Flashback arrestor pressure drop and flow capacity: how IS 11006 rates flow and why fine elements cost pressure

Every flashback arrestor costs pressure. The element stops a flame by forcing gas through passages too fine for a flame to survive, and the same fineness resists the normal flow of gas. IS 11006:2011 handles the trade-off by rating an arrestor's flow at the point where the drop across it reaches 10 % of its maximum operating pressure, so a buyer can compare devices on a common basis. This page explains that rating, how the main element types compare, and why stacked, undersized or fouled devices starve the torch or pipeline they protect.

Quenching is a heat-transfer duty. A flame dies when the walls around it take heat away faster than the reaction can replace it, which means small passages and a very large internal surface. For an oxyacetylene flame the passages must be finer than about 0.12–0.16 mm at 1 atm (two independent extrapolations, as compiled by S. A. Miller, 1965); for the decomposition flame of pure acetylene the quenching diameter measured by Cummings, Hall and Straker in 1962 was 0.6 cm at 2.02 atm absolute. Gas flowing through such a mass loses pressure by friction, and the loss rises with flow. The designer's job is to give the gas as much flow area as possible while keeping every individual passage below quenching size. The guide to how an arrestor works covers the quenching physics in detail.

The trade-off is old and well recorded. Tightly packed plugs of fine powder suppress a detonation reliably but resist flow so strongly that they suit only places where little gas moves, such as the connection to a pressure gauge (as compiled by S. A. Miller, 1965). Where the main flow passes, designers moved to sintered metal and, on large low-pressure lines, to packed towers.

IS 11006:2011 sets four requirements that together define what a buyer should be shown:

Two consequences follow. First, an arrestor run below its maximum operating pressure passes less than its rated flow, so a device rated at 1.5 bar but used behind a regulator set to 0.7 bar must be judged on the 0.5 point of its curve, not the headline figure. Second, a flow figure quoted without its pressure and pressure-drop basis cannot be compared with an IS 11006 rating at all.

Flow is normally measured on air or nitrogen, as IS 11006 cl 11.3 requires, and converted. Witt and Wilhelmsen/Unitor, two unrelated manufacturers, publish the same conversion factors: 1.04 for acetylene, 0.80 for propane, 0.95 for oxygen and 3.75 for hydrogen.

The C.P.I. described the dry granular arrestor as offering "only limited resistance" to gas flow, then drew the practical line: the loss is enough that dry arrestors "cannot as a rule be used in low-pressure installations" and belong on medium- and high-pressure systems, normally with dry gas from dissolved acetylene cylinders (C.P.I., Safety in the Production and Use of Acetylene, Paris 1968). The same code barred them from generator gas unless the acetylene is absolutely dry, because moisture attacks the packing. That is why a carbide generator outlet carries a wet seal; see wet flashback arrestors.

A wet seal has its own flow limit. The C.P.I. required any hydraulic device to stay effective at a delivery of about 30 % above normal, and warned that over-drawing a low-pressure hydraulic valve can pull its water into the torch line. Its identification plate must state both the maximum working pressure and the rated maximum delivery.

Manufacturers publish very different figures, and the pressure is usually the reason. One manufacturer's air-flow table for a regulator-outlet device (Wilhelmsen/Unitor W-66 S manual, air at 20 °C) shows how strongly flow depends on inlet pressure:

The acetylene column is an illustration, not a rating. It shows why acetylene devices look heavily derated: another maker rates its regulator-outlet unit at 10 m³/h on acetylene against 95 m³/h on oxygen, mainly because acetylene is held to 1.5 bar. IS 14814:2023 cl 7.2.1 sets the free-state acetylene limit lower still, at 0.612 atm, or 1.021 atm with PESO clearance, so most acetylene arrestors in India work well below their marked maximum. Size them on the curve at the pressure you actually use. The regulator and working pressures guide covers typical settings.

Losses add. A typical portable set has an arrestor at the regulator outlet and another at the torch inlet; a pipeline adds devices at the source, the branch and the point of use. IS 15190 (Part 1):2002 cl 4.12.1.2 calls for arrestors at the outlet of the gas source, the entrance and exit of any gasholder, each booster exit, each branch from main header to shop sub-header and the entry to each consuming unit. IS 15190 (Part 2):2002 cl 8.6 requires three flame arrestors with non-return valves in series on every cylinder charging or discharging manifold.

The standards anticipate the cost. IS 11006 cl 13 makes the manufacturer state an installation method that keeps total pressure drop low, and IS 15190 (Part 1) cl 4.12.1.5 says that where one unit cannot carry the flow, several are to be installed in parallel between common inlet and outlet headers. Parallel units are the correct answer to a flow shortfall; a larger-bore single device with coarser passages is not, because coarser passages may no longer quench.

Often more. The British Acetylene Association's 1962 code found that piping of 1–3 inches (2–8 cm) is adequate for up to 100 m³/h even over runs of hundreds of metres at the legally permitted pressures. IS 15190 (Part 2):2002 cl 6.4 avoids high-pressure pipe above DN 25 where possible and forbids anything above DN 40, recommending multiple DN 25 lines for more flow, sized at 2–4 m/s under cl 6.4.1. Small pipe is safer as well as adequate: the C.P.I. recorded that in 100 mm pipe a detonation is to be feared above 0.6 kg/cm². The pipe diameter and detonation guide explains why.

Because the element is a filter whether or not it was designed as one. Pipe scale, rust, carbide dust and thread debris collect on it, carbon from earlier flashbacks plugs it, and water from wet gas can block it entirely (Wilhelmsen/Unitor W-66 S manual). Every major manufacturer fits an inlet filter for this reason. A device that met its curve when new may starve the same torch years later, and a starved torch backfires more readily. Wilhelmsen/Unitor lists insufficient flow capacity among the causes of flashback, and the American Welding Society notes that arrestors reduce the gas volume available at the tip. The annual flow check described in testing and replacement exists to catch this.