Abstract
Decentralized energy planning (DEP) is looked upon as an indisputable opportunity for energy planning of villages, isolated islands, and far spots. Nonetheless, at this decentralized planning level, the value of demand-side resources is not fairly examined, despite enjoying great advantages. Therefore, the core task of this study is to integrate demand-side resources, as a competing solution against supply-side alternatives, with decentralized energy planning decisions and demonstrate the rewarding role it plays. Moreover, sustainability indicators (SIs) are incorporated into DEP attempts in order to attain sustainable development. It is emphasized that unless these indicators are considered at lower energy planning levels, they will be ignored at higher planning levels as well. Hence, to the best knowledge of the authors, this study for the first time takes into account greenhouse gas (GHG) emissions produced by utilization of renewable energies in DEP optimization models. To address the issues mentioned previously, multiobjective linear programming model along with a min-max goal programming approach is employed. Finally, using data taken from the literature, the model is solved, and the obtained results are discussed. The results show that DSM policies have remarkably contributed to significant improvements especially in terms of environmental indicators.
1. Introduction
Decentralized energy planning aims for efficient use of local resources to supply energy. DEP is an available option for satisfying energy needs of villages, isolated islands, and small-scale end users in a reliable, affordable, and environmentally sustainable way (Hiremath et al. [1]). To see some recent top researches conducted in the related area, the works of [2–7] are recommended. The DEP paramount feature is to provide an area-based DEP so that with the least economic and environmental costs develops alternate energy sources and satisfies energy requirements. References [8–11] are good example works done in the literature with focus on DEP modeling.
Energy demand management, also referred to as demand-side management (DSM), includes activities that modify end use energy demand; for example, the activities aim to reduce peak demand. It is worth mentioning that peak demand management does not necessarily lead to reduction in total energy consumption, but it defers the need for investments in establishing new power plants. The most generally accepted definition of DSM is the one presented by Gellings in 1984 [12].
As discussed by Kreith and Goswami [12], DSM alternatives must be compared to supply-side alternatives. Thus, it entails considering and evaluating both supply-side and demand-side options in energy planning efforts. In the late 1980s, Ontario Hydro of Canada estimated that satisfying its customer peak demand via supply-side measures costs four times that of demand-side measures [13]. As a result, any energy planning model without simultaneously considering both sides is incomplete.
Although developing a model which considers a DEP with evaluating both supply-side and demand-side management measures is of great importance, it is not yet inclusive. The shortcoming is related to not taking SIs into account. To overcome this, we need to have such indicators quantified, which is of course a very challenging task [14] (see also [15, 16]). Evans et al. [17] ranked four renewable energy technologies, namely, photovoltaic, wind, hydro and geothermal, against several SIs using data extracted from the literature. These indicators include electricity generation cost, GHG emissions, availability and limitations of each technology, efficiency of energy transformation, land use requirements, water consumption, and social impacts. Based on their findings, wind technology outranked the others followed by hydro, photovoltaic, and geothermal technologies, respectively. Nonetheless, it should be pointed out that the ranking was provided for the global international conditions, whereas each technology can be considerably geographically affected.
To cover the mentioned gaps in the literature, in this paper a multiobjective linear programming formulation for a DEP problem is proposed that not only takes both supply-side and demand-side alternatives into account but also considers a number of important SIs against which five energy resources, including DSM options, are evaluated. It is worth mentioning that using multi-objective formulations for energy planning is commonplace in the literature; see, for example, [10, 18–21].
In a nutshell, the contributions of this paper can be summarized as follows.(i)It is, to the best knowledge of the authors, for the first time that in DEP models supply-side and demand-side options are treated equally.(ii)It is, to the best knowledge of the authors, for the first time that in DEP models GHG emissions produced by renewable energy technologies during their life cycle are incorporated. It may be surprising that, for instance, for photovoltaic, the amount of produced GHG emissions is almost one quarter of natural gas emissions [22], while it had been a widespread supposition that GHG emissions of renewable energy technologies are insignificant.(iii)A variety of important sustainable indicators, namely water consumption, land use, cost of generation, employment, and GHG emissions, are considered.The rest of the paper is organized as follows. In the next section, materials and methods used are elaborated. In Section 3, results and discussions are provided. Finally, conclusion remarks are drawn in Section 4.
2. Materials and Methods
2.1. Materials
The data used in this paper are collected from the literature and are based on the global international conditions. To solve the model for a numerical example, Lingo 8.0 software was employed.
2.2. Mathematical Formulation
As stated before, five SIs are taken into account. These SIs are regarded as distinct objective functions. The notations used in this model along with the proposed integrated multi-objective formulation are presented below.
Notations. Sets are as follows: set of renewable energy technologies;: set of end uses.Parameters and variables as follows: : water consumption of th resource option (kg/kWh); : greenhouse gas emissions of the th resource option (g CO(2-e)/kWh); : land use characteristic of th resource option (m2/kWh); : electricity generation cost of resource for th end use (US$/kWh); : electricity saving cost using the implemented DSM program in the th end use (US$/kWh); : employment generation of th resource option (jobs/kWh); : total energy demand for th end use (kWh); : conversion efficiency for the th resource option for th end use; : availability of the th resource option for th end use (kWh); : the maximum possible saving using the DSM program in the th end use (kWh); : the required saving that is, to be achieved by the DSM program (kWh).Decision variables: : optimal amount of th resource option for th end use (kWh); : optimal saving in th end use using the implemented DSM program (kWh).
The objective function given in (1) minimizes the total electricity generation costs. As can be seen, costs associated with implementing the DSM program are also considered in the first objective function, and it is treated just like other energy resources. It means that saving a unit of electricity by implementing DSM can be treated like producing a unit of electricity by a power plant, provided that the data for the associated costs are available. Equation (2) maximizes total number of jobs created due to electrifying the area. It should be noted that since the data for the number of jobs created by implementing the DSM program were not found, assumed to be negligible in this paper. The objective function given in (3), (4), and (5) minimize total water consumption, greenhouse gas emissions, and total land used, respectively. Constraints (6) and (7) are demand and resource availability constraints, respectively. Constraint (8) denotes that the maximum expected saving that could be achieved by executing the DSM program is limited to a realistic upper bound. Constraint (9) forces energy planners to implement DSM programs and meet some predetermined energy efficiency goals [23]. Constraint (10) indicates that decision variables are nonnegative real numbers.
2.3. Methods
Before we can solve the model by Lingo 8.0 software, we must find a way to cope with the multi-objective nature of the proposed formulation. To do so, a min-max goal programming approach was applied to obtain the best mix of renewable energy sources. For further study about goal programming (GP) approaches, the respected reader is referred to [24]. Now, using GP approach the model presented in the previous subsection can be easily reformulated as a single-objective linear programming formulation as presented in the following:
As shown above, parameter denotes maximum deviation from every target . The parameter values and are set by decision maker. In some sense, this method is similar to the weighted Tchebycheff approach except that here the optimum is replaced by target value.
3. Results and Discussion
In this section, in order to demonstrate the applicability of the model to real case problems, a numerical example based on the data obtained from the literature is developed. It should be pointed out that for the unavailable data, assumed values are considered. Tables 1 and 2 show the parameter values of the model that are either taken from [10–12, 17] or assumed by the authors.
The values of parameter in Table 2 are taken from [12]. For example, it is stated that a successful DSM program (2002 Massachusetts Electric-Residential Lighting Program) cost was million while it resulted in 18037 MWh energy saving. Therefore, the unit cost of energy saving using the implemented DSM program in the domestic sector is calculated as follows:
The way of calculating the other values are the same. In this example, a rural area is assumed that is partly electrified, but the most parts are not yet provided with electricity. An energy saving target ranging from 5 to 15 percent in the already electrified parts of the area is considered.
To solve the goal programming model proposed in Section 2.3, all and values are set to be 0.5. The objective functions are also considered as presented below.
The first objective function:
The second objective function:
The third objective function:
The fourth objective function:
The fifth objective function:
As a decision maker, the following target values are assigned to the five objective functions mentioned above, respectively. The first objective function: 0.16. The second objective function: 0.7. The third objective function: 1.2. The fourth objective function: 0.3. The fifth objective function: 2.7.
Table 3 shows the outputs of Lingo 8.0 software for this numerical example.
Figure 1 shows the degree to which renewable energy sources have contributed to electrifying the area under study.
Using the solution given in Table 3, one can calculate the value of each individual objective function separately (Case 1). Also, by solving the original model presented in Section 2.2 once for each objective function (as a single-objective model) independently, the optimum value of each individual objective function is obtained (Case 2). Comparing the objective function values obtained in Case 1 with those obtained in Case 2 the following conclusions can be drawn.(1)Goal programming results are more practical to implement.(2)If the main limitations are water and land use requirements, minimizing GHG emissions and generation costs or maximizing employment will not be an attractive goal to attain.(3)If financial resources are scarce, minimizing water consumption and land used will not be an attractive goal to attain.(4)If maximizing employment is of the highest importance, similar to the previous, minimizing water consumption and land used will not be an attractive goal to attain.
From the results pointed out above, it can be summarized that provided that target values are determined carefully, the model could be useful under various geographical conditions.
4. Conclusions
In this study, an integrated multi-objective linear programming formulation was developed for decentralized energy planning such that demand-side management and environmental measures were also taken into account. Moreover, a number of important sustainability indicators were utilized upon which four renewable energy technologies were compared. Using goal programming approach, the proposed model was solved by Lingo 8.0 software. The findings indicate that fine choices of sustainability indicators for doing the decision-making process are extremely affected by the limitations the policy makers encounter. For example, poor countries suffer from scarce funding resources. Thus, when planning for their energy supplies, it is not so sensible to consider minimizing the land occupied by the supply technologies as an important criterion to pursue. Also, the obtained numerical results show that in the optimum solution hydro outranked the other renewable energy technologies followed by wind, photovoltaic, DSM resources, and geothermal, respectively. It is worth pointing out that DSM alternatives have greatly contributed together with the other energy supply technologies.
Conflict of Interests
The authors declare that they have no conflict of interests.