In short: The UK's industrial decarbonisation clusters are being built more or less from scratch — new hydrogen production, new CO2 pipelines, new capture plants bolted onto existing works. Unlike the legacy refineries they sit beside, they have the rare luxury of designing the connectivity layer in from the start, and private 5G is the obvious candidate.
Key Takeaways
- A blank-page site is a once-only chance — hydrogen and carbon-capture plants are being designed now, which means the wireless layer can be specified as core infrastructure rather than retrofitted later through a live, hazardous facility.
- New hazards need new monitoring — hydrogen and dense-phase CO2 bring leak, dispersion and asphyxiation risks that demand continuous sensing and connected-worker safety across very large outdoor footprints.
- The cluster is a corridor, not a point — capture, compression, pipeline and storage span tens of kilometres, so the network has to stretch along a corridor and into remote injection sites, not just cover a single fence line.
In a nutshell
The UK's plan to decarbonise heavy industry does not rest on a single flagship project. It rests on a handful of geographic clusters — places where existing industrial demand, suitable geology and coastal access happen to line up — into which the government has channelled support through the Track-1 and Track-2 cluster sequencing programme. HyNet in the North West, anchored on Stanlow and the Liverpool Bay depleted gas fields. The East Coast Cluster, spanning Teesside and the Humber, with Net Zero Teesside as its centrepiece. Acorn in the North East of Scotland, built around the St Fergus gas terminal. Between them they are meant to capture and store millions of tonnes of CO2 a year and stand up the country's first at-scale hydrogen production.
What makes these projects unusual, from a connectivity point of view, is that they are largely being built new. And that changes everything about how the network underneath them should be planned. This post sets out why the clusters are a different proposition from the refineries we wrote about previously, what new operational hazards they introduce, and why the smart move is to treat private 5G as foundational infrastructure rather than a later upgrade.
The retrofit trap the refineries are stuck in
We have argued before that the UK's surviving downstream refineries — Stanlow, the Humber sites, Fawley — carry the country's fuel security on networks largely designed in the 1990s, and that retrofitting modern connectivity into a live, ATEX-zoned, continuously operating plant is slow, expensive and disruptive. Every new cable run is a permit-to-work. Every radio install in a hazardous area is a certification exercise. You are threading new infrastructure through a facility that cannot stop and was never laid out with it in mind.
The decarbonisation clusters are the rare case where that trap can be avoided. A new hydrogen production unit or a new carbon-capture plant is being engineered right now, on a drawing board, by people who get to decide where the cable trays go, where the instrument racks sit, and — if anyone thinks to ask the question early enough — where the radio infrastructure lives. The marginal cost of designing in a private 5G network during FEED (front-end engineering design) is a rounding error against the cost of retrofitting one into the same plant five years after commissioning. This is the blank-page advantage, and it expires the moment the concrete is poured.
Hydrogen and CO2 are not natural gas
It is tempting to assume that because these sites sit next to, or grow out of, existing gas infrastructure, the operational risks are familiar. They are not, and the differences bear directly on what the network has to do.
Hydrogen is a small, light, fast-diffusing molecule with a very wide flammability range and a low ignition energy. It leaks through joints that would hold methane, it burns with an almost invisible flame, and it disperses upward and outward quickly. Detecting a hydrogen leak is a harder sensing problem than detecting a natural-gas leak, and the case for continuous, networked gas detection — rather than periodic manual survey — is correspondingly stronger.
Carbon dioxide brings a different and less intuitive danger. CO2 transported in pipelines for storage is moved in dense phase, at high pressure, and a release behaves nothing like a gas leak: it is colder, heavier than air, and pools in low-lying ground, displacing oxygen. A CO2 release is principally an asphyxiation and dispersion hazard, and managing it safely across a large outdoor site means knowing, in real time, where your people are, where the gas is, and how it is moving. That is a connected-worker and sensor-fusion problem of exactly the kind a deterministic wireless network is built to support — wearables with location and gas exposure, fixed dispersion sensors, and the bandwidth to bring it all together fast enough to act on.
Continuous sensing over a very large footprint
Decarbonisation plant is sensor-dense in the way modern process plant always is — temperature, pressure, flow, vibration and composition instrumentation by the thousand — but it is also spread out. Capture trains, compression, drying, conditioning and metering occupy large outdoor areas, and the whole point of the exercise is to feed verified, continuous emissions and capture data into reporting systems that regulators and offtake contracts depend on. A carbon-capture plant that cannot continuously and credibly measure what it has captured is not doing its job; the measurement is the product, much as the batch record is the product in pharmaceutical manufacturing.
Carrying that volume of continuous instrumentation reliably across a large, partly hazardous outdoor site is precisely where private 5G earns its place. It offers the coverage to reach the far corners of a sprawling plant from a small number of radios, the capacity to handle massive IoT density, and — through network slicing — the ability to keep safety-critical leak detection and mustering data on a guaranteed path, separate from the general process-data and CCTV load. ATEX-certified equipment makes it deployable in the classified zones where the hazard, and therefore the need for monitoring, is greatest.
The cluster is a corridor
The single biggest way these projects differ from a conventional plant is geography. A refinery is, broadly, a point — a fenced site you can cover from within. A decarbonisation cluster is a corridor. CO2 is captured at one or more emitter sites, gathered, compressed, and piped — often tens of kilometres, sometimes offshore — to an injection point above a storage reservoir. HyNet's stored CO2 goes out under Liverpool Bay; Acorn's goes from St Fergus out to North Sea geology; the East Coast Cluster gathers across Teesside and the Humber estuary.
Connectivity that stops at the capture plant's fence line is therefore only half a network. Pipeline integrity monitoring, cathodic protection telemetry, valve and metering stations along the route, and remote injection wellheads all need to report, and some of them sit well away from any existing connectivity. A private network designed for the cluster has to think in terms of a route, with coverage extended along the corridor and resilient backhaul — including satellite where terrestrial links do not reach — to the remote ends. This is the same hybrid, corridor-shaped thinking that rail and pipeline operators already wrestle with, applied to brand-new infrastructure where, helpfully, it can be planned rather than patched.
Construction is its own connected operation
There is an interim point worth making, because it changes the timeline of the investment. These sites spend years as enormous construction projects before they ever process a molecule. Major-projects construction is itself increasingly a connected operation — plant tracking, geofenced exclusion zones around lifts and heavy equipment, digital progress capture, connected-worker safety across a site with thousands of contractors who do not yet know the layout. A private network stood up to serve construction is not wasted effort that has to be torn down at handover; it is the same infrastructure that will run the operating plant, brought into service early and earning its keep through the most hazardous, highest-headcount phase of the project's life.
That continuity — construction network becomes operations network — is only available to those who decide early. It is the clearest practical expression of the blank-page advantage, and it is gone the moment a site reaches for temporary, throwaway connectivity to get the builders working.
What we conclude
The UK's industrial decarbonisation programme is, among other things, the largest piece of greenfield process-plant construction the country has undertaken in a generation, concentrated in a few clusters with a couple of decades of operational life ahead of them. The hazards they manage — hydrogen's leakiness, dense-phase CO2's asphyxiation risk — argue for continuous, networked sensing and connected-worker safety. Their geography — a capture-to-storage corridor rather than a single site — argues for a network planned along a route. And their newness argues, most strongly of all, for designing that network in now.
We conclude that the clusters at HyNet, Teesside, the Humber and St Fergus have a chance the legacy refineries never had: to make the wireless layer a designed-in part of the plant rather than a retrofit fought through a live facility years later. Private 5G is the natural fit for a large, hazardous, sensor-dense, corridor-shaped operation — and the window to specify it as foundational infrastructure, rather than pay several times over to add it afterwards, is open now and will not stay open long.