Dynamic Cost–Benefit Analysis of Digitalization in the Energy Industry

Jose Angel Leiva Vilaplana; Guangya Yang; Emmanuel Ackom; Roberto Monaco; Yusheng Xue

doi:10.1016/j.eng.2024.11.005

Engineering ›› 2025, Vol. 45 ›› Issue (2) :174 -187. DOI: 10.1016/j.eng.2024.11.005

Research Energy Systems Engineering—Article

research-article

Dynamic Cost–Benefit Analysis of Digitalization in the Energy Industry

Author information +

History +

PDF (2406KB)

Abstract

Assessing the benefits and costs of digitalization in the energy industry is a complex issue. Traditional cost–benefit analysis (CBA) might encounter problems in addressing uncertainties, dynamic stakeholder interactions, and feedback loops arising out of the evolving nature of digitalization. This paper introduces a methodological framework to help address the intricate inter connections between digital applications and business models in the energy industry. The proposed framework leverages system dynamics to achieve two primary objectives. It investigates how digitalization generally influences the value proposition, value capture, and value creation dimensions of business models. It also quantifies the financial and social impacts of digitalization from a dynamic perspective. The proposed dynamic CBA allows for a more precise quantification of the benefits and costs, associated with evidence-based decision-making. Findings from an illustrative case study challenge the static assumptions of conventional methods. These methods often presume continuous operation, neglecting reinvestment and operational feedback loops, and resulting in negative net present values. Conversely, the outcomes of the proposed method indicate positive net present values when accounting for factors such as reinvestment rates and the willingness to invest in digitalization projects. The principles outlined in this paper can enable a more accurate assessment of digitalization projects, thus catalyzing the development of new CBA applications and guidelines for digitalization.

Graphical abstract

Keywords

System dynamics / Digitalization / Cost–benefit analysis / Energy industry / Business modeling / Socioeconomic assessment

Cite this article

Download citation ▾

Jose Angel Leiva Vilaplana, Guangya Yang, Emmanuel Ackom, Roberto Monaco, Yusheng Xue. Dynamic Cost–Benefit Analysis of Digitalization in the Energy Industry. Engineering, 2025, 45 (2) : 174-187 DOI:10.1016/j.eng.2024.11.005

登录浏览全文

4963

注册一个新账户忘记密码

1. Introduction

The energy industry is currently undergoing a significant shift, driven by the interplay between energy systems and digitalization. Advancements in data gathering, analytics, systems automation, and data-driven decision-making can enhance process efficiency, safety, productivity, and sustainability [1]. Digital technologies in the energy sector are utilized in a diverse range of applications, such as the Internet of Things (IoT) for monitoring and managing electricity networks, artificial intelligence (AI) to facilitate data-driven decision-making, and digital twins to monitor, simulate, and optimize the performance of energy systems [2]. In particular, the power sector digitalization is intricately linked with the concept of smart grids, where information and communication technologies (ICTs) are combined with the electricity grid to facilitate bidirectional data exchange and the seamless integration of renewable energy sources (RES) [3]. These RES include wind [4], solar [5], biomass [6], hydropower [7], and hybrid power plants [8], [9], [10].

Digitalization delivers substantial benefits throughout the energy value chain, from resource extraction and energy conversion (e.g., power generation) to transmission, distribution, and end-use [11]. In renewable power generation, digital technologies improve the forecasting of the available power from sources such as solar and wind, monitor plant operations, and adjust operations to real-time conditions [12]. From a holistic perspective, digitalization holds potential for optimizing RES capacity factor and reducing their levelized costs of energy (LCOE). In transmission and distribution, digitalization enhances grid management through extensive data collection and modeling [13]. Such tools support asset condition monitoring and strategic grid planning and operation tools, which increase transparency in asset health and grid conditions [14]. These insights lead to reduced maintenance costs and unplanned outages, extended asset lifespans, minimized energy losses, and decreased energy theft [15]. As a result, digitalization can significantly lower both investment and operational expenses [16]. At the energy use stage, monitoring energy consumption at the final step of the energy value chain enhances efficiency and reduces costs [17]. Additionally, it enables more active participation from end users, such as through demand response programs, which allow consumers to adjust their energy usage in response to market signals, reducing peak demand and shifting consumption to off-peak periods [18].

The impact of digitalization varies based on the employed technologies, which may act as enablers (e.g., ICTs) or provide direct benefits through enhanced operational speed, improved planning, and better responses to malfunctions and critical events [19]. These technologies influence various business areas within an energy company, including construction, manufacturing, operation, maintenance, and recycling. From both individual and systems perspectives, digitalization induces efficiencies and reduces emissions and fuel consumption, thus diminishing the environmental footprint of the energy sector [20]. Additionally, digital technologies are essential for managing the increasingly complex energy systems driven by decarbonization and electrification [21]. This complexity necessitates advanced tools for demand forecasting, grid management, and interoperability between grid and market operations, all of which are facilitated by digitalization [22].

Despite the promising benefits of digitalization, its adoption, which is still in its early stages, may pose social, technical, and economic challenges [23]. Key challenges include ensuring cyber security and reliability, facilitating access to and sharing of data, and safeguarding data protection and privacy [24]. Furthermore, the energy consumption of digital technologies, represented by the gradual increase of data centers consumption [25], together with the carbon footprint of digital assets, is considered a relevant challenge [26]. Within the economic domain, one important challenge is the assessment of the benefits and costs of digitalization, referred to as digitalization impacts. This challenge arises from the intricate nature of digital technologies, which create diverse impacts across multiple stakeholders and sectors [27]. These impacts include both tangible and intangible aspects, permeating various functional areas within energy companies across the entire energy value chain. In addition to this complexity, failing to quantify these impacts could encourage energy stakeholders to be risk-averse. When compounded by budget constraints, this may impede the widespread adoption of digitalization [28], [29]. Thus, the role of economic analysis becomes essential to unlock digitalization investments.

Traditionally, economic assessment has relied on methods such as cost–benefit analysis (CBA) [30]. CBA involves comparison of the costs and benefits associated with a specific project against a business as usual (BaU) scenario [31]. This comparison is carried out through techniques such as discounted cash flow (DCF) analysis, whose results are the project metrics. Examples of CBA metrics are the net present value (NPV), the internal rate of return (IRR), and the LCOE [32]. Conventionally, energy companies have employed these metrics to assess the financial value of their investments, including impacts such as increased profitability and decreased staff hours. Nevertheless, with the advent of smart grid initiatives such as smart meters and innovations in AI and ICTs, a broader scope for CBA has been adopted. Referred to as social CBA, this method accounts for the financial, environmental, and multi-stakeholder benefits and costs [33]. As a result of the need for assessment tools in addition to the financial, two main sets of guidelines have emerged to appraise the social impacts of smart grid projects. The first is the EU Joint Research Centre (JRC) [34], and the second is the guide from the Electric Power Research Institute (EPRI) [35]. These guidelines offer comprehensive foundations for evaluating smart grid investments and have catalyzed the development of additional methodologies [36], [37].

Although previous guidelines and methodologies represent prominent efforts to quantify the impacts of digitalization, their approach has been reductionist [38], individually assessing specific investments and breaking down the multiple areas of impact into a set of measurable functionalities and benefit categories. This approach, which traces causality only to external influences beyond the assessment scope [39], inadequately captures the feedback-governed, self-organizing, and nonlinear impacts of digitalization. In this sense, digital technologies such as AI can furnish information to enhance decision-making capabilities, leading to subsequent decisions on reinvestment and the optimization of operation and maintenance (O&M) schedules. This in turn can facilitate the creation of complex interactions involving infrastructure, performance, new decisions, and infrastructure improvements. The problem of capturing these nonlinear relationships has led to difficulties in monetizing the benefits of grid modernization investments [40], highlighting the need for novel assessment tools [41].

Recognizing energy companies as entities influenced by multiple interdependent factors, and the actions of numerous stakeholders, a dynamic CBA is proposed in this paper. In this context, dynamic CBA refers to a form of economic assessment that takes into account the complex and nonlinear relationships, interactions, and influences between digitalization impacts. These interactions can arise from the internal structure of a business model, operational and managerial decision-making, and interactions among stakeholders. This dynamic CBA differs from the practice of treating a new project as the result of ex-ante static assumptions. Dynamic CBA deviates from the viewpoint that it is solely the action of exogenous linear inputs (e.g., static cost conditions and fixed investment rates) that causes digitalization impacts. The application of dynamic CBA is based upon system dynamics (SD), a modeling methodology that facilitates the understanding of complex system behaviors [42].

SD has served as a computer-aided method for strategic development and better decision-making [43], which has its roots in control systems theory [44]. A fundamental concept within the theory of SD revolves around feedback loops. These loops involve two or more causal links between elements, forming a causal chain where tracing causality from any element eventually leads back to the initial element. Two types of feedback loops can be distinguished: reinforcing and balancing. Reinforcing feedback loops amplify the increase or decrease in a variable, fostering exponential growth or decay when considered in isolation [45]. Balancing feedback loops counter changes, promoting system stability [42]. Both reinforcing and balancing feedback loops induce nonlinear behavior even when the underlying causal relationships are linear. Feedback loops are a central element in SD modeling, which comprises several steps, ranging from defining the model’s scope to qualitative and quantitative modeling, validation, testing, and sensitivity analysis [46]. These steps have been applied in the literature to unveil dynamic feedback loops qualitatively and quantitatively over time.

Qualitative SD models, such as causal loop diagrams (CLDs), have been employed in modeling and analyzing business models characterized by complexity and unpredictability [47]. CLDs have proved instrumental in facilitating the design and testing of strategies for performance management, sustainable development, and organizational change [48]. An example of qualitative business modeling is demonstrated in Ref. [49], in which the fundamental value dimensions essential to a sustainability-focused business model, specifically, value creation, value proposition, value capture, and environmental value proposition, are systematically displayed from a systemic viewpoint. However, although SD modeling enables analysis of the causal interdependencies among business model variables, an explicit analysis of digitalization as a value dimension impacting the business model remains an open issue [50].

Quantitative SD models, also known as stock-and-flow diagrams (SFDs), have been applied as a suitable tool in applications where empirical data is limited [51], being particularly well-suited for economic and regulatory impact assessments [52], [53]. For instance [54], utilized SD modeling to scrutinize the feedback loops within mining operation systems, quantifying variables, and interactions, and estimating project volatility. Additionally [55], developed an SD model to simulate energy savings by promoting the adoption of renewable energies and energy efficiency. In addition [56], constructed an SD model to assess the impacts of a new transportation project, revealing a broader spectrum of benefits and significantly shorter payback periods compared to conventional static financial analyses. Furthermore [57], presented a novel, open-source CBA model based on NPV analysis, specifically tailored for protection, automation, and control systems in electrical substations. The study indicates positive NPVs for virtualized substations from both financial and societal perspectives. Likewise [58], developed an SD-based model for firm valuation under several debt structures and tax rate scenarios. Within the domain of digitalization assessment, several studies have investigated the application of SD in evaluating business performance and fostering innovation [59], [60]. However, the systematic and generic integration of CBA and SD, and assessing digitalization’s impact on the energy industry remain unresolved challenges.

Building on the preceding theoretical groundwork, and recognizing current limitations, this paper introduces a methodological framework with two objectives. Primarily, it formulates an innovative framework and methodology with the potential to enhance the economic appraisal of digitalization projects. Secondarily, it describes the complex causal connections and feedback loops among diverse value dimensions within energy companies and the digitalization process. The study’s novelty lies in its systemic exploration of the mechanisms governing costs and benefits in digitalization projects. It conceptualizes a digitally transformed business model and delineates the financial and social CBA from an SD perspective in Section 2. This exploration, bridging CBA, digitalization in energy, SD, and business modeling, culminates in the identification of significant feedback loops influencing the impacts of digitalization. Following such a conceptual framework, the paper introduces the first methodology explicitly designed to conduct a dynamic CBA of digitalization in the energy sector, as detailed in Section 3. The efficacy of the proposed methodology is demonstrated through an illustrative use case conducted by the EU JRC in Isernia, Italy [61], as described in Section 4. This use case serves as a benchmark for the application of the proposed dynamic CBA, whose results are presented and discussed in Section 5. Finally, conclusions are drawn in Section 6.

2. Dynamic CBA framework

As digitalization reshapes an energy company’s offerings to its customers, as well as its targeted customer segments, distribution channels, core competencies, cost structure, and revenue models [62], a holistic perspective must be adopted. Thus, this paper introduces a framework displaying the interdependencies among digitalization and business models. In this framework, displayed in Fig. 1, each value dimension layer is characterized by a generic list of variables that can be added, removed, or extended depending on the specific case study in the economic analysis. While making such variable selections involves subjectivity, the general description presented here draws support from well-cited literature, as suggested in Ref. [39]. This CLD employs arrows and variables to link each of the layers, which divide the general economic analysis into a graphic set of value dimensions. From a high-level perspective, the most relevant causal links between the layers are displayed along with their positive or negative signs, signifying whether changes in connected variables are aligned to or opposed to one another. This layered approach expands on insights from Ref. [49] by including dimensions associated with digitalization (2.1 Value creation layers, 2.2 Value proposition layers). Furthermore, value capture layers are leveraged to graphically describe the main elements of a dynamic CBA, both from a financial and social perspective (Section 2.3).

2.1. Value creation layers

One of the fundamental goals of a business model is value creation. Such models show the resources, infrastructure, personnel, and assets that enable operational functionalities and that are aligned with business objectives, culminating in the production of a value proposition Ref. [63]. Here, as shown in Fig. 1, a distinction is made between digital value creation and conventional value creation to clarify the role of new digital technologies within the whole business model and the subsequent economic analysis.

2.1.1. Conventional value creation

In the realm of energy companies, relevant variables such as energy infrastructure capacity, assets, and the number of personnel, play a crucial role in shaping the company’s capacity to meet targeted production capacity and address demand. These variables in value creation have a direct impact on the customer value proposition (Causal link 1), thus facilitating the expansion of customer numbers through an increased portfolio of products and services. This is evident in the strategies employed by original equipment manufacturers in the renewable energy sector, where diversification of the project portfolio includes solar power plants, offshore wind power plants, and integration with energy storage technologies [9].

Generally, value creation is influenced by maintenance strategies, which are crucial in mitigating the likelihood of failures through distinct approaches such as corrective and preventive maintenance [64]. Corrective maintenance is initiated in response to specific failures, addressing issues as they arise. In contrast, preventive maintenance involves regular inspections designed to prevent potential equipment malfunctions. Within this context, internal feedback loops can emerge from the maintenance policies implemented. In fact, both maintenance approaches can dynamically alter the remaining useful life of these assets, which ultimately determines the probability of failure [65], resulting in corrective maintenance actions. These O&M activities incur deterministic and stochastic costs, including scheduled asset replacements, operational costs of personnel and assets, and unexpected expenses stemming from events such as supply disruptions or process failures caused by adverse weather conditions or unexpected equipment failures [66].

2.1.2. Digital value creation

This layer includes the role of data infrastructure, digital skills, and advanced technologies to modify existing processes, streamline value chains, and optimize decision-making within a business model [67]. Here, the digital dimension is represented as a diverse spectrum of tools and technologies, including ICTs, IoT, AI, big data (BD), robotics, cloud computing, virtual reality, and blockchain [68]. These technologies enable diverse functionalities within a company’s core processes, including process monitoring (e.g., state estimation and condition monitoring), thereby enhancing operational and planning efficiencies. Moreover, they enable predictive maintenance strategies that potentially extend the lifespan of assets [67]. For instance, in the context of power transmission and distribution grids, digital assets play a pivotal role in integrating flexibilities by enabling sector-coupling generators, consumers, and storage units [69]. This enhanced flexibility has the potential to extend assets’ lifespan, consequently deferring capital expenditure (CAPEX) [70] and fostering value creation (Causal link 2).

As digitalization enables a more data-driven approach to agile product development and market analysis, it fosters conventional value propositions (Causal link 3). Beyond improving operational efficiencies, digital technologies play a crucial role in introducing innovative service tools and enhancing customer experience and service (e.g., e-business and marketing) [71], enabling companies to augment their customer value propositions (Causal link 4). The impacts of digitalization on both value propositions and traditional value creation are undeniably connected to the scalability of digital technologies. Scalability refers to the ability of these technologies to handle growing data volumes and integrate additional resources while remaining adaptive and cost-effective [68]. Scalability is crucial for the cost-effectiveness and timely implementation of digitalization. Budget constraints and risk-aversion can hinder a full-scale roll-out of digital technologies at once, suggesting a more modular deployment. A scalable digital value creation process can leverage economies of scale, enabling cost-effective solutions and faster deployment through reduced hardware and software costs [72]. Technologies with high scalability also offer better compatibility and interoperability [73], allowing for modular expansion and integration with existing systems, further accelerating implementation. This is crucial as the variety, velocity, and volume of data generated in the energy sector expand.

Similar to conventional assets, digital technologies entail O&M costs associated with software updates, hardware replacement, data handling, and personnel training. Within the costs associated with digitalization, cyber security costs deserve special consideration. The interconnectedness of digitalized energy systems with intelligent devices creates numerous potential attack vectors used by cyber threat actors. Servers, networks, websites, mobile applications, software, firmware, web services, cloud platforms, sensors, motors, relays, and hardware are all susceptible to cyber attacks [74]. These attack vectors are related to several cyber security risks, including data breaches, malware and ransomware, denial of service (DoS) attacks, insider threats, supply chain attacks, and vulnerabilities in control systems, as detailed in Ref. [75]. These attacks can lead to financial losses, operational downtime, legal penalties, and reputational damage, resulting in losses in customer value proposition and conventional value creation [76]. Data breaches affect ICTs by compromising customer and operational data, while malware, ransomware, and DoS attacks disrupt service delivery and increase mitigation costs. Vulnerabilities in conventional value creation can cause catastrophic failures, posing significant safety and reliability risks and necessitating substantial security upgrades [77].

2.2. Value proposition layers

Value proposition serves as a representation of the value extended by the energy company to its customer base through the provision of diverse products and services. Following the structure defined in the value creation layers, Fig. 1 shows a dichotomy between digital and conventional value propositions.

2.2.1. Conventional value propositions

Within this layer, common variables to be mapped include customer management and the offering of products and vices. These include reductions in energy prices, revenue generation from energy and related services, reductions in energy consumption and electricity bills, enhanced power quality, better management of customers’ carbon footprints, alternative pricing strategies, and advocating for more sustainable energy sources [62]. In the energy sector, differences in value propositions arise from the heterogeneity in products, services, and markets. For instance, within multiple markets organized as natural monopolies such as gas and power transmission and distribution [67], revenues, value propositions, and product quality and performance are regulated by public organizations, in contrast to liberalized and competitive markets.

2.2.2. Digital value propositions

The digital value proposition transcends conventional energy offerings by leveraging digitalization to redefine customers’ interactions and experiences. Leveraging digital infrastructure and technologies (Causal link 4), these applications offer a diverse array of capabilities in the energy sector, ranging from customer analytics to advanced energy-efficiency monitoring and evaluation [78]. By augmenting product quality and introducing novel functionalities, companies have the potential to enhance customer attraction and innovation in value propositions (Causal link 5) [79]. For instance, digital innovations in energy supply, transmission, and distribution can impact energy efficiency, enabling energy to be generated and consumed more sustainably.

2.3. Value capture layers

Value capture layers include the revenue and cost streams arising from the business model’s activity and sales. Financial and social value capture layers enable a graphic representation of the CBA from the viewpoint of both the project promoter and society, including all market, environmental, and third parties with standing. These layers are susceptible to uncertainties stemming from assumptions made in the value creation and value proposition layers. Such uncertainties are associated with factors such as the implementation and technology readiness of digital technologies, market and supply-chain dynamics, and social development.

2.3.1. Financial value capture layer

The financial value capture layer aims to represent the aggregation of O&M costs, as well as the benefit categories arising from both the value propositions (Causal link 6) and the value creation layers (Causal link 7).

This involves the incorporation of initial investments, and reinvestments, in digital and conventional infrastructure and products (Causal links 8–11), providing the CBA applications with significant feedback loops. Financial value capture, value creation, and value proposition layers are relative to the project promoter’s perspective. However, in cases where digitalization impacts stem from the business models of multiple stakeholders, additional conceptualization may be needed. This involves replicating these value dimensions for each stakeholder. For instance, in a business-to-business scenario, an extended modeling effort can be undertaken by dynamically connecting the value proposition of the supply with the value creation of the demand, thereby linking both business models. This multi-stakeholder perspective may be incomplete without considering the environmental and social impacts, which are included in the social value capture layer.

2.3.2. Social value capture layer

Beyond singular or multi-corporate interests, this layer embraces a social approach, taking into account environmental and social impacts. Here, the financial impacts are included and aggregated (Causal link 12) with positive and negative externalities such as a reduction in primary energy consumption, emission volumes, and cost impacts for other stakeholders, such as the costs of potential cyber attacks [76].

Particularly relevant externalities are knowledge spillovers, as they foster the growth of startups, promote knowledge-sharing, enhance supply-chain efficiency, and contribute to human capital development [80]. The realization of both social and financial benefits contributes to expected profitability and a reduction in the risk-aversion of investors. Thus, metrics from social and financial CBA can hinder or prompt reinvestments (Causal link 13), creating a dynamic CBA that unfolds over time.

3. Dynamic CBA methodology

Building upon the framework layers, this paper proposes a methodology for applying dynamic CBA to digitalization. The proposed methodology involves multiple sequential steps, as illustrated in Fig. 2. Firstly, mapping digital technologies along their value dimensions is concerned with defining the type and purpose of digital assets under consideration, classifying them under the digital value creation and digital value proposition layers (Step 1). Once the digital infrastructure has been mapped onto the corresponding layers, the interdependencies between digitalization impacts and the entire business model can be identified (Step 2). As depicted by the generic framework in Fig. 1, conventional value creation, value proposition, and value capture will be affected by the new or enhanced functionalities (e.g., predictive maintenance and optimal management of energy storage systems).

The formal modeling process begins with a comprehensive assessment of the investment and operational costs associated with digital technologies, encompassing initial expenditures such as software development, IoT infrastructure, and ongoing updates and maintenance. These costs are evaluated, considering both deterministic and stochastic expenses, particularly those linked to low-probability, high-consequence events such as cyber security incidents. For instance, the financial implications of these expenditures are weighed against the projected costs stemming from cyber attacks, calculated as the product of their likelihood and potential impact. Despite the recognized significance of cyber security risks across the energy sector, their impacts are frequently overlooked in traditional CBAs. However, employing a dynamic CBA framework allows for the modeling of cyber security implications as feedback loops within the CBA: one reinforcing cyber security through targeted investments, and another potentially increasing cyber attack vulnerabilities due to expanded data collection, automation, and interconnected systems.

Subsequently, when identifying benefit categories, the functionalities of digital technologies (e.g., optimization and predictive maintenance) must be listed and linked to the variables in the conventional value layers (e.g., preventive maintenance inspections and remaining useful life). Then, as carried out in Ref. [33], functionalities are mapped onto benefits, and these costs and benefits are causally linked to the financial and social value capture layers (Step 3). This linkage encompasses both deterministic and uncertain benefits, including considerations such as the mitigation of cyber security risks. The resulting CLD, which illustrates the underlying feedback loops among value dimensions, requires validation through expert interviews and a review of the literature (Step 4) [81].

When the qualitative model is validated, the next step relies on the quantitative formulation and testing of the quantitative model, enabling both financial and social CBA (Step 5). Here, SFDs are defined by employing algebraic equations that depict the variables identified in each layer and that simulate system behavior across diverse scenarios and parameter values, as outlined by Refs. [46], [51]. These SFDs delineate the system’s state, capturing its status through stocks, which act as a memory, while illustrating the dynamics of change via inflows and outflows (i.e., flows) over time. According to Ref. [54], this relationship is mathematically described in Eq. (1):

(1)

S (t) = \int_{t_{0}}^{t} [I (t) - O (t)] d t + S (t_{0})

where the notation S(t) denotes the cumulative stock from the initial time t₀ to the current time t. S(t₀) represents the initial value of the stock, while I(t) and O(t) portray the inflow and outflow rates influencing the cumulative stock over time. Examples of such stocks within this framework include variables such as the number of customers, number of staff, digital and conventional asset capacity, and cumulative earnings.

Quantitative validation of the evolution of the various stocks and variables is essential to ensure the model accurately reflects the system’s actual behavior (Step 6). This validation involves comparing digitalization cost and benefit categories with existing data and expected values, as outlined by Ref. [82]. Validation may include verifying the model structure, testing its parameters, simulating extreme conditions, and assessing the model’s ability to replicate system behavior [44]. Such validation can be both qualitative and quantitative to guarantee the defined stocks and causal links represent the business dimensions and their incremental impacts with digitalization.

Following validation, the model’s simulation produces financial and social metrics, which serve as the outcomes of the dynamic CBA (Step 7). This is illustrated in Fig. 1, where the costs and benefits arise from the value creation and value proposition layers. To present a comprehensive formulation of both financial and economic CBA metrics, this methodology utilizes the expected value operator (

E

), acknowledging the stochastic nature of model assumptions. This stochasticity is employed in the methodology to model low-probability high-consequence events such as cyber attacks and power outages [83].

Within this methodology, a distinction is drawn between static and dynamic categories of costs and benefits. Dynamic impacts are manifested through multiple feedback loops triggered by decisions on reinvestment, maintenance, and operations, as noted by Ref. [48]. These loops lead to adjustments in the initial assumptions embedded within the economic models. In alignment with this perspective, the proposed methodology advocates for considering the dynamic impacts to compute the financial and social metrics. This conceptualization is illustrated by the calculation of the project NPV, expressed through the following formulation:

(2)

E {[NPV]}_{T} = E {[NPV]}_{s} + E {[NPV]}_{d}

where, the total expected NPV of digitalization initiatives (

E

[NPV]_T) is defined as the sum of the expected NPV obtained from static benefits and costs (

E

[NPV]_s), and the expected NPV, including the value arising from feedback loops (

E

[NPV]_d). This formulation encapsulates dynamic revenue and cost streams, which are often overlooked in traditional assessments. Overall, the proposed methodology offers significant advantages for assessing the long-term value and risks of digital technologies compared to state-of-the-art methods. These advantages stem from several factors, including the modeling basis, information handling, perspective considered, tool transparency, validation procedures, and the approach to addressing uncertainty, as detailed in Table 1.

Having identified both the dynamic and static impacts, the accountability of these cost and benefit categories among stakeholders becomes a relevant issue. Positive and negative impacts for other stakeholders than the project promoter must be carefully accounted for, avoiding double-counting by not including the transfer impacts, which are costs from one stakeholder accounted as benefits from other stakeholders [84]. For instance, a reduction in operational costs by one stakeholder can lead to an increase in revenue for a customer, and only one of them should be included to guarantee the accuracy of the final financial and social metrics. Acknowledging the stochastic nature of these metrics and varying profiles of assumptions and parameters, an uncertainty analysis must be carried out (Step 8). In this final step, various methods such as single point sensitivity analysis, scenario analyses, or Monte Carlo simulations [85] can be utilized. In particular, Monte Carlo simulations can provide valuable insights into the individual influence of model parameters on both financial and economic metrics, enabling a deeper understanding of the sensitivity of assumptions, model parameters, and simulation scenarios.

4. Application case

This section provides a practical demonstration of the methodology described here through an illustrative application case, presented in Section 4.1. Next, in Section 4.2, an SD model is constructed and validated against the results obtained through the EU JRC methodology outlined in Ref. [33]. Then emphasis is placed on considering the reinvestment feedback loops illustrated in Fig. 1 (Causal links 8–11). To formulate the equations of the SD model and simulate its yearly behavior over a time horizon of nine years, this application proposes to use of InsightMaker, an open-source and versatile tool for web-based modeling and simulation [86].

4.1. Overview of case study

The chosen case study shows the application of a CBA to assess the financial and social costs and benefits of implementing a smart grid initiative in the electricity distribution grid in the Isernia region of Italy [61]. This case holds relevance as a large-scale smart grid project, integrating both conventional and digital assets, and stands as a prominent example of both past and ongoing digitalization within the energy sector. Furthermore, it utilizes a robust and well-established CBA framework that considers regulatory interactions and end-user impacts, providing a comprehensive evaluation of the financial and societal advantages of digitalization. Additionally, the JRC publication offers a detailed set of assumptions, data, results, and sensitivity analyses, ensuring the replicability of this case and positioning it as a valuable benchmark for the proposed methodology.

Key stakeholders include the distribution system operator (DSO), the national regulatory authority (NRA), the final electricity users, and the environment. The DSO, serving as the project promoter, executed the smart grid investment. This project, initiated in 2011, aimed to enhance the protection, automation, and management of power distribution grids. To realize these benefits, CAPEX are incurred through the deployment of supervisory control and data acquisition (SCADA) systems, measurement devices, modems and routers, smart information interfaces, and electric vehicle (EV) charging infrastructure. Investments can yield both financial and social benefits. Firstly, financial benefits are derived through the regulated weighted average cost of capital (WACC), which serves as the remuneration rate for the financial CBA. In the Isernia case [61], two main scenarios are considered for the WACC: Scenario A with a 7.4% regulated WACC, and Scenario B with an extra remuneration of 2.0% WACC, a total of 9.4% WACC. Secondly, social benefits arise from electricity savings for customers, primary energy savings resulting from the increased use of renewables, and emissions reductions. Significantly, in the initial CBA study conducted by Ref. [61], it was assumed that financial benefits are included as social benefits. Thus, for the sake of establishing benchmarks, the same assumption is integrated into the SD model.

4.2. Dynamic CBA application

4.2.1. Qualitative model description

The conceptualization of the SD model is depicted in Fig. 3(a), in which a CLD portrays the CBA variables specific to the Isernia case. However, this conventional CBA does not consider the feedback loops arising from the decisions of reinvesting due to a positive financial NPV, or due to an increase in the regulated WACC. Thus, a CLD of the dynamic CBA is presented in Fig. 3(b). The latter treats the project as not operating continuously at a set timescale, which may reflect the development process of smart grid projects more accurately [87]. Within this schematic representation, both reinforcing and balancing feedback loops driving reinvestment rates are discernible.

Reinforcing feedback loops may arise when stakeholders observe the positive financial and social benefits, caused by the incremental benefits produced by the increasing stock of digital products and assets. These loops can enable an increased share of RES and electricity savings, subsequently resulting in an elevated remuneration from the NRA to the DSO. Consequently, the WACC rises, positively impacting the financial NPV. The project promoter interprets this phenomenon as a compelling incentive for further investments, which can lead to an increase in the willingness to invest (WTI) in smart grid technologies. This WTI is affected by the technology maturity and scalability of digital technologies. Conversely, balancing feedback loops are manifested from the expenses associated with investing in digitalization. These costs can hinder the adoption of new investments, given their adverse impact on both financial and social NPVs. The general SFD for the Isernia case is constructed and shown in Fig. 3(c). In this SFD, feedback loops and stock variables, such as the amount of investment in digital assets, drive benefits and introduce nonlinear relationships. As anticipated in the CLD, reinvestment is considered through the correlation between social DCFs and an additional WACC factor for the DSO.

4.2.2. Quantitative model description and validation

The model equations illustrate the causal chains of benefits and costs, and the model parameters are detailed in Table 2, Table 3. In this study, a rule-based approach is employed to articulate reinvestment decisions. Specifically, reinvestment is delineated by the correlation between the cumulative social DCFs, which in turn can trigger an additional WACC factor accrued by the DSO. This assumption applies to the original Isernia case, in which the NRA offers extra remuneration to the DSO for investment recovery. Consequently, this paper builds upon this assumption, positing that when social DCFs exceed zero, the WACC is gradually assigned to the DSO. This gradual assignment influences the DCFs for the DSO and, when these DCFs are positive, reinvestment rates from the project promoter escalate, leading to additional social benefits and costs. Due to space constraints, the table exclusively presents the key equations describing benefit, cost, financial and social metric calculations, and reinvestment feedback loops. The comprehensive set of equations, original data from Ref. [61], and the complete SD model can be found in Appendix A.

To evaluate and benchmark the simulation results, a first step in validation must be carried out to determine whether the SD model reproduces the original results of the Isernia CBA. In this paper, the yearly financial and social benefits are utilized as reference modes to test the capability of the model, as shown in Fig. 4. Here, the social benefits from customers’ electricity savings, primary energy savings, and emissions savings are shown in Figs. 4(a)–(d), whereas the financial benefits are simulated for the scenario of BaU WACC, and the scenario with extended WACC (see Figs. 4(e) and (f)). It is noted that the simulation results replicate the baseline, which suggests that the SD model is valid according to Ref. [82]. However, minor discrepancies in the validation arise due to differences in the application of discount rates (normal or real terms) and the absence of data on benefits. Despite these variations, the validation process confirms the validity of the SD model.

4.3. Limitations and uncertainties

The proposed application case leverages an existing real CBA to benchmark the potential of the dynamic CBA methodology. However, this comparison is subject to several limitations and uncertainties. Firstly, the assessment is constrained by numerous assumptions embedded in the original Isernia CBA. These include a fixed social discount rate of 2.5%, a financial discount rate of 4.0%, and an assumed CO₂ emission price of 15 EUR·t⁻¹ of CO₂, an optimistic estimate given the evolving electricity prices in the European Emission Trading System [88]. Additionally, the initial assumption of a regulated WACC of 7.4% and a time horizon of ten years imposes limitations on the benefit categories considered. Without these constraints, varying discount factors could be applied to better reflect evolving risk prices [89] and long-term benefits. Indeed, a longer time horizon could account for deferred investments in grid infrastructure and facilitate higher rates of RES integration. Furthermore, the costs associated with operating and replacing digital technologies were not included in the application, as these were beyond the scope of the original Isernia CBA [61]. In addition to model limitations, results are affected by several uncertainties related to parameter definition and the construction of the causal links in the SD model. A selection of the major sources of uncertainty, as carried out in Ref. [9], includes: ① the evolution of WACC and WTI, subject to uncertainty analysis; ② external inputs such as electricity prices and emission factors; and ③ uncertainty in the rules for reinvestments.

From an economic perspective, WACC is subject to changes due to evolving regulatory frameworks and the cost of debt. Additionally, extra WACC remunerations may arise to compensate for the costs of digital technologies or their effects on supply quality and power loss reduction. Such increases in WACC may impact DSOs differently, depending on their WTI, which is significantly influenced by DSO size, digital skills, financial sustainability, and organizational barriers [23]. Similarly, uncertainties related to the spot market are significant when considering long time horizons [90]. Finally, reinvestment rules are defined to react when metrics are positive, and investments are recovered, from both social and financial perspectives. However, these patterns may change depending on the risk perception of both the DSO and the NRA.

5. Results and discussion

This section involves simulating the SD model across four selected scenarios, which are described and discussed in Section 5.1. Simulation results are obtained using the fourth-order Runge–Kutta algorithm [91] provided in InsightMaker and recommended for SD modeling. Additionally, to take uncertainty into account, Section 5.2 includes the analysis of the relevant parameters influencing the DCFs.

5.1. Scenario analysis

The process of benchmarking the dynamic CBA results is carried out by comparing Scenarios A and B, which exclude feedback loops or reinvestment considerations, with Scenarios C and D, outlining the model’s evolution under specific reinvestment and WTI conditions. The general description of each scenario is detailed in Table 4. This paper investigates whether an expansion of WACC can lead to higher financial profitability, increased WTI, and, consequently, further reinvestments in digital technologies and EV charging infrastructure, thereby impacting both the financial and social benefits.

Model results for the selected scenarios are illustrated in Fig. 5, which presents the simulation of cumulative and early DCFs from both the financial and social perspectives for each scenario. It is important to note that the final value of the cumulative DCFs signifies the aforementioned metric, NPV. In the figure, from a top-down approach, the two first subplots portray the evolution over the nine-year simulation time horizon of social metrics. Remarkably, Scenarios A and C share an identical initial investment of 7.4 million EUR and maintain indistinguishable social DCFs until they reach the break-even point in the fifth year. Conversely, Scenarios B and D, simulated under a higher initial WACC rate (9.4%), achieve break-even by the fourth year. Following these break-even points, the dynamic costs and benefits come into play, resulting in additional WACC remuneration. This leads to the differences in NPVs, outlined in Table 5, among Scenarios A and B, and among Scenarios C and D.

Subsequently, as the cumulative financial DCFs turn positive due to an increasing WACC, the DSO may decide to invest in new digital technologies and EV chargers. While this may initially reduce the financial and social metrics due to added investment costs, this leads to a substantial increase in the social metrics. As depicted in the figure, the social metrics consistently surpass the financial metrics, since these include both the project promoter and multi-stakeholder benefits. Specifically, as summarized in Table 5, the differences between the dynamic and static scenarios are influenced by the original assumption of considering additional WACC for DSO as a social benefit. This is also illustrated in Fig. 5, Scenario B, in which this revenue stream represents the main driver since reinvestments are not triggered until the final year.

Besides the WACC contribution, the effect of reinvestment appreciates in Scenario D, where it impacts yearly DCFs. In this case, positive DCFs from environmental benefit categories offset additional reinvestment costs, making the case for considering both dynamic and static NPVs. This is consistent with the hypothesis described in Eq. (2), in which taking account of reinforcing and balancing feedback loops can lead to varying financial and social metrics, and thus change investment decisions. This scenario is exemplified in Scenario C, where, unlike its static counterpart, Scenario A, a positive NPV may lead the DSO to approve the investment decision, as the investment could potentially be recovered. In essence, the inclusion of feedback loops can significantly alter investment decisions. While conventional CBA may not result in a positive financial NPV, the simulation results show that scenarios incorporating reinvestment can effectively capture the nonlinear effects of the balancing and reinforcing feedback loops driving digitalization. Such outcomes demonstrate that project valuation varies significantly when considering factors such as reinvestment rates and WTI in digitalization projects. This inclusion allows for a more comprehensive understanding of the impact of conflicting feedback loops that may not be included in static approaches.

5.2. Uncertainty analysis

Inherent uncertainties arise from the reinvestment feedback loops, being contingent on specific uncertainty parameters. This uncertainty analysis aims to elucidate their influence on the evolution of cumulative DCFs. It encompasses a probabilistic assessment through a Monte Carlo simulation of the social and financial metrics, taking into account the variability of the WACC factor and the WTI factor, both of which significantly impact model dynamics. The WACC factor is assumed to follow a normal distribution with a mean (μ_wacc= 1.5) and standard deviation (σ_wacc = 1.0), while the WTI factor follows a normal distribution with a mean (μ_wti = 1.0) and standard deviation (σ_wti = 0.2) for the WTI factor. Subsequently, a Monte Carlo analysis employing 2500 samples is conducted by leveraging the sensitivity analysis toolbox in InsightMaker. The assumption of a normal distribution is applied for convenience, as is done in Ref. [92]. The selection of the number of samples is guided by observing the financial and social metrics until significant convergence is achieved. This analysis explores both financial and social cumulative DCFs for the dynamic Scenarios C and D.

The results of the uncertainty analysis are visualized in Fig. 6, which presents the time series of mean values of cumulative DCFs alongside the 95% lower and upper boundaries. This graphic representation effectively conveys the inherent uncertainties in Scenarios C and D. Significantly, Scenario D emerges with higher financial and social metric values when reinvestments are taken into account. However, it also exhibits greater variability, influenced by the impact of investment costs, and leading to a reduction in DCFs from year six. A noteworthy observation is drawn when comparing the sensitivity analysis of Scenario C with the results for Scenario A. Due to the high levels of uncertainty regarding reinvestments, cumulative DCFs consistently differ for both financial and social metrics in scenarios considering SD. This variation reflects the underlying benefits and costs driven by the balancing and reinforcing feedback loops. Whereas the balancing feedback loops increase costs and reduce the WTI, reinforcing feedback loops foster new investments, which subsequently can produce new benefits and spur reinvestment.

6. Conclusions

The profound influence of digitalization on the energy sector is evident through the introduction of competitiveness and cost reductions across the entire energy value chain. However, digitalization has faced limited adoption due to challenges in quantifying its benefits and costs. Traditional CBA approaches have considered static project operations and limited assessment scopes. This has hindered estimates of the complex and dynamically influential impacts of digitalization. Adopting an SD modeling approach, this paper has formulated an innovative framework and methodology with the potential to enhance the economic appraisal of digitalization projects. By introducing the concept of dynamic CBA, which treats both the financial and social perspectives, the proposed framework provides a comprehensive basis for assessing the effects of dynamic feedback structures that shape business models within the energy industry. This methodology enhances the capture of long-term value chains and risks associated with the adoption of digital technologies through its comprehensive assessment of uncertainties, model-based approach, and inclusion of nonlinear relationships that influence digitalization costs and benefits. Moreover, although this paper primarily focuses on the energy sector, the principles underpinning this methodology can also be extended to various other economic sectors. The application of this framework to an illustrative use case has demonstrated its practical and theoretical significance. Firstly, in practical terms, more accurate modeling and assessment of the dynamic financial and social impacts can lead to changes in managerial decisions and enhance the accuracy of decision-making. This can mitigate decision-maker’s risk-aversion and thus address a significant factor hindering the widespread adoption of digitalization. Likewise, this paper may aid in conceptualizing the impacts of digitalization on the value dimensions of a business model. Such an approach can facilitate a seamless identification of the causal chains driving the cost and benefit streams arising from digitalization. On a theoretical level, this paper emphasizes that digitalization impacts are profoundly influenced by interactions, feedback loops, and delays among different components of a business model. To address this complexity, the proposed framework may be particularly useful for the management and engineering community, as they must acknowledge the dynamic costs and benefits that are present in operational and managerial decisions, influencing the adoption of technology and its maturity.

While the dynamic CBA approach offers notable advantages, its adoption presents challenges such as defining variables for feedback loops, modeling complex correlations, and quantifying diverse impacts across stakeholders and sectors. Accurate definition of the scope and key variables, such as stocks, flows, and feedback loops, is essential for effective modeling, with proper characterization of business layers and their interconnections with digitalization. Additionally, the diversity of case studies and business models introduces variability into the specific SD equations, demanding expertise to cope with uncertain causal chains, technological innovation levels, firm types, and regional differences. For reducing uncertainty, rigorous calibration and validation of the SD model using historical data, alongside sensitivity analysis, probabilistic approaches, and clear documentation of feedback loops, ensures robust, reliable, and transparent results that align with real-world behavior and support informed decision-making. Overcoming these challenges demands multidisciplinary cooperation among stakeholders such as economists, engineers, managers, and regulatory authorities. Future work should focus on further enhancing the practical application of the proposed methodological framework. This involves identifying new feedback loop structures for digitalization and extending the framework for multiple stakeholders across the energy value chain. This promising avenue for further research can be instrumental in maximizing the net benefits of digitalization in energy.

Acknowledgments

The authors are grateful to reviewers and editors for their helpful and constructive comments on this paper. This study was conducted as part of the project Innovative Tools for Cyber–Physical Energy Systems (InnoCyPES), which has received funding from the European Union’s Horizon 2020 research and innovation program under the Marie Skłodowska-Curie (956433).

Compliance with ethics guidelines

Jose Angel Leiva Vilaplana, Guangya Yang, Emmanuel Ackom, Roberto Monaco, and Yusheng Xue declare that they have no conflict of interest or financial conflicts to disclose.

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.eng.2024.11.005.

References

Publishing order | Descend order by publishing year | Descend order by cited within

[1]	Heymann F, Milojevic T, Covatariu A, Verma P.Digitalization in decarbonizing electricity systems-phenomena, regional aspects, stakeholders, use cases, challenges and policy options.Energy 2023; 262:125521.

[2]	Palensky P, Cvetkovic M, Gusain D, Joseph A.Digital twins and their use in future power systems.Digital Twin 2021; 1:4.

[3]	Tuballa ML, Abundo ML.A review of the development of smart grid technologies.Renew Sustain Energy Rev 2016; 59:710-725.

[4]	Fathi NY, Yassin AS.Wind energy potential in Gaza Strip-Palestine state.Sol Energy Sustain Dev J 2021; 7(2):41-57.

[5]	Awad H, Nassar YF, Hafez A, Sherbiny MK, Ali AF.Optimal design and economic feasibility of rooftop photovoltaic energy system for Assuit University, Egypt.Ain Shams Eng J 2022; 13(3):101599.

[6]	Miskeen A, Elzer RS, Mangir IK, Nassar Y, El-khozondar H, Khaleel M, et al.Electricity from wastewater treatment plants.J Sol Energy Sustain Dev 2023; 12:24-37.

[7]	Quaranta E, Bejarano MD, Comoglio C, Fuentes-P JFérez, P JIérez-Díaz, Sanz-Ronda FJ, et al.Digitalization and real-time control to mitigate environmental impacts along rivers: focus on artificial barriers, hydropower systems and European priorities.Sci Total Environ 2023; 875:162489.

[8]	Nassar YF, Abdunnabi MJ, Sbeta MN, Hafez AA, Amer KA, Ahmed AY, et al.Dynamic analysis and sizing optimization of a pumped hydroelectric storage-integrated hybrid PV/wind system: a case study.Energy Convers Manage 2021; 229:113744.

[9]	Nassar YF, Alsadi SY, El-Khozondar HJ, Ismail MS, Al-Maghalseh M, Khatib T, et al.Design of an isolated renewable hybrid energy system: a case study.Mater Renew Sustain Energy 2022; 11(3):225-240.

[10]	El-Khozondar HJ, El-batta F, El-Khozondar RJ, Nassar Y, Alramlawi M, Alsadi S.Standalone hybrid PV/wind/diesel–electric generator system for a COVID-19 quarantine center.Environ Prog Sustain Energy 2023; 42(3):e14049.

[11]	International Energy Agency (IEA).Digitalization & energy. Technical report. Paris: IEA; 2017.

[12]	WindEurope.Wind energy digitalisation towards 2030 [Internet].Brussels: WindEurope; 2021 Nov 22 [cited 2023 Mar 13]. Available from: https://windeurope.org/intelligence-platform/product/wind-energy-digitalisation-towards-2030/.

[13]	Weigel P, Fischedick M, Viebahn P.Holistic evaluation of digital applications in the energy sector-evaluation framework development and application to the use case smart meter roll-out.Sustainability 2021; 13(12):6834.

[14]	Zhao J, Yue X.Condition monitoring of power transmission and transformation equipment based on industrial internet of things technology.Comput Commun 2020; 157:204-212.

[15]	Barja-Martinez S, Aragü Més-Peñalba, ÍMunn é-Collado, Lloret-Gallego P, Bullich-Massagu Eé, Villafafila-Robles R.Artificial intelligence techniques for enabling big data services in distribution networks: a review.Renew Sustain Energy Rev 2021; 150:111459.

[16]	Wilczek P.Connecting the dots: distribution grid investments to power the energy transition.In: Proceedings of the 11th Solar & Storage Power System Integration Workshop (SIW 2021), Hybrid Conference; 2021 Sep 28; online. London: IET Conference Publications; 2021. p. 1–18.

[17]	Verma P, Savickas R, Strüker J, Buettner S, Kjeldsen O, Wang X.Digitalization: enabling the new phase of energy efficiency. In: Proceedings of the Group of Experts on Energy Efficiency 2020; 2020 Sep 22–25; Geneva, Switzerland. New Delhi: DTU Library; 2020.

[18]	Albadi MH, El-Saadany EF.A summary of demand response in electricity markets.Electr Power Syst Res 2008; 78(11):1989-1996.

[19]	Leiva Vilaplana JA.Modernising electricity distribution grids in Spain: assessing the benefits of digitalisation in the context of the Spanish recovery, resilience, and transformation plans. Fundación de Estudios de Economía Applicada, Madrid (2023)

[20]	Xu Q, Zhong M, Li X.How does digitalization affect energy? international evidence.Energy Econ 2022; 107:105879.

[21]	Di Silvestre ML, Favuzza S, Riva Sanseverino E, Zizzo G.How decarbonization, digitalization and decentralization are changing key power infrastructures.Renew Sustain Energy Rev 2018; 93:483-498.

[22]	International Energy Agency (IEA).Unlocking smart grid opportunities in emerging markets and developing economies. Report. Paris: IEA; 2023.

[23]	Monaco R, Bergaentzl Cé, Leiva Vilaplana JA, Ackom E, Nielsen PS.Digitalization of power distribution grids: barrier analysis, ranking and policy recommendations.Energy Policy 2024; 188:114083.

[24]	European Commission.Digitalising the energy system—EU action plan [Internet].Brussels: European Commission; 2022 Oct 18 [cited 2023 Mar 13]. Available from: https://eur-lex.europa.eu/legal-content/EN/ALL/?uri=CELEX:52022DC0552.

[25]	Murino T, Monaco R, Nielsen PS, Liu X, Esposito G, Scognamiglio C.Sustainable energy data centres: a holistic conceptual framework for design and operations.Energies 2023; 16(15):5764.

[26]	Mondejar ME, Avtar R, Diaz HLB, Dubey RK, Esteban J, Gómez-Morales A, et al.Digitalization to achieve sustainable development goals: steps towards a smart green planet.Sci Total Environ 2021; 794:148539.

[27]	Vilaplana JL, Yang G, Dong Z, Ackom E, Martin F, Pavlov A.Revisit cost–benefit analysis in energy infrastructure: current methodology, practice, and limitations toward digital investment. SSRN Electr J 2023:4505225.

[28]	Flyvbjerg B, Budzier A, Lee JS, Keil M, Lunn D, Bester DW.The empirical reality of IT project cost overruns: discovering a power-law distribution.J Manage Inf Syst 2022; 39(3):607-639.

[29]	Electric Power Research Institute (EPRI).Estimating the costs and benefits of the smart grid: a preliminary estimate of the investment requirements and the resultant benefits of a fully functioning smart grid. Report. Beijing: Electric Power Research Institute; 2011.

[30]	Nordling A, Pädam S, af Burén C, Jörgensen P, Jörgensen P.Social costs and benefits of smart grid technologies. Report. Paris: International Energy Agency; 2018.

[31]	de Rus G.Introduction to cost–benefit analysis: looking for reasonable shortcuts. Edward Elgar publishing, Cheltenham (2010)

[32]	Abelson P.A partial review of seven official guidelines for cost–benefit analysis.J Benefit Cost Anal 2020; 11(2):272-293.

[33]	Giordano V, Onyeji I, Fulli G, Sanchez JM, Filiou C.Guidelines for cost benefit analysis of smart metering deployment. Report. Luxembourg: Publications Office of the European Union; 2012.

[34]	Giordano V, Onyeji I, Fulli G, Sanchez JM, Filiou C.Guidelines for conducting a cost–benefit analysis of smart grid projects. Guidelines for conducting a cost–benefit analysis of smart grid projects. Report. Luxembourg: Publications Office of the European Union; 2012.

[35]	Electric Power Research Institute (EPRI).Guidebook for cost/benefit analysis of smart grid demonstration projects: revision 1, measuring impacts and monetizing benefits.Palo Alto: EPRI; 2012.

[36]	Pilo F, Troncia M.MC-CBA toolkit: model and case study. IEA International Smart Grid Action Network (ISGAN), Rogaland (2013)

[37]	de Kinderen S, Kaczmarek-He Mß, Ma Q, Razo-Zapata IS.Model-based valuation of smart grid initiatives: foundations, open issues, requirements, and a research outlook.Data Knowl Eng 2022; 141:102052.

[38]	Hoover KD.Reductionism in economics: intentionality and eschatological justification in the microfoundations of macroeconomics.Philos Sci 2015; 82(4):689-711.

[39]	Richardson GP.Reflections on the foundations of system dynamics.Syst Dyn Rev 2011; 27(3):219-243.

[40]	Woolf T, Havumaki B, Bhandari D, Whited M, Schwartz LC.Benefit–cost analysis for utility-facing grid modernization investments: trends, challenges, and considerations. Report. Berkeley: Lawrence Berkeley National Laboratory; 2021.

[41]	Marnay C, Liu L, Yu J, Zhang D, Mauzy J, Ba S.Benefit analysis of smart grid projects: white paper 2014–2016. Energy Technologies Area, Berkeley (2016)

[42]	Schieritz N, Milling PM.Modeling the forest or modeling the trees: a comparison of system dynamics and agent-based simulation.In: Proceedings of the 21st International Conference of the System Dynamics Society; 2003 Jul 20–24; New York City, NY, USA. Southampton: University of Southampton; 2003.

[43]	Bale CS, Varga L, Foxon TJ.Energy and complexity: new ways forward.Appl Energy 2015; 138:150-159.

[44]	Forrester JW, Mass NJ, Ryan CJ.The system dynamics national model: understanding socio-economic behavior and policy alternatives.Technol Forecast Soc Change 1976; 9(1–2):51-68.

[45]	Pruyt E.Using small models for big issues: exploratory system dynamics modelling and analysis for insightful crisis management.Int J Inf Secur 2009; 23(1):51-61.

[46]	Bala BK, Arshad FM, Noh KM.System dynamics, springer texts in business and economics. Springer Nature, Singapore (2017)

[47]	Cosenz F, Rodrigues VP, Rosati F.Dynamic business modeling for sustainability: exploring a system dynamics perspective to develop sustainable business models.Bus Strategy Environ 2020; 29(2):651-664.

[48]	Sterman JD.System dynamics: systems thinking and modeling for a complex world. Cambridge: Massachusetts Institute of Technology; 2002.

[49]	Abdelkafi N, Tscher K.Business models for sustainability from a system dynamics perspective.Organ Environ 2016; 29(1):74-96.

[50]	Parida V, Sjödin D, Reim W.Reviewing literature on digitalization, business model innovation, and sustainable industry: past achievements and future promises.Sustainability 2019; 11(2):391.

[51]	Franco MA.A system dynamics approach to product design and business model strategies for the circular economy.J Clean Prod 2019; 241:118327.

[52]	He YX, Jiao J, Chen RJ, Shu H.The optimization of Chinese power grid investment based on transmission and distribution tariff policy: a system dynamics approach.Energy Policy 2018; 113:112-122.

[53]	Hsu CW.Using a system dynamics model to assess the effects of capital subsidies and feed-in tariffs on solar PV installations.Appl Energy 2012; 100:205-217.

[54]	Liu D, Li G, Hu N, Ma Z.Application of real options on the decision-making of mining investment projects using the system dynamics method.IEEE Access 2019; 7:46785-46795.

[55]	Shih YH, Tseng CH.Cost–benefit analysis of sustainable energy development using life-cycle co-benefits assessment and the system dynamics approach.Appl Energy 2014; 119:57-66.

[56]	Nguyen T, Cook S, Ireland V.Application of system dynamics to evaluate the social and economic benefits of infrastructure projects.Systems 2017; 5(2):29.

[57]	Vilaplana JAL, Kabbara N, Coste T, Morais H, Zerriffi H, Gibescu M.Virtualized protection, automation, and control in electrical substations: an open-source dynamic cost–benefit assessment model.IEEE Access 2024; 12:107488-107504.

[58]	Khan A, Qureshi MA, Davidsen PI.A system dynamics model of capital structure policy for firm value maximization.Syst Res Behav Sci 2021; 38(4):503-506.

[59]	Viswanathan R, Telukdarie A.A systems dynamics approach to SME digitalization.Procedia Comput Sci 2021; 180:816-824.

[60]	Qiao W, Ju Y, Dong P, Tiong RLK.How to realize value creation of digital transformation? A system dynamics model.Expert Syst Appl 2024; 244:122667.

[61]	Gianluca F, Silvia V, Gianluca F, Laura M, Jon S.Cost–benefit analysis of smart grid projects: Isernia: costs and benefits of smart grid pilot installations and scalability options. Publications Office of the European Union, Luxembourg (2018)

[62]	Niesten E, Alkemade F.How is value created and captured in smart grids? A review of the literature and an analysis of pilot projects.Renew Sustain Energy Rev 2016; 53:629-638.

[63]	Richardson J.The business model: an integrative framework for strategy execution.Strateg Change 2008; 17(5–6):133-144.

[64]	Mol Męda, Ma Błysiak-Mrozek, Ding W, Sunderam V, Mrozek D.From corrective to predictive maintenance—a review of maintenance approaches for the power industry.Sensors 2023; 23(13):5970.

[65]	Saxena A, Celaya J, Saha B, Saha S, Goebel K.Metrics for offline evaluation of prognostic performance.Int J Progn Health Manag 2010; 1(1):1-20.

[66]	Tazi N, Safaei F, Hnaien F.Assessment of the levelized cost of energy using a stochastic model.Energy 2022; 238:121776.

[67]	Council of European Energy Regulators (CEER).Digitalisation steering group CEER consultation on dynamic regulation to enable digitalization of the energy system: conclusions paper.Brussels: CEER; 2019.

[68]	Andoni M, Robu V, Flynn D, Abram S, Geach D, Jenkins D, et al.Blockchain technology in the energy sector: a systematic review of challenges and opportunities.Renew Sustain Energy Rev 2019; 100:143-174.

[69]	Weigel P, Fischedick M.Review and categorization of digital applications in the energy sector.Appl Sci 2019; 9(24):5350.

[70]	Bhela S, Kekatos V, Veeramachaneni S.Enhancing observability in distribution grids using smart meter data.2016. arXiv: 1612.06669.

[71]	St Bępień.Drivers of digitalization in the energy sector—the managerial perspective from the catching up economy.Energies 2022; 15(4):1437.

[72]	Liorca M, Soroush G, Glovannetti E, Jamasb T. Davi-Arderius D. Energy sector digitalisation, green transition and regulatory trade-offs. Working paper. Frederiksberg: Copenhagen Business School; 2024.

[73]

Vilaplana JAL, Sundsgaard K, Guerreiro GMG, Yang G.Challenges in the representation of digital applications in SGAM: overview and solutions.In: Proceedings of the 2022 IEEE PES Innovative Smart Grid Technologies Conference Europe (ISGT-Europe); 2022 Oct 10–12; Novi Sad, Serbia. New York City: IEEE; 2022.

[74]	Energy Expert Cyber Security Platform.Cyber security in the energy sector. EECSP report. Brussels: European Union; 2017.

[75]	United Nations Economic Commission for Europe (UNECE).Cyber resilience of critical energy infrastructure. Technical report. Geneva: UNECE; 2023.

[76]	Venkatachary SK, Prasad J, Samikannu R.Economic impacts of cyber security in energy sector: a review.Int J Energy Econ Policy 2017; 7:250-262.

[77]	Sani AS, Yuan D, Jin J, Gao L, Yu S, Dong ZY.Cyber security framework for internet of things-based energy internet.Future Gener Comput Syst 2019; 93:849-859.

[78]	Ahmad T, Zhang D, Huang C, Zhang H, Dai N, Song Y, et al.Artificial intelligence in sustainable energy industry: status quo, challenges and opportunities.J Clean Prod 2021; 289:125834.

[79]	Park C.Expansion of servitization in the energy sector and its implications.Energy Environ 2022; 11(4):e434.

[80]	Bouwman H, Nikou S, Molina-Castillo FJ, de Reuver M.The impact of digitalization on business models.Regul Gov 2018; 20(2):105-124.

[81]	Martinez-Moyano IJ, Richardson GP.Best practices in system dynamics modeling: introduction to the importance of best practices.In: Proceedings of the 19th International Conference of the System Dynamics Society; 2001 Jul 23–27; Atlanta, GA, USA. Albany: System Dynamics Society; 2001.

[82]	Barlas Y.Multiple tests for validation of system dynamics type of simulation models.Eur J Oper Res 1989; 42(1):59-87.

[83]	Sperstad IB, Kj Gølle, EØNorum .Accounting for uncertainties due to high-impact low-probability events in power system development.Electr Power Syst Res 2021; 193:107015.

[84]	European Commission.Guide to cost–benefit analysis of investment projects.Technical Report. Brussels: Directorate-General for Regional and Urban Policy; 2014.

[85]	Boardman AE, Greenberg DH, Vining AR, Weimer DL.Cost–benefit analysis: concepts and practice. (5th ed.), Cambridge University Press, Cambridge (2015)

[86]	Fortmann-Roe S.Insight maker: a general-purpose tool for web-based modeling & simulation.Simul Model Pract Theory 2014; 47:28-45.

[87]	Schachter JA, Mancarella P.A critical review of real options thinking for valuing investment flexibility in smart grids and low carbon energy systems.Renew Sustain Energy Rev 2016; 56:261-271.

[88]	Sitarz J, Pahle M, Osorio S, Luderer G, Pietzcker R.Eu carbon prices signal high policy credibility and farsighted actors.Nat Energy 2024; 9(6):691-702.

[89]	Kozak S, Santosh S.Why do discount rates vary?.J Financ Econ 2020; 137(3):740-751.

[90]	European Commission: Directorate-General for Energy; Smith M, Jagtenberg H, Lam L, Torres P, Gorner M, Ansarin M, et al.Study on energy prices and costs: evaluating impacts on households and industry—2023 edition. Final report. Luxembourg: Publications Office of the European Union; 2024.

[91]	Romeo G.Mathematics for dynamic economic models.Elements of Numerical Mathematical Economics with Excel, Academic Press, Cambridge 2020; 139-215.

[92]	Turner JA, Godínez-Olivares H, McCarthy DD.Determining discount rates required to fund defined benefit plans. Society of Actuaries, Chicago (2015)

RIGHTS & PERMISSIONS

THE AUTHOR