Smart Grid Project Planning and Cost/Benefit Evaluation

Abstract

The smart grid involves a set of interconnected ecosystems applications (electrical, electronic, computer and communications), so the modernization is needed of information, security and infrastructure systems that monitor, control and manage them are increasingly evident. The upgrading smart grid process is a complex interaction between different alternatives and adequate selection of assessment criteria, where tangible and intangible items must be chosen with little information or with uncertain data. This work presents a framework within the context of the Smart Grid, to provide electric energy companies with a tool for planning the modernization of their generation, transmission, distribution and marketing systems. The planning of the modernization networks under the smart grid concepts is represented based on Smart Grid Architecture Model (SGAM) reference model. Furthermore, it presents how to integrate a pilot project (Smart Grid information source) into the SGAM reference model, through the identification of key performance indicators defining it. Multi-criteria analysis (MCA) combined with cost/benefit analysis (CBA) concept is explored, providing a novel insight into the approaches used in smart grid research applied at a case study: a distribution grid for rural smart grids.

1. Introduction

Smart Grids (SG)refer to advanced energy networks that incorporate the advances, trends and needs of the 21st century.

The current electric power transmission, distribution and commercialization infrastructure allows SG to add the technological potential of electronics, communication and computing, achieving a bidirectional flow between the equipment installed in the user’s network area and the service providers. Therefore, SG seek to support a more efficient and reliable electric grid, which improves the security and quality of supply, according to the advances of the digital era [1]. Additionally, SG have been conceived in such a way that they can generate a positive environmental impact in the reduction of the “carbon footprint” due to their implementation through global policies and regulations.

SG emerge in their first generation to solve issues related to building more grids, installing automatic meters, developing a workforce oriented towards new communication technologies, and reducing losses and increasing system reliability. The second generation includes the concepts of stability, market design issues and the setting of hourly tariffs.

The third and fourth generations point towards a smart energy grid, addressing the new needs of the world’s sustainable energy system by making full use of new methods for optimization, the penetration of electric mobility with electric vehicles, renewable energies, storage, distributed generation, and distributed, interoperable and secure information systems. These opportunities, which are important for society and new fundamental research challenges, demand a breakthrough in the modernization of SG [2, 3].

The wide range impacts not only monetary aspects are of interest, clearly identify the impact allocation is difficult because indirect/side effects by intangible impacts and quantify all impacts is not possible. Data availability and reliability are necessary for strategic decision-making. One of the main aspects of decision making in planning is identifying the best options. The best option involves the assessment the option performance on several conflicting criteria, making trade-offs, considering the stakeholder perspective and typically achieves a comfortable level of performances by minimizing the related cost. The Smart grid planning calls for effective tools for complex decision- making problems.

Most Smart grid assessment frameworks descend from EPRI approach. Some methods are devised on the specific case study, while others are devised as general frameworks where only qualitative or quantitative criteria are considered [4].

For the majority of the methods for analytical frameworks, large amount of input data and high analyst know-how are main requirements. These make those methods have low replicability due to different context or different assets. Therefore, low comparability of results from different frameworks limited feedbacks from real smart grid projects. The gap between users’ requirements and methods for smart grid assessment is great and the lack of unprofitable projects reassessment grows a gap in how to deal with uncertainty and regulation [4].

The proposed methodology is based on the latency model [5]. In this model, every pilot project is considered a Smart Grid information source, where its data can be analyzed in different layers according to latency and storage. After the modeling of the pilot project, Key Performance Indicators (KPI) are defined for each of the layers described in the reference model used. Once the KPIs have been defined with the characteristics of effectiveness, efficiency, quality and economy, the Analytic Hierarchy Process (AHP) is applied for the multi-criteria evaluation including the cost/benefit analysis to obtain the weights relating the indicators to the criteria. This last phase allows the comparison between the different alternatives for prioritization.

2. Good practices in smart grid project assessment

Several factors at the global level, as well as the emerging technologies needed to establish the criteria and vision of a smart grid, lead electric energy companies to exchange information to ensure the reliability of the operation of interconnected electricity systems [6].

Advances in the integration of SG can be observed in the most important countries and economic groups in the world, among which we can mention:

• European Union: problems related to climate change, a need to generate clean and renewable energy, show greater competitiveness from energy efficiency, interest in issues related to national security viewed from energy independence and automatic demand management and cover the new needs of today’s society (electric vehicles, digital society).

• United States: national security issues viewed from energy independence, economic recovery, modernization of transmission and distribution (T&D) infrastructure, creation of new jobs, reliability, security and the growing need for renewable and distributed generation, treaties on climate change where direct actions are required to address this issue and automatic demand management as an inherent need for SG.

• Japan: treaties and issues associated with climate change, increasing integration of intermittent distributed generation (Distributed Photovoltaic), aging T&D infrastructure, (these factors affect competitiveness given the nation’s high levels of economic growth).

• China: accelerated growth of the electrical infrastructure and the large extension of land which requires long distance transmission (so efficiency in the energy sector is an urgency for this nation).

• India: accelerated growth of electricity infrastructure and the quest to reduce non-technical losses, which are around 40%, are factors that encourage their vision of the implementation of a smart grid as a basis for their development.

However, the infrastructure of electric grids is generally heterogeneous, i.e., there are different formats, technologies and management and storage systems with proprietary and closed formats that hinder interoperability between companies and even internally. The problem of having a large number of data interfaces, multiple processes for exporting and importing information, as well as diverse requirements for transforming the exchanged data, has become exponential. Likewise, several typical problems arise, such as duplicity of information and functions that occur when two or more systems contain the same data or perform the same function; data inconsistency is evident when two systems have different values for the same data; and incompatibility that occurs when information from two or more systems cannot be combined for technological, political, syntactic or semantic causes [7].

In 2011, the IEEE P2030 [8] international standard proposal for Smart Grid Interoperability was published with the objective of providing common understanding, terminology and definitions for the design and implementation of Smart Grid components and applications. P2030 offers three key viewpoints: the energy systems perspective, the communication technology perspective, and the information technology perspective. In addition, each perspective is composed of seven domains: generation, transmission, distribution, service, markets, control/operations and customers. Each domain is composed of a few entities that are logically connected with interfaces. P2030 is promising reference architecture for the standardization of interfaces. However, there is no evidence to determine whether smart grid concepts are appropriate in this approach and the links between the different perspectives are not presented. Therefore, it is not clear whether this approach is feasible for a rigorous analysis with respect to the functionalities envisioned for the smart grid, and their manifestation within the system. In the following subsections, reference models widely used in smart grid representation are presented.

2.1 Smart grid architecture model

The Smart Grid Architecture Model (SGAM) is a reference model for analyzing and visualizing the use of SG in a neutral way [9]. In addition, it supports the comparison between different approaches to smart grid solutions so that differences and similarities between different paradigms, roadmaps, and viewpoints can be identified.

The SGAM provides a systematic approach to deal with the complexity of SG, allowing the representation of the current state of implementations in the power grid, as well as the evolution of future scenarios of SG by supporting the principles of universality, localization, consistency, flexibility and interoperability. The current trends take as a reference model the one proposed by the European community called SGAM, which is shown in Figure 1.

Figure 1.

Model architecture for smart grid [9].

The ease of representation in this architecture allows the SG structure to be extended in one dimension of the complete electrical energy conversion chain, divided into 5 domains: generation, transmission, distribution, distributed energy resources (DER) and local customers and in the other dimension of the hierarchical levels of the power management system, divided into 6 zones: process, field, station, operation, enterprise and market. Interoperability as a key factor for SG is intrinsically addressed in SGAM by the overlapping of the 5 layers: components, communication, information, functions and business. The SGAM layers allow modeling the different business views, which are described below:

In the business layer, economic and political regulatory structures are mapped onto the models related to enterprises, business possibilities and the market players involved. Business processes can also be represented in this layer. Thus, the management group is supported in making decisions related to (new) business models and specific business projects (business case) as well as regulatory agents in defining new market models. The SGAM technical views are modeled in the four lower layers.

The function layer describes the functions and services that the business needs. The functions are represented independently of their physical implementation (represented by elements in the component layer).

The information layer contains the objects and information data models, their usage and the exchange mechanism between functions.

The emphasis of the communication layer is to describe the mechanisms and protocols for the interoperable exchange of information between functions.

The component layer describes all the elements involved. This includes the power system equipment (typically found in process and in the field), the teleoperation protection and control devices, the network infrastructure (wired/wireless communication connections, routers, switches) and any computers. For a specific use case implementation of the identified functions, they can be mapped to components that complement the relationships between all layers [9].

2.2 GridWise architecture council

The GridWise Architecture Council (GWAC) emerges as a conceptual reference model for the identification of standards and protocols needed to ensure interoperability, IT security and define architectures for systems and subsystems in a Smart Grid [10].

Technical interoperability: it covers physical connections and communications between devices or systems (electrical contacts, USB ports).

Informational interoperability: it covers the content, semantics and format of data or instruction streams (such as the accepted meaning of human and programming languages). It focuses on what kind of information is exchanged and its meaning.

Organizational interoperability: it covers the relationships between organizations and individuals and their parts of the system, including business relationships (contracts, properties, and market structures) and legal relationships (regulations, requirements, protection of physical and intellectual property). It emphasizes pragmatic aspects (context, regulations, laws), especially management and the electricity market.

2.3 Smart grid compass

The central objective of the Smart Grid Compass is to redefine the approach to Smart Grid planning that ensures successful technology deployment and maximizes operational resource efficiency. The framework created by the compass is based on an assessment of the key challenges of Smart Grid planning and the associated causes of failure in implementation [11]. Due to the complexity of the environment and market change, utilities face a myriad of business planning challenges. The approach to building and expanding a smart grid provides a 360o view across the entire core service domains:

• Smart grid operation.

• Intelligent customer service.

• Intelligent asset and human talent management.

• Smart energy.

• Intelligent organization, which acts as a driver and “controller” of the changes needed in the other domains. The use of these smart grid domains ensures consistency in language and collaboration across the organization.

2.4 Smart grid maturity model

The Smart Grid Maturity Model (SGMM) proposed by the Software Engineering Institute of Carnegie Mellon University [12], is a management tool that allows planning functions, a quantifiable measurement of evolution and a prioritization of strategies on the way to the implementation of SG.

The SGMM has the following characteristics: it provides a framework for analyzing and addressing modernization needs with a systemic and integrative approach, and with a balance between domains involving processes, people and technology.

This model uses domains and levels to assess and set aspirations for achieving smart grid maturity. The 8 domains of the SGMM represent logical groups of smart grid capabilities and characteristics:

• Strategy, Management and Regulation.

• Organizational Structure.

• Grid operation.

• Personnel and asset management.

• Technological Infrastructure.

• Customer.

• Business Value Chain.

• Society and environment.

In the SGMM there are 5 levels, which represent the stages that evaluate the maturity in each domain (see, Table 1).

Maturity level	Name	Maturity features
5	Leading	The organization drives new business models and moves towards the latest trends.
4	Optimizing	Smart grid implementations are defined to increase organizational performance.
3	Integrating	In the organization, the theme of Intelligent Networks is being integrated into its organizational structure..
2	Setting up	The organization is making progress in the implementation of functionalities, which will allow it to achieve and maintain modernization.
1	Getting started	The organization is in the process of implementation.
0	Basic	Level defined by default at the start of the study.

Table 1.

Maturity levels and features.

2.5 Methodology for the representation of a smart grid project on a SGAM reference model

Given the diversity and complexity of SG, each project must be described in detail to be represented on the SGAM reference model.

Figure 2 shows the roles and latencies for storage in the databases for each zone such that they model a smart grid source [5]. The roles of each zone or storage outlet are described as follows:

• Protection and control: this group stores and delivers indicators related to the protection and control systems of the smart grid system. The indicators of this block are related to the component layer of the SGAM reference model.

• Operation: the indicators obtained in this layer allow to evaluate the correct operation of the network. The indicators in this block are represented in the components layer of the SGAM reference model.

• Engineering analysis: in this block are the network design specifications (emissions reduction, installation size) and the indicators to define maintenance and/or network expansion actions. The indicators enable the evaluation of the communications layer in terms of the protocols used and cybersecurity issues. This block also evaluates the information layer. Since data integrity can be analyzed in the business model and the goods and services offered by the system.

• Consumption analysis: in this group the important indicators from the user’s perspective are related. The functions layer is evaluated with the indicators of this block.

• Business intelligence analysis: the indicators are focused on the management of physical resources, new business and human talent. Regulatory policies and business objectives are related to the business layer of the SGAM reference model and are evaluated with the indicators generated in this block.

Figure 2.

Roles, data storage framework [5].

2.5.1 Procedure for the representation of a Smart Grid source in the proposed model

The representation of a Smart Grid source requires following a basic procedure for modeling in the SGAM architecture. The procedure consists of four steps and can be repeated iteratively to refine the result. Each iteration can cover a different aspect of the class of systems:

Step 1. Smart Grid source selection, this step involves the analysis of available sources and the acquisition of design documentation. This step is crucial for obtaining the model and applying it to the reference architecture. The result of this step is a set of design data and specifications.

Step 2. Identification, creation or adoption of indicators for each stage defined in the reference model, the purpose of this step is to find through different key performance indicators those with the lowest level of abstraction in the representation in the reference model. The use of existing terms (derived, for example, from standardized glossaries) enables generalization in the implementation. The result of this step is a set of reference indicators.

Step 3. Modeling the reference level, in this step each KPI is projected onto the layer level and its respective latency. This translation of the KPIs to the layer level allows a better comparison of the individual KPIs, as well as accessibility to the relevant information of each layer. The result of this step is the relation of the KPIs with the layer level and their latency.

Step 4. Validation of the reference architecture, the creation of reference indicators and their assignment to the layers of the model may not be valid. Therefore, the task in this step is to present the resulting model to Smart Grid system experts, operations, commercial, financial and engineering managers from the power sector, and Smart Grid users to validate whether the Smart Grid source is represented correctly. The result of this step is a validated reference architecture.

3. Multi-criteria analysis for planning assessments

Decision making can be considered as a cognitive process resulting from the selection of a belief or a course of action among several possible options. Every day, all people are faced with different alternatives from which they must select and identify the one that seems to be the best alternative or the one that satisfies the greatest number of intended needs. It is common to find circumstances that lead to make decisions that are relevant in a specific context and the fact of facing the choice of one alternative over another, generates several sensations to the decision maker [13]. It is, therefore, an emotional reasoning or process that can be rational or irrational.

A decision can be considered good or safe, if it comes from an appropriate methodology, considering all related aspects. On the other hand, it is not possible to consider a decision as good if it has not provided an optimal result, or the source and the procedure in its adoption are unknown. The process used to decide becomes important now of choosing the best alternative, since in this way it is possible to support that the solution was the best possible within the options and resources available. The three main characteristics for making a good decision can be found in [14]:

• The objective to be achieved has been outlined.

• All relevant information has been gathered.

• The preferences of the decision maker have been considered.

In engineering projects, decision making is a daily activity. Therefore, the project leader must be clear about what will be the best decision so that the project can thrive and have the least number of inconveniences. It is common that during the development of engineering projects, complex decisions are made and that these have direct consequences on the stakeholders and affected by the decision-making process. Therefore, before making any decision, knowledge, facts, and experience must be gathered and evaluated in the context of the problem. The decision-making process usually relies on the experience of the decision maker or on the similarity to decisions previously made that led to good results.

3.1 Multi-criteria analysis

Multi-criteria analysis (MCA) is a type of decision analysis tool that is applicable to cases where mutually conflicting criteria must be assessed, tangible and intangible impacts must be evaluated simultaneously and allows qualitative assessments such as environmental and social impacts to which quantifiable values cannot be assigned [15]. Multicriteria analysis can help individuals to make decisions in complex situations, where the problem can be addressed from different points of view and the interests can be social, political, environmental, technical and financial. This methodology provides support for decision making as it helps to focus on what is most important, it is logical and consistent and easy to use. At its core, multi-criteria decision-making analysis is useful for:

Breaking the decision into smaller, more understandable parts.

Analyze each part of the problem.

Integrate the parts to generate a comprehensive solution.

The above, supported by mathematical, analytical, research and experimental foundations of management sciences [16]. The application of this type of techniques has been developed since the 50’s of the last century, where the main objective has been to help managers and leaders to make complex decisions. There are several techniques for multi-criteria decision making, among which the Scoring method, the multi-attribute utility, the Analytic Hierarchy Process (AHP), the Analytic Network Process (ANP), among others, stand out. The most widely used for solving problems related to the choice of technologies is the AHP. The AHP method is widely used to choose among certain technological options which would be the best, considering the characteristics of certain areas with their respective particularities.

In [17, 18] the method is used for the prioritization of microgrid generation plans considering resource uncertainties and efficient energy dispatch in smart microgrids. Finally, in [19, 20] AHP is used to obtain the best data information process of an energy metering system and the selection of a smart metering infrastructure according to the needs of an energy meter developer company.

3.1.1 Analytical hierarchical process

The Analytical Hierarchical Process (AHP) is a theory of measurement through pairwise comparison and subsequently organized by expert judgments to obtain priority scales. Developed by Thomas L. Saaty between 1971 and 1975 [21], it has been extensively studied and refined since then. This technique is applied in a wide variety of situations associated with the fields of public administration, industry, business, health, and education. To perform an AHP analysis, those who are involved require a thorough knowledge of the issue to be solved because the construction of the hierarchical structure must include enough relevant details to fully describe the problem. After being clear about the main objective of the problem, the first thing to do is to decompose it in a top-down hierarchy, position it at the top vertex and from there, place the criteria first, to make the selection of alternatives, the constituent parts of the problem, the sub-criteria and their fundamental relationships, as shown in Figure 3.

Figure 3.

AHP hierarchy. Goal, criteria and alternatives.

When the hierarchy is established, the decision-makers (panel of experts) methodically evaluate each of the elements to compare them with each other; these comparisons are made at each hierarchical level in pairs, as they seek to determine the importance of each of them to the higher-level element to which it is related. When making the comparisons, the experts can use concrete (quantifiable) data on the elements that are necessary, or they can use their own judgments according to their level of relevance. It is fundamental to the AHP method that judgments are used to make the assessments [22]. The comparisons assessments made by means of pairs are evaluated by preference indices if alternatives are compared, or importance indices if criteria are compared, which are subsequently evaluated according to a numerical scale proposed by Saaty, the scales for direct assignment are given in Table 2.

Verbal judgment	Saaty’s ratio scale (wj/wk)
Absolute preference for element wk	1/9
Demonstrate preference for element wk	1/7
Strong preference for element wk	1/5
Weak preference for element wk	1/3
Equal preference	1
Weak preference for element wj	3
Strong preference for element wj	5
Demonstrate preference for element wj	7
Absolute preference for element wj	9

Table 2.

Saaty’s score judgment.

The intermediate integer values (2, 4, 6, 8) can be used to express a preference between two adjacent judgments. Humans could establish relationships between objects or ideas so that they are consistent. For this reason, it is important to review the logical consistency of the resulting matrix, to verify whether a contradiction is generated between the values stipulated to the criteria, as a result of the pairwise comparisons. The number of required pairwise comparisons for AHP increases as the number of the criteria and/or of the alternatives increase by performing the process of paired comparison between criteria and alternatives, this leads to a relative scale of measurement of the priorities given to the problem. The AHP converts these evaluations into numerical values that must add up to the unit, giving them a respective weight, in order to be able to compare them with each other in a rational and consistent way. This is how the AHP distinguishes itself from other decision-making techniques. In the final part of the evaluation, numerical priorities are calculated for each of the decision alternatives. The numerical values obtained (see Table 3) represent which of the alternatives has a higher weight to achieve all the criteria of the main objective of the problem [23].

Alternatives	${\mathit{KPI}}_1$	.	.	.	${\mathit{KPI}}_n$	Weighting
$A_1$	w(1,1)	.			w(1,n)	w_1
.	.		.		.	.
.	.			.	.	.
.	.				.	.
$A_m$	w(m,1)				w(m,n)	$w_m$
	${\sum }_{k=0}^m{w}_{\left(k,1\right)}=b_1$				${\sum }_{k=0}^m{w}_{\left(k,n\right)}=b_n$	${\sum }_{k=0}^m{w}_k=1$

Table 3.

Decision matrix.

3.2 Multicriteria analysis & cost/benefit analysis for smart grid project

Strengths of a combined evaluation approach in smart grid project is to select profit each one. The cba is reliable tool for an economical/financial evaluation of tangible impacts, shows some fundamental shortcomings when a large share of intangible impacts is involved. The AHP allows for considering multiple heterogenous, even conflicting criteria, soft effects are directly evaluable and monetization for all impact is not required.

The transformation of traditional power grids to SG demands significant investment in technological infrastructure, certainly, the ability to effectively monitor and manage these technologies will determine the performance of SG and will be critical to the success of those involved in the energy sector. Specifically, in the United States and Europe, regulatory and governmental bodies have for some years defined tools to measure how “smart” current power infrastructures are [24, 25]. In the rest of the world, utilities and government agencies are beginning to work on quantifying and implementing them in the context of each region. These tools help to make important decisions at the organizational level, by allowing to have a clear vision of an implementation towards the future through the results obtained, and to execute in a reliable way, in order to increase the satisfaction of the company itself and its users.

SG are considered as a model that seeks to optimize energy supply, helping to improve efficiency, reliability factors and availability and security from its generation to its delivery to consumers [26].

Some authors are working into integrate MCA and CBA in the evaluation smart grit project, Celli et al. 2017, present a sequential MCA-CBA funded by the Italian Regulator to define the condition for remunerating DSOs which own and operates storage for network issues. Many plans involving storage devices are devised by using a multi-objective optimization approach. Then, the economic sustainability of the alternatives pertaining to the Pareto frons is assessed by a CBA [27].

Another job presented in 2016 for Marnay et al. [4], an MCA is used for evaluation the TEC smart grid demonstration project which is divided in three sub-projects: distributed automation, microgrid, and smart substation. Four different evaluation domains are considered: technological, economic, social, and practical. An index is assigned to each subproject according to the performances on each domain, an overall score is computed by using the proposed SG-MCA method which combines AHP and fuzzy evaluation method [4].

The upgrading plan of the Italian smart metering infrastructure is evaluated by means of MC-CBA approach. Three different areas of interest are investigated: economic, enhanced smartness of the grid, and externalities. Three different MCA techniques to investigate the effects on the provided result [28].

It will show the construction of the approach in smart grid project assessment based on MCA and CBA methods.

3.2.1 Key performance indicators

Key performance indicators (KPI) are tools that provides information on the measurement (management or results) in the delivery of products (goods or services) generated by an institution, covering quantitative or qualitative aspects [29, 30]. Indicators are measurable factors that facilitate decision making.

KPIs are the final phase of strategic planning, which implies an adequate evaluation, selection and definition in the context of SG. In the construction of the KPIs for SG projects, a review was made of the most widely used indicators and those that generate value and impact for the pilot projects evaluated were included [24].

Table 4 describes a KPI which is related to the impact of the alternatives on the quality of the electrical power supply service. The electrical power supply service.

KPI- Duration interruptions per customer, including climate related disruptions
Description	This criterion assesses the extent to which the project option contributes to reducing the duration of the interruptions. The system can improve the fault location and restoration procedures.
Index	System Average Interruption Duration index [occ/yr] (SAIDI)
Quantitative appraisal	The quantitative appraisal of the KPI is possible by comparing the values of the SAIDI before and after the project development.; KPI=$\frac{{\sum }_{\mathit{i}\mathit{=}\mathbf{1}}^{\mathit{n}}{\mathit{U}}_{\mathit{i}}\mathit{N}{\mathit{C}}_{\mathit{i}}}{{\sum }_{\mathit{i}\mathit{=}\mathbf{1}}^{\mathit{n}}\mathit{N}{\mathit{C}}_{\mathit{i}}}$
Qualitative appraisal	A qualitative appraisal can be indirectly made by estimation the level of the monitoring that can be achieved based on the monitored system features.

Table 4.

KPI description table.

3.2.2 Normalization for quantitative KPI

The normalization of the KPIs is used to obtain the weights when they are quantitative, since for qualitative KPIs the weights are obtained by means of paired comparison. The objective of normalization is to ensure that the sum of the KPIs for each alternative is 1. The ideal value for each KPI can be for its maximum or minimum value, taking into account this characteristic, it is necessary to select between Eq. (1) or (2).

$${\mathit{KPI}}_{\mathit{\text{Maximization}}}=\frac{{\mathit{KPI}}_i-\mathit{min}}{\mathit{max}-\mathit{min}}$$E1

$${\mathit{KPI}}_{\mathit{\text{Minimization}}}=\frac{\mathit{max}-{\mathit{KPI}}_i}{\mathit{max}-\mathit{min}}$$E2

where min and max are the lowest and the highest values of the KPI for each criterion, respectively.

3.2.3 Cost/benefit analysis

Cost/benefit analysis (CBA) is one the most acknowledged tool for assessing the financial viability of industrial projects. It aims to an optimal resource allocation in which the monetary benefic outclass cost, and for the most profitable investment alternative. It also provides an incremental analysis regarding a particular scenario and produces easy to read economic indicators. The economic performance indicators are the indexes obtained from a CBA:

• the Net present value (NPV) criterion measures the project profitability in terms of the net benefit. In general, an investment option is economically viable if NPV is positive. The profitability of the investment increases as the related NPV grows. It is a quantitative criterion measured in terms of currency.

• The Internal rate of return (IRR) criterion measures the quality of the investment option. An alternative is positively evaluated if its IRR is higher than the reference social discount rate. It is a quantitative criterion measured in percentage terms.

• The Cost Benefit ratio (CBR) criterion measures the efficiency of the investment option. An alternative is positively evaluated if its CBR is greater than one. It is a quantitative dimensionless criterion.

• Investment payback period (Pt) is the period of balancing net profits and all construction expenses. It is a significant index for reflecting the project’s investment return capacity. Payback period can be broken down into dynamic and static. To simplify the calculation, the static investment payback period is adopted for the calculations for the smart grid project. Those criteria are fulfilled according to the increasing values of the related indices (see Table 5).

Index	Calculation method	Evaluation Standard
NPV	$\mathit{NPV}={\sum }_{t=1}^n{\left(\mathit{CI}-\mathit{CO}\right)}_t{\left(1+i_c\right)}^{-t}$; CI: cash inflow; CO: cash outflow; n: calculation period; Ic: Specified discount rate (benchmark yield)	if NPV is ≥0, the project achieves its expected benefits.
IRR	$\mathit{IRR}={\sum }_{t=1}^n{\left(\mathit{CI}-\mathit{CO}\right)}_t{\left(1+\mathit{IRR}\right)}^{-t}=0$; CI: cash inflow; CO: cash outflow; n: (CI-CO)t; net cash flow of the t-th year; n: calculation period; IRR: internal rate of return.	if IRR is > benchmark yield, the project achieves its expected benefits.
Pt	Pt = (The year in which the accumulated net cash flow becomes a positive value −1) + |accumulated net cash flow previous year/the net cash flow of the current year|	if Pt is > benchmark yield, the project achieves its expected benefits.
BCR	BCR = The proposed total cash benefit/the proposed total cash cost. Prior to dividing the numbers, the NPV of the respective cash flows over the proposed lifetime of the project – considering the terminal values, including salvage/remediation costs – are calculated.	if a project has a BCR >1.0, the project is expected to deliver a positive NPV to a firm and its investors.; if a project’s BCR ≤ 1.0, the project’s costs outweigh the benefits, and it should not be considered.

Table 5.

Evaluation economics KPI [4].

Electric utilities invest large sums in dedicated utility equipment to review compliance with their regulatory or statutory obligations. For example, the benefits of extending service to new regions and planning for continued growth are generally accepted and implicit in mandatory regulations. These companies routinely meet these non-discretionary obligations and minimize their execution costs. Moreover, they are often well prepared to defend their decisions within this cost-minimization framework. Smart Grid projects, on the other hand, may not fit into this time-tested paradigm of cost minimization because they may be discretionary. For example, the decision to invest in a Smart Grid project to improve reliability beyond currently acceptable levels depends on how much to invest to obtain the improvement, and whether the improvement gained justifies the amount of money to be invested. This goes far beyond mere regulation, maintaining the stringent nature imposed by the regulation itself.

Many Smart Grid investments require going beyond cost minimization. In addition to their novelty, Smart Grid applications offer new benefits beyond basic or least-cost service. They can improve reliability and quality of service beyond currently accepted levels, in addition to providing customers with options and services never before experienced. Consequently, they are discretionary for the utility, and a feasible/positive scenario is needed to incorporate such innovations into the regulated business. Eventually, Smart Grid technologies are the only realistic alternatives to address technical issues that may arise when services such as distributed generation or electric vehicle charging become commonplace on distribution systems. However, these technical issues are mostly in the future, so today it remains the responsibility to devise a business and economically positive scenario, showing sufficient benefits to offset the costs [31].

The CBA is a methodology proposed by the Electric Power Research Institute [31], it aims to determine whether the benefits of a project or decision outweigh its costs. However, CBA analyzes costs and benefits from a particular point of view, which can range from the broad and societally impactful (public perspective) to the particular and focused (private perspective). General economic analyses adopt a social perspective, determining whether a project is a good allocation of social resources, without considering the distribution of benefits. This contrasts with financial analysis, as performed in private companies, which generally focuses on investment returns. This tool allows for a mid-point analysis, as the focus is on the costs incurred by the company, which are borne by the customers. The planning analysis of regulated companies minimizes the cost of reliable service while assuming the return on investment. When minimizing the cost of service is inconsistent with public policy goals, legislators and regulators can impose conditions designed to address those goals, in the hope of stimulating decisions that benefit society.

The methodological approach of the EPRI-generated guidance sets out a CBA methodology that is compatible with either the societal or customer approaches to weighing costs and benefits. This concept is more comfortably suited to fully integrated companies, as costs and benefits are easily aligned, and all are contained within a corporate environment (except for externalities that fall outside the electricity sector). Costs in one part of an enterprise can be offset by savings in another part of the same enterprise, minimizing or even eliminating the need for additional cost recovery. In addition, users are recognized as a variety of utility entity types, many of which participate in some of the functions of a generation, operation, transmission and/or distribution utility. Costs incurred within one entity may produce offsetting savings in a separate corporate entity. Although consumers may be indifferent to where costs and savings occur, the various corporate entities involved face varying levels of cost recovery risk depending on their regulatory situations and their position in the cost and savings chain. The latter is important from a private enterprise perspective. Figure 4 shows the steps to its implementation [32].

Figure 4.

Steps in describing the technology.

3.2.4 Sensitivity analysis

Factors of high uncertainty among costs and benefits are analyzed, the impact on the benefits of the project is analyzed quantitatively, and sensitive factors are identified to enable control and avoidance of risks. Multi-factor sensitivity analysis is carried out as required. The risks of not realizing the project’s benefits are determined according to the sensitivity analysis.

The AHP sensitivity analysis for each KPI(1..n) or sub-criterion is obtained from the decision matrix. Each alternative for the sensitivity analysis is calculated with the values of the lines obtained from the relative weights of each KPI and the overall score of the decision matrix, Eqs (3)–(5) is used for calculus. The values to calculate the KPI line are:

$$\left(x_2,y_2\right)=\left(100\%,\frac{w\left(i,j\right)}{b_j}\right)$$E3

$$\left(x_1,y_1\right)=\left(b_j,w_i\right)$$E4

$$S_i\left(X\right)=w_i+\frac{\left(X-b_j\right)\left(\frac{w\left(i,j\right)}{b_j-w_i}\right)}{w_i-b_j}$$E5

where $b_j$ is sum of the relative weigh of j-th KPI, $w\left(i,j\right)$ is weight of j-th KPI and i-th alternative, and $w_i$ is overall score for i-th alternative, $S_i$ is value for i-th alternative, X is percentage for KPI value (see Table 6).

${\mathit{KPI}}_i$	$A_i$	.	.	.	$A_m$	Sensitivity
0%	$S_i$(0%)	.			$S_m$ (0%)	$b_j$
.			.			.
.				.		.
.						.
100%	w(i,j)/$b_j$				w(m,j)/$b_j$	$b_j$
$b_j$	$w_i$	.	.	.	$w_m$	$b_j$

Table 6.

Sensitivity matrix for ${\mathit{KPI}}_{\mathit{i}}$.

4. Case study: distribution grid planning of a median voltage rural grid

The case study consists in distribution grid planning of the medium voltage rural grid with five possible alternatives (see Figure 5).

Figure 5.

Distribution grid in rural area. * Increased local economic development.

4.1 Smart grid deployment reference model for a rural distribution grid

Table 7 presents the KPIs for a rural distribution grid project. It uses the reference model based on roles, which has 5 layers that are associated with the interoperability layers of the SGAM model.

Criteria	Description
Protection and control	It fulfills the function of storing and delivering indicators related to the protection and control systems of the smart grid system. The KPIs of this block are related to the component layer of the SGAM reference model.
Subcriteria	Description
E1 [p.u] Min	Voltage variation index [p.u.] =${\sum }_{l=1}^n{\sum }_{h=1}^{\mathit{Nh}}\left|V_{\mathit{max},l}^{\left(h\right)}-V_{\mathit{min},l}^{\left(h\right)}\right|$
Criteria	Description
Operation Process	The indicators obtained with this block make it possible to evaluate the correct functioning of the network. The indicators of this block are represented in the component layer of the SGAM reference model.
Sub Criteria	Description
D1 [MW] Max	Active power available for black start =${\sum }_{j=1}^{\mathit{\text{NDES}}}{\sum }_{k=1}^{\mathit{Nh}}\mathit{min}\left(\mathit{\text{SoCn}},l\ast \mathit{\text{ndis}},l\ast \mathit{ph},l\right)$
D2 [MW] Max	Maximum use in power of dispatchable resources =${\sum }_{l=1}^{\mathit{\text{NDES}}}\frac{P_{\mathit{DES},l}^{\left(\mathit{\text{out}}\right)}+\left|p_{\mathit{DES},l}^{\left(h\right)}\right|}{2}$
Criteria	Description
Engineering Analysis	This block is fed by the network design specifications (reduction of the carbon footprint or environmental impact) and its own indicators to define maintenance and/or network expansion actions. The indicators make it possible to evaluate the communications layer in terms of the protocols used and cybersecurity-related issues. This block also evaluates the information layer since data integrity can be used to analyze the business model and the goods and services offered by the system.
Sub Criteria	Description
C1 [occ/yr] Min	System Average Interruption Duration Index –SAIDI = $\frac{{\sum }_{i=1}^n{U}_{i}N{C}_i}{{\sum }_{i=1}^{n}N{C}_i}$
C2 [hr/yr] Min	System Average Interruption Frequency Index -SAIFI=$\frac{{\sum }_{i=1}^n{\lambda }_{i}N{C}_i}{{\sum }_{i=1}^{n}N{C}_i}$
Criteria	Description
Consumption analysis	The important indicators from the user's perspective are listed in this block. The function layer is evaluated with the indicators in this block.
Sub Criteria	Description
B1 [%] Max	Reliability =$\frac{\#\mathit{\text{annual operating hours}}}{8760}$
B2 [MWh] Min	Expected network energy losses [Mwh] =${\sum }_{j=1}^{\mathit{Ne}}{\sum }_{k=1}^{\mathit{Nh}}{E}_{lj,k}$
Criteria	Description
Business intelligence	The indicators are focused on the management of physical resources, new business and human talent. Issues related to regulatory policies and business objectives are related to the business layer of the SGAM reference model and are evaluated with the indicators generated in this block.
Sub Criteria	Description
A1 [K$] Min	Investments and reactive power exchange net value – NPV = Cost (transmission investment) + Cost (DES investment) + Q (power exchange net value)
A2 [Ton/yr] Max	CO2 reduction =$\mathit{EG}\ast \mathit{EFG}$
A3	Increased local economic development

Table 7.

Evaluation criteria for a rural distribution grid.

n: number of busses; Nb: number of time intervals; Ndes: number of Des, Ne: number of network elements; $\mathit{EG}$: Total energy generated by PVs [kWh]; $\mathit{EFG}$: Emissions Factor by Generation.

The A3 sub-criterion is properly a qualitative KPI, because it is an aspect associated with the increase of the local economy, the other KPIs can be calculated from the operation records and projections of the five alternatives to be considered.

4.2 Multicriteria analysis for rural distribution grid projects

The multi-criteria analysis for rural distribution grid used is AHP where qualitative and quantitative data can combine. In the case of quantitative data is required to maximize or minimize the KPI and calculate the corresponding weighting. For qualitative data is need a paired comparison using the Saaty’s scale. Figure 6 shows AHP’s structure for rural distribution grid planning.

Figure 6.

AHP methodology. Goal, criteria, subcriteria and alternatives for rural distribution grid projects.

Table 8 presents the performance matrix of each KPI in relation to the alternatives.

4.2.1 Increased local economic development

The A3 is the sub-criteria associated to Local development that significantly contributes to national economic performance and has become more critical with increased global competition, population mobility, technological advances, and consequential spatial differences and imbalances. Effective local development can reduce disparities between poor and rich places, add to the stock of locally generated jobs and firms, increase overall private sector investment, improve the information flows with investors and developers, and increase the coherence and confidence with which local economic strategy is pursued [33]. This can also give rise to better diagnostic assessment of local economic assets and distinctive advantages, and lead to more robust strategy assessment. This indicator is evaluated using Saaty’s scale through pairwise comparison as shown in Table 9.

Criteria	Sub-criteria	Weights
Business intelligence; (56.1%)	A1	73.6%
	A2	5.3%
	A3	21.1%
Consumption analysis (18.3%)	B1	20.0%
	B2	80.0%
Engineering Analysis; (12.9%)	C1	25.0%
	C2	75.0%
Operation Process; (7.5%)	D1	33.3%
	D2	66.7%
Protection and control; (5.2%)	E1	100.0%

Table 8.

Corresponding weights for each layer and KPIs.

C2	Alt_1	Alt_2	Alt_3	Alt_4	Alt_5	Weighting
Alt_1	1	1/5	1/7	1/3	1/3	0,057
Alt_2	5	1	3	1	1	0,299
Alt_3	7	1/3	1	1	1	0,217
Alt_4	3	1	1	1	1	0,213
Alt_5	3	1	1	1	1	0,213

Table 9.

Performance matrix for increase local economic development.

The AHP requires that the weightings of the criteria, sub-criteria and alternatives be calculated. After making these assessments, the decision matrix is obtained, providing the result of prioritization among the alternatives, as shown in Table 10.

Alternatives.	Business intelligence			Consumption analysis		Engineering analysis		Operation process		Protection & control	Rank
	A1	A2	A3	B1	B2	C1	C2	D1	D2	E2
Alt_1	0.109	0.003	0.007	0.004	0.000	0.005	0.000	0.000	0.000	0.000	0.128
Alt_2	0.104	0.009	0.035	0.021	0.045	0.000	0.029	0.011	0.005	0.011	0.272
Alt_3	0.094	0.000	0.026	0.000	0.010	0.008	0.033	0.001	0.034	0.016	0.222
Alt_4	0.105	0.006	0.025	0.012	0.047	0.014	0.005	0.007	0.000	0.013	0.234
Alt_5	0.000	0.012	0.025	0.000	0.045	0.005	0.029	0.005	0.010	0.011	0.144

Table 10.

Decision matrix.

Figure 7 shows the sensitivity for A1(41%) and A2(3%) subcriteria, these are investment and reactive power exchange net value and CO₂ reduction, the sensitivity analysis is calculated with decision matrix.

Figure 7.

Sensitivity analysis: a) NPV criterion; b) CO₂ reduction criterion.

5. Conclusions

The proposed tool aims to simplify decision making processes, so a pilot project into SGAM reference model is represented. This tool adds one more step for identifying weaknesses and opportunities, realizing for sensitivity analysis to decision-making stability assessment. This implementation includes the tangibles and intangibles impacts and data collection and allows planning smart grid based on applications of an assessment framework.