Operating atmospheric vent collection headers using methane gas enrichment

Authors


Abstract

Tests at 60°C and 16 psia using ethylene, hydrogen and methyl alcohol “fuel vapors” showed that if an atmospheric vent collection header contains 25 vol% of methane and the only source of oxygen is the air, no possible mixture of fuel vapor, nitrogen and residual oxygen is flammable. Addition of these fuel vapors to a header containing 25% by volume of methane in all cases increases the 3.8 vol% oxygen safety factor that exists with zero fuel vapor in the gas stream. It is irrelevant that the fuel vapor has an upper flammable limit (UFL) greater than the methane enrichment gas. The minimum oxygen concentration to sustain a flame (MOC) increases with increased methane: nitrogen ratio in the gas stream, so that the “listed” MOC has no relevance under methane enriched conditions. These findings have important ramifications when applying Coast Guard Regulations in 33CFR.154 for Marine Vapor Control Systems, which implies the need to operate at 170% of the combined gas stream UFL and requires operation at less than the MOC (≤ 8% oxygen) when tanks have been partly inerted with nitrogen. Large reductions of enrichment gas usage with attendant environmental benefits are technically possible using flow control of methane rather than gas analysis downstream of the enrichment station. Operation above the UFL rather than below the MOC can cut enrichment gas usage by 50% or more while actually increasing the assumed 2 vol% oxygen safety factor. A negative flow control error of 7 vol% methane (−28% of target) is required to achieve flammability under worst case assumptions.

Exceptions: Users are cautioned that a comprehensive list of exceptions to the 25 vol% methane enrichment method has not been developed. Using the test protocol described the method appears to fail for decomposable fuel vapors, such as ethylene oxide, and fuel vapors susceptible to cool flames under vent header conditions, such as ethyl ether. Coast Guard regulations in 33CFR.154 have not changed and applications for variance are considered on a case-by-case basis. In the context of this paper, additional test data development may be justified where the cargo vapor UFL exceeds that of methane.

When operating at less than the MOC using nitrogen or other inert diluent, it is important to allow for errors in the MOC value used. NFPA 69 lists many MOCs measured in the 1930s using weak ignition in open ended 2 inch diameter glass tubes. Since MOC depends on test conditions it is optimistic to expect such values to apply to large flare headers. For example, the MOC measured for ethylene was 8% rather than 10% oxygen as listed in NFPA 69. This offsets the oxygen safety factor normally applied for continuously monitored gas streams. Methods for estimating MOC based on LFL can introduce additional errors. These errors are compounded in cases where the MOC is estimated for mixed gas streams using the Le Chatelier Rule. Example calculations are compared with measured MOC values listed in NFPA 69. Some practical considerations are given with respect to oxygen sources in vent collection headers and flame arresting devices used.

Ancillary