Gas flaring is the controlled combustion of gas that an industrial facility cannot immediately recover, process, transport, or use. It is most closely associated with petroleum production, where gas emerges with crude oil, but it also occurs in gas-processing plants, refineries, petrochemical works, and other installations handling combustible gases. A flare system collects these streams and directs them to a burner designed to operate over a specified range of flow rates and gas compositions.
Flaring serves two distinct purposes: disposing of gas during ordinary operations and providing a means of handling exceptional releases during equipment failures, maintenance, or emergencies. This distinction is central to its regulation and engineering. A flare can be an essential part of process safety while persistent use of that same system can indicate a lack of gas-recovery infrastructure. Flaring is different from venting, which releases gas without deliberate combustion.
The practice converts much of the combustible gas into carbon dioxide and water, but can also release unburned hydrocarbons, soot, and other pollutants. Its significance therefore extends beyond the visible flame: flaring involves industrial safety, resource conservation, local air quality, and climate change. Understanding its effects requires distinguishing the volume sent to a flare from the fraction actually burned, and routine disposal from the capacity retained for emergencies.
Gas streams and operating purposes
The principal source of flared gas in oil production is associated petroleum gas. In an underground reservoir, hydrocarbons may occur as an oil phase containing dissolved gas, a separate gas phase, or both. As fluids rise to the surface and their pressure falls, dissolved gas comes out of solution. Surface separators then divide the production stream into gas, oil, and produced water. Oil can often be stored and transported relatively easily, whereas gas normally requires compression, treatment, and a continuous outlet.
Associated gas is not a uniform substance. It commonly contains methane, ethane, propane, butanes, and heavier hydrocarbons, together with varying amounts of carbon dioxide, nitrogen, water vapor, and sometimes hydrogen sulfide. Composition determines heating value, flame behavior, treatment requirements, and potential emissions. A gas rich in heavier hydrocarbons tends to form soot more readily than methane under poorly mixed conditions. Gas heavily diluted with noncombustible components may require supplementary fuel to sustain stable combustion.
Routine flaring generally means burning associated gas during normal oil production because no suitable use or disposal route has been provided. This can occur at a remote field without a gathering pipeline, where treatment equipment has not been installed, or where the available gas exceeds the capacity of a sales system. It is not synonymous with every predictable flare event: regulatory definitions vary, and maintenance, startup, and well-testing releases may be classified separately.
Nonroutine flaring includes releases during plant startup, shutdown, equipment maintenance, compressor trips, and interruptions to downstream gas acceptance. Emergency flaring handles gas discharged to prevent dangerous overpressure or to reduce the inventory of a threatened installation. These categories can overlap operationally. A compressor failure may initially require a safety response, but repeated failures can produce a substantial continuing volume of avoidable flaring.
Well testing provides another distinct case. Before permanent production facilities are available, a test may bring reservoir fluids to the surface to determine flow capacity and fluid properties. Separated gas may then be flared for a limited period. Refineries and chemical plants, by contrast, may send diverse process streams to a common flare network, including hydrogen-rich gas, light hydrocarbons, and mixed vapors. A landfill or wastewater facility may also burn surplus biogas, although its gas supply, operating context, and regulatory treatment differ from those of an oilfield flare.
Historical development and geographical distribution
Flaring developed alongside commercial oil production in the nineteenth century. Early oil industries primarily sought a transportable liquid fuel. Gas arriving with the oil had little value where pipelines, local demand, or reliable compression were absent. Operators burned it or released it, while the expansion of gas distribution and chemical manufacturing gradually created uses for what had previously been treated as a by-product.
The imbalance between oil and gas infrastructure persisted as petroleum production moved offshore and into remote regions. Crude oil could leave a field by tanker, but gas required a pipeline, a local consuming industry, reinjection facilities, or conversion into a transportable product. Developing these systems often required coordination among several fields, utilities, governments, and customers. A field could therefore begin producing oil years before a regional gas market was ready to receive its associated gas.
Large collection projects demonstrate how this relationship changed. Saudi Arabia’s Master Gas System, developed from the 1970s, gathered and processed associated gas for domestic fuel and petrochemical production rather than treating it chiefly as a disposal stream. Norway’s petroleum framework, established during the development of its offshore industry, subjected flaring to government authorization and encouraged integrated gas handling. Such examples did not eliminate all flaring; they changed the conditions under which routine disposal was accepted.
Global flaring is concentrated in a relatively small number of producing countries but occurs across many different operating environments. The World Bank estimated that approximately 148 billion cubic metres of gas were flared worldwide in 2023. Russia, Iran, Iraq, the United States, Venezuela, Algeria, Libya, and Nigeria were among the largest contributors. These totals should not be read as a simple ranking of operational efficiency: countries differ substantially in oil output, gas composition, field geography, and infrastructure.
Flaring intensity—the volume flared per unit of oil produced—provides a different perspective from absolute volume. A large producer can have a high total but relatively low intensity, while a smaller producer can flare a large share of its associated gas. Rapid production growth can also temporarily outpace gathering systems, as occurred in some North American shale-oil regions during the 2010s. Elsewhere, persistent flaring reflects long-standing constraints on processing, investment, electricity supply, or access to gas markets.
Historical comparisons have measurement limitations. Older records frequently depended on operator reports with incomplete coverage, whereas satellite observations now provide a more consistent view of visible flare activity. Changes in reported totals can therefore reflect both real operating changes and improvements in observation.
Flare-system architecture and equipment
A flare is the terminal component of a collection and disposal system rather than simply a pipe with a flame. The system begins at equipment that may discharge gas: pressure-relief valves, depressurization valves, process vents, and selected control outlets. Branch lines connect these sources to one or more flare headers. Their arrangement must accommodate the expected fluids, pressures, temperatures, and simultaneous release scenarios without imposing excessive backpressure on protected equipment.
Before gas reaches the burner, it commonly passes through a knockout drum. This vessel removes entrained liquid droplets and collects liquid slugs. Liquid carryover is hazardous because it can produce burning droplets, unstable flames, and unexpectedly high thermal radiation. Header routing, drainage, and liquid-level management are consequently important parts of flare design. Separate flare networks may be needed for streams that are incompatible, unusually cold, corrosive, or at very different pressures.
The familiar elevated flare uses a stack to place the flame away from personnel and equipment. Its burner, or flare tip, distributes the gas and controls the region where fuel meets air. One or more pilot burners provide an ignition source; ignition equipment and flame detectors support reliable operation. A small purge flow may prevent air entering the stack. Seals or specialized tip arrangements can reduce purge requirements and help prevent a combustible mixture from forming inside the system.
Assisted flares introduce steam, air, or another medium to improve mixing. Steam-assisted designs are common in refineries, where hydrocarbon-rich gases can otherwise produce smoke. Air-assisted designs use blowers and burner geometry to bring combustion air into contact with the gas. Pressure-assisted or sonic designs use the momentum of the discharged gas itself to promote entrainment. These approaches are not interchangeable: their suitability depends on gas pressure, composition, required operating range, utility availability, and noise constraints.
Ground flares offer alternative arrangements. Multi-point ground flares distribute gas among many burners, often in stages, while enclosed flares place combustion within a surrounding structure. Such systems can reduce visible flame and limit some external radiation, but require appropriate air supply and careful management of the combustion space. “Enclosed” does not mean that all emissions are captured or that the device operates like a fully controlled furnace.
A flare-gas recovery unit can be connected upstream of the final burner to recover low-pressure gas during ordinary operation. Compressors return it to a fuel-gas or processing system, while a suitable arrangement preserves the emergency disposal route. The recovery system must not create an obstruction or pressure condition that compromises the flare’s protective function. Its design therefore integrates chemical engineering, machinery reliability, and the hydraulic behavior of the entire network.
Combustion chemistry and performance
For pure methane undergoing complete combustion, the ideal reaction is
This equation describes the chemical endpoint, not a guarantee of actual flare performance. Most industrial flares are nonpremixed flames: gas leaves the burner and reacts as it mixes with atmospheric oxygen. The effectiveness of that mixing, the time available for reaction, and the temperature of the reacting gases determine how closely the real system approaches complete combustion.
An idealized calculation illustrates the scale of the conversion. Burning one kilogram of methane completely produces approximately 2.75 kilograms of carbon dioxide, because the molecular masses of methane and carbon dioxide are about 16 and 44 respectively. At 0 °C and one atmosphere, one cubic metre of pure methane produces about 1.96 kilograms of carbon dioxide. Commercial gas volumes are reported under several different reference conditions, however, so an emissions calculation must state its temperature and pressure basis.
Two performance terms require careful distinction. Destruction efficiency measures how much of a specified incoming compound, such as methane, is destroyed. Combustion efficiency describes the extent to which combustible material is converted into fully oxidized products. A flare may destroy methane while forming carbon monoxide or soot rather than converting all of its carbon to carbon dioxide. Numerical efficiencies are therefore meaningful only when the measured compounds and calculation method are specified.
Soot forms through complex reactions in fuel-rich parts of the flame, especially when the gas contains heavier hydrocarbons. Steam or air assistance can reduce smoke by improving mixing, but excessive assistance can dilute and cool a small gas stream enough to impair combustion. Low heating value, strong crosswinds, low discharge momentum, and gas-flow fluctuations can likewise produce poor performance. At high discharge velocities, flame lift-off and instability become concerns. No single burner arrangement gives identical efficiency across all these conditions.
An efficiency of 98 percent has often been used in emissions inventories or regulatory contexts for adequately operated flares, but it is not a universal measured property. A 2022 aircraft-based study of flares in three major United States oil-producing basins reported an effective methane destruction efficiency of about 91.1 percent when inefficient and unlit flares were included. That result concerned a particular field population and operating period, rather than every flare technology.
The distinction between a lit flare’s performance and a fleet-wide effective efficiency is especially important. A small number of unlit flares carrying large gas flows can contribute disproportionately to methane emissions. Conversely, counting visible flames without measuring throughput gives little information about their relative contribution. Performance assessment must connect ignition status, gas composition, flow, and combustion behavior.
Pressure relief, emergency operation, and safety
Flare systems provide a controlled destination for releases that would otherwise discharge combustible or toxic material near equipment and people. A pressure-relief valve may open when a vessel exceeds its allowable operating conditions. A depressurization system may deliberately discharge gas to reduce pressure and hazardous inventory after a fire or major equipment failure. These are related but distinct functions: relief limits pressure, whereas depressurization can reduce the consequences of an escalating incident.
Sizing requires analysis of credible release scenarios. Examples include loss of cooling, blocked process outlets, external fire, utility failures, and combinations of equipment trips caused by a common event. The design load is not necessarily the sum of every individual valve’s maximum capacity. It depends on which releases can occur together and how the process responds over time. Conversely, assuming that only one valve can discharge may substantially underestimate the load during a plant-wide utility failure.
Backpressure is a central constraint. Flow through branch lines, headers, drums, seals, and the flare tip raises the pressure against which a relief valve must discharge. Excessive backpressure can reduce capacity or affect valve operation. Network calculations therefore use fluid properties, pressure losses, and transient conditions to verify that the system remains effective. Very cold releases can also create material and operating problems, including brittle-fracture risks and ice or hydrate formation.
At the flare itself, major hazards include thermal radiation, flame impingement, noise, and dispersion of unburned gas. Elevated stacks and exclusion distances are selected partly to keep radiation at occupied locations within design criteria. Wind can tilt the flame and move the plume toward structures or work areas. An unignited release may form a flammable cloud or expose nearby areas to toxic constituents, making pilot reliability and rapid detection important even when the flare is normally used only intermittently.
Air ingress presents a different hazard. If oxygen enters a flare header and mixes with fuel, combustion may propagate into equipment intended to contain gas rather than flame. Purging, seals, appropriate operating pressures, and isolation practices reduce this risk. These controls must remain effective at low flow as well as during major releases. A large emergency capacity alone does not ensure safe behavior during prolonged standby operation.
Industry documents such as API Standard 521 and ISO 23251 address pressure-relieving and depressuring systems, while API Standard 537 addresses flare details. Their application depends on the installation and governing requirements. Instrumented functions may be implemented through a safety instrumented system, with routine monitoring handled by plant controls. Because a flare is a protective layer, eliminating routine flaring does not generally mean removing emergency flaring capability.
Measurement, satellite observation, and emissions accounting
A complete account of flaring requires several quantities: gas volume, composition, heating value, operating duration, ignition status, and combustion performance. No single instrument measures all of them. Flow measurements establish how much gas enters the system, gas analyses describe what it contains, and observations of the flame or plume indicate what happens during combustion. Differences between these information streams are a major source of measurement uncertainty.
Direct metering is difficult because flare headers operate over exceptionally wide flow ranges. A line may carry a small purge flow for long periods and then experience a large emergency release. Low pressure, changing composition, wet gas, and disturbed flow profiles can complicate measurement. Ultrasonic and other flow-meter technologies are used, but installation conditions, appropriate calibration, and treatment of low-flow performance remain important. A meter accurate during a large release may provide much less certainty about routine low-volume activity.
Composition can be determined by periodic sampling or online analysis. These measurements allow conversion of volume into mass, energy, and potential emissions. For example, the carbon dioxide produced by complete oxidation depends on the number of carbon atoms in the incoming hydrocarbons, not simply the volume of gas. Sulfur content is needed to estimate sulfur dioxide formation. Where several process streams share a header, composition may change rapidly during an upset.
Remote sensing provides independent observations beyond the facility boundary. Satellite instruments detect the intense infrared radiation of high-temperature flare flames, particularly at night. The Visible Infrared Imaging Radiometer Suite, or VIIRS, has been important in global flare monitoring. Researchers relate detected radiant output to gas volumes using models and calibration datasets. Such estimates make comparisons possible in regions where operator reporting is incomplete or inconsistent.
Satellite flare observations nevertheless have limits. Clouds obstruct the view, small or intermittent flames may be missed, and several nearby sources can be difficult to separate. Crucially, an unlit flare does not produce the same high-temperature signal as a burning one. A satellite-derived flaring volume is therefore not a direct inventory of all gas released through flare equipment. Methane-sensitive instruments, aircraft surveys, ground-based measurements, and optical flame monitoring provide complementary information.
Emissions inventories combine these data with emission factors or measured efficiencies. A simplified methane estimate is
where is methane destruction efficiency. The equation is straightforward, but its inputs may be uncertain and correlated. Applying a high assumed efficiency to an unlit flare can severely underestimate emissions. Reliable accounting therefore separates measured quantities from assumptions and specifies whether the reported result describes an individual event, an installation, or a regional fleet.
Environmental effects and human exposure
Flaring affects the atmosphere through both its intended combustion products and its incomplete-combustion emissions. Carbon dioxide is the main long-lived greenhouse gas produced when hydrocarbon gas burns completely. Methane that escapes combustion adds a different climate influence: it is much more potent per unit mass over short time horizons, although it persists for a shorter time than carbon dioxide. The comparison depends on the chosen assessment period and atmospheric accounting method.
For a gas stream that would otherwise be vented, effective combustion generally reduces near-term warming by converting methane into carbon dioxide. This does not make flaring equivalent to preventing the release or recovering the gas. The relevant comparison might instead be reinjection, productive use, reduced production, or another feasible operating arrangement. A climate assessment must identify that alternative and include leakage, processing energy, transport, and any displaced fuel use.
Incomplete combustion can produce black carbon, carbon monoxide, and unburned hydrocarbons. Black carbon absorbs sunlight and can contribute to warming; deposition on snow and ice can also reduce surface reflectivity. Its production depends strongly on fuel composition and combustion conditions, so flared volume alone is an imperfect predictor. A visibly smoky flare indicates soot formation, but a clear-looking flame does not demonstrate the absence of methane or other emissions.
High-temperature combustion can generate nitrogen oxides, while sulfur-containing gas can produce sulfur dioxide. Unburned constituents may include volatile organic compounds, depending on the source stream. These emissions can contribute to local and regional air-quality problems through direct exposure and atmospheric chemistry. The resulting concentrations depend on stack height, plume rise, wind, atmospheric stability, terrain, and the location of nearby populations.
Flares also create light, noise, and thermal radiation. Large releases may be audible over substantial distances, and continuous nighttime flames can affect the surrounding visual environment. Local exposure is not determined by annual volume alone: a short, high-flow event and continuous low-flow operation can create different concentration patterns. Offshore platforms, remote onshore fields, and facilities near settlements consequently present different exposure settings.
Research on public health near oil and gas operations has examined respiratory outcomes, pregnancy outcomes, and other endpoints. Attributing effects specifically to flaring is difficult because nearby populations may also encounter traffic emissions, dust, equipment exhaust, leaks, and other industrial sources. Epidemiological associations therefore require careful interpretation alongside exposure measurements and information about confounding factors. The composition of the gas and the actual operating condition of the flare are essential to any site-specific assessment.
Gas recovery, utilization, and reinjection
Reducing flaring begins with identifying why gas has no available destination. A field without a gathering line presents a different problem from a refinery experiencing frequent compressor trips. Solutions can involve new infrastructure, changes in operating practice, greater equipment reliability, or coordinated development of several gas sources. The technically possible option is not always commercially viable, especially where flows are small, variable, or expected to decline rapidly.
Gathering and processing provide the conventional route. Pipelines collect associated gas, compressors raise its pressure, and treatment removes water, acid gases, or other constituents that would violate sales specifications. Recoverable liquids may be separated as additional products. The treated gas can then supply households, industry, or power generation. The full chain must be available: constructing a gathering pipe without dependable processing capacity or a buyer may merely move the bottleneck downstream.
Reinjection returns gas underground rather than selling or burning it. It can maintain reservoir pressure, store gas for possible later recovery, or support enhanced oil recovery. Feasibility depends on reservoir characteristics, injection pressure, gas compatibility, well integrity, and compressor capacity. Reinjected hydrocarbon gas should not be confused with carbon capture and storage, which concerns managing carbon dioxide; the substances, objectives, and accounting boundaries differ.
Local power generation can use associated gas in an internal combustion engine or gas turbine. Electricity may serve field equipment, nearby users, or an electrical grid, where a connection exists. Combined heat and power can recover useful heat as well. These options require sufficiently consistent gas quality and reliable demand. Engines may need gas conditioning, and maintenance arrangements must suit remote operating conditions.
Transportable gas products offer alternatives where pipelines are impractical. Compressed natural gas can be moved in pressure vessels, while small-scale liquefied natural gas plants reduce volume through cooling. Gas-to-liquids processes convert gas into liquid hydrocarbons, and chemical plants can use it to manufacture methanol or other products. Each introduces capital costs, energy consumption, operational complexity, and minimum economic scale. “Modular” equipment reduces some deployment barriers but does not remove the need for a stable supply chain and a market.
Facility-level flare-gas recovery is particularly useful where a plant already has a fuel-gas network. Improved compressor availability, adequate buffer capacity, and better coordination with downstream users can also reduce nonroutine events. However, recovered gas must be assessed across its eventual use: sale and combustion still produce carbon dioxide, while additional transport can introduce leakage. Projects achieve different environmental outcomes depending on whether they displace other fuels, avoid methane release, or enable additional hydrocarbon production.
Regulation, economics, and implementation
The economics of flaring reflect the separation between oil revenue and gas-handling costs. An operator may be able to sell oil immediately while a gas project requires years of planning and substantial investment. Where gas prices are low or access to a pipeline is restricted, the private return from capture may be insufficient even when the gas has a wider social value. Rules governing field development therefore influence whether gas handling is built before oil production begins.
Regulatory approaches include prohibitions on routine flaring, time-limited permits, volume limits, performance requirements, fees, and obligations to submit gas-utilization plans. Some systems distinguish emergency releases from other categories while requiring all significant events to be recorded. Effective implementation depends on how categories are defined: a broadly worded exemption for maintenance or operational necessity can encompass repeated releases that another jurisdiction would treat as avoidable.
Flaring and venting need to be regulated together. A rule that penalizes the visible flame but does not address unburned releases can create a misleading incentive to substitute venting for combustion. Similarly, a reduction in satellite-observed flare activity does not by itself establish that gas capture has increased. Reporting systems must follow the gas through production, use, reinjection, flaring, and venting, preferably using consistent measurement boundaries.
The World Bank launched the Zero Routine Flaring by 2030 initiative in 2015. Participating governments, companies, and development institutions commit to ending routine flaring within the initiative’s framework. The commitment distinguishes routine disposal during normal oil production from safety flaring and other nonroutine events. It is a voluntary initiative rather than a universally binding legal prohibition, and progress depends on national implementation, investment, and reporting.
Policy also addresses combustion quality. The United States Environmental Protection Agency’s 2015 refinery-sector rule included strengthened flare operating and monitoring provisions, reflecting the fact that gas reaching a flare is not necessarily destroyed at a fixed efficiency. More generally, regulation may specify flame presence, heating value, operating parameters, or monitoring obligations rather than relying solely on an assumed destruction percentage.
Implementation difficulties are often institutional as well as technical. Different entities may own the oilfield, associated gas, pipeline, processing plant, and electricity network. Uncertain payment, delayed permits, limited financing, and unstable demand can prevent a technically feasible project from proceeding. Shared infrastructure introduces additional questions about capacity allocation and who bears the cost when production declines. Assessing progress therefore requires absolute flaring volumes, intensity, event frequency, gas disposition, and the quality of the underlying measurements—not merely the number of recovery projects announced.