Performance-Based Maintenance and Operation of Multi-Campus Critical Infrastructure Facilities Using Supply Chain Multi-Choice Goal Programming

Abstract

Building maintenance is a critical component of ensuring long-term performance, safety, and cost-efficiency in both conventional and critical infrastructures. While traditional contracting approaches have often led to inefficiencies and rigid procurement systems, recent developments in performance-based maintenance, digital technologies, and multi-objective optimization provide opportunities to enhance both operational reliability and energy performance. From a resilience perspective, the ability to sustain functionality, adapt maintenance intensity, and recover performance under resource or operational stress is essential for ensuring infrastructure continuity and resilience. This study develops and validates an optimization model for the operation and maintenance of large campus infrastructures, addressing the persistent imbalance between over-maintenance, where costs exceed optimal levels by up to 300%, and under-maintenance, which compromises performance continuity and weakens resilience over time. The model integrates maintenance efficiency indicators, building performance indices, and energy-efficiency retrofits, particularly LED-based lighting upgrades, within a multi-choice goal programming framework. Using datasets from 15 campuses comprising over 2000 buildings, the model was tested through case studies, sensitivity analyses, and simulations under varying facility life cycle expectancies. The facilities were analyzed for alternative life cycles of 25, 50, 75, and 90 years, and the design life cycle was set for 50 years. The results show that the optimized approach can reduce maintenance costs by an average of 34%, with savings ranging from 1% to 55% across campuses. Additionally, energy retrofit strategies such as LED replacement yielded significant economic and environmental benefits, with payback periods of approximately 2–2.5 years. The findings demonstrate that integrated maintenance and energy-efficiency planning can simultaneously enhance building performance, reduce costs, and support sustainability objectives, offering a practical decision-support tool for managing large-scale campus infrastructures.

1. Introduction

The management of multi-campus infrastructures poses significant challenges due to their scale, complexity, and the need for uninterrupted service. Traditional maintenance approaches, typically based on prescriptive specifications or unit-price contracts, often lead to inefficiencies such as over-maintenance, which drive costs beyond optimal levels, or under-maintenance, which undermines safety, reliability, and long-term asset performance. These challenges are further manifested in critical infrastructures, such as universities, healthcare complexes, and public-sector facilities, where effective maintenance is crucial for ensuring the resilience and continuity of operations. In this context, resilience refers to the capacity of campus infrastructure systems to maintain and restore functionality under conditions of aging, resource constraints, or unexpected disruptions. Performance-based maintenance offers a proactive mechanism to enhance such resilience by balancing preventive actions, adaptive resource allocation, and risk-informed decision-making of maintenance.

At the same time, growing demands for sustainability and energy efficiency require facility managers to integrate maintenance strategies with energy retrofits, such as upgrading lighting systems, to reduce lifecycle costs and environmental impacts. Despite increasing research attention, most existing studies treat maintenance planning and energy efficiency separately, resulting in a gap in comprehensive decision-support tools that address both objectives simultaneously at the multi-campus scale. This study addresses this gap by developing and validating a Supply Chain Multi-Choice Goal Programming (MCGP) optimization model tailored for the performance-based maintenance of multi-campus infrastructures.

Recent advancements in urban stormwater management emphasize the critical role of integrated grey–green infrastructure (GGI) in enhancing resilience and cost-efficiency under climate change pressures. Multi-objective optimization frameworks have been developed to evaluate trade-offs between life cycle cost (LCC), technological resilience, and operational adaptability, offering robust planning tools for urban drainage systems. For instance, Wang et al. [1] demonstrated that decentralized GGI systems incorporating bioretention cells and porous pavements can reduce overflow risks and improve soil permeability, while maintaining economic viability. These findings support the claim that cost–performance integration is essential for sustainable infrastructure planning.

Moreover, the relevance of multi-stage planning in dynamic urban systems has been highlighted by Zhang et al. [2], who proposed a dual-pathway framework, as well as forward and backward planning, to optimize hybrid low-impact development (LID) and grey infrastructure (GREI) layouts in response to land-use changes. Their study showed that Forward Planning, which incrementally increases LID deployment, yields more resilient and economical drainage solutions over time, especially under extreme storm scenarios and pipeline failure risks. This staged approach aligns with the adaptive needs of multi-campus infrastructures and reinforces the applicability of the proposed model in evolving urban contexts.

This study introduces a novel optimization framework that uniquely integrates Maintenance Efficiency Indicators (MEIs), Building Performance Indicators (BPIs), and targeted energy retrofit strategies, specifically LED lighting upgrades, within a MCGP model. Unlike previous research that typically addresses maintenance planning and energy efficiency as separate domains, this model enables the simultaneous optimization of both aspects across a large-scale, multi-campus infrastructure comprising over 2000 buildings. By combining performance-based maintenance with energy-saving interventions, the framework offers a scalable and data-driven decision-support tool for facility managers and policymakers that fosters synergy between maintenance and energy efficiency.

The proposed framework incorporates the MEI, BPI, and retrofit measures to balance cost, fostering operational savings, sustainability, and reliability. The model is designed to guide resource allocation across distributed facilities, capturing trade-offs between maintenance intensity, the service regime, and energy-efficiency improvements. Its applicability is demonstrated through case studies involving 15 campuses and over 2000 buildings, supported by simulation and sensitivity analyses. The findings highlight the potential of integrated planning to significantly reduce costs, extend service life, and enhance the sustainability of critical infrastructures. By combining PBM with optimization techniques, the research provides a practical decision-support tool for policymakers, facility managers, and stakeholders responsible for ensuring the long-term performance and resilience of multi-campus building systems. The paper follows with a state-of-the-art review, overview, and detailed research framework, a detailed field survey, implementation of the framework on a comprehensive database of public facilities, validation of the proposed model, and recommendations for future research.

2. Literature Review

Building maintenance is not optional but a fundamental requirement for all facility types (e.g., residential, commercial, industrial, logistical, or infrastructure). For critical infrastructures, maintenance and operation are vital to ensuring performance continuity, operational efficiency, cost-effectiveness, and control. Achieving long-term, cost-effective performance requires structured maintenance strategies that account for varying levels of criticality: high (critical systems), standard (regular importance), and low (reduced importance). Properly designed plans improve reliability, energy efficiency, and safety while maintaining the cost-effectiveness of the maintenance and operations of the facilities.

Critical infrastructures (CIs) are highly complex, combining conventional building systems (e.g., structure, envelope, finishes, electrics, water, HVAC, fire safety) with advanced systems that secure the performance continuity of CIs: communications, IT, emergency power supply, and monitoring. Managing such facilities presents a multi-objective challenge that must balance (a) reliable and robust energy performance, (b) minimal maintenance costs (preventive and corrective), and (c) minimal operational costs through energy-efficiency measures. Addressing this requires optimal allocation of labor, equipment, and resources, alongside clear maintenance and energy policies.

This literature review is structured around four key themes aligned with the scope of the research: (1) definitions and the state of the art in building maintenance management, (2) maintenance and performance of critical buildings, (3) energy efficiency in maintenance practices, and (4) resource sharing strategies across multi-campus buildings and systems.

2.1. Building Maintenance Management: Definitions and Knowledge Base

Traditional approaches to maintenance contracting have been widely critiqued for their rigidity. Shohet and Straub [3] highlighted the limitations of conventional unit-price contracts, which rely on prescriptive specifications that constrain procurement flexibility and ultimately impair building performance. In contrast, performance-based maintenance (PBM) emphasizes outcome-oriented measures, allowing for greater adaptability and efficiency.

The introduction of PBM as an outsourcing alternative was first advanced by Stenbeck [4]. Subsequent comparative studies in the Netherlands and Israel by Shohet and Straub [3] demonstrated that PBM could reduce costs by approximately 20% while simultaneously enhancing performance in public facilities. Building on this, Shohet and Nobili [5] proposed an innovative model tested across 13 public building case studies. Their framework, based on six performance indicators and outsourced service provision, yielded improvements of 20–40% in the performance–cost ratio. Significantly, the model was found to be adaptable across building types and capable of extending into risk-based assessment frameworks, thereby strengthening its applicability in diverse contexts.

The broader paradigm of performance-based contracting (PBC) has similarly been explored across other sectors. Kostas and Finn [6], through a systematic review of 241 publications in operations and supply management (OSM), emphasized the importance of aligning performance specifications, incentive mechanisms, supplier behavior, and risk allocation. They also pointed to gaps in empirical grounding and theoretical integration, particularly in sustainability and innovation studies. This echoes a recurring theme in the facilities management literature: while conceptual frameworks abound, their implementation often faces practical barriers.

Recent studies illustrate how digital technologies are beginning to transform maintenance management. For instance, Signorini and Pomè [7] assessed facility management’s readiness for digital transformation via digital twins (DTs), identifying significant opportunities alongside challenges such as cost, interoperability, and skills gaps. Complementing this, advanced computer vision frameworks are being developed, such as the automated 2D–3D system by Wang and Gan [8], which integrates transfer learning, deep learning (ResNet-50 with Grad-CAM), and scene reconstruction to enhance defect detection. Similarly, predictive maintenance (PdM) frameworks that combine hybrid optimization algorithms, support vector machines, and recurrent neural networks have been shown to improve failure forecasting and maintenance scheduling [9]. Together, these innovations highlight the growing integration of AI and digitalization in maintenance management.

In terms of conceptual clarity, the literature provides several widely accepted definitions and classifications. Maintenance is broadly defined as the set of activities aimed at preserving building performance, aesthetics, and safety through proper asset use (Standards Institution of Israel [SII], 2002 [10], Wang et al. 2022 [11]). Israeli law distinguishes between defect liability and warranty periods, the former being unconditional and the latter conditional upon proof of fault (Israeli Ministry of Justice, 1973). Maintenance methods range from planned and preventive approaches to breakdown and condition-based strategies, each balancing reliability and cost differently. Life Cycle Cost (LCC) analysis is also central, integrating construction, operation, and maintenance expenses to support design-stage decision-making, while Multiple Criteria Decision Analysis (MCDA/MCDM) offers structured frameworks to balance cost, safety, quality, and sustainability.

The design life cycle of the facilities considered four scenarios: 25, 50, 75, and 90 years [12,13,14,15,16]. A design life cycle of 25 years is appropriate for short-term use with a high ratio of occupancy changes along the service life cycle, a 75-year design life cycle is appropriate for long-term use of critical facilities, such as hospitals, and a 90-year design life cycle is adaptable for strategic critical infrastructures.

Standards and regulatory guidelines formalize these practices internationally. In Israel, Standards 1525 Parts 1–4 specify requirements for finishes and service systems in residential and non-residential buildings, emphasizing as-built documentation [10]. These frameworks draw from British precedents while aligning with global benchmarks such as ASTM E1670 [17] and BS 8210 [18]. Complementary guidelines, including the “Blue Book” for construction works (Israeli Ministry of Housing, 2010 [19]) and NFPA 96 [20], further underline the importance of systematic classification and lifecycle- and safety-oriented practices. Collectively, these contributions establish a solid knowledge base for building maintenance management.

2.2. Maintenance of Critical Buildings and Infrastructure

The maintenance of critical facilities has increasingly been linked to both sustainability and risk reduction. For instance, Sharif and Hammad [21] emphasized that buildings account for more than 30% of global energy consumption and greenhouse gas emissions. Their work demonstrated how simulation-based multi-objective optimization, coupled with artificial neural networks, could enable cost-effective energy renovation strategies that reduce life cycle costs.

Specialized facilities, such as data centers, illustrate the complexity of balancing performance, cost-effectiveness, and efficiency. Zhu et al. [22] developed a hybrid fault-diagnosis model combining random forest classification with regression-based severity estimation, achieving diagnostic accuracy rates above 94% while significantly lowering false alarm rates. At a broader scale, Gunay et al. [23] reviewed the potential of big data analytics in operation and maintenance (O&M), noting that building automation and security data remain underutilized resources. They called for semantic data models and open-access datasets to unlock algorithmic innovation.

Beyond buildings, similar challenges are evident in infrastructure systems. For example, a Virtual Depot ontology has been introduced to enhance rolling stock maintenance management, integrating fleet knowledge, depot resources, and scheduling processes to optimize short-term planning [24]. In the manufacturing sector, a hybrid metaheuristic for joint decision-making in the remanufacturing-integrated operation and maintenance of complex manufacturing systems was developed [25]. Gan et al. [26] elucidate the need for optimization of maintenance and spare inventory, with dependence between the system and the environment. Operational readiness-oriented condition-based maintenance and spare part optimization for multi-state systems are introduced in [27]. Integrated maintenance and spare part inventory optimization with transshipments for multi-fleet systems emphasize the need for multi-goal programming optimization of maintenance and operations [28].

Comparable challenges are faced in transport infrastructure, where integrated frameworks using integer programming, Markov decision processes, and priority-based optimization have been applied to budget allocation and preventive maintenance decisions at the network level [29].

The water sector offers another critical case. Urban flood resilience planning increasingly relies on advanced optimization techniques to balance environmental, economic, and social objectives. Multi-objective optimization models have proven effective in designing sustainable urban drainage systems (SUDSs), enabling trade-offs between flood volume, duration, peak runoff, pollutant loads, and capital costs. Seyedashraf et al. [30] demonstrated that spatial slope variations across urban catchments significantly influence the distribution and equity of green infrastructure benefits, underscoring the need for context-sensitive design policies. Khalifi et al. [31] emphasized the importance of integrating life cycle assessment (LCA), life cycle costing (LCC), and social LCA into green infrastructure planning to capture the full spectrum of sustainability impacts. Such integration supports resilient urban water systems that not only mitigate flood risks but also promote long-term resource efficiency and equitable service delivery. These approaches align with broader sustainability goals by incorporating life cycle cost analysis and environmental performance metrics. A study of European wastewater treatment plants (WWTPs) [32] developed cost models to capture the impacts of facility deterioration, providing decision support for preventive maintenance and equipment replacement. Similarly, in the building sector, MCDM methods, including fuzzy axiomatic design, have been employed to determine the most suitable maintenance strategies for public buildings, shifting attention from traditional risk-based approaches toward sustainability-driven decision-making [33]. Taken together, these studies highlight the growing sophistication of maintenance strategies across diverse critical facilities, while underlining the challenges of aging assets, resource optimization, and sustainability imperatives.

2.3. Energy Efficiency in Maintenance

Energy efficiency during the operational phase of buildings is increasingly recognized as a decisive factor in sustainability, given that the built environment accounts for roughly half of global greenhouse gas emissions. Maintenance practices play a pivotal role in shaping this energy performance.

A range of approaches has been proposed to integrate energy efficiency into maintenance decision-making. Wang et al. [34] introduced an indicator-based framework that balances energy savings with investment returns, achieving a payback period of four years in a case study involving lighting system upgrades. In hospital settings, Dulce-Chamorro and Martinez-de-Pison applied [35] predictive analytics to optimize HVAC chiller operations, reducing startup events by 82.5% while stabilizing energy use. Loli and Bertolin [36] demonstrated the feasibility of implementing energy-efficient maintenance strategies in historically sensitive districts, showing that conservation and sustainability goals can coexist.

Similarly, Koroxenidis and Theodosiou [37] investigated the potential of green roof systems in Mediterranean climates, reporting energy savings of 8–31%. However, they also noted increased irrigation demands, underscoring the importance of climate-specific assessments in evaluating trade-offs. Beyond the building envelope, optimization methods drawn from waste management also demonstrate cross-sector relevance. Rough set theory has been applied to municipal solid waste systems, yielding nonlinear programming models that balance cost minimization and revenue maximization under uncertainty [38]. Scalarization-based goal programming methods have also been adapted to multi-objective optimization in facility location planning, integrating environmental and economic considerations [39].

Extending this line of research, studies have also introduced multi-choice goal programming models to support renewable energy capacity expansion, balancing objectives such as plant mix, location, investment costs, emissions reduction, and social acceptance [40]. Collectively, these works illustrate that while energy-efficient maintenance practices remain context-dependent, there is a clear trend toward multi-objective, system-wide approaches capable of balancing economic, environmental, and social goals.

There exist significant seasonal fluctuations in the energy consumption in residential and non-residential facilities. Fluctuations may be as high as four times between autumn lows and summer high peaks, depending on the climatic region [41,42]. This topic has not been the focus of the present research and is recommended for future research efforts.

2.4. Maintenance Resource Sharing Across Multi-Campus Systems

Despite growing interest in building maintenance management, a significant gap persists in the allocation of resources across multi-campus systems. Most existing approaches have focused on industrial and supply chain contexts, with limited transfer into higher education or large institutional infrastructures.

For example, Aalaei and Davoudpour [43] proposed a bi-objective model integrating supply chain management with dynamic cellular manufacturing, addressing inter-campus logistics, demand uncertainty, and workforce variability. Their model minimized maintenance costs while supporting operational flexibility, offering valuable insights for distributed facility management. Similarly, Karydas and Gifun [44] developed a prioritization framework for MIT campuses, integrating risk minimization, economic optimization, and academic operational requirements. This framework was validated during campus modernization projects, demonstrating its practical effectiveness.

Digital innovations are also beginning to support resource sharing. A DT-based decision support framework has been developed to integrate BIM, IoT, and GIS for the life cycle management of relocatable modular buildings, validated through South Korea’s modular school system [45]. Other logistics-focused models, such as multi-objective optimization frameworks for quarantine center operations [46] and GIS-integrated mixed integer programming for biofuel supply chains [47], further demonstrate the applicability of advanced optimization tools to distributed systems. Moreover, bi-objective optimization models for dynamic cellular manufacturing and supply chain design [48] underline the importance of balancing cost-efficiency with labor allocation under uncertainty, reinforcing parallels with campus-level maintenance planning.

Despite these advances, the literature confirms that designated decision-support tools for campus-wide maintenance remain scarce. Addressing this gap offers significant potential for enhancing resilience, efficiency, and sustainability in the management of multi-campus infrastructures.

3. Methods

This study develops a model for the operation and maintenance of large campus infrastructures, where current practices are often either excessive, leading to over-allocation of staff and resources at costs up to 300% above optimal levels, or insufficient, resulting in costly failures and loss of performance continuity. To address this imbalance, the model minimizes preventive and corrective maintenance costs, operational expenditures (energy, cleaning, outsourcing), and resource use (labor, equipment, spare parts, and materials), while ensuring minimum system performance and energy efficiency in lighting and HVAC systems. The model was fully formalized, implemented on a representative network of critical campuses, and validated through case studies and sensitivity analyses.

The research process included building performance and maintenance datasets (collected in comprehensive field survey), developing efficiency models for both maintenance and energy, and incorporating energy retrofits such as the replacement of low-grade lighting with LED systems, which extend life cycle, reduce corrective interventions, and lower electricity demand. Simulations were conducted across different facility lifespans, with results expressed as costs per square meter and measured against a Building Performance Indicator (BPI) target. The analysis showed that, in some cases, observed building performance exceeded optimal levels due to over-maintenance or the presence of new facilities. Finally, sensitivity tests were applied to assess robustness under varying assumptions.

Research Stages

The methodology followed the structured process delineated in Figure 1:

• Data gathering: Development of performance and maintenance databases for facilities and campuses, including estimated costs per square meter. This stage included a comprehensive field survey of the buildings and infrastructures using B.P.I. survey and estimated costs of maintenance based on the buildings age and occupancy (low, standard and high).
• Maintenance modeling: Establishment of a campus-level maintenance efficiency model. The design life cycle of campuses was assumed to be 50 years [12,13,14,15,16].
• Energy modeling: Development of an energy-efficiency model incorporating replacement of low-grade lighting with LED systems, accounting for life cycle, failure rates, and reduced corrective maintenance.
• Optimization: Application of the MCGP framework to integrate maintenance and energy costs under performance constraints.
• Simulation and sensitivity analysis: Evaluation of scenarios with varying facility life cycle expectancies to determine optimal trade-offs between costs and performance continuity.
• Validation: Case-study implementation of the model on campus infrastructures, with results compared against baseline assumptions and tested through sensitivity analysis.

Table 1. B.P.I performance categories.

B.P.I. Range	Building Performance
BPI ≥ 80	Good and above
70 ≤ BPI < 80	Borderline
60 ≤ BPI < 70	Deteriorating
BPI < 60	Poor or hazardous

presents building performance according to B.P.I. range. Companies should aim to have a B.P.I. of at least 80. Aiming for a BPI of 100 does not justify the substantial increase in maintenance costs. Each building has several systems. These systems do not have equal importance in the total BPI of the building.

Table 2. Building system component weight.

Building System Component	${\mathit{W}}_{\mathit{n}}$ [%]
Interior finishes	24
Structural frame	17
HVAC (Heating, Ventilation, Air Conditioning)	16
Exterior envelope	14
Electrical systems	13
Plumbing and sanitation	7
Communication and low voltage	5
Peripheral infrastructure	4

presents the weight of each system that will lead to the building’s total BPI using Equation (1). These values are based on the calculation of the percentage of each building system’s life cycle costs over the building life cycle (e.g., 24% of the total maintenance cost is allocated to interior finishes). The MEI depends on the building type; acceptable values are presented in

Table 3. Required MEI for different building types.

Building Type	MEI
Residential	0.17
Offices	0.19
Clinics	0.24
Education	0.20
Dining	0.27

. Each campus has many building types; the weighted MEI is calculated based on Equation (2).

$$B.P.I.={\sum }_1^n{P}_n·W_n$$

(1)

where $P_n$ is the system function and $W_n$ is the system weight (Table 2).

$$\mathrm{W}\mathrm{e}\mathrm{i}\mathrm{g}\mathrm{h}\mathrm{t}\mathrm{e}\mathrm{d}\mathrm{M}\mathrm{E}\mathrm{I}={\displaystyle \frac{\sum _{i=1}^N{MEI}_i·A_i}{\sum _i^N{A}_i}}$$

(2)

$$\mathrm{N}\mathrm{A}\mathrm{M}\mathrm{E}={\displaystyle \frac{AME}{OC·{AC}_y}}$$

(3)

$$\mathrm{M}\mathrm{E}\mathrm{I}={\displaystyle \frac{NAME}{BPI}}$$

(4)

where ${MEI}_i$ is the building MEI according to building type (Table 3) and $A_i$ is the area of building i.

4. Results

Table 4. Campuses data.

Campus	Number of Buildings	Built Area [m2]	Weighted MEI	Current Maintenance Cost [NIS]
1	221	18,745	0.160	1,474,234
2	226	17,667	0.164	1,542,791
3	221	16,409	0.164	1,319,795
4	104	12,400	0.190	643,228
5	170	12,801	0.196	747,589
6	149	7491	0.195	618,119
7	132	21,275	0.143	888,531
8	126	9607	0.195	873,426
9	102	5864	0.192	510,934
10	95	6732	0.203	631,259
11	117	5731	0.172	440,108
12	167	11,849	0.186	1,076,331
13	95	7497	0.194	582,709
14	70	5750	0.192	609,127
15	59	3800	0.198	492,594

presents the data for the 15 campuses. The weighted MEI is calculated for each campus based on Equation (2). The current maintenance costs for each campus are also presented. For example, the first campus includes 221 buildings with a total built area of 18,745 m², its weighted MEI for all 221 buildings is 0.160, and its current maintenance cost is NIS 1,474,234.

Table 5. Planned and current B.P.I for the 15 campuses and selected systems.

		Structural Frame	Exterior Envelope	Interior Finishes	Electrical Systems	Plumbing and Sanitation	HVAC	Communication and Low Voltage	Peripheral Infrastructure	Weighted B.P.I.
1	Planned	85.0	76.0	80.0	85	85	85.0	85	72.0	82.0
	Current	88.9	67.7	73.9	77.2	78.2	77.9	76.2	78.8	77.3
2	Planned	85.0	74.0	74.0	85	85	85.0	85	70.0	80.2
	Current	80.7	70.9	70.9	69.9	71	83.2	5.1	65.4	70.9
3	Planned	85.0	74.0	74.0	85	85	85.0	85	70.0	80.2
	Current	80.0	81.7	75.4	72.2	78.5	80.8	90	70.0	78.2
4	Planned	85.0	76.0	80.0	85	85	85.0	85	72.0	82.0
	Current	82.7	76.1	79.1	74.6	76.9	77.2	-	-	78.2
5	Planned	90.0	90.0	75.0	90	90	89.0	85	70.0	85.2
	Current	94.0	92.8	69.7	86.2	87.8	89.4	85	81.4	84.9
6	Planned	85.0	70.0	70.0	85	85	85.0	85	66.0	78.5
	Current	81.3	76.5	70.5	71.8	73.2	89.3	59.7	-	76.2
7	Planned	85.0	76.0	75.0	85	85	85.0	85	75.0	80.9
	Current	80.8	68.0	74.4	73.5	63.3	73.8	0	70.0	69.7
8	Planned	85.0	71.0	73.0	85	85	85.0	85	66.0	79.4
	Current	73.0	54.3	62.0	66.6	61.3	65.7	60	70.0	64.2
9	Planned	85.0	75.0	75.0	85	85	85.0	85	70.0	80.6
	Current	85.6	73.6	72.6	73.9	81.8	79.0	-	78.1	77.2
10	Planned	85.0	70.0	74.0	85	85	85.0	85	65.0	79.5
	Current	78.5	62.2	68.2	69.8	61.6	70.9	-	73.9	69.6
11	Planned	85.0	70.0	74.0	85	85	85.0	85	65.0	79.5
	Current	74.9	60.8	62.0	64.5	61.3	67.7	-	67.7	65.6
12	Planned	85.0	74.0	74.0	85	85	85.0	85	70.0	80.2
	Current	86.0	82.4	75.7	57.2	81.3	82.1	-	-	77.6
13	Planned	85.0	85.0	85.0	85	85	85.0	85	85.0	85.0
	Current	86.0	83.5	82.6	76.7	81.4	83.1	66.3	85.0	81.8
14	Planned	90.0	90.0	90.0	90	90	90.0	90	90.0	90.0
	Current	96.1	94.5	85.4	88.9	89.5	88.0	-	-	90.1
15	Planned	85.0	76.0	80.0	85	85	85.0	85	72.0	82.0
	Current	83.7	77.8	73.2	75.1	72	65.6	-	87.5	75.3

present the planned and current B.P.I. of the 15 campuses. For visual illustration, Figure 2 and Figure 3 present the current and planned BPI of the 15 campuses for interior finishes, electrical systems, HVAC, and weighted BPI. Eight systems are considered (Table 2) and, based on them, the weighted B.P.I. for each campus is calculated based on Equation (1). The eight systems are structural frame, external envelope, interior finishes, electrical systems, plumbing and sanitation, HVAC, communication and low voltage, and peripheral infrastructures. Some campuses have missing data for communication, voltage system, and peripheral systems; the current peripheral infrastructure of campus 4, for example, is missing. This is due to the absence of these systems during the campus inspection, and they were later added. In that case, the weighted BPI is calculated based on the remaining systems.

Table 6. Calculation of energy savings costs and return on investment in lighting for a typical kitchen and dining room, and a typical logistics building.

	Dining Facilities		Logistical Facilities
Item	Existing Fixture	Replacement Fixture	Existing Fixture	Replacement Fixture
Lighting Type	Fluorescent 2 × 36 W	LED Fixture 40 W	MH Discharge Lamp 400 W	LED Fixture 120 W
Electricity Consumption (KWh)	0.09	0.04	0.48	0.12
Building Area (m2)	600	600	200	200
Number of Fixtures	50	50	8	8
Fixtures per building area (m2 per unit)	12	12	25	25
Operating Hours per Day	15	15	10	10
Days per Year	365	365	250	250
Price per KWh (NIS)	0.5	0.5	0.5	0.5
Annual Electricity Cost (NIS)	12,319	5475	4800	1200
Annual Maintenance Cost (NIS)	–	0	–	0
Total Annual Cost (NIS )	12,319	5475	4800	1200
Annual Savings (NIS)	–	6844	–	3600
Lighting fixture cost (NIS)	–	230	–	1000
Replacement Cost (NIS )	–	70	–	70
Total Fixture Replacement Cost (NIS)	–	300	–	1070
Total Project Cost (NIS)	–	15,000	–	8560
ROI (Years)	–	2.2	–	2.4
Savings per m2 per Year (NIS)	–	11.41	–	18
Cost per m2 (NIS)	–	25	–	42.8

presents the technical and economic implications of replacing conventional fluorescent fixtures with LED technology for a typical kitchen and dining room, and for a typical logistics building. In the dining facility, fluorescent 2 × 36 W is replaced with LED fixture 40 W; a building area of 600 m² with 50 fixtures results in a fixture density of 12 m² per unit. Each fixture consumes 0.09 kWh per hour, with 15 operating hours per day and 365 days per year, at an electricity price of NIS 0.5 per kWh. The electricity price was constant for the research period for all 15 campuses. This leads to an annual electricity cost of NIS 12,319. When replaced with 40 W LED fixtures, which consume 0.04 kWh per hour, the annual cost is reduced to NIS 5475, resulting in annual savings of NIS 6844, equivalent to 11.41 NIS/m². The cost of each LED fixture is NIS 230, with an installation cost of NIS 70 per unit, resulting in a total project cost of NIS 15,000 for all 50 fixtures, or 25 NIS/m². When this investment is compared with the annual savings, the payback period, expressed as return on investment (ROI), is calculated at 2.2 years. Following the same procedure, replacing the MH discharge lamp 400 W with a 120 W LED fixture in logistical facilities yields annual savings of 18 NIS/m².

Table 7. Database for integrated calculations of energy efficiency and maintenance.

	Campus 1	Campus 2	Campus 3
MEI	0.160	0.164	0.164
Lighting energy savings for logistic facilities [NIS/m2]	18.00	18.00	18.00
Lighting energy savings for dining facilities [NIS/m2]	11.41	11.41	11.41
Building area [m2]	18,745	17,667	17,409
Energy upgrading area for dining facilities [m2]	5063	1045	1962
Energy upgrading area for logistical facilities [m2]	686	936	873
Energy investment for logistical facilities [NIS/m2]	48,680	10,045	18,861
Energy investment for dining facilities [NIS/m2]	3855	5255	4905
Energy benefit for dining facilities [NIS]	91,141	18,807	35,313
Energy benefit for logistical facilities [NIS]	7830	10,674	9963
Maintenance costs [NIS]	911,282	858,153	796,698
Total cost [NIS]	864,845	843,971	775,189
Current maintenance costs [NIS]	1,474,234	1,542,791	1,319,795
Total cost savings [%]	41	45	41

summarizes the total cost savings calculated for three of the fifteen campuses. For illustration, Campus 1 has a weighted MEI of 0.160. According to Table 6, the lighting energy savings per year for logistic facilities and for dining facilities are 18.00 NIS/m² and 11.41 NIS/m², respectively. The corresponding areas of these building types in Campus 1 are 5063 m² and 686 m². For energy savings, the cost for logistic facilities is calculated as the product of the building area (5063 m²) and the cost per square meter (42.8 NIS/m² from Table 6), amortized over a five-year period at an interest rate of 4%. This yields an energy investment of NIS 48,680 for logistic facilities. Applying the same procedure to kitchens and dining rooms results in an additional saving of NIS 3855. The economic benefits are the building areas multiplied by lighting energy savings (18.00 NIS/m² for logistic facilities and 11.41 NIS/m² for dining rooms). This results in economic benefits of NIS 91,141 for logistic facilities and NIS 7830 for dining facilities.

The maintenance cost is calculated using Equation (3). For example, Campus 1 is estimated at NIS 911,282. This value is derived by multiplying the MEI (0.160) by the Building Performance Index (B.P.I., 82.0 from Table 5), the total building area (18,745 m²), and the occupancy and age coefficients (both equal to 1), and then converting to NIS. The total cost is the maintenance cost (NIS 911,282) plus the energy investment (NIS 48,680 and 3855) minus the energy savings (NIS 91,141 and 7830), which equals NIS 864,845 (5.1% reduction for using energy savings).

Current maintenance cost is NIS 1,474,234 (Table 4). The maintenance cost received is NIS 911,282, which is 38% lower. When we consider energy savings, this cost reduction is 41%. These cost savings ranged from 1 to 55% for the 15 campuses, with an average of 34% and a standard deviation of 15%.

The objective function (Equation (5)) was defined to minimize the annual maintenance costs required for each facility to achieve optimal maintenance and the necessary energy efficiency improvements that would generate significant savings in energy consumption. Decision variables are the B.P.I. of each building and the percentage of logistic facilities and dining rooms that will undergo energy savings. The constraints on the decision variables are demonstrated in Equations (6) and (7). The minimum BPI of each building is according to the planned BPI from Table 5.

$$F_1=min{\sum }_{i=1}^n{\displaystyle \frac{{(MEI}_i·{BPI}_i·A_i+U_i·{AU}_i)}{n}}$$

(5)

$${BPI}_i\ge min{BPI}_i$$

(6)

$$1\ge U_i\ge 0$$

(7)

where ${B.P.I.}_i$ are the optimal building performance indicators for campus i; $U_i$ is the percentage of logistic facilities and dining rooms that will undergo energy savings; ${MEI}_i$ is the maintenance efficiency indicator from campus i (Table 3); $A_i$ and ${AU}_i$ are the campus area and building area where energy efficiency was implemented, respectively; n is the number of campuses; and $min{BPI}_i$ is the minimum building performance indicator.

Since energy savings have a positive effect on the total cost, the optimal solution was to have 100% of logistic facilities and dining rooms that will undergo energy savings for all campuses ($U_i=1$). And the BPI for each building was the planned BPI (the minimal allowed value according to Equation (6)). This result is visually illustrated in Figure 4, where there are three scenarios for the total cost for the 15 campuses. The first scenario has no energy savings, the second scenario involves replacing 50% of the lighting (limited energy upgrading), while the third scenario involves replacing all lighting (comprehensive energy upgrading). As can be seen, the higher the investment in energy savings, the lower the total cost.

Figure 5 presents the effect of increasing the BPI of all 15 campuses by one unit on the total maintenance cost for the 15 campuses and the NAME (dividing the total cost by the total area of all campuses). The results show that each increase in BPI by one unit increases the NAME by NIS 0.65.

The quantified energy savings presented in Table 7, ranging from 11.41 to 18 NIS/m² annually for dining and logistical facilities, are consistent with findings from recent studies on energy retrofits in institutional buildings. For example, Asdrubali et al. [48] evaluated retrofit interventions in Italian schools and found that replacing fixtures and adding insulation yielded significant energy and carbon payback benefits, with economic payback periods ranging from 2.5 to 4 years. Similarly, Alamin et al. [49] conducted a comprehensive energy audit of a university building in Malaysia and reported cumulative energy savings of 41% from combined retrofit measures, with payback periods between 0.8 and 8.9 years depending on the intervention.

Papangelopoulou et al. [50] reviewed financial evaluation methods for building retrofits and found that life cycle cost analysis (LCCA) and Cost–Benefit Analysis (CBA) are the most used tools, with LED lighting upgrades consistently showing favorable returns on investment. Azouz and Elariane [51] demonstrated that smart retrofitting of office buildings in Egypt achieved over 20% energy savings, with payback periods of approximately 3 years. These results align closely with the 2.2–2.4-year ROI observed in our model.

Moreover, Walter and Sohn [52] used regression-based modeling to estimate retrofit savings across a large building dataset, confirming that lighting upgrades are among the most predictable and cost-effective energy-saving measures. Kim and Medal [53] emphasized that institutional decision-making for retrofits is often driven by life cycle cost and environmental impact, reinforcing the relevance of integrating MEI and energy strategies in campus-scale planning.

These findings substantiate the economic and environmental benefits of comprehensive energy retrofits and support the model’s recommendation for full implementation across all campuses.

5. Conclusions

This study introduces an inclusive framework for integrated maintenance and energy efficiency of a multi-campus facility management decision-making problem. An objective function for the allocation of resources in maintenance and energy was developed. The framework seeks to minimize the maintenance costs and maximize the energy savings given performance constraints. The framework was validated on a case study of 15 campuses. The performance and costs of maintenance and energy were compared prior and after the implementation of the proposed model. The results deliver conclusive evidence that a performance-based and optimization-driven maintenance framework can significantly enhance the performance, cost-effectiveness and energy efficiency of complex FM problems. The framework contributes to the operational resilience, reliability, and sustainability of multi-campus critical infrastructures by reducing the failure frequency and assuring the reliability of critical systems. By integrating maintenance efficiency indicators (MEIs), building performance indicators (BPIs), and energy retrofit strategies, the model enables decision-makers to maintain essential functions under varying operational and environmental conditions. The results demonstrate that implementing the proposed framework can reduce maintenance costs by an average of 34%, with savings ranging from 1% to 55% across campuses, while energy retrofit strategies such as LED lighting upgrades yielded short payback periods of 2–2.5 years. These outcomes confirm that optimizing maintenance and energy performance together strengthens infrastructure adaptability, reduces vulnerability to system failures, and supports continuity of operations during disruptions.

From a resilience perspective, the model directly contributes to sustaining critical infrastructure performance by enhancing adaptability and ensuring service continuity under challenges such as supply shortages, extreme weather events, or seismic hazards. The findings emphasize the practical value of aligning preventive and corrective maintenance with energy efficiency to achieve both cost-effectiveness and long-term resilience. Additionally, effective energy planning must incorporate resilience measures to address unexpected disruptions, such as supply shortages caused by extreme weather events or earthquakes, ensuring continuity and reliability of operations. Although the framework proved robust across diverse scenarios, future research should incorporate life cycle carbon impacts, renewable energy integration, digital twin technologies, and predictive analytics to further strengthen proactive and adaptive maintenance strategies. Overall, the developed model not only enhances sustainability and cost-efficiency but also establishes a data-driven foundation for improving the resilience and long-term reliability of campus-scale critical infrastructure systems.