A HOLISTIC APPROACH TO SPACE SUSTAINABILITY: CLOSING THE LOOP BETWEEN GLOBAL SPACE HEALTH INDICATORS AND LIFE CYCLE ASSESSMENT

11th EUROPEAN CONFERENCE FOR AERONAUTICS AND AEROSPACE SCIENCES (EUCASS) A HOLISTIC APPROACH TO SPACE SUSTAINABILITY: CLOSING THE LOOP BETWEEN GLOBAL SPACE HEALTH INDICATORS AND LIFE CYCLE ASSESSMENT Cecilia Lanfredi Alberti†⋆, Yirui Wang†, Andrew R. Wilson‡§ and Massimiliano Vasile† †University of Strathclyde, 16 Richmond Street, Glasgow G1 1XQ, United Kingdom ‡Glasgow Caledonian University, 70 Cowcaddens Road, Glasgow G4 0BA, United Kingdom §Metasat UK Ltd, 21 Stravaig Path, Renfrewshire, PA2 0RZ, United Kingdom ⋆Corresponding author: [email protected] Abstract This paper presents an integrated framework that combines Life Cycle Assessment (LCA), utilizing the Strathclyde Space Systems Database (SSSD), with the NEtwork model for Space SustainabilitY (NESSY) to evaluate space sustainability from both terrestrial and orbital perspectives. NESSY simulates orbital debris dynamics under various policy scenarios, while the LCA assesses environmental impacts across all mission phases on Earth. The framework produces Sustainability Indices that quantify both systemwide and mission-specific impacts. Two case studies demonstrate the framework’s potential to inform environmentally responsible space mission design. The results underscore the critical role of collision avoidance and end-of-life disposal policies. Overall, the framework supports evidence-based decisionmaking to mitigate long-term environmental degradation in both Earth-based and orbital domains. 1. Introduction The space sector is currently one of the fastest-growing industries, leading to a continuous rise in the number of resident space objects. Collisions among these objects can generate debris clouds, potentially initiating a cascade effect, commonly known as the Kessler Syndrome, that could render certain orbital regions unusable for future missions.1Large satellite constellations significantly contribute to the evolution of the space debris environment,2while the presence of untracked debris further elevates the risk of frequent and hazardous collisions.3To ensure the long-term sustainability of space operations, it is imperative that mission planners and regulatory bodies incorporate environmental considerations during the design and approval phases of space missions. To address this challenge, the NESSY is introduced. NESSY models the space environment as a dynamic temporal network, where each node represents a group of space objects of a specific type, and the links capture their interactions. This network-based approach enables the analysis of the temporal evolution of object interactions and provides insights into the overall health of the orbital environment through its topological and dynamical properties. Complementing this orbital perspective, the SSSD offers an LCA framework tailored specifically for the space sector. The SSSD enables the quantification of environmental, social, and economic impacts across all phases of a space mission. This Earth-based assessment tool supports stakeholders in evaluating and minimizing negative impacts, thereby promoting responsible and sustainable space activities that align with broader goals of environmental stewardship and societal well-being.4 This paper proposes an integrated framework that combines NESSY and the SSSD to enable a comprehensive, end-to-end assessment of space mission impacts across both orbital and terrestrial domains. The study makes significant advancements to the foundational work reported by Wang (2024),5and addresses two primary objectives. Objective 1 (OB1) involves a quantitative evaluation of the trade-offfor the STRATHcube constellation between omitting the propulsion system, to reduce environmental impact, and including it to enable critical functions such as collision avoidance and end-of-life disposal. Objective 2 (OB2) focuses on analysing the evolution of the space environment and assessing how different launch traffic models influence both Earth-based and orbital environmental impacts. Copyright ©2025 by the authors. Published by the EUCASS association with permission. A HOLISTIC APPROACH TO SPACE SUSTAINABILITY 2. Strathclyde Space System Database The SSSD is a specialised database developed at the University of Strathclyde to support process-based LCA tailored to the specific needs of the space sector.4While this paper focuses on the application of LCA, the overarching ambition of the SSSD project is to facilitate the transition towards a comprehensive Life Cycle Sustainability Assessment (LCSA) framework. The database was initially created to address the absence of space-specific LCA datasets prior to the release of the European Space Agency (ESA) LCA Database, with a scope that extends beyond geographical boundaries. Ultimately, the tool can be used to facilitate the integration of eco-design and Life Cycle Engineering (LCE) principles into concurrent engineering practices for space mission development. 2.1 Impact Category Scores Computation While the SSSD supports multiple assessment methodologies, its default output presents results at the midpoint level, based on the IPAT equation. Midpoint indicator results are calculated as: IRc=X s CFcs ·ms(1) where IRcis the indicator result for impact category c, CFcs is the characterisation factor linking intervention sto category c, and msis the magnitude of intervention s. The full set of indicators are primarily based on the Product Environmental Footprint (PEF) approach.6This includes: Acidification (AC),Climate Change (CC),Ecotoxicity, Freshwater (EX-FW),Eutrophication (Freshwater, Marine, Terrestrial – EP-FW, EP-M, EP-T),Human Toxicity (Cancer and Non-Cancer – HT-C, HT-NC),Ionising Radiation (IC),Land Use (LU),Ozone Depletion (OD),Particulate Matter (PM),Photochemical Oxidant Formation (POF),Resource Use (Fossil Fuels and Minerals/Metals – RU-FF, RU-MM), and Water Use (WU). However, to expand upon the existing suite of Earth-based indicators, adaptations to the PEF method have been introduced in the SSSD to better reflect the space sector’s unique characteristics. This includes the definition of additional impact categories and new/updated characterisation factors for specific impact categories related to launcher emissions and re-entry. To facilitate streamlined decision-making within concurrent design environments, the SSSD also supports generating a single score using a Multi-Criteria Decision Analysis (MCDA) approach based on Multi-Attribute Value Theory (MAVT), which is particularly useful for constellation assessments. The overall score is computed as: v(a)= I X i=1 wi·vi(a) (2) where v(a) is the overall score for system a,wiis the weight assigned to impact category i, and vi(a) is the performance of system ain that category. Whilst various single score methodologies for space LCA are beginning to develop7,8 for environmental criteria within the SSSD, the default single score method applied relates to the PEF approach for normalisation and weighting based on EU-27 consumption data,9,10 with the planetary boundaries approach used as an alternative.11 2.2 Integration of the Space Environmental Impact into LCA Historically, LCAs in the space sector have primarily concentrated on Earth-based environmental impacts. Notable exceptions include the work of Thibaut Maury,12 who developed a framework to address the risks associated with longlived orbital debris. However, this approach has not been formally adopted by the ESA. Instead, ESA manages space debris through its Space Debris Mitigation Guidelines.13 Additionally, the technical complexity involved, particularly concerning ESA’s MASTER tool, has hindered broader integration of debris impact assessment within standard LCA practices. 2.2.1 Orbital Impact Indicators To overcome this limitation, two new space-based impact categories have been introduced into the SSSD and will be added to the default impact assessment framework. New environmental indicators were created to reflect spacecraft operations at varying altitudes and inclinations, referring to data provided by NESSY. The new orbital impact indicators include: •OR-AP (Orbital Risk): Represents the induced risk of the objects in a particular orbital regime with the rest of the space environment. This is a unit-less quantity ranging from 0 to 1, quantified using percolation theory. 2 A HOLISTIC APPROACH TO SPACE SUSTAINABILITY •OR-DR (Orbital Space Use and Depletion): Measures the spatial density of spacecrafts in a particular orbital regime, expressed in units of objects/m3. These indicators are then passed through the MCDA module to compare the relative criticality of Earth-based vs. space-based impacts, addressing the emerging Space Sustainability Paradox.14 3. NEtwork model for Space SustainabilitY (NESSY) NEtwork model for Space SustainabilitY (NESSY) was proposed to model the space environment as a stochastic dynamics model, where each node is a group of objects, belonging to a given class, and their relationship is represented by stochastic links. The stochastic dynamics network model of the space environment is introduced, followed by the percolation dynamics of the space environment, which quantifies the local/global health of the space environment. This provides a foundation for defining orbital impact indicators, offering a novel measure of orbital risk and supporting long-term sustainability assessments of space activities. 3.1 A Stochastic Dynamic Network Model of the Space Environment The space environment is modelled with a network consisting of nnodes. Each node Sirepresents xSi(tk) objects of a given class Sat time tk. Four species are considered: Payloads (P), Upper stages (U), Fragments (F) and Nonmanoeuvrable satellites (N), therefore, S={P,U,N,F}. Each class is further partitioned according to the altitude shell and inclination interval to which a given group of objects belongs. The number of objects xSi(tk) can change because of collisions, explosions, natural decay due to atmospheric drag, Post-Mission Disposal (PMD) strategies, operational lifetime duration, and new launches. Once the structure and topology of the network are defined one can derive the stochastic evolutionary equations, governing the evolution of each node Pi,Ui,Niand Fi. Let Xk= [xP1,xU1,xN1,xF1..., xPi,xUi,xNi,xFi, ..., xPn,xUn,xNn,xFn]Trepresent the vector that contains all the nodes xSi, with S∈ {P,U,N,F}and i=1, ..., n. Then a more compact form of the evolutionary equations is: Xk+1=Xk+Y(Xk)+g(tk) (3) where Y(Xk) contains the interactions among nodes (collisions and orbit change) and satellite failures, while g(tk) contains all time dependent terms, such as the launch traffic model Λi(tk). For further details on each component of the model, please refer to the paper.15 3.2 The Percolation Dynamics of the Space Environment Percolation is quantified differently for different networks and has its own dynamics. Together with percolation one can compute the speed of diffusion which measures how fast events are spreading.16 The connectivity between nodes represents the degree of interaction or the probability that two species or communities interact. The connection rate between node Siand node Sjwithin time ∆tk, denoted by χSiSj, depends on: the collision rate between the nodes (χC SiSj), the rate of the fragments generated from node Siflowing to node Sj(χF,F SiSj), the rate of an object flowing from node Sito node Sjdue to the natural decay (χF,D SiSj), and the rate of an object flowing from node Sito node Sjdue to the failure of PMD (χF,PMD SiSj). The event can be generated by one node and affect the other or can be due to the interaction of both nodes. The total flow rate defining a link from node Sito node Sjis: χSiSj=χC SiSj+χF,F SiSj+χF,D SiSj+χF,PMD SiSj(4) If only collision-related links are considered, then χSiSj=χC SiSj+χF,F SiSj. In order to account for the spatial distribution of nodes across multiple sites and the flow of the consequences of events across different parts of the space environment the expression is defined: pPiPj=1−e−∆tkPSi∈S PSj∈S χSiSj(5) and, δPiPj(ρ)=pPiPj≥ρ(6) which is different from zero if the site Piand Pjare connected. In a directed graph Gρ, the propagation of the effects of events happening in a given orbit site can be quantified via the reachability matrix Rc(Gρ). Each entry (i,j) of this matrix is 1 if there is a continuous path from site ito site j, 3 A HOLISTIC APPROACH TO SPACE SUSTAINABILITY indicating that site ihas an influence on the evolution of site j. The reachability matrix is the transitive closure of the n×nconnectivity matrix Apwhose entries are δPiPj. With Rc(Gρ) one can define the local connectivity of site ias: αn i=Z1 0Pn j=1ri j ndρ(7) Therefore, the global connectivity is: αn=Pn iαn i n(8) Considering the fragments induced by the collision events, the augmentation of the number of fragments generated by the collisions related to the nodes in the given orbital regime is introduced. Given the estimated number of fragments flowing from node Sito node Flgenerated by the collision between node Siand node Sj NFl SiSj=χC SiSj∆tkζSiSjξFl SiSj(9) where PlξFl SiSj=1. The updated number of objects in node Flis: x′ Fl=xFl+X S∈S X j NFl SiSj(10) Then the updated number of objects in each node is given by: X′ k=[xP1,xU1,xN1,x′ F1..., xPi,xUi,xNi,x′ Fi, ..., xPn,xUn,xNn,x′ Fn]T(11) Consequently, the impact score αn ican be recalculated based on the updated state vector X′ k, which accounts for the increase in fragment numbers due to the collisions involving nodes in site i. 4. OB1 - Quantitative Evaluation of the CAM&PMD Trade-Off The first objective involves a quantitative assessment of the environmental trade-offs associated with including or excluding propulsion systems in satellite constellations. While omitting propulsion may reduce the overall environmental footprint, it can compromise critical functions essential to space sustainability, such as collision avoidance manoeuvres (CAM) and post-mission disposal (PMD). This analysis systematically compares the life cycle impacts of constellations with and without propulsion, evaluating how the environmental burden of propulsion hardware and fuel consumption balances against the benefits of improved debris mitigation and operational safety. From this comparison, a set of weighting factors can be derived to reflect the relative significance of various impact categories across both terrestrial and orbital domains. These insights aim to support responsible design decisions by integrating Earth-based and Space-based environmental considerations within a unified sustainability framework. 4.1 STRATHcube Constellation The STRATHcube satellite has been selected as the reference for designing a constellation mission to compare the two scenarios, due to the availability of detailed design data. It is a 2U CubeSat developed through a student-led initiative at the University of Strathclyde, with mission objectives strongly focused on space sustainability. Originally conceived to demonstrate debris detection and re-entry analysis, the mission has since matured and was recently selected to participate in the European Space Agency’s Fly Your Satellite! programme. Figure 1: STRATHcube Satellite Configurations. 4 A HOLISTIC APPROACH TO SPACE SUSTAINABILITY 4.1.1 Mission Scenarios For this study, two scenarios were evaluated, each involving a constellation of 1,000 STRATHcube satellites deployed in orbits at altitudes between 800 and 850 km, with inclinations ranging from 40° to 60°. The operational life is assumed to start in June 2029 and last for five years, following a launch using a Falcon 9 launch vehicle. A conservative ride-share cap of 100 CubeSats was considered per launch, requiring a total of 10 launches for each scenario. The constellation layout consists of 20 orbital planes with 50 satellites per plane, ensuring global coverage and even distribution. The timing and sequence of operations are assumed to be identical in both scenarios and are summarized in the Concept of Operations (CONOPS) diagram shown in Figure 2. The mission phases are defined following the ESA convention, ensuring consistency with standardized space mission lifecycle definitions. In the figure, mission phases governed by LCA assumptions are highlighted in green, while those affected by the NESSY framework are blue, to clearly distinguish the environmental modelling boundaries. The first scenario considers an updated STRATHcube design, consistent with the latest Fly Your Satellite! Baseline Design Review (BDR) and the baseline configuration detailed by Wilson and Vasile,17 which assumes no collision avoidance manoeuvrers (CAM) and includes post-mission disposal (PMD) conducted passively, without any propulsion system. The second scenario presents the same model, but enhanced by the integration of an ExPace cold-gas propulsion system and the GomSpace OBC (On-Board Computer) to enable active CAM and PMD. These configurations are based on consistent design and operational assumptions, ensuring a fair and coherent comparison of the environmental impacts between the two STRATHcube variants. The input parameters used for the two tools applied in this study are detailed in Table 1 and Table 2. Figure 2: Concept of Operations (CONOPS) for the STRATHcube constellation mission. Table 1: NESSY Data w/o CAM&PMD w/CAM&PMD Altitude [km] 800–850 800–850 Inclination [degrees] 40–60 40–60 Number of satellites 1000 1000 Mass [kg] 2.726075 4.046075 Size [cm] 20 20 Lifetime [year] 5 5 CAM Success [%] 0 99.9 PMD Failure [%] 100 5 Table 2: SSSD LCA Data w/o CAM&PMD w/CAM&PMD Number of satellites 1000 1000 Mass [kg] 2.726075 4.046075 Launcher Falcon 9 Falcon 9 Payload-to-Launcher ratio 0.02478 0.03678 Re-entry [% survived mass] 45 0 Disposal space ocean 5 A HOLISTIC APPROACH TO SPACE SUSTAINABILITY 4.1.2 Earth Environmental Impact - Phase C+D According to the ESA mission lifecycle convention, Phase C involves the detailed design of the satellite platform and its sub-systems, while Phase D covers manufacturing, integration, and testing activities in preparation for launch. Together, Phases C and D represent the most resourceand energy-intensive stages of spacecraft development, as they encompass the full realization of the physical hardware. In this study, Phase C+D is of particular importance because it accounts for the upstream environmental impacts associated with the production and assembly of satellite components, especially those that vary between the two STRATHcube design scenarios. The CubeSat evaluated has been modelled using the latest available mass data and component breakdowns, ensuring that the environmental assessment reflects the current design baseline. Notably, Scenario 2 incorporates an ExPace cold-gas propulsion system and a GomSpace onboard computer (OBC) to support active collision avoidance and post-mission disposal. The integration of the propulsion subsystem increases the satellite’s overall mass and system complexity, leading to elevated resource consumption and emissions during both the design and manufacturing stages. Figure 3: Normalized environmental emissions-based impacts of STRATHcube components by category, shown as percentages compared to the baseline configuration. Figure 4: Normalized environmental resources-based impacts of STRATHcube components by category, shown as percentages compared to the baseline configuration. Under the CAM&PMD configuration, propulsion system components exhibit percentage rises across multiple environmental impact categories. Specifically, emissions-related impacts (Figure 3) show growths in Climate Change (6.67%), Ionising Radiation (6.97%), and Acidification (∼5.5%). These increases suggest that the propulsion subsystem is a not negligible contributor to these environmental pressures. The primary drivers are likely the energy-intensive manufacturing processes, the production and handling of propellants that emit greenhouse gases and acidifying substances, as well as materials or procedures associated with ionizing radiation during component fabrication or testing. 6 A HOLISTIC APPROACH TO SPACE SUSTAINABILITY In terms of resource use (Figure 4), the most pronounced increases occur in Water Use (7.37%), Total Cumulative Energy Demand (6.7%), and Resource Use, Fossils (6.65%). This indicates that the propulsion system demands considerable water and energy inputs, alongside reliance on fossil-based materials. Contributing factors include cooling and cleaning operations during propulsion assembly, energy-intensive production of specialised propellant chemicals, and the dependence on fossil fuels and fossil-derived materials throughout manufacturing and testing phases. Conversely, some impact categories, such as Human Toxicity and Particulate Matter show negligible increases, reflecting the stringent controls in production environments and minimal release of toxic substances or particulate emissions. Overall, these findings underscore the introduction of propulsion subsystem as a candidate driver in mission decision-making, particularly regarding energy demand, water use, and climate-related effects. 4.1.3 Earth Environmental Impact - Phase E1 Phase E1 of the mission lifecycle includes all launch-related operations and events, such as propellant consumption and engine testing firings. In this study, Falcon 9 is selected as the reference launcher due to its high reliability and the availability of detailed environmental data. Its capacity to deploy large numbers of CubeSats via ride-share missions makes it a representative and practical choice for constellation scenarios and ensures consistency with current industry practices. Recalling Figure 2, both STRATHcube constellation configurations are assumed to be deployed in 10 shared launch events, with up to 100 satellites per launcher mission. Configuration 1 (without CAM&PMD) has a total payload mass of 272.6 kg on the vehicle, while Configuration 2 (with CAM&PMD) totals 404.6 kg. Given Falcon 9’s LEO payload capacity of approximately 11,000 kg, these correspond to payload-to-launcher mass ratios of 0.02478 and 0.03678, respectively (Table 2). Although both ratios represent a small fraction of the vehicle’s capacity, they provide a basis for proportionally attributing environmental impacts.           P Lw/CAM&PMD −P Lw/o CAM&PMD P Lw/o CAM&PMD           ×100 (12) The percentual increment is calculated as the relative increase in payload-to-launcher ratio (Equation 12), as the difference in environmental impact between the two configurations during Phase E1 is driven entirely by the change in payload mass, since the launch system remains constant. Substituting the given values yields an approximate increase of 32.62%. This method assumes that launch-related impacts scale linearly with the share of payload mass, allowing a consistent and quantitative comparison of the environmental consequences of design changes. 4.1.4 Space Environmental Impact - Phase E2 Phase E2, as defined by the ESA, encompasses the operational lifetime of the satellite in orbit. Environmental considerations shift from terrestrial emissions to interactions within the orbital environment, with a focus on space traffic management and long-term orbital sustainability. This study assesses the Orbital Risk (OR),the risk objects pose in a given orbital regime during a satellite’s operational life, using the NESSY framework coupled with a launch traffic model. This combined approach allows for realistic estimation of collision probabilities and orbital congestion based on detailed mission parameters and traffic dynamics (Table 1). As shown in Figure 5, during the operational period (roughly June 2029 to June 2034), the STRATHcube constellation with CAM&PMD capabilities (Scenario 2) achieves a significantly lower Orbital Risk of 0.00083574, compared to 0.00249257 in the baseline configuration without CAM&PMD (Scenario 1). This reduction results from enhanced active collision avoidance and controlled post-mission disposal capabilities, which reduce both the duration and likelihood of close conjunctions. The relative reduction in Orbital Risk is calculated analogously to the payload-to-launcher ratio increment (Equation 12), as follows: ORw/CAM&PMD −ORw/o CAM&PMD ORw/o CAM&PMD !×100 ≈ −66.45 (13) This notable decrease highlights the beneficial role of propulsion-enabled subsystems in improving orbital safety and promoting space sustainability by minimizing passive orbital exposure and enabling timely evasive manoeuvres aligned with debris mitigation guidelines. Regarding Orbital Space Use, NESSY results indicate negligible differences between the two configurations. Given the equal number of satellites and the limited spatial footprint of the additional propulsion system, overall orbital crowding remains similar in both scenarios. Thus, while Scenario 2 substantially lowers Orbital Risk, its impact on orbital density is effectively negligible. 7 A HOLISTIC APPROACH TO SPACE SUSTAINABILITY Figure 5: Temporal evolution of NESSY environmental indices through 2060 for three STRATHcube scenarios: no constellation launch, Configuration 1 (without CAM&PMD, blue), and Configuration 2 (with CAM&PMD, red). The curves highlight differences in orbital risk and spatial occupation over the mission lifetime. 4.1.5 Space Environmental Impact - Phase F In this study, results computed using the NESSY framework (see Figure 5, after the end of operations in 2034) show that the STRATHcube configuration without CAM&PMD (Scenario 1) exhibits a higher Residual Orbital Risk (ROR) of approximately 0.00257, compared to 0.00113 for the configuration with CAM&PMD (Scenario 2). This substantial reduction, about −55.99% as calculated via the percentual increment expression in Equation 12, highlights the effectiveness of propulsion-enabled subsystems in reducing long-term collision risks through active post-mission disposal. A similar trend is observed for Orbital Resource Depletion (ORD), which quantifies the long-term occupation of valuable orbital slots by non-functional objects. Scenario 1 yields a significantly higher depletion rate of approximately 3.33 ×10−8objects/m3/year, compared to only 2.44 ×10−9objects/m3/year in Scenario 2. This corresponds to a percentual reduction of approximately −92.79%, further confirming the benefit of controlled de-orbiting in preserving orbital capacity. Together, these results emphasize the critical role of end-of-life disposal strategies in maintaining a sustainable orbital environment. 4.1.6 Earth Environmental Impact – Phase F Phase F of the mission life cycle encompasses post-mission disposal and end-of-life operations, thereby extending the environmental assessment beyond the orbital regime to include potential impacts on the Earth’s surface and atmosphere. Two critical parameters are quantified in this phase: •Mass Survival: the fraction of spacecraft mass surviving atmospheric re-entry and depositing on Earth. •Disposal Pathway: the environmental burden associated with the chosen final disposal strategy (e.g., controlled re-entry, ocean recovery, abandonment in orbit). As defined in the mission scenarios (Section 4.1.1), the without CAM&PMD case involves an uncontrolled disposal path, while Scenario 2 (with CAM&PMD) features a controlled re-entry manoeuvre. Although these end-oflife strategies differ, the satellite platform is a CubeSat, and in both cases, it is highly unlikely that any residual mass will survive atmospheric re-entry or reach the Earth’s surface. To estimate atmospheric emissions, the values in Table 2 are scaled based on the satellite mass difference between scenarios. This accounts for the increased material ablated and vaporised during re-entry of the heavier, CAM&PMDequipped satellite. While the relationship between mass and emissions is not strictly linear, being influenced by geometry, velocity, and drag, in the absence of detailed re-entry modelling, linear or slightly sub-linear scaling is a common assumption in early-stage LCA. In this study, atmospheric emissions are conservatively assigned at 45% of full burnup factors to the lighter satellite, reflecting a penalty for the lack of a targeted disposal strategy. This aligns with LCA principles that reward controlled end-of-life measures. Consequently, Scenario 2 incurs a proportional increase of approximately 45% in impact categories such as Climate Change,Human Toxicity (cancer), and Human Toxicity (non-cancer). 8 A HOLISTIC APPROACH TO SPACE SUSTAINABILITY Figure 6: Environmental emissions-based impacts of STRATHcube disposal to space strategy by category. Scenario 1 is nominally classified as “disposal to space.” In this context, disposal to space is considered to have no direct terrestrial environmental impact, as emissions and environmental burdens associated with this phase are assumed outside the scope of terrestrial impact assessment. Scenario 2 implements a controlled disposal strategy, directing re-entry targeting remote oceanic regions where, if any, surviving debris is assumed to be safely recovered. This approach involves an active manoeuvrer to initiate disposal earlier, accelerating orbital decay and minimizing the time spent in low Earth orbit. Consequently, it reduces long-term risks to the space environment and limits the occupation of valuable orbital slots, contributing to more sustainable space operations. The environmental impacts associated with the controlled disposal option (Scenario 2) are illustrated in Figure 6, depicting emissions-related effects by impact category. Reported values indicate the typical order of magnitude for each category, varying considerably as expressed using different units of measure. 4.2 STRATHcube CAM&PMD Trade-OffConsiderations This analysis presents a controlled comparison of the life cycle environmental impacts of a fixed satellite constellation architecture, evaluated under two operational strategies: one incorporating Collision Avoidance Manoeuvres and PostMission Disposal (CAM&PMD), and one excluding these measures. By isolating the effect of CAM&PMD, the LCA methodology highlights how end-of-life strategies influence mission-level environmental performance. The evaluation produces two complementary result types: •Absolute impact values, expressed in category-specific physical units (e.g., kg,CO2,eq for Climate Change, CTUh for Human Toxicity), quantify the total burden attributable to each scenario. •Relative changes (in %), denoted as approximate percentage differences, highlight the proportional increase or decrease in impact between the two scenarios within each impact category. This dual-layered approach enables robust intra-category comparisons across scenarios. Although direct comparisons across impact categories are not meaningful due to differing units and mechanisms, the method identifies where CAM&PMD induces the most significant relative shifts and which life cycle stages contribute most to overall burdens. These insights support informed design decisions, mitigation prioritization, and sustainability trade-offs. 9 A HOLISTIC APPROACH TO SPACE SUSTAINABILITY where xP(t), xU(t), xN(t) and xF(t) represent the number of payloads, upper stages, non-maneuverable satellites and fragments at time t, respectively. Table 4: An example of the population list used in NESSY ID a e i bstar mass radius alt ctrl life launch class node 1 7261.6 0.0936 28.05 8.47e-4 1840 0.0030 200 0 0 2.4530e6 2 2 2 6683.8 0.0102 58.47 0.3367 130 0.0014 200 0 0 2.4384e6 2 5 3 6598.1 8.93e-4 53.21 5.18e-4 1 1.22e-4 200 0 0 2.4596e6 3 3 4 6612.9 7.73e-4 51.63 0.0013 4 1.72e-4 200 0 0 2.4579e6 4 4 . . .. . .. . .. . .. . .. . .. . .. . .. . .. . .. . .. . .. . . 18346 1.6420e4 0.5941 26.8517 5.111e-4 4300 0.0036 2200 0 0 2.4599e6 2 5 The action vector atis at=[A,b,k] (16) where Aand bare parameters that govern the launch rate, as described in.18 The launch rate is modeled by the following equation: N(t)=n0+A·exp(d(t−t0)) b+exp(−c(t−t0)) (17) where t0=2023, c=0.2, d=0.0001 and n0=150. The parameter krepresents the proportion of newly launched objects with mass less than 100 kg that are maneuverable in a given year. The reward function is defined as: Rt=−w1·IEarth(t)−w2·ISpace(t)+w3·IBenefits(t) (18) where IEarth(t) represents the Earth-based environmental impact, such as Climate Change,Water Use, etc. ISpace(t) represents the space-based environmental impact such as the local/global connectivity factor proposed in this paper. IBenefits(t) represents the benefits generated by the launched missions, such as revenues from satellite services (e.g., communication, remote sensing, and navigation), and scientific outcomes, etc. The weights vector w=[w1,w2,w3] specifies the relative importance of each component. To improve computational efficiency during agent training, a surrogate model fis pre-trained such that [Rt,st+1]=f(st,at) (19) Considering A∈[1200,1800], b∈[0.1,0.5] and k∈[50,100]%, the following examples showcase how the proposed RL framework is used to optimize the space policy and launch schedule between 2025 and 2030. The objective is to achieve low IEarth (quantified by Climate Change), low ISpace (quantified by local connectivity factor αn i), and high IBenefits(t) (quantified by Launch rate Λ) over next 40 years. A reference case uses the following static parameters: A=1200, b=0.1, c=0.2, d=0.0001, t0=2023, n0=150 and k=100, which corresponds to the scenario with the lowest space environmental impact due to the high manoeuvrability ratio for a given launch rate. To illustrate the long-term effect across different impact categories, the cumulative Climate Change, cumulative Launch Rate and temporal local connectivity are presented, showing how these scores evolve over time. Assuming a weight vector w=[0,1,0], which indicates that the agent priorities minimizing space environmental impacts, Figure 14 and Figure 15 compare the optimized launch traffic and the corresponding long-term impacts with those of the reference scenario. As a results, both space-based and Earth-based environmental impacts are reduced, primarily due to a lower launch rate. This result highlights the importance of incorporating benefits into the reward function; otherwise, the trained agent will consistently favour actions that minimize the launch rate, potentially compromising benefits. Assuming an alternative weight vectorr w=[5×10−8,5×106,1], Figure 16 and Figure 17 compare the optimized launch traffic with the reference scenario. The simulation results suggest that the increase of mission benefits inevitably leads to a rise in both space and Earth environmental impacts, underscoring the trade-offbetween maximizing benefits and maintaining environmental sustainability. 16 A HOLISTIC APPROACH TO SPACE SUSTAINABILITY Figure 14: Comparison of launch rate. Figure 15: Comparison of scores. Figure 16: Comparison of launch rate. Figure 17: Comparison of scores. 6. Conclusions This study presents an integrated framework that combines the NESSY tool with the SSSD methodology, enabling a comprehensive assessment of space mission impacts across both terrestrial and orbital domains. Two case studies demonstrate the framework’s potential to inform environmentally responsible space mission design. Applied to the STRATHcube constellation, OB1 investigates the trade-offbetween minimizing Earth-based environmental impacts by omitting a propulsion system and enhancing mission functionality through its inclusion for collision avoidance and end-of-life disposal. The analysis reveals that while propulsion increases the environmental burden on Earth, it is essential for ensuring long-term orbital sustainability. The comparison between the CAM&PMD configuration and the non-propulsion scenario provides a quantitative basis for future mission designers to assign relative weights to different impact categories and system priorities. This enables more informed decisions when balancing environmental performance, mission safety, and compliance with space debris mitigation guidelines, particularly relevant in the context of large satellite constellations in LEO. OB2 examines how various launch traffic and related space policy scenarios influence the evolution of the space environment and affect cumulative impacts on both Earth and space. These results underscore the importance of integrated, forward-looking assessments to guide sustainable design under increasing space activity and regulatory pressure. As part of OB2, the proposed Reinforcement Learning (RL) block further demonstrate that autonomous policy optimization can successfully balance trade-offs between environmental and space progress objectives. When the reward function prioritizes only space impact, the RL agent minimizes launch rates, leading to reductions in both orbital and terrestrial impacts but at the expense of benefits. Conversely, when the positive return performance is weighted, the optimized policy increases launch activity, boosting revenues but also amplifying environmental impacts. 17 A HOLISTIC APPROACH TO SPACE SUSTAINABILITY These findings highlight the need for a balanced, multi-objective reward structure to support sustainability-aware space traffic strategies. Future steps will involve extending the framework from traditional LCA to LCSA in both objectives by testing the feasibility of incorporating economic and social dimensions alongside traditional LCA impact categories and the newly developed global space health indicators via MCDA. The complete set of quantified impacts will serve as input for defining the reward function in a reinforcement learning block, supporting automated, sustainability-aware mission architecture optimization. Together, these developments advance the foundation for a systemic and scalable approach to sustainable space system design. 18 A HOLISTIC APPROACH TO SPACE SUSTAINABILITY A. Impact Categories Weights Table 5: Phase C+D Type Impact Category Inside Weight Outside Weight Normalized Weight emission Acidification 0.088305489 0.097956266 0.073490385 emission Air acidification 0.088782816 0.097967091 0.073895796 emission Climate change 0.106125696 0.085004677 0.076643262 emission Human toxicity, non-cancer 0.000159109 0.995460523 0.001345641 emission Ionising radiation 0.110898966 0.050112535 0.047215487 emission Ozone depletion 0.085759745 0.084995549 0.061928458 emission Particulate matter 0.000159109 0.995308287 0.001345436 emission Photochemical ozone formation 0.072553699 0.229541713 0.141491966 resources Critical raw materials 0.029753381 0.328262978 0.082979121 resources Land use 0.087032617 0.122282530 0.090418451 resources Resource use, fossils 0.105807478 0.127237246 0.114377665 resources Resource use, minerals and metals, reserve base 0.000636436 0.969108373 0.005240077 resources Resource use, minerals and metals, ultimate reserve 0.000159109 0.992149263 0.001341165 resources Total cumulative energy demand 0.106603023 0.068686052 0.062208349 resources Water use 0.117263325 0.166702162 0.166078740 Table 6: Phase E1 Type Impact Category Inside Weight Outside Weight Normalized Weight emission Acidification 0.0625 0.895435743 0.093610062 emission Air acidification 0.0625 0.896127596 0.093682389 emission Climate change 0.0625 0.914987179 0.095653995 emission Human toxicity, cancer 0.0625 0.000320245 0.000033479 emission Human toxicity, non-cancer 0.0625 0.004434537 0.000463593 emission Ionising radiation 0.0625 0.949887465 0.099302518 emission Ozone depletion 0.0625 0.915003806 0.095655733 emission Particulate matter 0.0625 0.004691145 0.000490419 emission Photochemical ozone formation 0.0625 0.759133804 0.079360873 resources Critical raw materials 0.0625 0.671737022 0.070224295 resources Land use 0.0625 0.877717470 0.091757770 resources Resource use, fossils 0.0625 0.872762754 0.091239797 resources Resource use, minerals and metals, reserve base 0.0625 0.030891627 0.003229452 resources Resource use, minerals and metals, ultimate reserve 0.0625 0.007850737 0.000820727 resources Total cumulative energy demand 0.0625 0.931313948 0.097360818 resources Water use 0.0625 0.833297838 0.087114081 Table 7: Phases E2 Type Impact Category Inside Weight Outside Weight Normalized Weight space Orbital Risk 1.000000000 0.424946636 1.000000000 19 A HOLISTIC APPROACH TO SPACE SUSTAINABILITY Table 8: Phase F Type Impact Category Inside Weight Outside Weight Normalized Weight emission Climate change 0.333333333 0.000007781 0.000961 emission Human toxicity, cancer 0.333333333 0.002585077 0.319408 emission Human toxicity, non-cancer 0.333333333 0.000104923 0.012964 emission Acidification 0.142857143 0.006607991 0.092396 emission Air acidification 0.142857143 0.005905313 0.082571 emission Climate change 0.142857143 0.000000363 0.000005 emission Ozone depletion 0.142857143 0.000000645 0.000009 emission Particulate matter 0.142857143 0.000000568 0.000008 emission Photochemical ozone formation 0.142857143 0.011324483 0.158344 space Orbital Risk 0.376327463 0.575053364 0.281899 space Orbital Space Use 0.623672537 0.063310846 0.051434 References [1] D. J. Kessler and B. G. Cour-Palais, “Collision frequency of artificial satellites: The creation of a debris belt,” Journal of Geophysical Research: Space Physics, vol. 83, no. A6, pp. 2637–2646, 1978. [2] J. Zhang, Y. Yuan, K. Yang, and L. Li, “Long-term evolution of the space environment considering constellation launches and debris disposal,” IEEE Transactions on Aerospace and Electronic Systems, pp. 1–17, 2023. [3] A. C. Boley and M. Byers, “Satellite megaconstellations create risks in low earth orbit, the atmosphere and on earth,” Nature Astronomy, vol. 5, p. 123–125, 2021. [4] A. R. Wilson, Advanced Methods of Life Cycle Assessment for Space Systems. PhD thesis, University of Starthclyde, Glasgow, United Kingdom, 2019. [5] Y. Wang, A. Wilson, C. Wilson, and M. Vasile, “Closing the loop between space capacity and life cycle assessment: a network-theoretic approach,” in 75th International Astronautical Congress (IAC), (Milan, Italy), 2024. [6] A. R. Wilson, S. M. Serrano, K. J. Baker, H. B. Oqab, G. D. Dietrich, M. Vasile, T. Soares, and L. Innocenti, “From life cycle assessment of space systems to environmental communication and reporting,” in 72nd International Astronautical Congress (IAC), (Dubai, UAE), 2021. [7] M. Verkammen, A. R. Wilson, G. D. Calabuig, A. Menicucci, H. Svedhem, and M. Udriot, “A consensus-based single-score for life cycle assessment of space missions: Preliminary results,” in 10th European Conference for Aeronautics and Aerospace Sciences (EUCASS), (Lausanne, Switzerland), 2024. [8] M. Verkammen, A. R. Wilson, E. Tormena, T. Turchetto, and E. David, “Meta-study of current proposed life cycle assessment single-score methodologies for space missions’ eco-design,” in 75th International Astronautical Congress (IAC), (Milan, Italy), 2024. [9] S. Sala, E. Crenna, M. Secchi, and R. Pant, “Global normalisation factors for the Environmental Footprint and life cycle assessment,” tech. rep., European Commission, Joint Research Centre, 2017. [10] S. Sala, A. K. Cerutti, R. Pant, et al., “Development of a weighting approach for the Environmental Footprint,” tech. rep., European Commission, Joint Research Centre, 2018. [11] S. Sala, L. Benini, E. Crenna, and M. Secchi, “Global environmental impacts and planetary boundaries in LCA,” tech. rep., Publications Office of the European Union, 2016. [12] T. Maury, Consideration of space debris in the life cycle assessment framework. PhD thesis, University of Bordeaux, Bordeaux, France, 2019. [13] European Space Agency, “Esa space debris mitigation requirements,” Technical Report ESSB-ST-U-007 Issue 1, European Space Agency, 2023. [14] A. R. Wilson and M. Vasile, “The space sustainability paradox,” Journal of Cleaner Production, vol. 423, p. 138869, 2023. 20 A HOLISTIC APPROACH TO SPACE SUSTAINABILITY [15] Y. Wang, P. De Marchi, and M. Vasile, “A stochastic dynamic network model of the space environment,” arXiv preprint arXiv:2411.03173, 2024. [16] Y. Wang, C. Wilson, and M. Vasile, “Multi-layer temporal network model of the space environment,” in 2023 AAS/AIAA Astrodynamics Specialist Conference, 2023. [17] A. R. Wilson and M. Vasile, “Life cycle engineering of space systems: Preliminary findings,” Advances in Space Research, vol. 72, no. 7, pp. 2917–2935, 2023. [18] C. Wilson, M. Vasile, J. Feng, and A. Horstmann, “Modelling future launch traffic and the associated risk to new missions.” https://doi.org/10.13140/RG.2.2.13194.17604, Jan. 2025. License: CC BY-NC-ND 4.0. 21