Water utilities performance assessment Multicriteria methodologies contributions
The International Conference on Adapting to Climate Change
Water, waste and other local infrastructure
Sandra Tralhão Rita Martins
and João Paulo Costa
1
Agenda
1. Introduction
2. Methodological approach
2.1. Regulation and performance evaluation
2.2. Multiple Criteria Decision Aid Methodologies
3. Environmental sustainability of water supply services bulk operators in Portugal
3.1. Identification and characterization of indicators
3.2. Application and discussion of results
3.3. Sensitivity analysis
4. Conclusions
2
1. Introduction
The idea of unlimited water resources meant that for centuries' water was considered an inexhaustible resource. However:
 Significant increase in population
 Consumer habits changes
 Climate changes  Occurrence of more frequent extreme climatic events
 Significant increases in demand
 Changing drainage
 Lowering of groundwater levels and contraction of lakes  reduction of water resources availability
Irregularities of freshwater supply and decrease of the water quality 3
1. Introduction
The goal of this work is to assess the environmental
sustainability of water supply services bulk operators
in Portugal.
The basis for the present research is the performance
evaluation model adopted by ERSAR.
However, our analysis allows to adjust the importance
given to each criterion in a specific management period
according to the defined sustainability goals.
4
2. Methodological approach
• The methodological approach is supported by benchmarking tools
and indicators, used by many organizations at the international
level, and in Multiple Criteria Decision Aid Methodologies.
‐ Multiattribute Value Theory (MAVT) The problematics that fall within the decision‐making context Sorting
Ranking
‐ Outranking method ‐ ELECTRE TRI Multiattribute Value Theory (MAVT)
5
2.1. Regulation and performance evaluation
The need to regulate the quality of water supply services is consensual and
benchmarking tools are used by the most known regulators at the
international level.
6
2.2. Multiple Criteria Decision Aid Methodologies
Multiattribute Value Theory (MAVT) is a compensatory method of additive aggregation that allows:
‐ To associate an overall score to an alternative
‐ To build the complete ranking of the alternatives ‐ That a poor performance on a criterion may be compensated by high performance in another criterion
∑
is the overall value of alternative is the weight of criterion j
reflects the performance of alternative for the criterion j
7
2.2. Multiple Criteria Decision Aid Methodologies
•
The outranking ELECTRE TRI method is a multicriteria‐based outranking sorting
method which allows assigning alternatives into a set of predetermined
categories.
Category 1
Category 2
…
…
Category n
…
Source: Adapted Mousseau et al. (1999)
•
In the construction of the outrank relations the parameters of the criteria
defining the importance coefficients (wj), the performance of each alternative in
a given criterion and the preference (pj), indifference (qj) and veto (vj) thresholds
8
are used.
3. Environmental sustainability of water supply services bulk operators in Portugal
Bulk water supply
Abstraction
•
The designation of the water
supply sector in Portugal
depending on the type of
activities
carried
out by
operators:
‐ Bulk water supply
‐ Retail water supply
•
The specific economic regulator
in Portugal is the Water and
Waste
Services
Regulation
Authority ( ERSAR).
9
Treatment
Elevation
Adduction
Storage
Retail water supply
Distribution
Evaluation sheets – operator’s performance is exposed indicator to indicator.
Indicators of environmental sustainability.
10
3.1. Identification and characterization of the indicators
•
The real water loss (RWL) indicator is used to assess the level of
environmental sustainability of the efficiency in use of environmental
resources. Is measured by the water losses by total length of pipes .
RWL
Losses m3/year
Totallengthofpipes km ∗ 365
Good service
quality
[0,0; 5,0]
Acceptable service
quality
]5,0; 7,5]
Unsatisfactory
service quality
]7,5; +∞[
11
3.1. Identification and characterization of the indicators
•
The fulfillment of water abstraction licensing (FWAL) indicator allows the
evaluation of the environmental sustainability of service, in terms of the
safety of water abstractions and the environmental protection.
FWAL
Good service
quality
100
/
/
Acceptable service
quality
]90,0; 100[
∗ 100
Unsatisfactory
service quality
[0,0; 90,0[
12
3.1. Identification and characterization of the indicators
•
•
The standardised energy consumption (SEC) indicator intends to assess
the level of efficiency in the use of environmental resources, taking into
account the appropriate use of energy resources.
It is defined as the average power consumption of normalized pumping
facilities.
SEC
Good service
quality
[0,27; 0,40]
Acceptable service
quality
]0,40; 0,54]
/
∗
/
Unsatisfactory
service quality
]0,54; +∞[
13
3.2. Application and discussion of results
Standardised
energy
consumption
0,49
0,4
0,38
0,43
0,5
0,37
0,39
0,3
0,49
0,47
0,4
0,37
-0,38
0,69
Alternatives
AdSAndré
AdTMAD
AdAlgarve
AdCentro
AdCAlentejo
AdDPaiva
AdMondego
AdNoroeste
AdNAlentejano
AdOeste
AdVouga
AdZCoa
AdPAlentejo
EPAL
ICOVI
8,42
1,53
4,73
1,22
3,05
11,27
10,64
1,25
2,94
4,26
1,36
2,77
4,4
34,88
4,52
Real water losses
Fulfilment of the water
abstraction licensing
Standardised energy
consumption
b2
5,00
100,00
0,40
b1
7,50
90,00
0,54
Performance matrix based on the evaluation sheets of the quality of service published by ERSAR for the year 2013. Standard profiles
Fulfillment of water
abstraction
licensing
54,1
98,04
60,42
29,37
83,01
94,39
95,73
98,3
94,39
0
50,21
30,51
35,36
100
0
Real water
losses
14
3.2. Application and discussion of results ‐ MAVT
Global value
• Weigths ( withthe
sameimportance
adjusted by scale factors
• Global profiles adjusted by scale factors:
b1= 0,770495827
b2 = 0,657142857
0,5877
0,7929
0,6852
0,6144
0,6966
0,7194
0,7183
0,8433
0,7299
0,4995
0,6781
0,6332
0,3325
0,5372
0,3929
Good service Acceptable Unsatisfactory
quality
service quality service quality
0
1
0
0
0
0
0
1
0
0
0
0
0
0
0
0
1
1
0
1
1
1
1
1
0
1
0
0
0
0
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
AdSAndré
AdTMAD
AdAlgarve
AdCentro
AdCAlentejo
AdDPaiva
AdMondego
AdNoroeste
AdNAlentejano
AdOeste
AdVouga
AdZCoa
AdPAlentejo
EPAL
ICOVI
15
3.2. Application and discussion of results ‐ MAVT
•
From the application of the MAVT method results the ranking of the
environmental sustainability alternatives.
•
The highest overall value function obtained is 0,8433 ("AdNoroeste“) and
the second is 0,7929 ("AdTMAD“)
•
Most operators get unsatisfactory performance.
16
3.2. Application and discussion of results – ELECTRE TRI
Good service
quality
•Threshold ‐ Indifference ( )=0
‐ Preference ( )=0 ={10,55, ∞}
‐ Veto •Weigths
( )={0,33;0,33;0,34}
• Standard profiles
• Cutting level =0,5 Acceptable Unsatisfactory
service quality service quality
0
1
1
0
0
0
0
1
0
0
0
0
0
0
0
Good service quality
0
1
1
0
1
1
1
1
1
0
1
0
0
0
0
Acceptable service quality
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
AdSAndré
AdTMAD
AdAlgarve
AdCentro
AdCAlentejo
AdDPaiva
AdMondego
AdNoroeste
AdNAlentejano
AdOeste
AdVouga
AdZCoa
AdPAlentejo
EPAL
ICOVI
17
Unsatisfactory service quality
3.2. Application and discussion of results – ELECTRE TRI
• In this case the results affect the same operators ("AdTMAD“ and
"AdNoroeste“ of the MAVT) and "AdAlgarve" to the good
environmental sustainability performance category .
• About 46.6% of the operators are allocated to performance category
"unsatisfactory" and 33.3% to the "acceptable".
• For  > = 0,57 the classification gets significantly worse and allocates
80% of operators to the "unsatisfactory” performance category (in this
scenario are obtained similar results from the application of MAVT).
18
3.3. Sensitivity analysis
•
With the two methods use the results reveal themselves more favorable if a
higher weight is assigned to water loss.
•
With the two methods the worst case of environmental sustainability is the
one giving a higher weight to the water abstraction licensing criterion.
•
increasing the cut off level decreases environmental sustainability obtained
with ELECTRE TRI method (Cutting level > =0,69)
•
The scenarios obtained by changing the weights assigned to the criteria are
more favorable when the water loss and the energy efficiency criteria are
assigned a weight greater than or equal to 0.5.
19
3.3. Sensitivity analysis
Good service
quality
For cutting level >= 0,69 in ELECTRE TRI
Acceptable
service quality
Unsatisfactory
service quality
Operators
0
0
1
0
1
1
0
0
1
0
0
1
0
0
1
0
0
1
0
0
1
0
1
1
0
1
1
0
0
1
0
0
1
0
0
1
0
0
1
0
0
1
0
0
1
AdSAndré
AdTMAD
AdAlgarve
AdCentro
AdCAlentejo
AdDPaiva
AdMondego
AdNoroeste
AdNAlentejano
AdOeste
AdVouga
AdZCoa
AdPAlentejo
EPAL
ICOVI
20
3.3. Sensitivity analysis
Good service
quality
The most favorable scenario, in ELECTRE TRI is the one where it is attributed to the water loss indicator the weight the 0.9.
Acceptable
service quality
Unsatisfactory
service quality
Operators
0
0
1
1
1
1
1
1
1
0
1
1
1
1
1
0
0
1
0
0
1
1
1
1
1
1
1
0
1
1
1
1
1
1
1
1
0
0
1
0
0
1
0
0
1
AdSAndré
AdTMAD
AdAlgarve
AdCentro
AdCAlentejo
AdDPaiva
AdMondego
AdNoroeste
AdNAlentejano
AdOeste
AdVouga
AdZCoa
AdPAlentejo
EPAL
ICOVI
21
4. Conclusions
•
•
•
•
The exploration of the potential application of MCDA methodologies for
the assessment of performance in sorting and ranking problematics
showed they can be appropriate to this decision problem.
The results, although slightly more favorable in ELECTRE TRI, reveal a high
number of operators with unsatisfactory sustainability and environmental
performance.
The two methods highlight the good performance of the "AdTMAD" and
"AdNoreste".
The application of the additive aggregation method‐MAVT, and of the
prevalence ELECTRE TRI method resulted in similar results, since the
compensatory nature of MAVT is counteracted by the veto threshold in
ELECTRE TRI.
22
Water utilities performance assessment Multicriteria methodologies contributions
Thank you
Sandra Tralhão1, Rita Martins2, João Paulo Costa3
1 Faculty of Economics, University of Coimbra, [email protected]
2 Faculty of Economics and CeBER/GEMF, University of Coimbra, [email protected]
3 Faculty of Economics and INESC Coimbra, University of Coimbra, [email protected]
References
1.
2.
3.
4.
5.
6.
7.
Armstrong, M., Cowan, S., & Vickers, J. (1999). Regulatory Reform – Economic Analysis and British Experience. Cambridge: The MIT Press.
Artley, W., & Stroh, S. (2001). Establishing an Integrated Performance Measurement System (2). Performance‐Based Management Special Interest Group.
Belton, V., & Stewart, T. (2002). Multiple Criteria Decision Analysis: An Integrated Approach. Springer Science & Business Media.
Church, J., & Ware, R. (2000). Industrial Organization ‐ A Strategic Approach. United States: Irwin McGraw‐Hill.
Danielson, M., & Ekenberg, L. (2016). The CAR Method for Using Preference Strength in Multi‐criteria Decision Making. Group Decision and Negotiation, 25, 775–797.
Dezert, J., & Tacnet, J.‐M. (2012). Soft ELECTRE TRI outranking method based on belief functions. IEEE, 607 ‐ 614.
ERSAR, & LNEC. (2012). Guia técnico nº 19 ‐ Guia de avaliação da qualidade dos serviços de águas e resíduos prestados aos utilizadores ‐ 2.ª geração do sistema de avaliação. Entidade Reguladora dos Serviços de Águas e Resíduos.
8. Figueira, J., Roy, B., & Mousseau, V. (2005). ELECTRE Methods. In J. Figueira, S. Greco, & M. Ehrgott, Multiple Criteria Decision Analysis ‐ state of the arte surveys (pp. 133‐162). Boston: Springer Sciense.
9. Grupo de Trabalho Operacional e Comissão de Acompanhamento do PENSAAR 2020. (2014). PENSAAR 2020 ‐ Uma nova Estratégia para o Setor de Abastecimento de Água e Saneamento de Águas Residuais, 2. Plano Estratégico de Abastecimento de Água e Saneamento de Águas Residuais. Portugal.
10. Keeney, R. (2007). Enquadramento de decisões de política pública. In C. Antunes, & L. Dias, Decisão: Perspetivas interdisciplinares (pp. 173‐210). Coimbra: Imprensa da Universidade de Coimbra.
11. Keeney, R., & Raifa, H. (1976). Decisisons with multiple objectives: prefrences and value trade‐offs. New York: John Wiley & Sons.
12. Lourenço, R., & Costa, J. (2004). Using ELECTRE TRI outranking method to sort MOMILP nondominated solutions. European Journal of Operational Research, 153, 271–289.
13. Marques, R. (2005). Regulação de Serviços Públicos. Lisboa: Edições Sílabo, Lda.
14. Martins, R. (2007). Regulação económica no setor das águas ‐ Promoção da concorrência e sustentabilidade tarifária. Tese de Doutoramento. Faculdade de Economia da Universidade de Coimbra.
15. Melo‐Baptista, J. (2014). Uma abordagem regulatória integrada (ARIT‐ERSAR) para os serviços de águas e resíduos. Entidade Reguladora dos Serviços de Águas e Resíduos (ERSAR).
16. Mousseau, V., Slowinski, R., & Zielniewicz, P. (2000). A user‐oriented implementation of the ELECTRE‐TRI method integrating preference elicitation support. Computers & Operations Research, 27, 757‐777.
17. OCDE. (2015). OECD Principles on Water Governance ‐ Water Governance Initiative. OCDE.
18. Parlamento Europeu e Conselho da União Europeia. (2010). Directiva 2000/60/CE. Jornal Oficial das Comunidades Europeias.
19. Riabacke, M., Danielson, M., & Ekenberg, L. (2012). State‐of‐the‐Art Prescriptive Criteria Weight Elicitation. Hindawi Publishing Corporation ‐ Advances in Decision Sciences.
20. Roy, B. (1996). Multicriteria methodology for decision aiding. Kluwer Academic Plublishers.
21. Shleifer, A. (1985). A theory of yardstick competition. Rand Journal of Economics,16 (3), 319,327.
22. Tundisi, J. (2003). Água no século 21: enfrentando a escassez. RiMa, IIE.
23. Vincke, P. (1992). Multicriteria decision‐aind. Bruxelles: Jonh Wiley & Sons, Inc.
24. Watson, S., & Buede, D. (1987). Decision synthesis. The principles and practice of decision analysis. Cambridge: Cambridge University Press.
24
25. Zopounidis, C., & Doumpos, M. (2002). Multicriteria classification and sorting methods: A literature review. European Journal of Operational Research 138, 229–246.