Probabilistic deployment pathways of scaling up distributed green hydrogen systems for urban residential communities in North America

Probabilistic deployment pathways of scaling up distributed green hydrogen systems for urban residential communities in North America | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Probabilistic deployment pathways of scaling up distributed green hydrogen systems for urban residential communities in North America Lexuan Zhong, You Wu This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-4327177/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 02 Jan, 2025 Read the published version in Nature Cities → Version 1 posted You are reading this latest preprint version Abstract In the context of the firm and enthusiastic development of renewable-based distributed energy systems, high-profit household collaboration strategies are widely recognized as essential for scaling up decentralized green hydrogen systems in urban residential communities. Here we develop bottom-up energy models linking climate, human behavior, and community characteristics to assess the cost-effective impacts of system deployment pathways on community green hydrogen systems for 7 North American climate zones in the 2030 ~ 2050 periods. Despite lower energy costs in moderate climate zones compared to hot and cold zones, a consistent pattern in deployment pathway impacts on costs is observed across all zones. The study underscores the critical role of selecting the right deployment pathway for urban decarbonization, with potential cost discrepancies of up to 60% between optimal and suboptimal options. Furthermore, energy storage demands significantly influence energy costs, emphasizing the need to prioritize increased energy storage in pathway design. Physical sciences/Energy science and technology/Renewable energy Earth and environmental sciences/Climate sciences/Climate change Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Introduction According to the International Energy Agency (IEA), the global energy sectors need to achieve net zero emissions by 2050 to reach the target of limiting global warming to 1.5°C in 2100 1 . Currently, about one-third of global power is consumed in cities 2 . Rapid deployment of clean energy technologies is a core pathway to achieve urban energy transition from fossil fuels to cleaner energy sources 3 . With surging energy demand caused by accelerating urbanization, growing evidence shows that planning large-scale plant-central renewable energy production outside the cities faces enormous pressure from technically feasible lands due to their limited renewable energy density 4,5 . Thus, distributed energy systems that support integrating renewable technologies will effectively reduce external energy demand from urban residential communities, minimizing the land pressure of urban renewable energy infrastructure 6 . In addition, a large body of research supports the notion that communities with coordinated action provide critical spatial, temporal, and social pathways for energy integration to realize higher economic, environmental, and social interests of distributed energy systems 7,8 . Therefore, an accelerated market introduction of decentralized energy systems into urban communities is considered a pivotal option to utilize inner-city resources for decarbonizing cities. Distributed energy systems powered by renewable sources depend on cost-effective energy storage technologies to address the severe energy mismatch caused by high homogeneous production and demand in urban residential communities 9 , requiring high efficiency for short-term extreme energy events and high capacity for cross-seasonal energy patterns 10 . Growing research shows that the single battery technology as a cross-seasonal storage option may not be economically viable in long-term resilience systems due to their low energy density and high leakage rate 11 . Green hydrogen systems, a distributed energy system with green hydrogen and battery as storage options, offer higher storage flexibility and lower storage costs, making them more suitable for long-term community sustainability 12 . We consider green hydrogen systems a promising community energy solution that needs to be evaluated in detail. Energy models are widely used to understand the impacts of energy systems on residential communities and to inform potential development directions to energy planners 13 . However, academic and policy spheres highlight concern that deep uncertainties of energy input are fundamental modelling limitations for energy models, causing inappropriate assessment and adversely impacting policy targets when ignoring modelling limitations 14 . Part of energy input uncertainties stems from both climate conditions and human behaviors. In future periods, energy models will confront heightened challenges associated with the evolving climate conditions that result from continuing climate change 15 . On the other hand, these models exhibit limited incorporation of historical data-driven methods to measure potential human behavior 16 . Thus, how to effectively expand input scenarios for energy models becomes core to understanding the potential fluctuations of energy input stemming from uncertainties associated with climate-human systems. The sampling method 17 and generative adversarial network (GAN) method 18 are adapted to generate a substantial number of energy input scenarios for modeling potential climate-human uncertainties. Besides climate-human systems, another significant source of uncertainties emerges from the community design layer, including community compositions and renewable energy levels. Community compositions account for different residential building types and their corresponding numbers within residential communities, while renewable energy levels represent the deployment numbers of energy systems within residential communities 19 . These two factors impact the diversity of energy input scenario types for energy models in the design phase, influenced by various geographical, sociological, and psychological considerations. Current studies have considered that the uncertainties of the community design layer significantly impact on the overall economic and environmental benefits of the community 20,21 . Thus, energy models must integrate detailed community design layers into energy planning. However, meeting this requirement poses a dilemma for energy models, navigating the delicate equilibrium between spatial dimensions and calculating tractable. To strike balance between the diversity of household energy profiles with the robustness of urban energy planning, historical data-driven energy input scenarios 23 and city-related virtual communities 24 are developed. System deployment pathways of green hydrogen systems in communities determine different household participation forms, highly affecting shared benefits and participation motivation among interconnected households 25 . Thus, system deployment pathways are a centerpiece of understanding the full impact of green hydrogen systems on decarbonizing communities in the strategic energy planning phase. We reckon that such multiple-source uncertainties of energy models are particularly relevant for analyzing the impact on the system deployment pathways due to energy input uncertainties undoubtedly that affect the model output results of system deployment pathways. The multiple-source uncertainties in communities include climate-human uncertainties at the household level and community design uncertainties at the community level. In that context, quantifying these uncertainties into energy models presents a considerable challenge due to the complexity of multidimensional impacts. We consider that green hydrogen systems are at a nascent stage. Whether and how green hydrogen systems can expand fast enough in uncertain communities remains unclear. Thus, systematically analyzing potential expansion pathways of green hydrogen systems through high-fidelity energy models is essential to facilitate rapid deployment and unleash its potential for community energy transition. In this article, we introduce a bottom-up energy model linking climate, human behavior, community archetypes, and energy system models, aiming to ensure cost-effective deployment pathways of green hydrogen systems for highly diverse urban residential communities. Figure 1 depicts a summary of three interconnected phases of the proposed energy model. First, we developed a multi-scale combined uncertainty approach to obtain a set of community energy production and demand profiles, including detailed household energy information with potential occurrence probability to reflect inevitable uncertainties from household and community layers. Next, we formulated three system deployment pathways based on an urban planning perspective to characterize strategies for household collaboration within the community. These pathways were merged with community energy profiles to serve as community inputs to the energy system model related to green hydrogen systems. Lastly, the robust energy system model yielded a collection of optimal system designs that reflect the cost-effective impacts of deployment pathways on green hydrogen systems under multi-source uncertainties. We extensively investigated seven North American climate zones to understand the general and overall effects of deployment pathways on community energy system design. Results Multi-source uncertainty quantification of energy input scenarios at the North American scale The IEA report points out that advanced economies need to take the lead and reach net-zero emissions earlier to buy more time for the energy transition in developing economies 1 . However, the efforts undertaken by the United States and Canada in North America fall far short of what is needed to achieve net zero global energy-related carbon dioxide emissions by 2050 26 . These developed nations need to take further action to assess the cost and availability of green hydrogen systems as "emerging" technologies for widespread commercial deployment. The United States and Canada are divided into eight building climate zones, and most of their cities are located in the previous seven climate zones. Cities within each climate zone do not exhibit significant differences in building and energy regulations. Thus, we selected one sample of cities from each climate zone to capture the typical regional energy performance. The green hydrogen systems in these cities were evaluated using an energy model. The standard approach to obtaining dynamic energy input scenarios for energy models is to use typical urban stock community parameters, such as climate patterns, human activities, building characteristics, etc. However, this approach often overlooks the compounded uncertainty impacts from both the household and community levels. Neglecting these impacts may lead to an energy information gap between computed energy performance and reality. This study introduces multi-source uncertainties into bottom-up building energy modelling layer by layer depending on their impact pathway in the energy calculation process. Thus, we begin by quantifying the impacts of climate-human uncertainties on energy input scenarios at the household level (details about climate-human uncertainties are presented in Method). Figure 2 . shows the impact of climate-human uncertainties on energy input scenarios in 7 cities taken from climate zones 1 ~ 7. Climate and human uncertainties show a more uneven and significant effect on the energy demand side rather than on the energy production side from the median and interquartile range perspective. On the energy demand side, the annual median household energy demand exhibits a valley-shaped pattern across all climate zones, with varying interquartile ranges. Vancouver demonstrates the lowest median energy demand with a narrower interquartile range at the bottom of the valley, while Miami and Edmonton exhibit higher median energy demand values and more comprehensive interquartile ranges at the top of the valley. On the energy production side, the median annual household energy production decreases as climate zones progress backward, with a nearly uniform interquartile range. Miami and Edmonton have the highest and lowest energy production performance, respectively. The variation in urban building archetypes results in higher energy demand but similar energy production within these observed cities. The median annual energy demand of bungalows is approximately 5,000 kWh lower than that of two-story buildings, yet the annual energy production exhibits minimal variance between the two types of structures. We then superimpose the impact of community design layer uncertainties onto the energy input scenarios at the household level using a community characteristic matrix assembled from household scale and prosumer scale (detail about community design layer uncertainties is presented in Method). The household scale represents the clustering process of average and boundary community energy scenarios from previously obtained input scenarios, while the prosumer scale denotes the labeling process of independent households as prosumers or consumers within obtained community samples. Figure 3 . shows the impact of community design layer uncertainties on the energy input scenarios at the community level, depicted using the sharing level indicator. The sharing level reflects the expected potential for energy self-sufficiency of independent households within community samples. Results show that annual median sharing levels in all community samples exhibit a ridge-shaped pattern across all climate zones. Vancouver shows the highest sharing levels at the peak of the ridge, while Miami and Edmonton exhibit lower sharing levels at the base. In addition, sharing levels within the cities show a consistent decreasing trend among the three types of community samples, with bungalow community samples exhibiting the highest sharing levels, followed by average community samples, and finally, two-story community samples. This pattern suggests that energy demand plays a dominant role in shaping sharing levels across various climate zones, regardless of energy production levels, such as the higher energy production in Miami and lower energy production in Edmonton. However, we also observed that the annual peak-sharing levels vary among cities. Despite this, the peak value in moderate climate regions is higher than in extremely hot or cold climate regions. This suggests that, within urban communities, the climate zones to which the city belongs exert a more substantial influence on sharing levels than the internal community archetypes. Impact of system deployment pathways on levelized cost of electricity The green hydrogen systems in North American cities were assessed using an energy model that provides all the non-dominant sets of optimal system designs for initial commercial-scale deployment via Pareto front analysis. Our analysis centers on how different deployment pathways of energy systems affect energy affordability of urban communities, measured as levelized cost of electricity (LCOE) calculated from compromise solutions of system design options for each group. The primary approach for decision-making from the Pareto front utilizes the Euclidean-distance-based method 25 . LCOE provides a levelized present price per unit energy value, making it suitable for comparing differences in the cost of system power generation across various urban residential sectors and deployment pathways. Figure 4 shows the impact of three types of deployment pathways on potential changes in LCOE in the eight selected cities. Two key insights emerge. First, as observed from a horizontal perspective, household distributed program (HDP) follows a small upward trend as the prosumer scale increases within an equivalent community scale in all urban community cases. In contrast, household-centralized programs (HCP) and community-centralized programs (CCP) show the opposite trend, decreasing significantly. This is due to the fact that, with more household participation, HDP faces resource congestion caused by household competition, resulting in greater costs, while HCP and CCP can fully utilize the scale effect of household cooperation to reduce costs. Additionally, the similarity in the LCOE change trend across all cities indicates that climate zones have a trivial impact on the ways in which the three deployment pathways affect LCOE, although LCOE tends to be lower in mild climate zones compared to hot or cold climate zones. Furthermore, the increased rate of LCOE in HDP has remained relatively stable as the prosumer scale increases, while the decreased rate of LCOE in HCP and CCP gradually slows down. This indicates that despite increased household participation, the increased resource congestion in HDP does not accelerate cost deterioration, while the advantages of scale economies in HCP and CCP gradually diminish. Second, as observed from a vertical perspective, the LCOE of different community scales within each deployment pathway decreases in the order of OB, Ave, and TW community scales across all urban community cases. This reveals that urban communities with high sharing levels bring more cost advantages regardless of climate zone and deployment pathways. Additionally, in all cities except Edmonton, HDP demonstrates a greater cost advantage compared to HCP and CCP at a 20% prosumer scale under the same community scale. With a prosumer scale exceeding 40%, both HCP and CCP demonstrate a cost advantage. Notably, HCP slightly outperforms CCP in most urban community scales, with both reaching parity at a 100% prosumer scale. The subtle difference shows that consumer scale can contribute to the cost burden, which decreases as consumer scale decreases. HCP shows better cost advantage than HDP and CCP at low prosumer scale in certain cities, such as Vancouver, Las Vegas, and Houston. However, the cost benefits diminish as more households with low sharing levels join. This is attributed to the fact that consortia of households with high sharing levels can readily export energy for profit through scale effects, thus supplementing energy costs. Therefore, it can be concluded that HDP and HCP approaches are suitable for low-participation communities, while HCP and CCP are more appropriate for high-participation communities. Analysis of system design for resilient urban community The primary driver of LCOE is the initial capital investment derived from the optimal system design identified by the energy model. Analyzing the variations in energy design across different pathways is crucial for understanding how these pathways drive changes in LCOE from the system design perspective. Here, we now examine the impact of deployment pathways on the system design to account for our previous findings on LCOE changes. Figure 5 illustrates the potential changes in four system design parameters under three types of deployment pathways across the eight selected cities. First, we find that the drivers of LCOE change are not the same on the same consumer scale. HDP keeps system design parameters essentially unchanged as the prosumer scale increases. This shows that the slight increase in LCOE stems from continuous external resource congestion rather than a significant change in system investment costs. In contrast, both HCP and CCP significantly expand hydrogen tanks compared to other design parameters as the prosumer scale increases. This suggests that the continuous decrease in LCOE is driven by the substantial reduction in system investment costs, as cheap hydrogen tanks replace a significant portion of batteries in energy storage as the scale increases. Therefore, it can be concluded that the hydrogen tank becomes a primary design factor in driving the scale effect of HCP and CCP. Second, we notice that higher energy storage demands lead to a higher hydrogen storage ratio, leading to lower LCOE when comparing system design parameters at the same community scale across the three pathways. This is because the advantage of scale in hydrogen costs offsets the drawback of linear increases in battery storage costs as energy storage demand increases. In the initial phase, both HDP and HCP demonstrate higher energy storage demand compared to CCP, resulting in a higher hydrogen storage ratio and better LCOE performance. However, as the prosumer scale increases, the higher storage demand leads to an inevitable growth in the hydrogen storage ratio for both HCP and CCP. Meanwhile, the storage structure limitations of HDP result in its hydrogen storage ratio remaining nearly constant, ultimately causing it to be quickly surpassed by HCP and CCP in LCOE performance. This can be further justified by the trivial difference across different community scales within the same pathway, where the hydrogen storage ratio decreases slightly along OB, Ave, and TW due to the decrease in energy storage demand. Discussion Current energy models typically focus on urban and national scales, overlooking the influence of local private entities such as households and communities on energy markets 27 . However, the decentralized nature of distributed energy systems grants more agency to these entities in participating in the energy market 28 . This oversight may result in significantly lower market participation scales than those anticipated by macro energy models, posing a risk to the achievement of energy transition targets. In this work, we built a bottom-up energy model that established links between energy systems and local urban communities, which incorporate necessary technological details related to climate, human behavior, building archetypes, and energy system characteristics. This energy model is applied to the role of system deployment pathways in developing cost-effective green hydrogen systems for diverse urban residential communities from a design and operation perspective. The evaluation results will complement critical market participation information to inform fully potential expansion pathways to energy planners. Our study shows that differences in community sharing levels influenced by climate zones and community type significantly impact the overall value of energy costs rather than their change trend affected by deployment pathways. Energy costs are lower in moderate climate zones than in extremely hot or cold climate zones. Additionally, community samples with high sharing levels consistently exhibit lower costs relative to those with low sharing levels at the deployment pathway. These findings underscore the potential for communities characterized by high sharing levels and moderate climate zones to attain optimal energy costs. Our study underscores the critical importance of selecting the right deployment pathway for achieving cost-effective decarbonization in urban communities, with impacts consistent across all climate zones. The cost differential between the best and worst pathways for the same community scale can be as high as 60%. A household centralized program stands out as the preferred deployment pathway for most communities, consistently driving cost reductions by harnessing scale effects. In communities with low prosumer scale, household distributed programs may demonstrate better cost advantages initially, but as prosumers increase, resource competition escalates, leading to higher energy costs. Conversely, community centralized programs are better suited for communities with high prosumer scales, where consumer scales always significantly negatively impact energy costs. Our study shows that variances in energy storage demands across different deployment pathways serve as the principal determinant of their respective energy cost trends. In green hydrogen systems, increased energy storage demands trigger a scale effect in hydrogen storage costs, consequently mitigating system investment costs. Our findings indicate that household centralized programs exhibit higher energy storage demand than household distributed programs and community centralized programs, demonstrating greater cost advantages in most communities. This underscores the critical importance of incorporating more prosumers in the design of deployment pathways to expand energy storage demand. There are several essential avenues for expanding upon this work. Extending our research to include future climate periods beyond 2050, such as those projected for 2050–2100 under Representative Concentration Pathway (RCP) 8.5, would offer valuable insights into the long-term performance of various system deployment pathways for green hydrogen systems 15 . This extended analysis would help anticipate and address the challenges posed by further worsening climate change, including ongoing shifts in climate patterns and increased frequency of extreme climate events. By examining deployment strategies across a broader range of future climate conditions, we can better understand their cost-effectiveness and resilience, ultimately informing more robust and adaptive energy transition strategies. Additionally, expanding discussions on developing green hydrogen systems to include social justice and ethical concerns is crucial, as energy policy and technology decisions impact communities in diverse ways beyond technical feasibility and cost-effectiveness 29 . Energy inequalities arising from the energy transition can exacerbate disparities and hinder access to opportunities for specific communities within the energy market 30 . Future studies can integrate social concerns into energy models to enhance our understanding of the deployment pathways needed to support equitable and inclusive urban energy transitions, ensuring that the benefits of green hydrogen systems are accessible to all urban communities. Finally, the energy models in this paper can be used to expand other renewable energy sources to suit local community conditions, such as increased adoption of wind, geothermal, or biomass. Methods Analytical approach This study aims to explore the impacts and implications of system deployment pathways on deploying cost-effective green hydrogen energy systems for different urban communities. Therefore, energy system design should cover all potential community energy input scenarios subject to multiple uncertainties to ensure robust and optimal outcomes. However, climate change exhibits a distinct probabilistic nature, i.e., extreme climate conditions are significantly less likely to occur than typical climate conditions. The energy input scenarios should reflect this variability in weather probabilities to ensure reasonable realism. Therefore, we use stochastic optimization to formulate and integrate the objective function into the energy model 31 . In this process, each energy scenario is treated independently, and then their results are combined based on the probability of occurrence to obtain the mathematical expectation of the outcome. The overview of modelling multi-source uncertainties Energy models are subject to inherent input parameter uncertainties derived from complex real-world communities. It is crucial for energy models to transparently disclose the impacts of community uncertainties on energy system inputs and provide reliable energy input scenarios for subsequent evaluation. Parameterizing the compound impacts of these uncertainties on energy input scenarios for the energy model is challenging due to the complexity of sources and the differences in their influence pathways. Thus, we distinguish the community uncertainties into two levels: climate-human uncertainties at the household level and community design layer uncertainties at the community level (Supplementary Notes 1). As a bottom-up energy model, we quantify uncertainties layer by layer from the household level to the community level and superimpose them to quantify compound impacts. Modelling climate-human uncertainties To parameterize the climate-human uncertainties at the household level, we distinguish climate and human systems into two treatments based on their different sources. We then associate them into compound scenarios to account for their superimposed impacts on energy input. Climate uncertainty stems from the unpredictable nature of climate period paths and their corresponding extreme weather events, leading to the loss of potential risk information on weather-dependent energy production and demand in defined climate scenarios 32 . Current climate change further increases uncertainty, exacerbating climate risk to energy systems due to the increasing frequency of extreme climate events and ongoing shifts in climate patterns 33 . Thus, we developed a method to synthesize future representative weather datasets based on regional climate models (RCMs) data to capture potential climate uncertainty information related to weather patterns and extreme weather events (Supplementary Note 2). This method can derive an arbitrary number of stochastic hourly-resolution weather data to decrease the simulation time without destroying month-scale weather features. Thus, this method maximizes our ability to generate enough weather data to measure climate uncertainty information and then trims these data to representative weather scenario sets with occurrence probabilities, which facilitates compatibility with stochastic energy system optimization. Human uncertainty comes from the complex stochastic nature of human behavior and their interaction with energy systems, in which the greater control freedom for the residential occupants makes the uncertainty of energy demand profiles even more significant. Deterministic occupant behavior variables not only lead to overestimating or underestimating energy demand profiles at the household level but also decrease the diversity of energy demand profiles at the community level. Thus, in human systems, we sample human-related operational information from assumed probability distributions and parameter ranges based on urban statistical data, ignoring the behavioral logic of the occupants 17 . We explore seven aspects of human behavior uncertainties that affect building energy demand profiles. These features encompass the full range of human-involved operational parameters in building energy simulations. Due to the aleatory nature and weak relationship of uncertainty from the approach, we specified independent probability functions and then drew random samples from these features using Latin Hypercube Sampling (Supplementary Note 3) 34,35 . We specify uniform distribution as probability functions to equal probabilities of each interval in uncertainty variables because there is no explicit reason to value one probability distribution over another. Meanwhile, we extracted the boundary of these variables from the urban residential energy surveys 36 . Every random human behavior scenario was then associated with each climate scenario to create a compound scenario pool to treat the uncertainties in both climate and humans properly. In this study, we created eight sets of composite climate-human uncertainty scenario pools in North America (climate zones 1 ~ 7), where each scenario pool yielded a total of 50,000 household-level production and demand profiles, i.e., 50 climate scenarios and 1,000 human scenarios. Modelling regional household benchmark buildings Besides climate-human scenarios, an urban residential archetype is crucial in generating household energy scenarios using dynamic energy simulations. We extracted physics-based benchmark building models based on local urban information: urban building survey data and building standards of climate zones (Supplementary Note 4). In this study, we created a set of benchmark bungalows and two-story houses in each city for aggregating artificial urban communities. We combined building information modeling (BIM) and building energy modeling (BEM) to generate household energy scenarios systematically. Specifically, BIM crafts benchmark archetypes with Level of Detail-2 (LoD2), incorporating fundamental construction information such as building typologies, floors, roofs, external walls, and windows, all sourced from urban building surveys. Concurrently, BEM analyzes building energy output by associating benchmark archetypes and climate-human scenarios. This process incorporates energy characteristic factors like envelope construction, internal loads, HVAC systems, and renewable energy systems, with specifications adhering to building standards within the local climate zone. Once completed, the EnergyPlus engine executes energy simulations to analyze hourly-resolution dynamic energy performance. Modelling community design layer uncertainties Parameterizing uncertainties within the community design layer in bottom-up energy models entails creating detailed community models that encompass the full array of urban community features. Extracting community samples from cities to parameterize uncertainties falls into a dilemma between scenario scopes and feature completeness due to the intricate nature of urban community features. Due to the lack of boundary samples, the representative samples derived from regional-average community features may not provide complete uncertainty information for extrapolation to all urban communities. Thus, we create a set of artificial communities using community characteristic matrix: household scale and prosumer scale. The household scale quantifies uncertainties in community composition using boundary and mean urban community features, while the prosumer scale accounts for uncertainties in renewable energy using linear scale expansion. These matrix parameters define the boundary and node samples to mimic community feature information, which provides fully quantitative information on community design layer uncertainties. In this study, the community scale was assumed to be 50 households. Each community characteristic matrix for the local climate zone considered three types of household scales and five sets of prosumer scales. The community design layer uncertainties were parameterized in two sequential processes. Initially, we clustered aggregate household energy scenarios into designed community energy scenarios. Subsequently, we assigned household roles as either prosumers or consumers within these community energy scenarios. In the first process, we translated each type of household energy scenario into corresponding shares of community energy scenarios based on household scale using k-means clustering. Subsequently, in the second step, to mitigate the impact of the direction of renewable energy penetration, we arranged households within the community in descending order of energy sharing levels. This assumed that households with higher levels of sharing were more inclined to adopt renewable energy systems. Next, we categorized households into prosumer and consumer roles based on prosumer scale. Finally, this pipeline of uncertainty propagation returned a total of 750 community energy scenarios, i.e., 50 community scenarios under per-community features. The community energy profiles of all communities were aggregated to return 15 profiles for the entire city. Modelling system deployment pathways From a government planning perspective, we model three system deployment pathways that coordinate households into energy community, including household distributed programs, household centralized programs, and community centralized programs 37 . These potential community collaboration programs govern energy allocation and trades to ensure effective and equitable energy flows among households in their local communities. This requires that energy models include energy dispatch strategies and pricing mechanisms under different system deployment pathways. To distinguish the roles of households clearly in energy models, we defined intra-household layers and intra-community layers based on control logic relationships rather than physical relationships. The intra-household layers refer to energy interactions between modules within the system, while the intra-community layers refer to energy interactions between different systems. Consequently, energy dispatch strategies are present in both layers, whereas pricing mechanisms are exclusive to the intra-community layers. We detail energy dispatch strategies using a two-stage logical framework which defines energy flow structure in intra-household layers and energy sharing structures in intra-community layers from the rational household perspective (Supplementary Note 5). In this study, our analysis of multiple deployment pathways focuses on intra-community layers that reflect collaborative solutions in local community energy markets. Thus, each deployment pathway features the same energy flow structure but differs in energy sharing structure. The energy flow structure in three deployment pathways identifies twelve operating states of the intra-household layer to manage energy flow hierarchy in system modules under varying module capacity and power states, with the goal of eliminating module interference. To exploit the storage advantages of the green hydrogen systems in large storage capacity and long timescale, we managed the battery module as a short-term energy storage option and the hydrogen module as a long-term energy storage option. Energy sharing structures are used to govern the energy trade methods between prosumers and consumers in the intra-community layer. These structures distinguish trade entities based on ownership of energy systems and trade paths based on internal energy sharing pricing. Household distributed programs describe prosumer-owned deployment pathways which specify each prosumer as an energy trade entity with a three-layer trade path. Household centralized programs describe prosumer-grouped deployment pathways which combine all prosumers as energy trade entities with a two-layer trade path. Community centralized programs describe community-owned deployment pathways that coordinate all households as energy trade entities with a one-layer trade path. The pricing model is essential for determining energy sharing pricing between buyers and sellers in intra-community layers to facilitate energy sharing. We utilize the terms "buyers" and "sellers" to delineate household roles in the pricing model, as prosumers may adapt their behaviors to function as either sellers or buyers based on their net power profiles. We assume that all sellers have equal privilege and are equally influential participants within the energy communities, implying that all sellers should collectively determine energy sharing prices. In accordance with the basic principles of economics, disparities in demand response can result in price fluctuations. This signifies that the market price for buyers favors the grid output price, whereas the market price for sellers favors the grid input price. Thus, we use a dynamic internal pricing model that uses local feed-in tariffs to define trade prices using the production and demand ratio (SDR) of shared energy and price boundaries. Further details can be found in ref. 38 . To maintain consistency with the time resolution of the energy model, the internal prices of the pricing model are adjusted in tandem with the hourly SDR within the boundary limitations. Moreover, the predicted time horizon of the pricing model is set to one hour ahead, with local electricity prices acting as the benchmark prices 39,40 . In this study, electricity sales prices were assumed to be 70% of electricity purchase prices. Notably, we disregarded the energy price fluctuation and local government price incentives during the operating cycle of the energy model. Outline of green hydrogen systems in urban communities As in refs. 41,42 , energy systems, long-term climate resilience systems designed for urban communities, must be considered cross-seasonal and large-capacity energy storage due to the high homogeneity of households in both energy production and demand. Compared to conventional battery storage systems 43 , green hydrogen systems provide higher storage and operation flexibility, which can schedule multi-storage resources in the most efficient way to address uncertainty, especially during extreme energy events. Specifically, green hydrogen systems can achieve synergistic benefits from both the superior efficiency of battery storage and the high energy density and low leakage rate of hydrogen storage. From a technical perspective, the design of green hydrogen systems for urban communities must prioritize safety and portability. Safety is paramount, given the risk of hydrogen leakage and spontaneous combustion 44 . Portability calls for compact equipment and installation methods that are both straightforward and minimally invasive. To meet these criteria, we have chosen photovoltaic panels as the energy production modules and paired them with lithium-ion battery packages and Liquid Organic Hydrogen Carrier (LOHC) equipment for energy storage modules (Supplementary Note 6). Modelling design optimization for community green hydrogen systems Design optimization in energy models involves sizing optimal system components of green hydrogen systems to inform evaluation results for metrics of interest. As discussed earlier, multi-source uncertainties should be incorporated to ensure the robustness and stability of evaluation outcomes. Scenario pools are utilized as part of stochastic optimization in design optimization, capturing compound impacts of uncertainty factors. These scenario pools comprise a set of community energy scenarios with occurrence probabilities. Once the energy input scenarios are established, design optimization maps decision space variables onto the objective space through cyclic simulation under inherent system constraints. Specifically, decision variables encompass the rated capacity of the battery package, along with the rated power and capacity of both the fuel cell and electrolyzer. Meanwhile, the objective variables consist of system affordability and independence, quantified as annual life cycle costs (LCC) and grid interaction level (GIL), respectively. In this study, the multi-objective particle swarm algorithm was used to calculate the design optimization part, as it has demonstrated effectiveness in handling multi-scenarios and multi-objectives simultaneously. To enhance the efficiency of the search process and prevent convergence to local minima, we incorporated mechanisms like self-adaptive adjustment of inertia weights 45 and wavelet mutation 46 . Formulating objective functions The annual LCC represents the equivalent annual cost of green hydrogen systems over their lifespan in urban communities, including equivalent investment costs, equivalent operation and maintenance costs, carbon tax, and community trade costs, as formulated in Eq. ( 1 ). Notably, it is imperative to note that all the techno-economic data deployed in this study are derivatives of prevailing current social and technical conditions. The price uncertainties linked with technological upgrades and capital inflation are not considered in this study. The details of objective functions are shown in Supplementary Note 7, including each term of LCC and techno-economic data of green hydrogen systems. $$LCC=\sum _{\forall s\in {N}_{s}}{\rho }_{s}N\left({C}_{s,inv}+{C}_{s,om}\right)CRF+\sum _{\forall n\in N}\sum _{\forall s\in {N}_{s}}{\rho }_{s}\sum _{\forall t\in T}\left({C}_{n,s,t,tax}+{C}_{n,s,t,community}\right),\forall n\in N,\forall s\in {N}_{s},\forall t\in T$$ 1 where, n , s , and t denote the system number, expected community energy scenarios and time series; N , Ns , and T denote the total system number (prosumer number in HDP, 1 at HCP and CCP), total scenario number, and total simulation time; ρ s is the probability of scenario s ; CRF denotes the capital recovery factor; C inv and C om denote the initial investment cost and operation and maintenance cost per system; C n , s,t , tax , and C n , s , t , community denote the carbon tax, and community interaction cost at time t in scenario s for system n . GIL quantifies the level of independence of distributed energy systems by evaluating interactions with the power grid. The objective for urban residential communities is to minimize their interactions with the grid. This approach not only maximizes the role and benefits of communities but also helps to improve the stability of the urban power grid. This study evaluates GIL based on the amount of electricity imported from and exported to the grid, as shown in Eq. (2). \(GIL=\sum _{\forall n\in N}\sum _{\forall s\in {N}_{s}}{\rho }_{s}\sum _{\forall t\in T}\left({P}_{n,s,t,gridsell}+{P}_{n,s,t,gridbuy}\right),\) \(\forall n\in N,\forall s\in {N}_{s},\forall t\in T\) (2) where, P n , s,t,gridsell t and P n , s,t,gridbuy denote the amount of electricity purchased from the grid and the amount of electricity sold to the grid at time t in scenario s . Declarations Declaration of interests The authors declare no competing interests. Author contributions You Wu developed the model, performed the analysis, and wrote the paper. Lexuan Zhong supervised the study process and edited the paper. Acknowledgements The authors would like to express their gratitude for the partial financial support provided by the China Scholarship Council in collaboration with the University of Alberta and the Natural Sciences and Engineering Research Council of Canada (NSERC) Discovery program. Data availability Original data for this paper, as well as the code used for simulation and analyses, are available on Github. References Global Energy and Climate Model Documentation 2023. International Energy Agency, https://iea.blob.core.windows.net/assets/ff3a195d-762d-4284-8bb5-bd062d260cc5/GlobalEnergyandClimateModelDocumentation2023.pdf . Hong, T., Chen, Y., Luo, X., Luo, N. & Lee, S. H. Ten questions on urban building energy modeling. Building and Environment 168 (2020). https://doi.org/10.1016/j.buildenv.2019.106508 . Hepburn, C. et al. Towards carbon neutrality and China's 14th Five-Year Plan: Clean energy transition, sustainable urban development, and investment priorities. Environmental Science and Ecotechnology 8 (2021). https://doi.org/10.1016/j.ese.2021.100130 . Lopez, A., Cole, W., Sergi, B. et al. Impact of siting ordinances on land availability for wind and solar development. Nat Energy 8, 1034–1043 (2023). https://doi.org/10.1038/s41560-023-01319-3 . Tonelli, D., Rosa, L., Gabrielli, P. et al. Global land and water limits to electrolytic hydrogen production using wind and solar resources. Nat Commun 14, 5532 (2023). https://doi.org/10.1038/s41467-023-41107-x . Kammen, D. M. & Sunter, D. A. City-integrated renewable energy for urban sustainability. Science 352, 922–928 (2016). https://doi.org/10.1126/science.aad9302 . Lowitzsch, J., Hoicka, C. E. & van Tulder, F. J. Renewable energy communities under the 2019 European Clean Energy Package - Governance model for the energy clusters of the future? Renew. Sust. Energ. Rev. 122, 13 (2020). https://doi.org/10.1016/j.rser.2019.109489 . Bauwens, T. et al. Conceptualizing community in energy systems: A systematic review of 183 definitions. Renew. Sust. Energ. Rev. 156, 16 (2022). https://doi.org/10.1016/j.rser.2021.111999 . Levin, T. et al. Energy storage solutions to decarbonize electricity through enhanced capacity expansion modelling. Nature Energy 8, 1199–1208 (2023). https://doi.org/10.1038/s41560-023-01340-6 . Twitchell, J., DeSomber, K. & Bhatnagar, D. Defining long duration energy storage. Journal of Energy Storage 60, 7 (2023). https://doi.org/10.1016/j.est.2022.105787 . Parra, D. et al. An interdisciplinary review of energy storage for communities: Challenges and perspectives. Renew. Sust. Energ. Rev. 79, 730–749 (2017). https://doi.org/10.1016/j.rser.2017.05.003 . Le, T. S., Nguyen, T. N., Bui, D. K. & Ngo, T. D. Optimal sizing of renewable energy storage: A techno-economic analysis of hydrogen, battery and hybrid systems considering degradation and seasonal storage. Applied Energy 336, 19 (2023). https://doi.org/10.1016/j.apenergy.2023.120817 . Herbst, A., Toro, F., Reitze, F. et al. Introduction to Energy Systems Modelling. Swiss J Economics Statistics 148, 111–135 (2012). https://doi.org/10.1007/BF03399363 . Beiter, P., Mai, T. U., Mowers, M. & Bistline, J. Expanded modelling scenarios to understand the role of offshore wind in decarbonizing the United States. Nature Energy 8, 1240–1249 (2023). https://doi.org/10.1038/s41560-023-01364-y . Perera, A. T. D., Nik, V. M., Chen, D., Scartezzini, J.-L. & Hong, T. Quantifying the impacts of climate change and extreme climate events on energy systems. Nature Energy 5, 150–159 (2020). https://doi.org/10.1038/s41560-020-0558-0 . Leprince, J. et al. Can occupant behaviors affect urban energy planning? Distributed stochastic optimization for energy communities. Applied Energy 348, 14 (2023). https://doi.org/10.1016/j.apenergy.2023.121589 . Gabrielli, P., Fürer, F., Mavromatidis, G. & Mazzotti, M. Robust and optimal design of multi-energy systems with seasonal storage through uncertainty analysis. Applied Energy 238, 1192–1210 (2019). https://doi.org/10.1016/j.apenergy.2019.01.064 . Perera, A. T. D., Khayatian, F., Eggimann, S., Orehounig, K. & Halgamuge, S. Quantifying the climate and human-system-driven uncertainties in energy planning by using GANs. Applied Energy 328, 12 (2022). https://doi.org/10.1016/j.apenergy.2022.120169 . Huang, P. et al. Characterization and optimization of energy sharing performances in energy-sharing communities in Sweden, Canada and Germany. Applied Energy 326, 16 (2022). https://doi.org/10.1016/j.apenergy.2022.120044 . Kang, H. et al. Multi-objective sizing and real-time scheduling of battery energy storage in energy-sharing community based on reinforcement learning. Renew. Sust. Energ. Rev. 185, 18 (2023). https://doi.org/10.1016/j.rser.2023.113655 . Chen, X. et al. Multi-objective optimization of battery capacity of grid-connected PV-BESS system in hybrid building energy sharing community considering time-of-use tariff. Applied Energy 350, 15 (2023). https://doi.org/10.1016/j.apenergy.2023.121727 . Li, Y. et al. Optimization of integrated energy system for low-carbon community considering the feasibility and application limitation. Applied Energy 348, 19 (2023). https://doi.org/10.1016/j.apenergy.2023.121528 . Dong, S. Y., Kremers, E., Brucoli, M., Rothman, R. & Brown, S. Improving the feasibility of household and community energy storage: A techno-enviro-economic study for the UK. Renew. Sust. Energ. Rev. 131, 17 (2020). https://doi.org/10.1016/j.rser.2020.110009 . Perera, A. T. D. et al. Challenges resulting from urban density and climate change for the EU energy transition. Nature Energy 8, 397–412 (2023). https://doi.org/10.1038/s41560-023-01232-9 . Dai, R., Esmaeilbeigi, R. & Charkhgard, H. The Utilization of Shared Energy Storage in Energy Systems: A Comprehensive Review. Ieee Transactions on Smart Grid 12, 3163–3174 (2021). https://doi.org/10.1109/tsg.2021.3061619 . Wang, M. et al. Combined multi-objective optimization and robustness analysis framework for building integrated energy system under uncertainty. Energy Conversion and Management 208, 15 (2020). https://doi.org/10.1016/j.enconman.2020.112589 . Gjorgievski, V. Z., Cundeva, S. & Georghiou, G. E. Social arrangements, technical designs and impacts of energy communities: A review. Renewable Energy 169, 1138–1156 (2021). https://doi.org/10.1016/j.renene.2021.01.078 Carley, S. & Konisky, D. M. The justice and equity implications of the clean energy transition. Nature Energy 5, 569–577 (2020). https://doi.org/10.1038/s41560-020-0641-6 . Montañés, C. C. et al. Enabling and centering equity and justice in clean energy transition research. Joule 7, 437–441 (2023). https://doi.org/10.1016/j.joule.2023.02.005 . Mavromatidis, G., Orehounig, K. & Carmeliet, J. Comparison of alternative decision-making criteria in a two-stage stochastic program for the design of distributed energy systems under uncertainty. Energy 156, 709–724 (2018). https://doi.org/10.1016/j.energy.2018.05.081 . Xu, L., Feng, K., Lin, N. et al. Resilience of renewable power systems under climate risks. Nat Rev Electr Eng 1, 53–66 (2024). https://doi.org/10.1038/s44287-023-00003-8 . Moazami, A., Nik, V. M., Carlucci, S. & Geving, S. Impacts of future weather data typology on building energy performance - Investigating long-term patterns of climate change and extreme weather conditions. Applied Energy 238, 696–720 (2019). https://doi.org/10.1016/j.apenergy.2019.01.085 . Tian, W. et al. A review of uncertainty analysis in building energy assessment. Renew. Sust. Energ. Rev. 93, 285–301 (2018). https://doi.org/10.1016/j.rser.2018.05.029 . Balvedi, B. F., Ghisi, E. & Lamberts, R. A review of occupant behaviour in residential buildings. Energy and Buildings 174, 495–505 (2018). https://doi.org/10.1016/j.enbuild.2018.06.049 . Petkov, I. & Gabrielli, P. Power-to-hydrogen as seasonal energy storage: an uncertainty analysis for optimal design of low-carbon multi-energy systems. Applied Energy 274, 25 (2020). https://doi.org/10.1016/j.apenergy.2020.115197 . Li, Y., Qian, F. Y., Gao, W. J., Fukuda, H. & Wang, Y. F. Techno-economic performance of battery energy storage system in an energy sharing community. Journal of Energy Storage 50, 21 (2022). https://doi.org/10.1016/j.est.2022.104247 . Liu, N. et al. Energy-Sharing Model with Price-Based Demand Response for Microgrids of Peer-to-Peer Prosumers. Ieee Transactions on Power Systems 32, 3569–3583 (2017). https://doi.org/10.1109/tpwrs.2017.2649558 . Average energy prices for the United States. U.S. BUREAU of LABOR STATISTICS, https://www.bls.gov/regions/home.htm . Energy Fact Book 2023–2024. NRCan, https://energy-information.canada.ca/sites/default/files/2023-10/energy-factbook-2023-2024.pdf . Pan, G. S. et al. Optimal Planning for Electricity-Hydrogen Integrated Energy System Considering Power to Hydrogen and Heat and Seasonal Storage. Ieee Transactions on Sustainable Energy 11, 2662–2676 (2020). https://doi.org/10.1109/tste.2020.2970078 . Gabrielli, P., Gazzani, M., Martelli, E. & Mazzotti, M. Optimal design of multi-energy systems with seasonal storage. Applied Energy 219, 408–424 (2018). https://doi.org/10.1016/j.apenergy.2017.07.142 . Special Report: Potential Reliability Impacts of Emerging Flexible Resources. North American Electric Reliability Corporation https://www.nerc.com/pa/RAPA/ra/Reliability%20Assessments%20DL/IVGTF_Task_1_5_Final.pdf . Yang, F. Y. et al. Review on hydrogen safety issues: Incident statistics, hydrogen diffusion, and detonation process. International Journal of Hydrogen Energy 46, 31467–31488 (2021). https://doi.org/10.1016/j.ijhydene.2021.07.005 . Cui, Y. Y., Meng, X. & Qiao, J. F. A multi-objective particle swarm optimization algorithm based on two-archive mechanism. Appl. Soft. Comput. 119, 17 (2022). https://doi.org/10.1016/j.asoc.2022.108532 . Tian, D. P. & Shi, Z. Z. MPSO: Modified particle swarm optimization and its applications. Swarm Evol. Comput. 41, 49–68 (2018). https://doi.org/10.1016/j.swevo.2018.01.011 . Additional Declarations There is NO Competing Interest. Supplementary Files SupplementaryNote.docx NATCITIES24040439SupplementaryNote.docx Supplementary Information Cite Share Download PDF Status: Published Journal Publication published 02 Jan, 2025 Read the published version in Nature Cities → Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-4327177","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":309819273,"identity":"63dc5ef0-ff10-4a3d-8c05-99b9a2f7a4cb","order_by":0,"name":"Lexuan Zhong","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABCklEQVRIiWNgGAWjYBACPhjDAER8YGCQMSCkhQ1ZC+MMBgYeIIOxgWgtzDxEaWHvMZP4uYPB3py99/Br2zY7HiDj+YMfDHbyOLXwHEuT7D3DwGzZcy7NOrctmcey57hhYw9DsiEuq9gkko9J8LYxsBncyDEzzm1j5jG4kcbYwMNwAKfr2OQftkn+bQN6AaTFsq2ex+D+M8bGPwwH7HHbwnxMGmiLBFCL8WPGtsNAvWyMzUBbEnFq4UlLtpZtkzAwOHPGjLHn3HGgX9IYZ8sYJCfj0sLPfsbw5ts2G3uD4z3GH36UVcuZsx9j+Pimws4WlxYokIA4EiFAMA1AAPMH4tSNglEwCkbBSAMANSpMWn6Iw2IAAAAASUVORK5CYII=","orcid":"https://orcid.org/0000-0001-5211-8826","institution":"University of Alberta","correspondingAuthor":true,"prefix":"","firstName":"Lexuan","middleName":"","lastName":"Zhong","suffix":""},{"id":309819274,"identity":"9fd4ea89-e86f-4d5f-bb04-29b96539522e","order_by":1,"name":"You Wu","email":"","orcid":"","institution":"University of Alberta","correspondingAuthor":false,"prefix":"","firstName":"You","middleName":"","lastName":"Wu","suffix":""}],"badges":[],"createdAt":"2024-04-26 04:35:42","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-4327177/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-4327177/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s44284-024-00178-7","type":"published","date":"2025-01-02T05:00:00+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":57712257,"identity":"99228308-b682-4dcf-a893-79a57eb86b68","added_by":"auto","created_at":"2024-06-04 16:10:57","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":245339,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eA graphical overview of the energy model. \u003c/strong\u003e(A) A set of community energy scenarios with 8,760-time steps is derived by applying multi-source uncertainties to a simple building simulation model. (B) Each household in the community energy scenario is connected using system deployment pathways. (C) The energy system is integrated into the community using a bottom-up approach. A comprehensive description of the model is presented in the Methods Section.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-4327177/v1/c8a458021187a5dfba19e82d.png"},{"id":57711575,"identity":"b509469c-3f2c-412c-a830-4d17a3311459","added_by":"auto","created_at":"2024-06-04 16:02:57","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":54556,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eImpacts of climate-human uncertainties on annual energy production and demand at the household level in 7 cities in North America. \u003c/strong\u003eMiami, Houston, Las Vegas, Vancouver, Toronto, Montreal, and Edmonton are the selected representative cities in climate zone 1~7, representatively. Annual energy data are derived from energy data collected on an hourly scale.\u003cstrong\u003e \u003c/strong\u003eOS and TS, in the endnote of each city on the horizon label, mean bungalows and two-story houses. In the box plot, the central box shows the interquartile range, with the median depicted as a line in the middle. Dots, rectangles, and crosses represent outliers, the mean, and 99% and 1% data points, respectively. Additionally,lines indicate the maximum and minimum values, excluding outliers.\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-4327177/v1/ae4723b4ebf5d7d381de0498.png"},{"id":57712859,"identity":"ced3737e-35d6-46a0-bbeb-d3016df7045a","added_by":"auto","created_at":"2024-06-04 16:18:57","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":33029,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eImpacts of community design layer uncertainties on sharing level at the community level in 7 cities in North America. \u003c/strong\u003eThe Sharing level above 1 means energy production exceeds energy demand, and vice versa.\u003cstrong\u003e \u003c/strong\u003eOB and TB represent boundary community samples, which are composed of bungalows and two-story houses, respectively, while “Ave” represents regional characteristic samples, which are composed of regional-average building stock data. Notably, each box represents the energy characteristics of the household in the entire community scenario pools without distinguishing the sample occurrence probability information.\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-4327177/v1/c8364907c43d64c88e2869c9.png"},{"id":57710738,"identity":"47e92177-e3ec-4508-b457-47e8eec2b76d","added_by":"auto","created_at":"2024-06-04 15:54:57","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":133189,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eThe impact of deployment pathways on potential changes of LCOE. \u003c/strong\u003e(a-g) (a) Miami, (b) Houston, (c) Las Vegas, (d) Vancouver, (e) Toronto, (f) Montreal, and (g) Edmonton.\u003cstrong\u003e \u003c/strong\u003eWith each plot, HDP, HCP, and CCP include three types of household scales and their respective prosumer scale.\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-4327177/v1/9f20f8e3ab2dd23abca76b14.png"},{"id":57710742,"identity":"7299229d-1721-4662-8511-f46b7c31e834","added_by":"auto","created_at":"2024-06-04 15:54:57","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":78114,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eChanges of four system design parameters under three types of deployment pathways across the eight selected cities.\u003c/strong\u003e (a-g) (a) Miami, (b) Houston, (c) Las Vegas, (d) Vancouver, (e) Toronto, (f) Montreal, and (g) Edmonton. Each compromise system design includes four design parameters, including an electrolyzer, fuel cell, hydrogen tank, and battery package. Each point in each plot corresponds to a compromise value of certain design parameters at a special household scale and prosumer scale.\u003c/p\u003e","description":"","filename":"5.png","url":"https://assets-eu.researchsquare.com/files/rs-4327177/v1/eebc30a0bab9365209be2e9d.png"},{"id":72875698,"identity":"ca64496a-dcfa-40b9-b1fb-a7b845c10180","added_by":"auto","created_at":"2025-01-03 08:10:40","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1148315,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4327177/v1/89553d30-3f38-4bd4-97a0-6714742168bc.pdf"},{"id":57710743,"identity":"9bd9bdcd-aae6-4cc2-8371-0d5a8c85e11e","added_by":"auto","created_at":"2024-06-04 15:54:57","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":7654592,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cbr\u003e\u003c/p\u003e","description":"","filename":"SupplementaryNote.docx","url":"https://assets-eu.researchsquare.com/files/rs-4327177/v1/bc79a4a5861acba6fe62e4d4.docx"},{"id":57710744,"identity":"be1c146e-53d1-4b3a-801c-ea0004f4ab8e","added_by":"auto","created_at":"2024-06-04 15:54:58","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":7673692,"visible":true,"origin":"","legend":"\u003cp\u003eSupplementary Information\u003c/p\u003e","description":"","filename":"NATCITIES24040439SupplementaryNote.docx","url":"https://assets-eu.researchsquare.com/files/rs-4327177/v1/52b3bdfc59b94c8cbabc74ab.docx"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Probabilistic deployment pathways of scaling up distributed green hydrogen systems for urban residential communities in North America","fulltext":[{"header":"Introduction","content":"\u003cp\u003eAccording to the International Energy Agency (IEA), the global energy sectors need to achieve net zero emissions by 2050 to reach the target of limiting global warming to 1.5\u0026deg;C in 2100\u003csup\u003e1\u003c/sup\u003e. Currently, about one-third of global power is consumed in cities\u003csup\u003e2\u003c/sup\u003e. Rapid deployment of clean energy technologies is a core pathway to achieve urban energy transition from fossil fuels to cleaner energy sources\u003csup\u003e3\u003c/sup\u003e. With surging energy demand caused by accelerating urbanization, growing evidence shows that planning large-scale plant-central renewable energy production outside the cities faces enormous pressure from technically feasible lands due to their limited renewable energy density\u003csup\u003e4,5\u003c/sup\u003e. Thus, distributed energy systems that support integrating renewable technologies will effectively reduce external energy demand from urban residential communities, minimizing the land pressure of urban renewable energy infrastructure\u003csup\u003e6\u003c/sup\u003e. In addition, a large body of research supports the notion that communities with coordinated action provide critical spatial, temporal, and social pathways for energy integration to realize higher economic, environmental, and social interests of distributed energy systems\u003csup\u003e7,8\u003c/sup\u003e. Therefore, an accelerated market introduction of decentralized energy systems into urban communities is considered a pivotal option to utilize inner-city resources for decarbonizing cities.\u003c/p\u003e \u003cp\u003eDistributed energy systems powered by renewable sources depend on cost-effective energy storage technologies to address the severe energy mismatch caused by high homogeneous production and demand in urban residential communities\u003csup\u003e9\u003c/sup\u003e, requiring high efficiency for short-term extreme energy events and high capacity for cross-seasonal energy patterns\u003csup\u003e10\u003c/sup\u003e. Growing research shows that the single battery technology as a cross-seasonal storage option may not be economically viable in long-term resilience systems due to their low energy density and high leakage rate\u003csup\u003e11\u003c/sup\u003e. Green hydrogen systems, a distributed energy system with green hydrogen and battery as storage options, offer higher storage flexibility and lower storage costs, making them more suitable for long-term community sustainability\u003csup\u003e12\u003c/sup\u003e. We consider green hydrogen systems a promising community energy solution that needs to be evaluated in detail.\u003c/p\u003e \u003cp\u003eEnergy models are widely used to understand the impacts of energy systems on residential communities and to inform potential development directions to energy planners\u003csup\u003e13\u003c/sup\u003e. However, academic and policy spheres highlight concern that deep uncertainties of energy input are fundamental modelling limitations for energy models, causing inappropriate assessment and adversely impacting policy targets when ignoring modelling limitations\u003csup\u003e14\u003c/sup\u003e. Part of energy input uncertainties stems from both climate conditions and human behaviors. In future periods, energy models will confront heightened challenges associated with the evolving climate conditions that result from continuing climate change\u003csup\u003e15\u003c/sup\u003e. On the other hand, these models exhibit limited incorporation of historical data-driven methods to measure potential human behavior\u003csup\u003e16\u003c/sup\u003e. Thus, how to effectively expand input scenarios for energy models becomes core to understanding the potential fluctuations of energy input stemming from uncertainties associated with climate-human systems. The sampling method\u003csup\u003e17\u003c/sup\u003e and generative adversarial network (GAN) method\u003csup\u003e18\u003c/sup\u003e are adapted to generate a substantial number of energy input scenarios for modeling potential climate-human uncertainties.\u003c/p\u003e \u003cp\u003eBesides climate-human systems, another significant source of uncertainties emerges from the community design layer, including community compositions and renewable energy levels. Community compositions account for different residential building types and their corresponding numbers within residential communities, while renewable energy levels represent the deployment numbers of energy systems within residential communities \u003csup\u003e19\u003c/sup\u003e. These two factors impact the diversity of energy input scenario types for energy models in the design phase, influenced by various geographical, sociological, and psychological considerations. Current studies have considered that the uncertainties of the community design layer significantly impact on the overall economic and environmental benefits of the community\u003csup\u003e20,21\u003c/sup\u003e. Thus, energy models must integrate detailed community design layers into energy planning. However, meeting this requirement poses a dilemma for energy models, navigating the delicate equilibrium between spatial dimensions and calculating tractable. To strike balance between the diversity of household energy profiles with the robustness of urban energy planning, historical data-driven energy input scenarios\u003csup\u003e23\u003c/sup\u003e and city-related virtual communities\u003csup\u003e24\u003c/sup\u003e are developed.\u003c/p\u003e \u003cp\u003eSystem deployment pathways of green hydrogen systems in communities determine different household participation forms, highly affecting shared benefits and participation motivation among interconnected households\u003csup\u003e25\u003c/sup\u003e. Thus, system deployment pathways are a centerpiece of understanding the full impact of green hydrogen systems on decarbonizing communities in the strategic energy planning phase. We reckon that such multiple-source uncertainties of energy models are particularly relevant for analyzing the impact on the system deployment pathways due to energy input uncertainties undoubtedly that affect the model output results of system deployment pathways. The multiple-source uncertainties in communities include climate-human uncertainties at the household level and community design uncertainties at the community level. In that context, quantifying these uncertainties into energy models presents a considerable challenge due to the complexity of multidimensional impacts. We consider that green hydrogen systems are at a nascent stage. Whether and how green hydrogen systems can expand fast enough in uncertain communities remains unclear. Thus, systematically analyzing potential expansion pathways of green hydrogen systems through high-fidelity energy models is essential to facilitate rapid deployment and unleash its potential for community energy transition.\u003c/p\u003e \u003cp\u003eIn this article, we introduce a bottom-up energy model linking climate, human behavior, community archetypes, and energy system models, aiming to ensure cost-effective deployment pathways of green hydrogen systems for highly diverse urban residential communities. Figure\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e depicts a summary of three interconnected phases of the proposed energy model. First, we developed a multi-scale combined uncertainty approach to obtain a set of community energy production and demand profiles, including detailed household energy information with potential occurrence probability to reflect inevitable uncertainties from household and community layers. Next, we formulated three system deployment pathways based on an urban planning perspective to characterize strategies for household collaboration within the community. These pathways were merged with community energy profiles to serve as community inputs to the energy system model related to green hydrogen systems. Lastly, the robust energy system model yielded a collection of optimal system designs that reflect the cost-effective impacts of deployment pathways on green hydrogen systems under multi-source uncertainties. We extensively investigated seven North American climate zones to understand the general and overall effects of deployment pathways on community energy system design.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eMulti-source uncertainty quantification of energy input scenarios at the North American scale\u003c/h2\u003e \u003cp\u003eThe IEA report points out that advanced economies need to take the lead and reach net-zero emissions earlier to buy more time for the energy transition in developing economies\u003csup\u003e1\u003c/sup\u003e. However, the efforts undertaken by the United States and Canada in North America fall far short of what is needed to achieve net zero global energy-related carbon dioxide emissions by 2050\u003csup\u003e26\u003c/sup\u003e. These developed nations need to take further action to assess the cost and availability of green hydrogen systems as \"emerging\" technologies for widespread commercial deployment. The United States and Canada are divided into eight building climate zones, and most of their cities are located in the previous seven climate zones. Cities within each climate zone do not exhibit significant differences in building and energy regulations. Thus, we selected one sample of cities from each climate zone to capture the typical regional energy performance. The green hydrogen systems in these cities were evaluated using an energy model.\u003c/p\u003e \u003cp\u003eThe standard approach to obtaining dynamic energy input scenarios for energy models is to use typical urban stock community parameters, such as climate patterns, human activities, building characteristics, etc. However, this approach often overlooks the compounded uncertainty impacts from both the household and community levels. Neglecting these impacts may lead to an energy information gap between computed energy performance and reality. This study introduces multi-source uncertainties into bottom-up building energy modelling layer by layer depending on their impact pathway in the energy calculation process. Thus, we begin by quantifying the impacts of climate-human uncertainties on energy input scenarios at the household level (details about climate-human uncertainties are presented in Method).\u003c/p\u003e \u003cp\u003eFigure\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e. shows the impact of climate-human uncertainties on energy input scenarios in 7 cities taken from climate zones 1\u0026thinsp;~\u0026thinsp;7. Climate and human uncertainties show a more uneven and significant effect on the energy demand side rather than on the energy production side from the median and interquartile range perspective. On the energy demand side, the annual median household energy demand exhibits a valley-shaped pattern across all climate zones, with varying interquartile ranges. Vancouver demonstrates the lowest median energy demand with a narrower interquartile range at the bottom of the valley, while Miami and Edmonton exhibit higher median energy demand values and more comprehensive interquartile ranges at the top of the valley. On the energy production side, the median annual household energy production decreases as climate zones progress backward, with a nearly uniform interquartile range. Miami and Edmonton have the highest and lowest energy production performance, respectively. The variation in urban building archetypes results in higher energy demand but similar energy production within these observed cities. The median annual energy demand of bungalows is approximately 5,000 kWh lower than that of two-story buildings, yet the annual energy production exhibits minimal variance between the two types of structures.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eWe then superimpose the impact of community design layer uncertainties onto the energy input scenarios at the household level using a community characteristic matrix assembled from household scale and prosumer scale (detail about community design layer uncertainties is presented in Method). The household scale represents the clustering process of average and boundary community energy scenarios from previously obtained input scenarios, while the prosumer scale denotes the labeling process of independent households as prosumers or consumers within obtained community samples. Figure\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e. shows the impact of community design layer uncertainties on the energy input scenarios at the community level, depicted using the sharing level indicator. The sharing level reflects the expected potential for energy self-sufficiency of independent households within community samples.\u003c/p\u003e \u003cp\u003eResults show that annual median sharing levels in all community samples exhibit a ridge-shaped pattern across all climate zones. Vancouver shows the highest sharing levels at the peak of the ridge, while Miami and Edmonton exhibit lower sharing levels at the base. In addition, sharing levels within the cities show a consistent decreasing trend among the three types of community samples, with bungalow community samples exhibiting the highest sharing levels, followed by average community samples, and finally, two-story community samples. This pattern suggests that energy demand plays a dominant role in shaping sharing levels across various climate zones, regardless of energy production levels, such as the higher energy production in Miami and lower energy production in Edmonton. However, we also observed that the annual peak-sharing levels vary among cities. Despite this, the peak value in moderate climate regions is higher than in extremely hot or cold climate regions. This suggests that, within urban communities, the climate zones to which the city belongs exert a more substantial influence on sharing levels than the internal community archetypes.\u003c/p\u003e\u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003eImpact of system deployment pathways on levelized cost of electricity\u003c/h2\u003e \u003cp\u003eThe green hydrogen systems in North American cities were assessed using an energy model that provides all the non-dominant sets of optimal system designs for initial commercial-scale deployment via Pareto front analysis. Our analysis centers on how different deployment pathways of energy systems affect energy affordability of urban communities, measured as levelized cost of electricity (LCOE) calculated from compromise solutions of system design options for each group. The primary approach for decision-making from the Pareto front utilizes the Euclidean-distance-based method\u003csup\u003e25\u003c/sup\u003e. LCOE provides a levelized present price per unit energy value, making it suitable for comparing differences in the cost of system power generation across various urban residential sectors and deployment pathways. Figure\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e shows the impact of three types of deployment pathways on potential changes in LCOE in the eight selected cities. Two key insights emerge.\u003c/p\u003e \u003cp\u003eFirst, as observed from a horizontal perspective, household distributed program (HDP) follows a small upward trend as the prosumer scale increases within an equivalent community scale in all urban community cases. In contrast, household-centralized programs (HCP) and community-centralized programs (CCP) show the opposite trend, decreasing significantly. This is due to the fact that, with more household participation, HDP faces resource congestion caused by household competition, resulting in greater costs, while HCP and CCP can fully utilize the scale effect of household cooperation to reduce costs. Additionally, the similarity in the LCOE change trend across all cities indicates that climate zones have a trivial impact on the ways in which the three deployment pathways affect LCOE, although LCOE tends to be lower in mild climate zones compared to hot or cold climate zones. Furthermore, the increased rate of LCOE in HDP has remained relatively stable as the prosumer scale increases, while the decreased rate of LCOE in HCP and CCP gradually slows down. This indicates that despite increased household participation, the increased resource congestion in HDP does not accelerate cost deterioration, while the advantages of scale economies in HCP and CCP gradually diminish.\u003c/p\u003e \u003cp\u003eSecond, as observed from a vertical perspective, the LCOE of different community scales within each deployment pathway decreases in the order of OB, Ave, and TW community scales across all urban community cases. This reveals that urban communities with high sharing levels bring more cost advantages regardless of climate zone and deployment pathways. Additionally, in all cities except Edmonton, HDP demonstrates a greater cost advantage compared to HCP and CCP at a 20% prosumer scale under the same community scale. With a prosumer scale exceeding 40%, both HCP and CCP demonstrate a cost advantage. Notably, HCP slightly outperforms CCP in most urban community scales, with both reaching parity at a 100% prosumer scale. The subtle difference shows that consumer scale can contribute to the cost burden, which decreases as consumer scale decreases. HCP shows better cost advantage than HDP and CCP at low prosumer scale in certain cities, such as Vancouver, Las Vegas, and Houston. However, the cost benefits diminish as more households with low sharing levels join. This is attributed to the fact that consortia of households with high sharing levels can readily export energy for profit through scale effects, thus supplementing energy costs. Therefore, it can be concluded that HDP and HCP approaches are suitable for low-participation communities, while HCP and CCP are more appropriate for high-participation communities.\u003c/p\u003e\u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003eAnalysis of system design for resilient urban community\u003c/h2\u003e \u003cp\u003eThe primary driver of LCOE is the initial capital investment derived from the optimal system design identified by the energy model. Analyzing the variations in energy design across different pathways is crucial for understanding how these pathways drive changes in LCOE from the system design perspective. Here, we now examine the impact of deployment pathways on the system design to account for our previous findings on LCOE changes. Figure\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e illustrates the potential changes in four system design parameters under three types of deployment pathways across the eight selected cities.\u003c/p\u003e \u003cp\u003eFirst, we find that the drivers of LCOE change are not the same on the same consumer scale. HDP keeps system design parameters essentially unchanged as the prosumer scale increases. This shows that the slight increase in LCOE stems from continuous external resource congestion rather than a significant change in system investment costs. In contrast, both HCP and CCP significantly expand hydrogen tanks compared to other design parameters as the prosumer scale increases. This suggests that the continuous decrease in LCOE is driven by the substantial reduction in system investment costs, as cheap hydrogen tanks replace a significant portion of batteries in energy storage as the scale increases. Therefore, it can be concluded that the hydrogen tank becomes a primary design factor in driving the scale effect of HCP and CCP.\u003c/p\u003e \u003cp\u003eSecond, we notice that higher energy storage demands lead to a higher hydrogen storage ratio, leading to lower LCOE when comparing system design parameters at the same community scale across the three pathways. This is because the advantage of scale in hydrogen costs offsets the drawback of linear increases in battery storage costs as energy storage demand increases. In the initial phase, both HDP and HCP demonstrate higher energy storage demand compared to CCP, resulting in a higher hydrogen storage ratio and better LCOE performance. However, as the prosumer scale increases, the higher storage demand leads to an inevitable growth in the hydrogen storage ratio for both HCP and CCP. Meanwhile, the storage structure limitations of HDP result in its hydrogen storage ratio remaining nearly constant, ultimately causing it to be quickly surpassed by HCP and CCP in LCOE performance. This can be further justified by the trivial difference across different community scales within the same pathway, where the hydrogen storage ratio decreases slightly along OB, Ave, and TW due to the decrease in energy storage demand.\u003c/p\u003e\u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eCurrent energy models typically focus on urban and national scales, overlooking the influence of local private entities such as households and communities on energy markets\u003csup\u003e27\u003c/sup\u003e. However, the decentralized nature of distributed energy systems grants more agency to these entities in participating in the energy market\u003csup\u003e28\u003c/sup\u003e. This oversight may result in significantly lower market participation scales than those anticipated by macro energy models, posing a risk to the achievement of energy transition targets. In this work, we built a bottom-up energy model that established links between energy systems and local urban communities, which incorporate necessary technological details related to climate, human behavior, building archetypes, and energy system characteristics. This energy model is applied to the role of system deployment pathways in developing cost-effective green hydrogen systems for diverse urban residential communities from a design and operation perspective. The evaluation results will complement critical market participation information to inform fully potential expansion pathways to energy planners.\u003c/p\u003e \u003cp\u003eOur study shows that differences in community sharing levels influenced by climate zones and community type significantly impact the overall value of energy costs rather than their change trend affected by deployment pathways. Energy costs are lower in moderate climate zones than in extremely hot or cold climate zones. Additionally, community samples with high sharing levels consistently exhibit lower costs relative to those with low sharing levels at the deployment pathway. These findings underscore the potential for communities characterized by high sharing levels and moderate climate zones to attain optimal energy costs.\u003c/p\u003e \u003cp\u003eOur study underscores the critical importance of selecting the right deployment pathway for achieving cost-effective decarbonization in urban communities, with impacts consistent across all climate zones. The cost differential between the best and worst pathways for the same community scale can be as high as 60%. A household centralized program stands out as the preferred deployment pathway for most communities, consistently driving cost reductions by harnessing scale effects. In communities with low prosumer scale, household distributed programs may demonstrate better cost advantages initially, but as prosumers increase, resource competition escalates, leading to higher energy costs. Conversely, community centralized programs are better suited for communities with high prosumer scales, where consumer scales always significantly negatively impact energy costs.\u003c/p\u003e \u003cp\u003eOur study shows that variances in energy storage demands across different deployment pathways serve as the principal determinant of their respective energy cost trends. In green hydrogen systems, increased energy storage demands trigger a scale effect in hydrogen storage costs, consequently mitigating system investment costs. Our findings indicate that household centralized programs exhibit higher energy storage demand than household distributed programs and community centralized programs, demonstrating greater cost advantages in most communities. This underscores the critical importance of incorporating more prosumers in the design of deployment pathways to expand energy storage demand.\u003c/p\u003e \u003cp\u003eThere are several essential avenues for expanding upon this work. Extending our research to include future climate periods beyond 2050, such as those projected for 2050\u0026ndash;2100 under Representative Concentration Pathway (RCP) 8.5, would offer valuable insights into the long-term performance of various system deployment pathways for green hydrogen systems\u003csup\u003e15\u003c/sup\u003e. This extended analysis would help anticipate and address the challenges posed by further worsening climate change, including ongoing shifts in climate patterns and increased frequency of extreme climate events. By examining deployment strategies across a broader range of future climate conditions, we can better understand their cost-effectiveness and resilience, ultimately informing more robust and adaptive energy transition strategies. Additionally, expanding discussions on developing green hydrogen systems to include social justice and ethical concerns is crucial, as energy policy and technology decisions impact communities in diverse ways beyond technical feasibility and cost-effectiveness\u003csup\u003e29\u003c/sup\u003e. Energy inequalities arising from the energy transition can exacerbate disparities and hinder access to opportunities for specific communities within the energy market\u003csup\u003e30\u003c/sup\u003e. Future studies can integrate social concerns into energy models to enhance our understanding of the deployment pathways needed to support equitable and inclusive urban energy transitions, ensuring that the benefits of green hydrogen systems are accessible to all urban communities.\u003c/p\u003e \u003cp\u003eFinally, the energy models in this paper can be used to expand other renewable energy sources to suit local community conditions, such as increased adoption of wind, geothermal, or biomass.\u003c/p\u003e"},{"header":"Methods","content":"\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eAnalytical approach\u003c/h2\u003e \u003cp\u003eThis study aims to explore the impacts and implications of system deployment pathways on deploying cost-effective green hydrogen energy systems for different urban communities. Therefore, energy system design should cover all potential community energy input scenarios subject to multiple uncertainties to ensure robust and optimal outcomes. However, climate change exhibits a distinct probabilistic nature, i.e., extreme climate conditions are significantly less likely to occur than typical climate conditions. The energy input scenarios should reflect this variability in weather probabilities to ensure reasonable realism. Therefore, we use stochastic optimization to formulate and integrate the objective function into the energy model\u003csup\u003e31\u003c/sup\u003e. In this process, each energy scenario is treated independently, and then their results are combined based on the probability of occurrence to obtain the mathematical expectation of the outcome.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003eThe overview of modelling multi-source uncertainties\u003c/h2\u003e \u003cp\u003eEnergy models are subject to inherent input parameter uncertainties derived from complex real-world communities. It is crucial for energy models to transparently disclose the impacts of community uncertainties on energy system inputs and provide reliable energy input scenarios for subsequent evaluation. Parameterizing the compound impacts of these uncertainties on energy input scenarios for the energy model is challenging due to the complexity of sources and the differences in their influence pathways. Thus, we distinguish the community uncertainties into two levels: climate-human uncertainties at the household level and community design layer uncertainties at the community level (Supplementary Notes 1). As a bottom-up energy model, we quantify uncertainties layer by layer from the household level to the community level and superimpose them to quantify compound impacts.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003eModelling climate-human uncertainties\u003c/h2\u003e \u003cp\u003eTo parameterize the climate-human uncertainties at the household level, we distinguish climate and human systems into two treatments based on their different sources. We then associate them into compound scenarios to account for their superimposed impacts on energy input. Climate uncertainty stems from the unpredictable nature of climate period paths and their corresponding extreme weather events, leading to the loss of potential risk information on weather-dependent energy production and demand in defined climate scenarios\u003csup\u003e32\u003c/sup\u003e. Current climate change further increases uncertainty, exacerbating climate risk to energy systems due to the increasing frequency of extreme climate events and ongoing shifts in climate patterns\u003csup\u003e33\u003c/sup\u003e. Thus, we developed a method to synthesize future representative weather datasets based on regional climate models (RCMs) data to capture potential climate uncertainty information related to weather patterns and extreme weather events (Supplementary Note 2). This method can derive an arbitrary number of stochastic hourly-resolution weather data to decrease the simulation time without destroying month-scale weather features. Thus, this method maximizes our ability to generate enough weather data to measure climate uncertainty information and then trims these data to representative weather scenario sets with occurrence probabilities, which facilitates compatibility with stochastic energy system optimization.\u003c/p\u003e \u003cp\u003eHuman uncertainty comes from the complex stochastic nature of human behavior and their interaction with energy systems, in which the greater control freedom for the residential occupants makes the uncertainty of energy demand profiles even more significant. Deterministic occupant behavior variables not only lead to overestimating or underestimating energy demand profiles at the household level but also decrease the diversity of energy demand profiles at the community level. Thus, in human systems, we sample human-related operational information from assumed probability distributions and parameter ranges based on urban statistical data, ignoring the behavioral logic of the occupants\u003csup\u003e17\u003c/sup\u003e. We explore seven aspects of human behavior uncertainties that affect building energy demand profiles. These features encompass the full range of human-involved operational parameters in building energy simulations. Due to the aleatory nature and weak relationship of uncertainty from the approach, we specified independent probability functions and then drew random samples from these features using Latin Hypercube Sampling (Supplementary Note 3)\u003csup\u003e34,35\u003c/sup\u003e. We specify uniform distribution as probability functions to equal probabilities of each interval in uncertainty variables because there is no explicit reason to value one probability distribution over another. Meanwhile, we extracted the boundary of these variables from the urban residential energy surveys\u003csup\u003e36\u003c/sup\u003e. Every random human behavior scenario was then associated with each climate scenario to create a compound scenario pool to treat the uncertainties in both climate and humans properly.\u003c/p\u003e \u003cp\u003eIn this study, we created eight sets of composite climate-human uncertainty scenario pools in North America (climate zones 1\u0026thinsp;~\u0026thinsp;7), where each scenario pool yielded a total of 50,000 household-level production and demand profiles, i.e., 50 climate scenarios and 1,000 human scenarios.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eModelling regional household benchmark buildings\u003c/h2\u003e \u003cp\u003eBesides climate-human scenarios, an urban residential archetype is crucial in generating household energy scenarios using dynamic energy simulations. We extracted physics-based benchmark building models based on local urban information: urban building survey data and building standards of climate zones (Supplementary Note 4). In this study, we created a set of benchmark bungalows and two-story houses in each city for aggregating artificial urban communities. We combined building information modeling (BIM) and building energy modeling (BEM) to generate household energy scenarios systematically. Specifically, BIM crafts benchmark archetypes with Level of Detail-2 (LoD2), incorporating fundamental construction information such as building typologies, floors, roofs, external walls, and windows, all sourced from urban building surveys. Concurrently, BEM analyzes building energy output by associating benchmark archetypes and climate-human scenarios. This process incorporates energy characteristic factors like envelope construction, internal loads, HVAC systems, and renewable energy systems, with specifications adhering to building standards within the local climate zone. Once completed, the EnergyPlus engine executes energy simulations to analyze hourly-resolution dynamic energy performance.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003eModelling community design layer uncertainties\u003c/h2\u003e \u003cp\u003eParameterizing uncertainties within the community design layer in bottom-up energy models entails creating detailed community models that encompass the full array of urban community features. Extracting community samples from cities to parameterize uncertainties falls into a dilemma between scenario scopes and feature completeness due to the intricate nature of urban community features. Due to the lack of boundary samples, the representative samples derived from regional-average community features may not provide complete uncertainty information for extrapolation to all urban communities. Thus, we create a set of artificial communities using community characteristic matrix: household scale and prosumer scale. The household scale quantifies uncertainties in community composition using boundary and mean urban community features, while the prosumer scale accounts for uncertainties in renewable energy using linear scale expansion. These matrix parameters define the boundary and node samples to mimic community feature information, which provides fully quantitative information on community design layer uncertainties. In this study, the community scale was assumed to be 50 households. Each community characteristic matrix for the local climate zone considered three types of household scales and five sets of prosumer scales.\u003c/p\u003e \u003cp\u003eThe community design layer uncertainties were parameterized in two sequential processes. Initially, we clustered aggregate household energy scenarios into designed community energy scenarios. Subsequently, we assigned household roles as either prosumers or consumers within these community energy scenarios. In the first process, we translated each type of household energy scenario into corresponding shares of community energy scenarios based on household scale using k-means clustering. Subsequently, in the second step, to mitigate the impact of the direction of renewable energy penetration, we arranged households within the community in descending order of energy sharing levels. This assumed that households with higher levels of sharing were more inclined to adopt renewable energy systems. Next, we categorized households into prosumer and consumer roles based on prosumer scale. Finally, this pipeline of uncertainty propagation returned a total of 750 community energy scenarios, i.e., 50 community scenarios under per-community features. The community energy profiles of all communities were aggregated to return 15 profiles for the entire city.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003eModelling system deployment pathways\u003c/h2\u003e \u003cp\u003eFrom a government planning perspective, we model three system deployment pathways that coordinate households into energy community, including household distributed programs, household centralized programs, and community centralized programs\u003csup\u003e37\u003c/sup\u003e. These potential community collaboration programs govern energy allocation and trades to ensure effective and equitable energy flows among households in their local communities. This requires that energy models include energy dispatch strategies and pricing mechanisms under different system deployment pathways. To distinguish the roles of households clearly in energy models, we defined intra-household layers and intra-community layers based on control logic relationships rather than physical relationships. The intra-household layers refer to energy interactions between modules within the system, while the intra-community layers refer to energy interactions between different systems. Consequently, energy dispatch strategies are present in both layers, whereas pricing mechanisms are exclusive to the intra-community layers.\u003c/p\u003e \u003cp\u003eWe detail energy dispatch strategies using a two-stage logical framework which defines energy flow structure in intra-household layers and energy sharing structures in intra-community layers from the rational household perspective (Supplementary Note 5). In this study, our analysis of multiple deployment pathways focuses on intra-community layers that reflect collaborative solutions in local community energy markets. Thus, each deployment pathway features the same energy flow structure but differs in energy sharing structure. The energy flow structure in three deployment pathways identifies twelve operating states of the intra-household layer to manage energy flow hierarchy in system modules under varying module capacity and power states, with the goal of eliminating module interference. To exploit the storage advantages of the green hydrogen systems in large storage capacity and long timescale, we managed the battery module as a short-term energy storage option and the hydrogen module as a long-term energy storage option. Energy sharing structures are used to govern the energy trade methods between prosumers and consumers in the intra-community layer. These structures distinguish trade entities based on ownership of energy systems and trade paths based on internal energy sharing pricing. Household distributed programs describe prosumer-owned deployment pathways which specify each prosumer as an energy trade entity with a three-layer trade path. Household centralized programs describe prosumer-grouped deployment pathways which combine all prosumers as energy trade entities with a two-layer trade path. Community centralized programs describe community-owned deployment pathways that coordinate all households as energy trade entities with a one-layer trade path.\u003c/p\u003e \u003cp\u003eThe pricing model is essential for determining energy sharing pricing between buyers and sellers in intra-community layers to facilitate energy sharing. We utilize the terms \"buyers\" and \"sellers\" to delineate household roles in the pricing model, as prosumers may adapt their behaviors to function as either sellers or buyers based on their net power profiles. We assume that all sellers have equal privilege and are equally influential participants within the energy communities, implying that all sellers should collectively determine energy sharing prices. In accordance with the basic principles of economics, disparities in demand response can result in price fluctuations. This signifies that the market price for buyers favors the grid output price, whereas the market price for sellers favors the grid input price. Thus, we use a dynamic internal pricing model that uses local feed-in tariffs to define trade prices using the production and demand ratio (SDR) of shared energy and price boundaries. Further details can be found in ref.\u003csup\u003e38\u003c/sup\u003e. To maintain consistency with the time resolution of the energy model, the internal prices of the pricing model are adjusted in tandem with the hourly SDR within the boundary limitations. Moreover, the predicted time horizon of the pricing model is set to one hour ahead, with local electricity prices acting as the benchmark prices\u003csup\u003e39,40\u003c/sup\u003e. In this study, electricity sales prices were assumed to be 70% of electricity purchase prices. Notably, we disregarded the energy price fluctuation and local government price incentives during the operating cycle of the energy model.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003eOutline of green hydrogen systems in urban communities\u003c/h2\u003e \u003cp\u003eAs in refs.\u003csup\u003e41,42\u003c/sup\u003e, energy systems, long-term climate resilience systems designed for urban communities, must be considered cross-seasonal and large-capacity energy storage due to the high homogeneity of households in both energy production and demand. Compared to conventional battery storage systems\u003csup\u003e43\u003c/sup\u003e, green hydrogen systems provide higher storage and operation flexibility, which can schedule multi-storage resources in the most efficient way to address uncertainty, especially during extreme energy events. Specifically, green hydrogen systems can achieve synergistic benefits from both the superior efficiency of battery storage and the high energy density and low leakage rate of hydrogen storage. From a technical perspective, the design of green hydrogen systems for urban communities must prioritize safety and portability. Safety is paramount, given the risk of hydrogen leakage and spontaneous combustion\u003csup\u003e44\u003c/sup\u003e. Portability calls for compact equipment and installation methods that are both straightforward and minimally invasive. To meet these criteria, we have chosen photovoltaic panels as the energy production modules and paired them with lithium-ion battery packages and Liquid Organic Hydrogen Carrier (LOHC) equipment for energy storage modules (Supplementary Note 6).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003eModelling design optimization for community green hydrogen systems\u003c/h2\u003e \u003cp\u003eDesign optimization in energy models involves sizing optimal system components of green hydrogen systems to inform evaluation results for metrics of interest. As discussed earlier, multi-source uncertainties should be incorporated to ensure the robustness and stability of evaluation outcomes. Scenario pools are utilized as part of stochastic optimization in design optimization, capturing compound impacts of uncertainty factors. These scenario pools comprise a set of community energy scenarios with occurrence probabilities. Once the energy input scenarios are established, design optimization maps decision space variables onto the objective space through cyclic simulation under inherent system constraints. Specifically, decision variables encompass the rated capacity of the battery package, along with the rated power and capacity of both the fuel cell and electrolyzer. Meanwhile, the objective variables consist of system affordability and independence, quantified as annual life cycle costs (LCC) and grid interaction level (GIL), respectively. In this study, the multi-objective particle swarm algorithm was used to calculate the design optimization part, as it has demonstrated effectiveness in handling multi-scenarios and multi-objectives simultaneously. To enhance the efficiency of the search process and prevent convergence to local minima, we incorporated mechanisms like self-adaptive adjustment of inertia weights\u003csup\u003e45\u003c/sup\u003e and wavelet mutation\u003csup\u003e46\u003c/sup\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003eFormulating objective functions\u003c/h2\u003e \u003cp\u003eThe annual LCC represents the equivalent annual cost of green hydrogen systems over their lifespan in urban communities, including equivalent investment costs, equivalent operation and maintenance costs, carbon tax, and community trade costs, as formulated in Eq.\u0026nbsp;(\u003cspan refid=\"Equ1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Notably, it is imperative to note that all the techno-economic data deployed in this study are derivatives of prevailing current social and technical conditions. The price uncertainties linked with technological upgrades and capital inflation are not considered in this study. The details of objective functions are shown in Supplementary Note 7, including each term of LCC and techno-economic data of green hydrogen systems.\u003cdiv id=\"Equ1\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ1\" name=\"EquationSource\"\u003e\n$$LCC=\\sum _{\\forall s\\in {N}_{s}}{\\rho }_{s}N\\left({C}_{s,inv}+{C}_{s,om}\\right)CRF+\\sum _{\\forall n\\in N}\\sum _{\\forall s\\in {N}_{s}}{\\rho }_{s}\\sum _{\\forall t\\in T}\\left({C}_{n,s,t,tax}+{C}_{n,s,t,community}\\right),\\forall n\\in N,\\forall s\\in {N}_{s},\\forall t\\in T$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e1\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003ewhere, \u003cem\u003en\u003c/em\u003e, \u003cem\u003es\u003c/em\u003e, and \u003cem\u003et\u003c/em\u003e denote the system number, expected community energy scenarios and time series; \u003cem\u003eN\u003c/em\u003e, \u003cem\u003eNs\u003c/em\u003e, and \u003cem\u003eT\u003c/em\u003e denote the total system number (prosumer number in HDP, 1 at HCP and CCP), total scenario number, and total simulation time; \u003cem\u003eρ\u003c/em\u003e\u003csub\u003es\u003c/sub\u003e is the probability of scenario \u003cem\u003es\u003c/em\u003e; CRF denotes the capital recovery factor; \u003cem\u003eC\u003c/em\u003e\u003csub\u003e\u003cem\u003einv\u003c/em\u003e\u003c/sub\u003e and \u003cem\u003eC\u003c/em\u003e\u003csub\u003e\u003cem\u003eom\u003c/em\u003e\u003c/sub\u003e denote the initial investment cost and operation and maintenance cost per system; \u003cem\u003eC\u003c/em\u003e\u003csub\u003e\u003cem\u003en\u003c/em\u003e\u003c/sub\u003e,\u003csub\u003e\u003cem\u003es,t\u003c/em\u003e\u003c/sub\u003e,\u003csub\u003e\u003cem\u003etax\u003c/em\u003e,\u003c/sub\u003e and \u003cem\u003eC\u003c/em\u003e\u003csub\u003e\u003cem\u003en\u003c/em\u003e\u003c/sub\u003e,\u003csub\u003e\u003cem\u003es\u003c/em\u003e\u003c/sub\u003e,\u003csub\u003e\u003cem\u003et\u003c/em\u003e\u003c/sub\u003e,\u003csub\u003e\u003cem\u003ecommunity\u003c/em\u003e\u003c/sub\u003e denote the carbon tax, and community interaction cost at time \u003cem\u003et\u003c/em\u003e in scenario \u003cem\u003es\u003c/em\u003e for system \u003cem\u003en\u003c/em\u003e.\u003c/p\u003e \u003cp\u003eGIL quantifies the level of independence of distributed energy systems by evaluating interactions with the power grid. The objective for urban residential communities is to minimize their interactions with the grid. This approach not only maximizes the role and benefits of communities but also helps to improve the stability of the urban power grid. This study evaluates GIL based on the amount of electricity imported from and exported to the grid, as shown in Eq.\u0026nbsp;(2).\u003c/p\u003e \u003cp\u003e \u003cspan class=\"InlineEquation\"\u003e \u003cspan class=\"mathinline\"\u003e\\(GIL=\\sum _{\\forall n\\in N}\\sum _{\\forall s\\in {N}_{s}}{\\rho }_{s}\\sum _{\\forall t\\in T}\\left({P}_{n,s,t,gridsell}+{P}_{n,s,t,gridbuy}\\right),\\)\u003c/span\u003e \u003c/span\u003e \u003cspan class=\"InlineEquation\"\u003e \u003cspan class=\"mathinline\"\u003e\\(\\forall n\\in N,\\forall s\\in {N}_{s},\\forall t\\in T\\)\u003c/span\u003e \u003c/span\u003e(2)\u003c/p\u003e \u003cp\u003ewhere, \u003cem\u003eP\u003c/em\u003e\u003csub\u003e\u003cem\u003en\u003c/em\u003e\u003c/sub\u003e,\u003csub\u003e\u003cem\u003es,t,gridsell t\u003c/em\u003e\u003c/sub\u003e and \u003cem\u003eP\u003c/em\u003e\u003csub\u003e\u003cem\u003en\u003c/em\u003e\u003c/sub\u003e,\u003csub\u003e\u003cem\u003es,t,gridbuy\u003c/em\u003e\u003c/sub\u003e denote the amount of electricity purchased from the grid and the amount of electricity sold to the grid at time \u003cem\u003et\u003c/em\u003e in scenario \u003cem\u003es\u003c/em\u003e.\u003c/p\u003e \u003c/div\u003e"},{"header":"Declarations","content":"\u003ch2\u003eDeclaration of interests\u003c/h2\u003e \u003cp\u003eThe authors declare no competing interests.\u003c/p\u003e\u003ch2\u003eAuthor contributions\u003c/h2\u003e \u003cp\u003eYou Wu developed the model, performed the analysis, and wrote the paper. Lexuan Zhong supervised the study process and edited the paper.\u003c/p\u003e\u003ch2\u003eAcknowledgements\u003c/h2\u003e \u003cp\u003eThe authors would like to express their gratitude for the partial financial support provided by the China Scholarship Council in collaboration with the University of Alberta and the Natural Sciences and Engineering Research Council of Canada (NSERC) Discovery program.\u003c/p\u003e\u003ch2\u003eData availability\u003c/h2\u003e \u003cp\u003eOriginal data for this paper, as well as the code used for simulation and analyses, are available on Github.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eGlobal Energy and Climate Model Documentation 2023. International Energy Agency, \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://iea.blob.core.windows.net/assets/ff3a195d-762d-4284-8bb5-bd062d260cc5/GlobalEnergyandClimateModelDocumentation2023.pdf\u003c/span\u003e\u003cspan address=\"https://iea.blob.core.windows.net/assets/ff3a195d-762d-4284-8bb5-bd062d260cc5/GlobalEnergyandClimateModelDocumentation2023.pdf\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHong, T., Chen, Y., Luo, X., Luo, N. \u0026amp; Lee, S. H. Ten questions on urban building energy modeling. Building and Environment 168 (2020). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.buildenv.2019.106508\u003c/span\u003e\u003cspan address=\"10.1016/j.buildenv.2019.106508\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHepburn, C. \u003cem\u003eet al.\u003c/em\u003e Towards carbon neutrality and China's 14th Five-Year Plan: Clean energy transition, sustainable urban development, and investment priorities. Environmental Science and Ecotechnology 8 (2021). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.ese.2021.100130\u003c/span\u003e\u003cspan address=\"10.1016/j.ese.2021.100130\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLopez, A., Cole, W., Sergi, B. \u003cem\u003eet al.\u003c/em\u003e Impact of siting ordinances on land availability for wind and solar development. Nat Energy 8, 1034\u0026ndash;1043 (2023). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41560-023-01319-3\u003c/span\u003e\u003cspan address=\"10.1038/s41560-023-01319-3\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTonelli, D., Rosa, L., Gabrielli, P. \u003cem\u003eet al.\u003c/em\u003e Global land and water limits to electrolytic hydrogen production using wind and solar resources. Nat Commun 14, 5532 (2023). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41467-023-41107-x\u003c/span\u003e\u003cspan address=\"10.1038/s41467-023-41107-x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKammen, D. M. \u0026amp; Sunter, D. A. City-integrated renewable energy for urban sustainability. Science 352, 922\u0026ndash;928 (2016). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1126/science.aad9302\u003c/span\u003e\u003cspan address=\"10.1126/science.aad9302\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLowitzsch, J., Hoicka, C. E. \u0026amp; van Tulder, F. J. Renewable energy communities under the 2019 European Clean Energy Package - Governance model for the energy clusters of the future? Renew. Sust. Energ. Rev. 122, 13 (2020). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.rser.2019.109489\u003c/span\u003e\u003cspan address=\"10.1016/j.rser.2019.109489\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBauwens, T. et al. Conceptualizing community in energy systems: A systematic review of 183 definitions. Renew. Sust. Energ. Rev. 156, 16 (2022). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.rser.2021.111999\u003c/span\u003e\u003cspan address=\"10.1016/j.rser.2021.111999\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLevin, T. \u003cem\u003eet al.\u003c/em\u003e Energy storage solutions to decarbonize electricity through enhanced capacity expansion modelling. Nature Energy 8, 1199\u0026ndash;1208 (2023). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41560-023-01340-6\u003c/span\u003e\u003cspan address=\"10.1038/s41560-023-01340-6\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTwitchell, J., DeSomber, K. \u0026amp; Bhatnagar, D. Defining long duration energy storage. Journal of Energy Storage 60, 7 (2023). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.est.2022.105787\u003c/span\u003e\u003cspan address=\"10.1016/j.est.2022.105787\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eParra, D. \u003cem\u003eet al.\u003c/em\u003e An interdisciplinary review of energy storage for communities: Challenges and perspectives. Renew. Sust. Energ. Rev. 79, 730\u0026ndash;749 (2017). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.rser.2017.05.003\u003c/span\u003e\u003cspan address=\"10.1016/j.rser.2017.05.003\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLe, T. S., Nguyen, T. N., Bui, D. K. \u0026amp; Ngo, T. D. Optimal sizing of renewable energy storage: A techno-economic analysis of hydrogen, battery and hybrid systems considering degradation and seasonal storage. Applied Energy 336, 19 (2023). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.apenergy.2023.120817\u003c/span\u003e\u003cspan address=\"10.1016/j.apenergy.2023.120817\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHerbst, A., Toro, F., Reitze, F. \u003cem\u003eet al.\u003c/em\u003e Introduction to Energy Systems Modelling. Swiss J Economics Statistics 148, 111\u0026ndash;135 (2012). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/BF03399363\u003c/span\u003e\u003cspan address=\"10.1007/BF03399363\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBeiter, P., Mai, T. U., Mowers, M. \u0026amp; Bistline, J. Expanded modelling scenarios to understand the role of offshore wind in decarbonizing the United States. Nature Energy 8, 1240\u0026ndash;1249 (2023). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41560-023-01364-y\u003c/span\u003e\u003cspan address=\"10.1038/s41560-023-01364-y\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePerera, A. T. D., Nik, V. M., Chen, D., Scartezzini, J.-L. \u0026amp; Hong, T. Quantifying the impacts of climate change and extreme climate events on energy systems. Nature Energy 5, 150\u0026ndash;159 (2020). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41560-020-0558-0\u003c/span\u003e\u003cspan address=\"10.1038/s41560-020-0558-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLeprince, J. \u003cem\u003eet al.\u003c/em\u003e Can occupant behaviors affect urban energy planning? Distributed stochastic optimization for energy communities. Applied Energy 348, 14 (2023). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.apenergy.2023.121589\u003c/span\u003e\u003cspan address=\"10.1016/j.apenergy.2023.121589\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGabrielli, P., F\u0026uuml;rer, F., Mavromatidis, G. \u0026amp; Mazzotti, M. Robust and optimal design of multi-energy systems with seasonal storage through uncertainty analysis. Applied Energy 238, 1192\u0026ndash;1210 (2019). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.apenergy.2019.01.064\u003c/span\u003e\u003cspan address=\"10.1016/j.apenergy.2019.01.064\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePerera, A. T. D., Khayatian, F., Eggimann, S., Orehounig, K. \u0026amp; Halgamuge, S. Quantifying the climate and human-system-driven uncertainties in energy planning by using GANs. Applied Energy 328, 12 (2022). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.apenergy.2022.120169\u003c/span\u003e\u003cspan address=\"10.1016/j.apenergy.2022.120169\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHuang, P. et al. Characterization and optimization of energy sharing performances in energy-sharing communities in Sweden, Canada and Germany. Applied Energy 326, 16 (2022). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.apenergy.2022.120044\u003c/span\u003e\u003cspan address=\"10.1016/j.apenergy.2022.120044\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKang, H. \u003cem\u003eet al.\u003c/em\u003e Multi-objective sizing and real-time scheduling of battery energy storage in energy-sharing community based on reinforcement learning. Renew. Sust. Energ. Rev. 185, 18 (2023). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.rser.2023.113655\u003c/span\u003e\u003cspan address=\"10.1016/j.rser.2023.113655\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen, X. \u003cem\u003eet al.\u003c/em\u003e Multi-objective optimization of battery capacity of grid-connected PV-BESS system in hybrid building energy sharing community considering time-of-use tariff. Applied Energy 350, 15 (2023). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.apenergy.2023.121727\u003c/span\u003e\u003cspan address=\"10.1016/j.apenergy.2023.121727\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi, Y. \u003cem\u003eet al.\u003c/em\u003e Optimization of integrated energy system for low-carbon community considering the feasibility and application limitation. Applied Energy 348, 19 (2023). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.apenergy.2023.121528\u003c/span\u003e\u003cspan address=\"10.1016/j.apenergy.2023.121528\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDong, S. Y., Kremers, E., Brucoli, M., Rothman, R. \u0026amp; Brown, S. Improving the feasibility of household and community energy storage: A techno-enviro-economic study for the UK. Renew. Sust. Energ. Rev. 131, 17 (2020). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.rser.2020.110009\u003c/span\u003e\u003cspan address=\"10.1016/j.rser.2020.110009\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePerera, A. T. D. \u003cem\u003eet al.\u003c/em\u003e Challenges resulting from urban density and climate change for the EU energy transition. Nature Energy 8, 397\u0026ndash;412 (2023). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41560-023-01232-9\u003c/span\u003e\u003cspan address=\"10.1038/s41560-023-01232-9\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDai, R., Esmaeilbeigi, R. \u0026amp; Charkhgard, H. The Utilization of Shared Energy Storage in Energy Systems: A Comprehensive Review. Ieee Transactions on Smart Grid 12, 3163\u0026ndash;3174 (2021). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1109/tsg.2021.3061619\u003c/span\u003e\u003cspan address=\"10.1109/tsg.2021.3061619\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang, M. \u003cem\u003eet al.\u003c/em\u003e Combined multi-objective optimization and robustness analysis framework for building integrated energy system under uncertainty. Energy Conversion and Management 208, 15 (2020). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.enconman.2020.112589\u003c/span\u003e\u003cspan address=\"10.1016/j.enconman.2020.112589\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGjorgievski, V. Z., Cundeva, S. \u0026amp; Georghiou, G. E. Social arrangements, technical designs and impacts of energy communities: A review. Renewable Energy 169, 1138\u0026ndash;1156 (2021). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.renene.2021.01.078\u003c/span\u003e\u003cspan address=\"10.1016/j.renene.2021.01.078\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCarley, S. \u0026amp; Konisky, D. M. The justice and equity implications of the clean energy transition. Nature Energy 5, 569\u0026ndash;577 (2020). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41560-020-0641-6\u003c/span\u003e\u003cspan address=\"10.1038/s41560-020-0641-6\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMonta\u0026ntilde;\u0026eacute;s, C. C. \u003cem\u003eet al.\u003c/em\u003e Enabling and centering equity and justice in clean energy transition research. Joule 7, 437\u0026ndash;441 (2023). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.joule.2023.02.005\u003c/span\u003e\u003cspan address=\"10.1016/j.joule.2023.02.005\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMavromatidis, G., Orehounig, K. \u0026amp; Carmeliet, J. Comparison of alternative decision-making criteria in a two-stage stochastic program for the design of distributed energy systems under uncertainty. Energy 156, 709\u0026ndash;724 (2018). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.energy.2018.05.081\u003c/span\u003e\u003cspan address=\"10.1016/j.energy.2018.05.081\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXu, L., Feng, K., Lin, N. \u003cem\u003eet al.\u003c/em\u003e Resilience of renewable power systems under climate risks. Nat Rev Electr Eng 1, 53\u0026ndash;66 (2024). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s44287-023-00003-8\u003c/span\u003e\u003cspan address=\"10.1038/s44287-023-00003-8\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMoazami, A., Nik, V. M., Carlucci, S. \u0026amp; Geving, S. Impacts of future weather data typology on building energy performance - Investigating long-term patterns of climate change and extreme weather conditions. Applied Energy 238, 696\u0026ndash;720 (2019). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.apenergy.2019.01.085\u003c/span\u003e\u003cspan address=\"10.1016/j.apenergy.2019.01.085\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTian, W. \u003cem\u003eet al.\u003c/em\u003e A review of uncertainty analysis in building energy assessment. Renew. Sust. Energ. Rev. 93, 285\u0026ndash;301 (2018). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.rser.2018.05.029\u003c/span\u003e\u003cspan address=\"10.1016/j.rser.2018.05.029\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBalvedi, B. F., Ghisi, E. \u0026amp; Lamberts, R. A review of occupant behaviour in residential buildings. Energy and Buildings 174, 495\u0026ndash;505 (2018). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.enbuild.2018.06.049\u003c/span\u003e\u003cspan address=\"10.1016/j.enbuild.2018.06.049\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePetkov, I. \u0026amp; Gabrielli, P. Power-to-hydrogen as seasonal energy storage: an uncertainty analysis for optimal design of low-carbon multi-energy systems. Applied Energy 274, 25 (2020). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.apenergy.2020.115197\u003c/span\u003e\u003cspan address=\"10.1016/j.apenergy.2020.115197\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi, Y., Qian, F. Y., Gao, W. J., Fukuda, H. \u0026amp; Wang, Y. F. Techno-economic performance of battery energy storage system in an energy sharing community. Journal of Energy Storage 50, 21 (2022). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.est.2022.104247\u003c/span\u003e\u003cspan address=\"10.1016/j.est.2022.104247\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu, N. \u003cem\u003eet al.\u003c/em\u003e Energy-Sharing Model with Price-Based Demand Response for Microgrids of Peer-to-Peer Prosumers. Ieee Transactions on Power Systems 32, 3569\u0026ndash;3583 (2017). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1109/tpwrs.2017.2649558\u003c/span\u003e\u003cspan address=\"10.1109/tpwrs.2017.2649558\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAverage energy prices for the United States. U.S. BUREAU of LABOR STATISTICS, \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.bls.gov/regions/home.htm\u003c/span\u003e\u003cspan address=\"https://www.bls.gov/regions/home.htm\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEnergy Fact Book 2023\u0026ndash;2024. NRCan, \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://energy-information.canada.ca/sites/default/files/2023-10/energy-factbook-2023-2024.pdf\u003c/span\u003e\u003cspan address=\"https://energy-information.canada.ca/sites/default/files/2023-10/energy-factbook-2023-2024.pdf\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePan, G. S. \u003cem\u003eet al.\u003c/em\u003e Optimal Planning for Electricity-Hydrogen Integrated Energy System Considering Power to Hydrogen and Heat and Seasonal Storage. Ieee Transactions on Sustainable Energy 11, 2662\u0026ndash;2676 (2020). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1109/tste.2020.2970078\u003c/span\u003e\u003cspan address=\"10.1109/tste.2020.2970078\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGabrielli, P., Gazzani, M., Martelli, E. \u0026amp; Mazzotti, M. Optimal design of multi-energy systems with seasonal storage. Applied Energy 219, 408\u0026ndash;424 (2018). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.apenergy.2017.07.142\u003c/span\u003e\u003cspan address=\"10.1016/j.apenergy.2017.07.142\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSpecial Report: Potential Reliability Impacts of Emerging Flexible Resources. North American Electric Reliability Corporation \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.nerc.com/pa/RAPA/ra/Reliability%20Assessments%20DL/IVGTF_Task_1_5_Final.pdf\u003c/span\u003e\u003cspan address=\"https://www.nerc.com/pa/RAPA/ra/Reliability%20Assessments%20DL/IVGTF_Task_1_5_Final.pdf\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYang, F. Y. \u003cem\u003eet al.\u003c/em\u003e Review on hydrogen safety issues: Incident statistics, hydrogen diffusion, and detonation process. International Journal of Hydrogen Energy 46, 31467\u0026ndash;31488 (2021). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.ijhydene.2021.07.005\u003c/span\u003e\u003cspan address=\"10.1016/j.ijhydene.2021.07.005\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCui, Y. Y., Meng, X. \u0026amp; Qiao, J. F. A multi-objective particle swarm optimization algorithm based on two-archive mechanism. Appl. Soft. Comput. 119, 17 (2022). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.asoc.2022.108532\u003c/span\u003e\u003cspan address=\"10.1016/j.asoc.2022.108532\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTian, D. P. \u0026amp; Shi, Z. Z. MPSO: Modified particle swarm optimization and its applications. Swarm Evol. Comput. 41, 49\u0026ndash;68 (2018). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.swevo.2018.01.011\u003c/span\u003e\u003cspan address=\"10.1016/j.swevo.2018.01.011\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"nature-portfolio","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"","title":"Nature Portfolio","twitterHandle":"","acdcEnabled":false,"dfaEnabled":false,"editorialSystem":"ejp","reportingPortfolio":"","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-4327177/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-4327177/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eIn the context of the firm and enthusiastic development of renewable-based distributed energy systems, high-profit household collaboration strategies are widely recognized as essential for scaling up decentralized green hydrogen systems in urban residential communities. Here we develop bottom-up energy models linking climate, human behavior, and community characteristics to assess the cost-effective impacts of system deployment pathways on community green hydrogen systems for 7 North American climate zones in the 2030\u0026thinsp;~\u0026thinsp;2050 periods. Despite lower energy costs in moderate climate zones compared to hot and cold zones, a consistent pattern in deployment pathway impacts on costs is observed across all zones. The study underscores the critical role of selecting the right deployment pathway for urban decarbonization, with potential cost discrepancies of up to 60% between optimal and suboptimal options. Furthermore, energy storage demands significantly influence energy costs, emphasizing the need to prioritize increased energy storage in pathway design.\u003c/p\u003e","manuscriptTitle":"Probabilistic deployment pathways of scaling up distributed green hydrogen systems for urban residential communities in North America","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-06-04 15:54:52","doi":"10.21203/rs.3.rs-4327177/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"nature-cities","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"natcities","sideBox":"Learn more about [Nature Cities](https://www.springer.com/journal/44284)","snPcode":"44284","submissionUrl":"","title":"Nature Cities","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"79feadc2-eab8-4f98-834e-98a19a0dd434","owner":[],"postedDate":"June 4th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":32732551,"name":"Physical sciences/Energy science and technology/Renewable energy"},{"id":32732552,"name":"Earth and environmental sciences/Climate sciences/Climate change"}],"tags":[],"updatedAt":"2025-01-03T08:10:34+00:00","versionOfRecord":{"articleIdentity":"rs-4327177","link":"https://doi.org/10.1038/s44284-024-00178-7","journal":{"identity":"nature-cities","isVorOnly":false,"title":"Nature Cities"},"publishedOn":"2025-01-02 05:00:00","publishedOnDateReadable":"January 2nd, 2025"},"versionCreatedAt":"2024-06-04 15:54:52","video":"","vorDoi":"10.1038/s44284-024-00178-7","vorDoiUrl":"https://doi.org/10.1038/s44284-024-00178-7","workflowStages":[]},"version":"v1","identity":"rs-4327177","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-4327177","identity":"rs-4327177","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}