The answer to Allen Hoffman's question, "Why does anyone flare anything =ADanymore?" is that flares are part of safety systems (C&EN, April 27, page 2). Hoffman was referring to an article about oil refiners (C&EN, Jan. 26, page 27).
When the quantity and/or rate of flammable and/or toxic materials is such that it can't be handled by normal processing, it's sent to a remote location flare to be safely burned.
Any sudden process stream surge or diversion can potentially overwhelm or pressurize a scrubber or absorber and reduce its efficiency because the additional material is too great to contain as a result of high pressure, reactivity, concentration, and so on. What's more, capturing such a stream into a larger, lower-pressure vessel risks explosion and is thus less desirable than immediately flaring it. Capture is more typically used if a thermal oxidizer or other destruction device is present to safely store the surge until it can be =ADconsumed.
Consider the example of polymerizing butadiene into rubber. If the reaction runs away or doesn't start, the only way to safely stop it is to flare off the butadiene. You cannot just stuff it back into bulk storage.
Refinery streams are flammable generally and often contain H2S. Operators can't simply let them go, potentially creating a boiling liquid expanding vapor explosion or a vapor cloud explosion. Even routine maintenance comes into play for using flares. Long pipelines of hazardous material eventually have to be cleared. Without flaring, where do you send the purge gas stream, since it will be toxic or could ignite?
James A. Bozin
As I understand from more than 30 years as a process chemist with a major oil company, continuous flares are "pilot lights" in a properly designed pressure relief system for a refinery or petrochemical unit.
The normal load of combustible by-products is used for heat and power. However, the pressure relief valves on a unit are piped to the flare stack. When a release occurs, the hydrocarbons are burned off in a controlled fashion, with reliable ignition provided by the flare. If pressure relief valves are released without these precautions, hydrocarbons might accumulate where they could form explosive mixtures with air. Upstream of the flare, these hydrocarbons would be too rich to burn.
Maybe the reliable ignition provided by the flare can one day be superseded by a rapid, computer-controlled igniter, as it is in my home gas furnace. But until then, flares are an important part of plant safety systems.
Joseph P. Bartek