Integrated Multi-Energy Microgrids for Port Decarbonization: A Techno-Economic Assessment of CHP-Based Cold Ironing Under Grid Constraints

1. Introduction

1.1. Port Decarbonization and the Role of Shore Power

Maritime transport is responsible for approximately 3% of global greenhouse gas (GHG) emissions, with port areas representing localized emission hotspots where environmental impacts are highly concentrated [1]. During berthing operations, vessels typically rely on onboard auxiliary engines to supply electricity for hoteling and auxiliary services, resulting in continuous emissions of CO₂ and local air pollutants such as NOₓ, SOₓ, and particulate matter [2]. These emissions directly affect port workers, nearby urban populations, and coastal ecosystems.

Ports have, therefore, emerged as strategic leverage points for accelerating maritime decarbonization. Compared to propulsion-related measures, which often require long asset lifetimes and vessel retrofitting, port-side interventions can deliver immediate emission reductions during port stays. Shore power, also referred to as cold ironing or onshore power supply (OPS), enables berthed vessels to switch off auxiliary engines and connect to shore-based electricity. Previous studies have demonstrated that OPS can significantly reduce local air pollution and noise, while also contributing to GHG mitigation when electricity is supplied from low-carbon sources [3].

The importance of shore power has been reinforced by an increasingly stringent regulatory framework. At the international level, the International Maritime Organization (IMO) has introduced progressively tighter carbon intensity targets for shipping [4], while at the regional level the European Union has adopted regulatory instruments such as the FuelEU Maritime Regulation [5] and the Alternative Fuels Infrastructure Regulation (AFIR) [6], which mandate the availability and use of shore power for specific vessel categories within defined timelines. Although the regulations are less stringent, interest in these solutions are observable also in North America and parts of Asia, where port electrification and shore power requirements are increasingly embedded within broader decarbonization and air quality strategies [7,8].

This global convergence of regulatory drivers underscores that the challenge of shore power deployment represents a common challenge across many ports globally.

1.2. Limitations of Conventional Cold Ironing Implementations

Despite its environmental benefits and strong regulatory support, the large-scale deployment of conventional grid-connected shore power systems faces several structural limitations. A key challenge is the limited capacity of local electrical grids in many port areas. Supplying OPS to large vessels may require several megawatts per berth, often exceeding available grid capacity and necessitating costly and time-consuming infrastructure reinforcements [9,10,11].

The environmental effectiveness of OPS also is highly context dependent. In regions where electricity generation remains carbon-intensive, shore power may primarily shift emissions from the transportation to the power sector, resulting in limited net climate benefits despite improvements in local air quality [12].

Economic barriers further complicate OPS adoption. Electricity supplied through shore power from the local grid can be more expensive than onboard power generation, particularly in contexts characterized by high electricity prices or unfavorable tariff structures. As a result, the economic viability of OPS often depends on regulatory mandates or financial incentives rather than intrinsic cost competitiveness [11,13,14,15]. These challenges are widely recognized in regulatory impact assessments and industry reports.

Taken together, these limitations suggest that while conventional shore power is essential to meet regulatory requirements, it may not represent the most effective solution under all operating conditions , particularly where grid constraints and cost pressures are significant. In grid-constrained environments, relying exclusively on external electricity supply may not only delay implementation but also increase dependence on upstream infrastructure over which they have limited operational control.

1.3. Integrated Energy Systems and Microgrids as Enabling Concepts

In response to the limitations of single-vector electrification approaches, energy systems literature increasingly has emphasized the role of integrated or multi-energy systems. These systems are based on the coordinated management of multiple energy vectors, such as electricity and heat, with the objective of improving overall efficiency, flexibility, and resilience [16]. The concept has emerged to overcome inefficiencies associated with isolated sectoral planning and to facilitate the integration of variable renewable energy sources. At the local scale, microgrids may represent an operational architecture enabling the deployment of integrated energy systems by combining distributed generation, energy storage, and controllable loads under advanced control schemes [17]. These systems facilitate coordinated operation under both grid-connected and islanded conditions.

Within this framework, cogeneration (combined heat and power, CHP) can play an important role by simultaneously producing electricity and useful thermal energy, thereby increasing overall system efficiency compared to separate production. In port environments, where electrical loads (e.g., shore power, terminal operations) and thermal demands (e.g., building heating and cooling) coexist, cogeneration and trigeneration can help address infrastructure constraints while improving overall energy efficiency, reducing dependence on external grid reinforcements while recognizing the importance of waste heat streams [18].

1.4. Ports as Multi-Energy Systems: Research Gap and Contribution

Building on these concepts, recent studies have increasingly framed ports as complex energy systems rather than passive electricity consumers. Ports host diverse and energy-intensive activities, including terminal operations, building services, and vessel hoteling, which involve multiple energy vectors and highly variable load profiles. Several contributions have explored the application of microgrids and integrated energy management in port environments, highlighting their potential to enhance energy efficiency, resilience, and environmental performance [19].

Beyond technical integration, the “port as energy hub” perspective also encompasses operational and strategic dimensions: improved resilience against grid disturbances, enhanced control over energy costs, and increased autonomy in energy supply decisions [17]. Such a framing repositions ports from mere infrastructure nodes to active energy actors capable of managing production, consumption, and storage in a coordinated manner.

However, existing studies often focus on specific aspects, such as electrical microgrids or operational scheduling, and rarely provide comprehensive techno-economic and environmental comparisons between conventional shore power, onboard generation, and fully integrated multi-energy solutions [3,19]. Many contributions remain confined to purely electrical optimization frameworks, neglecting thermal integration and cogeneration potentials, while others rely on stylized or synthetic load data that limit the assessment of real-world infrastructural constraints. Moreover, the role of cogeneration systems in supporting shore power under grid-constrained conditions remains underexplored, despite their potential to supply electricity while simultaneously meeting port-side thermal demands [12].

This study addresses this gap by presenting a case study-based assessment of alternative cold ironing configurations for a grid-constrained port. By explicitly comparing conventional onboard power generation and an integrated microgrid incorporating cogeneration, renewable energy, and storage, the paper provides quantitative evidence on cost competitiveness, emissions reduction, and operational resilience under realistic infrastructure constraints.

1.5. Structure of the Paper

The remainder of the paper is organized as follows. Chapter 2 describes the research design, system configurations, and modeling framework. Chapter 3 presents the results of the base case assessment and sensitivity analyses. Chapter 4 discusses the findings in relation to existing literature and explores policy and planning implications. Chapter 5 concludes the paper and outlines directions for future research. Appendix A.1 and Appendix A.2 describe the modeling of the electrical and thermal loads.

2. Materials and Methods

2.1. Research Design and Analytical Framework

This study adopts a case study–based techno-economic framework to evaluate alternative energy supply strategies for cold ironing in a grid-constrained port environment. In this context, the analysis investigates whether an integrated on-site microgrid can provide a viable and scalable solution. The microgrid configuration is designed to compensate for the limited grid availability (3 MW) while meeting peak shore power demand of up to 20 MW through a combination of cogeneration units, photovoltaic generation, battery energy storage, and coordinated control systems.

Given that a grid-only OPS configuration cannot satisfy the full demand under the imposed infrastructure constraint, the comparison focuses on two feasible paradigms: (i) onboard auxiliary engine generation and (ii) integrated port-based generation through a multi-vector microgrid.

The system architecture is dimensioned ex ante based on engineering feasibility and operational requirements. Its performance then is evaluated through time-resolved simulation over a full annual cycle. The methodological structure clearly separates input data, dispatch modeling, and performance evaluation metrics, helping to ensure transparency and internal consistency in the comparison between configurations.

The methodological approach, depicted in Figure 1, consists of:

• Modeling annual electrical and thermal load profiles for the port;
• Defining two alternative energy supply architectures under identical demand conditions;
• Performing time-resolved dispatch simulation over 8,760 hours with a time resolution of 1 hour;
• Evaluating economic, environmental, and operational performance;
• Testing robustness through structured sensitivity analysis.

2.2. Case Study Context and Energy Demand Modeling

The reference case represents a commercial port operating under a maximum external grid capacity of 3 MW. Peak vessel demand during cold ironing operations reaches approximately 20 MW, creating a structural mismatch between required and available power. This mismatch defines the core problem addressed in this study: how to supply large berth loads without grid reinforcement.

To answer this research question, as a first step it is necessary to construct energy load profiles. In particular, both electrical and thermal load profiles are constructed over a full annual horizon (8,760 hours).

2.2.1. Electrical Load Modeling

The electrical load profile is constructed using the methodology described in Appendix A.1.

The model incorporates:

• Four vessel categories (container, cruise, RoRo/ferry, tanker);
• Vessel-specific power ranges and connection durations;
• Port baseline consumption (150–300 kW);
• Seasonal, weekly, and daily variation factors;
• Capacity constraints preventing unrealistic berth overlap.

The electrical load profile, depicted in Figure 2, exhibits high intermittency due to vessel arrival patterns, reinforcing the need for flexible dispatch and storage integration.

2.2.2. Thermal Load Modeling

The thermal load profile is modeled using the methodology detailed in Appendix A.2. The modeled terminal area is 10,000 m², with annual energy intensities assigned to:

• Space heating;
• Domestic hot water;
• Space cooling;
• Process heat.

Thermal demand follows seasonal and hourly distribution patterns and traditionally is supplied by gas boilers (55 MW_th installed capacity).

The inclusion of thermal demand is critical because it enables evaluation of the port as a multi-vector energy system. The possibility of recovering waste heat from cogeneration fundamentally alters system efficiency and economic performance.

2.3. Energy System Configurations

As previously introduced, two alternative configurations to fulfill electrical and thermal demands are evaluated:

• Configuration A – Onboard Auxiliary Engine Generation: In the baseline configuration, vessels continue to rely on onboard auxiliary engines. This configuration serves as the reference baseline for both economic and environmental comparison. Key parameters of configuration A are summarized in Table 1.

• Configuration B – Integrated Port Microgrid: The alternative configuration consists of an integrated on-site microgrid designed to overcome grid capacity limitations while maintaining redundancy and operational robustness. The selection of six 4.5 MW CHP units enables compliance with the N+1 (referred to the number of on-site power generation units, CHP) redundancy criterion to improve the microgrid availability in case of utility grid failure and allows operation within a 50% – 100% loading window. The battery system is dimensioned to:
  • Manage transient vessel connection surges in lower load conditions (< 2 MW);
  • Provide about 2 hours of critical load support;
  • Enable load shifting to maintain generator efficiency.

Thermal integration includes heat recovery from all CHP units and buffering through thermal storage, reducing boiler gas consumption by approximately 10%. This configuration is not merely an electrical solution; it redefines the port as an integrated energy hub capable of managing electricity and heat simultaneously. Key information describing configuration B is summarized in Table 2. For clarity, a graphical representation of configuration B is reported in Figure 3. The solar capacity is limited to 2 MW due to assumed local space constraints on the terminal roofs.

2.4. Simulation Model and Input Parameters

The techno-economic assessment was conducted using a commercial microgrid simulation and optimization software environment. The simulation horizon covers a full calendar year (8,760 hours) with hourly resolution. The modeling approach is deterministic, and time-series based. The optimization framework is based on a cost-minimization objective subject to component-level technical and operational constraints. The algorithm simultaneously determines the optimal system configuration and its operational strategy by minimizing the total life-cycle cost of the microgrid. This objective function incorporates both capital expenditures (CAPEX), associated with the sizing of generation, storage, and supporting infrastructure, and operating expenditures (OPEX), including fuel consumption and electricity purchases. Environmental criteria are not explicitly included in the optimization process. Consequently, environmental performance indicators are evaluated ex post based on the simulated energy flows and the corresponding emission factors associated with each energy source.

Electrical and thermal demands are defined exogenously, computed according to the methodology discussed in Section 2.2, and held constant within each run. The same applies to the technical parameters relating to CHP, PV, and BESS, which are summarized respectively in Table 3, Table 4, and Table 5.

Additional aspects defined in the simulation are:

• For the CHP units, operating range is constrained to 50%-100% of rated load (minimum stable loading assumption);
• CHP ramp rate constraints were applied according to representative manufacturer specifications (implemented as operational limits in the dispatch logic);
• The recoverable thermal energy from CHP units was modeled based on manufacturer performance specifications. The maximum available thermal output corresponds to the recoverable heat from exhaust gases and cooling circuits under operating conditions;
• PV is modeled as non-dispatchable with priority of dispatch within the microgrid;
• Grid import is capped at 3 MW;
• Thermal demand not covered by heat recovery is supplied via conventional boilers, with thermal storage used to buffer heat recovery and improve utilization;
• No grid export is allowed. Consequently, in cases where generation exceeds instantaneous electrical demand and storage capacity, surplus electricity is assumed to be curtailed;
• Heat recovery is limited by instantaneous demand and storage capacity. Consequently, excess thermal energy beyond storage limits is assumed to be dissipated and not credited in performance calculations.

A deterministic 24-hour grid outage scenario is simulated to evaluate resilience. No probabilistic reliability modeling is performed.

The simulation framework is parameterized using a consistent set of technical, economic, and environmental assumptions reflecting representative operating conditions for a grid-constrained commercial port. All input parameters are applied uniformly across the configurations analyzed to help ensure that performance differences arise exclusively from the energy supply architecture.

Key economic input parameters are summarized in Table 6. Capital costs are amortized over 20 years, and no residual value is considered.

For the model to provide environmental performance, key parameters related to emission factor are necessary. Key emissions-related input parameters are summarized in Table 7.

2.5. Performance Indicators

System performance is evaluated through a multidimensional framework encompassing economic, environmental, thermal, and operational criteria. This integrated evaluation approach reflects the multi-vector nature of the investigated system and enables a consistent comparison between onboard auxiliary engine generation and the integrated microgrid configuration.

From an economic perspective, the primary metric adopted is the levelized cost of energy (LCOE), which provides a comprehensive measure of lifetime electricity supply cost by integrating capital expenditures and operating costs over the project horizon. LCOE is defined by the following equation:

$$LCOE={\displaystyle \frac{CAPEX+{\displaystyle {\sum }_{t=1}^n}(OPE{X}_t+Fue{l}_t+Gri{d}_t)}{{\displaystyle {\sum }_{t=1}^n}{E}_t}}$$

(1)

where $CAPEX$represents total capital investment; $OPE{X}_t$, annual operational expenditures; $Fue{l}_t,$fuel costs; $Gri{d}_t,$purchased electricity costs; $E_t,$electricity supplied to vessel demand; and $n,$ the project lifetime. In addition to LCOE, total annual operating cost and fuel expenditure are quantified to provide further insight into cost structure. The relative cost reduction compared to the onboard baseline configuration is also computed to express economic competitiveness in percentage terms. For simplicity, the LCOE formulation adopted in this study does not apply discounting. All costs are expressed in real annual values and aggregated over the project lifetime.

Environmental performance is assessed within a Scope 1 and Scope 2 boundary. Scope 1 includes direct fuel combustion emissions from auxiliary engines (configuration A) and CHP units (configuration B), while Scope 2 accounts for indirect emissions associated with purchased grid electricity. The evaluation includes total annual CO₂ emissions (tCO₂/year), emission intensity expressed in kgCO₂ per kWh supplied, and percentage emission reduction relative to the onboard baseline. Upstream emissions related to fuel extraction, processing, transportation, and infrastructure manufacturing (Scope 3) are not included in this assessment.

Given the integrated nature of the proposed microgrid (configuration B), specific indicators are introduced to capture the effects of thermal integration. These include the percentage reduction in boiler fuel consumption due to heat recovery, the effective utilization rate of recovered thermal energy, and the combined electricity and heat cost savings achieved through cogeneration. These metrics explicitly quantify the added value of treating the port as a multi-vector energy system rather than as a purely electrical load.

Finally, operational performance is evaluated through an assessment of system resilience. Specifically, a grid outage scenario was simulated assuming a 24-hour interruption of the external electricity supply. During this period, the microgrid operates in island mode, with load supplied by CHP units and battery storage within their technical operating constraints. The system configuration follows an N+1 redundancy criterion (related to the amount of on-site power generation equipment), where installed generation capacity exceeds peak demand even in the event of a single unit failure. This helps ensure continuity of supply for critical shore power loads under grid-disconnected operation. No probabilistic reliability assessment (e.g., SAIDI/SAIFI-based modeling) was conducted; resilience assessment is based on deterministic outage simulation. Thus, although resilience is not quantified through probabilistic reliability metrics, these indicators provide a structured basis for assessing operational continuity under constrained infrastructure conditions. It should be noted that the resilience assessment adopted in this study reflects the specific design requirements of the analyzed case study. Depending on the characteristics of the port infrastructure and the objectives of the analysis, alternative availability and reliability assessments may be considered (e.g., different outage durations, multiple equipment failure scenarios or stress tests involving simultaneous disruptions affecting generation). Therefore, the methodology presented herein should be interpreted as one possible resilience assessment framework, which can be adapted and extended according to the operational requirements and risk profile of the specific port under investigation.

2.6. Sensitivity Analysis Design

A univariate (one-factor-at-a-time) approach was adopted. In each sensitivity run, only one parameter was changed while all others were kept at the base-case value. This structure improves interpretability and allows isolation of the marginal impact of each driver. Indeed, no multivariate or probabilistic uncertainty analysis was implemented, as the objective is to provide a clear, decision-oriented understanding of the robustness of results rather than exhaustive uncertainty propagation.

The selected parameters reflect key uncertainties that directly affect the feasibility of shore power deployment in grid-constrained ports. Specifically, to conduct the sensitivity analysis, the following parameters were varied:

• Grid electricity price;
• Grid carbon intensity;
• CHP CAPEX;
• Battery Energy Storage System (BESS) CAPEX;
• Natural gas price.

Each parameter was varied using a low–base–high structure to capture realistic uncertainty bands, consistent with common energy infrastructure planning practice. Percentage-based ranges are used where cost uncertainty dominates (prices and CAPEX). Absolute values are used for grid carbon intensity to reflect real-world differences between electricity systems. Table 8 provides a general overview of the scenarios that have been considered in the sensitivity analysis.

The following outputs were evaluated for each sensitivity scenario, both in absolute terms and in percentage reduction from the base case scenario:

• Levelized cost of energy (LCOE);
• Total annual system cost (fuel + grid electricity + fixed OPEX);
• Annual CO₂ emissions (Scope 1 + Scope 2 boundary).
• Percentage CO₂ reduction relative to the onboard auxiliary engine baseline configuration.

3. Results

3.1. Base Case Scenario

This section presents the base case scenario results for the two configurations analyzed under identical demand conditions. All results are computed over a full annual horizon (8,760 hours). Moreover, to provide insights regarding the microgrid functioning, Paragraph 3.1.2 complements the annual outputs with a representative weekly operational extract, also illustrating the system behavior in terms of energy flows and dispatching.

3.1.1. Configuration A: Onboard Auxiliary Engine Generation

In configuration A, vessels meet electricity demand, equal to 46.21 GWh on yearly base, through onboard auxiliary engines during port stays. Using the adopted specific fuel consumption and conversion assumptions, annual marine diesel oil (MDO) consumption is approximately 10,166 metric tons. Applying the selected emission factor, total annual CO₂ emissions are approximately 32,592 tCO₂/year. The corresponding total annual cost for onboard electricity supply is approximately 9,326,100 €, which translates into an average unit cost of about 0.202 €/kWh for the electricity generated onboard. These values, summarized in Table 9, provide the reference baseline for the comparative assessment.

3.1.2. Configuration B: Integrated Port Microgrid

In the integrated microgrid configuration, port-side generation and flexibility assets supply shore power demand under a hard grid import constraint. Annual electricity supply is primarily provided by the CHP units, complemented by solar PV generation and limited grid imports within the maximum import cap. In the base case, the annual electricity balance indicates that most energy is produced on site by cogeneration, while PV and grid imports contribute smaller shares of total energy served.

Within the assumptions adopted in this case study, the microgrid configuration achieves a reduction in levelized electricity supply cost compared to onboard generation, with an LCOE equal to 0.157 €/kWh. Specifically, the LCOE reduction is approximately 22% relative to the onboard baseline, leading to an absolute reduction equal to 0.045 €/kWh. In this case, total annual cost for energy procurement (fuel + electricity) is equal to 6.3 million €.

Under the modeled conditions, total annual CO₂ emissions are lower compared to onboard generation. Specifically, microgrid annual emissions are approximately 20,702 tCO₂/year, corresponding to a reduction of approximately 36% relative to the onboard baseline.

Outputs related to configuration B are summarized in Table 10, while Figure 4 graphically shows the comparison between the two configurations respectively regarding economic and environmental performance.

To illustrate the operational behavior of the integrated microgrid under realistic load variability, results are examined both at the annual level and through a representative weekly dispatch extract.

Over the full annual horizon (8,760 hours), the microgrid supplies vessel demand primarily through on-site cogeneration, with PV generation and limited grid imports complementing the supply mix. The annual electricity balance, illustrated in Figure 5, shows that most energy (62%) is produced by the CHP units, with grid imports consisting of the second supply source (33%), despite being constrained by the 3 MW infrastructure cap. PV contributes a smaller share, equal to 5%. Total annual energy supplied amounts to approximately 54.6 GWh, including vessel demand and baseline port loads.

To provide greater insight into dynamic system behavior, a representative week (168 hours) was extracted from the simulation results. During this week, electrical load varies significantly, with a maximum observed peak of 12.692 MW and an average load of approximately 5.293 MW. The dispatch logic follows the cost-minimizing objective defined in Chapter 2. Figure 6 illustrates how the electricity demand during the analyzed week is met, highlighting the specific technologies used to satisfy the load. According to model outputs, most electrical demand is fulfilled by CHP engines (55%) and the grid (37%), with a smaller contribution provided by the PV plant (7%). The BESS contribution is negligible.

With reference to the PV plant, during the selected week it reaches a maximum hourly output of approximately 1.59 MW. An analysis of the energy flows (Figure 7) shows that all PV generation is allocated to meeting on-site demand, either directly or indirectly through BESS, with no share being curtailed.

The battery energy storage system plays a secondary role in operational flexibility. Over the representative week, the battery discharges for 24 hours and charges for 20 hours. The maximum hourly discharge power reaches approximately 0.94 MW, while maximum charging power is approximately 1.053 MW. Figure 8 shows the BESS profile, assuming positive power values for the charging phase, and negative ones for the discharging phase. The BESS state of charge, constrained within the predefined operational window (20% - 80%), is represented in Figure 9.

A deterministic 24-hour grid outage scenario also was simulated to evaluate operational continuity under infrastructure failure. During the outage, the system transitions to island mode, with CHP units and battery storage supplying the entire load within their technical operating constraints. No load shedding occurs during the simulated outage, and the battery provides critical support during generator startup and transient phases. These results indicate that the configuration satisfies the N+1 redundancy criterion and maintains shore power availability even in the absence of grid support.

3.2. Sensitivity Analysis Results

To evaluate the robustness of the base-case configuration, a one-factor-at-a-time sensitivity analysis was conducted. In each scenario, a single parameter was varied while all others were held constant at their base-case values. The dispatch algorithm remained cost-minimizing in all simulations, and performance indicators were recalculated consistently with the methodology described in Chapter 2. The sensitivity baseline corresponds to an LCOE of approximately 0.157 €/kWh and annual CO₂ emissions of 20,702 tCO₂/year. Table 11 provides a general overview of how the input parameters were varied to conduct the sensitivity analysis, while Table 12 and Figure 10 show results, commented on in the following lines.

3.2.1. Grid Electricity Price

A ±30% variation in grid electricity tariff produces the largest observed variation in LCOE. Under the −30% scenario, LCOE decreases to 0.137 (-13%), whereas under the +30% scenario it increases to 0.177 (+13%). Annual CO₂ emissions remain unchanged at 20,702 tCO₂/year in both cases. This indicates that tariff variability directly affects economic performance but does not modify the simulated energy mix under cost-minimizing dispatch.

3.2.2. CHP CAPEX/OPEX

Variations in CHP capital and operational expenditures (−10% / +20%) produce moderate changes in LCOE, ranging from 0.154 (-2%) in the low-cost scenario to 0.162 (+3%) in the high-cost scenario. Annual CO₂ emissions remain constant at 20,702 tCO₂/year across these cases, confirming that investment cost variability affects levelized cost but not environmental performance within the modeled boundary.

3.2.3. BESS CAPEX

A ±20% variation in battery capital costs results in minimal changes in LCOE, from 0.156 to 0.157. Annual CO₂ emissions remain unchanged at 20,702 tCO₂/year. These results suggest that within the analyzed configuration, battery cost uncertainty has a limited impact on overall economic outcomes.

3.2.4. Natural Gas Price

A ±20% variation in natural gas price produces intermediate LCOE sensitivity, with values ranging from 0.147 to 0.156. Annual CO₂ emissions remain largely stable, with minor variation between 20,702 and 20,793 tCO₂/year. This reflects the significant role of CHP in the supply mix while indicating that fuel price fluctuations do not materially alter emission outcomes under fixed operational constraints.

3.2.5. Grid Carbon Intensity

Grid carbon intensity is the primary environmental driver. When the emission factor is reduced to 150 gCO₂/kWh, annual emissions decrease to approximately 15,350 tCO₂/year (-26%). Conversely, when the grid carbon intensity increases to 600 gCO₂/kWh, annual emissions rise to approximately 21,295 tCO₂/year (+3%). LCOE values associated with these scenarios are 0.170 (150 gCO₂/kWh) and 0.121 (600 gCO₂/kWh), consistent with the parameter structure adopted in the sensitivity dataset (this result reflects the assumptions embedded in the sensitivity dataset and does not imply a direct causal relationship between carbon intensity and electricity cost). This variation reflects an implicit carbon cost embedded in the electricity pricing structure under the tested carbon-intensity scenarios, rather than a direct mechanistic link within Equation (1).

Overall, the sensitivity analysis indicates that, within the explored parameter ranges, economic performance primarily is influenced by the conditions of the national grid, both in terms of electricity tariff and carbon intensity. Figure 10 shows the absolute variation of LCOE according to different parameters’ scenarios. As for environmental performance, it is strongly dependent on this latter factor. Technology cost uncertainty within the tested ranges produces comparatively limited variation in key performance indicators.

4. Discussion

4.1. Interpretation of Results

The results indicate that the integrated microgrid configuration achieves better performance within the scenario analyzed across economic, environmental, and operational dimensions under grid-constrained conditions. The most significant finding is that the proposed system achieves a marked reduction in LCOE while simultaneously lowering CO₂ emissions and maintaining full operational compliance with peak demand and redundancy requirements.

From an economic perspective, the approximately 22% reduction in levelized cost relative to onboard generation reflects the higher conversion efficiency of natural gas cogeneration systems compared to marine auxiliary engines. While onboard engines operate at average efficiencies around 30% - 35%, the CHP units modeled in this study achieve total system efficiencies exceeding 85% due to the simultaneous production of electricity and useful heat. The economic advantage, therefore, arises not only from fuel substitution but from structural efficiency gains inherent to cogeneration technology.

Thermal integration further enhances system performance. Although the reduction in installed boiler capacity (55 MW_th to 50 MW_th) may appear modest, the associated 10% reduction in boiler gas consumption produces measurable operational cost savings and emission reductions. This suggests that the value of cogeneration in port environments lies not only in electrical supply substitution but in its ability to couple electrical and thermal energy flows within a unified system architecture.

Operationally, the microgrid was able to meet a 20 MW peak demand under a 3 MW grid cap, supporting the study hypothesis that integrated on-site generation can help address grid bottlenecks. The representative weekly dispatch results illustrate stable coordination between CHP, grid imports, and PV, while battery systems play a secondary role. Furthermore, deterministic outage simulations confirm that the system maintains supply continuity in island mode under N+1 redundancy criteria.

The sensitivity analysis reveals a structural separation between economic and environmental drivers. Economic performance primarily is influenced by grid electricity tariffs and carbon intensity, with this latter strongly affecting environmental performance. Technology cost uncertainty (CHP and BESS CAPEX) exerts comparatively limited influence within realistic ranges, indicating robustness of the proposed configuration under moderate investment cost fluctuations.

Taken together, these findings suggest that, within the assumptions of the analysed case study, the integration of cogeneration within a port microgrid does not merely provide incremental improvement over conventional OPS but represents an alternative energy system approach capable of addressing both infrastructure constraints and multi-vector energy demands.

4.2. Comparison with Literature

The findings of this study align with and extend prior research on shore power limitations and integrated port energy systems.

Studies examining conventional shore power implementation consistently highlight grid capacity constraints, tariff barriers, and variability in environmental benefits as structural challenges [10,11,13]. The present study results are consistent with these limitations under a realistic grid cap scenario and suggest that relying exclusively on external grid electricity may not provide a scalable solution in constrained ports.

The literature on microgrids and integrated energy systems has emphasized the benefits of multi-vector coordination and distributed generation in improving flexibility and resilience [16,18]. However, many prior contributions focus either on electrical microgrids in isolation or on abstract system optimization without explicit techno-economic comparison against onboard baselines [3,19]. By directly comparing onboard generation with a fully integrated CHP-based microgrid under identical load conditions, this study provides a quantitative bridge between conceptual integrated energy frameworks and real-world port applications.

Moreover, existing port-focused studies often concentrate on operational scheduling or grid-connected microgrid optimization without explicitly modeling severe infrastructure constraints [3,19]. The present analysis explicitly embeds a hard 3 MW grid cap, thereby modeling conditions frequently observed in practice but less frequently represented in the literature.

The sensitivity results further contribute to the literature by clarifying that external electricity market conditions represent the dominant economic risk factor, whereas technology cost uncertainty plays a secondary role. This finding complements previous techno-economic analyses of shore power [9] and supports the argument that tariff structure and energy market exposure are critical determinants of project viability.

Overall, the study expands the existing body of knowledge by quantitatively indicating that cogeneration-based microgrids may help address grid limitations while supporting efficiency gains and integrated energy optimization.

4.3. Policy and Planning Implications

The results have several implications for port authorities, regulators, and infrastructure planners.

First, grid reinforcement may not always be the most suitable pathway for compliance with shore power mandates. In ports where grid expansion is costly or delayed, integrated on-site generation may provide a more immediately deployable alternative.

Second, tariff design and long-term electricity procurement mechanisms are critical to economic viability. The sensitivity analysis shows that electricity price variability has a stronger effect on LCOE than moderate changes in technology CAPEX. This suggests that regulatory stability and predictable tariff frameworks may be more important than short-term equipment cost fluctuations.

Third, grid decarbonization policy directly influences port-level emission outcomes. Even under a cost-minimizing dispatch strategy, residual grid imports expose the port system to upstream carbon intensity. Therefore, coordinated national energy and maritime policy frameworks may play an important role in supporting emission reductions.

Fourth, the results support the “port as energy hub” perspective introduced in Chapter 1. By integrating electricity and heat production, ports can move beyond compliance-driven electrification toward strategic energy system optimization. This integrated approach may improve resilience, reduce cost volatility exposure, and create pathways for progressive decarbonization, including future hydrogen blending.

4.4. Limitations of the Study

Several limitations should be acknowledged.

First, dispatch optimization is cost-minimizing and deterministic. While this reflects realistic operational priorities, it does not explore alternative objective functions or probabilistic uncertainty modeling.

Second, the sensitivity analysis adopts a univariate structure. Interaction effects between variables, such as simultaneous electricity price and gas price fluctuations, are not modeled.

Third, resilience assessment is qualitative and based on deterministic outage scenarios rather than probabilistic reliability metrics.

Fourth, the case study reflects a specific port scale and geographic context. While structural relationships are likely transferable to other grid-constrained ports, absolute performance values may vary depending on local energy prices, carbon intensity, and demand characteristics.

Despite these limitations, the study provides a structured and transparent evaluation of integrated cogeneration microgrids under realistic infrastructure constraints and contributes quantitative evidence to ongoing discussions on port decarbonization strategies.

5. Conclusions

This study assessed the techno-economic and environmental performance of an integrated port microgrid for cold ironing in a grid-constrained port environment. The analysis considered a case in which the available grid connection is limited to 3 MW, while peak shore power demand can reach approximately 20 MW. Under these conditions, the study compared continued onboard auxiliary engine generation with a port-side microgrid integrating cogeneration units, photovoltaic generation, battery storage, limited grid imports, and heat recovery.

The results show that, within the assumptions of the case study, the integrated microgrid can satisfy the modeled electrical demand under the imposed grid constraint while reducing both energy supply costs and CO₂ emissions compared with onboard generation. In the base case, the microgrid achieves an LCOE of 0.157 €/kWh, compared with 0.202 €/kWh for onboard auxiliary engines, corresponding to a reduction of approximately 22%. Annual CO₂ emissions decrease from 32,592 tCO₂/year to approximately 20,702 tCO₂/year, equal to a reduction of about 36% within the adopted Scope 1 and Scope 2 boundary. The simulated 24-hour grid outage scenario also indicates that the system can maintain supply continuity in island mode under the adopted N+1 redundancy criterion.

Beyond these quantitative benefits, the results highlight a broader system-level implication. The proposed configuration suggests that cogeneration should not be interpreted merely as an alternative electricity generation technology, but rather as an enabling mechanism for integrating multiple energy vectors within a constrained infrastructure environment. By coupling electricity production, thermal recovery, renewable generation, and storage within a coordinated microgrid architecture, the system may help address grid limitations while supporting overall energy system optimization. In this sense, the value of CHP extends beyond efficiency gains alone and lies in its ability to support cold ironing deployment where conventional grid-connected solutions would face significant technical or economic barriers.

The proposed microgrid, then, should not be interpreted solely as a natural gas-based alternative to onboard generation. Rather, it should be viewed as an energy infrastructure platform capable of evolving alongside the decarbonization of fuel supply chains. Subject to technical compatibility, safety requirements, future regulatory developments, economic viability, and fuel availability, the same architecture could progressively integrate renewable and low-carbon fuels such as biomethane, synthetic methane, or hydrogen blends, while preserving the underlying generation, storage, and control infrastructure.

This aspect is particularly relevant in the context of ports increasingly being considered as energy hubs. Ports are increasingly being considered not only as energy consumers for terminal operations and vessel services, but also to manage, distribute, store, and potentially produce alternative energy carriers for both maritime and landside applications. The integrated microgrid analyzed in this study is consistent with this perspective, since it combines electricity generation, grid imports, storage, and heat recovery within a coordinated local energy system. Its contribution, therefore, is not limited to reducing emissions during berthing operations but also supports greater operational autonomy and prepares port energy infrastructure for future energy transitions.

Future research should extend the proposed framework to alternative fuel scenarios, including renewable gases and hydrogen-compatible technologies, as well as broader life-cycle emission boundaries, multivariate uncertainty analysis, and different port typologies. Additional research also may investigate the interaction between port microgrids and wider energy systems, including participation in flexibility markets, sector coupling opportunities, and integration with emerging regional energy infrastructures.

Overall, the study demonstrates that integrated port microgrids represent a technically feasible and economically competitive pathway for supporting cold ironing in ports where grid reinforcement is constrained, delayed, or costly, under the conditions examined in this study. Their value should not be assessed solely in terms of current cost and emission reductions, but also in terms of infrastructure flexibility and long-term adaptability. If properly designed, port-side energy assets deployed today for cold ironing can become foundational elements of future port energy systems, supporting the transition from conventional electrification strategies toward integrated multi-energy hubs capable of coordinating renewable electricity, thermal recovery, storage, and progressively decarbonized fuels.

Appendix A.1

To model shore power electric load profile, the following elements were considered:

• Vessel classification parameters: The load profile incorporates four primary vessel categories with distinct operational characteristics. All relevant information is reported in Table A1.
• Port baseline load: The model incorporates a realistic baseline load to represent power consumption when no vessels are connected to shore power. This baseline helps ensure that the model reflects the reality that shore power infrastructure always consumes some power, even when no vessels are connected. All relevant information is reported in Table A2.
• Port capacity constraints: The model includes operational constraints that limit simultaneous connections and total power capacity. These constraints prevent unrealistic scenarios where excessive vessel arrivals would create power demands beyond practical port capabilities. All relevant information is reported in Table A3.
• Core modeling approach:

  iv.

  Probabilistic vessel arrivals: based on seasonal and weekly patterns;

  v.

  Capacity constraints: limited berths and total power capacity;

  vi.

  Baseline load: always-on infrastructure power consumption;

  vii.

  Full-year simulation: 8,760 hourly load values (365 days × 24 hours)

  viii.

  Calibration: adjusted to enable the system to operate below 50% capacity for 70% - 80% of hours.

• Temporal patterns:

  ix.

  Seasonal factors: higher cruise activity in summer; container peak in fall;

  x.

  Weekly patterns: cruise/ferry weekend focus; container weekday preference;

  xi.

  Daily baseline: higher during business hours (8 am – 4 pm), lower at night.

Key features of the modeled electrical load are:

• The shore power infrastructure itself (transformers, control systems, standby equipment) accounts for approximately 4% of the total annual energy consumption;
• Vessel power consumption represents about 96% of the total annual energy;
• During periods of low port activity (especially in winter), the baseline infrastructure load can represent a higher percentage of the instantaneous power consumption.

Figure A1 graphically represents the load duration curve. While beyond the scope of the current work, future enhancements could include:

• Dynamic vessel load profiles: implementing time-varying load profiles based on operational phases;
• Weather integration: adding weather-related disruption patterns;
• Berth scheduling logic: implementing more realistic scheduling constraints;
• Grid interaction modeling: incorporating demand response and grid constraints;
• Economic decision modeling: adding price-responsive connection decisions;
• Machine learning integration: implementing adaptive parameters based on operational data;

These enhancements would require additional data collection, stakeholder engagement, and computational resources for effective implementation.

Table A1. Operational characteristics of four vessel categories.

Vessel Type	Power Range (MW)	Connection Duration (hours)	Load Factor (%)	Justification
Container	1 - 7.5	8 - 36	60% - 65%	Based on typical container vessel sizes (1,000 - 14,000 TEU) with higher power demands during cargo operations and refer container support.
Cruise	4 - 12	6 - 14	50% - 55%	Reflects the significant hotel load requirements for passenger vessels with shorter port stays typical of cruise itineraries.
RoRo/Ferry	0.8 - 2.5	2 - 10	40% - 45%	Lower power requirements due to smaller vessel size and minimal cargo refrigeration needs; shorter stays typical of scheduled services.
Tankers	1.5 - 4	24 - 72	55 - 60%	Moderate power requirements with extended port stays due to longer cargo operations and regulatory inspections.

Table A1. Operational characteristics of four vessel categories.

Vessel Type	Power Range (MW)	Connection Duration (hours)	Load Factor (%)	Justification
Container	1 - 7.5	8 - 36	60% - 65%	Based on typical container vessel sizes (1,000 - 14,000 TEU) with higher power demands during cargo operations and refer container support.
Cruise	4 - 12	6 - 14	50% - 55%	Reflects the significant hotel load requirements for passenger vessels with shorter port stays typical of cruise itineraries.
RoRo/Ferry	0.8 - 2.5	2 - 10	40% - 45%	Lower power requirements due to smaller vessel size and minimal cargo refrigeration needs; shorter stays typical of scheduled services.
Tankers	1.5 - 4	24 - 72	55 - 60%	Moderate power requirements with extended port stays due to longer cargo operations and regulatory inspections.

Table A2. Relevant information related to port baseline load.

Parameter	Value Range	Justification
Minimum baseline	150 kW	Represents standby power for shore power connection points, control systems, port infrastructure, and transformer losses
Maximum baseline	300 kW	Accounts for peak background consumption during port operations
Daily pattern	Varies by hour (0.7 - 1.3 multiplier)	Higher demand during daytime working hours (8 am – 4 pm), lower demand during night hours, with gradual transitions

Table A2. Relevant information related to port baseline load.

Parameter	Value Range	Justification
Minimum baseline	150 kW	Represents standby power for shore power connection points, control systems, port infrastructure, and transformer losses
Maximum baseline	300 kW	Accounts for peak background consumption during port operations
Daily pattern	Varies by hour (0.7 - 1.3 multiplier)	Higher demand during daytime working hours (8 am – 4 pm), lower demand during night hours, with gradual transitions

Table A3. Relevant information related to port capacity constraints.

Constraint Type	Value	Justification
Max Container Vessels	2	Typical berth availability at medium-sized container terminals
Max Cruise Vessels	1	Limited dedicated cruise berths at most ports
Max RoRo/Ferry Vessels	2	Typical berth configuration for roll-on/roll-off operations
Max Tanker Vessels	1	Limited hazardous cargo berths with shore power capability
Max Total Power	20 MW	Typical shore-side infrastructure capacity considering substation and distribution limitations

Table A3. Relevant information related to port capacity constraints.

Constraint Type	Value	Justification
Max Container Vessels	2	Typical berth availability at medium-sized container terminals
Max Cruise Vessels	1	Limited dedicated cruise berths at most ports
Max RoRo/Ferry Vessels	2	Typical berth configuration for roll-on/roll-off operations
Max Tanker Vessels	1	Limited hazardous cargo berths with shore power capability
Max Total Power	20 MW	Typical shore-side infrastructure capacity considering substation and distribution limitations

Figure A1. Load duration curve.

Figure A1. Load duration curve.

Appendix A.2

To model port terminal thermal load profile, the following elements were considered:

• Annual energy intensity parameters: Annual energy intensity benchmarks reported in Table A4 were used as baseline values. These values represent the middle of the documented ranges for a typical Mediterranean port terminal. For process heat, which varies significantly by port type and industrial integration, a reasonable estimate was derived based on typical auxiliary thermal demands for port operations (focusing on terminal buildings only).
• Terminal size assumption: A reference terminal size of 10,000 m² was selected for the model. This size represents a medium-scale terminal facility and allows for easy scaling of results for different sized terminals. According to this size, a set of thermal needs were selected. All relevant information is reported in Table A5.
• Seasonal distribution:

  xii.

  Heating: winter-focused (Jan - Mar, Oct - Dec), peak in January;

  xiii.

  Cooling: summer-focused (May - Sep), peak in July - August;

  xiv.

  Domestic hot water (DHW): relatively consistent, slight winter increase;

  xv.

  Process heat: consistent year-round with minor variations.

• Daily variation:
  • Weekdays: 110% of average daily load;
  • Weekends: 80% of average daily load.
• Hourly patterns:
  • Heating: bimodal with morning and evening peaks;
  • Cooling: mid-day peak (10 am – 3 pm);
  • DHW: peak at 8 am and 12 pm – 1 pm;
  • Process heat: business hours concentration (8 am - 5 pm).
• Calculation process:
  • Annual demand allocated to months using seasonal factors;
  • Monthly demand distributed to days with weekday/weekend adjustments;
  • Daily demand distributed to hours using normalized patterns.

The methodology used to model the thermal load profile has some limitations, the most relevant of which are:

• Represents typical Mediterranean port terminal (specific ports may vary);
• Does not account for extreme weather events or operational anomalies;
• Process heat demand requires adjustment for specific port industrial processes.

Table A4. Energy intensity benchmarks used as baseline values.

Thermal Demand Type	Range from Document	Selected Value
Space heating	80 - 140 kWh/m²/year	110 kWh/m²/year
DHW	20 - 35 kWh/m²/year	28 kWh/m²/year
Space cooling	60 - 120 kWh/m²/year	90 kWh/m²/year
Process heat	Varies by port type	40 kWh/m²/year

Table A4. Energy intensity benchmarks used as baseline values.

Thermal Demand Type	Range from Document	Selected Value
Space heating	80 - 140 kWh/m²/year	110 kWh/m²/year
DHW	20 - 35 kWh/m²/year	28 kWh/m²/year
Space cooling	60 - 120 kWh/m²/year	90 kWh/m²/year
Process heat	Varies by port type	40 kWh/m²/year

Table A5. Information related to the terminal area.

Parameter	Value
Terminal area	10,000 m²
Total annual space heating	1,100,000 kWh
Total annual DHW	280,000 kWh
Total annual space cooling	900,000 kWh
Total annual process heat	400,000 kWh

Table A5. Information related to the terminal area.

Parameter	Value
Terminal area	10,000 m²
Total annual space heating	1,100,000 kWh
Total annual DHW	280,000 kWh
Total annual space cooling	900,000 kWh
Total annual process heat	400,000 kWh