The UK Government’s target to reduce greenhouse gas emissions by 80% of 1990 levels by 2050 requires the gas networks to explore alternative options for heat. Hydrogen has become an attractive alternative to natural gas because it does not emit carbon when combusted.
SGN are involved in a number of research and development projects associated with hydrogen ranging from transmission down to low pressure distribution. Before we can demonstrate that these hydrogen projects are viable, a leak detection instrument that can detect hydrogen at the same levels natural gas is required. The leak detection instrument is a critical safety feature to managing and operating a gas network.
Objectives
The objectives of this project are to
· Investigate sensor technologies that will detect hydrogen.
· Develop test and certify the prototype detector against relevant standards
· Produce user documentation and training guide for the gas detection instrument
· Further test the gas detection instrument in a field trail
Learnings
Outcomes
The project aimed to develop a gas detection instrument that can be used by SGN to trace internal and external hydrogen escapes and then confirm absence of hydrogen post repair. The project plan divided the project goal into 5 key project deliverables:
1. Conduct a review of IPR and prior work
This review found several patents relating to hydrogen detection and development of sensing elements. With the available project timescales none of the patents were considered further. Instead, the review concluded that existing technology would be assessed for suitability with no impact on any existing IP
2. Generate a technology assessment report
This deliverable was to recommend which sensor technology should be adopted. This report concluded that hydrogen measurements could be derived from two existing sensor technologies namely catalytic and thermal conductivity. Although changes would be required to the way these sensors typically operate, it was believed that measurements from 0ppm to 100% volume would be possible. The report also stated that the infrared sensor currently being used in the standard GS700 would be retained. This would not only allow natural gas readings to be made but would also allow the instrument to reliably determine the gas type being sampled.
3. Production and testing of a hydrogen leak detection instrument
This stage involved incorporating the proposed hydrogen sensors into a modified GS700 and to assess performance from 0ppm to 100% volume.
First steps were to redesign the sensor PCB to facilitate the hydrogen sensors. The hardware circuitry that was implemented used a novel configuration that allowed the catalytic sensor (%LEL) to operate at a temperature lower than that required to measure methane. This meant the sensor only reacted to hydrogen.
Linearisers were created for both the catalytic (PPM and %LEL) and thermal conductivity (%GAS) sensors. Additional hardware and firmware changes were included to generate an effective PPM range. The PPM range uses the same sensor as the %LEL range but is passed through a higher value gain stage in the circuitry to allow for higher resolution readings to be made for PPM hydrogen measurement.
During this stage of the project, there was critical learning on the effect of the background gas when measuring hydrogen in the %GAS range. During purging operations, a background of natural gas may be present when purging between hydrogen and natural gas. The natural gas has a thermal conductivity which is lower than hydrogen but significantly higher than air. Because the sensor is zeroed in air the presence of methane/natural gas will be measured as a shift in the zero. Compensation algorithms were successfully developed and tested to compensate for this effect.
Temperature effects on gas readings were also compensated. The performance of the compensation for the %LEL range is excellent with very little change in signal measured at 50%LEL hydrogen across the temperature range -10°C to +50°C.
A simulator was developed that allowed the operation of the instrument firmware to be tested without the need to use an actual instrument. The simulator is completely software based and mirrors the operation of a real instrument in how it responds to user interaction and ‘gas application’.
The simulator was an invaluable tool that accelerated the instrument firmware development and it also facilitated constructive virtual demonstrations.
4. Testing, verification and certification of the prototype hydrogen detection instrument
This deliverable for this stage was a performance tested, ATEX certified, hydrogen leak detection instrument.
Performance testing was conducted by an independent UKAS accredited testing laboratory. The %LEL and %GAS ranges were tested BS-EN60079-29-1:2016. (Note – This BS EN standard does not cover PPM therefore the INQ4 requirements were used as the benchmark for the calibration and adjustment test). All required performance tests were passed.
The modified GS700, together with the modifications to the sensor PCB and the hydrogen sensors, were submitted for the relevant ATEX certification. This was fully achieved, and the instrument is certified with a protection level of Flameproof Ex db and Intrinsically Safe Ex ia.
The GS700 was submitted to Eurofins York in Grangemouth to perform the required EMC testing. This was required since hardware changes had been implemented. Tests were conducted in accordance with standard EN50270:2015 but the instrument failed both the radiated immunity and ESD tests. Following extensive investigation, the bonding between the metal can of the infrared sensor and the ground plane on the PCB was improved. This resulted in the sensor’s metal enclosure providing better shielding to RF interference entering the instrument. This modification was prototyped successfully and subjected to a repeat set of radiated immunity testing, again at Eurofins York, and passed. The modification had no detrimental impact on the ATEX certification or the safety rating of the product.
5. Field trial assessment
SGN engaged with NGN and it was agreed that the instrument would be used extensively at Spadeadam to support their H21 project while essential field data was collected to support the hydrogen gas detection instrument project. In addition to the work undertaken at Spadeadam, it was agreed with SGN that any other relevant hydrogen projects in the UK should be assessed, and if applicable, should be supported to gain further field exposure and feedback. In total, five field trial were supported.
This field trial was established with the aim of obtaining regular feedback. The instrument was used for a wide variety of applications including survey, barhole, leak investigation and extensively for purging. NGN commented that calibration “was checked periodically” and accuracy of the instrument was “within 1%”. This is in alignment with SGN’s current calibration policy in INQ.
From field visits by GMI, it was evident that the instrument quickly became the instrument of choice for operatives on-site. No feedback was received on the instrument’s user interface suggesting that the configuration supplied was suitable for the application, however, this should have been formally recorded.
Notably, instrument drift was of up to 4% was reported. However, the instrument was not calibrated over the four-month period in question which would have rectified the issue. Automatic calibration was not in scope for the project but needs to be included if the product proceeds to commercial release, outside of the SGN H100 Fife trial. As part of the H21 project, both Baxi Boiler and Worcester Bosch installed 100% hydrogen boilers into the houses at Spadeadam. Following requests from both boiler manufacturers, GMI supported their commissioning activities. During the short one-day activities, the instrument was used for leak investigation (ppm range) and purging
Steer Energy Solutions were researching domestic purging with hydrogen in associating with BEIS in relation to the Hy4Heat project and requested a GS700-Hydrogen instrument to support their on-going testing of hydrogen releases into a room and purging.
Steer conducted comparative tests on the volume range using their Bronkhorst flow controllers. There was strong correlation of results and typically less than 1% difference, but only up to 57%. Above this concentration the difference increased and suggests that the instrument lineariser may need reviewed around this region.
Steer also performed LEL testing and compared performance with “other sensors and detectors being used”. Since the detail of this comparison was not recorded, it again suggests that trial feedback data was not closely monitored. However, performance is perhaps validated as Steer comment “the GS700H was successful in capturing the required data”.
The field trial at the hydrogen homes near Gateshead was requested by NGN and is like the house set-up at Spadeadam. NGN advised that an instrument would be desirable to support typical FCO activities including survey and purging.
A variety of users, in varying locations, have evaluated the unit, with excellent performance feedback and with little operational difficulties noted. This field testing, coupled with the laboratory testing and associated product certification received, indicates the GS700-Hydrogen satisfies the project requirements.
Lessons Learnt
The project has evidenced that Infrared techniques, currently used for sensing flammable gases will not work for hydrogen. Sensing hydrogen from ppm up to 100% GIA will require a combination of catalytic and thermal conductivities. The full sensor assessment report can be provided upon request.
The project has proven a dual gas detection instrument that is fully certified and can detect natural gas and hydrogen is possible. The hydrogen gas detection instrument aims to be commercially available by end of 2022.
