Learnings

Outcomes

The final project synthesis report has completed its HSE review and has since been updated to reflect the outcomes of this review process. Some of the high-level conclusions of the project are as follows:

Contaminant Selection

The chemical characteristics of the range of possible contaminants were considered, and the contaminants taken forward for further testing were:

·       Mono Ethylene Glycol (MEG): a commonly used agent in natural gas to help prevent leakage from the network.

·       Toluene: representative of aromatic hydrocarbons.

·       Methane: as a main component in natural gas and as a representative aliphatic hydrocarbon.

·       Sulphur-containing compounds: to represent contaminants associated with odorization of the hydrogen supply and components that may be adsorbed onto pipe walls.

·       Nitrogen: as a representative inert gas that may be present from purging activities (combustion testing only).

Material Selection and Performance

The material selections were based on information on pipework from UK domestic property surveys and prior test work, together with information from stakeholders, relevant standards, and other ongoing research projects. The materials selected for testing were:

·       Copper: primary material used in UK gas pipework installations (particularly domestic).

·       Carbon Steel: a common material used in UK gas pipework installations.

·       Brass: Copper alloy of which most pipework fittings are made of (e.g., soldered, compression).

·       Aluminium: primary material used in pressure regulators and in some appliances.

Lead and stainless steel were also considered. Lead was the focus of a separate study being led by DESNZ and will be removed as part of the conversion process to hydrogen regardless, whilst stainless steel is known to be more resistant than carbon steel under the operating conditions of gas installation pipework, and as such the selection of carbon steel is thought to represent the “worst case”.

Copper, carbon steel, and brass all displayed consistent SSRT results and trends irrespective of the gaseous environment they were exposed to. This represents a positive outcome of the research suggesting that all three materials are suitable for repurposing to low pressure hydrogen service.

The results for the cast aluminium samples were less consistent, with sometimes noticeable variations in results between different samples. There was however no consistent pattern to indicate that exposure to hydrogen, or a hydrogen-contaminant mixture had any impact on the material properties, and it was concluded that the inconsistencies in the results were most likely down to variability in the case aluminium material itself.

With practical, experimental constraints, limited numbers of tests (two or three) for each material at a given environment were performed. The results showed that there were uncertainties in the quoted standard deviation, with some of the average results in hydrogen environments lying just outside the range of the repeatability in measurement. Although recognising that there was a limited range of testing undertaken, the net effect of hydrogen on metal properties is evidently small and should not result in material performance impacts. The overall conclusion here is that the presence of the contaminants in hydrogen did not show material integrity impacts.

Combustion Burner Selection and Performance

For the combustion studies, a glass fronted, balanced flue gas fire and a modified natural gas cooker formed the basis of the test programme developed to provide information on the overall combustion efficiency, emissions, and operability. These burners provide a good test environment to study the impact of contaminants. Other appliances could be considered in the future with tests focused by the outcomes of the current research.

For the gas fire, the measured nitrogen oxide (NOx) emissions were around 20 ppm for tests at 4.5 kW and less than 2 ppm for tests at 1.5 kW. Some of the tests at 4.5 kW had NOx emissions greater than 130 mg/kWh (a current, proposed limit) but these were not the influence of the contaminants and more associated with main hydrogen combustion. Once operating at steady state, there were no unburnt hydrogen emissions or any noticeable difference in the optical flame pictures. The infrared (IR) radiation and the peak radiation temperature were also consistent across all test gases.

A 2 kW natural gas cooker hob burner was modified to operate on hydrogen to understand if there are impacts of using hydrogen with possible trace contaminants on gas cooking. The modification changed the original design into a diffusion combustion burner with an orange flame that was clearly visible in all tests. The air flow to the burner was restricted to support development of stable flames. This modification could be applied to any natural gas cooker hob as they all use premixed combustion with similar design principles.

The hydrogen emissions in the combustion products were near zero for all fuels (PAS 4444 requirement <1000 ppm), with methane and toluene hydrocarbon trace components burnt to >99% efficiency. NOx emissions were similar for all gas compositions apart from the mixture with 2% N2 in hydrogen, where the NOx was 59% of that for pure hydrogen, on a zero-oxygen basis. A further study is required to understand the impact of all cooking emissions (not just NO2) from the various domestic cooking options available to end users before policy decisions are made in this area.

Lessons Learnt

The project was awarded via a mini-tender event to a contractor/sub-contractor consortium which is not an organisational structure that will be used for this type of research project in the future. The need for various suppliers couldn’t naturally have been avoided due to the different test programmes that had to be delivered; however, the sub-contractors that were responsible for delivery of the testing were often the source of project delays. Despite attempts to get a tighter grip on delivery by convening regular update meetings throughout the test phase of the project, the sub-contractors were not particularly responsive to any pressure exerted and often preferred to liaise directly with their own client. 

The preference moving forward will be to enter contracts directly with any/all suppliers. While this may be more involved from a contractual perspective up front, the NIA contract template used by the gas networks is designed to be a multi-party agreement, this will enable far more control of delivery once a project is up and running.