• Presentation

Application of a financial model for determining optimal management of non revenue water in developing countries

Citation

Wyatt, A. S., & Romeo, K. J. (2010, January). Application of a financial model for determining optimal management of non revenue water in developing countries. Presented at 3rd Arab Countries Water Utilities Association (ACWUA) Best Practice Conference on Non-Revenue Water Management in the Arab Region, Rabat, Morocco, January 20-21,, .

Abstract

Non-revenue water (NRW) includes 1) physical losses (pipe bursts and leaks) and 2) commercial losses (illegal connections, un-metered public use, meter error, and data recording errors). NRW levels are high in many developing countries, and can be expensive to reduce. Various authors (Lambert, Farley, McKenzie, Trow, and others in the International Water Association [IWA]) developed a framework—the Economic Level of Leakage (ELL)—which outlines the optimum level of physical losses based on the costs of water production, physical loss control costs, and other engineering inputs. However, the ELL approach is less useful in developing countries, as it ignores 1) commercial losses, 2) the annualized cost of water supply capacity expansion, and 3) situations where production capacity does not meet demand. Also, the computerized ELL requires data which are not readily available in many countries. A new model which addresses these issues would allow individual, regional, or national utility managers and regulators to establish NRW targets and to optimally allocate resources to NRW management. This paper presents a financial model which addresses the limitations noted above and provides acceptably accurate values of optimal steady-state NRW. The model uses an NRW framework consistent with the IWA water balance and solves the algebraic optimality conditions for commercial and physical losses. The spreadsheet form of the model computes optimal NRW from a modest, commonly known set of inputs. The paper examines the sensitivity of the model to the accuracy of input data. Next, the paper presents both generic results for optimal NRW and specific results for 10 countries in Asia, Africa, and Latin America. Input data come mostly from secondary sources. Key results and their implications are reviewed. The paper closes with conclusions and recommendations for further research and model application.