Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Accueil du site → Doctorat → États-Unis → 1999 → Monte Carlo design of residential water rates under uncertainty : The case of Dakar, Senegal

University of North Carolina at Chapel Hill (1999)

Monte Carlo design of residential water rates under uncertainty : The case of Dakar, Senegal

Cueva, Alfredo Hernan

Titre : Monte Carlo design of residential water rates under uncertainty : The case of Dakar, Senegal

Auteur : Cueva, Alfredo Hernan

Etablissement de soutenance : University of North Carolina at Chapel Hill,

Grade : Doctor of Philosophy (PhD) 1999

Résumé
This dissertation is the first application of Monte Carlo simulation (MCS) to design of water rates in a developing country. A deterministic net revenue model and its stochastic counterpart were developed and calibrated using data from a financial study and a contingent valuation household survey from Dakar, Senegal. One hundred prices were randomly chosen over a selected range to produce deterministic estimates of net revenue. The prices were then used in MCS to produce mean net revenues and ninety-nine percent confidence intervals around means from the stochastic model. For each price and for two sets of deterministic model parameters (means and medians), net revenues from the deterministic model were compared to the confidence intervals around means from the stochastic model. For the comparisons performed with means, the results from the deterministic and stochastic models were statistically equal only two times ; with medians, they were statistically equal eight times. Variable correlations not included in the deterministic model appear to be the main reason for discrepancies between the deterministic and the MCS results. For revenues to cover costs in Dakar, an eight-percent price increase to private-connection subscribers was needed using the deterministic model with mean parameters. The deterministic model with median parameters required a 39 percent increase, and the stochastic model about 34 percent. Confidences that those prices would produce non-negative net revenues were 28 percent, 44 percent, and 42 percent, respectively. Hence, the deterministic model based on mean parameters runs the risk of recommending too low prices. The MCS model was used in a series of experiments to investigate the influence of rate level and rate structures on fulfilling four policy objectives : financial self-sufficiency, affordability by the poor, health promotion, and water conservation. These objectives were respectively measured by four indicators : net revenue, fraction of household income spent on water, per capita consumption, and relative change in forecasted production. MCS modeling provides detailed insights on the influence of current and alternative rate levels and tariff structures for the case of Dakar.

Search Oxford Libraries Online (SOLO)

ProQuest Dissertations Publishing

Page publiée le 23 décembre 2017, mise à jour le 1er février 2019