18.1 Introduction
18.2 MCDA for Sustainability Assessment
18.2.1 Stakeholder Integration
18.2.2 Sustainability Criteria and Indicators
18.2.3 Selection of MCDA Method
18.2.4 Classification of MCDA Methods
18.2.5 WSM
18.2.6 TOPSIS
18.2.7 PROMETHEE
Preference function | Thresholds | Graphical representation |
---|---|---|
Type I – Usual | None | A graph of P versus d with a horizontal line at the vertical axis on Point, 1.
|
Type II – U-shape | q– Indifference | A graph of P versus d with a horizontal line at (q, 1) and beyond it. The starting point is connected to the points by dashed lines.
|
Type III – V-shape | p– Preference | A graph of P versus d has a line rising from the origin to (p, 1), then remains constant. Point 1 and p from the axes is connected to (1, p) by dashed lines.
|
Type IV – Level | q,p | A graph of P versus d has 2 parallel lines from (1, p) to beyond and (Half, q) to (Half, p). Point 1 and p and point Half and q from the axes are connected to (1, p) and (Half, q) by dashed lines.
|
Type V – V-shape with indifference (linear) | q,p | A graph of P versus d has a line that starts from the origin to (0, q), then rises to (p, 1), and then remains constant. Point 1 and p from the axes is connected to (1, p) by dashed lines.
|
Type VI – Gaussian | S – Gaussian threshold | A graph of P versus d depicts a rising S-shaped curve originating from the origin, passing through point S on the horizontal axis, and reaching its peak at 1 on the vertical axis.
|
18.3 Properties of MCDA Methods for Sustainability Assessment
-
Handling qualitative and quantitative data: when conducting sustainability assessment, different information can be obtained in different forms, i.e., ordinal, cardinal, or mixed.
-
Type of weights: within an MCDA model, there are two types of weights: trade-offs when the weights reflect intensity of preference and importance coefficients which represent voting power [44]. In the case of sustainability assessment, the weights should be modeled as importance coefficients. Therefore, special attention should be paid when selecting the methods for preference elicitation.
-
Partial/null compensation between criteria: compensation implies the existence of trade-offs in the aggregation of criteria, i.e., the extent to which bad performance of one criterion can be offset by good performance of another. Compensation is associated with the concept of weak sustainability and low compensation with strong sustainability (a more detailed description can be found in Ziemba [60]).
-
Threshold values: these can be useful in complex preference models where not all preferences have the same intensity or relevancy.
-
Ease of use: simple structure facilitates the experience of the users. Some methods are commonly preferred because of their simplicity. For example, full compensatory methods such WSM are easier to implement compared to low-compensatory methods that could require high cognitive effort such ELECTRE III or PROMETHEE II. However, it is a task of the analyst to properly understand the methods and be able to explain it to stakeholders.
-
Handling uncertainty: sustainability issues are inherently related to uncertainty. In order to account for this imprecision or vagueness in the information, the multicriteria evaluation needs to either model the uncertainty of the input data, i.e., stochastic analysis, or include sensitivity analysis [42].
-
Software support and graphical representation: several software exist that facilitate the implementation of different MCDA methods. Given their importance on the implementation of MCDA methods, an additional subchapter is dedicated to this topic.
MADM methods | |||
---|---|---|---|
Elementary methods | Single synthesizing criteria | Outranking methods | |
Properties/characteristics for sustainability assessment | WSM | TOPSIS | PROMETHEE II |
Handle qualitative and quantitative data | Quantitative | Quantitative | Quantitative, qualitative |
Weights as importance coefficients | Trade-offs | Trade-offs | Relative importance coefficients |
Threshold values | No | No | Preference, indifference |
Partial/null compensation between criteria | Full | Full | Null, partial |
Handling uncertainty | Yes | Yes | Yes |
Ease of use | High | High | Medium |
Software support and graphical representation |
18.4 MCDA Software
18.4.1 MCDA KIT Tool
18.5 MCDA for Sustainability Assessment in the Field of Batteries
Source | Technologies /alternatives | Application | Criteria | MCDA methods (w = weighting, a = aggregation) | Number and categories of stakeholders involved (participant/category description) | Results |
---|---|---|---|---|---|---|
Haase et al. [24] | Battery electric vehicle (BEV), fuel cell electric vehicle (FCEV), internal combustion engine vehicle (ICEV) | Electric vehicles | Environmental (LCA) Economic (total costs) Social (domestic value added) | TOPSIS (a) | None | Under given assumptions, the BEV with wind power was assessed as most sustainable option in 2020 as well as in 2050 |
Domingues, Marques, Garcia, Freire, and Dias [16] | BEV, FCEV, ICEV, plug-in hybrid (PHEV) | Electric vehicles | Life cycle impact assessment | ELECTRE TRI (a) | None | BEV and PHEV are the only vehicles that can achieve the top class |
Baumann, Peters, and Weil [7] | Lithium–iron–phosphate (LFP), lithium–iron–phosphate/lithium titanate (LFP–LTO), lithium–manganese oxide (LMO), nickel–cobalt–aluminum oxide (NCA), and nickel–cobalt–manganese oxide (NMC) | Utility scale Battery storage technologies: Primary regulation, energy time shifting, wind energy support, decentralized grid | Environmental (damage to ecosystem DE, damage to human health DHH, damage of resources availability DRA) Economic (life cycle costs LCC) Social (socioeconomic values, acceptance, regulation, and frame) Technological (maturity, technology performance, tech. flexibility) | AHP (w), TOPSIS (a) | 72 people/civil society, regulation, policymakers, researchers (university), municipal utility, network operator, utility company, RES production/retail, automotive sector, battery manufacturer, energy storage business | LIBs seem to be the most Recommendable technology among the evaluated BESS for most application areas (with exception of the LIB–LTO, which is only suitable for low E/P ratios) |
Ma et al. [36] | Nas battery, lead-acid (LA) battery, NiMH battery, and Li-ion battery (LIB) | Electrochemical energy storage for renewable energy-based power generation stations | Environmental (CO2 intensity) Economic (capital intensity and operation cost) Social (social acceptance and electric power system reserve capacity reduction) Technological (cycle life, energy efficiency, and self-discharge rate) | Bayesian BMW (w), TOPSIS (a) | Five people/experts | LIB is the optimal solution |
Albawab, Ghenai, Bettayeb, and Janajreh [1] | LA batteries, LIBs, super-capacitors, hydrogen storage, compressed air energy storage, pumped hydro, and thermal energy storage | Not given | Environment (area and material intensities, energy, CO2, and capital intensities of the construction, life cycle greenhouse gas emissions) Economic (operating cost, current installed capacity, growth rate) Social indicators (health and safety issues) Technological: cycle lifetime, cycle efficiency, discharge time at rated power, and adaptability for mobile systems Resource (specific energy, specific power, energy density) | Extended SWARA (w)/ARAS (a) | Three people/experts from engineering, energy storage, sustainability, energy and climate change, renewable energy | Ranking: (1) thermal energy storage, (2) compressed air, (3) LIBs, (4) pumped hydro, (5) LA batteries, (6) hydrogen storage (onboard), and (7) supercapacitors |
Salameh et al. [49] | PV-BES based on nickel–iron (Ni–Fe), LIB, and LA battery technologies at different depths of discharges (DOD) | PV/battery technology microgrid system for a desalination plant | SDG 7: Affordable and clean energy (round-trip efficiency, energy density, number of life cycles, lifetime years, storage duration) SDG 8: Decent work and economic growth SDG 15: Life on land (environmental impact) Economic (annual levelized cost, levelized cost of energy) Size: NPV, NBAT | TOPSIS, WSM, NEW (a) | None | The PV-LIB at 50% DOD was the best option among all cases |
Murrant and Radcliffe [46] | Power to gas, a distributed battery system, battery storage integrated with solar PV and demand from an airport, liquid air energy storage, battery storage integrated with wave energy, and thermal energy storage at a new residential development | Energy storage projects | Environmental co-benefits Economic growth (innovation), economic viability, increasing self-consumption, economic co-benefits Deferral of grid upgrades, technology viability | MAVT (w,a) | 18 people/local and national businesses, academia, community energy groups, and CC | Top-ranking project: Battery storage integrated with solar PV and demand from an airport |
18.6 Use Case MCDA Sustainability Assessment for Early-Stage Cathode Materials for Sodium Ion Batteries
18.6.1 Stakeholder Integration
18.6.2 Problem Definition
18.6.3 Selection of Criteria
Sustainability issues | Criteria | Indicator | Unit | Description | Methods for quantification/source of data |
---|---|---|---|---|---|
Resource management (global supply concentration, country governance, import reliance, trade restriction, recycling) | Raw material criticality (criticality) | Supply risk (SR) for the EU | SREU/kWh | Collective term describing the economic value and dependency on certain materials as well as the probability of supply chain disruptions [50] | SR for Europe [17] |
Global warming, emissions to air and water | Carbon footprint (CF) | GHG emissions | kg CO2eq./Wh | Greenhouse gas emissions of the CAM precursors and their synthesis process | LCA |
Competitiveness | CAM cost (cost) | Costs | €/kWh | Costs of raw materials and precursor materials | Literature and market search inflations and inflation adjusted median values of costs from the last 11 years |
18.6.4 Definition of Alternatives
No. | CAM name | Theoretic capacity (mAh/g) | Reversible capacity (mAh/g) | Reversible specific energy without anode (Wh/kg) | Cost (€/kWh) | Criticality (SREU/kWh) | CF (kg CO2eq.-kWh) |
---|---|---|---|---|---|---|---|
Layered oxide materials | |||||||
1 | LiCoO2 | 274 | 150 | 585 | 39.45 | 2.77 | 72.19 |
2 | LiNi0.8Co0.15Al0.05O2 (NCA) | 279 | 188 | 696 | 20.06 | 0.86 | 27.39 |
3 | LiNi0.5Mn0.5O2 | 280 | 150 | 585 | 15.98 | 0.90 | 16.51 |
4 | LiNi0.33Mn0.33Co0.33O2 (NMC111) | 278 | 160 | 592 | 23.62 | 1.51 | 34.92 |
5 | LiNi0.6Mn0.2Co0.2O2 (NMC622) | 276 | 170 | 629 | 21.25 | 1.12 | 29.40 |
6 | LiMn2O4 (LMO) | 148 | 110 | 472 | 8.01 | 1.29 | 5.04 |
7 | LiNi0.5Mn1.5O4 (LNMO) | 147 | 140 | 644 | 10.33 | 0.67 | 10.62 |
8 | P2-Na0.67CoO2 | 168 | 115 | 369 | 44.06 | 3.76 | 99.05 |
9 | a-NaMnO2 | 244 | 185 | 509 | 1.97 | 0.88 | 2.16 |
10 | ß-NaMnO2 | 244 | 190 | 523 | 1.92 | 1.14 | 2.77 |
11 | Na0.44MnO2 | 122 | 120 | 336 | 3.22 | 1.52 | 2.86 |
12 | P2-Na0.67MnO2 | 175 | 175 | 490 | 2.14 | 0.99 | 2.08 |
13 | P2-Na0.67Mn0.72Mg0.28O2 | 191 | 220 | 572 | 1.74 | 1.16 | 2.23 |
14 | P2-Na0.67Mn0.95 Mg0.05 O2 | 177 | 175 | 455 | 2.29 | 1.13 | 2.33 |
15 | P2-Na0.67Mn0.5 Fe0.5O2 | 174 | 190 | 523 | 1.26 | 0.72 | 1.51 |
16 | O3-NaMn0.5Fe0.5O2 | 243 | 110 | 303 | 2.12 | 1.16 | 2.92 |
17 | P2-Na0.67Ni0.33 Mn0.67O2 | 173 | 161 | 596 | 5.16 | 0.69 | 7.78 |
18 | P2-Na0.8Li0.12Ni0.22 Mn0.66O2 | 214 | 118 | 415 | 7.14 | 0.97 | 9.26 |
19 | P2-Na0.83Li0.07 Ni0.31Mn0.62O2 | 214 | 140 | 490 | 6.60 | 0.80 | 9.40 |
20 | P2-Na0.83Li0.25 Mn0.75 O2 | 237 | 185 | 500 | 4.35 | 0.85 | 3.54 |
21 | O3-NaFe0.5Co0.5O2 | 238 | 160 | 502 | 15.61 | 1.55 | 35.18 |
22 | O3-NaNi0.33Co0.33 Fe0.33O2 | 238 | 165 | 487 | 15.12 | 1.24 | 31.68 |
23 | O3-NaNi0.5Mn0.5O2 | 240 | 125 | 377 | 10.13 | 0.93 | 16.28 |
24 | Na[Mn0.4Fe0.5Ti0.1]O2 | 244 | 110 | 308 | 2.38 | 1.18 | 6.30 |
25 | NaMn0.33Fe0.33Ni0.33O2 | 240 | 100 | 481 | 5.51 | 0.66 | 9.00 |
26 | Na0.6Fe0.11Mn0.66Ni0.22O2 | 159 | 120 | 324 | 7.25 | 1.29 | 10.50 |
27 | NaMn0.3Fe0.4Ni0.3O2 | 241 | 130 | 390 | 6.19 | 0.80 | 10.16 |
28 | P2-Na0.6Fe0.2 Mn0.65Ni0.15O2 | 158 | 200 | 620 | 2.97 | 0.67 | 4.16 |
29 | Na0.6Ni0.22Al0.11Mn0.66O2 | 164 | 225 | 675 | 3.63 | 0.62 | 5.33 |
Polyanionic materials | |||||||
30 | LiFePO4 (LFP) | 170 | 165 | 569 | 5.70 | 0.81 | 5.08 |
31 | Na3V2(PO4)3 | 118 | 110 | 381 | 15.33 | 1.59 | 34.99 |
32 | Na4MnV(PO4)3 | 111 | 110 | 380 a) | 8.01 | 1.30 | 18.79 |
33 | Na3MnTi(PO4)3* | 117 | 114 | 410a | 1.87 | 1.14 | 5.55 |
34 | Na3MnTi(PO4)3** | 176 | 172 | 506b | 1.51 | 0.93 | 4.50 |
35 | Na3MnZr(PO4)3 | 107 | 110 | 402b | 1.41 | 1.12 | 3.23 |
36 | NaFePO4 | 154 | 152 | 410 | 0.57 | 0.87 | 3.06 |
37 | Na1.702Fe3(PO4)3 | 87 | 140 | 406 | 0.55 | 0.93 | 2.95 |
38 | Na4Fe3(PO4)P2O7** | 152 | 129 | 406 | 0.62 | 0.87 | 3.27 |
39 | Na2MnPO4F* | 249 | 178 | 651 | 0.99 | 0.76 | 2.29 |
40 | NaV(PO4)F | 143 | 82 | 303 | 23.42 | 2.52 | 51.79 |
41 | Na1.5VPO4.8F0.7 | 130 | 134 | 509 | 12.69 | 1.30 | 28.27 |
42 | Na2Fe(PO4)F | 124 | 110 | 360 | 0.78 | 1.09 | 3.52 |
43 | Na3MnPO4CO3* | 192 | 125 | 490 | 1.08 | 0.61 | 2.80 |
44 | Na2MnFe(CN)6* | 171 | 140 | 490 | 1.32 | 0.50 | 2.06 |
45 | Na0.61Fe[Fe(CN)6]0.941 | 61 | 170 | 493 | 0.80 | 0.58 | 2.06 |
46 | Na0.81Fe[Fe(CN)6]0.791 | 90 | 149 | 447 | 0.88 | 0.46 | 1.67 |
47 | Na2FeSiO4* | 276 | 181 | 724 | 0.87 | 0.44 | 3.14 |
48 | Na2MnSiO4* | 278 | 210 | 630 | 1.66 | 0.68 | 4.00 |
49 | Na2Fe2(SO4)3* | 120 | 102 | 418 | 0.40 | 0.45 | 0.87 |
18.6.5 Preference Modeling
18.6.6 Weighting
18.6.7 Preference Function and Parameters
Criteria | Cost (€/kWh) | Criticality (SREU/kWh) | CF (kg CO2eq.-kWh) |
---|---|---|---|
Preference function | Linear | Linear | Linear |
Maximum value | 44.06 | 3.76 | 99.05 |
Minimum value | 0.40 | 0.44 | 0.87 |
Q | 0 | 0 | 0 |
P | 0.89 | 0.07 | 2 |
18.6.8 Results
18.6.9 Comparison and Evaluation of Alternatives (Ranking)
18.6.10 Sensitivity Analysis
Equal weights (original case) | Sensitivity analysis | |||||
---|---|---|---|---|---|---|
50% costs | 50% criticality | 50% CF | ||||
Rank | Net flow | CAM No. | CAM name | CAM No. | ||
1 | 0.921 | 49 | Na2Fe2(SO4)3 | 49 | 49 | 49 |
2 | 0.818 | 46 | Na0.81Fe[Fe(CN)6]0.79a | 46 | 46 | 46 |
3 | 0.757 | 45 | Na0.61Fe[Fe(CN)6]0.94a | 45 | 47 | 45 |
4 | 0.713 | 47 | Na2FeSiO4 | 47 | 45 | 44 |
5 | 0.710 | 44 | Na2MnFe(CN)6 | 44 | 44 | 15 |
6 | 0.647 | 43 | Na3MnPO4CO3 | 43 | 43 | 47 |
7 | 0.638 | 15 | P2-Na0.67Mn0.5Fe0.5O2 | 15 | 15 | 43 |
8 | 0.593 | 39 | Na2MnPO4F | 39 | 39 | 39 |
9 | 0.492 | 36 | NaFePO4 | 36 | 48 | 36 |
10 | 0.473 | 38 | Na4Fe3(PO4)P2O7 | 38 | 36 | 38 |
11 | 0.443 | 37 | Na1.702Fe3(PO4)3 | 37 | 38 | 37 |
12 | 0.399 | 48 | Na2MnSiO4 | 42 | 28 | 9 |
13 | 0.365 | 9 | a-NaMnO2 | 48 | 29 | 12 |
14 | 0.294 | 42 | Na2Fe(PO4)F | 9 | 37 | 48 |
15 | 0.276 | 28 | P2-Na0.6Fe0.2 Mn0.65Ni0.15O2 | 35 | 9 | 13 |