Assessment and Control of the Quality of Data Used during Dam Reviews by Using Expert Knowledge and the ELECTRE TRI Method

Abstract

Generally, experts use very large quantities of data during the analysis of civil works. These data are of four main types: visual observations, monitoring data, data calculated from models, and design and construction data. They are frequently “imperfect”: they contain uncertainty, imprecision, incompleteness. These imperfections have various roots, for instance, material location and sampling, number of measurement instruments, maintenance, such as scraping downstream slope, and frequency of monitoring measurement, etc. The method proposed in this paper first permits the identification of the various sources of imperfection and their definition as criteria; second, it provides a reliable way of assessing these criteria; third, it applies the multicriteria analysis ELECTRE TRI method to aggregate the values resulting from the assessment of the criteria. Finally, corrective actions are proposed to maintain or replace the measurement system under standard operating conditions. The method was applied to a real-case study.