Selection of Metal for Oxygen Valves

Abstract

The document ASTM G.94.92 “Standard Guide for Evaluating Metals for Oxygen Service ” gives a complete guidance for the selection of metals. The different test methods are explained and a metal ranking table is given for each method. The best-known test is promoted combustion as described in ASTM G.124-95 “ Standard Test Method for Determining the Combustion Behaviour of Metallic Material in Oxygen Enriched Atmospheres ”. In this ranking aluminium bronze with only 9 to 13% aluminium is near aluminium and is therefore considered less resistant than any stainless steel. This result is not consistent with the paper “ Ignition Testing of Hollow Vessels Pressurised with Gaseous Oxygen ” [1]. The particle impact test performed by NASA as described by Benz, et al [2] also gives a better ranking for aluminium bronze, between Inconel 600 and 625. Analysing how ignition occurs in oxygen pipeline networks, we considered that the most probable cause is promoted ignition by particles. New rankings of metals are established in the case of impact on bulk materials by burning particles. Two theoretical rankings are based on interface temperature associated with a critical temperature, while another ranking is based on the results of tests made on the gaseous oxygen impact apparatus, with the addition of particles. Measurements of interface temperatures for different metals confirm the theoretical values. A comparison of the different rankings is discussed for metal selection for oxygen valves and it appears that stainless steel and aluminium bronze with 9 to 11% aluminium are equivalent choices for oxygen valves.

Investigation of a large industrial oxygen valve fire

“…Metal particles carried by flowing oxygen have impacted a metal part at a velocity larger than about 50 m/s (Williams, Benz, & McIlroy, 1988). Fano, Faupin, and Barthelemy (2000) theorise that approximately 90% of the fires in oxygen pipelines with a length ranging from 1 to 500 km have been ignited by particle impact. Newton and Forsyth (2003) conclude that the fire of a 152-mm (6-in) ball valve was due to the ignition of a hydrocarbon lubricant.…”