Project Summary
How DPLA meets the aim of SIF Challenge (Eligibility Criterion 1 and 3)
The DPLA addresses the whole system integration and data and digitalisation Innovation Challenges set by Ofgem, whilst simultaneously addressing a range of user needs. Comprising of eight work-packages (WP), the Project expects to be delivered over a two-and-a-half-year time frame. The Project will enhance the coordination between the distribution networks and regulatory bodies as they work towards a common goal: reduce carbon emissions, realise customer benefits, and improve safety in a cost-effective manner. By combining upgraded modelling capabilities, the Project will deliver the next generation of user driven digital processes accelerating progress in methane leakage detection, as well as unlock opportunities across hydrogen leakage detection. The DPLA will directly improve data monitoring and insights improving efficiency and resilience of the networks through the innovative system architecture unique to the Project.
The underlying system architecture has evolved to include an expert system and user interface
Prior to the DPLA project (Discovery and Alpha), an expert system based on weighted rule-based inferences combining individual leak detection was designed. Across Alpha, the Project Team demonstrated the provenance of the idea and the opportunities to synthesised outputs into a digitally driven user interface. It was determined that the user interface will enable Cadent's workforce to view and interact with leakage data quickly, easily, and effectively. Together with the hybrid hydraulic model and advanced analytics, the expert system and user interface should enhance network coordination, reduce operational complexities, and improve user experiences.
The perception of the problem evolving, demonstrated through a range of detailed user needs:
Whilst in the first instance the ambition of the DPLA was to address gas leakage, through engagements with all Great Britain's distribution networks, the use cases have expanded into a non-mutually exclusive set of regulatory, operational, and business use cases. Whilst eight key use cases were identified, the five with the highest relative importance were as follows:
Gas Leakage Regulatory Reporting -- a need to provide more accurate annual reports of gas leakage quantity to Ofgem, compared to the current Shrinkage and Leakage Model (SLM)
Condition based monitoring -- a need to improve the understanding of the state of network assets by proactively detecting leaks rather than relying on models alone
Regulatory Performance and Revenue Generation -- A need to accurately measure network performance in reducing shrinkage/leakage and develop a fair incentive mechanism for reward and/or penalty
Proactive emergency intervention -- a need to reduce the risk to humans and properties via smarter and earlier identification, characterisation, and localisation of gas leaks
Improved Asset Replacement and Maintenance -- a need to leverage an improved understanding of the network and leakage hotspots to tailor the schedule of the Mains Replacement Programme and better target maintenance cycles and AGI replacement
Moreover, informed by the above use cases, the user interface will leverage a dashboard to display key information providing a priority view to the intervention engineers. Outputs are planned to include real-time alerts of critical leaks, visual heatmaps and leakage reports.
The key learning from the Alpha phase was the importance of the hybrid hydraulic model and machine learning based approach to offsetting the cost of physical in-field sensors, to achieve the same level of emissions reduction. This is now a fundamental principle of the DPLA and this philosophy will underpin the concept build in Beta.
Scalable solution
The DPLA project recognises the importance of a scalable solution, and hence, an area reflective of asset availability (including low, medium, and high-pressure tiers), type of assets (plastic, cast iron, steel) and physical spatial features (e.g., Urban versus Rural, Coastal versus Inland). Areas across North London and East England should demonstrate the viability of the system architecture. During the project mobilisation phase (August -- October 2023), these regions should be reaffirmed based on available data (WP2). There are opportunities to expand the scope of the DPLA and include advanced criteria such as the application of the solution across the transmission system, as well as the proactive detection of hydrogen and carbon.
Participation from a range of stakeholders (Eligibility Criterion 6)
As detailed across the Beta plans, the DPLA will enable knowledge dissemination both internally and externally with regulatory bodies, the GDNs as well as customer/third sector groups (e.g., Citizens Advice, IGEM, and Fuel Bank) and governing bodies (e.g., Ofgem, Environmental Agency and HSE). Engagement with regulatory bodies and shippers have already commenced (Question 10). Opportunities to realise interconnected benefits across SIF funded projects are already being identified. For example, aligning DPLA with the LeakVISION, a patented in-pipe gas leakage detection sensor.
Innovation Justification
Current processes and available state of the art solutions
As part of their licence condition, Gas Distribution Networks (GDNs) report their annual leakage emissions using the Shrinkage and Leakage Model (SLM). The SLM provides a static theoretical value of total gas leakage and cannot identify actual leak locations or volumes. The SLM is based on legacy data and studies from 1994 and 2002. Small improvements have been made over the years as the GDNs are obliged to review and try to improve the SLM. However, progress has been slow, and it has remained a static, theoretical approach which lacks the accuracy and granularity needed to inform a cost-effective, strategic emissions reduction plan. In addition, the SLM is not compatible with a hydrogen blended network, which is likely to be a major issue for all GDNs in future.
More recently, companies have started to develop innovative in-field products, sensors, and tools to detect leak locations. Although several of the solutions have been successful in accurately locating leaks, they have narrow focus on specific assets and are challenging to scale cost effectively.
This project seeks to revolutionise the GDNs' approach to leakage by replacing the SLM with an innovative, dynamic approach. It will build on past leak detection projects such as NGN's "LeakVision", National Grid's "Monitoring of real-time fugitive emissions" and Cadent's "ThermalTrax", as well as global oil and gas leak detection projects like the Environment Defense Fund's "Methane Detectors Challenge".
DPLA's innovative nature and transformational approach
DPLA proposes a first-of-a-kind data driven approach to dynamically detect and locate methane leaks on distribution networks. The solution applies a combination of physics based hydraulic models and machine learning based models that leverage existing and new data to detect and locate gas leaks more accurately than the SLM. Although elements of the solution e.g., machine learning models, hydraulic models, and/or innovative in-field products have been demonstrated in isolation, they have never been integrated in an analytics platform to successfully deliver a holistic approach to leak detection. Q#5_Appendix Innovation Justification provides an overview of the novel solution architecture to be developed and trialled during the Beta phase.
Post-project IRL and CRL
The need for SIF funding and large-scale trial are also aligned with the current assessment of the integration readiness level (IRL) and commercial readiness level (CML) of the DPLA.
Data integration is a key component and challenge when it comes to large numbers of datasets from disparate sources. Although some extracts have been successfully delivered in Alpha phase, the Project estimates DPLA's IRL at level 2. DPLA's Beta phase is expected to successfully demonstrate full data integration within the trial area hence DPLA is expected to reach at least IRL 6. While IRL 7 is achievable for Cadent, further work is required for other GDNs.
Considering commercial readiness, all GDNs are supportive of the value from the DPLA and have demonstrated this by partnering on the project, however no commercially viable product is available, hence at Alpha conclusion the Project estimated the CRL at level 3. Following the Beta phase, the value of DPLA will be demonstrated on a trial area, however the platform will need development before scaling up to a Great Britain scale production solution, hence DPLA is expected to reach CRL 6-7.
DPLA and competitive markets development
DPLA is a project supported by all Great Britain's GDNs and National Gas Transmission (NGT). If proven successful during trials, the solution can be rolled out across GDNs. The Project will develop and demonstrate a blueprint for how data, analytics and models can be used to identify and locate leaks. The blueprint can be replicated by other GDNs using their own data and preferred technology vendors, hence enabling competitive markets to drive value for Great Britain's gas distribution customers.
Why is the proposed scale appropriate?
The proposed project activities will enable Cadent to successfully demonstrate the DPLA on a live gas distribution network. A live large-scale trial across multiple asset types and geographies is needed to build the confidence in the DPLA results, which will drive key operational and planning decisions in future and replace a solution (the SLM) used for circa 20 years.
Why is the SIF the right funding mechanism?
Primarily, the GDN licence conditions mandate that the SLM is used, therefore there is no regulatory incentive for GDNs to consider adopting alternative solutions until these have been successful proven and mandated by Ofgem. If GDNs had the opportunity to develop new approaches and solutions, the innovative nature and associated risk in development of DPLA would prevent GDN's from funding the innovation. Furthermore, the scale of the project, focus on net zero transition, and collaboration across GDNs means that SIF funding is applicable rather than NIA funding.
Benefits
Environmental impacts
Key quantifiably evidenced benefits from DPLA implementation will be the 1) avoided gas loss and 2) avoided Greenhouse gas emissions. The Project performed a detailed cost benefit analysis in the Alpha phase, building on work in the Discovery phase with updated assumptions and more accurate costings. Assuming all GDNs were to implement the DPLA, updated values were:
12,435 GWh of avoided natural gas and/or hydrogen loss volumes by 2050 (benefit to the end consumer)
14,856 ktCO2e of avoided greenhouse gas emissions from distribution network shrinkage and leakage by 2050 (carbon reductions)
DPLA will reduce Greenhouse gas emissions from two gases with far greater global warming potential (GWP) than CO2: methane (GWP of 25 kg CO2e/ kg CH4) and hydrogen (GWP of approximately 11 kg CO2e/ kgH2). The DPLA will therefore continue to have a significant carbon reduction impact as GDNs begin to incorporate hydrogen into their networks, particularly since hydrogen has been shown to leak more readily than natural gas.
The avoided loss of natural gas/hydrogen and avoided equivalent greenhouse gas emissions will continue to be used as the metrics to track benefits as the business case is further refined and realised in the Beta phase.
Cost reductions in operating the network and wider energy system
Accounting for the total costs of DPLA deployment and assuming a representative forecasted cost of natural gas/hydrogen and carbon dioxide equivalent (see Appendix for trends and full list of assumptions), the above avoidance figures amount to net cumulative discounted financial benefits across GB of up to £2.86 billion by 2050 for our core modelled scenario in the Alpha phase. This is a fivefold increase on the initial financial benefits value calculated during the Discovery phase, largely due to the sustained long-term increase in forecasted gas prices.
The benefits are dominated by the cost of avoided emissions (~90% of the £2.86B), given the high GWP of the gases whose leakage is being minimised.
The precise magnitude of the benefits will depend on the future evolution of gas and carbon prices, the exact technology mix used for the remote leak sensing, and the pace of DPLA rollout across gas networks nationwide. The core assumed technology mix combines Picarro for rural and urban pipelines with SeekOps for large AGIs. The core scenario pace is that Cadent starts to test and deploy first, with other GDNs following a year later. Across three assumed paces and three technology options, the cumulative discounted net financial benefits range from £0.8 billion to £3.2 billion. Almost all combinations of the four factors listed above lead to the project breaking even by 2027.
Cost savings for network services users
The avoided volume losses could feed through as direct customer benefit by decreasing the cost of shrinkage and leakage gas which is passed on from GDNs to consumers' bills. At the very least, the DPLA will result in no net impact on customer bills, and could even lead to reductions in bills if less in-field monitoring is required as part of the DPLA's realisation, or if the baseline of methane emissions is found to be higher than the current modelled amount (through the SLM).
Impacts on consumers (individual and collectively) of the whole energy system
Further benefits to energy system consumers include:
Reduced carbon costs, representing a further environmental benefit to customers
Improved health and safety, with DPLA reducing the number of site visits required and thus also the risk of accidents
Lower fire risk, as DPLA allows networks to take immediate action to prevent ignition by providing real-time alerts.
Reduced disruption, as better localisation of leaks reduces the noise and time taken to fix them. This will improve public confidence and social acceptance.
Economic benefits to users and any other parts of the supply chain, broader (UK) economy
DPLA will provide improved certainty on operating conditions and modelled outlooks, resulting in gas shippers having to build less risk into their pricing strategies. This will lead to a narrower spread of prices which can feed through to consumer bills.
Impacts on government priorities
The estimated decrease in methane emissions from pipes and AGIs between 2020 to 2030 via DPLA is up to 58%, which supports the government priority to tackle methane emissions as a Participant of the Global Methane Pledge which was committed to at COP 26 in November 2021.
DPLA sends a strong message that GDNs are doing the right thing and sets an example for other networks and even other countries to do the same. It paves the way for best-in-class approaches to regulatory and operational methane emissions reduction.
Expected regional or wider energy supply resilience benefits
In future, the platform could be developed to perform predictive leak prevention as well as leak detection. This would further improve energy supply resilience by minimising system losses.
