Project Summary

FastPress is an analyst‑in‑the‑loop decision‑support tool designed to improve long‑term planning for the GB gas transmission network. It combines hydraulic simulation with advanced AI/ML optimisation to explore complex, constraint‑heavy network scenarios at scale. FastPress enables NESO and NGT to assess resilience, demand uncertainty, and infrastructure reconfiguration options more efficiently than existing tools. Evolving from earlier NIA projects, FastPress has shifted from solving individual scenarios to supporting whole‑system risk assessment. The project supports evidence‑based decisions on asset repurposing, reducing cost, risk, and uncertainty for energy consumers.

Innovation Justification

FastPress addresses Round 5, Innovation Challenge 7 (Whole System Optimisation) by enabling NESO and NGT to assess future gas network capability, resilience, and infrastructure configuration at a scale, consistency and level of integration that is not achievable using current manually intensive approaches. By automating scenario solving, flow analysis, and infrastructure timeline optimisation (while retaining analyst oversight), FastPress enables a whole system assessment of trade-offs between operability, cost, resilience, and transition pathways. This supports integrated, evidence-led whole-system planning rather than isolated, asset-level decision-making.

Innovation

FastPress represents a radical step-change in gas network planning through the novel application of AI-driven optimisation to physically validated, deterministic hydraulic simulation within a regulated environment. The innovation goes beyond incremental efficiency gains by fundamentally changing how planning problem spaces are explored. Key innovative aspects include:

• Bespoke genetic AI algorithm to optimise over large, non-linear, constraint-heavy solution spaces, where discrete decisions and continuous variables must be solved simultaneously. Solves are accelerated using machine learning (ML) approaches on historic solutions.
• Automated solving of static and time-coupled transient scenarios, transforming a stagnant workflow which struggles with the complexities of transient scenarios.
• Smart solving of entry and exit flow analysis that finds novel solutions using adaptive ML methods.
• Infrastructure timeline optimisation, enabling sequential identification of assets that can be removed/repurposed (and when), without compromising system resilience.
• Presenting novel visualisations of results in a custom frontend that gives analysts a better understanding of solutions and brings most disparate activities into a streamlined, user-friendly space.
• Encoding analyst knowledge, enabling advanced heuristics to ensure validity, safety, and speed improvements.
• FastPress is platform-agnostic, where suitable integration can be applied to other physics-based solvers, mitigating SIMONE lock-in.

Compared to current practice, FastPress enables scalable scenario analysis (demonstrated in Alpha+) by delivering 100+ overnight static solves where none are currently possible, with consistent, repeatable outputs not available in existing planning tools.

Counterfactuals

The approach to FastPress was chosen because it balances innovation, technical ambition, and operational realism.

Several alternative approaches were also considered and rejected:

• Business-as-usual process optimisation does not address fundamental scalability limits.
• Generic optimisation tools lack integration with gas-specific physics, topology and operational constraints.
• Standalone modelling approaches approximate physical outputs instead of integrating with deterministic simulations and cannot meet operational requirements.
   

Learning and stakeholder engagement

FastPress Beta will build directly on prior NIA projects (equivalent to SIF Discovery and Alpha), that demonstrated the feasibility of core features, progressed these to Proof-of-Concept, and de-risked cloud deployment.

The below lessons learned informed the Beta focus on innovative transient solving and infrastructure timeline optimisation:

• Automation must be analyst-in-the-loop to ensure trust/adoption.
• Algorithmic solving matches (or exceeds) analyst performance while enabling scale.
• Performance varies by scenario type, requiring adaptive AI optimisation strategies.
• Deployment considerations must be addressed early. 

The current solution has been iteratively refined through engagement with NESO via user testing, ensuring it goes beyond incremental innovation and addresses existing pain points.

Readiness levels

This phase represents clear progression:

TRL: 5-6 to 7 (from validated components to system prototype in operational environment)
IRL: 4-5 to 6-7 (from partial to full workflow integration into NESO workflows)
CRL: 2-3 to 4-5 (from no defined BAU adoption to clear route to operationalisation)

Scale

At Beta, deployment within NESO gas planning delivers transformational system-level capability while managing risk, focusing on high-value features to ensure evidenced adoption within the timeframe.

SIF funding

FastPress cannot be funded through price control or BAU mechanisms because it:

• Involves material technical and adoption risk.
• Delivers system-wide benefits, not attributable to a single function.
• Explores AI approaches with uncertain outcomes.

SIF Beta funding is therefore requested to responsibly develop, validate, and de-risk this innovation.

Impacts and Benefits

Financial: future reductions in the cost of operating the network

Total benefits: £88m gross lifetime.

1. FTE cost equivalent of increased productivity

Current position: NESO/NGT incur material operating expenditure on gas planning workflows. Analysts' effort concentrates on manual static/transient scenarios iteration, constraining the number of assessed demand days and network configurations without increasing headcount.

Baseline metrics:

• Analyst hours/scenario
• Number of demand days analysed
• Cost of incremental analyst/FTEs to expand scenario coverage

Forecast BAU benefits: Automation of static/transient solving, flow adjustments, batch runs, and Infrastructure Timeline Optimisation (ITO) increases scenario throughput, freeing analyst time for higher-value planning activities. FastPress expects to deliver productivity gains equivalent to at least 12 additional FTEs, avoiding up to £1.1m/yr (£18m gross lifetime) labour costs.

Benefits already realised: Alpha+ demonstrated that automated static solving achieves analyst-comparable results while also enabling unattended overnight/weekend batch execution (e.g. \>100 scenarios per night), showcasing order-of-magnitude throughput increases and decreased manual effort.

2. Decreased CAPEX spend via pipeline repurposing / avoided new build

Current position: Future hydrogen/CO2 transport networks require significant pipeline investment. Without robust evidence, the default solution is new-build pipeline, driving high capital expenditure.

Baseline metrics:

• Project Union ambition: 2400km of hydrogen transmission pipeline
• Conservative build assumption (20% of ambition): 480km
• New-build cost: £3.74m/km
• Conservative repurposing cost: £0.52m/km

Forecast BAU benefits: ITO identifies candidate assets for repurposing and validates feasibility under capability and resilience constraints. Assuming FastPress enables just 5% of the pipeline requirement (1.2km/yr, or equivalently a 12km section-of-pipe every 10yrs) for repurposing vs. new-build, this conservatively delivers £3.82m/yr (£61.1m gross lifetime) CAPEX savings. Though not quantified, this also supports associated emissions reduction and increased alignment with whole-system energy vector planning.

Benefits already realised: Alpha+ pipe removal and asset importance analysis demonstrated the feasibility of narrowing asset sets for further evaluation, providing evidence of potential CAPEX savings.

3. Reduced decommissioning spend via evidence-based identification

Current position: NGT plans £56.1m of decommissioning spend in RIIO-GT3 to remove redundant assets. However, redundancy classification is uncertain -- e.g. during RIIO-GT2, the Redundant Assets Price Control Deliverable (PCD) reduced from 80 to 70 (12.5% decrease) following updated operational evidence, demonstrating that improved evidence materially alters decommissioning scope/timing.

Baseline metrics:

• Planned RIIO-GT3 decommissioning spend: £56.1m
• Historical redundancy change: 12.5% reduction

Forecast BAU benefits: FastPress enhances the upstream "redundancy check" process via timely and quantified capability/resilience evidence feeding into targeted decommissioning decisions. Assuming a conservative reduction of misclassification-driven scope churn of 20% of the historical reference (i.e. 2.5% of spend), this results in a £0.28m/yr (£4.4m gross lifetime) savings through more targeted, evidence-based decommissioning and reduced mis-timed works.

Benefits already realised: Alpha+ demonstrated the ability to assess asset removal feasibility under different operating scenarios, providing early evidence of potentially more accurate decommissioning plans.

4. Reduced maintenance spend, improved resilience

Current position: NGT plans £180m of "Other Non-Load" spend in RIIO-GT3, of which £123.9m relates to maintainability, network capability and security of supply (excluding decommissioning). Planning uncertainty (e.g. via limited scenario coverage), can lead to precautionary or mis-timed interventions, increasing maintenance, resilience, and asset-health expenditure.

Baseline metrics:

• Planned RIIO-GT3 "Other Non-Load" spend (excl. decommissioning): £123.9m
• Frequency/timing of maintenance- or resilience-driven interventions

Forecast BAU benefits: FastPress directly tests network resilience/capability through broader scenario analysis under asset use restrictions to evaluate resulting network capability reduction. This enables more targeted, sequenced or deferred interventions, avoiding unnecessary maintenance. Assuming a conservative 1% improvement, this equates to £0.25m/yr (£3.9m gross lifetime) savings through improved asset-health and resilience planning.

Benefits already realised: Alpha+ demonstrated scalable capability and resilience analysis across RIIO zones (National Transmission System regional planning groupings) and demand scenarios, providing evidence to inform resilience-driven decisions.