Fossil fuels, including coal, petroleum, natural gas, and nuclear energy, are the primary electricity sources currently. However, with depletion of fossil fuels, global warming, nuclear crisis, and increasing environmental consciousness, the demand for renewable energy resources has skyrocketed. Solar energy is one of the most popular renewable energy resources for meeting global energy demands. Even though there are abundant studies on various solar technology developments, there is a lack of studies on solar technology evaluation and selection. Therefore, this research develops a model using interpretive structural modeling (ISM), benefits, opportunities, costs, and risks concept (BOCR), and fuzzy analytic network process (FANP) to aggregate experts' opinions in evaluating current available solar cell technology. A case study in a photovoltaics (PV) firm is used to examine the practicality of the proposed model in selecting the most suitable technology for the firm in manufacturing new products.
Energy is an essential element for civilization development of mankind and quality improvement of life. The use of fossil fuels and other natural resources has resulted in detrimental impacts on the environment, especially through the damage to the air, climate, water, land, and wildlife [
Photovoltaic (PV) solar cells are semiconductor devices that transfer solar radiation into electricity by converting the energy of the sunlight to direct current electricity after photovoltaic effect [
Regardless of the effort in making proper cost-benefit solar cells by using new materials and new technologies, the predominant wafer-based silicon technology, or so-called first-generation solar cell [
With the huge growth expectations of thin film solar cell technologies, the competing market price of crystalline-silicon has slowed the development of thin film solar cell. The latter is expected to grow anyway at a lower rate because of its potential reduction of production costs, low material consumption, lower energy consumption, and a shorter energy payback time [
Advanced production technologies can help reduce production cost, improve product quality, and increase yield rate, and these technologies can be related to process engineering, system integration, production automation, and process equipment [
Even though there are abundant studies on the development of various solar technologies, there are very few studies on solar technology evaluation and selection. To survive in the global competition, firms need continuously to develop new products these days. Before a new technology is introduced, a firm needs to consider and evaluate available technologies first and then select the most suitable one in an efficient way. It is a multidisciplinary process to make the decision in choosing the most appropriate technology to fabricate products since this process needs to integrate different professional knowledge, production management, and market trend. Therefore, the decision should not only consider the expected benefits a technology can bring in making the final products with specified quality, but also cover other aspects, such as opportunities, costs, and risks. As a result, the technology selection is a sophisticated evaluation process which must consider multiple attributes.
Technology selection is not a new research topic; however, very little research has examined the interrelationship of the criteria in the decision making process and considered the positive and negative aspects of the alternatives simultaneously. Thus, this paper, based on the model proposed by Lee et al. [
Many approaches on technology evaluation have been presented in previous studies. Internal rate of return (IRR), net present value (NPV), return on investment (ROI), and payback period (PB) have been traditionally applied to evaluate technology alternatives from financial viewpoint [
Technology evaluation problem is a multicriteria decision making (MCDM) problem, and it should involve the opinions of multiple experts and consider fuzzy assessments [
To summarize, even though technology evaluation is a MCDM problem in nature, relatively few models have been proposed. This study applies and revises the model proposed by Lee et al. [
The solar industry in Taiwan has a great potential in the global market due to the technical advantages gained from the semiconductor industry and the TFT-LCD industry [
A model, based on the model proposed by Saaty [
Form a committee of experts in a solar firm to define the solar cell technology evaluation problem.
Develop a control hierarchy for the solar cell technology evaluation problem. The control hierarchy has several strategic criteria and four merits, benefits
The control hierarchy for solar cell technology selection.
Develop a network with BOCR subnetworks. Based on literature review and interview with the experts, the problem can be decomposed into a network, as depicted in Figure
The BOCR-ANP network for solar cell technology selection.
Develop a relation matrix for the criteria under each merit. Experts are asked to identify whether there is a relation between any two criteria and what the direction of the relation is. Then, a relation matrix
Construct an initial reachability matrix for each merit. By adding
Calculate the final reachability matrix
Develop a subnetwork for the criteria under each merit based on
Prepare a questionnaire based on Figures
Based on the control hierarchy, two kinds of information are collected through the questionnaire. First, experts are asked to pairwise compare the importance of the strategic criteria with seven linguistic terms, as shown in Figure
Fuzzy numbers for relative importance/performance.
Fuzzy numbers for ranking.
Calculate the importance of the strategic criteria. Aggregate experts’ responses by employing geometric average approach, and a synthetic trapezoid fuzzy number is calculated as follows:
Defuzzify fuzzy number
|
|
|
---|---|---|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Check the consistency of the aggregated comparison matrix. The consistency index (CI) and consistency ratio (CR) are [
The experts are asked to revise the part of the questionnaire if there is an inconsistency, and the calculations in Step
Calculate the importance of each merit
Calculate the priorities of the merits. The priority of a merit is obtained by multiplying the importance of the merit on each strategic criterion from Step
Calculate the relevant priorities on the BOCR-ANP network in Figure
Prepare an unweighted supermatrix for each merit using the priorities obtained from Step
Unweighted supermatrix for merit
Obtain the weighted supermatrix for each merit subnetwork. One approach is to determine the relative importance of the clusters in the supermatrix with the column cluster (block) as the controlling component [
Obtain the priorities of technology alternatives under each merit subnetwork. Calculate the limit supermatrix for each merit subnetwork by raising the weighted supermatrix to powers. The priorities of the technology alternatives under each merit are found in the alternative-to-goal column of the limit supermatrix of the merit.
Obtain overall priorities of the technology alternatives by synthesizing the priorities of each alternative under each merit from Step
Perform sensitivity analysis to examine the robustness of the outcomes. By changing the priorities of the merits, Steps
The proposed model is applied to an anonymous solar cell manufacturer in Taiwan to select the most suitable solar cell technology. A comprehensive literature review is done first, and some experts in the solar cell technology field are interviewed. A control hierarchy and an ANP-BOCR network are developed by the authors and verified by the experts, as shown in Figures
The control hierarchy for solar cell technology selection.
The BOCR-ANP network for solar cell technology selection.
The interrelationship among the criteria under the same upper-level merit is determined through the ISM. Experts’ consensus is obtained through the Delphi method, and a relation matrix under each merit is prepared. The relation matrix among the criteria under benefits,
Using Step
After the calculation, the final reachability matrix
Based on
Subnetwork for the criteria under the benefits merit.
Subnetwork for the criteria under the opportunities merit.
Subnetwork for the criteria under the costs merit.
Subnetwork for the criteria under the risks merit.
Based on the network in Figure
The fuzzy aggregated pairwise comparison matrix for the strategic criteria is
A defuzzified comparison matrix is prepared by the Yager [
Then, the priority vector, that is,
By applying Step
Because CR is less than 0.1, the consistency test is passed. In the opinions of the experts, manufacturing capability
Next, the importance of each merit to each strategic criterion is determined. A seven-level linguistic scale is used to collect the experts’ opinions in the questionnaire, and each linguistic scale is assigned a trapezoid fuzzy number. The outcomes from the experts are aggregated using the geometric average method, and the fuzzy numbers are defuzzified by the Yager [
Integrated fuzzy weights of the merits on strategic criteria.
Manufacturing capability ( |
Market demand ( |
Financial performance ( |
Social responsibility ( |
|
---|---|---|---|---|
Benefits | (0.715, 0.828, 0.865, 0.883) | (0.684, 0.795, 0.831, 0.865) | (0.682, 0.761, 0.0832, 0.866) | (0.704, 0.805, 0.0854, 0.877) |
Opportunities | (0.625, 0.733, 0.766, 0.831) | (0.382, 0.481, 0.526, 0.621) | (0.278, 0.402, 0.408, 0.530) | (0.526, 0.609, 0.676, 0.755) |
Costs | (0.695, 0.817, 0.842, 0.871) | (0.535, 0.654, 0.684, 0.759) | (0.595, 0.701, 0.744, 0.805) | (0.673, 0.773, 0.820, 0.860) |
Risks | (0.392, 0.521, 0.536, 0.636) | (0.627, 0.716, 0.777, 0.822) | (0.828, 0.865, 0.883, 0.763) | (0.300, 0.425, 0.432, 0.555) |
Priorities of the merits
|
|
|
|
Overall priorities | Normalized priorities | |
---|---|---|---|---|---|---|
(0.381) | (0.358) | (0.165) | (0.096) | |||
Benefits | 0.823 | 0.794 | 0.785 | 0.81 | 0.805 | 0.295 |
Opportunities | 0.739 | 0.502 | 0.405 | 0.642 | 0.590 | 0.216 |
Costs | 0.806 | 0.658 | 0.663 | 0.782 | 0.727 | 0.266 |
Risks | 0.520 | 0.735 | 0.65 | 0.428 | 0.610 | 0.223 |
Based on the opinions of the experts collected from the questionnaire, we can further calculate the priorities of the criteria with respect to each merit, the interrelationship among the criteria under each merit, and the expected relative performance of the technologies under each criterion. That is, the collected data are synthesized into aggregated pairwise comparison matrices using the geometric average method first, and the Yager [
Unweighted supermatrix for the benefits merit.
G |
|
|
|
|
|
|
|
|
|
|
---|---|---|---|---|---|---|---|---|---|---|
G | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
|
||||||||||
|
0.179 | 0.634 | 0.347 | 0.161 | 0.186 | 0 | 0 | 0 | 0 | 0 |
|
0.218 | 0.366 | 0.653 | 0.173 | 0.213 | 0 | 0 | 0 | 0 | 0 |
|
0.274 | 0 | 0 | 0.367 | 0.274 | 0 | 0 | 0 | 0 | 0 |
|
0.329 | 0 | 0 | 0.299 | 0.327 | 0 | 0 | 0 | 0 | 0 |
|
||||||||||
|
0 | 0.252 | 0.252 | 0.286 | 0.301 | 1 | 0 | 0 | 0 | 0 |
|
0 | 0.231 | 0.231 | 0.294 | 0.277 | 0 | 1 | 0 | 0 | 0 |
|
0 | 0.205 | 0.205 | 0.190 | 0.172 | 0 | 0 | 1 | 0 | 0 |
|
0 | 0.201 | 0.201 | 0.148 | 0.156 | 0 | 0 | 0 | 1 | 0 |
|
0 | 0.111 | 0.111 | 0.082 | 0.094 | 0 | 0 | 0 | 0 | 1 |
Weighted supermatrix for the benefits merit.
G |
|
|
|
|
|
|
|
|
|
|
---|---|---|---|---|---|---|---|---|---|---|
G | 0.5 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
|
||||||||||
|
0.089 | 0.317 | 0.173 | 0.081 | 0.093 | 0 | 0 | 0 | 0 | 0 |
|
0.109 | 0.183 | 0.327 | 0.086 | 0.107 | 0 | 0 | 0 | 0 | 0 |
|
0.137 | 0 | 0 | 0.183 | 0.137 | 0 | 0 | 0 | 0 | 0 |
|
0.165 | 0 | 0 | 0.150 | 0.163 | 0 | 0 | 0 | 0 | 0 |
|
||||||||||
|
0 | 0.126 | 0.126 | 0.143 | 0.151 | 1 | 0 | 0 | 0 | 0 |
|
0 | 0.115 | 0.115 | 0.147 | 0.138 | 0 | 1 | 0 | 0 | 0 |
|
0 | 0.103 | 0.103 | 0.095 | 0.086 | 0 | 0 | 1 | 0 | 0 |
|
0 | 0.101 | 0.101 | 0.074 | 0.078 | 0 | 0 | 0 | 1 | 0 |
|
0 | 0.056 | 0.056 | 0.041 | 0.047 | 0 | 0 | 0 | 0 | 1 |
Limit supermatrix for the benefits merit.
G |
|
|
|
|
|
|
|
|
|
|
---|---|---|---|---|---|---|---|---|---|---|
G | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
|
||||||||||
|
0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
|
0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
|
0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
|
0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
|
||||||||||
|
0.243 | 0.232 | 0.179 | 0.267 | 0.272 | 1 | 0 | 0 | 0 | 0 |
|
0.256 | 0.232 | 0.233 | 0.277 | 0.266 | 0 | 1 | 0 | 0 | 0 |
|
0.199 | 0.208 | 0.217 | 0.194 | 0.185 | 0 | 0 | 1 | 0 | 0 |
|
0.184 | 0.205 | 0.214 | 0.165 | 0.170 | 0 | 0 | 0 | 1 | 0 |
|
0.118 | 0.123 | 0.157 | 0.098 | 0.106 | 0 | 0 | 0 | 0 | 1 |
The relative performances of solar cell technology alternatives under each merit are shown in Table
Performance of alternatives under the four merits.
Merits | Benefits | Opportunities | Costs | Risks | ||||
|
||||||||
Priorities | 0.295 | 0.216 | 0.266 | 0.223 | ||||
|
||||||||
Alternatives | Normalized | Normalized | Normalized | Reciprocal | Normalized | Normalized | Reciprocal | Normalized |
Reciprocal | Reciprocal | |||||||
|
||||||||
Dye sensitized solar cell (DSSC) ( |
0.243 | 0.236 | 0.173 | 5.773 | 0.221 | 0.235 | 4.250 | 0.159 |
Crystalline silicon cell ( |
0.256 | 0.193 | 0.169 | 5.902 | 0.226 | 0.118 | 8.447 | 0.317 |
Concentrating cell (GaAs) ( |
0.199 | 0.180 | 0.159 | 6.304 | 0.242 | 0.206 | 4.843 | 0.182 |
Thin film cell ( |
0.184 | 0.193 | 0.226 | 4.420 | 0.170 | 0.205 | 4.881 | 0.183 |
Organic cell ( |
0.118 | 0.199 | 0.273 | 3.670 | 0.141 | 0.235 | 4.256 | 0.160 |
By aggregating the scores of each alternative under
Final priorities of alternatives.
Methods | Additive | Probabilistic additive | Subtractive | Multiplicative priority powers | Multiplicative | |||||
|
||||||||||
Alternatives | Priorities | Ranking | Priorities | Ranking | Priorities | Ranking | Priorities | Ranking | Priorities | Ranking |
|
||||||||||
Dye sensitized solar cell (DSSC) ( |
0.2171 | 2 | 0.5133 | 2 | 0.0240 | 2 | 0.2144 | 2 | 1.4070 | 2 |
Crystalline silicon cell ( |
0.2479 | 1 | 0.5348 | 1 | 0.0455 | 1 | 0.2443 | 1 | 2.4583 | 1 |
Concentrating cell (GaAs) ( |
0.2023 | 3 | 0.4985 | 3 | 0.0091 | 3 | 0.2009 | 3 | 1.0921 | 3 |
Thin film cell ( |
0.1819 | 4 | 0.4794 | 4 |
|
4 | 0.1817 | 4 | 0.7666 | 4 |
Organic cell ( |
0.1507 | 5 | 0.4421 | 5 |
|
5 | 0.1480 | 5 | 0.3661 | 5 |
No matter which aggregation method is used, crystalline silicon cell
A sensitivity analysis is performed next to examine the robustness of the outcomes. The software Super Decisions [
Sensitivity analysis when applying the additive method.
The management needs to understand the importance of the criteria when making the solar cell technology selection decision, and such information can be obtained from the calculation results. The importance of the criteria under each of the four merits is shown in Table
Importance of criteria.
Merits | Criteria | Criterion priorities | Integrated priorities in the network | Integrated ranking |
---|---|---|---|---|
Benefits |
( |
0.179 | 0.0528 | 8 |
( |
0.218 | 0.0643 | 6 | |
( |
0.274 | 0.0808 | 3 | |
( |
0.329 | 0.0971 | 1 | |
|
||||
Opportunities |
( |
0.127 | 0.0274 | 16 |
( |
0.125 | 0.0270 | 17 | |
( |
0.112 | 0.0242 | 18 | |
( |
0.214 | 0.0462 | 13 | |
( |
0.225 | 0.0486 | 10 | |
( |
0.197 | 0.0426 | 14 | |
|
||||
Costs |
( |
0.141 | 0.0375 | 15 |
( |
0.242 | 0.0644 | 5 | |
( |
0.328 | 0.0872 | 2 | |
( |
0.290 | 0.0771 | 4 | |
|
||||
Risks |
( |
0.086 | 0.0192 | 19 |
( |
0.270 | 0.0602 | 7 | |
( |
0.211 | 0.0471 | 12 | |
( |
0.222 | 0.0495 | 9 | |
( |
0.212 | 0.0473 | 11 |
In the case study, the expected performance of alternatives with respect to each criterion can be learned. The performance results are found in the (3,2) block of the limit supermatrix. For instance, the relative performances of
Performance of alternatives with respect to each criterion.
Criteria | Alternatives | ||||
---|---|---|---|---|---|
|
|
|
|
|
|
( |
|
|
0.208 | 0.205 | 0.123 |
( |
0.179 |
|
0.217 | 0.214 | 0.157 |
( |
0.267 |
|
0.194 | 0.165 | 0.098 |
( |
|
0.266 | 0.185 | 0.170 | 0.106 |
( |
|
0.153 | 0.169 | 0.179 | 0.183 |
( |
|
0.148 | 0.155 | 0.179 | 0.194 |
( |
|
0.153 | 0.171 | 0.190 | 0.184 |
( |
0.177 |
|
0.200 | 0.200 | 0.206 |
( |
0.192 |
|
0.182 | 0.198 | 0.208 |
( |
0.203 |
|
0.184 | 0.199 | 0.202 |
( |
0.159 | 0.221 | 0.140 |
|
0.251 |
( |
0.151 | 0.169 | 0.136 |
|
0.301 |
( |
0.205 | 0.144 | 0.192 | 0.218 |
|
( |
0.163 | 0.174 | 0.150 | 0.220 |
|
( |
|
0.114 | 0.213 | 0.204 | 0.225 |
( |
|
0.107 | 0.223 | 0.206 | 0.225 |
( |
|
0.128 | 0.192 | 0.212 | 0.225 |
( |
0.224 | 0.133 | 0.172 | 0.198 |
|
In conclusion, the criteria of solar technology can be logically identified, organized, reviewed, and concluded by this model to avoid the rank bias and halo effect [
With natural resource scarcity and environmental protection, renewable energy sources have been recognized as the last resort for future economic development, and solar energy is one of the most promising renewable energy sources from the perspective of environmental sustainability. However, the PV market is facing a rather volatile market cycle in response to the global economic condition. A PV firm must have a solid foundation in its technology in order to survive and to lead the market in the future. Therefore, good evaluation and selection of the most appropriate technology become a complicated decision that a firm often encounters.
This research constructed an integrated model, which incorporates interpretive structural modeling (ISM), benefits, opportunities, costs, and risks concept (BOCR), and fuzzy analytic network process (FANP), for facilitating the evaluation of technologies. The model consists of three phases. In the first phase, a control hierarchy and a BOCR-ANP network are constructed. In the second phase, the relevant priority weights in the control hierarchy and the BOCR-ANP network are calculated. In the last phase, the technology alternatives are ranked. The proposed model is implemented in a solar cell firm to help select the most suitable solar cell technologies.
This evaluation model is constructed under known technology of solar cell for technology selection for mass production. By applying the proposed model, experts can understand the expected performance of technology alternatives based on different criteria and merits. The overall ranking of the technologies can be calculated as a result. Further studies can be conducted for the conceptualized or developing stage of solar cell technology to facilitate decision making on new technology development to ensure the success of new product introduction. In addition, based on the special needs of a firm, the proposed model can be adjusted as required by the firm in the PV industry or in another industry to help select the most suitable technology.
The authors declare that there is no conflict of interests regarding the publication of this paper.