This study presents a routing-and-sizing framework for hydrogen pipeline networks that minimises max regret across uncertain demand. An obstacle-aware genetic algorithm generates corridors over a weighted-GIS surface; a multi-period hydraulic sizing step selects commercial diameters subject to pressure, velocity, and wall-thickness constraints. Decisions are taken in a rolling-horizon so topology and capacity adapt as information arrives. Applied to the UK Humber cluster from 2030 to 2050, built length reaches 165-200 km, with a Spine-First routing strategy averaging 185 km. Least-regret oversizing adds 40 pound m compared to a myopic approach but cuts 2040 worst-case incremental outlay from 260 pound m to < 80 pound m. By 2050, Spine-First achieves 2.07 pound m/km, LCOT 54 pound/kt and regret 12 pound/kt, rivalling a Perfect-Foresight strategy (44 pound/kt; 3 pound/kt). The results show how a short, centrally aligned trunk combined with anticipatory sizing reduces stranded-asset risk and budget shocks, providing a transferable least-regret template for hydrogen pipelines under deep uncertainty.

The housing sector contributes 18% of the UK's greenhouse gas emissions, which poses a challenge for meeting the 1.5 degrees C carbon budget. Key strategies for decarbonisation of the built environment include the retrofit of building efficiency measures, and decarbonised heat sources such as heat pumps and district heat networks (DHNs). Here, we argue that neither of these approaches in isolation is likely to be cost-optimal to meet the carbon budget; rather, a careful balance must be struck between deep retrofit and decarbonisation of heat sources. We used archetype-based housing stock models to predict the heat demand of housing districts for different retrofit scenarios. We then considered the interaction of different levels of housing retrofit with district heat network deployment. GIS was used to map heat demand and plan DHN layout; sizing and costing of the DHN were determined from annual and peak heat demand for different outcomes of retrofit scenarios. A case study was conducted for Sheffield wherein the heat network was supplied by an energy-from-waste facility. We present results illustrating the implications of building retrofit for the techno-economics of the DHN, including net present cost, levelized cost of heat and carbon emissions. The preliminary results indicate that deep retrofit and electrification are preferable to the DHN in both environmental and cost terms; however, more fidelity must be added to the model to interrogate this conclusion. The cost-optimal strategy for meeting the carbon budget likely requires targeted retrofit at different levels for different building archetypes.

Rapid power system decarbonization requires massive renewable investment, with the UK anticipating $\boldsymbol{\sim} \mathbf{4 0}$ billion annually through $\mathbf{2 0 3 0}$. However, traditional models often decouple investment from market feedback, ignoring construction lead times and grid bottlenecks. Here we present an annual closed-loop framework that combines the AMIRIS market simulator based on the UK electricity market with agent-based investors. We find that Mid- Capital Expenditure (CAPEX) total capacity (435.6 GW) exceeds the Low-CAPEX case (371.3 GW) by 2050. This divergence occurs because low costs favor high-utilization offshore wind, whereas higher wind costs shift investor budgets toward cheaper solar and gas, buying larger total capacities. Consequently, high solar penetration causes a “cannibalization” effect, reducing its value factor to $\mathbf{0. 7 4 6}$. Furthermore, increasing annual grid connection limits from 4 GW to 6 GW boosts final connected capacity by 34.8 GW. These results highlight that infrastructure policy and investor diversity are vital for reaching net-zero goals. Future work will integrate capacity markets and policy subsidies to further enhance model realism.

Disruptions to manufacturing processes can result in wasted resources and missed orders, ultimately leading to reduced revenue and increased costs. Siemens estimates that, in 2022, unplanned downtime was responsible for an 11% reduction in turnover for Fortune 500 companies. It is therefore essential to consider the potential for disruptions in planning and scheduling decision making to encourage operational agility and improve resilience. In this study, we present a scheduling framework that incorporates realistic constraints for complex multistage manufacturing processes with process or equipment failures and resource limitations. The problem is formulated as a rolling horizon, with the framework emulating real-time decision making in response to a dynamic scenario, allowing for assessments to be made for key metrics such as cost and total manufacturing time. The framework is applied to a vaccine manufacturing scenario, with a demonstration of the impact of simulated disruptions indicating the frameworks potential for identifying key areas for equipment investment and enhanced maintenance. The impact of material reorder policies during intermittent material availability is assessed, with the results indicating that larger orders placed at times of greater current inventory result in fewer delays and lower order completion times but incur higher costs as a result of unused materials. For the policies investigated, a 62.2% reduction in mean completion times was found to correspond to a 66.3% increase in costs. The framework provides a basis from which future works could make further assessments of process design and operation policy.

Fires from lithium-ion cells are a major heat transfer pathway, promoting thermal runaway propagation and spreading flames to surrounding cells. This study presents new experimental insights into the fire behaviour of lithium iron phosphate (LFP) cells by quantifying radiative heat flux, gas temperature, and cell surface temperature during venting and flaming. The key contributions are: (i) the first spatially and time-resolved measurements of radiative heat flux from single-cell LFP fires, revealing non-uniform radiation with peak intensities varying up to similar to 10 & times; across the horizontal plane; and (ii) a controlled comparison of 18650 and 26650 cylindrical cells, demonstrating fundamental differences in combustion behaviour and peak radiative output. Fully charged cells were heated under constant power to induce thermal runaway, with and without an external ignition source. Three gas emission stages were identified: (1) steady emission post-venting, (2) increased gas production with self-heating, and (3) rapid gas release at thermal runaway. Forced ignition sustained flames in stages 1 and 3 for 18650 cells, whereas 26650 cells maintained combustion throughout. Peak radiative heat fluxes reached similar to 150 kW/m2 for 18650 cells and 350-500 kW/m2 for 26650 cells. These findings quantify the dynamic flammability of off-gases and highlight the importance of spatially resolved measurements for fire hazard assessment.

High penetration of distributed energy resources increasingly creates congestion in low-voltage distribution networks, while local energy markets (LEMs) optimise community welfare without explicitly internalising network constraints. This paper investigates whether a profit-seeking aggregator embedded within a welfare-oriented LEM can partially internalise distribution-level congestion through market participation. We develop a post-clearing, price-protected intervention in which the aggregator injects additional supply and triggers re-clearing, with network feasibility validated using nonlinear AC power flow subject to a non-deterioration constraint on maximum line loading. The mechanism is benchmarked against Distribution System Operator (DSO)-only corrective control and a hybrid regime with residual DSO action following aggregator intervention. Results on a UK LV feeder show that aggregator participation reduces thermal loading and preserves community welfare relative to DSO-only control, though it does not fully restore compliance under severe stress. The hybrid regime achieves the strongest technical performance while maintaining lower welfare loss. Overall, aggregator intervention remains privately profitable, indicating partial incentive alignment.

Renewable Energy Communities (RECs) with peer-to-peer (P2P) trading can lower bills and unlock flexibility, but they also reshape supplier revenue and distribution-network headroom. We couple an agent-based REC market model to a Transformer Headroom Index (THI) to evaluate welfare and grid outcomes across seasons. A 20-household feeder in South-East England with PVs, EVs, and heat pumps is simulated with and without P2P under four retail tariffs: Flat, Economy-7, Agile, and a wholesale-linked Dynamic tariff. Metrics include community net bills, supplier profit, and import/export headroom for 50 and 100 kVA transformers.Results reveal a persistent distributional trade-off: tariffs that minimise community bills compress supplier margins. P2P accentuates this by reducing grid purchases in PV-rich periods. P2P improves import headroom via EV load shifting; however, in PV-rich summer it can synchronise prosumer behaviour around price spikes, concentrating exports into short windows that erode export headroom. Nonetheless, across seasons, P2P is consistently part of the Pareto-optimal solution set, though the preferred supporting tariff varies by season.Policy implications derived from these trade-offs include: the need to re-base residual cost recovery onto fixed or capacity charges, the adoption of facilitation fees and THI-based flexibility contracts for value-sharing, the addition of Dynamic Operating Envelopes (DOEs) as guardrails for real-time tariffs, and the regulatory enablement of EVs as flexibility assets.

Lithium-ion batteries (LIBs) are critical for renewable energy storage, and accurate charge and health estimation remains a significant challenge. Acoustic sensing offers a unique method to observe lithium-ion movement between electrodes during battery operation. However, both the charge state and the internal temperature of the battery affect the acoustic response. This study systematically investigates the interactions between temperature and charge state on acoustic signals through a novel thermal cycle methodology. Using a global sensitivity analysis, we demonstrate that temperature has a non-negligible and dominant effect on the acoustic signal, with a largely insignificant cooperative interaction with charge state. The results reveal temperature-induced variations in the acoustic signal that increase with charge level, though not uniformly. Our findings underscore the critical importance of temperature compensation in acoustic-based LIB estimation techniques. By quantifying the independent and cooperative effects of temperature and charge, this research provides the possibility of independently measuring thermal and SOC effects on the acoustic signal without the need for additional thermal sensing equipment.

Hydropower plants degrade from mechanical wear and stress, reducing reliability and increasing maintenance to sustain high efficiency. Climate change intensifies this by increasing hydrological variability, causing design deviations that reduce performance, particularly in storage-limited run-of-river systems. Periodic turbine replacement offers opportunities for climate adaptation, but traditionally, it replicates the original design, without assessing its performance under future conditions.This study introduces a novel framework for run-of-river plants that explicitly accounts for climatic uncertainties to efficiently identify optimal turbine replacement solutions. Unlike conventional approaches, the framework (i) optimizes turbine configuration, including non-identical turbines, to maximize production and (ii) evaluates performance across a wide range of plausible future climate scenarios. It represents the first future-proof approach to run-of-river hydropower retrofitting. Applied to a real-world plant in Scotland, our framework shows that non-identical turbine replacements, characterized by increased capacity, varied operating ranges, and optimization for net present value, outperform identical retrofits in energy generation, profitability, and financial robustness under climate uncertainty.This framework, and the insights from the case-study application, are particularly relevant at a time when over 20 % of global hydropower units are projected to exceed 55 years of age by 2030. They also highlight the urgency of rethinking approaches to maintenance and retrofitting in hydropower infrastructure beyond run-of-river plants.

Li-ion batteries (LIBs) are integral to modern society, driving the electrification of transport and supporting renewable energy generation to meet Net Zero. However, LIBs suffer from the potential to undergo thermal runaway (TR) which can lead to fire and explosions. Computational modelling of TR is essential to understanding its hazards, and to accurately quantify risks there is a need to account for the uncertainty in TR behaviour. To adequately predict the safe limits of battery operation we incorporate the stochasticity of thermo-physical and kinetic reaction parameters in module thermal runaway propagation (TRP) analysis. A 0-dimension heat transfer model for TRP predictions is validated against experimental findings. From this, Monte Carlo simulations are undertaken to determine the uncertainty in the predicted cell temperatures, times to cell TR and times to TRP. The critical heat dissipation coefficient to prevent TRP considering cell uncertainty was found to be 2.5 and 4.6 times larger for LFP and NMC stacks, respectively, compared to the scenario where cell uncertainty was not considered. For the LFP stack, the less severe TR events mean, in theory, that TRP can be prevented by heat pipe or submersion cooling thermal management systems. Without considering cell stochasticity there is a significant overestimate of TRP time and an underestimate of critical heat dissipation coefficient to prevent TRP. Hence, the predicted safe time for evacuation and appropriate thermal management methods are inaccurate. This work highlights the need to incorporate uncertainty in predictions of risk.

Energy-intensive plants must navigate technology, fuel-price, and policy uncertainty when selecting least-cost decarbonization pathways. We develop a two-stage mixed-integer linear programming (MILP) framework for industrial decarbonization planning that co-optimizes multi-technology capacity expansion and commissioning schedules with hourly site-energy dispatch, while explicitly modeling emissions-allowance procurement and intertemporal banking under an emissions trading system (ETS). The formulation combines multi-decadal investment decisions to hourly operations; includes an allowance-accounting module for ETS-consistent compliance cost calculation with banking; represents correlated trajectories for grid-carbon intensity, fuel and electricity prices, and ETS design; and incorporates time-varying technology costs for electrification, hydrogen, carbon capture and storage (CCS), and bio-energy. We demonstrate the framework on a UK epoxy-resin facility (2025–2055) under five market–policy scenarios to illustrate how scenario-driven stress-testing alters technology choice and timing. Results show that electrification is least-cost/lowest-emissions only if grid intensity falls below 50gCO2 kWh−1 by 2035; a slower trajectory increases cumulative emissions by up to 745ktCO2. A carbon-price corridor alone is insufficient to close the green-fuel cost premium, motivating long-dated hedging instruments (e.g., power purchase agreements and forward/swap positions) to reduce exposure to price volatility. Allowing credit banking enables additional cost-effective abatement (up to 110ktCO2) by valuing early over-compliance, whereas prohibiting banking increases compliance cost and weakens the cost–emissions trade-off. Overall, the framework provides a practical, scenario-driven tool for regulators and operators to evaluate robust industrial decarbonization pathways.

This paper introduces innovative approaches for robust and computationally efficient optimal design of run-of-river hydropower plants. Compared with existing design software, it (1) integrates optimized turbine operations into design optimization instead of following predefined operational rules, and (2) combines this with a regular sampling of the flow duration curve to significantly reduce data inputs. Our rigorous benchmarking demonstrates that (1) operation optimization improves design performance at low computational cost, whilst (2) data input reduction slashes computational costs by over 92% with minimal impact on design recommendations and key robustness analysis insights. Taken together, these innovations make integrated design and operation optimization, complete with in-depth robustness analysis, laptop-accessible. They also reinforce sustainability efforts by minimizing the need for high-performance computing and large associated embodied greenhouse gas emissions.

Lithium-ion batteries, widely used in modern technology, degrade with use, leading to reduced capacity and power output. Monitoring and diagnosing this degradation is essential, and ultrasound has emerged as a potential tool because of its low cost and non-destructive nature. Studies have noted changes in ultrasound behaviour with battery degradation, making it potentially valuable for tracking battery health. However, the behaviour of ultrasound as a battery degrades has been an issue within studies. This paper explores the relationship between state-of-health (SOH) loss and permanent ultrasonic signal changes over 100 charge cycles. A strong correlation was found between SOH reduction, observed to be caused by the loss of lithium inventory (LLI), and shifts in ultrasound signal responses. The analysis of individual peaks within a single acoustic signal showed consistent shifts in time-of-flight (TOF), often trending towards shorter TOFs. In particular, the rate of degradation was not entirely linear, with fluctuations observed across the cycles. These findings suggest that ultrasound can effectively monitor the rate of lithium-ion battery degradation. Future work could expand on these results by inducing varied degradation conditions and cathode chemistries to determine specific TOF shifts, enhancing detection methods for different degradation mechanisms in lithium-ion batteries.

Adding to the empirical body of evidence on residential flexibility responsiveness, this paper presents a field experiment utilizing social norms and pecuniary incentives to promote residential load shifting among households in the United Kingdom, occurring from spring 2024 to spring 2025. The experiment is a randomised controlled trial, where recruited customers were randomly assigned to an intervention and control groups. A third group of non-participants consumers, shielded from the observation effect and the selection bias of the recruitment, is also considered. Four treatment effects, each measuring a desirable outcome of interest are introduced. A purposive sample of the intervention and control groups took part in semi-structured interviews one to two months after the launch of the field experiment. This provided qualitative data on consumer motivations, barriers and enablers as part of a mixed methods approach to explain the resultant energy usage behaviour. Using a set of quantitative methods for causal inference, we examine if normative information can be inducive of a reduction in peak consumption. We assess if the effect of the monetary incentives is crowded out by the normative ones when used simultaneously. A thematic analysis of the qualitative interview data indicated lifestyle factors that could influence behaviour such as predictability of routines and level of control over when energy-consuming tasks could be performed. It also highlighted influential features of information provision such as the frequency of the challenges set by the energy provider and how performance data could be presented to be meaningful to them. Finally, a technical-economic study was carried out to simulate the effect of mass adoption of this scheme on the scale of the electrical system.

The emergence and design of hydrogen transport infrastructures are crucial steps towards the development of a hydrogen economy. However, pipeline routing remains underdeveloped in hydrogen infrastructure design models, despite its significant impact on the resultant cost and network configuration. Many previous studies assume uniform cost surfaces on which pipelines are designed. Studies that consider a variable cost surface focus on designing candidate networks rather than bespoke routes for a given infrastructure. This study proposes a novel multi-stage approach based on a graph-based Steiner tree with Obstacles Genetic Algorithm (StObGA) to route pipelines on a complex cost surface for multi-source multi-sink hydrogen networks. The application of StObGA results in cost savings of 20-40% compared to alternative graph-based methods that assume uniform cost surfaces. Furthermore, this publication presents an in-depth methodological comparative analysis of different pipeline routing and sizing methods used in the literature and discusses their impact. Finally, we demonstrate how this model can generate design variations and provide practical insights to inform industry and policymakers.

Li-ion batteries up to the MWh capacity are increasingly adopted in marine applications, wherein the fire, explosion and toxicity hazards of thermal runaway (TR) events present a unique and complex problem relative to other applications. As such, to perform a critical risk assessment of these hazards this work analyses past incidents. Short circuits related to water ingress and coolant leakage have been the most prominent cause of TR while the majority of TR incidents led to fire or explosion. HAZID analysis was carried out on the battery system, the battery space and the electronic system. Risks were significantly reduced by considering transferable technologies (from automotive and stationary storage sectors) and future technologies. Bow-tie analysis was used to assess the barriers along the threat-consequence pathways of an electrical abuse event leading to the TR hazard and a TR event leading to the battery space failure hazard. The analysis showed that the consequences of battery TR significantly increase if it leads to battery space failure as complete loss of capability, dangers to passengers, and complete ship loss can occur. Further quantitative assessment of proposed improvements is required to determine their effectiveness in hazard reduction for ongoing safety developments.

District heating networks (DHNs) are essential for decarbonising urban heating but face adoption barriers due to stakeholder complexities, infrastructure legacies, and consumer resistance. Existing models overlook the dynamic, iterative nature of infrastructure development and multi-stakeholder interactions. To address these limitations, this study develops an agent-based model (ABM) that continuously simulates DHN adoption and expansion. Unlike discrete optimisation approaches, this integrates evolving household decision-making, project developer strategies, and dynamic market conditions. Socio-economic factors, spatial dynamics, and temporal evolution are incorporated to provide a comprehensive understanding of DHN diffusion. The model projects that adoption in the studied urban area could reach 30.4% +/- 2% of households by 2050. While this projection exceeds the UK's national target of 20% for DHN adoption, it highlights the significant potential to contribute to national decarbonisation goals. Heating bill savings range from 41.9% to 56.3%, while CO2 emissions are reduced by 24.3% compared to existing gas heated systems. Through the quantification of temporal dynamics and multi-stakeholder interactions, this study offers insights for urban energy transition strategies. These findings can inform policies that accelerate DHN adoption, mitigate risks associated with large-scale infrastructure investments, and contribute to sustainable urban heating strategies aligned with international climate goals.

Small hydropower plants (SHPs) present an eco-friendly and economically viable alternative to conventional dam-based plants. With only 36% of their worldwide capacity currently tapped, there is potential for substantial global expansion including in industrialised nations. Most SHPs follow run-of-the-river (RoR) scheme, depending on the fluctuating flow of rivers because of their negligible storage capacity. They are deployed in a world characterised by a changing hydro-climate and unpredictable socio-economic evolutions. Due to their inability to regulate discharge fluctuations as well as their dependency on selling energy at higher rates, these plants are significantly vulnerable to these changes. Design alternatives are generated through traditional approaches relying on cost-benefit analysis that use past hydroclimatic conditions and disregard operational considerations, without assessing investment robustness in the face of changes. What is more, optimization and robustness analysis of these systems typically require a significant amount of computing time and resources necessitating high performance computing.We introduce a new framework for robust hydropower system design to address these issues. This framework uses and extends HYPER, a state-of-the-art toolbox that computes technical performance, energy production, maintenance and operational costs of a design. It combines HYPER with many-objective robust decision making (MORDM) to define robust alternatives. Our implementation involves a systematic four-step process: (1) Introducing a two-objective formulation to identify design parameters balancing cost and revenue. (2) Creating alternative futures by sampling deeply uncertain factors, encompassing socio-economic (electricity prices, interest rate, cost overrun) and hydroclimatic factors (median, coefficient of variation, the 1st percentile of flows). These streamflow statistics are then transformed into flow duration curves using an innovative approach. (3) Robustness quantification of alternative designs using two newly introduced financial robustness metrics based on the probability of making the plant financially viable. (4) Identify the most critical parameters influencing robustness through sensitivity analysis and scenario discovery. We then employ a computationally efficient approximation approach to streamline resource-intensive steps in optimisation and robustness analysis.Results indicate that employing the MORDM approach in the design of RoR hydropower plants offers valuable insights into the trade-offs between cost and revenue, while supporting design with a range of viable alternatives aiding in the determination of the most robust and reliable design. Maximising the benefit cost ratio yields more robust and financially viable solutions than maximising NPV, as it leads to less costly designs that generate slightly less revenue on average but tend to better exploit low flows. Traditional design approaches employing identical turbine configurations and focusing on NPV maximisation, have proven to be less effective when compared to designs incorporating non-identical turbines. Moreover, such designs have demonstrated greater vulnerability to climate change, primarily attributable to their less flexible configuration. Combined optimization and robustness analysis of a RoR design, initially taking 120 hours, is also made computationally inexpensive through a novel method involving strategic data input reduction. This innovation resulted in a significant 95% reduction in processing time, while maintaining nearly identical outcomes in both steps. An open-source Python version of this methodology is scheduled to be available by July 2024.

Hydropower stands out as an economical, reliable, sustainable, and renewable source of energy. It has been the leading source of renewable energy across the world, generating more than 15 % of total electricity in 2022. Therefore, it will likely play a crucial role as the energy system shifts towards a carbon-free future. The turbine system is at the heart of the hydropower plant and converts flowing water into mechanical energy. Remarkably, around 154 gigawatts, or one-fifth of the installed hydropower turbines, will be more than 55 years old by 2030 globally. Modernising these aged turbines is essential for sustaining optimal plant performance and this will create opportunities to retrofit hydropower facilities to improve their adaptability to changing hydrological conditions. A well-defined methodology is necessary to evaluate feasibility and select optimal solutions for upgrades. This study addresses this critical necessity in the context of run-of-river (RoR) hydropower plants with the HYPEROP toolbox to efficiently evaluate and choose optimal turbine replacement or upgrade options. HYPEROP provides operational optimization capabilities coupled with design flexibility and expanded simulation features for complex turbine configurations. It facilitates the selection of turbine systems featuring large and small turbines. The effectiveness of this toolbox is illustrated through the case study of the Bonnington RoR hydropower plant, commissioned in 1927 on the upper reaches of the River Clyde in Scotland, United Kingdom. Bonnington RoR features a pair of two identical Francis turbines, each designed for a discharge of 12 m³/s and equipped with an installed capacity of 5.5 MW.Our analysis indicates that, by prioritising Net Present Value (NPV) maximisation through a single objective function and considering historical discharge records, HYPEROP offers a novel configuration featuring non-identical Francis turbines with design discharges of 16.13 and 9.13 m³/s. Optimal design increases power production by approximately 3.4 GWh (~7 %) annually by providing operational flexibility and retaining high efficiency over a range of discharge values. The optimal design yields an NPV of approximately 3 million dollars (USD), factoring in the additional energy increase as revenue, turbine replacement cost, and lifetime operation cost. The payback period for this investment is projected to be 15 years when considering only the additional energy as revenue. It's worth highlighting that the optimised design notably outperforms the current configuration, particularly in response to variable streamflows, including both high and low flows. Therefore, optimal design is expected to be less vulnerable to climate change due to higher efficient configuration.

To meet future resource requirements for the uptake of bioenergy with carbon capture and storage (BECCS) technologies to meet Net Zero targets, a range of biomass feedstocks are required to ensure the security of supply, utilise waste materials, and promote the circular economy. This study investigates the potential for blending forestry/agricultural residues and waste wood products with woody biomasses in combustion-based Power-BECCS, using a process model developed in Aspen Plus and validated against literature data. The base case assessment highlights the key performance indicators (KPIs) and energy penalty associated with CCS for next-generation BECCS plants. The results of a comparative study show the impact of blending various biomass species on plant KPIs. This research provides the basis for decision making on feedstock selection and optionality. Several alternate fuels produce similar KPIs to the base case, in some cases generating more net power (Fuels B, D, E, F, H, and I) or capturing more CO2 (Fuels C and G). Overall, blending biomass fuels is a promising option to utilise alternate feedstocks, improve plant performance, and enhance the Carbon Dioxide Removal (CDR) potential.