ETI Project Manager Paul Winstanley looks at the safe future of hydrogen and hydrogen blended gas

Paul Winstanley Project Manager

The project began in 2011 in partnership with Imperial College London and the Health & Safety Laboratory which is now the research division of the Health and Safety Executive, based in Buxton. The project carried out a full literature review to identify what work had been performed to date by industry and academia on the use of hydrogen-based fuels in reciprocating and turbine powered generators. It was clear the work carried out had been minimal and there was a clear risk of downstream equipment such as heat recovery steam generating (HRSG) boilers being filled with unburnt high hydrogen fuels above their lower flammable exposure limits, if the generator suffered a combustion failure. We felt this posed a risk of an explosive event in the system.

From the literature review, typical syngas mixtures were tested in laboratory equipment to identify areas of interest. The downstream equipment heat recovery steam generating boiler (HRSG) used was a 1/7th scale model of a (HRSG) typical GE 350MW gas turbine which was developed to understand when the actual issues might be if a combustion event occurred in the system. To ensure a cleaner combustion system, the hot gas flow was generated using a Rolls Royce Viper 301 gas turbine, converted to run on Butane. In this test rig we have conducted 150 experiments using various hydrogen mixtures from pure hydrogen to mixtures of hydrogen, carbon monoxide and methane.

The results showed that hydrogen and hydrogen and carbon monoxide mixtures were more reactive than methane or hydrogen and methane mixtures. Our research also showed hydrogen and hydrogen and carbon monoxide mixtures could auto-ignite under certain circumstances, particularly when exhaust temperatures were above 350°C. On the other hand, methane had a moderating effect on these more reactive mixtures. It was also observed that flammable fuels when ignited, could result in damaging overpressures within the HRSG, depending on the fuel concentration as expressed by its EQR value. These pressures took the form of a single pressure pulse which was seen to originate in the region within and immediately downstream of the heat exchanger tubes where the turbulence was greatest. This pulse was then responsible for the pressure peaks seen at other regions of the system by normal wave transport and reflection processes. Limiting EQR values were therefore determined for the fuels tested, below which the overpressures generated could be safely contained within the HRSG.