HDT 68: Decision Analysis Methodology For Selecting Alternative Treatment and Use Of Wastewater
Eng. Marco Antonio Almeida de Souza
Spanish version: HDT 68: Metodología de análisis de decisiones para seleccionar alternativas de tratamientos y uso de aguas residuales
Brasilia University. Department of Civil Engineering. University Campus – Asa Norte. 71910 – 090 Brasilia – DF, brasil.
April 1997
General index
- Introduction
- Background information
- Considerations for selection
- Proposed methodology for selection
- Analysis methods of multiple objective decision
- Bibliographic references
Introduction
The concern for sustainable development has led to a change in who conduct research programs and planning of water resources. This change has led them to consider alternatives to the rational use of natural resources, including waste minimization and water reuse.
The selection of technologies for the collection and treatment of waste water must increasingly consider to a greater extent alternatives that include the reuse of water. The question that arises is: How consider all decision variables simultaneously and how to combine them to get an answer that satisfies interested persons?
To answer this question, this paper proposes and discloses a methodology for the selection of integrated treatment, recovery and use of waste water. This methodology is based on the use of decision analysis methods with multiple objectives and criteria allowing a holistic treatment technology selection (Schumacher, 1973; Willoughby, 1990).
These methods replace the economic and optimization methods criticized for monetizing and materialize the factors involved. The proposed methodology can be used in any other environmental and sanitation area.
Background Information
The application of decision analysis method for the selection of alternative waste water treatment is first recorded in 1987 when Wolf (1987) presented a simple methodology based on the weighted average method. Subsequently, Tecle et al. (1988) introduced decision analysis techniques with multiple objectives, such as the “Compromising Programming” and Electre-I.
In a similar manner Souza (1992) developed PROSEL-I model (“Process Selection Version I”) to select waste water treatment processes using the principles of appropriate technology and decision analysis with multiple objectives.
In a project aimed at solve a specific case, Gobbetti (1993) and Gobbetti & Barros (1993), applied multi-objective analysis techniques “Compromise Programming”, “FunçÆo Utilidade Multidimensional” Electre-I and Promethee for the revision of Drains master plan SÆo Paulo, Brazil .
We can cite many applications in other areas of environment and water resources. For example, Perlac & Willis (1985) used analysis decision methods with multiple objectives to solve a problem of solid waste management.
Merkhofer & Keeney (1987) applied the multi-attribute utility analysis to define the place of nuclear waste disposal. Briggs et al. (1990) applied the methods Gaia and Promethee for managing nuclear waste. Hokkanen et al. (1995), Caruso et al. (1993), Maystre & Simos (1987), and Simos (1990) used diverse techniques of decision analysis with multiple objectives and criteria for the selection of alternatives for solid waste management.
Duckstein et al. (1994) used analysis techniques with multiple criteria decision Compromise Programming, FunçÆo Utilities Multidimensional, Electre-III, and UTA (from French Utilité Additive) for the selection of alternatives for ground water management.
These methods constitute a tool which helps to make choices when you have to consider tangible and intangible factors and when community participation is desired.
Considerations For Selection
To pose the problem of planning is appropriate to review the integrated systems of treatment and use of waste water to generate multiple alternatives with percentages of use for agriculture, forestry, livestock, aquaculture and public areas irrigation. These alternatives shall also address the distribution technology of recovered waste water for different users.
To achieve the technical and economic development of water reuse projects in developing countries and their application in aquaculture and agriculture, some researchers defend the use of stabilization ponds in series due to its low cost and high security offered for health (León & Moscoso, 1996; Moscoso &; Florez Muñoz, A., 1991; Moscoso et al., 1991).
In the case of potable direct reuse, it is required that the system of wastewater recovery has a high degree of efficiency and reliability, being necessary to employ combinations of processes and unit operations which generally include clarification with chemical precipitation, nutrient removal, recarbonation, filtration, activated carbon adsorption, reverse osmosis demineralization and chlorine disinfection, ozone or both (Metcalf & Eddy, 1991).
In selecting technology of waste water recovery, operational reliability and performance of all processes and unit operations are important factors (Metcalf & Eddy, 1991). However, the selection of a technology should be considered as a problem both particular as home. The solutions can not be generalized (there is only a particular case) and you have to to examine the influence on ambits the social, economic, cultural, legal, environmental, educational, etc.
Proposed Methodology For Selection
Some decision analysis methods that allow the participation of the population and holistic treatment is proposed. This methodology has the following steps:
- Definition of the objectives of the, water use and quality of treated effluent system
Planning for water reuse should be conducted at the hidrographic watershed _and consider cooperation between different agencies and stakeholders. It should include the following points:
- inventory of the treatment needs and waste water disposal;
- inventory of demand and water supply;
- inventory of the benefits of water supply through reuse;
- market analysis for recovered waste water;
- technical and economic analysis of alternatives (above) and
- reuse implementation plan with financial analysis (Metcalf & Eddy, 1991).
In Asano (1991), is a list of activities and inventories to be performed in planning the waste water reuse
A existing Market for recovered waste water is essential. You should make a list of potential customers and their purchasing capacity. The inventory should also include the needs of potential users.
- Definition of the quality of natural waste water or treated effluent
You need to determine the characteristics of water quality according to the demands of predefined usage patterns and municipal and industrial contributions in the basin of waste water collection. For industrial water reuse is necessary to make an inventory of all possible contaminants.
The definition of the characteristics of water quality is not an easy task and can lead to errors in the later stages. The choice of variables depends on local epidemiological analysis and examination of the actual situation, having to be wary of simplistic studies that only work with NMP or BOD of fecal coliform, unless is needed for the study.
- Definition of alternative integrated systems of treatment and use of waste water
To define the universe of alternatives must take into account the criteria of reliability and effectiveness of each alternative to rule out non-viable and to not scrap some improperly. There may be many alternatives due to the different combinations of processes and unitary operations and forms of distribution and use of the recovered water.
The location and type of reuse should be considered as part of each alternative, including the water distribution of in different consumption points.
- Definition of decision variables to select the alternative waste water recovery
The decision variables may include principles of appropriate technology and sustainable development, such as:
- maximization of the use of local materials resources and labor;
- minimization of energy consumption;
- maximization of the quality and quantity of the final effluent for reuse;
- minimization of environmental impact;
- economic profitability;
- ease of operation and maintenance;
- minimization of risk to the health of workers and the public;
- social participation; and
- public acceptance.
You can have more than one criteria to measure each objective and can use discrete or continuous, numerical or ordinal variables. Variables can be quantified, such as cost, area occupied, etc., or subjective, as preference for one or another aspect of each alternative. The variables can be compared to each other with value judgments such as good, average, poor, better, worse, etc. For each objective can be used efficiency criteria, reliability, endurance, flexibility, etc.
- Comparison of the alternatives for waste water recovery with the decision variables
In the analysis of multiple objective decision (Souza, 1992; Souza & Forster, 1996), the method for the selection of technology more currently recommended should be included the called results matrix (payoff matrix) which compares feasible alternatives according to the consideration degree to all the decision variables from the previous step.
This matrix is composed by the criteria and activities. The evaluation of alternatives can be performed with an executive decision group or a team of reuse specialists and popular participation should be requested _ through a “waste water user council”.
The success of the method depends on how judicious result the evaluation of alternatives. It Can be made based on questions and weights attribution, so that evaluators do not distinguish the relationship between their responses and the final result. Methods have been developed, such as Delphi, to answer scientifically. The definition of objectives and variables to be considered in each case must be performed by the same decision agent.
- Choosing an auxiliary method for the decision
To escape the excessive monetization of the decision variables and reach a point of satisfaction of the objectives outlined in the choice of technology, several authors advocate using of decision analysis with multiple objectives (Souza, 1992; Souza & Forster, 1996).
The most widely used methods when you have a discrete variable, which is the easiest to operate, are those of the Electre series (version I, II, III and IV-A), “Compromising Programming”, the Promethee, that of “Multidimensional Utility Function “and the ME-MCDM (“Multiple ExpertMultiple CriteriaMultiple Decision Makers”). There are computer programs available to them, which facilitates its use in low capacity microcomputers (Souza, 1996 and 1992; Goicoechea et al., 1982).
Some specialists in treatment and wastewater reuse are not interested in the analysis of decisions and resist using the proposed methodology because it seems complicated. However, methods of analysis of simple decision and easily understandable exsists. Those interested should consult the recommended literature (Goicoechea et al .., 1982; Souza, 1994).
Another point that should be emphasized is the use of several methods (two, three or four) to prioritize the solution of the problem (Souza, 1992; Tecle et al .. 1988).
- Hierarchization of the alternatives for waste water recovery and acceptance of the result
With any of the auxiliary methods of decision a list of alternatives in accordance to achieving the objectives. As stated earlier, it is better to use more than one auxiliary method to compare the results.
In some methods is possible to do a sensitivity analysis to know the decision factors that most contribute to the solution. The process allows to adapt the values of one of the variables and check the effect of such variations in the method responses . With the sensitivity analysis it is also possible to know the reasons that lead the preference of one or another alternative.
- Process repetition with modifications if you have not reached a decision
The ordered list of alternatives and an explanation of the organization process is discussed with the decision agents to determine the acceptability of the proposed solution. Usually, the result is questioned because only then some faults are recognized in one or some of the previous stages since many of these are interdependent.
If there is no consensusshould be considered all suggestions and criticisms in the methodological process of relevant previous stages and restart the cycle. The process is repeated until reach a satisfactory decision
It’s necessary to understand that there is no an optimal solution for the problem and should reach a “compromise” that can not be easily defined in numerical terms.
Is more than the “Pareto optimal” or “point of satisfaction”, which is a possible solution (ie, a viable alternative) for which there is no other viable solution that causes an improvement in any of the proposed objectives without aggravating at least one of the other goals.
Analysis methods of multiple objective decision
The main methods with the respective references for consultation are:
- Additive weighting method (Wolf, 1987; Goicoechea et al., 1982).
- Electre-I method (Benayoun et al, 1966;. Tecle et al., 1988).
- Electre-II method (Roy, 1973; Hokkanen et al., 1995).
- Electre-III method (Roy, 1991; Duckstein et al., 1994).
- Electre-IV and IV-A method (Roy, 1991).
- “Compromising Programming” method (Zeleny, 1973; Tecle et al., 1988; Duckstein et al., 1994).
- Method of the Theory of Cooperative Games (Szidarovsky, et al 1984;. Tecle et al 1984;. Tecle et al., 1988).
- Promethee method (Briggs et al., 1990).
- Method of Multidimensional Utility Function (Keeney & Raiffa, 1976; Merkhofer & Keeney, 1987; Duckstein et al., 1994).
- UTA method (Net Additive) (JacquetLegrèze & Siskos, 1982; Duckstein et al., 1994).
With so many methods available, choosing one of them is a challenging task. It is proposed to use three or four methods selected in accordance to the experience and ease that represents the analyst and by its adaptation to the case of water reuse. The result obtained with the various methods must be analyzed for compatibility check. Experience has shown that most methods converge at a common result.
The additive weighting method is based on its utility, a relationship between the value of a decision variable and its value attributed in the decision and, crucially, on the weighted average of the values of all criteria, in which the weights attributed importance to each of them. The selected alternative is the one which obtains the highest value of additive utility (that is, the weighted average). This method is widely criticized, but it is the simplest and allows the inexperienced understand the justification for the decision.
The Electra-I method (of the acronym in French translation of Reality for Elimination of Choice) compares alternatives in pairs between all the alternatives, so that eliminates a subset of alternatives and select those that meet most of the criteria. Involves three concepts mathematically defined: concordance, discordance and limit values of concordance and discordance. The method was not designed to organize alternatives, but Souza (1992) proposes a simple and effective way to rank the alternatives.
The Electra-II method is a modification of the first version and was developed to orient the alternatives. Create comparison relationships between strong and weak, concordance terms and different discordance, and an alternative ordination process in three stages. In the first stage is sorted in descending order (as best to worst alternative), in the second stage in ascending order and in the third stage as the arithmetic average of the above.
The Electra-III method is based on fuzzy comparison relations. Creates the pseudocriterio concept to measure the preferences of the decision agent Allows working with indifference relationships “not very strict” or “strict” and combines the limits of indifference and preference with the traditional criteria.
Add these partial preferences and forms a fuzzy comparison relationship that establishes two preliminary ordination of the alternatives (up and down) with a “distillation” technique. The method performs the final ordination from these two preliminary arrangements.
This method is the most acceptable when there is doubt and uncertainty in the evaluation of alternatives and allows comparison. Does not use comparisons provided directly by the decision agent, but creates them internally from the preferences.
Also makes a partial compensation of the comparison from the choice of the indexes of concordance and discordance as well as allowing the decision agent to express their preferences, indifference and opposition to choose weights that measure the degree of importance of the diverse criteria.
The Comprising Programming method is a method developed to identify the closest to the ideal solution, therefore not viable formed by a vector of the best recorded values for all criteria. The degree of closeness is measured by a pattern of distances that considers another vector formed by the worst values of the various criteria recorded in the results matrix. It is a quick and easy method of application.
Finally, because of its importance, is quoted Duckstein et al. (1994) “The purpose of applying the methods of decision analysis with multiple objectives and criteria is to help the decision agent.decision and not to replace it.”
Bibliographic references
Asano, T. Planning and implementation of water reuse projets. Water Science and Technology 24(9):1-10; 1991.
Benayoun, R.; Roy, B.; Sussman, B. Electre: une methode pour guider le choix en presence de points de vue multiples. Note de Travail 49, París: SEMA, Metra International, 1996.
Briggs, T.; Kunsh, P.L.; & Marshal, B. Nuclear waste management: an application of the multicriteria Promethee methods. European Journal of Operational Research 44:1-10; 1990.
Caruso, C.; Colorni, A.; Paruccini, M. The Regional Urban Solid Waste Management System: a modelling approach. European Journal of Operational Research 70:16-30; 1993.
Duckstein, L.; Treichel, W.; Magnouni, S.E. Ranking groundwater management by multicriterion analysis. Journal of Water Resources Planning and Management 120(4):546-565; jul.-ago. 1994.
Gobetti, L.E.C. Análise multiobjetivo aplicada ao planejamento de sistemas de recursos hídricos. DissertaçÆo de mestrado. SÆo Paulo: Escola Politécnica, Universidade de SÆo Paulo; 1993.
Gobetti, L.E.C.; Barros, M.T.L. Análise multiobjetivo aplicada ou planejamento de sistemas de recursos hídricos. En: Anais do X Simpósio Brasileiro de Recursos Hídricos. Gramado, Rio Grande do Sul: ABRH-AssociaçÆo Brasileira de Recursos Hídricos; 1993. pp. 317-326;
Goicochea, A.; Hansen, D.R.; Duckstein, L. Multiobjective decision analysis with engineering and business applications. New York: J. Wiley & Sons; 1982.
Hokkanen, J.; Salminen, P.; Rossi, E.; Ettala, M. Choice of a solid waste management system using the Electre-II decision-aid method. Waste Management & Research 13(2):175-193; 1995.
Jacquet-legrèze, E.; Siskos, J. Assesing a set of additive utility functions for multicriteria decision-making, the UTA method. European Journal of Operational Research 10(2):151-164; 1982.
Keeney, R. F.; Raiffa, H. Decision with multiple objectives: preferences. 1976.
León Suematsu, G.; Moscoso, J. Curso de Tratamiento y Uso de Aguas Residuales. Lima: CEPIS, Centro Panamericano de Ingeniería Sanitaria y Ciencias del Ambiente; 1996.
Maystre, L. Y.; Simos, J. L’aide à la décision dans l’organization. Paris: Association Française pour la Cybernétique Économique et Technique; 1987.
Merkhofer, M. W.; Keeney, R. L. A multiattribute utility analysis of alternative sites for the disposal of nuclear waste. Risk Analysis 7:173-194; 1987.
Metcalf & Eddy, Inc. Wastewater Engineering: Treatment, Disposal, and Reuse, 3ed. New York: McGraw-Hill; 1991.
Moscoso, J.; Flórez Muñoz, A. Reuso en acuicultura de las aguas residuales tratadas en las lagunas de estabilización de San Juan. Sección I: Resumen ejecutivo. Lima: CEPIS, Centro Panamericano de Ingeniería Sanitaria y Ciencias del Ambiente; 1991.
Moscoso, J.; León Suematsu, G.; Gil, E. Reuso en acuicultura de las aguas residuales tratadas en las lagunas de estabilización de San Juan. Sección II: Tratamiento de las aguas residuales y aspectos sanitarios. Lima: CEPIS, Centro Panamericano de Ingeniería Sanitaria y Ciencias del Ambiente; 1991.
Perlac, R. D.; Willis, C.E. Multi-objective decision-making in waste disposal planning. Journal of the Environmental Engineering Division, Proceedings of ASCE 11(EE3):373-385; 1985.
Roy, B. The outranking approach and the foundations of Electre methods. Theory and Decision 31:49-73; 1991.
Schumacher, E.F. Small is beautiful: economics as if people mattered. Londres: Harper & Row; 1973.
Simos, J. Évaluer l’impact sur l’environement. Lausanne: Presses Polytechniques et Universitaires Romandes; 1990.
Souza, M.A.A. Programas computacionais para alguns métodos de análise de decisÆo com múltiplos objetivos: PonderaçÆo Aditiva, Compromising Programming, Electre-I, y Electre-II. Software para microcomputadores. Brasília: Departamento de Engenharia Civil, Universidade de Brasília; 1996.
Souza, M.A.A. Análise de decisÆo com múltiplos objetivos. Brasília: Departamento de Engenharia Civil, Universidade de Brasília; 1994.
Souza, M.A.A. Methodology for selection of wastewater treatment processes. Ph.D Thesis. Birmigham: School of Civil Engineering, University of Birmingham; 1992.
Souza, M.A.A.; Forster, C.F. Metodologias para seleçÆo de processos de tratamento de águas residuárias. Engenharia Sanitária y Ambiental (ABES/Brasil) 1(2):19-31; jun. 1996.
Szidarovsky, F.; Duckstein, L.; Bogardi, Y. Multiobjective management of mining under water hazard by games theory. European Journal of Operations Research 15(2):251-258; 1984.
Tecle, A.; Fogel, M.; Duckstein, L. Multicriterion selection of wastewater management alternatives. Journal of Wastewater Resources Planning and Management Division, Proceedings of ASCE 114(4):383-398; jul. 1988.
Willoughby, K.W. Technology choice; a critique of the appropiate technology movement. London: Intermediate Technology Publications; 1990.
Wolf, P. Auswahl-und Bewert-ungskriterien für Kleine Klarangem. Abwassertechnik 38(2):5-7; 1987.
Zeleny, M. Comprising Programming. En: J. L. Cochrane & M. Zeleny, Eds., Multiple criteria decision making. Columbia: University of South Carolina Press; 1973. pp. 263-301.
