Fire and gas detection systems for hydrocarbon processing plants

Hydrocarbon processing facilities handle volatile substances under conditions of elevated temperature and pressure. Liquefied petroleum gas, naphtha, kerosene, diesel, and residual fuel oil can release flammable vapours at ambient conditions, while enclosed processing units concentrate both fuel and ignition sources in close proximity. A single undetected leak, a missed spark, or a delayed operator response can escalate into a fire, vapour cloud explosion, or catastrophic asset loss within seconds. This reality shapes the entire engineering philosophy of any modern refinery, gas plant, or petrochemical complex.

OrdaSintez Gaz LLP, an integrated oil and gas producer headquartered in the Kyzylorda Region of Kazakhstan, applies this same hazard-driven logic to its refining and marketing activities. The company produces aviation fuel, motor fuels, diesel, LPG, LNG, bitumen, mazut, petcoke, sulphur, and urea, and supplies buyers across Asia, Europe, and the Americas. Because its product slate spans highly flammable and pressurised materials, the firm has invested in fire and gas detection architectures that meet internationally recognised expectations. Its approach offers a useful reference point for operators in other regions, including the Australian downstream sector.

In Australia, downstream assets stretch from the Geelong refinery near Melbourne and Lytton in Brisbane to the Kwinana complex south of Perth and the Bulwer Island site in Gladstone. Operators there work under stringent state and federal oversight, and they also import a steady stream of refined products and petrochemicals. Regardless of origin, every cargo of crude, condensate, or finished fuel that enters an Australian port passes through facilities that rely on the same fire and gas detection principles used by OrdaSintez Gaz. Understanding those principles helps procurement teams, EPC contractors, and safety professionals make informed choices.

Foundational philosophy of fire and gas detection

A fire and gas system is more than an arrangement of sensors on a wall. It is a structured engineering discipline that begins with a quantified understanding of what could go wrong. Process hazard analyses, including HAZOP and what-if studies, identify credible release scenarios, while layers of protection analysis translate those scenarios into target frequencies for the safety instrumented function. Designers then choose detector coverage, response times, and availability figures that reduce the residual risk of each scenario to a tolerable level.

Three performance goals sit at the centre of every fire and gas design. The first is coverage, which means that any credible release or fire scenario has at least one detector within the appropriate distance and orientation. The second is speed, since flame propagation and vapour cloud movement often outpace human reaction. The third is availability, expressed through safety integrity levels that quantify the probability of the detection chain functioning on demand. Together, these targets create a measurable performance envelope for the system.

Local regulators and standards bodies reinforce the same philosophy. In Australia, Safe Work Australia publishes model work health and safety regulations, while Standards Australia maintains the AS 1851 suite for routine service of fire protection systems, AS 60079 for equipment in explosive atmospheres, and AS 1670 for fire detection and alarm systems. Offshore operators additionally fall under the National Offshore Petroleum Safety and Environmental Management Authority, which expects engineers to demonstrate that detection coverage aligns with the registered safety case for the facility.

Sensor technologies and field placement

The detector mix on a hydrocarbon unit reflects the failure modes that the hazard analysis identified. Catalytic bead sensors respond to combustible gases at the lower explosive limit, electrochemical cells measure toxic species such as hydrogen sulphide, and infrared point detectors identify hydrocarbon vapours in oxygen-rich atmospheres. Ultrasonic gas detectors, an Australian-developed technology refined for offshore platforms in Bass Strait and elsewhere, listen for the characteristic acoustic signature of a pressurised leak and are particularly valuable in open modules where convection could carry a gas cloud past conventional sensors before detection.

For flame detection, ultraviolet and single-band infrared units remain common, but multispectral infrared detectors now dominate new builds because they discriminate real fires from sunlight, hot exhaust, and reflective backgrounds. Open-path gas detectors project an infrared beam across a process aisle or around a storage tank, covering large volumes with a single device and providing earlier warning for buoyant gases that accumulate near ceilings. Heat detectors, including linear fibre-optic systems along pipelines and cable trays, complement the gas and flame coverage where smouldering or low-flame conditions are plausible.

Placement is as important as device selection. Engineers position point detectors near flanges, pump seals, compressor stations, tank vents, and loading arms where historical leak frequencies are highest. They also account for ventilation patterns, since a stiff cross-breeze in a coastal facility such as Geelong or Kwinana can push a plume away from a fixed sensor. Computational fluid dynamic modelling has become a routine tool to confirm detector layouts, especially for low-lying heavy gases and lighter-than-air releases that behave differently depending on the surrounding structures.

Integration with control, alarm, and shutdown systems

Detection is valuable only if it triggers a defined response. The detector output travels to a logic solver within a safety instrumented system, which evaluates the signal against a voting logic, typically one-out-of-two or two-out-of-three, to suppress nuisance trips while still achieving the required safety integrity level. From there, the logic solver initiates actions such as closing block valves, stopping pumps and compressors, de-energising electrical equipment, starting firewater pumps, opening deluge valves, and initiating ventilation isolation or emergency depressurisation.

Alarm management is a discipline of its own, and one where Australian operators have made notable progress. The EEMUA 191 framework, widely adopted in refineries around Melbourne and Brisbane, prioritises alarms so that operators receive only a small number of high-value notifications during an incident. Each fire and gas alarm should correspond to a documented operator action, an automatic control response, or both. Where alarms are merely informational, they belong in the control system trending view rather than on the operator console.

Mapping software links the fire and gas database to the three-dimensional plant model, which simplifies coverage reviews during brownfield modifications. When a new pipeline is added in a unit at Lytton or a turnaround project rearranges equipment in Kwinana, the mapping tool overlays the existing detector coverage on the new geometry and highlights any gaps. This integration reduces the risk that a small construction change will quietly disable a critical detection function.

Supply chain, procurement, and lifecycle considerations for end users

Downstream operators and procurement teams rarely buy detection hardware in isolation. They purchase a system that includes engineering design, certification, installation, commissioning, training, spare parts, and long-term support. For buyers in Australia, this often means evaluating vendors against both local standards and the requirements of remote or hazardous sites, where logistics add a layer of cost and complexity that a procurement plan in a port-adjacent facility may not face.

OrdaSintez Gaz markets crude, refined products, and related services to buyers in Asia, Europe, and the Americas, and its supply chain is therefore accustomed to coordinating with international counterparts. A fire and gas detection package, like a cargo of aviation fuel or diesel, moves through documentation, classification, and acceptance procedures that mirror those used for petroleum products. The same attention to specification, traceability, and after-sales support applies to detection systems that protect the assets handling those products.

Lifecycle cost is a deciding factor in many tenders. A detector that costs a little more at purchase but offers a longer calibration interval, lower false alarm rate, and easier proof testing can save significant operational expense over a twenty-year horizon. Buyers in Sydney, Adelaide, or Perth routinely weigh these trade-offs alongside the technical performance metrics. The most successful projects treat the detection system as a long-term partnership rather than a transactional purchase, with shared responsibility for upgrades, obsolescence management, and regulatory alignment.

Operational discipline and the human element

Even a well-designed system can drift out of specification if maintenance is neglected. Detectors lose sensitivity as optical windows cloud and as chemical cells age. Proof testing, calibration, and functional verification therefore follow documented intervals set out in AS 1851 and aligned with the safety case for the facility. A typical proof test introduces a known stimulus at the sensor and confirms that the logic solver and final elements respond within the expected time, generating the evidence needed to demonstrate continued compliance.

Human factors deserve equal attention. Operators must understand what each alarm means, what they should do in the first thirty seconds, and when to evacuate. Drills, simulator sessions, and competency assessments are standard practice in Australian refineries and at major LNG terminals near Gladstone. The same approach supports the multinational workforce at OrdaSintez Gaz, where staff from the Kyzylorda operations and the company's marketing offices in Asia and Europe must share a common vocabulary around fire and gas emergencies.

Incident learning closes the loop. Near-miss reports, alarm logs, and maintenance records feed back into the hazard analysis, prompting revisions to coverage, replacement of underperforming detector models, and updates to procedures. Facilities that treat fire and gas detection as a living programme rather than a one-off engineering exercise tend to maintain low incident rates over decades, regardless of whether the plant sits beside the Caspian lowlands, the Swan Coastal Plain, or Port Phillip Bay.