In short: The UK has five working refineries after the closure of INEOS Grangemouth. They run on networks largely designed in the 1990s, in environments classified for explosive atmospheres, with workforces under pressure and assets well past their original design life. Private 5G is becoming the foundation layer for the inspection, monitoring and connected-worker programmes that refineries need to operate safely and economically for the next twenty years.
Key Takeaways
- The UK's refining footprint is now critical national infrastructure with no spare capacity — five plants supply the bulk of UK demand for petrol, diesel, jet and petrochemical feedstock, and any unplanned downtime is felt nationally
- ATEX zoning is the dominant network constraint — the rules that govern equipment in Zone 0/1/2 areas drive every choice about radios, cameras, sensors and tablets, and most existing kit was certified a decade ago
- Drones, crawlers and connected workers replace the most dangerous tasks — flare-stack inspection, confined-space entry and shutdown turnarounds are exactly where the safety case for private 5G pays for itself
In a nutshell
A smaller UK refining sector that has to keep working
When INEOS confirmed the wind-down of Grangemouth refining operations, the UK lost the only refinery in Scotland and one of six on the mainland. What remains is a tight set of facilities that together meet most UK demand for road fuels, jet fuel and petrochemical feedstock:
- Stanlow (Essar Oil UK), Cheshire — the third-largest UK refinery by capacity, supplying around a sixth of UK road fuels
- Humber (Phillips 66), Lincolnshire — the largest single-site refinery in the UK, also producing the specialty needle coke used in EV battery anodes
- Fawley (ExxonMobil/Esso), Hampshire — the UK's largest integrated refinery and petrochemical site, supplying much of Southern England's jet fuel via the pipeline to Heathrow and Gatwick
- Pembroke (Valero Energy), West Wales — a deep-water coastal refinery with a major export book
- Lindsey (Prax Group), Lincolnshire — sitting next to the Humber complex, with shared infrastructure into the Killingholme terminal
Combined, these five carry the UK's fuel security in a way that wasn't quite the case ten years ago when there were nine refineries. The economic and energy-security case for keeping them operating reliably is now stronger than the case for capacity expansion.
That changes what matters operationally. The conversation at every UK refinery is no longer about new capex on production units; it's about extending the safe operating life of existing units, reducing turnaround duration, and getting more useful work out of fewer field operators. All three of those depend, increasingly, on the network.
The network most refineries are actually running on
Most UK downstream sites are still operating networks whose architecture was laid down in the 1990s and early 2000s. The control system uses proprietary fieldbuses (Foundation Fieldbus, Profibus PA, Modbus RTU) into the DCS — typically Honeywell Experion, Emerson DeltaV or Yokogawa CENTUM, each with its own air-gapped network for the integrity of the process control loop. That layer is fine; it does what it needs to do and isn't where the network problem sits.
The problem sits in the operations and maintenance layer above the DCS:
- Field operators carry radios on a UHF voice channel and paper checklists
- Permit-to-work and isolation control runs on a desktop application back in the central control room
- Mobile devices, where they exist, are intrinsically safe (IS) handhelds with limited data rates
- WiFi exists in the offices, the warehouse and a few specific units, but not as a continuous, refinery-wide layer
- Cameras for security and process monitoring run on separate cabled CCTV networks
- Drone inspection footage gets reviewed back in the office hours after the flight, because there's no way to stream it live from the unit
Layered on top of that are the vendor-specific radios for vibration monitoring, gas detection, acoustic leak detection, and ultrasonic thickness monitoring — each from a different vendor, each with its own basestation, each with its own coverage map. The total cost of running and maintaining this patchwork quietly exceeds what a single managed private network would cost, but it's distributed across enough budget lines that nobody owns the total.
ATEX is the framing for every conversation
The defining constraint on any refinery network is the ATEX/IECEx zoning of the site. Process units that handle hydrocarbons are classified into Zone 0 (continuous explosive atmosphere), Zone 1 (likely to occur in normal operation) and Zone 2 (not likely in normal operation, and if it does, only briefly). The rules governing electrical equipment in each zone are strict and inspected: certified-intrinsically-safe or explosion-proof enclosures, surface temperature limits, and grounding requirements that are non-negotiable.
That has three implications for private 5G:
- Radio equipment has to be ATEX-rated for the zones where it's deployed. Vendors including Nokia, Ericsson and several specialist integrators now offer Zone 2-rated small cells and Zone 1-rated handhelds. This is no longer a theoretical category — the certified kit exists
- Antenna and small-cell siting follow the zoning map. Most of the radio infrastructure can live in safe-area locations (above the unit, on lighting columns, on the unit's structural frame outside the cloud envelope) with coverage projected into the zone. Only the user devices and the gas-detection sensors need to be inside the zone
- The certification work is paperwork-heavy. This is normal for refinery capital projects. The point is that it's now possible to do the paperwork once for a private 5G network rather than re-doing it for every vendor-specific radio that gets added
The ATEX framing also makes private 5G's case easier than WiFi's. Industrial WiFi vendors don't have the same depth of ATEX-certified product lines, and the licensed-spectrum case for safety-critical applications is much stronger than the unlicensed-spectrum case.
The three use cases that pay back first
Refineries are conservative environments. Network projects need to show measurable, attributable benefits against existing operating costs — not just future-looking productivity claims. Three use cases consistently pass that test in UK downstream:
Autonomous inspection. Flare stacks, storage tanks, process columns and pipe racks all need regular inspection, and the existing methods are dangerous (rope access at height, confined-space entry into tanks) and expensive (scaffolding for column inspections, lost production for tank entries). Drones, magnetic crawlers and inspection robots can do most of this work, but only if the network can carry 4K video, thermal imaging and ultrasonic data back to the inspection engineer in real time. Private 5G is what makes that work at scale. Stanlow's recent inspection programme and the Net Zero Technology Centre's work in Aberdeen have both pointed at this exact pattern
Connected worker safety. Lone-worker monitoring, gas exposure tracking, real-time location, and biometric monitoring (heart rate, fall detection) for field operators are now standard expectations under HSE major-hazards regulation. The existing implementations use a mix of IS handhelds on UHF and proprietary radio meshes. A single managed network covering the whole site is dramatically simpler and cheaper to operate, and it removes the dead spots that voice-only systems have always had inside process units
Shutdown and turnaround coordination. A refinery turnaround involves thousands of contractors working concurrently across multiple units for four to six weeks, with permit-to-work, isolation, hot-work and confined-space approvals flowing continuously. The current state of the art is paper-and-iPad, which doesn't scale and which creates the conditions for the safety incidents that turnarounds are notorious for. A connected turnaround on private 5G — with permits issued and revoked in real time, isolation checks logged automatically, contractor location tracked, gas readings streamed — pays for the network in a single turnaround cycle
The wider energy-transition story
UK refineries aren't standing still on decarbonisation. Stanlow is the anchor site for HyNet North West, the cluster building hydrogen production and carbon capture infrastructure across the North West. Phillips 66 Humber is part of the East Coast Cluster, with carbon capture planned at the Lincolnshire petrochemical complex. Fawley has been investing in lower-carbon feedstocks and renewable diesel. Pembroke is connected into the Celtic Sea floating offshore wind plans and has been studied for green hydrogen production.
Every one of these transitions adds connected infrastructure to sites that were originally engineered as standalone process plants. CCS pipelines, hydrogen electrolysers, renewable feedstock terminals, and the monitoring required to verify them all generate data, and that data has to live on a network. A private 5G layer that's already covering the existing refinery footprint extends naturally to cover the new units as they're built. A bolt-on solution for each project ends up costing more and integrating poorly with the existing site systems.
What we'd build, and why
Refineries are exactly the kind of operational environment where a managed private network has a clearer business case than at most industrial sites. The safety improvements are real, measurable and reportable to the HSE. The operational savings — drone inspection vs scaffolding, connected turnarounds vs paper-based ones, single network vs vendor patchwork — are large enough to fund the network multiple times over. And the energy-transition projects add value on top of the legacy refining business case.
Aerix builds private 5G networks for environments where conventional networking has failed and where safety and uptime are non-negotiable. Refineries fit that envelope precisely. We work with ATEX-certified equipment vendors, design the radio plan around the zoning map, and manage the network as a service so the refinery's engineering team doesn't have to build a 5G capability from scratch.
If you're running a UK downstream site and you're looking at the next decade of inspection regimes, workforce expectations and energy-transition obligations on a network that was designed for a different era, the network is the bottleneck. Private 5G is what unblocks it.