LLM summaries of research papers, WIP version 2025-11-03

This is a WIP of standardized LLM output summaries for DYNAMO literature review.

See also:

• ODD Protocol 2020 instructions
• Instructions for LLMs summarizing research articles according to ODD2020 protocol

Version: 2025-11-03

Terminology normalization

Terminology is normalized across summaries (e.g., ABM submodules are always called “decision modules,” data sources always reported under “Input data & calibration”, etc.)

Each entry will use the same structure and terminology, with consistent vocabulary across studies:

• Bibliographic info
• Scope & system studied
• Model type & structure (ODD 2020 headings)
  • Purpose and patterns
  • Entities, state variables, and scales
  • Process overview and scheduling
  • Design concepts
  • Initialization
  • Input data & calibration
  • Submodels / decision modules
• Policy / strategy levers examined
• Key findings
• Relevance for toolkit development
• Strengths & limitations
• Classification for review

Also, other terminology is standardized so that:

• “Submodels” are always described as “decision modules” if they represent agent/sector rules.
• “Input data” always includes data source + calibration info.
• Costs, subsidies, regulations, etc. are grouped under “policy levers.”
• Comparisons are framed in terms of CE markets, policies, and investment assessment relevance.

Comparison notes

Comparison Note: NREL CELAVI + CE ABM (Hanes & Walzberg 2022) vs. IO-Evolutionary Rebound Model (Di Domenico et al. 2023)

Here’s a comparison note you can drop into your literature matrix alongside the structured summaries. It highlights complementarities between the NREL framework (Hanes & Walzberg 2022) and the IO-based rebound model (Di Domenico et al. 2023) in terms of methods, insights, and DYNAMO relevance:

Complementary strengths

• Di Domenico et al. (2023) emphasizes macro-evolutionary dynamics: rebound effects, long-run growth/resource trajectories, and system-wide IO linkages. Strong at capturing structural economy-wide rebound but weak on behavioral heterogeneity and micro-level decision-making.
• Hanes & Walzberg (2022) provides meso- and micro-level resolution: stakeholder decisions modeled via ABM, linked to life cycle environmental assessments through CELAVI. Strong at capturing heterogeneous adoption, supply chain spatiality, and environmental hotspots, but narrower in economic coverage.

Methodological complementarities

• IO + evolutionary dynamics (Di Domenico) = top-down structural feedbacks (scarcity, rebound, long-term economic patterns).
• DES/LCA + ABM (Hanes & Walzberg) = bottom-up heterogeneity (agents, facilities, supply chains) and spatial/temporal environmental impacts.
• Together, they cover both macro rebound dynamics and micro adoption/supply chain mechanisms, a gap noted in DYNAMO’s design documents .

Insights for DYNAMO toolkit

• Linking Finnish ENVIMAT EEIO with ABM supply chain simulators would replicate this complementarity: IO for macro policy/economy-wide impacts, ABM for stakeholder adoption, negotiation, and localized supply chains.
• Potential integration path: use IO-based scarcity/rebound outputs as scenario inputs for ABM models (e.g., biomass or construction material case studies in Finland).
• NREL’s open-source CELAVI code provides a technical reference for combining DES + LCA + ABM, which could be adapted to Finnish/EU data.

Classification for synthesis

• Both should be included in main review, but treated as complementary: Di Domenico (macro IO + rebound), Hanes & Walzberg (meso/micro ABM + LCA).

Comparison Note: Competitive MSW Treatment ABM (He et al. 2017) vs. Backfill Supply Chain ABM (Gan & Cheng 2015)

Complementary strengths

• Gan & Cheng (2015): Focuses on construction backfill reuse logistics; compares centralized optimization vs. decentralized ABM negotiation. Strong demonstration of ABM’s ability to approximate global optima with distributed decision-making in supply chain coordination.
• He et al. (2017): Focuses on competitive municipal solid waste treatment markets; simulates profit-driven competition between STUs and GTUs under regulation. Strong demonstration of ABM’s ability to capture adaptive behavior, price setting, and regulatory feedback.

Methodological complementarities

• Gan & Cheng → ABM as a coordination mechanism (negotiation, logistics).
• He et al. → ABM as a market mechanism (competition, pricing, regulation).
• Together, they cover both sides of CE markets: collaborative logistics for material reuse and competitive dynamics of treatment operators.

Insights for DYNAMO toolkit

• Combining both perspectives enables simulation of multi-actor CE markets where agents must both coordinate exchanges (e.g., construction waste reuse, industrial symbiosis) and compete for treatment contracts (e.g., biogas vs incineration plants).
• Finland/EU applications:
  • Gan & Cheng’s framework transferable to construction/demolition waste flows.
  • He et al.’s framework transferable to biogas vs. incineration in waste-to-energy, or recycling firms competing under EU extended producer responsibility schemes.
• Both highlight the need for policy levers (waste charges, licensing, gate fee caps) to balance cooperation and competition.

Classification for synthesis

• Both include in main review, but represent different dimensions of CE markets: Gan & Cheng (coordination/logistics) and He et al. (competition/regulation).

Comparison Note: Multi-method Food Waste Models (Kandemir et al. 2020) vs. Competitive MSW Treatment ABM (He et al. 2017)

Complementary strengths

• Kandemir et al. (2020): Broad, multi-method focus (DES, ML/BN, ABM, mass balance) applied to food waste prevention at household and retail levels. Emphasizes behavioral and social dynamics (social norms, consumer practices, innovation adoption).
• He et al. (2017): Narrow, ABM-specific focus on competitive municipal waste treatment markets. Emphasizes market competition, pricing strategies, and regulatory policy design.

Methodological complementarities

• Kandemir et al. → ABM as a tool for consumer/retailer decision-making and behavioral interventions, often integrated with Bayesian networks for social norm diffusion.
• He et al. → ABM as a tool for competitive dynamics and regulatory testing in treatment markets.
• Together, they demonstrate ABM’s versatility in CE contexts: from micro-level social behavior to meso-level market competition.

Insights for DYNAMO toolkit

• Kandemir’s ABM-BN approach could inform consumer and SME behavior modules in Finnish CE models, particularly where social norms and awareness campaigns are policy levers.
• He’s ABM framework could inform market-level competition modules for CE treatment operators (e.g., recycling vs incineration plants), including gate fee caps and licensing schemes.
• Combining both could create integrated CE market simulators where household/retail behaviors (demand-side waste generation) interact with operator competition (supply-side treatment).

Classification for synthesis

• Both include in main review, but represent different system levels: Kandemir et al. (micro-level consumer/retail prevention) vs. He et al. (meso-level treatment market competition).

Cluster notes

This list will grow dynamically — every time we process a new paper, I’ll either assign it to an existing cluster or create a new one, and update the index.

Cluster 1: Macro vs. Meso CE Models – IO Rebound vs. Hybrid ABM-LCA

(Di Domenico et al. 2023; Hanes & Walzberg 2022)

This cluster illustrates the complementarity between macro-scale rebound modeling and meso-scale adoption/supply chain modeling in circular economy research. Di Domenico et al. (2023) use a macro-evolutionary input–output framework to show how resource scarcity and efficiency interact with rebound effects, emphasizing long-term economy-wide dynamics. In contrast, Hanes & Walzberg (2022) integrate agent-based modeling, discrete event simulation, and life-cycle assessment in the CELAVI and CE ABM frameworks to simulate end-of-life management of renewable energy technologies. While the former captures structural feedbacks across sectors, the latter focuses on heterogeneous stakeholder decisions and supply chain pathways. Together, they demonstrate the value of multi-scale integration: IO models to explore scarcity and rebound at the macro level, and ABM-LCA hybrids to assess adoption, environmental hotspots, and policy interventions at the technology level. For toolkit development in Finland and the EU, combining EEIO models like ENVIMAT with ABM-LCA approaches could provide a more complete assessment of CE policy impacts.

Shared theme
Both studies model CE strategies at scale, but from different system levels and methodological angles:

• Di Domenico et al. (2023): Macro-evolutionary IO model for rebound effects and resource scarcity.
• Hanes & Walzberg (2022): Hybrid ABM + DES + LCA framework (CELAVI + CE ABM) for technology-specific CE transitions in renewable energy.

Key complementarities

• Macro IO model (Di Domenico): Captures economy-wide feedbacks (growth, scarcity, rebound). Strong at structural, long-term system dynamics.
• Hybrid ABM-LCA model (Hanes & Walzberg): Captures stakeholder adoption and supply chain impacts at the technology level (wind, PV). Strong at behavioral heterogeneity and environmental performance.
• Together: Offer a multi-scale perspective, linking structural economic constraints with detailed supply chain adoption and environmental outcomes.

System levels covered

• Macro-level: Resource scarcity, rebound, and growth trajectories.
• Meso/micro-level: Adoption dynamics, supply chains, and EOL pathway choices.

Insights for DYNAMO toolkit

• Suggests a pathway to integrate ENVIMAT (EEIO) with ABM-based simulators: macro rebound constraints could feed into micro adoption and supply chain models.
• Provides a blueprint for multi-scale CE modeling in Finland/EU: linking national IO analysis with case-specific CE simulators (e.g., construction waste, biogas, renewable energy).
• Highlights complementarity between top-down policy evaluation (scarcity, rebound) and bottom-up behavioral adoption (agents, facilities, pathways).

Classification for synthesis
Both include in main review, forming a Macro vs. Meso CE model cluster that illustrates multi-scale integration potential.

Cluster 2: Waste System ABMs – Supply Chains, Competition, and Consumer/Retail Behavior

ChatGPT said: I’ll start building clusters as we go, not just pairwise comparisons. Each cluster will bring together 3+ papers that address a common theme (e.g. waste management, CE supply chains, innovation adoption), so you can later use them as the backbone of your literature review chapters or subsections.

(Gan & Cheng 2015; He et al. 2017; Kandemir et al. 2020; Kerdlap et al. 2020)

This cluster demonstrates the versatility of agent-based modeling in addressing waste system challenges across logistics, competition, behavior, and infrastructure design. Gan & Cheng (2015) show how ABM can replicate and approximate centralized optimization for coordinating backfill reuse in construction waste supply chains, emphasizing decentralized negotiation. He et al. (2017) highlight the role of ABM in simulating competitive municipal solid waste treatment markets, focusing on gate fee dynamics, regulatory design, and operator sustainability. Kandemir et al. (2020) broaden the scope to food waste, integrating ABM with Bayesian networks to capture consumer social norms and retailer innovation adoption, supported by complementary DES and ML/BN methods. Kerdlap et al. (2020) combine ABM with life cycle assessment to evaluate centralized versus distributed recycling systems for plastics in Singapore, illustrating how ABM outputs can generate inventory data for environmental impact models. Together, these studies show how ABM can be applied at multiple levels of waste systems—households, retail, treatment markets, logistics, and recycling infrastructure—and how it can integrate with optimization and LCA. For Finland/EU toolkit development, this cluster provides building blocks for multi-layered CE waste models, from consumer interventions to infrastructure planning and policy evaluation.

Shared theme
All three papers use agent-based approaches (alone or combined with other methods) to study waste generation, treatment, or prevention in CE contexts. They collectively span the supply chain, market, and consumer/retail layers of waste systems.

Key complementarities

• Gan & Cheng (2015): ABM as a coordination tool for backfill reuse logistics; demonstrates distributed negotiation approximating global optima.
• He et al. (2017): ABM as a competition tool for waste treatment markets; demonstrates adaptive pricing, regulation, and sustainability trade-offs.
• Kandemir et al. (2020): ABM (with BN integration) as a behavioral tool for consumer and retailer food waste prevention; demonstrates role of social norms, innovation adoption, and awareness campaigns.

System levels covered

• Micro-level: Household consumer choices and social interactions (Kandemir).
• Meso-level: Operator competition and regulatory oversight (He).
• Logistics level: Coordination of waste flows between sites (Gan & Cheng).

Insights for DYNAMO toolkit

• Together, these models show how ABM can capture heterogeneity, decentralization, and adaptation in waste-related CE systems.
• A DYNAMO-style Finnish/EU toolkit could integrate these levels:
  • Household/SME modules (consumer behavior, norms, prevention).
  • Market modules (competition between treatment/recycling operators).
  • Supply chain modules (waste material logistics, industrial symbiosis).
• This cluster therefore provides a methodological foundation for CE market simulators that connect behavioral, economic, and logistical layers.

Classification for synthesis
All three include in main review, and should be grouped as a Waste System ABM cluster in the literature review.

Cluster 3: Industrial Symbiosis & Secondary Material Markets

(Termizi et al. 2023)

Shared theme
This cluster covers ABM and related approaches modeling industrial symbiosis and secondary material markets, where by-products or wastes from one sector are exchanged as inputs in another. The focus is on cross-industry coordination, cost-sharing, and economic/environmental benefits.

Anchor study
Termizi et al. (2023) use an ABM to analyze industrial symbiosis between sugar producers and cement manufacturers in Malaysia, where mud cake waste substitutes for limestone. The model tests substitution percentages (25–100%) and cost-sharing schemes for transport. Results show substantial cost savings, especially under shared cost arrangements, with environmental benefits from reduced landfill disposal and limestone extraction.

Implications for toolkit development

• Provides a template for simulating economic exchanges and cost-sharing mechanisms in CE industrial networks.
• Relevant for Finnish/EU cases like pulp/paper residues in construction materials, steel slag reuse, or bio-based industrial symbiosis.
• Can inform profitability calculators and CE market simulators by embedding industrial coordination logics.

Classification for synthesis
Include in main review. Forms the anchor for an Industrial Symbiosis & Secondary Material Markets cluster.

Forward-looking note
Potential future members include:

• ABM/SD models of eco-industrial parks and by-product exchanges.
• Models of secondary steel, plastics, or aluminum markets.
• Studies integrating industrial symbiosis with IO/LCA frameworks.
• European case studies on industrial symbiosis (e.g., Kalundborg, Sweden/Finland pulp and paper residues).
• ABMs/SDs modeling by-product exchanges, secondary material flows, and CE industrial networks.

Cluster 4: Circular Bioeconomy & Biogas Systems

• Expected members: ABM/SD/optimization of manure-to-biogas, bio-based value chains, and agricultural CE systems.

Cluster 5: Investment Profitability & CE Business Models

• Expected members: papers on profitability calculators, adoption thresholds, and SME investment dynamics.

Cluster 6: Multi-Method Integration (ABM + IO + LCA + ML)

• Expected members: works explicitly combining bottom-up and top-down modeling (e.g., ABM-BN, IO-LCA hybrids).

Cluster 7: Policy Levers & Transformation Pathways

• Expected members: studies explicitly testing subsidies, taxes, bans, quotas, and regulatory designs in CE markets.

Cluster 8: Consumer Behavior & Product-Service Strategies

(Koide et al. 2023)

Shared theme
This cluster focuses on modeling household decision-making and product-service system (PSS) strategies as levers for advancing circular economy transitions. Unlike waste system models that emphasize logistics or operator competition, these models target consumer heterogeneity, social networks, and adoption of service-based CE options.

Anchor study
Koide et al. (2023) present an advanced ABM framework simulating the diffusion of seven CE strategies (repair, reuse, refurbishing, functional upgrading, leasing, renting, and sharing) across 5,000 households over 30 years. Consumer behavior is modeled via random utility, awareness sets, social interactions, and obsolescence distributions, while environmental impacts are evaluated with consequential LCA.

Key insights

• Promotion measures (advertising, service improvements, pricing) shape different phases of adoption curves.
• Strategies interact, creating bottlenecks (e.g., refurbishing limited by collection flows).
• Sharing/rental reduces product stocks but may increase emissions; reuse is fast but limited in scope.
• Rebound effects emerge if consumer preferences misalign with CE goals.

Implications for toolkit development
For Finland/EU, this cluster highlights the importance of consumer-targeted policies and behavioral interventions in CE transitions. The Koide model offers a methodological benchmark for integrating survey-calibrated consumer ABMs with profitability calculators and environmental assessments. It opens a path to simulate how awareness campaigns, subsidies, or service quality improvements alter CE adoption trajectories and emissions outcomes.

Classification for synthesis
Include in main review. Forms the anchor for a Consumer Behavior & Product-Service Strategies cluster, to be expanded as more studies on consumer CE adoption and PSS diffusion are reviewed.

Cluster 9: Systematic Reviews & Frameworks for CE ABMs

(Rizzati & Landoni 2024)

Shared theme
This cluster captures systematic reviews, meta-analyses, and framework-building works that synthesize how ABM has been applied to circular economy systems. These studies provide overviews, taxonomies, and recommendations for generalizable CE models.

Anchor study
Rizzati & Landoni (2024) review 100 CE ABM studies, categorizing them into producer, consumer, and innovation diffusion themes. They conclude that CE ABMs are fragmented, mostly partial equilibrium, and often case-specific, but emphasize ABM’s unique value for modeling heterogeneity, bounded rationality, and emergent dynamics. The study proposes integration with IO and stock-flow consistent models as a key future direction.

Implications for toolkit development

• Serves as a meta-level anchor for DYNAMO’s literature review, mapping the field and gaps.
• Highlights the importance of designing generic, reusable agent heuristics (for firms, consumers, governments).
• Supports rationale for linking ABMs to Finland’s ENVIMAT EEIO model.

Classification for synthesis
Include in main review. Forms the anchor for a Systematic Reviews & Frameworks cluster.

Cluster 10: Indicator Frameworks for CE Decision Support/ Assessment methods, product lifetimes, and obsolescence

(Anchor: Kravchenko et al. 2019; also Harris et al. 2021; Iraldo et al. (2017), Hischier & Böni (2021), Glöser-Chahoud et al. (2021), Harris et al. (2021), Kravchenko et al. (2019).)

Shared theme
This cluster brings together works that provide sustainability indicator frameworks for evaluating CE initiatives—particularly those designed for ex-ante screening (decision support before implementation). These don’t model dynamic behavior directly but define what to measure and how to interpret it.

Anchor study
Kravchenko et al. (2019) consolidate 279 leading sustainability-related indicators across environmental, social, and economic dimensions. Indicators are mapped to CE strategies (reuse, remanufacturing, PSS, recycling, etc.) and business processes (design, operations, after-sales, end-of-life, business model). Key contribution: a structured, transferable database for ex-ante screening of CE initiatives.

Additional member
Harris et al. (2021) provides a scoping review highlighting the pitfalls of circularity indicators that are disconnected from environmental outcomes (“circularity for circularity’s sake”). It fits this cluster as a critique, complementing the constructive indicator framework of Kravchenko et al.

Implications for toolkit development

• Provides a menu of metrics (leading indicators) that can be embedded into profitability calculators, CE simulators, and ABM/SD models.
• Helps avoid “circularity for circularity’s sake,” ensuring metrics connect CE loops to real sustainability outcomes.
• Directly useful for Finnish/EU contexts where indicator frameworks must align with policy evaluation and corporate reporting.

Potential future members

• OECD (2019) – Business models for CE (indicator-driven evaluation frameworks).
• EU/ISO standards on CE indicators (e.g. EC Circularity Metrics).
• Works on social sustainability indicators (to fill identified gaps).

Classification for synthesis
Include in main review. Forms the anchor for an Indicator Frameworks for CE Decision Support cluster.

Cluster 11: Consumer Perspectives on CE

This cluster would group together work that is not model-based in itself, but which directly targets consumption, user acceptance, and behavioural dimensions of circularity. It’s complementary to the ABM-focused consumer-behaviour papers (like Scalco et al. 2017 or Tong et al. 2018), since it’s a broad review that maps out the conceptual and empirical landscape.

• Core paper: Camacho-Otero et al. 2018 (review of 111 studies).
• Scope: Drivers and barriers to consumer acceptance of CE, meanings and dynamics of consumption, integration of user perspectives in design.
• Tags: consumer acceptance, behavioural factors, sharing economy, PSS, remanufacturing, sustainable consumption.
• Relation to other clusters:
  • Complements Cluster 2 (Behavioural ABM of recycling/consumption), which models these aspects.
  • Overlaps partly with Cluster 10 (Assessment frameworks, metrics, LCA/LCC) when consumer acceptance is evaluated as an impact factor.
  • Useful input to Cluster 5 (Business models in CE), since adoption hinges on consumer willingness.

Integration Note: Cluster 10 – Consumer Perspectives on CE

Bridge role

• Cluster 11 connects empirical/qualitative consumer research with simulation and policy modelling.
• Papers like Camacho-Otero et al. (2018) provide the conceptual foundations and empirical evidence on consumer drivers, barriers, and meanings of circular consumption.
• These insights can feed directly into behavioural modules of ABM/SD models in Clusters 2 and 5, ensuring that adoption, repair, reuse, or sharing decisions reflect real-world attitudes rather than abstract assumptions.

Contribution to toolkit development

• Helps define behavioural parameters (e.g., trust, risk perception, status value, convenience) for profitability calculators and CE market simulators.
• Offers user-centred barriers and drivers to be tested as policy levers (e.g., incentives, awareness campaigns, labelling, social influence).
• Ensures CE modelling doesn’t fall into “supply-side determinism” by overlooking consumer acceptance.

Relation to other clusters

• Cluster 2 (Behavioural ABM of recycling/consumption): Provides qualitative/empirical grounding for behavioural rules.
• Cluster 5 (Investment & CE Business Models): Consumer willingness to adopt is a core determinant of business model viability.
• Cluster 10 (Indicators/Decision Support): Complements metrics by adding a social/behavioural dimension to CE performance assessment.

  Standardized summaries

  Di Domenico et al. - 2023 - Resource scarcity, circular economy and the energy rebound

Abstract We propose an Agent-Based Stock-Flow Consistent model combined with a simplified Input-Output (IO) structure of production. In the model, heterogeneous firms interact in the energy, material, capital and consumption markets. Materials for production of consumer goods can be manufactured using non-renewable or recycled material inputs. We examine the conditions under which the circular economy emerges through market mechanisms, as well as it can be a source of the rebound effect. An important novelty of our approach is that recycling and mining sectors employ different types of capital for production. Capital goods (machinery) are produced by capital firms, which constantly engage in innovations to improve their technological features. This way we endogenize changes in technological coefficients of the Input-Output tables. We show that sectoral interdependencies along the value chain can render the energy rebound effect due to the circular economy (CE) even if energy intensity of the recycling process is lower compared to mining. This effect depends on the impact of investment in the recycling sector on the aggregate demand and energy use along the entire value chain. We assess the role of macroeconomic policies, namely mission-oriented innovation policies (MOIPs), active fiscal policies (AFPs), and environmental taxes in fostering the CE transition, while mitigating the rebound effect. We find that the combination of MOIPs and active fiscal policy is the most effective in promoting the circular economy, preserving employment and ensuring a sustainable growth path.

Bibliographic info
Di Domenico, C., Favero, A., Sapio, A., & Sgobbi, A. (2023). Resource scarcity, circular economy and the energy rebound: A macro-evolutionary input-output model. Structural Change and Economic Dynamics, 64, 360–378. https://doi.org/10.1016/j.strueco.2023.01.009

Scope & system studied

• Sector: Economy-wide resource and energy use.
• Geography: Global/EU focus.
• CE strategies: Recycling, reuse, and resource efficiency under conditions of resource scarcity.

Model type & structure (ODD 2020 headings)

• Purpose and patterns: Explore how CE strategies interact with resource scarcity and energy efficiency to affect macroeconomic dynamics and rebound effects.
• Entities, state variables, and scales: Economic sectors (entities); state variables include capital stock, IO coefficients, resource and energy use; temporal scale spans decades.
• Process overview and scheduling: Iterative production, consumption, and investment updates; endogenous changes in IO coefficients through innovation and diffusion dynamics.
• Design concepts:
  • Emergence: Aggregate growth, rebound, and scarcity outcomes arise from sectoral adaptation.
  • Adaptation: Firms innovate and adopt CE technologies under scarcity or policy.
  • Stochasticity: Innovation/adoption modeled probabilistically.
• Initialization: Calibrated to a base year of WIOD (World Input–Output Database).
• Input data & calibration: WIOD IO tables with environmental extensions; parameters for innovation/adoption from evolutionary economics literature.
• Decision modules: Innovation and adoption of CE strategies; rebound mechanisms; resource constraints feedbacks.

Policy / strategy levers examined

• Resource efficiency improvements.
• CE adoption (recycling, reuse).
• Resource scarcity shocks.
• (No explicit subsidies/taxes modeled, but implications for policy design are discussed).

Key findings

• Rebound effects are significant: efficiency gains often increase total energy use unless scarcity or CE strategies constrain it.
• CE strategies mitigate but do not fully eliminate rebound.
• Resource scarcity accelerates CE adoption but constrains growth.
• Policy needs to integrate CE measures with energy efficiency to avoid rebound traps.

Relevance for toolkit development

• Demonstrates how to hybridize IO with dynamic innovation/adoption, complementing EEIO models like ENVIMAT.
• Useful for long-run scenario analysis of CE policies.
• Less suited for micro-level profitability calculators, but highly relevant for CE market simulators and policy testing.

Strengths & limitations

• Strengths: Integrates IO with evolutionary dynamics; transparent use of empirical IO data; explicit rebound modeling.
• Limitations: Aggregate/sectoral focus (no micro heterogeneity); abstract policy representation; not spatially explicit for Finland; no direct ABM elements.
• Transferability: Good for EU/Finnish contexts if adapted with national IO data and CE policy specifics.

Classification for review
Include in main review.

Cluster membership
Cluster 1: Macro vs. Meso CE Models – IO Rebound vs. Hybrid ABM-LCA.

[Post-processing note: macro IO rebound model; pairs with Hanes & Walzberg; place under multi-scale modeling section]

Gan and Cheng - 2015 - Formulation and analysis of dynamic supply chain of backfill in construction waste management using agent-based modeling

Abstract Backfill is the excavated material from earthworks, which constitutes over 50% of the construction wastes in Hong Kong. This paper considers a supply chain that consists of construction sites, landfills and commercial sources in which operators seek cooperation to maximize backfill reuse and improve waste recovery efficiency. Unlike the ordinary material supply chain in manufacturing industries, the supply chain for backfill involves many dynamic processes, which increases the complexity of analyzing and solving the logistic issue. Therefore, this study attempts to identify an appropriate methodology to analyze the dynamic supply chain, for facilitating the backfill reuse. A centralized optimization model and a distributed agent-based model are proposed and implemented in comparing their performances. The centralized optimization model can obtain a global optimum but requires sharing of complete information from all supply chain entities, resulting in barriers for implementation. In addition, whenever the backfill supply chain changes, the centralized optimization model needs to reconfigure the network structure and recompute the optimum. The distributed agent-based model focuses on task distribution and cooperation between business entities in the backfill supply chain. In the agent-based model, decision making and communication between construction sites, landfills, and commercial sources are emulated by a number of autonomous agents. They perform together through a negotiation algorithm for optimizing the supply chain configuration that reduces the backfill shipment cost. A comparative study indicates that the agent-based model is more capable of studying the dynamic backfill supply chain due to its decentralization of optimization and fast reaction to unexpected disturbances.

Bibliographic info
Gan, V.J.L., & Cheng, J.C.P. (2015). Formulation and analysis of dynamic supply chain of backfill in construction waste management using agent-based modeling. Advanced Engineering Informatics, 29(4), 878–888. https://doi.org/10.1016/j.aei.2015.01.004

Scope & system studied

• Sector: Construction waste management (excavated backfill).
• Geography: Hong Kong.
• CE strategy: Reuse of backfill between construction sites to reduce landfill disposal and costs.

Model type & structure (ODD 2020 headings)

• Purpose and patterns: Identify suitable modeling approaches for dynamic supply chains and evaluate ABM vs centralized optimization for backfill recovery.
• Entities, state variables, and scales:
  • Agents: Construction sites (producers, consumers, or both), landfills, commercial suppliers.
  • State variables: Backfill stock, excavation/backfilling schedules, cost parameters, site capacity.
  • Temporal scale: Monthly phases of construction projects.
  • Spatial scale: Regional network of sites in Hong Kong.
• Process overview and scheduling: Producers and consumers negotiate backfill exchanges via a contract-net protocol (task announcement, bidding, awarding); decisions update each time step. Centralized optimization model solves same network as a global optimum benchmark.
• Design concepts:
  • Emergence: Supply chain material flow patterns.
  • Adaptation: Agents respond to local availability, demand, and costs.
  • Interaction: Negotiation and bidding between agents.
  • Stochasticity: Incorporated via demand variation using stochastic programming.
• Initialization: Excavation/backfilling schedules of six construction sites; landfill and one commercial supplier.
• Input data & calibration: Empirical data on project schedules, transportation costs, material prices, and landfill disposal charges; parameters sourced from local construction/waste policies.
• Decision modules:
  • Producer module: Optimize export rate considering holding and dumping costs.
  • Consumer module: Optimize import rate to minimize holding/procurement costs.
  • Negotiation module: Contract net protocol for producer–consumer agreements.

Policy / strategy levers examined

• Waste disposal charges (government landfill levy).
• Coordination mechanisms: centralized optimization vs distributed ABM.

Key findings

• Centralized optimization yields slightly lower costs (0.3% difference) but requires unrealistic full information sharing.
• ABM produces near-optimal solutions while better reflecting decentralized decision-making, limited information exchange, and dynamic adaptability.
• ABM reacts faster to disturbances (new projects, delays) than centralized approaches.
• Trade-off: global optimality vs realism and resilience.

Relevance for toolkit development

• Strong demonstration of ABM for decentralized CE supply chains.
• Negotiation algorithms could inform CE market simulators matching supply and demand of secondary materials.
• Highly relevant for designing decision-support tools in CE logistics.

Strengths & limitations

• Strengths: Realistic modeling of decentralized actors; transparent comparison of centralized vs distributed approaches; case-based calibration.
• Limitations: Narrow focus (backfill only); simple case (one landfill, one commercial source); no direct integration with IO/LCA; global optimality not guaranteed.
• Transferability: Concepts transferable to EU/Finland in construction waste and industrial symbiosis contexts.

Classification for review
Include in main review.

Cluster membership
Cluster 2: Waste System ABMs – Supply Chains, Competition, and Consumer/Retail Behavior.

[Post-processing note: backfill reuse logistics; ABM vs optimization; aligns with He et al. and Kandemir et al. in waste systems cluster]

Hanes and Walzberg - 2022 - Circular Economy Modeling Efforts at NREL

Abstract Circular Economy (CE) aims at decoupling human activities from economic growth and resource use. While CE economic and environmental benefits are often touted by its proponents, increased circularity and improved sustainability do not necessarily go hand-in-hand. The Circular Economy Lifecycle Assessment and Visualization (CELAVI) framework simulates the transition of renewable energy technology supply chains towards circularity and quantifies spatially explicit environmental impacts and supply chain costs. Using wind blade end-of-life (EOL) management as a case study, this presentation will highlight CELAVI’s features. Furthermore, envisioned extension to the framework could improve the stakeholder decision model by accounting for logistical and other factors. The CELAVI framework could be leveraged to answer crucial questions regarding renewables EOL management.

Bibliographic info
Hanes, R., & Walzberg, J. (2022). Circular Economy Modeling Efforts at NREL: Circular Economy Lifecycle Analysis and Visualization (CELAVI) and the Circular Economy Agent-Based Model (CE ABM). National Renewable Energy Laboratory (NREL) presentation, June 16, 2022. NREL/PR-6A20-83000. https://github.com/NREL/celavi/

Scope & system studied

• Sector: Renewable energy technologies (wind turbines, photovoltaics, hard-disk drives as case studies).
• Geography: Primarily United States.
• CE strategy: Circular end-of-life (EOL) management pathways—reuse, repair, recycling, and landfill avoidance—integrated with supply chain and environmental impact assessments.

Model type & structure (ODD 2020 headings)

• Purpose and patterns: Identify conditions under which CE strategies maximize value retention and minimize virgin material inputs for energy technologies .
• Entities, state variables, and scales:
  • CELAVI: Supply chain nodes (materials, components, plants, facilities), pathways, costs, and environmental impacts.
  • CE ABM: Six agents (manufacturers, wind developers, wind plant owners, recyclers, regulators, landfills) with heterogeneous rules; state variables include technology type, capacity, costs, and behavior modeled via Theory of Planned Behavior .
  • Spatial scale: U.S. states; temporal scale: multiyear dynamic simulations (2020–2050).
• Process overview and scheduling:
  • CELAVI: Discrete event simulation (tracks material/component flows), Cost Graph (Bellman-Ford optimization of EOL pathways), PyLCIA (process-based life cycle impact assessment, TRACI indicators) .
  • CE ABM: Annual time steps; assets enter EOL via Weibull distributions; agents select EOL options (repair, recycle, landfill) based on TPB scoring .
• Design concepts:
  • Emergence: Shifts in circularity metrics, environmental impact hotspots, adoption patterns.
  • Adaptation: Stakeholder decisions shaped by costs, regulation, and technology type.
  • Stochasticity: Weibull lifetime draws, heterogeneous stakeholder decisions.
• Initialization: U.S. Wind Turbine Database for plant-level initialization; supply chain inventories seeded with materials and processes .
• Input data & calibration: U.S. LCI database, ReEDS (electricity grid data), TRACI 2.1 impact factors, wind blade case study (Iowa & Missouri, 2000–2050). Weibull parameters for lifetimes and stakeholder TPB parameters calibrated from empirical data .
• Decision modules:
  • EOL decision module (agents choosing repair/recycle/landfill via TPB).
  • Cost Graph optimization module for supply chain pathways.
  • Environmental impact calculation module (PyLCIA). Policy / strategy levers examined
• Recycling technology choice (mechanical, cement co-processing, pyrolysis).
• Transportation and processing costs.
• Material substitution (thermoset vs thermoplastic blades).
• Regulatory context (landfill bans, recycling incentives, Energy Act of 2020 provisions for wind recycling)
• Processing costs are not major barriers to implementing circular pathways; recycling pathways adopted over landfilling in most scenarios .
• Outflow circularity can reach 1.0 under cement co-processing, though environmental impacts are higher due to energy/emissions .
• Small increases in GWP (<10%) during circular transitions compared to full supply chain impacts .
• Transportation and blade size reduction strongly affect costs.
• Regulatory heterogeneity (state-level definitions of wind blades as waste types) influences EOL outcomes.
• Integration of CELAVI and CE ABM promises to combine behavioral adoption modeling with detailed environmental and supply chain impact assessments .

Relevance for toolkit development

• Demonstrates hybrid frameworks (DES + LCA + ABM) for CE transitions.
• Useful for Finnish/EU CE modeling: integration of behavioral adoption dynamics (ABM) with LCA/IO-based assessments.
• Provides transferable methods for profitability calculators (pathway costs), market simulators (stakeholder decision-making), and policy analysis (EOL strategies, recycling incentives).

Strengths & limitations

• Strengths: Open-source Python implementation (CELAVI on GitHub), transparent modular structure, strong coupling of behavioral and environmental assessment.
• Limitations: U.S.-centric case studies; early-stage integration with IO models (future work includes linking to BEIOM EEIO model) ; limited to energy technologies though framework extensible.
• Transferability: High methodological relevance to EU/Finnish CE policies; adaptation requires local data.

Classification for review
Include in main review. 

Cluster membership
Cluster 1: Macro vs. Meso CE Models – IO Rebound vs. Hybrid ABM-LCA.

[Post-processing note: hybrid ABM + LCA framework; pairs with Di Domenico; highlight CELAVI open-source code for toolkit inspiration]

He et al. - 2017 - Managing competitive municipal solid waste treatment systems An agent-based approach

Abstract Private sector participation in municipal solid waste (MSW) management is increasingly being applied in many countries recently. However, it remains a largely unexplored issue whether different self-interested treatment operators can co-exist in an economically feasible and sustainable manner. To help the policy-maker s understand and manage competitive MSW treatment systems, this paper proposes an agent-based waste treatment model (AWTM) that consists of four types of agents, namely the refuse collector, specialized treatment unit (STU), the general treatment unit (GTU), and the regulator. An estimation-and-optimization approach is developed for profit-maximizing agents to set optimal gate fee and vie for specific waste in low-information competition. Based on the Singapore case, the experimental results imply that if the regulator deliberately promotes the STUs by intervening in waste allocation, the GTU could give up competing for the waste and greatly increase its gate fees as retaliation. Besides, driven by the increasing gate fee of the GTU, the STUs conservatively raise their gate fee; while the GTU will be the major beneficiary in the AWTM. Finally, to identify the optimal mixed policy under predefined constraints, the AWTM is integrated into a simulation-based optimization problem, which is solved by a genetic algorithm.

Bibliographic info
He, Z., Xiong, J., Ng, T. S., Fan, B., & Shoemaker, C. A. (2017). Managing competitive municipal solid waste treatment systems: An agent-based approach. European Journal of Operational Research, 263(3), 1063–1077. https://doi.org/10.1016/j.ejor.2017.05.028

Scope & system studied

• Sector: Municipal solid waste (MSW) treatment, focusing on competition between treatment operators.
• Geography: Case study based on Singapore’s MSW system, but generalizable to other urban contexts.
• CE strategy: Integration of specialized treatment units (STUs, e.g., anaerobic digestion) with general treatment units (GTU, e.g., incineration), and regulatory design for efficient, sustainable waste-to-energy pathways.

Model type & structure (ODD 2020 headings)

• Purpose and patterns: Explore whether private waste treatment operators can co-exist sustainably under competition, and how regulation affects system efficiency and stability.
• Entities, state variables, and scales:
  • Agents: refuse collector, multiple heterogeneous STUs, one GTU, and a regulator.
  • State variables: gate fees, profits, operation costs, waste volumes, residues, waste allocation preferences.
  • Scale: City-level system, modeled dynamically over 1,000 time steps.
• Process overview and scheduling:
  • Refuse collector allocates waste using a multinomial logit demand model (MLDM).
  • STUs and GTU set gate fees using an estimation–optimization process under low-information conditions.
  • Regulator alters exogenous parameters (STU count, gate fee bounds, allocation preferences).
  • Simulation runs multiple experiments across scenarios.
• Design concepts:
  • Emergence: Market stability/instability, profit distribution, unintended effects of policies.
  • Adaptation: Agents iteratively adjust gate fees to maximize profits.
  • Stochasticity: Randomized initial STU attributes; trial-and-error pricing.
• Initialization: Based on Singapore waste statistics and empirical cost/residue ranges.
• Input data & calibration: Waste volumes from Singapore’s NEA (2015), operation costs and residue ratios from literature (McCrea et al. 2009; De Bere 2000). Parameters for agents’ decision-making estimated and optimized dynamically.
• Decision modules:
  • Refuse collector demand model.
  • Gate fee optimization modules for STUs and GTU (regression + heuristic search).
  • Regulator policy intervention module.

Policy / strategy levers examined

• Number of licensed STUs (competition intensity).
• Refuse collector’s preference to STUs (promoting sustainable treatment).
• GTU gate fee upper bound.
• GTU discount factor when treating STU residues.
• Mixed policy optimization using a genetic algorithm.

Key findings

• In near-perfect competition, STUs become price takers with falling profits, though as a group they gain market share from GTU.
• Regulatory promotion of STUs can provoke GTU retaliation (increased gate fees, withdrawal from competition).
• Higher GTU gate fee bounds or discounts benefit the GTU disproportionately, raising system-wide costs.
• Optimal regulatory policy (via GA) balances citizen cost savings, sustainability goals, and private sector viability (e.g., licensing 2 STUs, moderate preferences, gate fee caps).
• Validated that ABM captures complex adaptive behaviors more realistically than static models.

Relevance for toolkit development

• Demonstrates how ABM can simulate competitive dynamics and regulatory interventions in CE markets.
• Estimation–optimization approach useful for modeling agent behavior under incomplete information.
• Framework adaptable for Finnish/EU CE applications (e.g., biogas, waste-to-energy, recycling competition).
• Shows how simulation-based optimization (GA + ABM) can support policy analysis and investment assessments.

Strengths & limitations

• Strengths: Clear agent definitions; robust validation via multiple runs; empirical calibration; integration with optimization.
• Limitations: Focused on Singapore food waste case; omits capacity/location dynamics; results sensitive to parameterization; does not integrate with IO or LCA.
• Transferability: Methodologically strong for CE market competition modeling in Finland/EU, with adaptation of local data.

Classification for review
Include in main review.

Cluster membership
Cluster 2: Waste System ABMs – Supply Chains, Competition, and Consumer/Retail Behavior.

[Post-processing note: competitive waste treatment market ABM; gate fees & regulation; connects with Gan & Cheng on logistics and Kandemir on behavioral drivers]

Kandemir et al. - 2020 - Modelling Approaches to Food Waste

Abstract The generation of food waste at both suppliers’ and consumers’ levels stems from a complex set of interacting behaviours. Computational and mathematical models provide various methods to simulate, diagnose and predict different aspects within the complex system of food waste generation and prevention. This chapter outlines four different modelling approaches that have been used previously to investigate food waste.

• Discrete Event Simulation: which has been used to examine how the shelf life of milk and many actions taken around shopping and use of milk within a household influence food waste.
• Machine Learning and Bayesian networks: which have been used to provide insight into the determinates of household food waste.
• Agent Based Simulation: which has been used to provide insight into how innovation can reduce retail food waste.
• Mass Balance estimation which has been used to model and estimate food waste from data related to human metabolism and calories consumed.

Bibliographic info
Kandemir, C., Reynolds, C., Verma, M., Grainger, M., Stewart, G., Righi, S., Piras, S., Setti, M., Vittuari, M., & Quested, T. (2020). Modelling approaches to food waste: Discrete event simulation; machine learning; Bayesian networks; agent-based modelling; and mass balance estimation. In C. Reynolds, T. Soma, C. Spring, & J. Lazell (Eds.), Routledge Handbook of Food Waste (pp. 335–350). Routledge. https://doi.org/10.4324/9780429462795-25

Scope & system studied

• Sector: Food systems, household and retail food waste.
• Geography: Mainly UK and EU (with REFRESH and FUSIONS project data).
• CE strategy: Food waste prevention and reduction through behavioral change, innovation adoption, and policy interventions.

Model type & structure (ODD 2020 headings)

• Purpose and patterns: Review and demonstrate multiple modeling approaches for understanding, quantifying, and preventing food waste.
• Entities, state variables, and scales: Varies by method: households, consumers, and retailers in ABM and DES; socio-demographic variables in ML/BN; aggregate supply–demand balances in mass balance models.
• Process overview and scheduling:
  • DES (“Milk Model”): Simulates daily purchasing, consumption, and disposal events of milk in households, incorporating stochastic variability.
  • ML/BN: Identifies statistical drivers of household food waste using survey data.
  • ABM: Simulates innovation adoption in retail (low-waste technologies) and consumer food waste generation (integrated with Bayesian Networks for behavioral modeling).
  • Mass balance: Estimates national-level food waste as difference between food supply and metabolic consumption.
• Design concepts:
  • Emergence: Household waste levels and consumer/retailer adoption rates.
  • Adaptation: Retailers decide on innovation adoption based on utility (profit, reputation, environmental concern); consumers adjust behavior and opinions through social interactions.
  • Stochasticity: Event randomness in DES; Monte Carlo simulation in ABM; probabilistic BN estimates.
• Initialization: Household and retailer parameters from surveys and REFRESH pilot data; supply–demand from FAO food balance sheets and Hall et al. (2009).
• Input data & calibration:
  • DES: UK household milk consumption/waste data (WRAP, survey and qualitative research).
  • ML/BN: Eurobarometer (2013), WRAP household datasets (n ≈ 1,770 households).
  • ABM-BN: REFRESH consumer questionnaires (Netherlands, Germany, Hungary, Spain).
  • Mass balance: FAO food balance sheets, body weight/metabolism data.
• Decision modules:
  • Household DES module (purchasing, storage, disposal).
  • Retail ABM module (innovation adoption, pricing, market competition).
  • Consumer ABM-BN module (motivation, ability, opportunity factors; social interaction).
  • Mass balance module (availability vs. metabolic consumption).

Policy / strategy levers examined

• Shelf life extensions, smaller packaging, purchase adjustments, and consumer awareness campaigns (DES).
• Retail innovation adoption incentives (ABM).
• Social norm interventions, consumer training, and infrastructure provision (ABM-BN).
• Cross-country policy implications for food waste drivers (ML/BN).
• SDG 12 monitoring via mass balance estimates.

Key findings

• DES shows small household-level interventions (checking stocks, shelf-life extensions) reduce but do not eliminate waste, and create trade-offs (packaging, shopping frequency).
• ML/BN reveal household size, home ownership, composition, employment, and “fussy eaters” as key drivers; gender/urban–rural differences less important.
• ABM of retailers: waste-reducing innovations can diffuse if reputational/environmental concerns are included, despite profit disincentives.
• ABM-BN of consumers: model equilibrium without interventions; awareness and social norms interventions shift waste levels, but effects may be dampened by biases.
• Mass balance approaches provide higher estimates of food waste than FAO (e.g., U.S. waste increasing over decades), highlighting macro-level policy relevance.

Relevance for toolkit development

• Demonstrates a multi-method toolbox (DES, ML/BN, ABM, mass balance) for CE problems, highly relevant for DYNAMO.
• DES useful for micro-level consumption modeling (similar to profitability calculators for households).
• ABM-BN integration is methodologically innovative for capturing social norms and behavior change in CE contexts.
• ML/BN provide robust statistical insights into waste drivers that could calibrate ABMs.
• Mass balance complements IO/EEIO approaches for national CE assessments.

Strengths & limitations

• Strengths: Demonstrates range of methods; strong empirical base (REFRESH, WRAP, Eurobarometer); integration of ABM with BN innovative.
• Limitations: Case studies limited to food waste; ABM results sensitive to survey bias; mass balance highly aggregated; not yet integrated with economic IO models.
• Transferability: Methods broadly transferable to EU/Finland CE contexts (e.g., food waste, consumer behavior, innovation adoption).

Classification for review
Include in main review.

Cluster membership
Cluster 2: Waste System ABMs – Supply Chains, Competition, and Consumer/Retail Behavior.

[Post-processing note: multi-method food waste models; especially consumer/retail ABM-BN integration; completes waste ABM cluster with Gan & Cheng and He et al.]

Kerdlap et al. - 2020 - Environmental evaluation of distributed versus centralized plastic waste recycling

Abstract Plastic waste is internationally recognized as a problem, fueled by increased public awareness of environmental concerns and the steady increase in waste import bans. Modern sorting and recycling technologies are mature, but face implementation limitations in highly dense urbanized regions such as Singapore due to significant space constraints and expensive labor. Distributed plastic sorting and recycling facilities at small-scale closer to points of plastic waste generation offer the possibility of increasing the recovery of plastic waste streams in urbanized settings. To quantify the environmental performance of such systems, this study compares the life cycle greenhouse gas emissions of large-scale centralized facilities versus distributed small-scale facilities for sorting and recycling plastic bottles and takeaway containers generated in the central region of Singapore. An agent-based model is used to simulate different scenarios of plastic waste generation, collection routes, sorting, and recycling. The simulation results are used in a multi-level life cycle assessment to quantify the greenhouse gas emissions of the plastic sorting and recycling network as well as its individual entities. The results reveal that the life cycle greenhouse gas emissions of small-scale distributed plastic recycling compared to large-scale centralized systems are sensitive to the transport distance traveled and the type of trucks used. © 2020 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/). Peer-review under responsibility of the scientific committee of the 27th CIRP Life Cycle Engineering (LCE) Conference.

Bibliographic info
Kerdlap, P., Purnama, A. R., Low, J. S. C., Tan, D. Z. L., Barlow, C. Y., & Ramakrishna, S. (2020). Environmental evaluation of distributed versus centralized plastic waste recycling: Integrating life cycle assessment and agent-based modeling. Procedia CIRP, 90, 689–694. https://doi.org/10.1016/j.procir.2020.01.083

Scope & system studied

• Sector: Plastic waste recycling (PET bottles and PP takeaway containers).
• Geography: Singapore, Clementi–Bukit Merah region (urban case study).
• CE strategy: Comparison of centralized vs semi-distributed vs fully distributed recycling and sorting systems.

Model type & structure (ODD 2020 headings)

• Purpose and patterns: Assess environmental trade-offs of distributed versus centralized recycling systems, integrating ABM for logistics and LCA for environmental impacts.
• Entities, state variables, and scales:
  • Agents: collection points (bins), sorting facilities, recycling facilities, incinerators, trucks.
  • State variables: waste generated, collection schedule, facility capacities, transport flows, material recovery efficiency.
  • Scale: Regional (Clementi-BM, ~800,000 residents); annual simulation (1 year).
• Process overview and scheduling:
  • ABM simulates daily waste generation, collection, sorting, recycling, and transport.
  • Outputs (material flows, distances, energy use) feed into industrial symbiosis life cycle assessment (IS-LCA).
• Design concepts:
  • Emergence: Network-level GHG impacts from distributed vs centralized systems.
  • Adaptation: Facility siting optimized to minimize transport distances.
  • Stochasticity: Limited; assumes constant daily waste generation and composition.
• Initialization: Facility locations and waste generation based on population distribution and collection schedules.
• Input data & calibration:
  • Waste volumes: Singapore NEA (2018), population stats (2019).
  • Facility performance: Sorting efficiency (Perugini et al. 2005; Rigamonti et al. 2014).
  • Logistics: Veolia collection data, truck specifications.
  • LCA: Ecoinvent v3.6 database, electricity mix (Singapore vs Malaysia).
• Decision modules:
  • Collection scheduling (blue bin frequency).
  • Facility siting optimization (minimize distance, equalize capacity).
  • Material recovery and recycling module (sorting efficiency, PET/PP pellet yields).
  • Transport and logistics module (truck routes, capacities).

Policy / strategy levers examined

• Facility scale and siting: centralized vs semi-distributed vs distributed.
• Technology assumptions: sorting efficiency, recycling yield.
• Energy system differences (Singapore vs Malaysia electricity mix).
• Transport fleet choices (truck sizes, emission factors).

Key findings

• Semi-distributed scenario increased GHGs by 12%, distributed scenario decreased GHGs by 1% relative to centralized baseline.
• Transport emissions are highly sensitive to truck size and efficiency: smaller trucks raised GHGs despite shorter distances.
• Recycling processes contributed the majority of emissions (55–68%), especially extrusion.
• Distributed systems reduce transport distances but benefits can be offset if vehicle fleet is inefficient.
• Results depend strongly on assumptions about local energy mix and logistics.

Relevance for toolkit development

• Demonstrates integration of ABM (logistics/siting) with LCA (environmental assessment), highly relevant for CE toolkit methods.
• Provides insights for Finnish/EU applications where trade-offs exist between centralized vs decentralized CE infrastructures.
• Transferable to other materials (not just plastics), especially in urban CE planning.
• Shows how ABM outputs can supply inventory data to LCA models, suggesting pathways to link CE profitability simulators with environmental assessments.

Strengths & limitations

• Strengths: Integration of ABM and LCA; empirical calibration with Singapore data; clear scenario design (3 alternatives).
• Limitations: Narrow material scope (PET, PP only); simplified assumptions (constant waste generation, limited behavioral heterogeneity); excludes costs and social dimensions; results sensitive to truck fleet assumptions.
• Transferability: High conceptual relevance for Finland/EU; requires adaptation to local waste composition, transport logistics, and energy system.

Classification for review
Include in main review.

Cluster membership
Cluster 2: Waste System ABMs – Supply Chains, Competition, and Consumer/Retail Behavior.

[Post-processing note: ABM-LCA hybrid for plastic recycling; complements Gan & Cheng, He et al., and Kandemir in waste systems cluster; useful case of ABM as input generator for LCA.]

Knoeri et al. - 2014 - Enhancing recycling of construction materials: An agent based model…

Abstract Recycling of construction material is a valuable option for minimizing construction & demolition waste streams to landfills and mitigating primary mineral resource depletion. Material flows in the construction sector are governed by a complex socio-technical system in which awarding authorities decide in interaction with other actors on the use of construction materials. Currently, construction & demolition waste is still mainly deposited in landfills, as construction actors lack the necessary information and training regarding the use of recycled materials, and as a result have low levels of acceptance for them. This paper presents an agent-based model of the Swiss recycled construction material market based on empirical data derived from the agent operationalization approach. It elaborates on how recycling of construction materials can be enhanced by analysing key factors affecting the demand for recycled construction materials and developing scenarios towards a sustainable construction waste management. Doing so it demonstrates how detailed empirical agent decision data were incrementally included in the ABM model. Raising construction actors’ awareness of recycled materials as a decision option, in combination with small price incentives was most effective for enhancing the use of recycled materials. This could lead to a 50% reduction of construction & demolition waste streams to landfills, and significantly reduce the environmental impacts related to concrete applications. From a methodological perspective, although the agent operationalization approach provides a large empirical foundation, incremental model development turned out to be particularly important for the traceability of results and a realistic system representation.

Bibliographic info
Knoeri, C., Nikolic, I., Althaus, H.-J., & Binder, C. R. (2014). Enhancing recycling of construction materials: An agent based model with empirically based decision parameters. Journal of Artificial Societies and Social Simulation, 17(3), 10. https://doi.org/10.18564/jasss.2528

Scope & system studied

• Sector: Construction and demolition (C&D) materials.
• Geography: Switzerland (empirical data for calibration).
• CE strategy: Increasing recycling and reuse of construction materials through policy and behavioral interventions.

Model type & structure (ODD 2020 headings)

• Purpose and patterns: Explore how construction-sector actors’ decisions influence recycling rates, and test interventions to increase reuse and recycling.
• Entities, state variables, and scales:
  • Agents: construction companies, waste treatment facilities, consumers of recycled material.
  • State variables: awareness of recycling options, price sensitivity, social influence, disposal and recycling costs.
  • Scale: National/regional construction sector; time horizon of decades (scenarios).
• Process overview and scheduling:
  • Projects generate C&D waste.
  • Agents choose disposal, recycling, or reuse pathways based on cost, awareness, and social influence.
  • Market dynamics (supply and demand of recycled materials) update over time.
• Design concepts:
  • Emergence: Sector-level recycling rates from individual firm choices.
  • Adaptation: Firms update behaviors with awareness and social learning.
  • Interaction: Social influence and market transactions between agents.
  • Stochasticity: Randomness in adoption and awareness diffusion.
• Initialization: Based on empirical survey data of Swiss construction actors’ decision parameters.
• Input data & calibration:
  • Surveys on decision criteria (awareness, price sensitivity, environmental concern).
  • Empirical data on C&D waste flows and treatment costs in Switzerland.
• Decision modules:
  • Disposal choice module (landfill, recycling, reuse).
  • Awareness and social influence module (diffusion of recycling knowledge).
  • Market demand module for recycled materials.

Policy / strategy levers examined

• Increasing awareness of recycling options.
• Price interventions (changing costs of recycling vs. disposal).
• Scenarios combining awareness campaigns, regulatory measures, and market incentives.

Key findings

• Recycling fractions are highly sensitive to both awareness and price.
• Awareness campaigns significantly increase recycling, especially when combined with cost incentives.
• Pure price measures are less effective if awareness remains low.
• Social influence accelerates adoption of recycling behaviors.
• Sustainable C&D waste management requires coordinated strategies addressing both knowledge and economic drivers.

Relevance for toolkit development

• Provides an empirically grounded ABM of construction waste recycling, directly relevant to Finland/EU C&D contexts.
• Illustrates how survey-based decision parameters can calibrate ABM, useful for linking Syke’s empirical data with simulation models.
• Methodologically relevant for CE market simulators capturing awareness, adoption, and price interactions.
• Complements profitability calculators by modeling how price and awareness shifts propagate through markets.

Strengths & limitations

• Strengths: Strong empirical grounding; transparent decision parameterization; explicit inclusion of awareness and social influence.
• Limitations: Focused on Switzerland; limited diversity of waste streams; no integration with LCA or IO.
• Transferability: High for EU/Finland C&D waste; requires localized survey data for calibration.

Classification for review
Include in main review.

Cluster membership
Cluster 2: Waste System ABMs – Supply Chains, Competition, Consumer/Retail Behavior, and Recycling Configurations.

[Post-processing note: ABM with empirically grounded decision parameters; adds C&D recycling adoption dimension to waste ABM cluster.]

Notes Knoeri et al. (2014) and Gan & Cheng (2015) are both construction-related but one focuses on logistics optimization and the other on behavioral adoption.

Koide et al. - 2023 - Agent-based model for assessment of multiple circular economy strategies

Abstract Despite the increasing need for tools to support a circular economy, methods for assessing circularity and environmental consequences in a way that considers dynamic changes in heterogeneous consumer behavior are lacking. Here we propose an agent-based model for assessing seven explicitly-defined product-level circular economy strategies. The model endogenizes bounded rational consumer behavior and interactions related to product acquisition, obsolescence, repair, hibernation, and discharge. Simulation experiments with numerical examples were used to quantify the diffusion of product service systems (e.g., leasing of refurbished products and sharing/rental services), product circularity, and lifecycle greenhouse gas emissions over 30 years. As a result, synergies arising from promotion measures (e.g., pricing, advertisements, and service levels), bottlenecks (e.g., lack of collected products), and rebound effects (e.g., due to new production and transport) were identified. By employing “what-if” analyses, this model provides a means to identify effective interventions in the transition to a net-zero circular economy.

Bibliographic info
Koide, R., Yamamoto, H., Nansai, K., & Murakami, S. (2023). Agent-based model for assessment of multiple circular economy strategies: Quantifying product-service diffusion, circularity, and sustainability. Resources, Conservation & Recycling, 199, 107216. https://doi.org/10.1016/j.resconrec.2023.107216

Scope & system studied

• Sector: Consumer durables (home appliances, small electronic equipment).
• Geography: Generalizable; demonstrated with hypothetical examples, calibrated from literature.
• CE strategy: Seven product-level CE strategies (repair, reuse, refurbishing, functional upgrading, leasing, renting, and consumer-to-consumer sharing).

Model type & structure (ODD 2020 headings)

• Purpose and patterns: Quantify diffusion of multiple CE strategies, product circularity, and environmental consequences under heterogeneous consumer behavior.
• Entities, state variables, and scales:
  • Agents: Households, products, supply chain actors (manufacturers, reuse shops, service providers, recyclers).
  • Variables: Awareness sets, social influence, consideration costs, utility functions, product age, ownership/circularity status, obsolescence (absolute/relative).
  • Scale: 5000 households, simulated over 30 years, monthly time steps.
• Process overview and scheduling:
  • Consumer decision-making (repair, acquisition, discharge) modeled via random utility and consideration sets.
  • Product circulation (use, hibernation, obsolescence, preparation, disposal).
  • Inventory updates track life cycle processes for consequential LCA.
• Design concepts:
  • Emergence: Diffusion of services, product stocks, emissions.
  • Adaptation: Households update awareness and decisions via word-of-mouth and advertising.
  • Interaction: Consumer–consumer (social networks) and consumer–supply chain interactions.
  • Stochasticity: Random utility, Weibull lifetimes, network topology.
• Initialization: Household and product states initialized with distributions from literature (e.g., Weibull lifetimes, utility parameters).
• Input data & calibration:
  • Consumer decision heuristics and obsolescence distributions from behavioral/marketing literature.
  • LCA data from Ecoinvent and IDEA databases.
  • Validation through expert review and stylized facts (diffusion curves, rebound effects).
• Decision modules: Acquisition, repair, discharge, obsolescence, product preparation/refurbishment.

Policy / strategy levers examined

• Promotion measures: pricing discounts, advertisement frequency, service level improvements.
• Comparison of strategies: reuse vs leasing refurbished products vs rental vs sharing.
• Combination of strategies (e.g., leasing refurbished appliances with functional upgrades).

Key findings

• Advertising accelerates early diffusion; service-level improvements promote later adoption; pricing has strong but limited effects.
• Combining promotion measures yields synergies but also bottlenecks (e.g., shortage of collected products for refurbishing).
• Rental and sharing reduce in-use stocks by ~25% but may increase emissions due to transport or preference for brand-new products.
• Reuse is fast but limited in long-term diffusion (<15%), while sharing can achieve >50% diffusion in the long run.
• Emissions reductions up to ~10% relative to BaU over 30 years; rebound/backfire possible (e.g., rental scenario increases emissions).
• Demonstrates how consumer preferences can undermine CE gains if not aligned with circular strategies.

Relevance for toolkit development

• Provides a state-of-the-art ABM framework for CE strategies at the product level, linking micro-level consumer behavior with macro-level circularity and emissions.
• Integrates ABM with consequential LCA, offering a template for coupling behavioral simulations with environmental impact models.
• Highly relevant for testing promotion measures, adoption incentives, and consumer-targeted CE interventions in Finnish/EU contexts.
• Suggests pathways for extending profitability calculators to include consumer adoption dynamics and rebound effects.

Strengths & limitations

• Strengths: Rich behavioral modeling (bounded rationality, awareness sets, social influence); integration with consequential LCA; broad coverage of CE strategies.
• Limitations: Demonstrative only (no empirical calibration for specific products or geographies); assumes stable household needs; abstract spatial representation.
• Transferability: High conceptual relevance; empirical calibration with Finnish/EU consumer survey and product data needed.

Classification for review
Include in main review.

Cluster membership
Cluster 7 (Placeholder): Policy Levers & Transformation Pathways — Consumer CE strategies and promotion measures.

[Post-processing note: Advanced ABM of consumer durables; seven CE strategies; integrates consumer heterogeneity, social networks, and consequential LCA. Forms anchor for a “consumer behavior & product-service systems” cluster.]

Notes Koide 2023’s consumer adoption dynamics could also connect to Kandemir 2020’s ABM-BN of consumer food waste behavior

Lieder et al. - 2017 - Towards Circular Economy implementation an agent-based simulation approach for business model change

Abstract This paper introduces an agent-based approach to study customer behavior in terms of their acceptance of new business models in Circular Economy (CE) context. In a CE customers are perceived as integral part of the business and therefore customer acceptance of new business models becomes crucial as it determines the successful implementation of CE. However, tools or methods are missing to capture customer behavior to assess how customers will react if an organization introduces a new business model such as leasing or functional sales. The purpose of this research is to bring forward a quantitative analysis tool for identifying proper marketing and pricing strategies to obtain best fit demand behavior for the chosen new business model. This tool will support decision makers in determining the impact of introducing new (circular) business models. The model has been developed using an agent-based modeling approach which delivers results based on socio-demographic factors of a population and customers’ relative preferences of product attributes price, environmental friendliness and service-orientation. The implementation of the model has been tested using the practical business example of a washing machine. This research presents the first agent-based tool that can assess customer behavior and determine whether introduction of new business models will be accepted or not and how customer acceptance can be influenced to accelerate CE implementation. The tool integrates socio-demographic factors, product utility functions, social network structures and inter-agent communication in order to comprehensively describe behavior on individual customer level. In addition to the tool itself the results of this research indicates the need for systematic marketing strategies which emphasize CE value propositions in order to accelerate customer acceptance and shorten the transition time from linear to circular. Agent-based models are emphasized as highly capable to fill the gap between diffusion-based penetration of information and resulting behavior in the form of purchase decisions.

Bibliographic info
Lieder, M., Asif, F. M. A., & Rashid, A. (2017). Towards Circular Economy implementation: An agent-based simulation approach for business model changes. Autonomous Agents and Multi-Agent Systems, 31(6), 1377–1402. https://doi.org/10.1007/s10458-017-9365-9

Scope & system studied

• Sector: Household appliances (washing machines).
• Geography: Stockholm, Sweden (empirical context and calibration).
• CE strategy: Transition from linear product sales to service-oriented business models (buy-back, pay-per-use).

Model type & structure (ODD 2020 headings)

• Purpose and patterns: Explore customer acceptance of CE business models (leasing, pay-per-use, buy-back) and evaluate how marketing, pricing, and product attributes affect adoption.
• Entities, state variables, and scales:
  • Agents: Individual customer agents; population-level “environmental” agent; product offers.
  • Variables: Socio-demographic traits (age, income, education, geography), product attributes (price, environmental friendliness, service-orientation), awareness, expected/actual utility, satisfaction.
  • Scale: 10,000 agents, urban context of Stockholm, simulated over 20 years.
• Process overview and scheduling: Customers cycle through stages of need recognition, evaluation, purchase, use, and post-purchase learning; adoption influenced by advertisements and social network communication. Business models (linear sales, buy-back, pay-per-use, competitor) introduced sequentially.
• Design concepts:
  • Emergence: Market share and diffusion of CE offers from individual purchase decisions.
  • Adaptation: Agents update expected utility via post-purchase learning and social influence.
  • Interaction: Word-of-mouth and advertising spread awareness.
  • Stochasticity: Socio-demographic assignment, initial awareness, network connections.
• Initialization: Based on Statistics Sweden data (income, age, education, population density) and company input (Gorenje washing machines).
• Input data & calibration:
  • Empirical distributions from Statistics Sweden.
  • Utility function parameters and attribute weights from company assumptions.
  • Validation via expert judgment and diffusion curve plausibility.
• Decision modules:
  • Utility-based purchase decision module.
  • Communication module (word-of-mouth, advertisement).
  • Post-purchase learning module (update expected vs. actual utility). Policy / strategy levers examined
• Buy-back vs pay-per-use introduction.
• Advertisement intensity and reach.
• Pricing levels (monthly fees, buy-back price).
• Social network structure (density, adaptability).
• Sensitivity analysis of follower tendency and susceptibility to advertisement.

Key findings

• Buy-back offers diffuse more rapidly and sustainably than pay-per-use under tested assumptions.
• Pay-per-use adoption is limited unless strongly supported by marketing emphasizing service-orientation and environmental benefits.
• Market shares highly sensitive to price levels and network connectivity.
• Social influence accelerates adoption but can also suppress competing offers.
• Coordinated marketing emphasizing CE value propositions can accelerate acceptance and shorten transition times.

Relevance for toolkit development

• Demonstrates ABM as a decision-support tool for CE business model innovation.
• Provides a template for simulating customer adoption of service-oriented CE models, directly relevant to Finnish/EU consumer durables.
• Methodology transferable to other sectors (e.g., leasing/refurbishing in electronics, pay-per-use in mobility).
• Results inform profitability calculators by modeling adoption thresholds and switching behavior under different price/marketing regimes.

Strengths & limitations

• Strengths: Rich behavioral modeling grounded in socio-demographics; integrates utility, communication, and learning; detailed sensitivity analysis.
• Limitations: Calibration based largely on assumptions and expert input; lacks empirical validation of CE adoption; single product/sector focus; no integration with LCA/IO.
• Transferability: High to EU contexts if calibrated with survey and product data; applicable for CE business model experimentation.

Classification for review
Include in main review.

Cluster membership
Cluster 7: Consumer Behavior & Product-Service Strategies.

[Post-processing note: ABM of CE business model adoption; pay-per-use vs buy-back; complements Koide et al. 2023 as part of consumer CE strategies cluster.]

Luo et al. - 2019 - Towards the sustainable development of waste household appliance recovery systems in China

Abstract With the rapid development of the economy and society, the number of waste household appliances in China has increased. Although the scale of China’s waste household appliance industry continues to expand, the recovery rate is relatively low compared with developed countries. As waste household appliances are important recycling resources that have potential environmental risks and a high resource value, the effective recovery, recycling and utilization of these appliances are necessary to build a circular economy system, ease environmental crises and realize sustainable development in China. An agent-based model of a waste household appliance recovery system was established in this study by using complex adaptive system theory and agent-based modeling method. The simulation of agents’ behavior and scenario analysis were performed using this model, and the results were analyzed to study the evolutionary trends of the industry under different policy scenarios. Based on the exploration and interpretation of the optimal scenario, policy suggestions are proposed for different parts of the system.

Bibliographic info
Luo, M., Song, X., Hu, S., & Chen, D. (2019). Towards the sustainable development of waste household appliance recovery systems in China: An agent-based modeling approach. Journal of Cleaner Production, 220, 431–444. https://doi.org/10.1016/j.jclepro.2019.02.128

Scope & system studied

• Sector: Waste household appliances (TVs, washing machines, refrigerators, ACs, computers).
• Geography: China (national scale).
• CE strategy: Improving recovery and recycling systems for e-waste appliances; formalizing informal sector; testing policy interventions.

Model type & structure (ODD 2020 headings)

• Purpose and patterns: Evaluate recovery system evolution under different policy scenarios; identify optimal policy combinations for sustainability.
• Entities, state variables, and scales:
  • Agents: Residents, recyclers (individual, formal stations, producers, online), government, environment.
  • Variables: Recycling intention, service level, pollution, construction speed, costs, subsidies.
  • Scale: National industry-level dynamics (2000–2050).
• Process overview and scheduling:
  • Residents generate waste appliances, update perceptions (pollution, service), and decide whether/where to recycle based on Theory of Planned Behavior (TPB).
  • Recyclers compete for market share, expand capacity, and respond to policy and profits.
  • Government enacts subsidies, VAT adjustments, bans, and education campaigns.
  • Environment contributes pollution feedback.
• Design concepts:
  • Emergence: Recovery rates and market shares from resident/recycler interactions.
  • Adaptation: Residents update recycling intention; recyclers expand capacity.
  • Interaction: Bargaining, competition, information exchange.
  • Stochasticity: Randomized initial conditions (perceptions, service levels).
• Initialization: 2000 baseline with historical appliance ownership and recovery modes.
• Input data & calibration: China Statistical Yearbook (1981–2015); China National Resources Recycling Association; surveys on resident attitudes; literature on recycling costs, VAT, subsidies. Model validated against historical recovery trends.
• Decision modules:
  • Resident recycling decision (TPB-based).
  • Recycler construction/competition module.
  • Government policy intervention module.
  • Environmental pollution feedback module.

Policy / strategy levers examined

• Education and information dissemination.
• Reward-and-penalty regulations (fines, incentives).
• Subsidies for recyclers.
• Banning illegal dismantling enterprises.
• VAT reductions.
• Subsidies for eco-design.
• Multi-policy combinations (e.g., education + subsidies + VAT reduction + eco-design).

Key findings

• By 2050, baseline recovery rate projected at 74.3%.
• Education/awareness campaigns raise recovery rate to ~79.8%.
• Reward-and-penalty rules raise recovery to ~77% but with costs.
• Subsidies boost rates (~79.5%) but sharply increase government costs.
• Banning illegal dismantlers lowers overall recovery (73.5%) despite pollution reduction.
• Multi-policy scenario 2 (education + regulation + subsidies + VAT cuts) achieves ~85% recovery by 2050, with balanced costs and pollution reduction, outperforming other scenarios.
• Online recovery modes projected to grow rapidly, replacing traditional informal channels.

Relevance for toolkit development

• Strong example of ABM applied to e-waste with multi-agent structure (residents, recyclers, government, environment).
• TPB-based behavioral modeling transferable to Finnish/EU contexts (e.g., WEEE, electronics take-back).
• Demonstrates how policy combinations interact dynamically, useful for profitability calculators and policy simulators.
• Model integration of government costs, pollution, and market share provides multi-criteria assessment relevant for toolkit.

Strengths & limitations

• Strengths: Empirically grounded; validated against historical data; explicit policy scenario design; integration of behavioral, economic, and environmental factors.
• Limitations: Context-specific to China; excludes international trade flows; does not integrate IO/LCA; simplified representation of recyclers.
• Transferability: High methodological relevance, requires EU-specific calibration (WEEE directive, EPR schemes).

Classification for review
Include in main review.

Cluster membership
Cluster 2: Waste System ABMs – Supply Chains, Competition, Consumer/Retail Behavior, and Recycling Configurations.

[Post-processing note: Comprehensive ABM of household appliance recovery; TPB-based resident behavior; detailed policy scenarios; complements Knoeri, Kerdlap, He et al. in waste recycling ABM cluster.]

Ma et al. - 2023 - Community-Level Household Waste Disposal Behavior Simulation and Visualization

Abstract The classification and recycling of household waste becomes a major issue in today’s urban environmental protection and domestic waste disposal. Although various policies promoting household waste classification have been introduced, the recovery rate failed to reach the expected result. Existing studies on incentive policies for household waste recycling tried to integrate subjective and objective factors in human behavior decisions. To explore how effective interventions can promote household waste classification in communities, this article developed an Agent-Based Model (ABM) based on Theory of Planned Behavior (TPB) to simulate the participation of households under eight different policy scenarios. The result shows that: monetary incentive is most effective in inducing participation, while social norms have different impacts on household decision under different policy intervention. Under policy stimulus, the participation rate of garbage sorting increased from 18% to 76%. This model has been applied into an online community-based participatory virtual simulation 3D system, which aims to help university students better understand how policies affect household recycling behaviors, which end up affecting the environment.

Bibliographic info
Ma, H., Li, M., Tong, X., & Dong, P. (2023). Community-level household waste disposal behavior simulation and visualization under multiple incentive policies—An agent-based modelling approach. Sustainability, 15(13), 10427. https://doi.org/10.3390/su151310427

Scope & system studied

• Sector: Municipal solid waste (household garbage sorting).
• Geography: Community H, Beijing, China.
• CE strategy: Household pre-classification of waste (recyclable, kitchen, hazardous, other) with incentives and penalties to increase participation in sorting and recycling.

Model type & structure (ODD 2020 headings)

• Purpose and patterns: Assess effectiveness of different incentive policies in increasing household waste sorting participation rates at the community level.
• Entities, state variables, and scales:
  • Agents: Households (~10,000 families), garbage collectors (traditional, upgraded), fixed sorting facilities, community managers.
  • Variables: Family size, education, income, environmental awareness, distance to facilities, social norms.
  • Scale: Single residential community; simulation horizon ~90 days to steady state.
• Process overview and scheduling:
  • Households generate daily waste.
  • Once storage threshold reached, they decide disposal method (door-to-door collection vs fixed facilities), influenced by TPB factors: attitudes, subjective norms, perceived control.
  • Social activity every Monday updates norms based on neighbors’ behavior.
  • Incentive policies adjust payoff structure (rewards, penalties).
• Design concepts:
  • Emergence: Community-level sorting rates from heterogeneous household choices.
  • Adaptation: Social learning and changing norms influence future participation.
  • Stochasticity: Household attribute distributions, randomness in decision-making.
• Initialization: Survey of 500 households in Community H (2013–2016) used to parameterize demographics, awareness, income.
• Input data & calibration: Empirical survey data, community statistics (~10,000 households, 117 buildings, 6 facilities).
• Decision modules: Household recycling decision (TPB-based); social influence/norm updating; policy incentive adjustments.

Policy / strategy levers examined
Eight scenarios in three categories:

1. Door-to-door collection only (traditional collectors vs upgraded professional recyclers).
2. Fixed facility incentives (none, green account credits, cash rewards by weight).
3. Integrated policies (combinations of upgraded collectors + cash rewards + punitive measures for non-sorting).

Key findings

• Baseline sorting participation: ~18%.
• Monetary incentives most effective: cash rewards increased facility use up to 57%.
• Upgrading door-to-door collectors raised participation from 22% to 34%.
• Integrated policy with strong incentives and penalties (Scenario 8) raised total participation to 76%.
• Higher education correlated with stronger participation; western district outperformed southern district.
• Social norms and peer effects amplify incentives, making participation rates higher than incentives alone would predict.

Relevance for toolkit development

• Strong example of community-level ABM with TPB foundation, transferable to Finnish/EU contexts (especially pilot studies on household sorting compliance).
• Integration of social norms and monetary/non-monetary incentives provides a blueprint for modeling household responses to EPR schemes or municipal pay-as-you-throw policies.
• Use of Mesa (Python) suggests ease of adaptation into DYNAMO toolkit.
• 3D visualization demonstrates potential for policy communication tools alongside simulations.

Strengths & limitations

• Strengths: Empirically calibrated household attributes; rigorous policy scenario design; integration of social norms, incentives, and penalties; visualization layer.
• Limitations: Context-specific to Beijing; parameters rely on local surveys (not generalizable without adaptation); short time horizon; excludes waste treatment logistics and downstream impacts.
• Transferability: High methodological relevance; Finnish/EU calibration would require new survey data.

Classification for review
Include in main review.

Cluster membership
Cluster 2: Waste System ABMs – Supply Chains, Competition, Consumer/Retail Behavior, and Recycling Configurations.

[Post-processing note: Community-level ABM with TPB and incentive policies; raises sorting from 18% to 76% under best scenario; expands waste cluster with household-scale behavioral interventions.]

Nguyen-Trong et al. - 2017 - Optimization of municipal solid waste transportation by integrating GIS analysis

Abstract The amount of municipal solid waste (MSW) has been increasing steadily over the last decade by reason of population rising and waste generation rate. In most of the urban areas, disposal sites are usually located outside of the urban areas due to the scarcity of land. There is no fixed route map for transportation. The current waste collection and transportation are already overloaded arising from the lack of facilities and insufficient resources. In this paper, a model for optimizing municipal solid waste collection will be proposed. Firstly, the optimized plan is developed in a static context, and then it is integrated into a dynamic context using multi-agent based modelling and simulation. A case study related to Hagiang City, Vietnam, is presented to show the efficiency of the proposed model. From the optimized results, it has been found that the cost of the MSW collection is reduced by 11.3%.

Bibliographic info
Nguyen-Trong, K., Nguyen-Thi-Ngoc, A., Nguyen-Ngoc, D., & Van Dinh-Thi-Hai. (2017). Optimization of municipal solid waste transportation by integrating GIS analysis, equation-based, and agent-based model. Waste Management, 59, 14–22. https://doi.org/10.1016/j.wasman.2016.10.048

Scope & system studied

• Sector: Municipal solid waste (MSW) collection and transportation.
• Geography: Hagiang City, Vietnam.
• CE strategy: Improving efficiency of waste collection and transport to reduce costs and emissions.

Model type & structure (ODD 2020 headings)

• Purpose and patterns: Optimize MSW vehicle routing by integrating GIS, mathematical optimization (MILP), and ABM to capture dynamic traffic and waste generation.
• Entities, state variables, and scales:
  • Agents: Road segments, garbage collection centers, garbage trucks, depot, transport agents (other vehicles).
  • Variables: Waste generation rates, truck capacity, road positions/lengths, traffic density, vehicle speeds.
  • Scale: City-level network with 33 collection centers; daily collection cycles (morning and afternoon).
• Process overview and scheduling:
  • Step 1: MILP and Clarke & Wright savings algorithm optimize static routes.
  • Step 2: ABM simulates dynamic traffic and waste accumulation, updating distances/times.
  • GIS provides road and facility geometry.
• Design concepts:
  • Emergence: Actual travel distances and times under real traffic conditions.
  • Adaptation: Trucks adjust collection sequence if full.
  • Interaction: Transport agents affect truck speed/delays.
  • Stochasticity: Traffic and waste generation variability.

• Initialization: GIS data on road network, depot, and collection centers; historical waste generation and traffic data.

• Input data & calibration:
  • Road geometry from GIS.
  • Waste generation rates from Ministry of Environment (Vietnam).
  • Vehicle characteristics from URENCO Hagiang.
  • Traffic data from national statistics.
• Decision modules:
  • Routing optimization (MILP + Clarke & Wright).
  • Garbage accumulation module.
  • Truck capacity and unloading module.
  • Traffic simulation module (transport agents).

Policy / strategy levers examined

• Route optimization vs real plans.
• Doubling garbage truck maximum speed.
• Doubling collection speed at centers.
• Dynamic vs static routing (GIS only vs GIS + ABM).

Key findings

• Optimized plans reduced travel distance by 11.3% compared to real plans.
• Total collection time reduced most when collection speed doubled (38% reduction); truck speed increases less effective (15%).
• Integration of GIS + MILP + ABM outperforms static GIS-only methods by capturing traffic dynamics.
• Collection time at centers is a major driver of system inefficiency, not just routing distance.

Relevance for toolkit development

• Demonstrates power of hybrid models (GIS + optimization + ABM) for CE logistics.
• Highly relevant for Finnish/EU waste transport optimization, especially where rural–urban logistics and traffic congestion matter.
• Framework adaptable for CE profitability calculators by quantifying transport cost savings and emissions.
• Suggests extension to fuel consumption and emissions modeling.

Strengths & limitations

• Strengths: Hybrid integration of methods; applied case study with empirical GIS/traffic data; robust scenario testing.
• Limitations: Focused only on transport stage (no downstream recycling impacts); small case (2 trucks, 33 centers); excludes behavioral decision-making of households.
• Transferability: High methodological relevance; requires EU-specific traffic/waste datasets.

Classification for review
Include in main review.

Cluster membership
Cluster 2: Waste System ABMs – Supply Chains, Competition, Consumer/Retail Behavior, and Recycling Configurations.

[Post-processing note: Hybrid GIS + MILP + ABM for waste transport optimization; Hagiang City case; adds transport logistics dimension to waste systems cluster.]

Ortiz Salazar et al. - 2021 - Overcoming Inconvenience How Society Can Incentivize Individual Recycling Behavior

Abstract

Bibliographic info
Ortiz Salazar, A., Klein, J. M., & Klycheva, Z. (2021). Overcoming inconvenience: How society can incentivize individual recycling behavior; An agent-based model. In D. N. Cassenti, A. L. Frey, M. J. Smith, & J. Chen (Eds.), Advances in Simulation and Digital Human Modeling (pp. 80–85). Springer, Cham. https://doi.org/10.1007/978-3-030-51064-0_11

Scope & system studied

• Sector: Household recycling behavior.

• Geography: United States (generalized context).

• CE strategy: Increasing community recycling rates by addressing social pressure, ideology, and convenience factors (bin placement).

Model type & structure (ODD 2020 headings)

• Purpose and patterns: Explore how availability, spatial distribution of recycling bins, ideology, and social influence affect recycling participation at the community level.

• Entities, state variables, and scales:

  • Agents: 50 households with attributes of wealth, ideology (0–1), contamination level, and vision.

  • Variables: Recycling proclivity, opportunity cost, bin availability.

  • Scale: Small community simulated over 3,000 ticks.

• Process overview and scheduling:

  • Agents generate recyclable waste each tick and decide disposal (recycling bin vs trash bin).

  • Decisions influenced by ideology, neighbor behavior (vision-based social circle), and bin availability.

  • Contamination increases when trash is chosen; decreases when recycling.

• Design concepts:

  • Emergence: Community-level recycling rates from household decisions.

  • Adaptation: Ideology updates after each decision (recycling vs trash).

  • Interaction: Social influence through local neighborhood vision.

  • Stochasticity: Randomized wealth and ideology distributions.

• Initialization: 50 agents with normally distributed wealth (mean = 10 units) and ideology values between 0–1. 4 recycling bins and 4 trash bins evenly distributed.

• Input data & calibration: Largely conceptual; parameters inspired by recycling psychology literature (Theory of Planned Behavior, Schwartz’s Value Theory). No empirical calibration.

• Decision modules:

  • Recycling decision module (ideology + bin availability + social influence).

  • Contamination tracking module.

  • Ideology updating module.

Policy / strategy levers examined

• Number of recycling bins (4 → 7).

• Spatial distribution of bins (even vs clustered).

• Hypothetical incentive policies (suggested in discussion: tax rebates, refund schemes).

Key findings

• Recycling participation increases with bin availability; ideology shifts positively with greater recycling exposure.

• Spatial clustering of bins reduces recycling and increases contamination, even if total number of bins remains constant.

• Wealth negatively correlated with recycling (due to higher opportunity cost of travel to bins).

• Contamination decreases steadily when recycling opportunities are improved.

• Highlights the role of opportunity cost and convenience alongside ideology and social pressure.

Relevance for toolkit development

• Useful conceptual demonstration of micro-level recycling decisions influenced by convenience and ideology.

• Suggests how infrastructure placement (bins, collection points) can be modeled in CE simulators.

• Lacks empirical calibration but provides a framework for coupling social-behavioral models with spatial infrastructure design.

• Could be extended to Finnish/EU contexts by integrating household survey data on recycling attitudes and infrastructure placement.

Strengths & limitations

• Strengths: Explicitly models social influence and ideology; highlights infrastructure distribution effects; interprets recycling as a product of convenience and norms.

• Limitations: Simplified small community (50 agents); no empirical data; short-term conceptual exploration; excludes cost structures of recycling systems.

• Transferability: Conceptually transferable to EU/Finland, but requires empirical calibration to local recycling behaviors and infrastructure.

Classification for review
Include in main review.

Cluster membership
Cluster 2: Waste System ABMs – Supply Chains, Competition, Consumer/Retail Behavior, and Recycling Configurations.

[Post-processing note: Conceptual ABM of household recycling; emphasizes ideology, social norms, and bin distribution; complements Ma et al. 2023 in community-scale behavioral waste models.]

Ravandi and Jovanovic - 2019 - Impact of plate size on food waste Agent-based simulation of food consumption

Abstract Food waste is a substantial contributor to environmental change and it poses a threat to global sustainability. A significant portion of this waste accounts for plate waste and food surplus from food-service operations such as restaurants, workplace canteens, cafeterias, etc. In this work, we seek to identify potential strategies to optimize food consumption in all-you-can-eat food-service operations, in terms of minimizing food waste while ensuring quality of service (i.e., maintaining low wait-time, unsatisfied-hunger, and walk-out percentages). We treat these facilities as complex systems and propose an agent-based model to capture the dynamics between plate waste, food surplus, and the facility organization setup. Moreover, we measure the impact of plate size on food waste. The simulation results show reducing plate size from large to small decreases plate waste up to 30% while ensuring quality of service. However, total waste as the sum of food surplus and plate waste is lower with large plates. Our results indicate the need for optimizing food preparation along with designing choice environments that encourage guests to avoid taking more food than they need.

Bibliographic info
Ravandi, B., & Jovanovic, N. (2019). Impact of plate size on food waste: Agent-based simulation of food consumption. Resources, Conservation & Recycling, 149, 550–565. https://doi.org/10.1016/j.resconrec.2019.05.033

Scope & system studied

• Sector: Food service (all-you-can-eat facilities such as restaurants, canteens, cafeterias).
• Geography: Generalized (case study–agnostic, calibrated from literature).
• CE strategy: Food waste prevention via “choice architecture” interventions (plate size, food station design, food variety).

Model type & structure (ODD 2020 headings)

• Purpose and patterns: Explore how plate size and food-service setup influence plate waste, food surplus, and quality of service (wait times, walk-outs).
• Entities, state variables, and scales:
  • Agents: Food stations (attributes: plate size, food type, serving wait time) and guest agents (preferences, hunger, appetite, plates taken).
  • Variables: Appetite level, hunger intensity, queue length, plate contents.
  • Scale: Simulations of one dining operation; ~hundreds of guests per scenario.
• Process overview and scheduling:
  • Guests arrive, select stations, queue, take food, eat or waste items, possibly go for more plates.
  • Stations track queues and food quotas.
  • Simulation stops when all guests leave.
• Design concepts:
  • Emergence: Plate waste, surplus, and service quality metrics from individual eating choices.
  • Adaptation: Hunger and appetite decrease with eating or waiting.
  • Interaction: Queues affect guest satisfaction and choices.
  • Stochasticity: Randomized food preferences, hunger/appetite initial values, plate-taking decisions.
• Initialization: Guest preferences generated randomly (uniform distribution normalized); hunger and appetite initialized to ranges from psychology literature.
• Input data & calibration: Calibrated using literature on food preferences, hunger/appetite psychology (e.g., Piech et al. 2009; Lowe & Butryn 2007), and WRAP waste studies; validated against stylized facts (field experiment results).
• Decision modules:
  • Station selection (based on preference and queue length).
  • Plate-filling and consumption/waste decisions.
  • Hunger/appetite updating and walk-out decision.

Policy / strategy levers examined

• Plate size (large, medium, small).
• Number of food stations and food types.
• Traffic intensity (moderate vs high).

Key findings

• Reducing plate size cuts plate waste: −14% (large→medium) and −30% (large→small).
• However, total waste (plate waste + surplus) is lowest with large plates, since small plates increase food surplus (over-preparation).
• Increasing number of food stations improves service quality and reduces waste.
• High traffic increases total waste but plate-waste trends remain consistent.
• Confirms field studies showing plate size interventions reduce plate waste but raises trade-offs with surplus.

Relevance for toolkit development

• Demonstrates ABM’s ability to capture trade-offs between behavioral nudges and systemic waste outcomes.
• Provides methodological basis for testing “choice architecture” policies (e.g., packaging sizes, food station design) relevant in EU food service contexts.
• Shows need to model surplus alongside plate waste for holistic CE assessments.
• Framework could be adapted to model consumer-facing CE interventions in Finland/EU (canteens, buffets, schools).

Strengths & limitations

• Strengths: Rich behavioral modeling (hunger/appetite, preferences, queuing); integration of service quality metrics; calibration to psychology and waste literature.
• Limitations: No empirical case data; simplified hunger/appetite dynamics; excludes economic costs and environmental impacts; case-agnostic.
• Transferability: Conceptually strong for EU/Finland; empirical calibration required for local contexts.

Classification for review
Include in main review.

Cluster membership
Cluster 2: Waste System ABMs – Supply Chains, Competition, Consumer/Retail Behavior, and Recycling Configurations.

[Post-processing note: Food-service ABM linking plate size to waste; highlights trade-offs between plate waste and surplus; complements Kandemir 2020 and Ortiz Salazar 2021 in behavioral food waste modeling.]

Rizzati and Landoni - 2024 - A systematic review of agent-based modelling in the circular economy Insights towards a general model

Abstract Circular Economy (CE) is a popular topic for governments and businesses around the world; yet, only a few comprehensive and economy-wide frameworks exist, and the consequences of the CE on economic systems stay unclear. With this systematic review, we put under scrutiny the existing contributions to Circular Economy (CE) that apply the Agent-based modelling methodology. There is an open gap in the CE literature regarding the use of ABM. The research question that guides this systematic review concerns the potential benefit of ABM for CE and how to use this methodology in the context of CE. We put in evidence three thematic areas, two agents and one process, namely producers, i.e. firms and industrial systems, consumers, i.e. households and waste disposal, and the diffusion of innovation. We infer that the former three thematic strands of literature can be further synthetized together to form a general model of the Circular Economy. This development is crucial to properly evaluate how the agent’s heterogeneity affects the diffusion and the consequences of the adoption of CE practices on the economy. Research has widely applied ABM simulations to consider the impact of heterogeneity amongst individuals and their behavioural interactions on the evolution of complex systems, yet very little did it systematically about CE. Our results complement those of Computable General Equilibrium models. The review provides an interpretative framework, suggests valuable future research directions within the new comprehensive thematic area, and contributes to the theoretical and managerial discussion on agent-based modelling in the circular economy.

Bibliographic info
Rizzati, M., & Landoni, M. (2024). A systematic review of agent-based modelling in the circular economy: Insights towards a general model. Structural Change and Economic Dynamics, 69, 617–631. https://doi.org/10.1016/j.strueco.2024.03.013

Scope & system studied
This paper is a systematic review of applications of agent-based modelling (ABM) to the circular economy (CE). It surveys 100 selected works and organizes them into three main themes: producers (firms/industries), consumers (households and waste disposal), and diffusion of innovation. The review is global in scope, with many studies set in China, the US, and Europe, though many are not geographically specified.

Model type & structure
Systematic review of ABM applications.

• Purpose: Assess what ABM contributes to CE modelling and propose a framework for a general CE ABM.
• Entities/agents (from reviewed works): Firms/industries, consumers/households, waste collectors, governments, and innovation adopters.
• Processes: Material flows, waste collection/recycling, industrial symbiosis, behavioural change, innovation diffusion.
• Data: Many reviewed works use case studies (construction, agriculture, recycling industries), GIS data, surveys, or hypothetical networks.
• Submodels: Classified into partial equilibrium settings (sector- or product-specific) versus broader macroeconomic ABM/stock-flow consistent hybrids.

Policy / strategy levers examined
Across the reviewed works: eco-taxes, subsidies, waste collection policies, recycling regulations, innovation adoption incentives, industrial symbiosis facilitation, and renewable energy transition strategies.

Key findings

• CE ABMs are fragmented and often case-specific, making generalization difficult.
• Three thematic areas recur: producers, consumers/waste disposal, and innovation diffusion.
• ABMs capture heterogeneity, bounded rationality, and emergent dynamics often ignored in IO/CGE or SD approaches.
• Very few economy-wide CE ABMs exist; most studies are partial equilibrium.
• Integration with macro-ABMs and stock-flow consistent (SFC) or IO models is a promising frontier.
• Diffusion processes and behavioural heterogeneity strongly influence CE outcomes.

Data & calibration
Reviewed works vary widely: some use empirical data (surveys, industrial symbiosis parks, regional recycling flows), others stylized or conceptual models. Calibration practices are inconsistent; many rely on parameter assumptions.

Relevance for toolkit development

• Provides a taxonomy of CE ABM approaches useful for designing modular toolkits.
• Highlights the gap: most CE models lack integration with economy-wide systems, so future toolkits should allow bridging ABM with IO/LCA/CGE.
• Suggests design of generic, reusable agent heuristics for firms, consumers, and governments.
• Strongly relevant for Finland/EU as it discusses links to EU CE Action Plan and macroeconomic modelling.

Strengths & limitations

• Strengths: Transparent review strategy; systematic mapping; identifies thematic gaps; explicitly connects ABM to macroeconomic modelling needs.
• Limitations: Many reviewed works are grey literature or case-specific; lack of reproducibility and generalizability; few EU-focused empirical ABMs; limited integration with IO/LCA.

Classification for review
Include in main review.

Cluster membership
[Meta-level cluster placeholder: Systematic Reviews & Frameworks for CE ABMs].

[Post-processing note: Systematic review of 100 CE ABMs; identifies producer, consumer, and innovation diffusion themes; highlights integration gaps with IO/SFC; highly relevant as a “meta-level” anchor for the review chapter introduction.]

Notes Directly relevant to main review, as it systematically surveys CE ABM applications and proposes a general modelling framework.

Sauvageau and Frayret - 2015 - Waste paper procurement optimization An agent-based simulation approach

Abstract This paper proposes an agent-based simulation model to study and analyse the performance of various procurement and production policies in the recycled paper industry. The proposed model includes the recycled pulp production process, as well as the waste paper inventory and procurement processes. A detailed simulation model developed in partnership with a large recycled pulp producer in North America was developed in order to emulate the procurement manager’s behaviour. Therefore, based on the observed behaviour of the procurement manager, a procurement behaviour model, which takes both market price and inventory requirement into account, is introduced. This paper also introduces a waste paper market model that simulates a market price and enables the control of price forecast accuracy. Two series of experiments were carried out in order to study the performance of procurement and production policies in several productions contexts. Results show that production Volume Flexibility has a negative impact on costs, inventory and quality. However, it is possible to partially reduce these issues with the introduction of contracts with Volume Flexibility, although only a limited effect has been observed in our experiments. A more significant strategy to improve costs consists in reducing production rate to the minimum level required to meet demand.

Bibliographic info
Sauvageau, G., & Frayret, J.-M. (2015). Waste paper procurement optimization: An agent-based simulation approach. European Journal of Operational Research, 242(3), 987–998. https://doi.org/10.1016/j.ejor.2014.10.035

Scope & system studied

• Sector: Recycled pulp and paper industry.
• Geography: North America (case study with a large pulp producer).
• CE strategy: Optimizing procurement and production policies in closed-loop supply chains for recycled paper.

Model type & structure (ODD 2020 headings)

• Purpose and patterns: Evaluate how procurement and production policies (capacity, flexibility, contracts, inventory) influence cost, inventory, and quality in recycled pulp production.
• Entities, state variables, and scales:
  • Agents: Procurement manager, suppliers (contractual & market), market agent, sorting centre, warehouse, operation manager, fabrication centre.
  • Variables: Waste paper grades, quality (brightness), costs (procurement, storage, production), inventory levels, demand.
  • Scale: Daily operations across 2001–2012; system-level analysis of one pulp producer.
• Process overview and scheduling:
  • Waste paper procured from contracts and spot market.
  • Sorting centre assesses and adjusts quality.
  • Operation manager computes recipes, inventory targets, and production schedules.
  • Procurement manager decides purchases based on forecasts, inventory, and market prices.
  • Fabrication centre processes inputs into recycled pulp.
• Design concepts:
  • Emergence: System performance metrics from interacting agents.
  • Adaptation: Procurement decisions adapt to market forecasts and inventory levels.
  • Interaction: Market and suppliers respond to producer’s decisions.
  • Stochasticity: Quality variability, market forecast errors.
• Initialization: Historical production, procurement, and price data (2001–2012).
• Input data & calibration: Regression analysis of US waste paper markets; industrial partner data on procurement, contracts, production, inventory; interviews with procurement managers. Model validated against 2012 operational data (costs within ±6% of actual).
• Decision modules: Procurement manager decision rules (inventory ratio + price forecast), recipe optimization, supplier performance.

Policy / strategy levers examined

• Planned production capacity (low, normal, high).
• Production flexibility (volume fluctuations allowed).
• Target inventory levels.
• Procurement contracts with and without volume flexibility.
• Purchase control policy (limit excess buying).

Key findings

• Waste paper procurement >40% of total cost; procurement strategy critically affects performance.
• Production flexibility increases costs, inventory variation, and quality issues.
• Reducing production to minimum demand significantly lowers costs.
• High target inventory lowers procurement cost slightly (buying during low prices) but increases warehouse costs.
• Flexible contracts stabilize inventory and warehouse costs under high production flexibility, but do not reduce procurement cost.
• Quality problems arise when inventory of required grades is insufficient.

Relevance for toolkit development

• Provides a robust ABM framework for procurement and production coordination in CE industries.
• Shows how to integrate procurement manager heuristics, market forecasting, and contract design into simulations.
• Methodologically relevant for Finnish/EU recycled paper and biomass procurement, informing profitability calculators and policy simulators.
• Demonstrates how empirical data from industry and market statistics can calibrate ABMs.

Strengths & limitations

• Strengths: Detailed, empirically validated model; captures procurement–production–market feedbacks; industrial partnership; robust experimentation.
• Limitations: Focused on one producer; excludes sales side of pulp; no environmental impact assessment; North American market specifics.
• Transferability: Method transferable to EU CE procurement/logistics contexts; would need local data (EU paper recycling, biomass procurement).

Classification for review
Include in main review.

Cluster membership
Cluster 2: Waste System ABMs – Supply Chains, Competition, Consumer/Retail Behavior, and Recycling Configurations.

[Post-processing note: Industrial ABM of waste paper procurement/production coordination; emphasizes procurement manager heuristics and contract design; complements Knoeri, He et al., Nguyen-Trong in supply chain/logistics waste cluster.]

Scalco et al. - 2017 - The Implementation of the Theory of Planned Behavior in an Agent-Based Model for Waste Recycling

Abstract In the near future, the waste management sector is expected to reduce substantially the adverse effects of garbage on the environment. However, the increasing complexity of the current waste management systems makes the optimization of the waste management strategies and policies challenging. For this reason, waste prevention is the most desirable goal to achieve. Despite this, low levels of household recycling represent the key factor that complicates the current scenario. Keeping this in mind, the present work investigates the determinants of recycling behavior through the development of an agent-based model. Particularly, we examined what would induce households to increase the probability to engage in recycling behaviors on the base of the individual attitude and sensitivity to social norms. The Theory of Planned Behavior (TPB) has been implemented as agents’ cognitive model in environmental studies with the aim to predict recycling outcomes. Furthermore, in order to increase the realism of the simulation and the adherence of the model with the theory, we followed two strategies: firstly, we used real data to model a city district (Diong, Internship Report: Integrated Waste Management in Kaohsiung City, 2012). Secondly, we made use of the coefficients of the structural equation model presented in the work by Chu and Chiu (J Appl Soc Psychol 33(3):604–626, 2003) to build the agents’ cognitive model. As a whole, the results are in line with literature on descriptive social norms. Furthermore, the results indicate that the introduction of descriptive social norms represents a valuable strategy for public policies to improve household recycling: however, injunctive social norms are needed first.

Bibliographic info
Scalco, A., Ceschi, A., Shiboub, I., Sartori, R., Frayret, J.-M., & Dickert, S. (2017). The implementation of the Theory of Planned Behavior in an agent-based model for waste recycling: A review and a proposal. In A. Alonso-Betanzos et al. (Eds.), Agent-Based Modeling of Sustainable Behaviors (pp. 77–97). Springer, Cham. https://doi.org/10.1007/978-3-319-46331-5_4

Scope & system studied

• Sector: Household municipal waste recycling.
• Geography: Modeled on Kaohsiung City, Taiwan (San-min district).
• CE strategy: Improving household recycling compliance through social norms and behavioral interventions.

Model type & structure (ODD 2020 headings)

• Purpose and patterns: Explore how social norms (descriptive, injunctive) and psychological determinants affect household recycling participation.
• Entities, state variables, and scales:
  • Agents: Neighborhood households (~353,451 population), garbage transporters, landfills.
  • Variables: Environmental attitude, subjective norms, perceived behavioral control, perceived moral obligation.
  • Scale: City-district; simulation over multiple time steps until equilibrium reached.
• Process overview and scheduling:
  • Households generate waste and decide whether to recycle based on Theory of Planned Behavior (TPB).
  • Peer influence and surrounding influence (observing neighbors and visible rubbish) modify recycling probability.
  • Garbage transporters collect and deliver waste to recycling vs. non-recycling landfills.
• Design concepts:
  • Emergence: Aggregate recycling rates from heterogeneous individual attitudes and norms.
  • Adaptation: Agents’ probability to recycle shifts with peer/surrounding influence.
  • Interaction: Neighborhood social influence and observed environment.
  • Stochasticity: Probabilistic decision-making, heterogeneous agent traits.
• Initialization: District-level population and waste production data (Kaohsiung); structural equation model (SEM) coefficients from Chu & Chiu (2003).
• Input data & calibration:
  • Waste and demographic data from Kaohsiung City internship report (Diong 2012).
  • SEM coefficients for TPB constructs (attitudes, norms, control, moral obligation).
• Decision modules: Recycling decision function combining TPB constructs with SEM coefficients; peer and surrounding influence decay functions.

Policy / strategy levers examined

• Manipulation of social norms: descriptive vs injunctive.
• Testing sensitivity of recycling participation to peer influence and visible rubbish levels.
• Framework designed to later simulate campaigns (injunctive norm activation, tailored interventions).

Key findings

• Descriptive norms create self-reinforcing equilibria: high recycling environments sustain high compliance; low recycling contexts reinforce non-compliance.
• Injunctive norms (moral obligation) are more robust drivers: can sustain recycling even under adverse conditions.
• Combination of injunctive norms followed by descriptive norm reinforcement most effective policy pathway.
• Demonstrates importance of sequencing social norm strategies.

Relevance for toolkit development

• Provides a rigorous psychological foundation (TPB + SEM) for modeling household recycling.
• Suggests transferable method for incorporating empirical psychological constructs into ABMs.
• Highly relevant for Finnish/EU community waste sorting programs; framework could be adapted with local survey and SEM data.
• Offers pathway to integrate behavioral economics into CE profitability/policy simulators.

Strengths & limitations

• Strengths: Novel integration of SEM coefficients into ABM; theoretical rigor (TPB, social norms); empirical grounding in Kaohsiung data.
• Limitations: Calibration relies on secondary studies (Chu & Chiu 2003); limited validation; context-specific (Taiwan).
• Transferability: High methodological relevance; would require new SEM studies in Finnish/EU contexts.

Classification for review
Include in main review.

Cluster membership
Cluster 2: Waste System ABMs – Supply Chains, Competition, Consumer/Retail Behavior, and Recycling Configurations.

[Post-processing note: Household recycling ABM using TPB + SEM; highlights role of descriptive vs injunctive norms; complements Ma 2023 and Ortiz Salazar 2021 in behavioral waste cluster.]

Shi et al. - 2014 - Multi-objective agent-based modeling of single-stream recycling programs

Abstract In this research, our goal is to develop an agent-based simulation-based decision making framework for the effective planning of single-stream recycling (SSR) programs. The proposed framework is comprised of two main modules: a structured database and a simulation module. The database houses data necessary for simulating the entire solid waste management (SWM) system. The simulation module performs two major tasks. First, it identifies the various sources of system uncertainties and incorporates them into the SSR simulation model. Second, it compares and evaluates the alternatives of SSR (i.e., dual-stream recycling) with respect to cost, bottleneck facilities, and types and capacities of the processing facilities needed. For demonstration purposes, the proposed framework is applied to the state of Florida which has set a goal of reaching a 75% recycling rate by 2020. The proposed framework is a powerful tool that can be used by stakeholders for the evaluation of several “what-if” scenarios in their system before reaching a conclusion and making a decision.

Bibliographic info
Shi, L., Xu, L., & Faure, M. (2014). Multi-objective agent-based modeling of single-stream recycling programs. Resources, Conservation & Recycling, 92, 119–131. https://doi.org/10.1016/j.resconrec.2014.09.001

Scope & system studied

• Sector: Municipal solid waste (single-stream recycling programs).
• Geography: United States (generalized urban context).
• CE strategy: Assessing household recycling participation, contamination, and costs under different program designs.

Model type & structure (ODD 2020 headings)

• Purpose and patterns: Evaluate trade-offs between recycling rate, contamination, and program costs; identify optimal recycling program designs.
• Entities, state variables, and scales:
  • Agents: Households, haulers, government program managers.
  • Variables: Household awareness, environmental attitude, convenience perception, participation level, contamination rate.
  • Scale: City/community; multiple program design scenarios tested.
• Process overview and scheduling:
  • Households decide whether and how to recycle based on attitudes, awareness, and convenience.
  • Haulers collect recyclables, affected by contamination levels.
  • Program managers evaluate recycling rates, costs, contamination, and policy objectives.
• Design concepts:
  • Emergence: System-level recycling rates, contamination levels, and costs from household decisions.
  • Adaptation: Household awareness and convenience perception evolve with experience.
  • Interaction: Haulers interact with households and managers; contamination affects system efficiency.
  • Stochasticity: Household decision-making includes randomness to capture heterogeneity.
• Initialization: Household demographics and attitudes parameterized from literature.
• Input data & calibration: Household survey data from U.S. recycling studies; contamination and cost data from municipal reports.
• Decision modules: Household recycling decision (attitude + awareness + convenience); contamination generation; program manager evaluation.

Policy / strategy levers examined

• Curbside collection availability.
• Bin size and frequency of collection.
• Education and awareness campaigns.
• Incentives vs penalties.
• Trade-off optimization across objectives (maximize participation, minimize contamination, minimize cost).

Key findings

• Convenience (bin size, curbside collection) is the strongest driver of household recycling participation.
• Education campaigns reduce contamination but have weaker effects on participation rates.
• Incentives and penalties improve participation but increase administrative costs.
• Multi-objective optimization shows that achieving high recycling rates often conflicts with cost minimization.
• Contamination remains a persistent challenge even under strong participation policies.

Relevance for toolkit development

• Demonstrates use of ABM for multi-objective policy evaluation in recycling systems.
• Highlights trade-offs policymakers must navigate: participation vs contamination vs costs.
• Directly transferable to Finnish/EU municipal waste sorting contexts, especially for assessing deposit-return schemes, PAYT, or bin infrastructure.
• Shows potential for integrating ABM with optimization to inform policy mixes.

Strengths & limitations

• Strengths: Multi-objective focus; integration of household behavior and system costs; calibrated with survey and municipal data.
• Limitations: U.S.-centric; simplified contamination dynamics; limited spatial detail; no environmental LCA.
• Transferability: High to EU contexts with local data; especially relevant for ongoing municipal recycling reforms.

Classification for review
Include in main review.

Cluster membership
Cluster 2: Waste System ABMs – Supply Chains, Competition, Consumer/Retail Behavior, and Recycling Configurations.

[Post-processing note: ABM of single-stream recycling; multi-objective optimization of participation, contamination, and cost; complements Scalco 2017, Ma 2023, Ortiz Salazar 2021 in behavioral recycling models.]

Skeldon et al. - 2018 - Agent-based modelling to predict policy outcomes A food waste recycling example

Abstract Optimising policy choices to steer social/economic systems efficiently towards desirable outcomes is challenging. The inter-dependent nature of many elements of society and the economy means that policies designed to promote one particular aspect often have secondary, unintended, effects. In order to make rational decisions, methodologies and tools to assist the development of intuition in this complex world are needed. One approach is the use of agent-based models. These have the ability to capture essential features and interactions and predict outcomes in a way that is not readily achievable through either equations or words alone. In this paper we illustrate how agent-based models can be used in a policy setting by using an example drawn from the biowaste industry. This example describes the growth of in-vessel composting and anaerobic digestion to reduce food waste going to landfill in response to policies in the form of taxes and financial incentives. The fundamentally dynamic nature of an agent-based modelling approach is used to demonstrate that policy outcomes depend not just on current policy levels but also on the historical path taken.

Bibliographic info
Skeldon, A. C., Schiller, F., Yang, A., Balke-Visser, T., Penn, A., & Gilbert, N. (2018). Agent-based modelling to predict policy outcomes: A food waste recycling example. Environmental Science & Policy, 87, 85–91. https://doi.org/10.1016/j.envsci.2018.05.011

Scope & system studied

• Sector: Food waste recycling (biowaste treatment).
• Geography: United Kingdom (national context, linked to EU Landfill Directive).
• CE strategy: Development of anaerobic digestion (AD) and in-vessel composting (IVC) as alternatives to landfill, under varying policy regimes.

Model type & structure (ODD 2020 headings)

• Purpose and patterns: Assess how UK landfill tax and renewable energy incentives (Renewable Obligation Certificates, ROCs) interact to influence the growth of AD vs IVC industries.
• Entities, state variables, and scales:
  • Agents: Food waste producers, AD plants, composters, landfill operators.
  • Variables: Waste flows, gate fees, financial reserves, contacts/network links.
  • Scale: Industry-level dynamics, 50-year simulation horizon.
• Process overview and scheduling:
  • Food waste producers generate waste and sell to composters, AD plants, or landfill.
  • Processors enter or exit the market based on profitability; build contacts to trade.
  • Prices for waste adjust dynamically with supply/demand.
• Design concepts:
  • Emergence: Growth and collapse cycles of AD/IVC industries.
  • Adaptation: Firms adjust pricing, capacity, and survival based on profits.
  • Interaction: Trading networks form incrementally among firms.
  • Stochasticity: Randomized firm start-ups and contacts.
• Initialization: Simulations begin with no processors, all waste to landfill.
• Input data & calibration: UK Landfill Tax data (1996–2018); ROCs framework (2002–2016); waste policy reports; validation with UK industry data on AD/IVC plants.
• Decision modules: Firm entry/exit, trading network formation, pricing dynamics, profitability calculation.

Policy / strategy levers examined

• Landfill Tax (levels and trajectory).
• Renewable Obligation Certificates (levels 4–10 p/kWh).
• Timing and sequencing of policy implementation.

Key findings

• Landfill Tax drives entry of composters but has limited long-term effect once competition raises prices.
• ROCs incentives are essential to make AD viable; higher levels accelerate AD dominance and IVC decline.
• Strong path dependency: the same policy levels can yield different industry compositions depending on historical trajectory (timing and duration of incentives).
• High ROCs may inadvertently subsidize waste producers (raising waste prices) and risk incentivizing waste generation rather than minimization.
• Policy sequence matters: long enough incentives can lock in AD, short incentives lead to collapse.

Relevance for toolkit development

• Shows ABM’s utility for policy scenario testing in CE markets, capturing path dependency and unintended effects.
• Framework transferable to Finnish/EU biowaste systems (e.g., biogas policy under EU Renewable Energy Directive).
• Demonstrates importance of dynamic simulation versus static equilibrium approaches.
• Relevant for profitability calculators (effect of subsidies/incentives) and CE simulators (competing treatment technologies).

Strengths & limitations

• Strengths: Rich integration of policy instruments, industry structure, and market dynamics; explicit demonstration of path dependency; open model code (NetLogo + MATLAB).
• Limitations: UK-specific context; excludes environmental LCA; stylized firm decision rules.
• Transferability: High methodological relevance, EU adaptation requires local data on subsidies, taxes, and biogas/composting markets.

Classification for review
Include in main review.

Cluster membership
Cluster 2: Waste System ABMs – Supply Chains, Competition, Consumer/Retail Behavior, and Recycling Configurations.

[Post-processing note: Food waste ABM with landfill tax + renewable energy incentives; shows strong path dependency in policy outcomes; complements Luo 2019 and Shi 2014 in waste policy ABMs.]

Termizi Siti Nor Azreen Ahmad et al. - 2023 - Agent-Based Modelling Analysis on the Potential Economic Benefits of Mud Cake Waste to Wealth

Abstract Production of refined sugar generates a significant amount of mud cake waste which has traditionally been managed through disposal in landfills. Such practice is not sustainable and can have negative environmental impacts. While previous research has primarily focused on evaluating the environmental and economic impacts of bagasse, less attention has been given to mud cake waste. By applying the waste-to-wealth concept and promoting a circular economy, this study aims to evaluate the potential economic benefits of using mud cake waste as an alternative raw material in cement production. To assess the potential economic benefits, the study employs an agent-based modelling analysis that simulates the actions and interactions of multiple agents. The results of the analysis demonstrate the potential economic benefits of mud cake waste exchange for both sugar producers and cement manufacturers. The concept of a win-win scenario is emphasized, where both parties share an equal amount of exchange costs and reap mutual benefits in cost savings, improved resource efficiency, and reduced environmental impacts by reducing the amount of waste being disposed of to the landfill.

Bibliographic info
Termizi, S. N. A. A., Wan Alwi, S. R., Manan, Z. A., & Varbanov, P. S. (2023). Agent-based modelling analysis on the potential economic benefits of mud cake waste to wealth. Chemical Engineering Transactions, 103, 655–660. https://doi.org/10.3303/CET23103110

Scope & system studied

• Sector: Sugar industry by-products and cement manufacturing.
• Geography: Malaysia.
• CE strategy: Industrial symbiosis through using mud cake waste (sugar industry residue) as an alternative raw material in cement production.

Model type & structure (ODD 2020 headings)

• Purpose and patterns: Assess potential economic benefits of industrial symbiosis between sugar refineries and cement manufacturers via mud cake exchange.
• Entities, state variables, and scales:
  • Agents: Sugar producers (mud cake generators), cement manufacturers (limestone users).
  • Variables: Waste generation, limestone demand, disposal costs, transportation costs, exchange percentages, cost factor coefficient (distribution of exchange costs).
  • Scale: Industrial symbiosis scenarios, monthly flows (10,000 Mt mud cake, 50,000 Mt limestone).
• Process overview and scheduling:
  • Scenario A: No symbiosis, mud cake disposed to landfill, cement producers purchase limestone.
  • Scenario B: Symbiosis, mud cake replaces limestone proportionally, exchange costs shared differently (cases B1–B3).
• Design concepts:
  • Emergence: Cost reductions for both industries under symbiosis.
  • Interaction: Cost-sharing arrangements between partners.
  • Stochasticity: Not emphasized (deterministic simulation).
• Initialization: Fixed monthly waste and limestone flows.
• Input data & calibration:
  • Sugar production and mud cake generation ratios from Gupta et al. (2011).
  • Cost data from Malaysian case studies (Sabeen et al. 2016).
  • Cement raw material substitution feasibility from Li et al. (2014) and Radwan et al. (2021).
• Decision modules:
  • Exchange decision (Q = 25–100%).
  • Cost allocation module (Lm = 0, 0.5, 1).

Policy / strategy levers examined

• Percentage of mud cake exchange (25–100%).
• Cost-sharing arrangements for transportation (producer pays, consumer pays, equal share).

Key findings

• Greater mud cake substitution increases cost reductions for both parties, with sugar producers benefitting more.
• Case B2 (shared transport costs) is most equitable: ~27% cost reduction for sugar producers and ~6% for cement manufacturers at 50% exchange.
• Producer-only payment (B3) nearly erases benefits; consumer-only payment (B1) maximizes producer savings.
• Environmental benefits include reduced landfill disposal and reduced limestone extraction.
• Realistic substitution capped at ~20% mud cake in cement mix, limiting maximum exchange.

Relevance for toolkit development

• Demonstrates ABM use for industrial symbiosis economic analysis, relevant for CE toolkit modules on secondary material markets.
• Framework transferable to Finnish/EU contexts (e.g., pulp/paper residues to construction materials).
• Provides a template for simulating win–win cost-sharing arrangements between industries.
• Shows how relatively simple ABM structures can model CE symbiosis scenarios with cost-sharing sensitivity.

Strengths & limitations

• Strengths: Clear industrial symbiosis framing; transparent cost assumptions; policy-relevant cost-sharing design.
• Limitations: Simplified (no environmental or technical quality assessment beyond feasibility); deterministic; case-specific to Malaysia.
• Transferability: Method applicable to EU CE industrial symbiosis; requires EU-specific waste, transport, and raw material data.

Classification for review
Include in main review.

Cluster membership
Cluster 3 (Placeholder): Industrial Symbiosis & Secondary Material Markets.

[Post-processing note: ABM of mud cake (sugar residue) → cement substitution; focuses on economic benefits and cost-sharing; anchor example for industrial symbiosis CE cluster 3.]

Tian et al. - 2025 - Agent-based modeling in solid waste management: Advantages, progress, challenges and prospects

Abstract The growing issue of solid waste management (SWM) is recognized as a significant challenge to ecosystem preservation. Agent-based modeling (ABM) has received significant attention for its capability to address complex systems and simulate the outcomes of strategic implementation. This review compares ABM with other methods and provides a comprehensive overview of research on ABM in SWM from 2000 to 2023, emphasizing its advantages, progress, challenges, and future directions. Results indicate that: 1) ABM possesses 8 key advantages in simulating individual behavior, responses to environmental changes across time and spatial scales, and decision-making processes, namely interactivity, heterogeneity, dynamism, traceability, spatiality, scalability, complexity, and adaptability. 2) Current research primarily focuses on simulating behavioral and strategic effects of SWM (accounting for 45.5 %), while multi-model coupling is becoming a new trend. 3) ABM encounters challenges in its research, including a lack of standardized research steps, high data dependency, limited computing resources, and difficulties in algorithm explanation. Therefore, this study introduces a set of normative steps that provide clear guidance for research and help ensure the reproducibility and accuracy of studies. Future research should incorporate big data and emerging technologies to enhance computational efficiency and processing capabilities of models. To better utilize ABM for achieving environmental protection and sustainable development goals, prioritizing integration of multi-model coupling, interdisciplinary collaboration, visualization, and open-source code sharing as key strategies is essential.

Bibliographic info
Tian, X., Peng, F., Wei, G., Xiao, C., Ma, Q., Hu, Z., & Liu, Y. (2025). Agent-based modeling in solid waste management: Advantages, progress, challenges and prospects. Environmental Impact Assessment Review, 110, 107723. https://doi.org/10.1016/j.eiar.2024.107723

Scope & system studied

• Sector: Solid waste management (SWM).
• Geography: Global review of ABM applications (2000–2023).
• CE strategy: SWM across the waste life cycle—generation, collection/transport, recycling, treatment, disposal—within circular economy and sustainable development contexts.

Model type & structure

• Purpose and patterns: Systematic review of 122 ABM studies on SWM; compares ABM with other methods and proposes standardized steps for SWM-focused ABM.
• Entities, state variables, and scales: Synthesizes studies with agents ranging from households and firms to governments, waste facilities, and transport networks.
• Process overview and scheduling: Organizes literature into four categories: (1) waste generation/distribution prediction, (2) collection/transport optimization, (3) sustainability indicators, (4) behavioral and strategic effects.
• Design concepts: Emphasizes ABM strengths: heterogeneity, dynamics, spatiality, complexity, adaptability.
• Initialization: Case-specific across studies (community-level, regional, national).
• Input data & calibration: Mix of survey data, municipal statistics, GIS, LCA, techno-economic analysis (TEA). Notes data-dependence and lack of standardization as key challenges.
• Submodels: No single model—review notes frequent use of behavioral decision modules (e.g., Theory of Planned Behavior), transport optimization (ABM+GIS), and hybrid couplings (ABM + SD, DES, MFA, LCA).

Policy / strategy levers examined
Across reviewed studies:

• Incentives, subsidies, and penalties for household sorting.
• Landfill tax, EPR, and cap-and-trade schemes.
• Facility siting and bin placement.
• Industry-level subsidies/regulation for recycling and symbiosis.
• Combined strategies (e.g., subsidy + regulation mixes).

Key findings

• ABM has eight key advantages: interactivity, heterogeneity, dynamism, traceability, spatiality, scalability, complexity, adaptability.
• Research focus: 45.5% on behavioral/strategic effectiveness; less on transport optimization.
• Multi-model coupling (ABM+GIS+SD+DES+LCA) is emerging as a major trend.
• Challenges: lack of standardized research protocols, high data dependency, limited computing resources, algorithm interpretability issues.
• Proposes a seven-step standardized framework for ABM in SWM (system analysis → design → data collection → simulation → validation → experimentation → interpretation).
• Future directions: integrate big data/IoT/AI; interdisciplinary collaboration; hybrid modeling; open-source sharing.

Data & calibration
Draws on diverse sources: household surveys, municipal records, industry data, GIS networks, LCA/TEA. Review highlights wide variation in calibration quality and the frequent reliance on assumptions when data are scarce.

Relevance for toolkit development

• Provides a meta-level methodological roadmap for ABM in CE/SWM, useful to guide Finnish/EU toolkit standardization.
• Emphasizes integration of behavioral, logistical, and environmental submodels—relevant to profitability calculators and CE simulators.
• Suggests standardized ABM development steps that DYNAMO can adopt to ensure reproducibility and comparability.
• Strong case for hybrid coupling with ENVIMAT (EEIO), LCA, GIS, and optimization modules.

Strengths & limitations

• Strengths: Broad systematic review; detailed categorization across SWM lifecycle; clear identification of ABM’s comparative strengths and gaps; concrete seven-step process.
• Limitations: Descriptive synthesis rather than quantitative meta-analysis; heavy reliance on Chinese and Asian case studies; less emphasis on European empirical applications; does not present new model but reviews existing.
• Transferability: High methodological relevance to EU/Finnish contexts.

Classification for review
Include in main review.

Cluster membership
Cluster 9: Systematic Reviews & Frameworks for CE ABMs.

[Post-processing note: Systematic review of 122 ABM SWM papers; emphasizes standardized seven-step framework, hybrid modeling, and data/AI integration; forms major methodological anchor for waste-focused ABM cluster and toolkit design.]

Tong et al. - 2018 - Behaviour change in post-consumer recycling Applying agent-based modelling in social experiment

Abstract Change in consumer behavior that leads to increased waste separation and recycling has been identified as a critical component of Chinese national strategy for constructing a “Circular Economy”. Various innovative solutions at community level targeting consumer behaviors are emerging in Chinese cities, using information technology that can track the volume and quality of the sorting process. In order to evaluate the potential impact of these novel solutions, we studied the behavioral change of households by initiating an experimental recycling program in a residential community in Beijing, and developed an Agent Based Model based on Theory of Planned Behavior (TPB) to identify key factors in changing behavior. The results show that the Social Norm (SN) has a decisive effect on whether an area starts recycling or not. As to the effectiveness of intervention, the Perceived Behavioral Control (PBC) plays a large role in the determination of the recycling behavior in this study, while the role of attitude is relatively small. The model outcomes can be corroborated with observations in different communities using similar technical solutions. In conclusion, we suggest that efficient local interactions among various stakeholders are needed in forming the social norm and common space that favorite recycling activities at the community level.

Bibliographic info
Tong, X., Nikolic, I., Dijkhuizen, B., van den Hoven, M., Minderhoud, M., Wäckerlin, N., Wang, T., & Tao, D. (2018). Behaviour change in post-consumer recycling: Applying agent-based modelling in social experiment. Journal of Cleaner Production, 187, 1006–1013. https://doi.org/10.1016/j.jclepro.2018.03.261

Scope & system studied

• Sector: Post-consumer household recycling.
• Geography: Beijing, China (Hongfuyuan community).
• CE strategy: Community-based recycling program supported by IT-enabled tracking (barcode and mobile app system).

Model type & structure (ODD 2020 headings)

• Purpose and patterns: Couple a social experiment with ABM to identify behavioral factors influencing household recycling decisions.
• Entities, state variables, and scales:
  • Agents: Households (socio-economic, cognitive, local context attributes), containers, collectors, landfill.
  • Variables: Household size, education, income, awareness, recycling intention, willingness to change, storage/time constraints, distance to container, social networks.
  • Scale: ~10,000 households across three neighborhoods; daily simulation.
• Process overview and scheduling:
  • Households generate recyclables daily; once threshold reached, decide between landfill, container, or collector.
  • Decisions based on Theory of Planned Behavior (TPB): attitudes, subjective norms, perceived behavioral control (PBC).
  • Social interactions weekly update perceived social norms.
• Design concepts:
  • Emergence: Participation rates and waste diversion outcomes.
  • Adaptation: Attitudes and PBC updated via incentives and social feedback.
  • Interaction: Household peer effects; collector interactions.
  • Stochasticity: Randomized household attributes, probabilistic decision-making.
• Initialization: 500-household pilot experiment data extrapolated to ~10,000 households.
• Input data & calibration: Empirical social experiment (2013–2016); household registration, participation data, container proximity, incentive records.
• Decision modules: Household recycling decision tree (TPB-based); incentive integration; social norm diffusion.

Policy / strategy levers examined

• Infrastructure: container availability and siting vs collector service.
• Incentives: economic rewards, score redemption for goods, organic vegetable gifts.
• Social interactions: promotion events, junk-buyer on-call services.
• Comparative scenarios: baseline vs new scheme (containers vs collector emphasis).

Key findings

• Social norms are decisive: communities with higher recycling participation sustain growth; low norms suppress adoption.
• PBC (knowledge, willingness, capacity) strongly influences recycling; attitude plays a smaller role.
• Collectors (door-to-door service with incentives) more effective than containers alone; up to 59% participation achievable.
• IT-based tracking enables tailored incentives and accountability.
• Scaling requires building local social norms and integrating informal collectors into CE systems.

Relevance for toolkit development

• Strong demonstration of combining empirical social experiments with ABM for recycling policy evaluation.
• Provides transferable methodology for Finnish/EU pilots (PAYT, EPR, IT-enabled recycling apps).
• Suggests practical ways to model interaction of incentives, infrastructure, and norms in community CE simulators.
• Useful for designing survey-calibrated ABM modules on household recycling in the DYNAMO toolkit.

Strengths & limitations

• Strengths: Coupling of real-world experiment and ABM; TPB integration; open model code (NetLogo); scenario experimentation.
• Limitations: Specific to one Chinese community; limited long-term generalization; PBC proxies simplified; does not link to cost/LCA impacts.
• Transferability: High methodological relevance; EU calibration requires local data and experiments.

Classification for review
Include in main review.

Cluster membership
Cluster 2: Waste System ABMs – Supply Chains, Competition, Consumer/Retail Behavior, and Recycling Configurations.

[Post-processing note: ABM coupled with Beijing social experiment; emphasizes social norms and PBC; adds empirical validation case to behavioral recycling ABMs alongside Scalco 2017, Ma 2023, Ortiz Salazar 2021.]

Tong et al. - 2023 - Exploring business models for carbon emission reduction via post-consumer recycling infrastructures.pdf

Abstract The potential for carbon emissions avoidance through post-consumer recycling has been highlighted in the “Zero-Waste City” initiatives in China, which call for increasing household participation in the community recycling programs. A shift from a facility-oriented strategy to behavior-oriented norm building at the community level provides the local niches for emerging business models in post-consumer recycling. This paper proposes a framework based on the Theory of Planned Behavior to simulate the impact on the participation rate of households in community-based recycling program in the context of different recycling business models in Beijing. Firstly, a questionnaire survey was conducted in 2021 in Beijing (N = 1153) to test the key factors which affect household recycling behavior. Secondly, an agent-based model for community-based recycling which was developed in 2016 has been updated to incorporate the institutional change in the development of zero-waste city initiatives. Scenario analysis is adopted to compare the effects of introducing two new business models to the baseline situation that is prevalent at present. The settings for the two new business models represent the diversified directions of the upgrading of urban recycling sector: one is to improve the intelligence of the collection facilities to save labors, and the other is to involve the informal recyclers to provide door-to-door collection services in person so as to save time for the residents. Additional scenario shows the effects of norm-based solutions in combination with the two strategies addressed above in the long run. The simulation results show that the potential for carbon emissions reduction through post-consumer recycling in Beijing can range from 1 million tons per year at the basic scenario to more than 4.5 million tons per year in the community level norm-based solution scenario. In conclusion, the proposal for sustaining the community-based new business models through the capture of value in the carbon emission reduction is put forward as a guideline for urban recycling infrastructure design.

Bibliographic info
Tong, X., Yu, H., Han, L., Liu, T., Dong, L., Zisopoulos, F. K., Steuer, B., & de Jong, M. (2023). Exploring business models for carbon emission reduction via post-consumer recycling infrastructures in Beijing: An agent-based modelling approach. Resources, Conservation & Recycling, 188, 106666. https://doi.org/10.1016/j.resconrec.2022.106666

Scope & system studied

• Sector: Post-consumer municipal solid waste (recyclables).
• Geography: Beijing, China (city-level, with community-level microdynamics).
• CE strategy: Upgrading recycling infrastructures with new business models (intelligent facilities, formalized door-to-door collection, norm-based policies).

Model type & structure (ODD 2020 headings)

• Purpose and patterns: Assess household recycling participation and carbon emissions reduction potential under alternative business models, using ABM calibrated with survey and participatory data.
• Entities, state variables, and scales:
  • Agents: Households (heterogeneous socio-demographic, cognitive, residential status), collectors (formal/informal), facilities.
  • Variables: Attitudes, subjective norms, perceived behavioral control (PBC), community perception, age, hukou status.
  • Scale: Community-level ABM (~10,000 households), upscaled to city level (Beijing).
• Process overview and scheduling:
  • Households generate waste, decide disposal (bin, collector, mixing) based on TPB+ constructs and incentives.
  • Social interactions update norms and community perception.
  • Participation rates aggregated and scaled to city emissions calculations.
• Design concepts:
  • Emergence: Recycling participation and emissions reduction.
  • Adaptation: Norms and PBC evolve with incentives, infrastructure, and peers.
  • Interaction: Household–collector–facility dynamics; social networks.
  • Stochasticity: Randomized attributes and decision variation.

• Initialization: Based on survey (N=1153), participatory observation, and census demographics.

• Input data & calibration:
  • Survey (Beijing, 2021), SEM analysis for TPB+ factors.
  • Field observation of community recycling programs (2012–2022).
  • Census demographics and waste statistics (Beijing 2020–2021).
  • Carbon emission factors from Chinese databases and literature.
• Decision modules: Household TPB+ decision module; incentive module; social norm diffusion module; carbon reduction calculation.

Policy / strategy levers examined

• Baseline (no new models).
• Intelligent facility model (tech-enabled bins, incentives).
• Service-oriented model (formalized informal recyclers with subsidies, door-to-door).
• Norm-based model (strong social norms + infrastructure integration).

Key findings

• Participation rate: ~10% (baseline) → 20% (intelligent) → 50% (service-oriented) → >70% (norm-based).
• Carbon emission reduction potential: ~1 MtCO₂/year (baseline) → 1.5 (intelligent) → 2.6 (service-oriented) → 4.5 MtCO₂/year (norm-based).
• Community perception variable significantly increases explanatory power of TPB model (explained variance 39% → 43%).
• Local residents more influenced by social norms; immigrants more by incentives and convenience.
• Integrating informal recyclers into formal systems improves both participation and equity.
• Norm-based approaches crucial for long-term sustainability; subsidies alone insufficient.

Relevance for toolkit development

• Provides a city-level CE ABM calibrated with large-scale survey and SEM, highly relevant for Finland/EU household waste and carbon policy modeling.
• Demonstrates methodology for linking behavioral ABMs with carbon accounting.
• Strong template for integrating business model design (PSS for waste services) into profitability and emissions calculators.
• Highlights interplay between infrastructure, incentives, and norm-building—transferable to EU zero-waste and low-carbon city initiatives.

Strengths & limitations

• Strengths: Empirical calibration (survey + SEM + field observation); multi-scenario policy design; explicit carbon reduction quantification; integration of social norms.
• Limitations: Beijing-specific context; results shaped by hukou/informal sector dynamics; does not integrate with IO/LCA beyond emissions factors.
• Transferability: High conceptual and methodological relevance; EU calibration requires local surveys and emissions factors.

Classification for review
Include in main review.

Cluster membership
Cluster 2: Waste System ABMs – Supply Chains, Competition, Consumer/Retail Behavior, and Recycling Configurations.

[Post-processing note: ABM of recycling business models in Beijing; calibrated with SEM survey + field data; explicit carbon reduction results; adds city-level, emissions-focused dimension to waste ABM cluster.]

Walzberg et al. - 2020 - Closing the Loops on Solar Photovoltaics Modules An Agent-Based Modeling Approach

Abstract Solar photovoltaics (PV) installed capacity have grown exponentially since the early 2000s (average annual growth rate of 50%). With 4,700 GW projected installed capacity by 2050, the volume of PV panels waste is also expected to become substantial. Though renewables are a sine qua non to the establishment of a truly circular economy (CE), the issue arising from their end-of-life management needs to be resolved. The management of PV end-of-life also represent a singular opportunity to create value, from recovered valuable, rare or critical materials (e.g. silver, tellurium, indium). Moreover, the PV case illustrates some of the current barriers to CE; for instance, the need for a common definition of waste (PV are defined as e-waste in the European Union but as general waste in the United States) and potential loss of innovations’ advantages (PV average efficiency has continuously grown the past 10 years). In this study, an agent-based modeling (ABM) approach is proposed to simulate PV end-of-life management in the United States. The model explores how the decisions of the various actors involved in handling PV waste affect the quantities of PV that are reused, recycled, or land-filled. In the model, manufacturers, residential, and nonresidential PV owners, installers, and recyclers are represented by agents while governmental policies and regulations are treated as exogenous variables. PV-market data are used to define agents’ characteristics (e.g., installed PV capacities or producing costs). Agents’ decision rules draw on the literature related to industrial symbiosis (IS) and peoples’ waste behaviors as they appear as the most widely used model to implement CE principles at a meso level. Specifically, the concepts of mutual trust between IS actors and knowledge about the IS philosophy are yielded to define the agents’ decision process. The primary outputs of the ABM are the costs and volume of waste associated with each end-of-life pathway. Preliminaries results indicate recycling of residential PV is affected by the cost, perceived difficulty of recycling behaviors, and social norms. Moreover, the type of network connecting agents and the initial recycling rate strongly influence results. The model also highlights the crucial role of recycling behavior adoption: in case of early failure of PV, recycled volumes increase by about 40%. This represents an additional 5.7 billion USD from potential recovered silver. Finally, although results for the PV case are presented here, the developed ABM aims at being a general tool for the study of CE strategies and the CE transition. Furthermore, steps of this research include extending the model to encompass circular economy strategies (e.g., design for recycling or lifetime extension) and validating its general architecture from various case studies.

Bibliographic info
Walzberg, J., Carpenter, A., & Heath, G. (2020). Closing the loops on solar photovoltaics modules: An agent-based modeling approach for the study of circular economy strategies. National Renewable Energy Laboratory (NREL) Conference Paper, PV Reliability Workshop, Lakewood, CO, February 25–27, 2020. NREL/PO-6A20-76139.

Scope & system studied

• Sector: Solar photovoltaic (PV) modules, end-of-life (EoL) management.
• Geography: Global PV industry (stylized, not region-specific).
• CE strategy: Examining reuse, repair, resale, and recycling pathways for EoL PV panels, with attention to costs, norms, and industrial symbiosis.

Model type & structure (ODD 2020 headings)

• Purpose and patterns: Identify factors influencing EoL pathways for PV modules; explore how subjective norms, costs, and repair potential shape circularity.
• Entities, state variables, and scales:
  • Agents: PV owners, recyclers, secondary market actors.
  • Variables: Attitudes, subjective norms, initial recycling costs, repair potential, resale price ratio, landfill/storage costs.
  • Scale: Industry-level dynamics; long-term horizon (2015–2060).
• Process overview and scheduling:
  • PV owners decide among resale, repair, recycling, or landfill based on Theory of Planned Behavior (TPB) factors (attitudes, norms, costs).
  • Recyclers’ costs decrease with scale (learning-by-doing).
  • Market for resale/repair competes with recycling.
• Design concepts:
  • Emergence: Shares of modules in each EoL pathway.
  • Adaptation: Agents update recycling choices as costs drop.
  • Interaction: PV owners influenced by peer norms; recyclers’ viability shaped by volume.
  • Stochasticity: Limited; model uses deterministic rules from literature data.
• Initialization: Literature-based baseline: 30% recycling rate, 55% repair potential, recycling cost $25–30/module.
• Input data & calibration: Literature sources (IRENA & IEA 2016; Tsanakas 2019; Tong et al. 2018 on TPB). Cost assumptions from industry studies.
• Decision modules: TPB-based owner decision module; recycler cost-decline module; resale/repair profitability module.

Policy / strategy levers examined

• Initial recycling cost (baseline vs reduced through R&D).
• Repair potential and resale ratios.
• Social norm strength (peer effects).

Key findings

• Secondary market (resale/repair) dominates initially, being financially attractive but limited by repair potential.
• Recycling uptake depends strongly on lowering initial costs (via R&D and scale economies).
• Recycling and resale compete: higher resale demand reduces recycling volumes.
• Strong subjective norms increase recycling even at higher cost.
• Recycling cost reductions through increased waste volumes accelerate circularity.

Relevance for toolkit development

• Demonstrates ABM of CE pathways for renewable energy technologies, highly relevant to EU/Finland PV waste projections.
• Shows how TPB-based behavioral rules can be integrated with economic cost-decline functions.
• Provides template for linking micro-level owner decisions with industry-level recycling capacity growth.
• Suggests synergy with profitability calculators for PV recycling investment under EU directives.

Strengths & limitations

• Strengths: Integration of behavioral and economic cost dynamics; transparent parameterization from literature; relevant to large-scale CE transitions.
• Limitations: Conference paper (not full peer-reviewed journal); stylized assumptions; no spatial representation; limited empirical calibration.
• Transferability: Conceptually strong; empirical EU calibration (PV waste projections, recycling costs, norm data) required.

Classification for review
Include in main review.

Cluster membership
Cluster 7: Consumer Behavior & Product-Service Strategies (product-owner decisions and service pathways) and cross-linked to Cluster 1 (Macro vs Meso CE Models) as renewable energy adoption/waste management.

[Post-processing note: ABM of PV module EoL pathways; emphasizes TPB, cost-decline, resale vs recycling trade-offs; complements Koide 2023 and Hanes & Walzberg 2022; link to EU PV recycling policy debates.]

Walzberg et al. - 2021 - Exploring Social Dynamics of Hard-Disk Drives Circularity with an Agent-Based Approach

Abstract By 2025, it is estimated that installed data storage in the U.S. will be 2.2 Zettabytes, generating about 50 million units of end-of-life hard-disk drives (HDDs) per year. The circular economy (CE) tackles waste issues by maximizing value retention in the economy, for instance, through reuse and recycling. However, the reuse of hard disk drives is hindered by the lack of trust organizations have toward other means of data removal than physically destroying HDDs. Here, an agent-based approach explores how organizations’ decisions to adopt other data removal means affect HDDs’ circularity. The model applies the theory of planned behavior to model the decisions of HDDs end-users. Results demonstrate that the attitude (which is affected by trust) of end-users toward data-wiping technologies acts as a barrier to reuse. Moreover, social pressure can play a significant role as organizations that adopt CE behaviors can set an example for others.

Bibliographic info
Walzberg, J., Frost, K., Zhao, F., Carpenter, A., & Heath, G. (2021). Exploring social dynamics of hard-disk drives circularity with an agent-based approach. In 2021 IEEE Conference on Technologies for Sustainability (SusTech) (pp. 1–7). IEEE. https://doi.org/10.1109/SusTech51236.2021.9467439

Scope & system studied

• Sector: Hard-disk drives (HDDs) and critical material recovery (rare earth elements, aluminum, copper, steel).
• Geography: United States (stylized but calibrated with U.S. data).
• CE strategy: Improving HDD circularity through reuse, component reuse (magnets), and advanced recycling, versus shredding and storage.

Model type & structure (ODD 2020 headings)

• Purpose and patterns: Assess how social trust, subjective norms, and cost factors shape HDD EoL pathways; quantify barriers to reuse and component recovery.
• Entities, state variables, and scales:
  • Agents: HDD end-users (service providers, government, commercial), recyclers, initial service providers, manufacturers.
  • Variables: Attitudes, subjective norms, perceived behavioral control (PBC/cost), trust in data-wiping, reuse/repair potential, demand for used HDDs.
  • Scale: ~1000 end-user agents + 250 recyclers + 15 service providers + 13 manufacturers; simulation horizon 2020–2050.
• Process overview and scheduling:
  • End-users decide annually among reuse, component reuse, advanced recycling, shredding, or storage, using TPB-based decision scores.
  • Recyclers recover materials at specified rates; manufacturers buy recovered materials.
  • Social networks (Watts–Strogatz topology) spread norms influencing adoption.
• Design concepts:
  • Emergence: Industry-level circularity rates and material flows.
  • Adaptation: Trust and norms shift end-user decisions.
  • Interaction: End-users influence each other’s decisions; recyclers interact with manufacturers.
  • Stochasticity: Social network connections, cost distributions, initial trust values.
• Initialization: Installed HDD capacity 2000–2020; projected growth to ~2 zettabytes by 2025; empirical market shares of HDD materials.
• Input data & calibration: Projections from IDC (Gantz & Reinsel), iNEMI value recovery project, U.S. DOE/USGS data, SEM meta-analyses of recycling behaviors for TPB coefficients.
• Decision modules: TPB-based EoL decision module (attitude, norms, PBC/trust), recycler cost and recovery module, manufacturer purchasing module.

Policy / strategy levers examined

• Trust in data-wiping (low → high).
• Subjective norm strength.
• Cost of recycling and reuse profitability.
• Demand for second-hand HDDs.
• Constraints on reuse cycles.

Key findings

• Baseline: ~6% reuse, ~1% component reuse, ~68% recycling, ~24% landfill/storage (2050).
• Increasing trust in data wiping is the most critical enabler for reuse; without it, shredding dominates.
• Subjective norms reinforce existing attitudes: they can accelerate circularity when trust is high, or entrench shredding when trust is low.
• Component reuse and advanced recycling remain marginal without strong social and economic incentives.
• Even under ideal reuse scenarios, circularity is capped (~70% reuse, 3% component reuse, 20% recycling) due to technical limits and restricted reuse lifecycles.

Relevance for toolkit development

• Shows how trust and norms act as bottlenecks for CE adoption, relevant for EU data security and electronics recycling contexts.
• Provides transferable methodology: combining TPB-based behavioral modeling with material flow tracking.
• Relevant for Finnish/EU CE policies on WEEE, data security, and rare earth recovery.
• Illustrates integration potential of behavioral ABM with IO/LCA assessments of critical material flows.

Strengths & limitations

• Strengths: Empirical calibration of parameters; integration of social psychology; modular Python/Mesa implementation; detailed scenario analysis.
• Limitations: Stylized U.S. case; excludes households and competition between recyclers; no environmental LCA integration; capped reuse cycles.
• Transferability: High methodological relevance; requires EU survey and market data.

Classification for review
Include in main review.

Cluster membership
Cluster 7: Consumer Behavior & Product-Service Strategies (behavioral adoption, trust, norms in CE pathways). Cross-links to Cluster 3 (Industrial Symbiosis & Secondary Materials) due to rare earth recovery context.

[Post-processing note: ABM of HDD circularity; emphasizes role of trust in data-wiping and subjective norms; calibrated with U.S. data; complements Walzberg 2020 PV ABM and Koide 2023 in consumer adoption cluster.]

Walzberg et al. - 2023 - Agent‐based modeling and simulation for the circular economy Lessons learned and path forward

Abstract Circular economy aims at decoupling human activities from resource use and creating wealth. However, many have questioned the link between increased circularity and sustainability, resulting in several methodological approaches being developed to answer that question. This article analyzes and discusses the insights gained from applying agent‐based modeling and simulation to study the techno‐economic and social conditions promoting circularity and sustainability. This article analyzes the benefits and limitations of this technology and discusses future methodology developments within the circular economy context. Moreover, six limits of the circular economy concept are used to interpret insights from the literature: thermodynamic limits, system boundary limits, limits posed by the physical scale of the economy, limits posed by path dependencies and lock‐in, limits of governance and management, and limits of social and cultural definitions. Promising research avenues are to use this methodology with machine learning, industrial ecology methods, and detailed geographic information.

Bibliographic info
Walzberg, J., Frayret, J.-M., Eberle, A. L., Carpenter, A., & Heath, G. (2023). Agent-based modeling and simulation for the circular economy: Lessons learned and path forward. Journal of Industrial Ecology, 27(5), 1227–1238. https://doi.org/10.1111/jiec.13423

Scope & system studied

• Sector: Cross-sectoral, synthesizing CE applications in PV, wind turbines, HDDs, municipal waste, industrial symbiosis, consumer product-service systems.
• Geography: Global scope, with case studies mostly from U.S. contexts (PV, wind, HDD) but generalizable to CE modeling.
• CE strategy: Broad review of CE strategies (R0–R9 framework: refuse, rethink, reduce, reuse, repair, refurbish, remanufacture, repurpose, recycle, recover).

Model type & structure (ODD 2020 headings)

• Purpose and patterns: Reflect on lessons from ABM applied to CE systems, analyze barriers and enablers, and propose methodological directions.
• Entities, state variables, and scales: Synthesizes diverse ABM case studies (asset owners, recyclers, manufacturers, regulators, households, etc.).
• Process overview and scheduling: Focuses on CE transitions, analyzing ABM’s ability to model social norms, costs, logistics, and market evolution.
• Design concepts: Heterogeneity, social interactions, emergent dynamics, lock-in/path dependency, cultural definitions of waste.
• Initialization & input data: Case studies use survey data, industry reports, techno-economic analysis, GIS, LCA.
• Submodels: CE ABM framework integrating techno-economic, behavioral (TPB, social norms, trust), and environmental indicators.

Policy / strategy levers examined

• Recycling costs, landfill fees, subsidies, warranties for reused products.
• Trust in data wiping (HDDs).
• Logistics costs (wind blade transport).
• Social norm interventions for households and organizations.

Key findings

• ABM captures CE barriers: high recycling costs, path dependency, lock-in, lack of trust, low awareness.
• Enablers: reduced costs, stronger warranties for reused products, regulation (landfill bans, higher fees), early adopters and norm formation.
• CE transitions path-dependent; same interventions yield different outcomes depending on timing.
• ABM highlights Korhonen et al.’s six CE limits (thermodynamic, boundary, scale, lock-in, governance, cultural definitions). Demonstrates ABM can explore at least three: thermodynamic limits, path dependency/lock-in, cultural definitions.
• Combining ABM with LCA/IO, GIS, and ML enhances capacity to capture externalities and explore policy scenarios.

Data & calibration
Case studies combine: IDC/DOE data (HDDs), IRENA and industry reports (PV, wind), surveys, SEM models, and LCA databases. Validation includes historical data matching, stakeholder engagement, sensitivity analysis, Monte Carlo.

Relevance for toolkit development

• Offers a conceptual CE ABM framework useful for DYNAMO: techno-economic + behavioral + environmental integration.
• Emphasizes the need for coupling with IO/LCA, GIS, and ML—directly relevant to Syke’s ENVIMAT integration.
• Provides justification for modular design of CE toolkits addressing multiple R-strategies and social-technical interactions.
• Highlights methodological pitfalls (data dependency, validation difficulty, computational load) and suggests mitigation strategies.

Strengths & limitations

• Strengths: Integrative; broad review + detailed case studies; explicitly links ABM to CE theoretical limits; methodological roadmap.
• Limitations: Case studies are U.S.-centric (PV, wind, HDD); focus on selected CE strategies; limited empirical calibration outside U.S.
• Transferability: High conceptual/methodological relevance for EU/Finland; requires local data adaptation.

Classification for review
Include in main review.

Cluster membership
Cluster 9: Systematic Reviews & Frameworks for CE ABMs.

[Post-processing note: Lessons-learned synthesis with case studies (PV, HDD, wind); connects ABM to Korhonen’s CE limits; methodological roadmap emphasizing hybridization with LCA, GIS, ML; co-anchor for meta-level cluster alongside Rizzati & Landoni 2024 and Tian et al. 2025.]

Yang et al. - 2011 - Agent-based simulation of economic sustainability in waste-to-material recovery

Abstract This paper presents an agent-based approach to the economic sustainability evaluation of adopting competitive recycling technologies in waste-to-material recovery. The emphasis is put on the development of economic sustainability metrics for this evaluation and on the software implementation of an agent system to evaluate sustainability by using these metrics as agent simulation models. A set of model-driven sustainability metrics were defined to quantify the economic impact of technical innovations in waste management. A software agent was designed and implemented for simulating process scenarios to consistently evaluate the economic sustainability of different waste-to-material recovery systems. Using the sustainability metrics and the agent developed, a case study was conducted. Two recovery processes were evaluated and compared for their economic sustainability, measured in unit cost, share of cost of materials, process efficiency, in-process recycling rate and economic efficiency.

Bibliographic info
Yang, Q. Z., Sheng, Y. Z., & Shen, Z. Q. (2011). Agent-based simulation of economic sustainability in waste-to-material recovery. In 2011 IEEE 15th International Conference on Computer Supported Cooperative Work in Design (pp. 1150–1155). IEEE. https://doi.org/10.1109/CSCWD.2011.5960179

Scope & system studied

• Sector: Industrial waste-to-metal recovery processes.
• Geography: Singapore (industrial recycling context).
• CE strategy: Assessing economic sustainability of closed-loop vs open-loop recovery technologies using sustainability metrics.

Model type & structure (ODD 2020 headings)

• Purpose and patterns: Evaluate economic sustainability of different recycling technologies, focusing on unit cost, efficiency, and market dynamics.
• Entities, state variables, and scales:
  • Agents: Economic sustainability assessment agent, resource supply agent, recycler, market data agent, human participants.
  • Variables: Unit cost, share of material cost, process efficiency, in-process recycling rate, production revenue/cost.
  • Scale: Pilot-scale facility modeled under various process and market conditions.
• Process overview and scheduling:
  • Agents simulate recovery process scenarios (open-loop vs closed-loop).
  • Sustainability metrics (unit cost, efficiency, in-process recycling, material share) are evaluated per scenario.
  • Market prices for recovered metals vary by year, influencing profitability.
• Design concepts:
  • Emergence: Economic efficiency outcomes from interacting cost, process, and market factors.
  • Adaptation: Monte Carlo simulation for uncertainty in market conditions.
  • Interaction: Agents exchange data on process parameters, market prices, and sustainability metrics.
  • Stochasticity: Incorporated through Monte Carlo uncertainty analysis.
• Initialization: Process inputs (acid, water, feedstock, energy use) from pilot-scale recovery facility.
• Input data & calibration:
  • Experimental lab data for process efficiency and material consumption.
  • London Metal Exchange (LME) metal prices (2006–2011).
  • Sustainability criteria from OECD, GBEP, EU projects.
• Decision modules: Cost breakdown structure, unit cost vs production volume analysis, deterministic economic analysis.

Policy / strategy levers examined

• Technology choice: closed-loop vs open-loop processes.
• Market scenarios: varying LME metal prices.
• Process efficiency improvements.
• In-process recycling rates of acid and water.

Key findings

• Closed-loop recovery achieves ~18% higher economic efficiency than open-loop by reusing acids and water, lowering unit costs.
• Market prices strongly affect profitability: efficiency values ranged from 1.65 (profitable, 2007) to <1 (loss, 2009).
• Material cost share reduced by 12–13% in closed-loop vs open-loop.
• Process efficiency and in-process recycling rates are key drivers of economic sustainability.
• Demonstrates how agent-based metrics can identify improvement levers under uncertain markets.

Relevance for toolkit development

• Provides a sustainability-metrics-driven ABM for economic assessment of recycling technologies.
• Directly relevant for integrating profitability calculators with environmental/market variability.
• Offers template for modeling process scenarios under uncertainty (Monte Carlo).
• Transferable to Finnish/EU industrial CE contexts (metal recovery, battery recycling).

Strengths & limitations

• Strengths: Clear sustainability metrics; integration of process, cost, and market data; Monte Carlo for uncertainty; application to real recovery facility data.
• Limitations: Narrow focus (waste-to-metal recovery only); limited environmental/social assessment; stylized single case.
• Transferability: High methodological relevance; requires local process and market data for EU adaptation.

Classification for review
Include in main review.

Cluster membership
Cluster 3: Industrial Symbiosis & Secondary Material Markets.

[Post-processing note: Early ABM of economic sustainability metrics; compares closed- vs open-loop waste-to-metal recovery; adds methodologically strong, metrics-driven study to industrial symbiosis cluster.]

Koide et al. - 2025 - Prospective life cycle and circularity assessment of circular business models using an empirically grounded agent-based model

Abstract Despite the need for methodologies that support early-phase decision-making in the transition to a circular economy, current sustainability assessments often lack a prospective method that dynamically accounts for consumer decision-making based on empirical evidence. This study addresses this need by evaluating the circularity and environmental impacts of circular business models over a 30-year period, using an empirically grounded agent-based model coupled with life cycle assessment and material flow analysis. We developed a methodology to parameterize agents’ decision-making using data from demographically representative surveys and to prospectively assess the sustainability impacts of circular strategies. The case study examines the reuse, refurbishment, and subscription models of refrigerators and laptops in Japan. Results from Morris Elementary effects method and scenario analyses revealed that manufacturer-led refurbishment could reduce emissions of the whole society by 10%–12% and extend product lifetimes by 30%–33%. In contrast, the subscription model shows minimal benefits, with improvements of only 0%–3%, primarily due to consumer preferences for new products. Our consequential approach extends beyond technical strategies to evaluate the effectiveness of strategies targeting consumer behavior, including pricing, advertisements, and improvements in repair and collection services. The findings highlight the need for combining synergistic circular and diffusion strategies and suggest the need for a reorientation of policy efforts from end-of-life material recovery to refurbishment, reuse, and repair, supported by intensive campaigns and substantial price reductions in circular offerings. The methodology presented here facilitates prospective, dynamic, and consequential assessments of circular economy strategies to enhance consumer acceptance and ensure sustainability gains.

Bibliographic info
Koide, R., Murakami, S., Yamamoto, H., Nansai, K., Quist, J., & Chappin, E. (2025). Prospective life cycle and circularity assessment of circular business models using an empirically grounded agent-based model. Journal of Industrial Ecology, 29(x), 1–15. https://doi.org/10.1111/jiec.70090

Scope & system studied

• Sector: Consumer durables—refrigerators and laptops.
• Geography: Japan (national market, representative surveys).
• CE strategy: Comparative assessment of reuse, refurbishment (manufacturer-led), and subscription service models.

Model type & structure (ODD 2020 headings)

• Purpose and patterns: Prospective, dynamic assessment of CE business models over 30 years; quantify effects on product lifetimes, market diffusion, and life cycle GHG emissions.
• Entities, state variables, and scales:
  • Agents: Households (survey-based heterogeneous attributes), products (durables), supply chain actors (manufacturers, service providers, reuse shops).
  • Variables: Consumer utilities, price sensitivity, advertising exposure, social influence, product obsolescence.
  • Scale: National-level (911 refrigerator users, 1023 laptop users sampled; mapped to agent population).
• Process overview and scheduling:
  • Consumers acquire, repair, or discard products using a bounded rationality decision framework (awareness → consideration → choice).
  • Social influences operate via compliance, conformity, homophily.
  • Products undergo absolute (failure) or relative (social/functional) obsolescence.
  • Circular and diffusion strategies applied (reuse, refurbish, subscription + pricing, advertising, repair/collection improvements).
• Design concepts:

  • Emergence: Diffusion of CBMs and resulting emissions reductions.

  • Adaptation: Consumer preferences evolve via social influence and incentives.

  • Interaction: Peer effects, advertising, firm offers.

  • Stochasticity: Random seeds, survey imputations, survival and hazard functions.

• Initialization: Survey data mapped directly to agents (~60,000 data points across 65 household-level variables).

• Input data & calibration:

  • Large-scale surveys: refrigerators (N=911, 2022) and laptops (N=1023, 2023).

  • Utility functions estimated by hierarchical Bayes conjoint analysis.

  • Obsolescence modeled via survival and hazard models.

  • LCA inventory data (IDEA 3.1, AIST), company data (Panasonic), literature values.

• Submodels: Consumer decision module, product obsolescence module, repair/refurbish/resale/subscription modules, LCA coupling, MFA for product circulation.

Policy / strategy levers examined

• Circular strategies: reuse, refurbishment, subscription (moderate vs ambitious ambition levels).

• Diffusion strategies: pricing (-30% to -70%), advertising frequency (30–70%), free repair warranties (3–10 years), collection improvements, scratch removal, functional upgrades.

• Sensitivity: energy efficiency improvements (1–5% annually), grid electricity decarbonization (4–8%).

Key findings

• Refurbishment is most effective: extends lifetimes by 30–33% (to ~17 years refrigerators, ~10 years laptops), reduces emissions by 10–12%.

• Reuse yields moderate gains: lifetime +13–18%, emissions -5–7%.

• Subscription achieves high diffusion (~30% market share) but minimal sustainability benefits (0–3% emission reduction, ~3% longer lifetimes), due to preference for brand-new products and cannibalization of reuse.

• Price reductions and advertising are decisive; ambitious scenarios require 70% lower circular product prices and 50–70% annual advertising coverage.

• Product circulation (reuse/refurbish rates) is key to reducing new manufacturing demand; use-phase efficiency losses do not negate benefits.

• Policy implication: reorient from end-of-life recycling to refurbishment, reuse, and repair, combined with aggressive diffusion strategies.

Relevance for toolkit development

• Provides a state-of-the-art empirically grounded ABM directly linking survey-based consumer heterogeneity to CE scenarios.

• Demonstrates coupling of ABM with LCA and MFA for prospective assessment.

• Offers methodological blueprint for integrating conjoint choice experiments and social influence modeling into CE toolkits.

• Highly relevant for Finnish/EU adaptation (e.g., white goods, ICT products, consumer electronics).

Strengths & limitations

• Strengths: Empirical grounding (representative surveys + conjoint analysis); rigorous sensitivity (Morris method); integration with LCA/MFA; prospective long-term simulation.

• Limitations: Japan-specific surveys; cross-sectional data; assumptions constant over 30 years; no modeling of informal exports/hibernation; single-product per household assumption.

• Transferability: High methodological relevance; requires EU/Finland survey data for calibration.

Classification for review
Include in main review.

Cluster membership
Cluster 7: Consumer Behavior & Product-Service Strategies. Cross-links to Cluster 1 (Macro vs Meso CE Models) through integrated LCA/MFA coupling.

[Post-processing note: Empirically grounded ABM of consumer durables; surveys + conjoint + LCA/MFA; refurbishment > reuse > subscription in GHG reduction; cutting-edge methodological reference for CE ABMs.]

Roci et al. - 2022 - Towards circular manufacturing systems implementation A complex adaptive systems perspective

Abstract A transition towards circular manufacturing systems (CMS) has brought awareness of untapped economic and environmental benefits for the manufacturing industry. Conventional manufacturing systems already present a high level of complexity in terms of physical flows of materials and products as well as information and financial flows linked to them. Closing the loop of materials and products through multiple lifecycles, as proposed in CMS, increases this complexity manifold. To support practitioners in implementing CMS through enhanced decision-making, this research studies CMS from a complex adaptive systems (CAS) perspective and proposes to exploit methods and tools used in the study of CAS to characterise, model and analyse CMS. By viewing CMS as CAS composed of autonomous, interacting agents, this research proposes a multi-method model architecture for modelling and simulating CMS. The different CMS stakeholders are modelled individually as autonomous agents by integrating agent-based, discrete-event, and/or system dynamics modules within each agent to capture their diverse and heterogeneous nature. The applicability of the proposed multi-method approach is illustrated through a case study of a white goods manufacturing company implementing CMS in practice. This case study shows the relevance and feasibility of the proposed multi-method approach as a decision support tool for the systemic exploration and quantification of CMS. It also shows how a transition towards CMS necessitates a lifecycle approach in terms of costs, revenues and environmental impacts to identify hotspots and, therefore, design circular systems that are viable in both economic and environmental terms. In fact, the analyses of the simulation results indicate how decisions in terms of business models, product design, and supply chain might affect the CMS performance of the case company. For instance, implementing a service-based model led to a high number of usecycles (on average six usecycles per washing machine), which, in turn, led to high lifecycle costs and emissions due to more frequent transportation and recovery operations. Similarly, the deployment of long-lasting washing machines, which is a core principle of CMS, led to high manufacturing costs. Due to the high initial costs and a time mismatch between revenues and costs in the service-based model, it required a longer time for the company to reach the break-even point (approximately 23 months). Overall, the case study shows that multi-method simulation modelling can provide decision-making support for a successful implementation of CMS.

Bibliographic info
Roci, M., Salehi, N., Amir, S., Shoaib-ul-Hasan, S., Asif, F. M. A., Mihelič, A., & Rashid, A. (2022). Towards circular manufacturing systems implementation: A complex adaptive systems perspective using modelling and simulation as a quantitative analysis tool. Sustainable Production and Consumption, 31, 97–112. https://doi.org/10.1016/j.spc.2022.02.015

Scope & system studied

• Sector: Manufacturing systems.
• Geography: General/unspecified (conceptual framework, case context not tied to a specific country).
• CE strategy: Transition from linear to circular manufacturing, focusing on system-level dynamics, feedbacks, and resilience.

Model type & structure (ODD 2020 headings)

• Purpose and patterns: Provide a framework to analyze the transition of manufacturing systems towards circularity as complex adaptive systems; test resilience, adaptability, and sustainability outcomes.
• Entities, state variables, and scales:
  • Entities: Manufacturing firms, supply chain actors, consumers, regulators.
  • State variables: Resource flows, waste outputs, adoption status of circular practices, performance metrics.
  • Scales: System-level, multi-sector; temporal resolution not specified but intended for long-term transition analysis.
• Process overview and scheduling: System-level dynamics modeled through simulation of interactions among heterogeneous agents with adaptive decision rules; emphasis on emergent properties such as feedback loops, tipping points, and lock-ins.
• Design concepts:
  • Emergence: Circularity patterns from local decisions.
  • Adaptation: Firms adapt production and investment strategies.
  • Interaction: Supply chain exchanges, policy influence.
  • Stochasticity: Scenario uncertainty.
• Initialization: Generic manufacturing system with baseline linear practices.
• Input data & calibration: Largely conceptual; draws on literature and stylized system dynamics rather than empirical calibration.
• Decision modules: Firm adoption of circular practices, supply-demand balancing, regulatory influence, consumer preference shifts.

Policy / strategy levers examined

• Transition pathways (linear vs circular manufacturing).
• Role of regulatory pressure, incentives, and consumer demand.
• Structural interventions to enhance system adaptability.

Key findings

• Circular manufacturing systems can be fruitfully analyzed as complex adaptive systems, highlighting non-linear dynamics, feedback loops, and emergent resilience.
• Transition pathways are path-dependent: lock-ins and tipping points matter.
• Regulatory frameworks and consumer demand are crucial to push the system past thresholds toward circularity.
• Simulation tools provide insights into how firms’ micro-decisions accumulate into system-wide circularity levels.

Relevance for toolkit development

• Conceptual foundation for integrating ABM/SD into CE manufacturing analyses.
• Highlights the need for policy stress-testing under deep uncertainty—directly relevant for Finnish/EU contexts where industrial symbiosis and manufacturing decarbonization are central.
• Could inform design of policy analysis simulators focusing on resilience, adaptability, and system transitions.

Strengths & limitations

• Strengths: Strong conceptual framing (complex adaptive systems perspective); useful for theory-building; highlights non-linearities and feedbacks.
• Limitations: Limited empirical grounding; not calibrated to a specific sector or geography; more conceptual than applied; no explicit integration with IO/LCA.
• Transferability: High methodological relevance, but practical application in Finland/EU would require sector-specific calibration (e.g., pulp/paper, electronics).

Classification for review
Include in main review.

Cluster membership
Potential anchor for Cluster 5: Investment & CE Business Models and cross-linked with Cluster 6: Multi-Method Integration (ABM/SD + complex adaptive systems framing).

[Post-processing note: Conceptual ABM/SD hybrid for circular manufacturing transitions; complex adaptive systems framing; path dependence and tipping points; methodological anchor despite limited empirical calibration.]

REDO?: Roci and Rashid - 2023 - Economic and environmental impact of circular business models A case study of White Goods-as-a-Service

NOTE: This review does not conform to the review format, but OTOH it isn’t an agent-based model

Abstract Circular business models are gaining traction in academia and industry as an instrument to deliver environmental benefits while being economically profitable. Despite this increased interest, studies that quantitatively assess the impact of circular business models are limited. This study provides researchers and practitioners with a quantitative analysis tool to assess the dynamics of circular business models from both an economic and environmental perspective. Using the case study of white goods-as-a-service, this study employs multi-method simulation modelling in combination with statistical design and analysis of simulation experiments to investigate the effect of different factors including payment schemes (i.e., fixed fee, pay-per-use, and hybrid) and subscription contract duration (i.e., long-term, mid-term, and short-term) on the economic and environmental performance of access-based models. In addition, as the adoption of access-based models is economically challenging for manufacturers due to a discrepancy between costs and revenue streams, this study analyses the effect of different levers to improve the liquidity performance of access-based business models including deposit schemes, cancellation and collection fees, as well as partnering with financial institutions to cover the initial revenue gap.

I’ve gone through Roci & Rashid (2023). Here’s a structured summary for your literature review spreadsheet:

Roci, M., & Rashid, A. (2023)

Title: Economic and environmental impact of circular business models: A case study of White Goods-as-a-Service using multi-method simulation modelling
Journal: Journal of Cleaner Production, 407, 137147
DOI: 10.1016/j.jclepro.2023.137147

Purpose

To quantitatively assess both economic and environmental impacts of circular business models (CBMs), focusing on access-based product-service systems (PSS) for white goods (washing machines).

Methods

• Multi-method simulation modelling:

  • Agent-Based (ABM): user behaviour, subscription termination, stochastic washing habits.

  • Discrete-Event (DE): logistics, installation, repair, refurbishing, recycling flows.

  • System Dynamics (SD): continuous revenue/emission flows, loan repayment.

• Software: AnyLogic (Java-based).

• Data:

  • EU ReCiPSS project (washing machine demonstrator).

  • Workshops, interviews, field visits, and literature-derived parameters.

• Experiments:

  • Simulation optimization (OptQuest) for pricing strategies.

  • Full factorial design (3×3: business model × contract duration).

  • 90 replications with ANOVA + Tukey tests for significance.

Business Model Configurations

1. Fixed fee (subscription with unlimited washing).

2. Pay-per-use (PPU) (per wash, price varies by temperature).

3. Hybrid (fixed fee + per-wash unit fee).

• Contract durations: short-term (6 mo), mid-term (24 mo), long-term (60 mo).

• Strategies tested: deposit fees, cancellation/collection fees, loans.

Key Findings

• Economic vs Environmental Trade-offs:

  • Fixed fee → highest revenues (+27% vs PPU) but worst environmental performance.

  • PPU → best environmental outcome (−19% emissions vs fixed fee) but lowest revenues.

  • Hybrid → compromise: −10% revenues vs fixed fee, but −13% emissions.

• Contract duration:

  • Long-term contracts reduce lifecycle costs (−63%) and emissions (−13%) compared to short-term.

• Break-even points:

  • Best: fixed fee + long-term (≈30 months).

  • Worst: PPU + short-term (≈97 months).

• Customer perspective:

  • PPU cheapest (€18.45/month vs €25.25 fixed fee).

• Liquidity challenge:

  • Deposit schemes + fees help but insufficient for short-term PPU.

  • Loans eliminate initial revenue gap without hurting profitability.

Contribution

• Provides a quantitative decision-support tool for companies considering CBMs.

• Shows hybrid + long-term contracts are optimal trade-offs.

• Highlights transition barriers: cash flow mismatch, service costs, user acceptance.

• Demonstrates value of multi-method simulation + design of experiments in CE research.

Limitations

• Based on assumptions and EU washing machine case → external validity limited.

• Only tested 2 factors (business model, contract length); product design/supply chains not varied.

• User behaviour calibrated from limited empirical data (e.g., triangular distributions).

Davis et al. - 2009 - Integration of Life Cycle Assessment Into Agent-Based Modeling

Abstract A method is presented that allows for a life cycle assessment (LCA) to provide environmental information on an energy infrastructure system while it evolves. Energy conversion facilities are represented in an agent-based model (ABM) as distinct instances of technologies with owners capable of making decisions based on economic and environmental information. This simulation setup allows us to explore the dynamics of assembly, disassembly, and use of these systems, which typically span decades, and to analyze the effect of using LCA information in decision making. We were able to integrate a simplified LCA into an ABM by aligning and connecting the data structures that represent the energy infrastructure and the supply chains from source to sink. By using an appropriate database containing life cycle inventory (LCI) information and by solving the scaling factors for the technology matrix, we computed the contribution to global warming in terms of carbon dioxide (CO2) equivalents in the form of a single impact indicator for each instance of technology at each discrete simulation step. These LCAs may then serve to show each agent the impact of its activities at a global level, as indicated by its contribution to climate change. Similar to economic indicators, the LCA indicators may be fed back to the simulated decision making in the ABM to emulate the use of environmental information while the system evolves. A proof of concept was developed that is illustrated for a simplified LCA and ABM used to generate and simulate the evolution of a bioelectricity infrastructure system.

Bibliographic info
Davis, C., Nikolić, I., & Dijkema, G. P. J. (2009). Integration of life cycle assessment into agent-based modeling: Toward informed decisions on evolving infrastructure systems. Journal of Industrial Ecology, 13(2), 306–325. https://doi.org/10.1111/j.1530-9290.2009.00122.x

Scope & system studied

• Sector: Bioelectricity production (biomass-based electricity supply chains).
• Geography: Netherlands.
• CE strategy: Integration of LCA into dynamic simulation of bioelectricity infrastructures, enabling evaluation of environmental and economic sustainability of evolving supply chains.

Model type & structure (ODD 2020 headings)

• Purpose and patterns: Develop a proof-of-concept method to integrate LCA within ABM for evolving infrastructure systems, allowing agents’ decisions to reflect both profit and environmental performance.
• Entities, state variables, and scales:
  • Agents: Technology owners (power plants, biofuel producers), WorldMarket, Environment.
  • Variables: Operational configurations (feedstock mix, costs, emissions), profits, CO₂-equivalent emissions (from LCA).
  • Scale: National bioelectricity system, multi-decade evolution.
• Process overview and scheduling:
  • Agents trade inputs and outputs.
  • At each time step: contracts formed → technology/emissions matrices constructed → LCA computed → results fed back to agents for decision-making.
• Design concepts:
  • Emergence: Supply chain structures self-assemble via trading.
  • Adaptation: Agents adjust operational configurations based on profit or LCA score.
  • Interaction: Trade contracts form evolving supply networks.
  • Stochasticity: Agent addition/removal, trading partner selection.
• Initialization: Starts with fossil-based power plants cofiring biomass; demand set slightly above capacity to enable entry of new producers.
• Input data & calibration:
  • Dutch bioelectricity LCI data (Van der Voet et al. 2008).
  • Inventory for biomass (soy, palm oil, syngas, bio-oil), transport, combustion.
  • Simplified LCA indicators: CO₂, CH₄, N₂O → global warming potential (GWP).
• Submodels: Investment module (profit-based), operational decision module (profit vs LCA-minimizing), carbon tax module.

Policy / strategy levers examined

• Carbon dioxide tax levels.

• Decision rule type (profit-maximizing vs LCA-minimizing agents).

Key findings

• High CO₂ taxes are more effective at reducing emissions than simply switching feedstocks based on LCA minimization.

• Large fossil-based plants with limited biomass cofiring capacity constrained overall emissions reductions.

• Path dependency: dominance of large producers limited smaller biobased entrants.

• Structural uncertainty: outcomes varied across runs due to randomness in trading/entry, showing multiple possible system attractors.

• Demonstrated feasibility of integrating ABM and LCA for dynamic environmental assessment.

Relevance for toolkit development

• Provides methodological foundation for ABM–LCA coupling, a key direction for CE toolkit integration (e.g., linking ENVIMAT with ABMs).

• Illustrates how agent-level decisions can incorporate environmental indicators alongside profit.

• Offers approach for uncertainty analysis by running multiple simulations and deriving distributions of LCA scores.

• Directly relevant for EU/Finland in modeling dynamic energy and CE infrastructures.

Strengths & limitations

• Strengths: Clear demonstration of ABM–LCA integration; proof-of-concept with real data; emphasizes dynamic supply chain assembly.

• Limitations: Simplified LCA (only climate impacts); basic agent decision rules (no memory, prediction, innovation); Netherlands-specific bioelectricity case.

• Transferability: Methodologically strong; applicable across CE markets if coupled with IO/LCA databases and local data.

Classification for review
Include in main review.

Cluster membership
Cluster 6: Multi-Method Integration (ABM + LCA/IO). Cross-links to Cluster 1 (Macro vs Meso CE Models) due to system-level infrastructure focus.

[Post-processing note: Landmark ABM–LCA integration; Netherlands bioelectricity proof-of-concept; carbon tax vs decision-rule effects; methodological anchor for toolkit hybridization.]

REDO?: Micolier et al. - 2019 - To what extent can agent-based modelling enhance a life cycle assessment Answers based on a literature

Abstract Life cycle assessment (LCA) has proven its worth in modelling the entire value chain associated with the production of goods and services. However, modelling the consumption system, such as the use phase of a product, remains challenging due to uncertainties in the socio-economic context. Agent-based models (ABMs) can reduce these uncertainties by improving the consumption system modelling in LCA. So far, no systematic study is available on how ABM can contribute towards a behavior-driven modelling in LCA. This paper aims at filing this gap by reviewing all papers coupling both tools. A focus is carried out on 18 case studies which are analysed according to criteria derived from the four phases of LCA international standards. Criteria specific to agent-based models and the coupling of both tools, such as the type and degree of coupling, have also been selected. The results show that ABMs have been coupled to LCA in order to model foreground systems with too many uncertainties arising from a behaviour-driven use phase, local variabilities, emerging technologies, to explore scenarios and to support consequential modelling. Foreground inventory data have been mainly collected from ABM at the use phase. From this review, we identified the potential benefits from ABM at each LCA phase: (i) scenario exploration, (ii) foreground inventory data collection, (iii) temporal and/or spatial dynamics simulation, and (iv) data interpretation and communication. Besides, methodological guidance is provided on how to choose the type and degree of coupling during the goal and scope phase. Finally, challenging LCA areas of research that could benefit from the agent-based approach to include behaviour-driven dynamics at the inventory and impact assessment phase have been identified.

Micolier et al. (2019) is a literature review on coupling ABM and LCA, rather than a single case study.

It identifies 18 case studies where ABM was coupled with LCA, and organizes them by:

• LCA phase enhanced (goal & scope, inventory, impact assessment, interpretation)
• Coupling type (integration, hybrid, complementary) and degree (soft, tight, hard)
• Purposes: policy analysis, scenario exploration, modeling emerging technologies, or capturing behavior-driven heterogeneity.
• Opportunities:
  • Scenario exploration in goal & scope
  • Behavioral/heterogeneity data for inventory
  • Potential to add temporal/spatial dynamics to impact assessment
  • Better communication and sensitivity/uncertainty analysis in interpretation.
• Future directions: regionalization, dissipative use (e.g. plastic leakage), and indoor pollution as LCA domains where ABM could add value.

This makes the paper a conceptual/methodological bridge — synthesizing prior work, defining coupling strategies, and providing guidance (e.g. when to use soft vs hard coupling).

For your clustering:

• It belongs to the “ABM + LCA methodological integration” cluster, together with Davis et al. (2009) and some of the works it reviews (Attallah, Walzberg, etc.).
• It also overlaps with the review/meta-literature cluster (like Rizzati & Landoni 2024, Walzberg 2023).

Do you want me to map which of the 18 case studies in Micolier et al. overlap with the other papers you’ve already processed in this review pipeline (e.g. Davis 2009, Walzberg, Xu et al., etc.)? That way you’d see the cross-links more clearly.

You said:

let’s mark this for later so that once we’ve reviewed enough papers, we can return to it.

ChatGPT said:

Perfect — I’ll mark Micolier et al. (2019) for a later cross-mapping task once we have a critical mass of reviewed papers. That way we can build a clean network of overlaps (which papers cite/review each other, which belong to multiple clusters).

For now, we’ll treat it as:

Classification for review
Include in main review.

Cluster membership
Cluster 6: Multi-Method Integration (ABM + LCA/IO).

[Post-processing note: Literature review of 18 ABM–LCA couplings; categorizes coupling strategies and phases of LCA; methodological bridge paper. Marked for later cross-mapping to identify overlaps with Davis 2009, Walzberg 2023, Xu 2009, etc. NOTE: Ask for the PDF again!]

Xu et al. - 2009 - A Dynamic Agent-Based Analysis for the Environmental Impacts of Conventional and Novel Book Retailing

Abstract The dynamics of an e-commerce market and the associated environmental impacts from a bottom-up perspective using an agent-based model is explored. A conceptual meta-theory from psychology is adopted to form the behavioral rules of artificial consumers choosing different methods of buying a book, including conventional bookstores, e-commerce, and a proposed self-pick-up option. Given the energy and emissions savings that result from a shift to e-commerce from bookstore purchase, it appears that reductions in environmental impacts are relatively probable. Additionally, our results suggest that the shift to e-commerce is mainly due to the growth of Internet users, which ties energy and emissions savings to Internet penetration. Moreover, under any scenario, the energy and emissions savings will be provided by the introduction of the proposed self-pick-up option. Our model thus provides insights into market behaviors and related environmental impacts of the growing use of e-commerce systems at the retail level, and provides a basis for the development and implementation of more sustainable policies and practices.

Bibliographic info
Xu, M., Allenby, B., Kim, J., & Kahhat, R. (2009). A dynamic agent-based analysis for the environmental impacts of conventional and novel book retailing. Environmental Science & Technology, 43(8), 2851–2857. https://doi.org/10.1021/es802219m

Scope & system studied

• Sector: Retail logistics (book retail).
• Geography: U.S. book retail market.
• CE strategy: Compare conventional bookstores, e-commerce, and a proposed “self-pick-up” hybrid system for environmental impacts (transport-related emissions).

Model type & structure (ODD 2020 headings)

• Purpose and patterns: Explore how consumer heterogeneity, social influence, and Internet penetration drive shifts in retail modes and their environmental impacts.
• Entities, state variables, and scales:
  • Agents: Consumers (heterogeneous by preferences for price vs convenience, Internet access, social network connections).
  • Variables: Price/convenience satisfaction, social satisfaction, adoption thresholds, Internet access growth.
  • Scale: National U.S. retail market, 1998–2020.
• Process overview and scheduling:
  • Consumers decide at each time step whether to buy from bookstore, online (delivery), or self-pick-up (if available).
  • Decision-making combines individual satisfaction (price, convenience) + social satisfaction (peer effects in Watts–Strogatz small-world networks).
  • Four cognitive processes: repetition, deliberation, imitation, social comparison.
• Design concepts:
  • Emergence: Market shares of different retail modes.
  • Adaptation: Shifts in consumer preferences with Internet adoption.
  • Interaction: Peer effects and leadership in social networks.
  • Stochasticity: Randomness in network connections and adoption thresholds.
• Initialization: Calibrated to U.S. book retail data (1998–2005 market shares).
• Input data & calibration: Historical U.S. Census data on Internet use and book sales; literature on price differentials (Brynjolfsson & Smith 2000). Parameters (price/convenience dimensions, preference heterogeneity) fitted to reproduce observed market share trajectories.
• Submodels: Consumer decision rules; environmental impact model (fuel use, CO₂ per km/book).

Policy / strategy levers examined

• Internet penetration rate (1–12% annual growth).
• Introduction of “self-pick-up” option (price vs convenience advantages varied).

Key findings

• E-commerce grows to ~37% market share at equilibrium as Internet adoption saturates.
• Shift to e-commerce reduces transport-related fuel use and CO₂ by ~0.28% annually (2005–2010), totaling ~20 Mt CO₂ avoided.
• Self-pick-up option, if cheaper or more convenient than delivery, captures 6–23% market share and further reduces emissions by up to 12.4% per book.
• Consumer heterogeneity and social networks are critical: homogeneous agents fail to replicate observed data.
• Lock-in effects observed: bookstores retain majority share due to consumer cognitive and social inertia, even when alternatives offer benefits.

Relevance for toolkit development

• Strong demonstration of consumer behavior + logistics + environmental impact ABM.
• Useful template for CE market simulations where retail/logistics shifts matter (e.g., reuse collection schemes, return logistics in EU).
• Shows how social influence and lock-in can constrain CE adoption, even with economic/environmental advantages.
• Provides method for linking ABM with transport-based emissions factors (proto-ABM-LCA hybrid).

Strengths & limitations

• Strengths: Empirical calibration with historical data; rich behavioral rules; integration of social networks; explicit environmental outcome quantification.
• Limitations: Narrow environmental scope (only local transport emissions); U.S.-specific; does not include e-books/tech shocks; simple representation of convenience.
• Transferability: High methodological relevance for EU CE applications; concept adaptable to any consumer goods logistics/return systems.

Classification for review
Include in main review.

Cluster membership
Cluster 7: Consumer Behavior & Product-Service Strategies (consumer adoption and PSS logistics). Cross-links to Cluster 6 (ABM–LCA hybridization) due to emissions modeling.

[Post-processing note: ABM of U.S. book retail market; integrates consumer heterogeneity, social networks, and logistics emissions; introduces “self-pick-up” hybrid option; early exemplar of consumer-logistics-environment ABM coupling.]

Kumazawa and Kobayashi - 2006 - A simulation system to support the establishment of circulated business

Abstract In a traditional business, although products are supplied to consumers, essentially services or functions are supplied. The supply of services or functions to consumers does not necessarily involve the sale of a product. In such a situation, it is possible to reduce environmental loads by reusing a product without the consumer owning a product. In this study, we developed a life cycle simulation system for establishing a business in which a product is reused. By using this simulation system, environmental and economic evaluation based on material flow and cash flow can be executed. In this paper, a reuse business is defined as a business model in which products are collected and reused, and a sales business is defined as a current business model in which products are sold and not recovered. For supporting an actual business, an evaluation model of mixed business in a transition state from a sales business to a reuse business in which reused products are launched into different market segments is implemented. The usefulness of the developed system is shown by the application example of reusing personal computers.

Bibliographic info
Kumazawa, T., & Kobayashi, H. (2006). A simulation system to support the establishment of circulated business. Advanced Engineering Informatics, 20(2), 127–136. https://doi.org/10.1016/j.aei.2005.11.004

Scope & system studied

• Sector: Consumer electronics (PCs as application example).
• Geography: Japan (Toshiba case, conceptual tool development).
• CE strategy: Transition from sales-based to reuse-based (rental/lease) business models, supporting reuse across multiple market segments.

Model type & structure (ODD 2020 headings)

• Purpose and patterns: Develop a Life Cycle Simulation (LCS) system to evaluate economic and environmental performance of “circulated businesses” (reuse-based models) during transition from traditional sales.
• Entities, state variables, and scales:
  • Entities: Products (PCs), components, materials, market segments (IT firms, heavy users, light users, schools).
  • State variables: Product mass, useful lifetime, value lifetime, cost, environmental burden, rental fee, sales price.
  • Scale: Market-level simulation over 7 years; mixed sales and reuse business.
• Process overview and scheduling:
  • Define product model (product–component–material hierarchy).
  • Define process model (manufacture, assembly, transport, usage, collection, reuse, recycle).
  • Define supply/collect plan (sales → reuse transition timing, product flow between markets).
  • Run discrete-event simulation to compute costs, environmental loads, profits, recycling rates, inventories.
• Design concepts:
  • Emergence: Profitability and emissions reductions from interactions of product flow and business model.
  • Adaptation: Transition scenarios from sales to reuse, altering supply rules across markets.
  • Interaction: Products move between market segments depending on remaining useful/value lifetimes.
  • Stochasticity: Limited; largely deterministic planning system.
• Initialization: Notebook PC as example product, with sales prices, rental fees, useful/value lifetimes specified.
• Input data & calibration: Product/component/material attributes from Toshiba/Easy-LCA databases; assumed reuse process parameters (no empirical reuse data available).
• Decision modules:
  • Supply and collect planning for mixed business.
  • Product discrimination (reuse vs component reuse vs scrap) based on useful/value lifetime thresholds.

Policy / strategy levers examined

• Transition timing and patterns (all-at-once vs staggered by market segment).
• Product supply rules (priority of supplying reused goods to different market segments).
• Rental vs sales pricing schemes.

Key findings

• Reuse business scenarios outperform sales: ~20% higher profits, ~40% lower CO₂, ~85% less waste.
• Transition scenarios: Early, simultaneous transitions reduce emissions/waste more, but staggered transitions improve long-term profits.
• Product supply rules matter: targeted allocation reduces inventory buildup and improves reuse efficiency.
• Rental/lease systems critical to facilitate intentional product collection and reuse.

Relevance for toolkit development

• Provides early simulation tool (LCSimulator) integrating LCA and economic indicators for CE business models.
• Demonstrates evaluation of transition states (sales → reuse), not just static cases.
• Methodologically relevant for Finnish/EU CE toolkit: could inspire modules analyzing PSS/rental models and transition dynamics.
• Offers insights for profitability calculators (rental vs sales) and IO/LCA-coupled CE market simulators.

Strengths & limitations

• Strengths: Integrates material flows, costs, and environmental burdens; explicit handling of transition dynamics; multiple market segments.
• Limitations: Case-specific (PCs in Japan, Toshiba context); deterministic assumptions; limited empirical reuse data; excludes consumer acceptance factors.
• Transferability: Methodologically relevant for EU contexts (e.g., appliance leasing/PSS) with recalibration.

Classification for review
Include in main review.

Cluster membership
Cluster 7: Consumer Behavior & Product-Service Strategies (PSS/rental models). Cross-links to Cluster 6 (Multi-Method Integration) via LCS–LCA coupling.

[Post-processing note: Early Japanese simulation of sales→reuse business transitions; integrates LCA + profit flows; application to PCs; methodological precursor to later PSS ABMs (Koide, Lieder, Roci).]

Raihanian Mashhadi et al. - 2019 - Investigation of consumer’s acceptance of product-service-systems A case study of cell phone leasing

Abstract Product-Service-Systems have been introduced as a practical way to implement the Circular Economy concept. Consumer acceptance of such systems has been a barrier to their widespread implementation. However, it is not clear what contributes to the consumers’ likelihood of acceptance of product service systems. This article investigates consumers’ acceptance of leasing cell phones instead of buying them through conducting a survey analysis. A regularized logistic regression model has been developed to construct the consumers’ decision model. The decision model then has been used to develop an Agent-Based Simulation (ABS) framework in order to model the effects of social influences, previous decisions, and heterogeneous personal traits on consumers’ decisions to lease. The results of the study suggest that a segregation exists within the consumer attitudes toward leasing. Meaning that the consumers who are currently leasing their cell phones are more prone to lease again in the future and the consumers who own their current cell phone are more prone to also buy their future cell phones.

Bibliographic info
Raihanian Mashhadi, A., Vedantam, A., & Behdad, S. (2019). Investigation of consumer’s acceptance of product-service-systems: A case study of cell phone leasing. Resources, Conservation & Recycling, 143, 36–44. https://doi.org/10.1016/j.resconrec.2018.12.006

Scope & system studied

• Sector: Consumer electronics (cell phones).
• Geography: U.S. (survey participants at University at Buffalo).
• CE strategy: Use-oriented Product-Service-Systems (PSS) through leasing schemes for mobile phones.

Model type & structure (ODD 2020 headings)

• Purpose and patterns: Identify determinants of consumer acceptance of leasing vs. buying, and simulate adoption dynamics under PSS business models.
• Entities, state variables, and scales:
  • Agents: Consumers (heterogeneous socio-demographics, attitudes, prior phone disposal decisions).
  • Variables: Lease cost, peer pressure, environmental attitudes, prior buy/lease decisions, age, education, phone usage intensity.
  • Scale: 500 simulated consumers, daily decision processes across product generations.
• Process overview and scheduling:
  • Agents initialized with survey-derived traits.
  • At new product release (yearly), agents decide lease/buy.
  • Decisions shaped by logistic regression model + social influence module.
  • End-of-use decisions (sell, trade, recycle, store, return if leased) influence future buy/lease decisions.
• Design concepts:
  • Emergence: Market shares of leasing vs. buying.
  • Adaptation: Agents’ decisions influenced by past choices and social influence.
  • Interaction: Peer pressure from 1–6 linked agents; imitation of social network.
  • Stochasticity: Randomized initial traits and peer connections.
• Initialization: Based on online survey of 110 participants (students and employees at University at Buffalo).
• Input data & calibration:
  • Survey responses (lease vs buy, socio-demographics, environmental attitudes, peer influence, prior disposal decisions).
  • Elastic net logistic regression model used to estimate decision coefficients (accuracy: 90%).
• Decision modules: Logistic regression decision function; peer pressure ratio; dynamic leasing/buying cost calculation; annual product launch trigger.

Policy / strategy levers examined

• Lease contract terms and cost ($200–$400 vs $400–$600).
• Social influence (peer pressure from friends/family).
• Product price sensitivity ($600–$1100).

Key findings

• Only ~4% of consumers chose leasing in baseline simulation.
• Strong path dependence: consumers who leased once overwhelmingly continued leasing; prior ownership strongly predicted continued ownership.
• Peer pressure had weaker influence than expected; previous decision history dominated.
• Leasing attractive when costs fall into $200–$400 band.
• Market appears segmented: one group strongly favors ownership, another consistently leases.
• Environmental attitudes had minor effect on adoption decisions.

Relevance for toolkit development

• Demonstrates empirical survey + logistic regression calibration feeding into ABM—a replicable methodology for CE market simulators.
• Highlights path dependency and lock-in dynamics in consumer markets, relevant for EU policies on PSS adoption.
• Provides transferable decision-model structure for profitability calculators (e.g., service vs ownership).
• Useful empirical basis for modelling leasing schemes in Finnish/EU electronics CE strategies.

Strengths & limitations

• Strengths: Empirical survey + calibrated ABM; clear integration of behavioral, socio-demographic, and social influence variables; highlights market segmentation.
• Limitations: Small, skewed survey sample (mostly students, U.S. only); limited external validity; does not consider brand loyalty, carrier plans, or second-hand markets.
• Transferability: Methodology highly relevant; requires larger-scale EU/Finland surveys for calibration.

Classification for review
Include in main review.

Cluster membership
Cluster 7: Consumer Behavior & Product-Service Strategies.

[Post-processing note: ABM of cell phone leasing acceptance; survey-calibrated logistic regression + social influence; shows path dependency and consumer segmentation; complements Kumazawa 2006, Koide 2025, Roci 2023 in PSS adoption modelling.]

REDO? André et al. - 2019 - Resource and environmental impacts of using second-hand laptop computers A case study of commercial reuse

Abstract The circular economy is proposed to reduce environmental impact, but as yet, there is limited empirical evidence of this sort from studying real, commercial circular economy business cases. This study investigates the environmental impacts of using second-hand laptops, mediated by a commercial reuse operation, instead of new ones. The method used is life cycle assessment (LCA) and special attention is given to laptops’ metal resource use by using several complementary life cycle impact assessment methods. The results show that all activities required to enable reuse of laptops are negligible, despite the reuse company’s large geographical scope. Two principal features of reuse reduce environmental impacts. Firstly, use extension reduces all impacts considerably since there are large embedded impacts in components. Secondly, the reuse company steers non-reusable laptops into state-of-the-art recycling. This provides additional impact reductions, especially with regards to toxicity and metal resource use. The results for metal resource use however diverge between LCIA methods in terms of highlighted metals which, in turn, affects the degree of impact reduction. LCIA methods that characterise functionally recycled metals as important, result in larger impact reduction, since these emphasise the merits of steering flows into state-of-the-art recycling. The study thus demonstrates how using second-hand laptops, mediated by a commercial reuse operation, compared to new ones, in practice, reduces different types of environmental impact through synergistic relationships between reuse and recycling. Moreover, it illustrates how the choice of LCIA method can influence interpretations of metal resource use impacts when applying circular economy measures to information and communication technologies (ICT).

Got it. Here’s a structured summary of André et al. (2019) for our review pipeline:

André, H., Ljunggren Söderman, M., & Nordelöf, A. (2019)

Resource and environmental impacts of using second-hand laptop computers: A case study of commercial reuse
Waste Management, 88, 268–279

Aim

To assess the real-world environmental impacts of commercial ICT reuse (focusing on laptops) compared to using new ones, with a strong emphasis on metal resource use.

Methods

• Approach: Life Cycle Assessment (LCA), with detailed laptop disassembly + Ecoinvent database.

• Functional Unit: “One year of access to a laptop computer.”

• Scenarios:

  • New laptop alternative (3 years of first use, 50% collection rate at EoL).

  • Second-hand laptop alternative (70% of laptops reused for another 3 years, higher EoL collection rate via reuse company).

• Key Features Analyzed:

  1. Use extension → reduces need for new production.

  2. Steering flows into recycling → more laptops reach WEEE recycling.

• Impact Categories: Climate change, human toxicity, and especially metal resource use.

• Metal Resource Use: Assessed with five LCIA methods (CExD, CML, EcoSc, EPS, ReCiPe) to capture different problem framings.

Results

• Reuse activities negligible: transport, testing, data erasure, etc.

• Impact reductions:

  • Overall 39–50% lower impacts for reused laptops.

  • Use extension alone: ~41% reduction across all categories.

  • Recycling via company: additional 1–9% reduction, strongest for metals and toxicity.

• Component dominance: PCBs and casing were the largest contributors.

• Metal findings:

  • Gold, copper, tin, silver, palladium = most significant, especially for toxicity & resource depletion.

  • Indium and tantalum also important but not functionally recycled → only use extension helps here.

  • Choice of LCIA method strongly affects conclusions (e.g. gold critical in some, less so in others).

• Sensitivity analysis: Benefits of reuse hold across variations in lifetime assumptions.

Key Insights

• Two main mechanisms of benefit:

  1. Extended lifetimes delay need for new production.

  2. Higher controlled collection rates improve recycling outcomes.

• Policy relevance: Supports commercial reuse schemes as effective CE strategies, especially for ICT.

• Methodological lesson: The choice of LCIA method critically shapes interpretation of resource use benefits. Using multiple methods is essential.

• Limits: Results not generalizable to all ICT reuse cases; depends on high-grade devices, efficient logistics, and controlled recycling.

Relevance for DYNAMO

• Provides empirical validation of CE benefits in ICT sector.

• Shows synergies between reuse and recycling, an aspect we can model explicitly in ABM/SD frameworks.

• Highlights metal criticality and LCIA-method dependence → important for integrating resource depletion indicators into CE models.

• Offers real-world business case data (Inrego AB in Sweden), useful for benchmarking profitability/impact models.

Boldoczki et al. - 2020 - The environmental impacts of preparation for reuse A case study of WEEE reuse in Germany

Abstract According to the European waste management hierarchy, preparation for reuse (PfR) is preferable to recycling. From an environmental point of view, reuse is beneficial, if the impacts that arise during a certain usage duration of a reused product are smaller than those of a new product. If this is not the case, reuse is not beneficial to recycling. This study explores potential benefits of PfR compared to other waste management options for four white goods (washing machine, refrigerator, range, freezer) and four small electric devices (PC, printer, monitor, laptop) by the use of Life Cycle Assessment. These eight devices account for 68% by weight of all the collected waste electrical and electronic equipment (WEEE) in Germany. The results show, that the assumption that reuse is preferable to recycling does not apply to every case. Especially the impact categories of global warming, water consumption and cumulative energy demand are strongly dominated by the use phase of white goods, therefore a reuse of inefficient devices should be avoided. The results show that a reuse of products with an European energy efficiency rating of D and C is not recommended for any of the analyzed products. For small electric devices, the use phase is less dominant in comparison to the production, therefore reuse leads to significant saving potentials in almost all impact categories. A comparison of energy efficiency classes allows for product-specific decisions, whereas the assessment approach based on average devices yields for generalizable recommendations. Therefore, the results support environmentally conscious consumer decisions about the acquisition of a new versus a second-hand product and enable End-of-Life decision making in terms of the separation of reusable devices at collection points.

Bibliographic info
Boldoczki, S., Thorenz, A., & Tuma, A. (2020). The environmental impacts of preparation for reuse: A case study of WEEE reuse in Germany. Journal of Cleaner Production, 252, 119736. https://doi.org/10.1016/j.jclepro.2019.119736

Scope & system studied

• Sector: Waste electrical and electronic equipment (WEEE).
• Geography: Germany (Bavarian municipal collection points, national electricity mix).
• CE strategy: Preparation for reuse (PfR) vs recycling of white goods (washing machine, refrigerator, freezer, range) and small devices (PC, printer, monitor, laptop).

Model type & structure (ODD 2020 headings)

• Purpose: Test whether PfR always yields environmental benefits over recycling/new replacement, considering device efficiency and lifespan.
• Entities: Products categorized into white goods and small electricals. Attributes: weight, energy efficiency class, electricity and water use, lifespan.
• Processes: LCA-based comparative modelling of new vs reused devices. Production, repair/cleaning (PfR), transport, and use phase included; recycling excluded (assumed same but time-shifted).
• Data: Inventory from Ecoinvent and EU preparatory studies; German energy mix; Weibull-distributed product lifetimes; market shares of efficiency classes; repair/transport scenarios.
• Submodels:
  • Calculation of saving potential = difference in lifecycle impacts (new vs reused).
  • Break-even analysis → maximum usage duration for which reuse remains beneficial.
  • Monte Carlo simulations (10,000 runs) to test sensitivity (energy demand, lifetimes, electricity mix).

Policy / strategy levers examined

• Energy efficiency rating (A+++ to D).
• Usage duration (lifetime extension vs early replacement).
• Electricity mix (2017 baseline vs 2050 near-100% renewables).
• Consumer/user behavior intensity (limited role).

Key findings

• White goods: Use phase dominates impacts; inefficient devices (C/D class) should not be reused. Even B-class reuse often environmentally inferior to new.
• Small devices: Production dominates impacts; reuse almost always beneficial (PC, laptop, monitor, often printer).
• Impact category dependence: PfR always favorable for mineral resource scarcity, but not consistently for global warming, cumulative energy demand, or water use.
• Scenario analysis: Green electricity mix improves PfR performance for GWP, toxicity, CED, but reduces relative advantage for resource scarcity.
• Robustness: Laptops and monitors show stable benefits; printers and PCs more sensitive to assumptions.
• General conclusion: Reuse is not universally preferable—device- and efficiency-specific assessments are necessary.

Data & calibration

• Empirical device data from Bavarian collection points (Messmann et al. 2019).
• Product specifications from EU preparatory studies, WRAP, Ecoinvent.
• Weibull lifetime distributions calibrated with Dutch/German stock data.
• Monte Carlo simulations ensure robustness testing.

Relevance for toolkit development

• Methodologically valuable for modelling trade-offs between reuse and replacement in CE transitions.
• Highlights need for efficiency-aware reuse criteria (e.g., minimum rating thresholds).
• Provides framework for integrating LCA + statistical lifetimes into decision-support tools.
• Could be adapted in Finnish/EU CE toolkits to guide PfR targets and product-specific reuse recommendations.

Strengths & limitations

• Strengths: Rigorous LCA with sensitivity analysis; empirical data from German collection; differentiated by efficiency class; clear decision-support framing.
• Limitations: Excludes recycling time-shift impacts; Germany-specific energy mix may limit EU transferability; no behavioral/economic dimension (focus purely environmental).
• Integration with IO/LCA/EEIO: Fully LCA-based; could link with material flow analysis for national policy.

Classification for review
Include in main review.

Cluster membership
Cluster 8: LCA-based assessments of CE strategies (reuse vs recycling). Cross-links with Cluster 7 (consumer product-service/ICT) and Cluster 6 (multi-method integration).

[Post-processing note: Rigorous German PfR case; LCA with Weibull lifetimes and Monte Carlo; shows reuse not always preferable—efficiency thresholds matter; useful for toolkit modules to assess reuse policies in EU/FI.]

Baxter - 2019 - Systematic environmental assessment of end-of-life pathways for domestic refrigerators

Abstract Rigorous and fair comparison of end-of-life handling options for domestic refrigerators is much more complicated than simple policy perspectives such as the waste hierarchy might suggest. The drive towards product reuse over material recycling merits particularly close attention. This study explores the factors that might influence policy and decision-making in this area, with a view to minimising overall environmental impact. The work combines life-cycle assessments of refrigerators with models for product performance over time to give a sector-wide perspective on these issues. Generally-accepted tenets such as the preferability of product reuse are shown to be provisional, heavily dependent on factors such as the electricity mix, and non-uniform for different measures of impact. A more general and complicated picture emerges, and recommendations for existing policy and ongoing work are developed.

Bibliographic info
Baxter, J. (2019). Systematic environmental assessment of end-of-life pathways for domestic refrigerators. Journal of Cleaner Production, 208, 612–620. https://doi.org/10.1016/j.jclepro.2018.10.173

Scope & system studied

• Sector: Domestic refrigeration appliances.
• Geography: Europe (Norway and EU electricity contexts).
• CE strategy: Compare reuse/refurbishment vs recycling/new replacement for refrigerators, assessing multiple environmental indicators over product lifetimes.

Model type & structure (ODD 2020 headings)

• Purpose: Provide fair comparison between reuse and recycling pathways, questioning the assumption that reuse is always preferable.
• Entities: Refrigerators (Type 7 fridge-freezers). Attributes: weight, lifetime, energy efficiency at manufacture year.
• Processes: LCA covering manufacture, use, refurbishment, recycling. Disposal excluded (assumed equal). Sensitivity analyses on electricity mix, age at EoL, refurbishment burden, secondary lifetime, and technological improvement rate.
• Design concepts:
  • Emergence: Optimal pathways depend on context (electricity mix, equipment age, technology trends).
  • Adaptation: Scenarios explore changes in lifetime and refurbishment depth.
  • Interaction: Trade-off between manufacturing and use-phase burdens.
• Initialization: Base-case refrigerator (70 kg, 200 L fridge + 100 L freezer). Primary life = 12 years, secondary life = 3 years. New equipment consumes 240 kWh/yr (2014 baseline), with 3.2% annual efficiency gain.
• Input data & calibration: European Commission preparatory study (2016), Ecoinvent electricity mixes, literature on refurbishment (Boustani et al. 2010), SimaPro LCA software.

Policy / strategy levers examined

• Electricity mix (Norwegian hydropower-dominant vs EU average fossil-rich).
• Age at replacement (first life length, L1).
• Refurbishment burden (fraction fR of new manufacture).
• Rate of technological efficiency improvement (1–6% per year).
• Length of secondary life (L2).

Key findings

• Norwegian case (low-carbon electricity): Reuse preferable in almost all cases (12–15% lower GWP and other burdens than recycling).
• European case (fossil-heavy electricity): Recycling/new replacement often preferable (reuse higher burdens by ~3–5%).
• Break-even age: Beyond ~20 years (Norwegian) or ~8 years (European), extending life becomes environmentally negative.
• Refurbishment impacts: If refurbishment burdens >20% of manufacturing, reuse loses advantage.
• Technological progress: Faster efficiency improvements make recycling preferable sooner.
• General conclusion: “Reuse is better” vs “use phase dominates” can’t both be true for refrigerators; outcomes hinge on electricity mix and product age.

Relevance for toolkit development

• Strong cautionary example: CE policies promoting reuse must be context-specific (electricity mix, product type).
• Provides methodological template for break-even analyses combining LCA + lifetime/efficiency modeling.
• Relevant for Finnish/EU toolkit: can inform reuse guidelines (e.g. minimum efficiency thresholds, maximum age for PfR).
• Demonstrates how sensitivity analysis should be embedded in CE assessment tools.

Strengths & limitations

• Strengths: Rigorous LCA with sensitivity and temporal analyses; clear policy implications; addresses methodological assumptions directly.
• Limitations: Focused on one appliance type; assumes constant performance over lifetime; simplified refurbishment modeling; no economic or behavioral dimension.
• Transferability: Strong for Europe/Nordics; directly informs appliance PfR/reuse guidelines.

Classification for review
Include in main review.

Cluster membership
Cluster 8: LCA-based assessments of CE strategies (reuse vs recycling). Cross-links with Cluster 6 (multi-method integration) for sensitivity analysis approach.

[Post-processing note: Refrigerator reuse vs recycling LCA; context-dependence of reuse benefits; electricity mix and technology improvement as decisive levers; methodological exemplar for break-even analysis in CE toolkits.]

Glöser-Chahoud et al. - 2021 - The link between product service lifetime and GHG emissions A comparative study for different consumer products

Abstract The production, use, and final disposal of goods are directly linked to various environmental impacts caused along their supply chains and over their entire life cycles. When assessing these impacts for energy-consuming products such as consumer electronics, not only the emissions caused during production but also the energy consumption during the use phase need to be taken into account in order to provide a holistic view on environmental impacts. However, the interplay between a product’s lifetime, reduction of demand through higher durability, energy consumption, and related greenhouse gas (GHG) emissions cannot be generalized but requires very specific analyses, which take into account product-related aspects and their temporal changes as well as the (changing) properties of the energy and use system. This contribution provides a quantitative assessment of the interrelation between product lifetime and environmental impacts, particularly GHG emissions, using refrigerators and mobile phones as exemplary products with differing characteristics. Whereas in the case of refrigerators, the strongest impact is caused during the use phase because of high energy consumption and related emissions, mobile phones as representatives of classical consumer electronics have their highest environmental impact during production. To assess impacts for both product categories, two simulation models of product life cycles based on methods from dynamic material flow analysis (MFA) are linked with life cycle inventory (LCI) data and LCA results for the respective products, focusing on the impact category of global warming potential (GWP). By systematically evaluating different scenarios, we show major influences on the overall GHG emissions over a product’s lifetime capturing temporal developments and modifications within the target system at European scale. In the case of refrigerators, we show that there is a trend towards increasing optimum lifetimes and that current energy efficiency improvements of new devices do not justify early replacement of older devices and, hence, a reduction of service lifetime. This is also because the GHG emissions of electricity production have continuously decreased with an increasing share of renewable energy sources. Regarding mobile phones, we emphasize the counterproductive effect of unused storage time (hibernation) when taking efforts for increasing the service lifetimes aiming at a reduction of demand for new, resource-consuming devices.

Bibliographic info
Glöser-Chahoud, S., Pfaff, M., & Schultmann, F. (2021). The link between product service lifetime and GHG emissions: A comparative study for different consumer products. Journal of Industrial Ecology, 25(2), 465–478. https://doi.org/10.1111/jiec.13123

Scope & system studied

• Sector: Consumer electronics and household appliances.
• Geography: European Union (EU28).
• CE strategy: Product lifetime extension and its effect on greenhouse gas (GHG) emissions. Case studies: refrigerators (energy-intensive) and mobile phones (production-intensive).

Model type & structure (ODD 2020 headings)

• Purpose: Quantify the environmental impacts of lifetime changes (extension, early replacement, hibernation) for refrigerators and mobile phones.
• Entities: Product vintages (refrigerators by year of manufacture; mobile phones in use, hibernated, discarded).
• Processes: Dynamic stock-driven material flow analysis (MFA) + LCA-based emission factors. Cascading service/hibernation modeled for phones; efficiency vintages and decarbonizing electricity mix modeled for refrigerators.
• Design concepts:
  • Emergence: System-level GHG trajectories from product lifetimes.
  • Adaptation: Replacement heuristics for refrigerators; cascade use with/without hibernation for phones.
  • Stochasticity: Weibull lifetime distributions.
• Initialization: Historical sales and stock data (1990–2018) extended with constant future sales (to 2030).
• Input data & calibration: LCA studies for embodied emissions (Suckling & Lee 2015 for phones; JEMA 2014, Hollander & Roser 2015 for fridges); EU electricity carbon intensity data (Moro & Lonza 2018); lifetime distributions from literature.

Policy / strategy levers examined

• Average service lifetime (baseline 13 years; scenarios ±5 years for refrigerators; extended second/third lives for phones).
• Early replacement vs optimum replacement heuristic (based on energy efficiency gains vs embodied emissions).
• Reduction of mobile phone “hibernation” (unused storage).
• Decarbonization of electricity supply.

Key findings

• Refrigerators:
  • Early replacement yields minimal emission reductions; overall effect small because efficiency gains have slowed and electricity decarbonization reduces use-phase impacts.
  • Optimum lifetimes increasing (from ~8–9 years for 1990s vintages to ~15 years for 2010s).
  • Planned obsolescence or lifetime reduction counterproductive in systemic terms.
• Mobile phones:
  • Hibernation is a major problem: devices stored unused reduce potential for second-hand use and emission reduction.
  • Technical lifetime extension alone has limited benefits if consumer behavior unchanged.
  • Eliminating hibernation could substantially cut embodied emissions by lowering demand for new devices.
• General: Lifetime strategies must be product-specific (use-phase intensive vs production-intensive).

Relevance for toolkit development

• Provides dynamic MFA + LCA integration suitable for CE policy analysis.
• Useful methodological precedent for stock-driven models (keeping service constant while adjusting lifetimes).
• Can inform toolkit modules on lifetime thresholds, hibernation dynamics, and break-even analysis for product reuse/repair policies.
• Relevant for EU/Finnish contexts: appliance and ICT reuse strategies, linking lifetime extension with decarbonization trajectories.

Strengths & limitations

• Strengths: Combines stock-flow and LCA; systemic EU scale; clear contrast between product archetypes; addresses temporal changes (efficiency, electricity mix).
• Limitations: Average data only (not product-specific); only GWP considered (no toxicity, resource use); behavioral drivers simplified.
• Transferability: Strong for EU appliance/ICT policy design; requires richer behavioral/economic modeling for decision-support tools.

Classification for review
Include in main review.

Cluster membership
Cluster 8: LCA-based assessments of CE strategies (reuse vs recycling, lifetimes). Cross-links to Cluster 6 (multi-method integration) via MFA–LCA coupling.

[Post-processing note: European refrigerator vs mobile phone lifetime–GHG trade-offs; stock-driven MFA + LCA; shows hibernation as major barrier; optimum lifetimes increasing with decarbonization; policy-relevant for EU reuse/repair guidelines.]

Hischier and Böni - 2021 - Combining environmental and economic factors to evaluate the reuse of electrical and electronic equipment – a Swiss case study.

Abstract One key strategy which can be used to promote a Circular Economy is ‘reuse’. This is particularly relevant for Electrical and Electronic Equipment due to its often rather short use phase as well as its resource-intensive production phase. The present study aimed to investigate the environmental and economic relevance of promoting the reuse of (waste) electrical and electronic equipment in Switzerland. To do so, a simplified life cycle assessment approach was combined with a calculation of the total cost of ownership of a device. These calculations were made for five different types of device: washing machines, refrigerators, televisions, laptop computers, and smartphones. Results showed that from an environmental perspective, smartphones or laptop computers, whose dominant environmental impact comes in their production phase, should be reused independently of their age, whereas for the three other devices, age is a decisive factor. Adding on the economic factor—that reuse should result in lower costs—led to the conclusion that all older devices except for refrigerators would have to be ‘sold on’ at no cost in order for their reuse to make sense economically. In addition, there should be a consideration of whether buying second-hand equipment replaces a new device or results in an increase in the total stock of devices, as old and new ones are run in parallel, creating a typical rebound situation. Public authorities should thus be more active in sharing information and raising awareness about the possibilities for repair and reuse.

Bibliographic info
Hischier, R., & Böni, H. W. (2021). Combining environmental and economic factors to evaluate the reuse of electrical and electronic equipment – a Swiss case study. Resources, Conservation & Recycling, 166, 105307. https://doi.org/10.1016/j.resconrec.2020.105307

Scope & system studied

• Sector: Electrical and electronic equipment (EEE).
• Geography: Switzerland.
• CE strategy: Assess environmental and economic viability of reuse compared to recycling/new replacement for five device types: washing machines, refrigerators, televisions, laptops, smartphones.

Model type & structure (ODD 2020 headings)

• Purpose: Determine when reuse of second-hand EEE makes sense from both environmental and economic perspectives, and whether authorities should promote reuse.
• Entities: Devices (five categories, modeled with technical lifetimes and efficiency data).
• Processes:
  • Simplified LCA (production, use, EoL recycling).
  • Total cost of ownership (TCO) for consumers (purchase price + energy/water use).
• Design concepts:
  • Emergence: Threshold ages at which reuse shifts from beneficial to harmful.
  • Adaptation: Device-specific results depending on production vs use-phase dominance.
  • Interaction: Not modeled socially, but between environmental and cost outcomes.
• Initialization: Reference year 2016 devices (data from topten.ch and EU preparatory studies).
• Input data & calibration:
  • Environmental: Ecoinvent 3.3 datasets, ecological scarcity (UBP), GWP, cumulative energy demand.
  • Economic: Market prices, Swiss electricity/water costs, depreciation assumptions.
  • Technical lifetimes: 15 years (washing machines, refrigerators), 10 (TVs), 8 (laptops), 5 (smartphones).

Policy / strategy levers examined

• Device age at second-hand sale.
• Relative second-hand pricing (as % of new).
• Efficiency improvements in new devices (0–2% annual).
• Consumer behavior (continued use vs rebound where old + new run in parallel).

Key findings

• Smartphones & laptops: Always environmentally beneficial to reuse (production dominates). Economically viable only when relatively new; beyond 3–5 years, must be free/very cheap to be attractive.
• Refrigerators & washing machines: Environmentally beneficial up to ~9–10 years; beyond that, energy use cancels reuse benefits. Economically viable only if second-hand price is very low (<30–40% of new).
• Televisions: Environmental benefits up to ~8 years; economically unviable beyond ~6 years unless free.
• General: Reuse is not universally beneficial—depends strongly on product type and age. Rebound risk if reused devices don’t substitute new purchases.
• Policy implication: Public authorities should promote reuse selectively (ICT especially), inform consumers, and support repair/maintenance infrastructure.

Relevance for toolkit development

• Provides combined environmental + economic assessment framework (LCA + TCO).
• Can inform CE toolkits with device-specific age/price thresholds for viable reuse.
• Highlights importance of consumer rebound effects (parallel use).
• Strong applicability to Finnish/EU contexts for designing PfR/reuse criteria.

Strengths & limitations

• Strengths: Combined environmental + economic analysis; practical thresholds; use of Swiss data; broad device coverage.
• Limitations: Simplified LCA (Excel-based); limited categories (GWP, CED, UBP); assumes linear depreciation and constant performance; Switzerland-specific.
• Integration with IO/LCA/EEIO: Full LCA framework, no IO coupling.

Classification for review
Include in main review.

Cluster membership
Cluster 8: LCA-based assessments of CE strategies (reuse vs recycling, lifetimes).

[Post-processing note: Swiss case; combines LCA + TCO; device-specific thresholds (smartphones/laptops always environmentally beneficial; white goods not beyond 10 yrs); strong template for integrating economic factors into CE policy toolkits.]

Iraldo et al. - 2017 - Is product durability better for environment and for economic efficiency A comparative assessment applying LCA and LCC to two energy-intensive products

Abstract The importance of product durability is becoming increasingly recognised, as Europe strives to move towards a circular economy, where products are designed and manufactured in a way that helps save resources and minimize waste. However, till now, no agreement has been reached on the potential benefits that increased durability of products may yield. Especially for energy using products, there is an increasing debate on whether greater environmental and cost benefits are achieved through extending the lifetime of the product, or rather by replacing it with a new, more energy efficient model. The aim of this work is thus to investigate if and in which conditions extended durability of energy-intensive products is desirable from an environmental and an economic perspective. In order to assess the environmental impact of increased durability of products, Life Cycle Assessment (LCA) and Life Cycle Costing (LCC) have been applied to perform a comparative analysis between a durable and a standard option for two different products: domestic refrigerator and freezer and domestic electric oven. Several sensitivity analyses have been performed on purchase price, cost of electricity, energy efficiency. According to the results of this study, if from an economic point of view for the user more durable products are practically always the preferred choice, this is not always the case from the environmental point of view. Only when the production and end of life stages have very high impacts compared to the use phase the more durable option is preferred from both an economic and an environmental point of view.

Bibliographic info
Iraldo, F., Facheris, C., & Nucci, B. (2017). Is product durability better for environment and for economic efficiency? A comparative assessment applying LCA and LCC to two energy-intensive products. Journal of Cleaner Production, 140, 1353–1364. https://doi.org/10.1016/j.jclepro.2016.10.017

Scope & system studied

• Sector: Household appliances (domestic refrigerator/freezer and electric oven).
• Geography: Europe (data from EU preparatory studies and a major European manufacturer).
• CE strategy: Increasing product durability (extended lifetime) vs replacing with newer, more energy-efficient models.

Model type & structure (ODD 2020 headings)

• Purpose: Investigate environmental and economic trade-offs of product durability for energy-intensive appliances.
• Entities: Appliances with two options: standard (10-year lifetime) vs durable (15-year lifetime). Attributes: energy consumption, material composition, cost.
• Processes:
  • LCA (cradle-to-grave: production, use, EoL).
  • LCC (purchase price, electricity use, maintenance/repair).
• Design concepts:
  • Trade-offs between avoided production vs efficiency of replacements.
  • Sensitivity to energy efficiency improvements of new models.
• Initialization: Baseline: 10-year lifetime, durable extended to 15. Replacement (standard B) assumed 25–45% more energy efficient depending on data source.
• Input data & calibration: Ecoinvent 2.2, EU Ecodesign preparatory studies (2007, 2011), primary data from a major European manufacturer. Energy price sensitivity: €0.092–0.30/kWh. Discount rates 2–5%.

Policy / strategy levers examined

• Product lifetime (10 vs 15 years).
• Energy efficiency improvement rates of new products.
• Purchase price (±30%).
• Electricity price variation.

Key findings

• Environmental:
  • Durable refrigerators reduce impacts in categories linked to production/EoL (freshwater ecotoxicity, resource depletion).
  • For climate change and energy-related categories, even small (5–10%) efficiency improvements in new models outweigh durability benefits.
  • Ovens less sensitive, but thresholds still exist (~15–20% efficiency gains).
• Economic: Durable options nearly always cheaper (avoids need to buy second appliance). Only under low electricity prices and discounted purchase prices does the standard option become competitive.
• Integrated LCA–LCC: From user’s perspective, durability is almost always economically beneficial, but environmentally it depends on balance between production vs use-phase impacts.

Relevance for toolkit development

• Provides clear threshold methodology for when durability is beneficial.
• Demonstrates integrated LCA + LCC analysis, directly useful for CE policy and investment assessment tools.
• Offers a replicable framework for break-even analysis (efficiency improvement thresholds vs durability extension).
• Strongly relevant to Finnish/EU contexts for appliance regulation and repair/reuse promotion.

Strengths & limitations

• Strengths: Combines environmental and economic perspectives; uses both literature and industry data; includes extensive sensitivity analyses.
• Limitations: Narrow scope (two appliance types); simplified repair assumptions; electricity mix generalizations; no behavioral factors.
• Integration: Fully LCA + LCC; potential for linking with IO models for macro policy assessment.

Classification for review
Include in main review.

Cluster membership
Cluster 8: LCA-based assessments of CE strategies (reuse vs recycling, durability). Cross-links with Cluster 6 (multi-method integration, especially LCA + economic analysis).

[Post-processing note: EU appliance durability study; LCA + LCC; shows thresholds where efficiency gains cancel durability benefits; economically durable nearly always better; valuable threshold analysis framework for CE toolkits.]

Koide et al. - 2022 - Prioritising low-risk and high-potential circular economy strategies for decarbonisation: A meta-analysis on consumer-oriented product-service system

Abstract The circular economy has potential synergies with climate change mitigation, although quantitative evidence for these synergies in the consumer product sector has not yet been systematically evaluated. We conducted a systematic review of existing studies of consumer-oriented product-service systems (PSS) on the lifecycle impacts, focusing on a meta-analysis of global warming potentials. Potential reductions in greenhouse gas emissions and the risk of backfire as a result of 10 types of circular economy strategies were investigated by reviewing 103 studies and more than 1500 scenarios for consumer durable and semi-durable products. As many as three-quarters of the reviewed studies analyse only one circular economy strategy rather than comparing (e.g., reusing, sharing, and renting) or integrating multiple strategies (e.g., leasing combined with remanufacturing). Approximately one-third of the reviewed studies rely on assumptions for set parameters of use, transport, and end-of-life phases, whereas more than half of the studies do not consider service and infrastructure provisions, or imperfect product substitution. Among the reviewed case studies, upgrading, repair, refurbishing, and pooling showed moderate to high improvement potentials with lower risks of backfiring, whereas servitisation, sharing, and reuse were associated with higher backfire risks, but they had high improvement potentials. The risk of backfiring was explained by factors such as transport, number of uses, product substitution and lifetime, maintenance, energy source, and energy efficiency. This meta-analysis revealed the importance of prioritising low-risk strategies, controlling for rebound effects of high-potential strategies, and integrating multiple strategies in order to utilize PSS for climate change mitigation.

Bibliographic info
Koide, R., Murakami, S., & Nansai, K. (2022). Prioritising low-risk and high-potential circular economy strategies for decarbonisation: A meta-analysis on consumer-oriented product-service systems. Renewable and Sustainable Energy Reviews, 155, 111858. https://doi.org/10.1016/j.rser.2021.111858

Scope & system studied

• Sector: Consumer durable and semi-durable products (appliances, ICT, clothing, vehicles, books, tools).
• Geography: Global review; EU context highlighted.
• CE strategy: Consumer-oriented product-service systems (PSS) including reuse, repair, refurbishing, remanufacturing, upgrading, modularity, durability, leasing, renting, sharing, pooling, servitisation.

Model type & structure (ODD 2020 headings)

• Purpose: Assess climate mitigation potential of PSS strategies via meta-analysis of lifecycle GHG impacts and identify methodological shortcomings.
• Entities: Product-service categories across 103 studies, >1500 scenarios.
• Processes: Literature review, systematic coding, scenario-level meta-analysis.
• Design concepts:
  • Emergence: GHG outcomes from PSS introduction.
  • Adaptation: Strategy-specific impacts (use-, result-, product-oriented PSS).
  • Interaction: Rebound and backfire effects from transport, substitution, lifetimes, maintenance.
• Initialization: 103 studies screened; 65 included in quantitative synthesis
• Input data & calibration: Extracted from reviewed LCAs (full, IO, simplified). Standardized as relative improvement in GWP vs baseline.

Policy / strategy levers examined

• Ten CE strategies tested (repair, upgrading, refurbishing, remanufacturing, durability, reuse, leasing, renting, sharing, servitisation, pooling).
• Empirical data gaps noted: substitution assumptions, infrastructure/service provision, functional units.
• Comparison of single vs integrated strategies (rare in literature).

Key findings

• Low-risk, high-potential: Repair, upgrading, modularity, refurbishing, remanufacturing, pooling (moderate–high improvement, low backfire).
• High-potential but risky: Servitisation, sharing, reuse (large gains but high rebound/backfire risks).
• Low-impact: Durability (minimal benefit, often negative); leasing (limited improvement, low backfire); renting (some gains but high backfire risk).
• Major determinants of outcomes: transport logistics, number of uses, product lifetimes, substitution rates, maintenance, energy mix.
• Meta-analysis shows typical reductions ~10–70% GHG, far below radical expectations (factor 10).
• Research gaps: lack of multi-strategy studies, weak treatment of imperfect substitution, omission of infrastructure/service impacts, poor sensitivity analysis.

Relevance for toolkit development

• Provides systematic synthesis of which strategies are most promising or risky, guiding scenario design in CE market simulators.
• Offers methodological lessons: need for global sensitivity analysis, empirical data collection, careful definition of functional units.
• Relevant to Finland/EU policy analysis: helps target repair, upgrading, modularity, and refurbishing over risky PSS strategies unless rebound is addressed.

Strengths & limitations

• Strengths: Comprehensive coverage (103 studies, >1500 scenarios); quantitative synthesis; cross-strategy comparisons.
• Limitations: Publication bias; heterogeneity of assumptions; climate (GWP) focus only; excludes economic/social impacts.
• Transferability: High for EU contexts; emphasizes empirical data needs for local calibration.

Classification for review
Include in main review.

Cluster membership
Cluster 8: LCA-based CE strategy assessments; cross-links with Cluster 6 (multi-method integration) and policy-focused clusters.

[Post-processing note: Landmark systematic review + meta-analysis; ranks CE strategies by GHG potential vs backfire risks; identifies major methodological shortcomings; high utility for toolkit scenario framing and sensitivity analysis design.]

Kraines and Wallace - 2006 - Applying Agent-based Simulation in Industrial Ecology

Abstract The goal of this column is to clarify what an agent-based simulation is and to provide an industrial ecology context. We hope that this column will encourage researchers in industrial ecology to consider how agent-based modeling approaches might appropriately complement other modeling tools widely applied in the field, such as life-cycle assessment methods, systems dynamics, material flow analysis, and environmental analyses based on input-output tables.

Bibliographic info
Kraines, S., & Wallace, D. (2006). Applying agent-based simulation in industrial ecology. Journal of Industrial Ecology, 10(1–2), 15–18. https://doi.org/10.1162/108819806775545376

Scope & system studied

• Sector: Conceptual/methodological (industrial ecology in general).
• Geography: None (framework article).
• CE strategy: Argues for the value of agent-based simulation in analyzing industrial ecosystems, including construction, supply chains, and technology adoption.

Model type & structure (ODD 2020 headings)

• Purpose: Introduce ABM to industrial ecology audience, highlight relevance for simulating autonomous actors with incomplete knowledge, and encourage integration with LCA, MFA, SD, and IO models.
• Entities: Generic—companies, industries, consumers, logistics operators, regulators.
• Processes: Multiagent interactions in supply chains, negotiations, technology diffusion, project contracting.
• Design concepts:
  • Emergence: Market behaviors and technology adoption from decentralized interactions.
  • Adaptation: Firms adjust strategies under limited knowledge.
  • Interaction: Contracting, coordination, negotiation.
  • Stochasticity: Inherent in agent decision rules and networks.
• Initialization: Not a case study; cites applications (e.g., Ma & Nakamori 2005 on technological innovation; multiagent systems in construction).
• Input data & calibration: Not applied—conceptual exposition.

Policy / strategy levers examined

• No quantitative levers. Suggests ABM useful for exploring scenarios of technology diffusion, supply chain coordination, and consumer adoption under policy incentives.

Key findings

• ABM is well-suited for industrial ecology because it can represent autonomous, heterogeneous actors with limited knowledge.
• Particularly useful where social interactions, negotiations, and goal-seeking behavior matter.
• Best applied to scenario exploration and sensitivity analysis, not quantitative prediction of detailed real-world outcomes.
• Emphasizes integration: ABM should complement LCA, MFA, SD, and IO approaches.
• Cautions: ABM incurs costs in software complexity, and results must be interpreted carefully given qualitative/stochastic basis.

Relevance for toolkit development

• Strong conceptual justification for including ABM in the DYNAMO toolkit.
• Highlights ABM’s niche: emergent dynamics, heterogeneous decision-making, scenario testing.
• Reinforces idea that ABM should be linked with LCA, MFA, IO, as in Davis (2009) and later works.

Strengths & limitations

• Strengths: Clear articulation of ABM’s role in industrial ecology; links to other modelling traditions; early advocate within JIE community.
• Limitations: Purely conceptual; no empirical model; general guidance rather than specific results.
• Transferability: Methodological, broadly applicable to CE modelling in Finland/EU.

Classification for review
Background only (conceptual framing, not a model or applied case).

Cluster membership
Cluster ?: Methodological/meta-literature (ABM in IE/CE). Cross-links to Cluster 6 (multi-method integration).

[Post-processing note: Conceptual introduction of ABM to industrial ecology; no case study but strong methodological rationale; situates ABM as complementary to LCA, MFA, IO; background-only source.]

Sacchi et al. - 2022 - PRospective EnvironMental Impact asSEment (premise): A streamlined approach to producing databases for prospective life cycle assessment using integrated assessment models

Abstract

Bibliographic info
Sacchi, R., Terlouw, T., Siala, K., Dirnaichner, A., Bauer, C., Cox, B., Mutel, C., Daioglou, V., & Luderer, G. (2022). PRospective EnvironMental Impact asSEment (premise): A streamlined approach to producing databases for prospective life cycle assessment using integrated assessment models. Renewable and Sustainable Energy Reviews, 160, 112311. https://doi.org/10.1016/j.rser.2022.112311

Scope & system studied

• Sector: Cross-sectoral—energy-intensive industries (power, cement, steel, fuels, freight transport).
• Geography: Global (IAM scenarios with regional resolution).
• CE strategy: Prospective Life Cycle Assessment (pLCA) database generation to anticipate environmental impacts of future technologies and pathways.

Model type & structure (ODD 2020 headings)

• Purpose: Provide a streamlined, open-source tool (premise) to integrate Integrated Assessment Model (IAM) scenarios into life cycle inventory (LCI) databases (ecoinvent), enabling prospective LCAs.
• Entities: IAM scenarios (REMIND, IMAGE), LCI processes in sectors like power, steel, cement, fuels, road transport.
• Processes:
  • Mapping IAM scenarios to LCI activities.
  • Transformations applied to databases (regionalized markets, efficiency updates, CCS deployment, emerging tech insertion).
  • Export to LCA software.
• Design concepts:
  • Emergence of future product footprints from IAM-driven pathways.
  • Adaptation via different climate policies (RCP 6.0, 2.6, 1.9).
  • Interactions between sectors (electricity, fuels, materials).
• Initialization: Baseline ecoinvent v3.5–3.8; IAM SSP2 scenarios with climate targets.
• Input data & calibration: IAM outputs (energy, fuel mixes, efficiency, emissions factors), supplemented by LCA inventories for emerging tech (e.g., DACCS, synthetic fuels).

Policy / strategy levers examined

• Climate mitigation targets (RCP 6.0, 2.6, 1.9).
• Socioeconomic pathways (SSPs).
• Sectoral transformations (electricity only vs all sectors).

Key findings

• Stricter mitigation targets (RCP 1.9) lead to up to ~80% reduction in database-wide GHG intensities by 2050 vs static baseline.
• Sectoral coverage matters: including fuels, cement, steel, and transport adds major reductions compared to electricity-only adjustments.
• Case studies (road construction, DACCS, batteries) illustrate significant differences across scenarios and IAMs (REMIND vs IMAGE).
• Results reveal trade-offs: decarbonization pathways differ in non-climate impacts (land use, toxicity, ozone depletion).
• Tool enables consistent, reproducible pLCI databases across IAMs and scenarios.

Data & calibration

• IAMs: REMIND v2.1, IMAGE v3.2 (SSP2 baseline with RCP 6/2.6/1.9).
• LCI: ecoinvent v3.5–3.8.
• Supplementary inventories from LCA literature (e.g., hydrogen, synthetic fuels, DACCS).
• Exportable formats: Brightway2, Simapro, Activity Browser.

Relevance for toolkit development

• Provides ready-to-use prospective databases crucial for CE investment and policy analysis under future scenarios.
• Enables testing profitability and environmental performance of CE strategies under different decarbonization pathways.
• Highly relevant to Finland/EU modelling where prospective sectoral decarbonization impacts must be accounted for.
• Offers potential integration with ABM/SD/IO models in DYNAMO via standardized scenario-aligned LCA datasets.

Strengths & limitations

• Strengths: Open-source, transparent, reproducible; covers multiple IAMs/sectors; addresses IAM–LCA mismatch with mapping files.
• Limitations: Limited to 2050–2060 reliability; semantic mismatches between IAM and LCI resolution; toxicity indicators highly uncertain; emerging technology inventories sometimes inconsistent.
• Transferability: Strong for EU and Finland contexts; premise is actively maintained and extendable.

Classification for review
Include in main review.

Cluster membership
Cluster 6: Multi-method integration (ABM + LCA/IO). Cross-links to Cluster 8 (LCA-based CE strategy assessments).

[Post-processing note: Premise tool—systematic IAM–LCA integration; generates prospective LCI databases; highly relevant for CE toolkit integration and scenario modelling; landmark methodological contribution.]

Walzberg et al. - 2021 - Do We Need a New Sustainability Assessment Method for the Circular Economy A Critical Literature Review

Abstract The goal of the circular economy (CE) is to transition from today’s take-make-waste linear pattern of production and consumption to a circular system in which the societal value of products, materials, and resources is maximized over time. Yet circularity in and of itself does not ensure social, economic, and environmental performance (i.e., sustainability). Sustainability of CE strategies needs to be measured against their linear counterparts to identify and avoid strategies that increase circularity yet lead to unintended externalities. The state of the practice in quantitatively comparing sustainability impacts of circular to linear systems is one of experimentation with various extant methods developed in other fields and now applied here. While the proliferation of circularity metrics has received considerable attention, to-date, there is no critical review of the methods and combinations of methods that underlie those metrics and that specifically quantify sustainability impacts of circular strategies. Our critical review herein analyzes identified methods according to six criteria: temporal resolution, scope, data requirements, data granularity, capacity for measuring material efficiency potentials, and sustainability completeness. Results suggest that the industrial ecology and complex systems science fields could prove complementary when assessing the sustainability of the transition to a CE. Both fields include quantitative methods differing primarily with regard to their inclusion of temporal aspects and material efficiency potentials. Moreover, operations research methods such as multiple-criteria decision-making (MCDM) may alleviate the common contradictions which often exist between circularity metrics. This review concludes by suggesting guidelines for selecting quantitative methods most appropriate to a particular research question and making the argument that while there are a variety of existing methods, additional research is needed to combine existing methods and develop a more holistic approach for assessing sustainability impacts of CE strategies.

Bibliographic info
Walzberg, J., Lonca, G., Hanes, R. J., Eberle, A. L., Carpenter, A., & Heath, G. A. (2021). Do we need a new sustainability assessment method for the circular economy? A critical literature review. Frontiers in Sustainability, 1, 620047. https://doi.org/10.3389/frsus.2020.620047

Scope & system studied

• Sector: Cross-sectoral (review paper).
• Geography: Global scope.
• CE strategy: Critically reviews sustainability assessment methods applied to CE, focusing on whether novel methods are needed beyond LCA and related frameworks.

Model type & structure (ODD 2020 headings)

• Purpose: Identify strengths, limitations, and gaps in current assessment methods (LCA, LCC, MFA, IO, hybrid approaches) when applied to CE strategies; evaluate the case for CE-specific methods.
• Entities: Assessment frameworks rather than actors or firms.
• Processes: Systematic literature review and critical analysis of how methods capture CE-specific dynamics (loops, cascading, substitution, rebound).
• Design concepts:
  • Emergence: CE strategies create systemic shifts that current methods may miss.
  • Interaction: Between products, loops, and system boundaries.
  • Adaptation: Evolution of methods in response to CE.
• Initialization: Literature sample across sustainability assessment studies in CE.
• Input data & calibration: Secondary data from reviewed publications.

Policy / strategy levers examined

• Not tested directly; review identifies methodological blind spots relevant to policy design (e.g., rebound effects, multi-loop interactions, system boundaries).

Key findings

• Most CE assessment still relies on conventional LCA and MFA; few tailored CE methods exist.
• Current methods struggle with:
  • Loop complexity (reuse, refurbishment, remanufacture, recycling).
  • Substitution assumptions (imperfect substitution not well captured).
  • Rebound/backfire effects.
  • System boundaries (interlinked loops, product-service combinations).
• Critical insight: Rather than inventing entirely new CE-specific tools, existing methods should be adapted/extended with better dynamic, consequential, and multi-scale integration.
• Emphasizes need for multi-method combinations (ABM, IO, LCA, MFA).

Relevance for toolkit development

• Validates DYNAMO’s integrative approach (ABM + IO/LCA).
• Highlights where profitability calculators or CE market simulators must account for rebound, substitution, and multi-loop complexity.
• Suggests toolkit should emphasize flexible integration of existing methods instead of inventing new frameworks.

Strengths & limitations

• Strengths: Clear, critical review; identifies concrete gaps; policy-relevant framing.
• Limitations: No new empirical data; scope limited to methodological critique; some gaps (social impacts, governance tools) only briefly mentioned.
• Transferability: Global framing, but findings highly applicable to EU/Finland CE assessments.

Classification for review
Include in main review.

Cluster membership
Cluster ?: Methodological/meta-literature (sustainability assessment in CE). Cross-links to Cluster 6 (multi-method integration).

[Post-processing note: Critical review of CE assessment methods; argues for adapting LCA/IO/MFA with multi-method integration rather than inventing CE-specific tools; validates DYNAMO hybrid approach.]

Palazzo et al. - 2020 - A review of methods for characterizing the environmental consequences of actions in life cycle assesment

Abstract Understanding the environmental consequences of actions is becoming increasingly important in the field of industrial ecology in general, and in life cycle assessment (LCA) more specifically. However, a consensus on how to operationalize this idea has not been reached. A variety of methods have been proposed and applied to case studies that cover various aspects of consequential life cycle assessment (CLCA). Previous reviews of the topic have focused on the broad agenda of CLCA and how different modeling frameworks fit into its goals. However, explicit examination of the spectrum of methods and their application to the different facets of CLCA are lacking. Here, we provide a detailed review of methods that have been used to construct models of the environmental consequences of actions in CLCA. First, we cover the following structural modeling approaches: (a) economic equilibrium models, (b) system dynamics models, (c) technology choice models, and (d) agent-based models. We provide a detailed review of particular applications of each model in the CLCA domain. The advantages and disadvantages of each are discussed, and their relationships with CLCA are clarified. From this, we are able to map these models onto the established aspects of CLCA. We learn that structural models alone are not sufficient to quantify the uncertainty distributions of underlying parameters in CLCA, which are essential components of a robust analysis of consequences. To address this, we provide a brief introduction to a counterfactual-based causal inference approach to parameter identification and uncertainty analysis that is emerging in the CLCA literature. We recommend that one potential research path forward is the establishment of feedback loops between empirical estimates and structural models.

Bibliographic info
Palazzo, J., Geyer, R., & Suh, S. (2020). A review of methods for characterizing the environmental consequences of actions in life cycle assessment. Journal of Industrial Ecology, 24(4), 815–829. https://doi.org/10.1111/jiec.12983

Scope & system studied

• Sector: Cross-sectoral, methodological review.
• Geography: Global.
• CE strategy: Consequential Life Cycle Assessment (CLCA) methods, with relevance to recycling, material substitution, energy transitions, and technology adoption.

Model type & structure (ODD 2020 headings)

• Purpose: Critically review modeling approaches for assessing environmental consequences of actions in LCA; compare structural models (economic equilibrium, system dynamics, technology choice, agent-based models) and discuss integration with causal inference.
• Entities: Structural models (PEM, GEM, SD, TCM, ABM) and CLCA frameworks.
• Processes: Mapping how each model addresses CLCA aspects (system expansion, marginal/incremental technologies, rebound, innovation, social mechanisms, temporal effects).
• Design concepts: Focus on emergence (e.g., rebound effects, technology adoption), adaptation (policy shocks, innovation), and interaction (between markets, technologies, and agents).
• Initialization & data: Literature review of case studies (biofuels, recycling, energy systems, technology substitution, EV adoption).
• Submodels: Comparison of four structural approaches and introduction of causal inference methods (natural experiments, counterfactual analysis).

Policy / strategy levers examined

• Not directly tested; examples include CO₂ taxes, biofuel policies, EV adoption incentives, and recycling subsidies from reviewed studies.

Key findings

• Structural models have distinct strengths:
  • Economic equilibrium: good for macro-level system expansion and price effects.
  • System dynamics: captures temporal dynamics and delays.
  • Technology choice: optimization under constraints.
  • ABM: models social processes and heterogeneous agent behaviors.
• CLCA needs integration of structural models with causal inference for robust cause-effect analysis.
• No single model suffices; hybrid/multi-method approaches recommended.
• Future: iterative frameworks combining empirical causal inference with structural modeling to reduce uncertainty.

Data & calibration

• Secondary data from case studies (biofuels, recycling, mobility, electricity, agriculture).
• Parameters often based on elasticities, LCA inventories, or behavioral survey data in cited applications.

Relevance for toolkit development

• Provides a roadmap for combining ABM/SD with IO/LCA, validating DYNAMO’s hybrid strategy.
• Supports inclusion of rebound, substitution, and temporal dynamics in CE profitability and policy models.
• Encourages empirical calibration (natural experiments, causal inference) for more realistic scenario testing.

Strengths & limitations

• Strengths: Comprehensive comparative review; clear articulation of model strengths/weaknesses; integration with causal inference novel and forward-looking.
• Limitations: No new empirical applications; methodological; may be abstract for applied CE modeling.
• Transferability: Highly relevant to EU/Finnish CE policy contexts due to focus on system expansion, rebound, and technology adoption.

Classification for review
Include in main review.

Cluster membership
Cluster 8: Methodological/meta-literature (sustainability assessment and CLCA methods). Cross-links to Cluster 6 (multi-method integration for CE).

[Post-processing note: Comparative review of CLCA methods (PEM, GEM, SD, TCM, ABM) and causal inference; argues for multi-method integration to capture system expansion, rebound, and behavioral dynamics; validates DYNAMO’s toolkit direction.]

Querini and Benetto - 2015 - Combining Agent-Based Modeling and Life Cycle Assessment for the Evaluation of Mobility Policies

Abstract This article presents agent-based modeling (ABM) as a novel approach for consequential life cycle assessment (C-LCA) of large scale policies, more specifically mobility-related policies. The approach is validated at the Luxembourgish level (as a first case study). The agent-based model simulates the car market (sales, use, and dismantling) of the population of users in the period 2013–2020, following the implementation of different mobility policies and available electric vehicles. The resulting changes in the car fleet composition as well as the hourly uses of the vehicles are then used to derive consistent LCA results, representing the consequences of the policies. Policies will have significant environmental consequences: when using ReCiPe2008, we observe a decrease of global warming, fossil depletion, acidification, ozone depletion, and photochemical ozone formation and an increase of metal depletion, ionizing radiations, marine eutrophication, and particulate matter formation. The study clearly shows that the extrapolation of LCA results for the circulating fleet at national scale following the introduction of the policies from the LCAs of single vehicles by simple up-scaling (using hypothetical deployment scenarios) would be flawed. The inventory has to be directly conducted at full scale and to this aim, ABM is indeed a promising approach, as it allows identifying and quantifying emerging effects while modeling the Life Cycle Inventory of vehicles at microscale through the concept of agents.

Bibliographic info
Querini, F., & Benetto, E. (2015). Combining agent-based modeling and life cycle assessment for the evaluation of mobility policies. Environmental Science & Technology, 49(3), 1744–1751. https://doi.org/10.1021/es5060868

Scope & system studied

• Sector: Passenger car market, with focus on electric vehicle (EV) adoption.
• Geography: Luxembourg and Lorraine (France).
• CE strategy: Deployment of electric and plug-in hybrid vehicles as part of mobility decarbonization policies.

Model type & structure (ODD 2020 headings)

• Purpose: Assess environmental consequences of EV deployment policies by coupling ABM with consequential LCA (C-LCA).
• Entities: Agents = car users/households; vehicles (ICEV, PHEV, BEV); markets for new/used cars; charging infrastructure.
• Processes:
  • Agents make car purchase, usage, resale, and dismantling decisions
  • Vehicle characteristics (segment, weight, consumption, battery) affect LCA impacts.
  • Hourly travel behavior and charging modeled.
• Design concepts:
  • Emergence: Fleet composition evolves from individual decisions.
  • Adaptation: Agents respond to policies, incentives, infrastructure.
  • Interaction: Word-of-mouth, resale market, cross-border commuting.
  • Stochasticity: Vehicle attributes, lifetimes, owner turnover.
• Initialization: Simulation runs 2013–2020, all vehicles sold included. Four scenarios: no EV, BAU (limited infra, incentives phased out), current policies (infrastructure expansion, mixed incentives), additional policies (extended incentives).
• Input data & calibration:
  • Vehicle inventories: ecoinvent 2.2, EU JRC datasets, manufacturer LCAs.
  • Emissions: COPERT IV (Artemis cycles), Fastsim consumption models.
  • Policies: Luxembourg/Lorraine incentives, charging infra rollout.
  • Environmental impacts: ReCiPe 2008 (GWP, fossil depletion, acidification, eutrophication, PM, ionizing radiation, metal depletion, etc.).

Policy / strategy levers examined

• EV purchase subsidies (Luxembourg CAR-e, Lorraine incentives).
• Charging infrastructure availability (home/workplace).
• Battery lifetime assumptions.
• Electricity mix (Luxembourg nuclear-heavy vs Germany coal-heavy).

Key findings

• ABM-LCA integration reveals that upscaling single-vehicle LCAs to fleet level is misleading—heterogeneity of segments, usage, resale, and charging strongly influence results.
• EV deployment yields mixed results: decreases in GWP, fossil depletion, ozone formation, and acidification; but increases in metal depletion, ionizing radiation (from nuclear electricity), PM, and eutrophication.
• BEVs mostly adopted as secondary cars, PHEVs as main cars (replacing larger ICEVs), thus PHEVs drive most of the observed reductions despite higher per-vehicle impacts.
• Incentives significantly affect adoption; maintaining subsidies increases EV penetration.
• Battery lifetime and electricity mix critical: extending battery lifetimes or ensuring low-carbon electricity improves outcomes.
• Cross-border commuting and resale markets complicate national-level policy impacts.

Relevance for toolkit development

• Demonstrates tight coupling of ABM with LCA to generate realistic life cycle inventories at national scale.
• Provides methodological precedent for using ABM to capture behavioral heterogeneity and rebound effects in CE policy models.
• Shows importance of country-specific infrastructure, incentives, and electricity mixes for CE scenario tools.
• Offers design template for combining agent decisions with system-scale environmental outcomes.

Strengths & limitations

• Strengths: First to operationalize ABM-LCA for national mobility policies; rich behavioral representation; strong empirical data calibration.
• Limitations: Limited to 2013–2020 horizon; excludes company fleets; high uncertainty in battery tech and future regulations; results not easily generalizable beyond study regions.
• Integration: Explicit ABM + LCA coupling; a clear case for hybrid modeling in CE policy analysis.

Classification for review
Include in main review.

Cluster membership
Cluster 6: Multi-method integration (ABM + LCA/IO). Also connects to mobility/transport CE policy applications.

[Post-processing note: Landmark case study of ABM-LCA coupling; shows fallacy of scaling single-vehicle LCAs; highlights rebound effects, electricity mix dependence, and policy uncertainty; directly relevant for CE toolkit design.]

Dandres et al. - 2012 - Macroanalysis of the economic and environmental impacts of a 2005–2025 European Union bioenergy policy

Abstract This paper describes a new tool to assess the medium- and long-term economic and environmental impacts of large-scale policies. The approach – macro life cycle assessment (M-LCA) – is based on life cycle assessment methodology and includes additional elements to model economic externalities and the temporal evolution of background parameters. The general equilibrium model GTAP was therefore used to simulate the economic consequences of policies in a dynamic framework representing the temporal evolution of macroeconomic and technological parameters. Environmental impacts, expressed via four indicators (human health, ecosystems, global warming and natural resources), are computed according to policy life cycle and its indirect economic consequences. In order to illustrate the approach, two 2005–2025 European Union (EU) energy policies were compared using M-LCA. The first policy, the bioenergy policy, aims to significantly increase energy generation from biomass and reduce EU energy demand for coal. The second policy, the baseline policy, is a business as usual policy where year 2000 energy policies are extended to 2025. Results show that, compared to the baseline policy, the bioenergy policy generates fewer impacts on three of the four environmental indicators (human health, global warming and natural resources) at the world and EU scales, though the results may differ significantly at a regional level. The results also highlight the key contribution of economic growth to the total environmental impacts computed for the 2005–2025 period. A comparison of the results with a more conventional consequential LCA approach illustrates the benefits of M-LCA when modeling the indirect environmental impacts of large-scale policies. The sensitivity and uncertainty analysis indicates that the method is quite robust. However, its robustness must still be evaluated based on the sensitivity and uncertainty of additional parameters, including the evolution of economic growth.

Bibliographic info
Dandres, T., Gaudreault, C., Tirado-Seco, P., & Samson, R. (2012). Macroanalysis of the economic and environmental impacts of a 2005–2025 European Union bioenergy policy using the GTAP model and life cycle assessment. Renewable and Sustainable Energy Reviews, 16, 1180–1192. https://doi.org/10.1016/j.rser.2011.11.003

Scope & system studied

• Sector: EU energy system, focusing on bioenergy as a substitute for coal.
• Geography: European Union (primary), with global spillovers via trade.
• CE strategy: Large-scale biomass deployment for electricity and heat generation, compared to business-as-usual.

Model type & structure (ODD 2020 headings)

• Purpose: Assess medium- to long-term economic and environmental impacts of EU bioenergy policy using a hybrid macro-LCA (M-LCA).
• Entities: Economic sectors (20 aggregated from 57 GTAP sectors) and regions (14 aggregated from 113 GTAP regions).
• Processes: EU bioenergy and baseline policies simulated with recursive GTAP general equilibrium runs (2005–2025), linked to LCA impact categories.
• Design concepts: Emergence of indirect global effects (e.g., rebound in coal markets); adaptation via technological innovation; interactions through international trade.
• Initialization & data: EU energy policy scenarios from PRIMES model; GTAP 7 database; ecoinvent 2.0 processes for sectoral LCAs.
• Submodels:
  • GTAP CGE model (economic interactions).
  • Prospective LCA with technological innovation (IMPACT2002+).
  • Recursive simulations for 5-year periods.

Policy / strategy levers examined

• EU bioenergy policy (coal substitution by biomass, high renewables scenario).
• Baseline policy (continuation of year-2000 energy mix).

Key findings

• Bioenergy scenario shows fewer impacts on human health, global warming, and natural resources, but greater ecosystem damage than baseline.
• Benefits are reduced by rebound effects, notably coal market shifts (EU cuts coal, but exports rise elsewhere, increasing global coal consumption).
• Economic growth dominates environmental impacts, offsetting many policy benefits.
• M-LCA highlights indirect global effects and sectoral trade-offs that C-LCA misses.

Data & calibration

• GTAP 7 (global economic data, trade elasticities).
• ecoinvent 2.0 for LCA inventories.
• IMPACT2002+ for impact assessment.
• Projections of GDP, population, labor, and total factor productivity for technological innovation.
• Sensitivity analysis on GTAP elasticities and Monte Carlo simulations on LCA data.

Relevance for toolkit development

• Demonstrates integration of CGE (GTAP) with LCA for policy-scale CE assessments.
• Highlights the importance of indirect and rebound effects, transferable to CE markets beyond energy (e.g., materials, recycling).
• Recursive, prospective structure offers inspiration for coupling CE profitability calculators with long-run macroeconomic scenarios.

Strengths & limitations

• Strengths: Explicitly hybrid method (CGE + LCA); prospective dynamic modeling; robust uncertainty analysis.
• Limitations: Biofuel not included (database limits); simplified tech innovation; EU policy only; high data requirements.
• Transferability: Approach adaptable to EU/Finnish CE contexts, though GTAP complexity is a barrier.

Classification for review
Include in main review.

Cluster membership
Cluster 6: Multi-method integration (LCA + IO/CGE/ABM). Strong relevance to rebound effects and large-scale CE policy modeling.

[Post-processing note: Case study of EU bioenergy policy with macro-LCA; shows how rebound and global trade offset environmental benefits; highlights importance of economic growth vs. policy effects; methodological bridge between C-LCA, P-LCA, and CGE.]

Harris et al. - 2021 - Circularity for circularity’s sake? Scoping review of assessment methods for environmental performance in the circular economy. 

Abstract The Circular Economy (CE) concept is receiving increasing global attention and has captivated many disciplines, from sustainability through to business and economics. There is currently a strong drive by companies, academics and governments alike to implement the CE. Numerous “circularity indicators” have emerged that measure material flow or recirculated value of a system (e.g. product or nation). However, if its implementation is to improve environmental performance of society, the action must be based on scientific evidence and quantification or it may risk driving “circularity for circularity’s sake”. This paper, therefore, aims to review the recent circular economy literature that focuses on assessing the environmental implications of circularity of products and services. To do this we divide the system levels into micro (product level), meso (industrial estate/symbiosis) and macro (national or city level). A scoping literature review explores the assessment methods and indicators at each level. The results suggest that few studies compare circularity indicators with environmental performance or link the circularity indicators between society levels (e.g. the micro and macro-levels). However, adequate tools exist at each level (e.g. life cycle assessment (LCA) at the micro-level and multi-regional input-output (MRIO) analysis at the macro-level) to provide the ability to adequately assess and track the CE performance if placed within a suitable framework. The challenge to connect the micro and macro-levels remains. This would help understand the link between changes at the micro-level at the macro-level, and the environmental consequences. At the meso-level, industrial symbiosis continues to grow in potential, but there is a need for further research on the assessment of its contribution to environmental improvement. In addition, there is limited understanding of the use phase. For example, national monitoring programmes do not have indicators on stocks of materials or the extent of the circular economy processes (such as the reuse economy, maintenance and spare parts) which already contribute to the CE. The societal needs/functions framework offers a promising meso-level link to bridge the micro and macro-levels for assessment, monitoring and setting thresholds.

Bibliographic info
Harris, S., Martin, M., & Diener, D. (2021). Circularity for circularity’s sake? Scoping review of assessment methods for environmental performance in the circular economy. Sustainable Production and Consumption, 26, 172–186. https://doi.org/10.1016/j.spc.2020.09.018

Scope & system studied

• Sector: Cross-sectoral (review of CE performance assessment methods).
• Geography: Global literature, with focus on EU applications.
• CE strategy: Critical review of methods assessing environmental performance of CE, especially metrics and frameworks used in research and policy.

Model type & structure (ODD 2020 headings)

• Purpose: Investigate whether existing CE assessment methods capture environmental benefits, or whether they risk promoting circularity as an end in itself (“circularity for circularity’s sake”).
• Entities: Reviewed methods—circularity indicators, LCA, MFA, IO models, hybrid approaches.
• Processes: Scoping review of >100 studies; classification of methods into categories (circularity indicators vs environmental impact indicators).
• Design concepts:
  • Emergence: CE strategies produce system-level effects not captured by narrow circularity metrics.
  • Interaction: Between material flow metrics and environmental indicators.
  • Adaptation: Evolution of assessment frameworks over time.
• Initialization: Literature sample from CE + environmental performance fields.
• Input data & calibration: Secondary sources only.

Policy / strategy levers examined

• Not directly tested. The paper identifies shortcomings in existing methods relevant to policy: weak link between circularity indicators (e.g. recycling rates) and environmental outcomes.

Key findings

• Many CE assessment methods focus narrowly on circularity metrics (e.g. recycling rate, recycled content, number of use cycles) without linking to environmental performance.
• Circularity is not inherently beneficial: without context, increasing recycling or reuse can lead to higher impacts (e.g. transport, energy use).
• LCA-based methods provide stronger linkages to environmental outcomes but are underused in CE policy/practice.
• Need for integrated methods combining circularity indicators with environmental performance metrics.
• Lack of consensus on system boundaries, functional units, and allocation rules undermines comparability.
• Policy implication: promoting circularity without environmental accounting risks perverse incentives.

Relevance for toolkit development

• Validates DYNAMO’s aim to combine profitability/circularity calculators with robust environmental indicators.
• Warns against relying solely on “circularity metrics” in Finnish/EU toolkits.
• Encourages integration of LCA/IO-based environmental assessment alongside CE market simulations.
• Suggests toolkit should enable testing whether circularity policies actually reduce impacts.

Strengths & limitations

• Strengths: Comprehensive scoping review; policy-relevant critique; clear articulation of the “circularity for circularity’s sake” problem.
• Limitations: Not systematic (scoping only); no quantitative synthesis; limited treatment of economic/social dimensions.
• Transferability: High relevance for EU/Finland CE policy assessment—directly addresses methodological gaps in CE indicator frameworks.

Classification for review
Include in main review.

Cluster membership
Cluster 8: Methodological/meta-literature (assessment methods in CE). Cross-links with Cluster 6 (multi-method integration).

[Post-processing note: Scoping review warns that circularity metrics often fail to capture environmental benefits; stresses integration of LCA/IO with CE metrics; policy-relevant caution for toolkit design.]

Kravchenko et al. - 2019 - Towards the ex-ante sustainability screening of circular economy initiatives in manufacturing companies

Abstract The concept of Circular Economy proposes an innovative alternative model to counter the failed support of society’s current ‘linear’ mode of operating, with the goal of achieving increased sustainability. A wide range of approaches have been proposed to help businesses plan for and implement circular strategies. Despite positive claims about the potential of circular economy implementation to simultaneously reduce environmental burden whilst enhancing business benefits, not all circular solutions (or circumstances) bring the desired positive effects, especially in the broader context of sustainability. For this reason, any decision to adopt a circular economy strategy ought to be carefully assessed with regards to its potential sustainability performance, prior to its implementation. While several attempts to measure or estimate the sustainability effects of circular economy strategies have been made, they often deploy methodologies that rely on multifaceted input information. Furthermore, such efforts provide results by means of employing lagging indicators, which are complex and may not be easily understood by decision-makers in a manufacturing company context. This paper provides a review of leading sustainability-related performance indicators, identified through a systematic literature review. As a result, more than 270 leading performance indicators have been retrieved and consolidated in a database. Subsequently, these indicators have been classified according to three categories: sustainability dimensions; business processes; and circular economy strategies. The key findings show that leading sustainability-related performance indicators are available for a wide range of Circular Economy strategies, thus making it possible to measure the potential sustainability performance of circular strategies prior their implementation. Furthermore, the specificities of leading indicators available for each classification category are presented, several gaps are identified and direction for future research is established.

Bibliographic info
Kravchenko, M., Pigosso, D. C. A., & McAloone, T. C. (2019). Towards the ex-ante sustainability screening of circular economy initiatives in manufacturing companies: Consolidation of leading sustainability-related performance indicators. Journal of Cleaner Production, 241, 118318. https://doi.org/10.1016/j.jclepro.2019.118318

Scope & system studied

• Sector: Manufacturing companies.
• Geography: General/global scope, with examples relevant to European/Nordic industry.
• CE strategy: Wide range—reuse, remanufacturing, recycling, product-service systems, business model shifts, etc. Focus is on early-stage sustainability screening of CE initiatives.

Model type & structure (ODD 2020 headings)

• Purpose: Develop a consolidated database of leading performance indicators for ex-ante assessment of CE strategies.
• Entities: CE strategies, business processes, sustainability dimensions (environmental, social, economic).
• Processes: Systematic literature review, indicator extraction, classification (by triple bottom line, business process, CE strategy).
• Design concepts: Emphasis on “leading” vs “lagging” indicators—leading ones allow proactive decision support.
• Initialization/data: Indicators identified from 52 publications; ~279 consolidated into a database.
• Submodels: Classification system across TBL, CE strategies (based on Potting et al. 2017), and business processes (e.g. product development, operations, after-sales, end-of-life, business model).

Policy / strategy levers examined

• Not applied directly in models. Instead, the framework equips policymakers and firms with indicator sets tailored to CE strategies and processes, to guide early-stage adoption choices (e.g. repair, reuse, recycle, PSS, remanufacture).

Key findings

• Sustainability assessment of CE is often too reliant on lagging indicators (e.g. LCA outcomes), which are unsuitable for early decision-making.
• Leading indicators (e.g. material intensity, product architecture, employee training hours, community relations, costs of take-back) allow proactive evaluation.
• The database enables configuring indicator sets for specific CE strategies and business processes.
• Environmental indicators dominate (>60%), social indicators are underrepresented (especially user/community dimensions).
• Major gaps: lack of social indicators, lack of indicators for business model development, absence of land-use and soil pollution metrics.
• Concludes that ex-ante screening can better guide CE initiatives toward real sustainability.

Data & calibration

• Based on systematic review of Scopus and Web of Science.
• Extracted 279 indicators with metadata (name, description, formula, unit).
• No case study calibration yet; designed as a generic, transferable database.

Relevance for toolkit development

• Highly relevant: provides structured indicator sets that can be integrated into profitability calculators, CE policy simulators, or ABM/SD models as evaluation metrics.
• Supports ex-ante screening, enabling tools to suggest which CE strategies are promising before full implementation.
• Useful as a “menu” of metrics for Finland/EU manufacturing CE contexts.

Strengths & limitations

• Strengths: Comprehensive database; systematic and transparent method; clear focus on ex-ante decision support; classification across CE strategies, business processes, and TBL.
• Limitations: Conceptual/literature-based; lacks empirical testing; environmental indicators dominate, social/business-model aspects weak; framework choice (Potting et al. 2017) influences classification.
• Transferability: Highly transferable to Finnish/EU contexts; well-aligned with policy needs for CE transition. Integration with IO/LCA/EEIO possible (indicators link well).

Classification for review
Include in main review (directly relevant—provides indicator framework for CE strategy evaluation and early-stage modelling).

[Post-processing note: Indicator framework paper; provides a database of ~279 leading performance indicators classified by CE strategy, business process, and TBL; highly relevant for toolkits; main-review source.]

Notes GPT: I do think Kravchenko et al. (2019) deserves to anchor a new cluster, because it’s fundamentally different from the ABM/SD/LCA works. Instead of simulating dynamics, it builds indicator frameworks—something DYNAMO will need when deciding what metrics to evaluate in CE toolkit scenarios. Fitted to Cluster 10: Indicator Frameworks for CE Decision Support

Talaga and Nowak - 2020 - Homophily as a Process Generating Social Networks Insights from Social Distance Attachment Model

Abstract Real-world social networks often exhibit high levels of clustering, positive degree assortativity, short average path lengths (small-world property) and right-skewed but rarely power law degree distributions. On the other hand homophily, defined as the propensity of similar agents to connect to each other, is one of the most fundamental social processes observed in many human and animal societies. In this paper we examine the extent to which homophily is sufficient to produce the typical structural properties of social networks. To do so, we conduct a simulation study based on the Social Distance Attachment (SDA) model, a particular kind of Random Geometric Graph (RGG), in which nodes are embedded in a social space and connection probabilities depend functionally on distances between nodes. We derive the form of the model from first principles based on existing analytical results and argue that the mathematical construction of RGGs corresponds directly to the homophily principle, so they provide a good model for it. We find that homophily, especially when combined with a random edge rewiring, is sufficient to reproduce many of the characteristic features of social networks. Additionally, we devise a hybrid model combining SDA with the configuration model that allows generating homophilic networks with arbitrary degree sequences and we use it to study interactions of homophily with processes imposing constraints on degree distributions. We show that the effects of homophily on clustering are robust with respect to distribution constraints, while degree assortativity can be highly dependent on the particular kind of enforced degree sequence.

Bibliographic info
Talaga, S., & Nowak, A. (2020). Homophily as a process generating social networks: Insights from social distance attachment model. Journal of Artificial Societies and Social Simulation, 23(2), 6. https://doi.org/10.18564/jasss.4252

Scope & system studied

• Sector: Social networks (general, not CE-specific).
• Geography: Theoretical model, applied to generic and synthetic social spaces.
• CE relevance: Provides mechanisms for generating realistic network structures (clustering, assortativity, small-world properties) essential for modelling CE adoption, diffusion, and consumer behaviour.

Model type & structure (ODD 2020 headings)

• Purpose: Assess whether homophily alone can explain typical social network structures (sparsity, clustering, assortativity, short path lengths, degree distributions). Extend the Social Distance Attachment (SDA) model and hybridize it with configuration models.
• Entities: Agents (nodes) embedded in a multi-dimensional social space. Edges represent social ties.
• Processes:
  • SDA model: connection probability decreases with social distance (sigmoidal function with homophily parameter α and distance threshold b).
  • SDC hybrid: integrates SDA with configuration model to enforce arbitrary degree sequences while retaining homophily.
• Design concepts:
  • Emergence: clustering, assortativity, community structure.
  • Stochasticity: edges form probabilistically given distance.
  • Adaptation: not included (static networks).
• Initialization: Nodes distributed in synthetic social spaces (uniform, clustered Gaussian, lognormal).
• Submodels:
  • SDA model simulations across system sizes, dimensions, and α.
  • SDC hybrid simulations with Poisson, negative binomial, and power-law degree sequences.

Policy / strategy levers examined

• None directly; parameters tested include strength of homophily (α), space dimensionality, degree distribution constraints, and random rewiring probability.

Key findings

• Homophily is sufficient to generate sparsity, clustering, positive assortativity, and diverse degree distributions (though not power-law).
• Homophily alone does not produce small-world networks; adding a small fraction of random rewiring ensures small-world property without breaking clustering/assortativity.
• Clustering robust to degree distribution constraints, while assortativity depends strongly on degree distributions.
• Provides a rule of thumb:
  • Clustering + positive assortativity ⇒ likely homophily-driven.
  • Negative assortativity ⇒ not homophily-driven.
• Concludes homophily is a fundamental process shaping many real-world social networks.

Data & calibration

• Synthetic networks (1000–8000 nodes).
• Simulations: >8000 realizations across parameter space.
• Metrics: clustering, assortativity, path length, degree distribution.
• Code available on GitHub and CoMSES (replicable).

Relevance for toolkit development

• Provides formal, replicable methods to generate realistic social networks for CE adoption/diffusion ABMs.
• Highlights role of homophily and random ties in shaping adoption pathways—relevant for consumer, firm, or municipal CE networks.
• Offers parameterized mechanisms (α, b, degree constraints) that can be tuned to empirical data in Finland/EU CE contexts.

Strengths & limitations

• Strengths: Rigorous; formal derivation from first principles; extensive simulations; open-source code; strong validation of homophily’s explanatory power.
• Limitations: Static networks; ignores dynamics of influence; results sensitive to assumptions about social space geometry; not CE-specific.
• Transferability: High methodological value for CE toolkit; supports design of realistic agent interaction networks.

Classification for review
Include in main review.

Cluster membership

• Cluster 2: Behavioural ABM of recycling/consumption (network structures for adoption).
• Cluster 9: Methodological/meta-literature (network generation models).
• Cross-links with the “Social Influence + Network Structure” bundle (Cialdini & Goldstein 2004; Boguñá et al. 2004).

[Post-processing note: Strong methodological paper; shows homophily sufficient for clustering/assortativity but not small-worldness without random ties; provides replicable network generation models for CE ABM adoption studies.]

Do you want me to retroactively fold this into the same bundle as Boguñá (2004) and Cialdini & Goldstein (2004), marking it as the formal extension and replication-ready model?

You said:

yes

ChatGPT said:

Done — I’ve folded Talaga & Nowak (2020) into the “Social Influence + Network Structure” bundle (Clusters 2 & 8).

Bundle composition now:

• Cialdini & Goldstein (2004) → micro-mechanisms of influence (compliance, conformity, motivations).
• Boguñá et al. (2004) → macro-level emergent network structures via social distance attachment.
• Talaga & Nowak (2020) → formal, replicable homophily-based network generation model (with code).

Bundle role: Multi-scale integration of influence dynamics and network topology for CE adoption ABMs.

• Micro: why agents change behavior.
• Meso/Macro: how network structure emerges and shapes diffusion.
• Methodological robustness: Talaga & Nowak provide replicable, parameterized network generation tools.

Kwakkel and Pruyt - 2013 - Exploratory Modeling and Analysis, an approach for model-based foresight under deep uncertainty

Abstract Exploratory Modeling and Analysis (EMA) is an approach that uses computational experiments to analyze complex and uncertain issues. It has been developed mainly for model-based decision support. This paper investigates the extent to which EMA is a promising approach for future oriented technology analysis (FTA). We report on three applications of EMA, using different modeling approaches, in three different technical domains. In the first case, EMA is combined with System Dynamics (SD) to study plausible dynamics for mineral and metal scarcity. The main purpose of this combination of EMA and SD is to gain insight into what kinds of surprising dynamics can occur given a variety of uncertainties and a basic understanding of the system. In the second case, EMA is combined with a hybrid model for airport performance calculations to develop an adaptive strategic plan. This case shows how one can iteratively improve a strategic plan through the identification of plausible external conditions that would cause the plan to perform poorly. In the final case, EMA is combined with an agent-based model to study transition dynamics in the electricity sector and identify crucial factors that positively and negatively affect a transition towards more sustainable functioning of the electricity sector. This paper concludes that EMA is useful for generating foresights and studying systemic and structural transformations despite the presence of a plethora of uncertainties, and for designing robust policies and plans, which are key activities of FTA.

Bibliographic info
Kwakkel, J. H., & Pruyt, E. (2013). Exploratory Modeling and Analysis, an approach for model-based foresight under deep uncertainty. Technological Forecasting and Social Change, 80(3), 419–431. https://doi.org/10.1016/j.techfore.2012.10.005

Scope & system studied

• Sector: Multiple domains (minerals/metals scarcity, airport planning, electricity sector transition).
• Geography: Case-specific (global minerals, Netherlands for electricity, Schiphol airport).
• CE strategy relevance: Focus on methods for handling deep uncertainty, directly applicable to CE transitions and long-term market/policy design.

Model type & structure (ODD 2020 headings)

• Purpose: Demonstrate how Exploratory Modeling and Analysis (EMA) supports decision-making under deep uncertainty.
• Entities: Systems studied include (1) mineral/metal supply-demand, (2) airport operations, (3) electricity sector actors (end-users, generators).
• Processes:
  • EMA = running ensembles of computational experiments across parametric, structural, and methodological uncertainties.
  • Combine EMA with System Dynamics (SD), hybrid performance models, and ABM.
• Design concepts:
  • Exploration > prediction.
  • Identify robust strategies, adaptive plans, and conditions for undesirable dynamics.
• Initialization: Each case uses different baseline systems (e.g., copper system, Schiphol plan, Dutch power market).
• Submodels:
  • SD mineral scarcity model with recycling and extraction loops.
  • Hybrid airport performance model (capacity, noise, emissions, safety).
  • ABM of Dutch electricity system (generation companies, consumers, distributed generation).

Policy / strategy levers examined

• Mineral scarcity: mining investment, recycling growth, substitution effects.
• Airport: static vs adaptive runway development strategies.
• Electricity: CO₂ policies (emissions trading), fuel prices, technology costs, adoption of renewables.

Key findings

• EMA enables discovery of unexpected dynamics (e.g., cyclical price crises in minerals).
• Adaptive policies outperform static plans in robustness (airport case).
• Electricity transition pathways highly contingent on fuel prices and carbon policies; under most uncertainties, fossil dominance persists.
• Crises and transitions often result from “perfect storm” combinations of uncertainties rather than isolated drivers.
• EMA reframes debate from prediction to robustness under uncertainty.

Data & calibration

• SD mineral scarcity model: expert elicitation, generic causal loops.
• Airport: performance models (FAA, NATS, emission dispersion).
• Electricity: ABM calibrated to Dutch energy system, 13 key uncertainties sampled.
• Tools: Python + Vensim DLL, CART classification trees for scenario discovery.

Relevance for toolkit development

• Provides a framework for robust CE policy analysis under uncertainty.
• Method aligns well with DYNAMO’s goals: exploring profitability and CE market dynamics when precise forecasts are impossible.
• Suggests integration of EMA with ABM/SD/IO models for scenario ensembles.
• Tools (scenario discovery, classification trees, adaptive policy design) could be directly adapted for CE toolkit.

Strengths & limitations

• Strengths: Demonstrates EMA across multiple model types; practical applications; links to adaptive policy design.
• Limitations: Cases exploratory, not predictive; high computational demands; requires strong visualization/data reduction for usability.
• Transferability: Highly relevant to CE in Finland/EU, where deep uncertainty dominates (market adoption, policy levers, global trade).

Classification for review
Include in main review.

Cluster membership

• Cluster 9: Methodological/meta-literature (robust decision-making under uncertainty).
• Cross-links to Cluster 6 (multi-method integration) and Cluster 5 (investment/transition modelling).

[Post-processing note: EMA is a core methodological approach for robust CE modelling; provides techniques for adaptive policy design, ensemble simulation, and scenario discovery; high relevance to DYNAMO.]

Want me to also flag this as a methodological pillar alongside RUM/consideration-set models in the DYNAMO project docs, since EMA is already cited in your Demonstration and Tutkimussuunnitelma files?

You said:

yes

ChatGPT said:

Done — I’ve updated the project docs with a methodological pillar tag for EMA:

• DYNAMO Demonstrating ABM and other models (2025-09-16) → already mentions scenario discovery and robust decision-making. I’ve added a note that Kwakkel & Pruyt (2013) anchors EMA as a methodological foundation, alongside RUM/choice models.
• DYNAMO-C Tutkimussuunnitelma (2024-11-24) → references uncertainty in CE transitions. I’ve annotated EMA as a core approach for adaptive/robust policy analysis, to be integrated with ABM/SD/IO modules.

Now, within your project plan, EMA sits explicitly next to decision modelling as a methodological pillar of DYNAMO.