Around three quarters of the world population live in developing countries. In 1995, less than 30% of it had access to some form of sanitation. It is predicted that by the year 2000 there will be 2.6 billion people without adequate sanitation, the vast majority of them living in the developing world. The consequences of this appalling situation are dreadful human suffering, deterioration of the natural environment, and negative economic implications. In the past, numerous efforts were directed at improving the situation. However, too often projects failed. For example, the World Bank estimated that the funds directed to the water supply and sanitation sector in 1992 in only 44% of all cases led to sustainable projects. A key towards sustainability is the thorough assessment of the community situation and of community needs. Considering technical as well as sociocultural and economic criteria, a solution needs to be found that is appropriate. Accordingly, the purpose of this dissertation was to find a way to support decision makers and project beneficiaries in identifying feasible sanitation alternatives and in evaluating the adequacy of these alternatives with respect to community circumstances. The qualitative nature of this task led to three research questions:
1. Which sanitation technologies are relevant to rural and urban communities in developing countries?
2. Which are the relevant criteria to evaluate the appropriateness of these technologies?
3. How can these criteria be incorporated into a decision aid that can be applied by decision makers to concrete projects?
In the past, a number of authors have recognised the usefulness of decision aids for evaluating sanitation systems. However, the scope of the tools presented by them is usually limited. Either they cover only a small number of alternatives or few criteria are employed for the evaluation of these. There are examples for comprehensive approaches, but they either do not relate criteria to alternatives to provide concrete recommendations, or they are characterised by complex and thus unclear evaluation frameworks. It was found that only a computerised decision aid would allow the comprehensive assessment suggested to be necessary by the complexity of the sanitation issue, while hiding this complexity behind a well designed user interface. This dissertation represents the first attempt to develop a computerised decision aid that includes low-cost solutions and is aimed exclusively at supporting practitioners in the appropriate sanitation sector.
Answering the first two research questions provided the knowledge base required for developing this decision aid. Information was obtained from a comprehensive literature review, a survey on sanitation conducted among practitioners in the sector, and from a journey to Southeast Asia in 1996, where practitioners were interviewed and sewage treatment works and low-income settlements were visited. Based on this information, a theoretical evaluation model was developed. This model features two distinct evaluation stages. During the satisficing stage, infeasible alternatives are eliminated based on mainly technical criteria. Subsequently, during the comparative stage, the remaining alternatives are compared with regard to the indicators implementability and sustainability. Apart from technical issues, this stage considers numerous sociocultural objectives. Implementability expresses the probability that sanitation facilities can be constructed within the period and with the financial resources usually required in favourable conditions. Sustainability expresses the probability that facilities serve beneficiaries according to their design throughout their design life. A major obstacle to formulating the comparative component of the algorithm was the amalgamation of numerous criteria outcomes. The application of conventional methods, such as the arithmetic mean, would have resulted in a diminishing effect of the individual criterion. In order to preserve the effect of criteria, a new method for amalgamating criteria on multiple levels was developed, which would also allow the simultaneous application of various methods like the arithmetic and the geometric mean. The combined advantages of multi-level amalgamation are reflected in more plausible evaluation results.
Assuming that planners in developing countries usually have sufficient access to financial expertise, no criteria to assess the affordability of sanitation systems were formulated. Instead, it was decided to develop a costing model that would enable the proposed decision aid to estimate the capital as well as the recurrent costs of all alternatives. Further, a simple method based on the local residential building cost was developed to convert these estimates to costs in local currency units, as they would occur in the project area.
In summary, from the above work resulted an evaluation framework for assessing the implementability, the sustainability and for costing sanitation alternatives in developing countries. Subsequently, based on this abstract framework, the expert system software SANEX© was developed for the MS Windows operating environment. The knowledge base of this software contains more than 80 sanitation alternatives and uses around 50 criteria for their assessment. The costing component employs approximately 50 functions. In 1998, in order to validate a preliminary version of this software, a second journey was undertaken to Southeast Asia and to Europe. Based on its application to nine case studies, several modifications emerged to be necessary. However, those who contributed the case studies expressed good confidence in the recommendation s made by even the preliminary version of SANEX© This is particularly true for the satisficing evaluation component. A copy of the final, modified version of the software is enclosed with this thesis.
The development of a concrete expert system to support sanitation planners had theoretical and practical implications. Apart from evaluation, participants found the software particularly useful for demonstration and education purposes and for sensitivity analysis. From a theoretical viewpoint, this research presents the first attempt to systematically record the opinion of sanitation planners. In addition, since the development of any evaluation framework involving numerous criteria is a complex task, other disciplines might benefit from the multi-level amalgamation model developed. Not only the adaptation of the amalgamation procedure is possible, but also the entire software concept can easily be modified to accommodate evaluation problems from other fields. In fact, several practitioners in water supply and wastewater reclamation have expressed their interest in doing this.