitors such as the Dräger X-zone 5500 or Industrial Scientific Radius BZ1, placed around hot work, confined space entries and blinding points, alongside the personal detectors everyone already wears.
These units are good at alarming locally, and several can link to each other by their own short-range radio so that one detecting gas sets off its neighbours. Getting their readings to the central control room is another matter. Often they report to a laptop in a contractor cabin, or to a cloud portal over a public cellular modem, or to nobody at all. The control room operator, who holds the permits and can stop the job, is frequently the last person to know that an area monitor on the crude unit went into alarm at 14:20.
The same gap separates gas detection from the permit system. We covered the case for digital permit to work during turnarounds separately; the gas readings are the evidence that should sit next to each live permit, and they rarely do because the data has no route back.
Which wireless options exist for connecting gas detectors?
There are four broad approaches in use on UK sites today, each with real strengths and real limits.
Firstly, industrial wireless mesh using WirelessHART (IEC 62591) or ISA100.11a at 2.4 GHz. These are mature, low-power and well understood by instrument engineers, and several fixed detector makers offer wireless variants. However, mesh throughput is modest, update rates are typically several seconds, and each mesh still needs gateways cabled back to the control system, so on a large tank farm the cable problem shrinks rather than disappears.
Secondly, proprietary radio built into transportable area monitors. It is quick to deploy and works out of the box, but each manufacturer's radio speaks only to its own kit, and it rarely integrates with the site's distributed control system (DCS) without a separate gateway.
Thirdly, public cellular, usually LTE-M, used by connected personal detectors and some area monitors from makers such as Blackline Safety and MSA. This works well where the public signal is good. On the inside of a steel structure, at the far end of a jetty or in a valley site, it often is not, and the operator has no control over coverage, capacity or when the mobile network changes its configuration. When the UK networks switched off 3G in 2024 and 2025, any detector with a 3G-only modem had to be replaced.
Fourthly, a site-owned private 4G/5G network. Using an Ofcom shared access licence, typically in the 3.8 to 4.2 GHz band for 5G, a site can put its own coverage across tank farms, jetties and unit interiors, with its own SIMs, its own core and a direct link into the control room network. The same network can also carry backhaul for mesh gateways and area monitor hubs, which turns approach one and two into something the control room can see.
Where does a private 5G network fit, and where should it stay out?
A private network is a good fit for three jobs: (i) carrying alarm, fault and concentration readings from fixed detectors that were never cabled, (ii) bringing transportable area monitors into the control room for the duration of a turnaround, and (iii) giving the same site a single, managed path for the rest of the connected worker load, from the lone worker devices to ATEX-rated phones and tank gauging.
It is a poor fit, at least for now, as the transport for executive safety functions. Where a gas detector initiates an automatic shutdown, isolation valve closure or deluge, that loop is a safety instrumented function under IEC 61511, and BS EN 50402 sets the functional safety requirements for the fixed gas detection system that feeds it. The safety integrity level (SIL) claims for those functions are built on hard-wired, fail-safe signalling, and we would not advise any operator to move a SIL-rated trip onto a radio link to save cable. The sensible pattern is to keep trips hard-wired and use wireless to extend detection, alarm and health monitoring to every place the cable never reached.
There are practical trade-offs too. Many LTE-M devices will only attach to a network that runs an LTE carrier configured for Cat-M1, which a 5G-only private network will not provide, so the device list needs checking before the radio design is fixed. Battery-powered fixed detectors trade update rate against battery life, and a site has to decide whether a reading every 10 seconds or every 60 seconds meets its alarm response time. Radio equipment in a hazardous area must itself be certified for the zone, or sited outside it with the antenna placed accordingly.
How should a site decide whether wireless gas detection is right for it?
The case usually stands or falls on four questions, and each can be answered with a site drawing and a day on the ground.
- Where are the detectors that alarm only locally? Mark every fixed detector with a local beacon but no control room input, plus every location where transportable monitors go during a turnaround. On a typical refinery or large terminal this list is longer than the operations team expects.
- What would the cable cost? Price the cable route to the nearest marshalling point for the ten most remote detectors. If the civils exceed the cost of the detectors several times over, wireless is worth a serious look.
- What does the control room need to see? Alarm, fault, concentration and battery status into the DCS or a standalone alarm display, rationalised in line with EEMUA 191 so that a turnaround does not flood the operator with nuisance alarms.
- Which devices are on the site, and what radios do they use? Mesh, proprietary radio, LTE-M and 5G each need different things from the network, and a site with a mix needs gateways as well as coverage.
We design and run private 4G/5G networks for UK industrial sites, with radio planning based on measured coverage rather than a desktop prediction. For a COMAH or DSEAR site, our first step is a survey of the tank farm, jetty and unit interiors against the list of detectors that currently alarm only locally, followed by a short trial that carries one turnaround's area monitors into the control room. If you are planning a turnaround for 2027, the time to run that trial is before the scaffolders arrive.