Every system in this series earns its keep by running. Separators separate, compressors compress, meters count money. This one is different. The relief and flare system is designed and maintained for credible demanding events, with the hope that its full emergency capacity will never be needed. It is the engineering of the worst day.

Start with one of the industry’s most misread images: a flare can serve safety, depressuring, start-up, shutdown, and other defined duties. A small visible flame may be a pilot or a low flare load; a large flame does not, by itself, explain the event. The operating objective is to minimise avoidable flaring while preserving the disposal capacity the facility needs.

Safety is a ladder, not a wall

Overpressure protection uses layers: process control, alarms and operator response, protective shutdowns, and pressure-relief devices where required. Their independence, reliability, and limitations matter more than the number of layers drawn on a diagram. A conventional spring-loaded relief valve can respond directly to pressure without an external power supply, but it is not infallible. Correct selection, installation, inspection, and testing are essential. It is one engineered layer within the complete protection strategy.

Sizing the worst day

Relief loads are calculated from credible scenarios. Blocked discharge is one: if an outlet closes, the assessment asks whether upstream pressure or continued energy input can overpressure the equipment, and what protection is needed. Utility failures, control failures, tube rupture, thermal expansion, and other scenarios may also need evaluation, depending on the system.

External fire can be a major case. Heat input may generate vapour and weaken exposed equipment; the assessment must consider more than valve capacity alone. For a shared flare system, credible simultaneous loads and common-cause events are also examined. Fire is important, but it does not automatically govern every relief valve or the entire flare header.

And a lesson written into industry practice with expensive ink: every modification must be re-examined against the relief case. A pump upgrade, a rerouted line, a new operating mode: each can quietly raise the load on a valve sized decades ago for a different plant. A change analysed only for normal flow is a change half analysed.

One pipe network for the worst moments

Designated hydrocarbon relief and depressuring streams commonly enter a flare header. Other services may require a different safe disposal route. The header is checked for credible combined loads, backpressure, drainage, and the behaviour of the discharged fluids. Purge and sealing arrangements help limit air ingress; their design and monitoring are essential to avoiding a flammable mixture in the system.

A knockout drum removes entrained liquid before the flare tip. Liquid carryover can impair combustion and, in severe cases, create burning-liquid fallout. Ignition, flame monitoring, and re-ignition arrangements are specified for the flare design. An unlit flare releasing combustible or toxic material is a serious failure mode, so availability depends on the complete system, rather than the pilot alone.

The flame's future

The direction of travel is to reduce routine and avoidable flaring while retaining necessary safety functions. Indonesia’s Minister of Energy and Mineral Resources Regulation No. 17 of 2021 provides a framework for flare-gas management, with distinctions between activities and flaring categories. Its actual provisions and the facility’s approvals govern. A one-line summary cannot be applied to every plant. Recovery systems can return suitable gas to useful service where technically and economically feasible. A quieter flare can reflect good operation, but it is not, on its own, proof that a facility is safe or compliant.

Nine episodes, one machine

This closes the season. The map; the separator's quiet gravity; a platform standing in the sea; compressors and the liquid that kills them; ice that is not ice; poison turned into yellow product; the oil train's argument with water and salt; the water plant that outlasts the oil; and the system built for the day everything else fails. Across all nine, the same four variables did nearly all the work: pressure, temperature, time, and level. Every vessel in this series is one more way of arranging those four in steel.

I’m a practising oil & gas facilities engineer. The playlist runs in order from the map to this flame, and it is built to be watched that way. Corrections are the most welcome comments of all.

Further reading

• HSE: relief and vent systems
• JDIH ESDM: Regulation 17/2021 on flare-gas management