Exploring “commercial” water losses in an urban water supply system in Uganda: A geostatistical modeling approach
2008 (English)In: Applied Spatial Analysis and Policy, ISSN 1874-463XArticle in journal (Other academic) Submitted
Like many urban water utilities, Kampala Water (KW) in Uganda has faced challenges in dealing with Non-Revenue Water, defined as the difference between water produced and water sold. One particular challenge has been in quantifying and localizing the “apparent” or “commercial” losses fraction of the water balance relating to the KW service area. This paper presents a study in which geostatistical modelling was used to investigate the possible causes of stagnation in watersales volumes within a selected part of the KW service area, despite continued growth in thecustomer base resulting from connection of new customers. In the study, a variogram model ofwater sales volumes was established. This model was then used to predict water sales volumes ona raster grid throughout the study area. The predicted volumes were compared with actual watersales for a selected period, and a difference map was generated. Subsequently, post plots of thelocations of various reported field anomalies (aged and defective meters, water supplyinsufficiencies, leaks, illegal consumption, disconnected accounts, and accounts billed onestimated consumption) were overlaid in turn on the difference map of water sales volumes.These map overlays were used to explore the spatial correlations between the prevalence of thedifferent categories of field anomalies and occurrences of large differences between actual andpredicted water sales volumes. The comparisons served to highlight the locations and possiblecauses of significant drops in water sales volumes. Through the study, a geostatistical analysis toolwas developed consisting of computer code written using R, an open-source statistical computingand graphics language and environment. This tool will be incorporated as a module within aspatial decision support system prototype being developed for the KW service area.
Place, publisher, year, edition, pages
Water loss analysis, Geostatistical modelling, Variogram analysis, Non-revenue water, Decision support systems
Oceanography, Hydrology, Water Resources Water Engineering
IdentifiersURN: urn:nbn:se:kth:diva-8671OAI: oai:DiVA.org:kth-8671DiVA: diva2:14053
QS 201203142008-06-042008-06-042012-03-14Bibliographically approved