Learnings

Outcomes

Workstream 1 & 2

1.     Pressure Swing Adsorption (PSA), gas separation membranes and cryogenic separation are all mature processes for recovery of hydrogen from off-gases within the chemical and petrochemical sector.

2.     For recovery of hydrogen from a dilute gas stream at a concentration of below 40% hydrogen, PSA is not suitable due to the underlying principles of operation that would require the major bulk of the natural gas to be separated from the minor hydrogen component into a low-pressure stream that would require considerable compressor power to re-deliver to the NTS grid.

3.     Cryogenic separation is capital intensive and most suitable for large stable gas separation requirements where the economies of scale can be exploited. The focus of this piece of work is to identify processes and equipment that would be suitable for smaller and intermittent gas flows and be flexible in both construction and operation to accommodate a wide range of large industrial consumers.

4.     Polymeric gas separation membranes can recover hydrogen from natural gas mixtures at low concentration and provide an upgraded stream enriched in hydrogen. However, polymer membranes are not suitable for separating and upgrading gas mixtures to an industrial hydrogen quality (>98%).

5.     Three scenarios have been identified for NTS customers that would be affected by the introduction of hydrogen into the NTS, namely:

a.     High hydrogen case – where a high concentration (>98%) of hydrogen is sought with the NTS providing a service connecting sources of hydrogen with demand.

b.     Low hydrogen case – where existing assets can accept some hydrogen in the fuel/feed gas (up to 5%) but where a stable and constant blend is required.

c.     No hydrogen case – where a customer cannot accept hydrogen in the offtake above existing GS(M)R standards (<0.1%).     

6.     Through discussions with solutions providers and systems integrators in the gas processing sector, this work has developed high level concepts for process plant to satisfy each of the three scenarios and has established indicative performance and costs for a deblending plant (service) in real-world applications.

7.     The combination of polymer membranes for bulk hydrogen recovery followed by pressure swing adsorption for hydrogen purification can provide a satisfactory solution for recovery of hydrogen at >98% from a 20% feed. Based upon an assumed 3500 hours of operation per year an indicative cost for the deblending service of ~€1 per kg has been determined. At a higher load factor of around 8500 hours/year this cost reduces to ~€ 0.75/kg of hydrogen recovered. However, the electrical power demands for recovery and upgrading of hydrogen from a 20% blend are significant and are of the order of 2.3- 2.4 kWh/kg.

8.     A simple gas separation membrane plant is able to reduce the hydrogen content from 20% to below 5% without high electrical power load. An indicative cost of around €13 – 17 per thousand normal cubic metres of gas delivered has been determined for a large gas fired power plant. The separated hydrogen can either be upgraded to an industrial quality or else re-injected into the NTS. The cost for upgrading of the gas has been determined to be around € 1.5 – 1.6/kg H2, however if the costs are shared between the lean and rich gas consumers, the specific costs of the deblending service resolve to € 13/1000 Nm3 for the lean gas consumer and € 0.75/kg H2 for the rich gas consumer.

9.     A “no-hydrogen” product gas (i.e. <0.1% v/v H2) can be deblended using gas separation membranes. However, because of the low partial pressure driving forces the membrane area becomes large and the cost of the deblending service increases significantly. A deblending cost of the order of € 37 – 75/1000 Nm3 deblended gas has been determined. The wide variation in deblending cost for this application is understood to arise from the differing performance of vendors membranes on this specific application.

10.  Gas separation membranes can be operated at ambient temperature, however when the feed gas is pre-heated higher trans-membrane fluxes can be achieved and are possible without leading to some of the conditions that might result in reduced performance or damage of the membrane (namely local condensation, plasticisation or concentration polarisation). The pre-heat required is of a low grade and might be available in the form of low grade waste heat from many of the processes that demand deblended gas.

11.  There is little industry experience in operating hydrogen deblending plant (gas separation membranes and PSA units) against variable loads with flexible operation and start-up. However, it is anticipated that many of the challenges can be resolved.                                                                                        

Workstreams 1 & 2 were peer reviewed by independent experts from Imperial College London with the below summary:

The reports provided for review are well written and informative. They clearly and succinctly 

highlight some of the key issues associated with technologies for deblending and implications 

for system design and structure. Attributes of these reports worth mention are summarised 

below:

·       The reporting covers a broad set of technologies, providing good coverage of a number of the key technology options

·       Analysis is spread across an informative range of scenarios, exploring the issues for three customer types, with two to three scenarios for each.

·       There is detailed spatial analysis of key offtakes from NTS, contributing to the evidence for scenario and customer selection

·       Useful schematic diagrams of deblending system designs are provided in the DNV report

·       The report identifies a number of knowledge gaps and opportunities for future research, 

There are some areas of the reporting that might be improved by additional observation or highlight.     

Workstream 3                                                                                    

The completed report by NPL provided a literature and technology review of novel and emerging techniques for gas separation, with particular attention to technologies that could disrupt the existing gas separation marketplace relevant to H2 separation from NG. In the summary shown in the report, 14 distinct technologies were highlighted, and a further 14 were identified in section 4 including 7 from existing companies. The review of new, innovative, or potentially disruptive technology demonstrates a wide field of research into gas separation, and specifically H2. Most technologies identified were applied to the broader application of separating different gas pairs for example H2/CO2, but several did investigate H2 separation from CH4 or NG. As each technology is developed further it is probable that their degree of applicability to separation of H2 from NG will be established. 

Gas separation research is profuse with membrane-based technologies, typically new materials, designs and processes that require scaling. Predominantly these are under development at laboratory, start-up, or small business scale. For each technology the evaluation criteria outlined in section 3 was considered. No single, standout technology with a high propensity for rapid growth or scaling for transmission use has been identified based on the applied criteria. 

Overall, there was a limited amount of information apparent to satisfy the criteria and in several cases no evidence was identified. The most promising opportunities identified in this review were either technologies at a very early stage of development offering the broadest potential applications or more mature technologies under-development at private companies. It is evident that more experimental data is required to be gathered on each type of technology before the application to H2 separation from NG can be fully assessed, and this reflects that the topic of gas separation, especially H2 from NG is a key target for many companies and it is expected that there will be new developments hereafter.

Workstream 4

Through our work with stakeholders, we identified the following key groups of issues which would

need to be resolved to enable the use of deblending technologies:

·       Network planning and customer needs: If customers have certain gas quality needs, networks could deliver this either through deblending, alternative solutions such as managing upstream hydrogen and methane injections, or customers could take action themselves (for example, adapting their equipment). Deblending is also one of the tools available for system operators to manage network blend levels. The key issue in this regard is how to ensure that decisions to invest in deblending to meet customer needs or network needs are least cost from a wider system perspective.

·       Managing system impacts: In some cases, deblending may require reinjection of unwanted gas back into the network. If this reinjected gas is hydrogen, or a hydrogen-rich blend, this can cause issues around downstream blend levels coming close to the blend cap. New processes will be required to ensure the impacts of the use of deblending on hydrogen blend levels downstream of deblending facilities are appropriately managed.

·       Network charging: Decisions will need to be taken regarding how to recover the costs incurred by network companies, if they build and operate deblending facilities. In addition, changes to charging methodologies will be needed to ensure that customers owning and operating deblending do not face ‘double charging’6 of gases that are not consumed, but simply withdrawn and reinjected into the grid.

We have considered a range of solutions to these issues with stakeholders, and have selected those likely to deliver reasonably efficient outcomes in the near term, while involving proportionate and low-regrets changes to existing frameworks. This focus on pragmatic solutions is important, given the uncertainty regarding how extensive future hydrogen blending (and deblending) might be before the transition to a fully net zero system.

The need for deblending is likely to be minimal in the near term, and it is possible that deblending will only have minimal usage throughout the transition to net zero. We have therefore focussed on solutions suitable for the near-term period. While these conditions may well persist through the transition, we have also sought to ensure that the proposed solutions avoid creating barriers to alternative solutions that could be needed to deal with potentially more complex circumstances, should they arise in future.

Recommended changes to the gas commercial frameworks:

Network Planning 

·       Gas customers with a specific gas quality need communicate their need through an application to the relevant network operator, including supporting evidence to justify their need.

·       If Government decides to socialise deblending costs through network charges, we recommend that networks be required to justify any expenditure on deblending using a cost-benefit analysis (CBA).

·       If Government decides to target deblending costs such that they are recovered from the customer requiring deblended gas, we recommend no changes to existing frameworks. 

·       In the case where deblending is used for networks’ own needs (e.g. to manage blend levels or billing impacts), we recommend continuing to rely on existing totex incentives

Managing system impacts

·       We recommend that reinjection points are connected to the grid subject to an enhanced pre-connection Impact Assessment (IA) by the relevant network operator.

·       If it is accepted, we recommend including conditions in the Network Entry Agreement (NEA) such that reinjection can be constrained by the network operator if needed for blend management purposes.

·       We recommend that, in the early stages of ‘minimal deblending’, an administrative approach is used to enable operators to choose between different tools.

·       Finally, we also recommend that where reinjections on the transmission network cause constraints on downstream networks (and vice versa), the relevant networks should coordinate to apply the chosen solutions described above in a ‘whole system’ manner.

Network charging

·       We recommend that where networks secure funding allowances for investment in deblending facilities (i.e. via allowed revenues) to serve a customer need, the costs of building and running such deblending facilities should be recovered through targeted charges.

·       Finally, we also recommend that, where customers own deblending facilities, the network charges they face are levied on a ‘net-exit’ basis (i.e. the net amount that leaves the network less any amount that re-enters from the deblending facility) such that users do not face entry charges on reinjected gases.

Building on the findings of DNV report, NGT has asked Frontier Economics to provide projections of future deblending technology costs across the range of deblending technologies taking into account changes in market conditions, economies of scale and expected efficiency learnings accrued over this time. In doing so, we have engaged closely with DNV and the supplier.

Our findings are that the costs of deblending depend largely on what use case deblending is being used to support: methane recovery or hydrogen recovery. For methane recovery use cases, estimated average production costs range between £1 and £5 per MWh of methane recovered, assuming a 20% hydrogen blend by volume. For hydrogen recovery costs, estimated costs range between £7 and 60 per MWh of hydrogen recovered. 

Within these ranges, precise costs depend on:

·       the precise technology used (higher costs are associated with technologies capable of delivering relatively pure streams of both hydrogen and methane); 

·       the purity of hydrogen or methane delivered (delivery of higher concentrations is associated with higher costs);

·       assumed economic lifetime of deblending facilities (a shorter lifetime drives a higher average unit cost, as construction costs must be recovered over a shorter timeframe); and

·       assumed energy prices (higher energy prices result in higher costs).

The cost ranges above are based on current construction costs and energy efficiency. Further reductions in construction costs and energy costs over time may be possible.

As this NIA project was classed as desktop research there were no direct benefits to be tracked, the project however does contribute to the understanding surrounding blending and deblending in the transmission system and if the technology can help the network transition to a greener gas earlier by protecting sensitive customers, then the benefits will be significant.

Lessons Learnt

Progress 21/22

The stakeholder led workstream relating to the commercial frameworks side has worked very well so far as it has given ample opportunity for the group to input ideas, concerns and recommendations. The set up of multiple virtual calls supported by a strong slide deck has kept the pace of the workshops up with a little update and reflection on last months each time. As the concepts covered affect a number of stakeholders this has given them a good opportunity to be apart of the project instead of just consulting at the end.      

Closure 22/23

The final webinar hosted by Frontier worked well with an update from the Engineering workstream by DNV and the Commercial update from Frontier. This was a really effective way to discuss and share some complex outputs from the project and offer an opportunity to ask further questions.

Having a collection of suppliers (DNV, Frontier, NPL) worked well in that each could share their expertise on the overall subject and with clear deliverables for each there was no overlap. The change control was required as a slight delay in one workstream had a knock on impact on following activities which is a risk with multiple different suppliers.

Having the independent peer review offered another perspective to the findings and highlighted some topics for further work.

Part of the delay was around seeking some NDAs from the OEMs which was not expected, a lesson learnt was that these could have been agreed at the start to prevent any delays.