Optimization of non revenue water management for Livingstone Town-Zambia: A case study of Lizuma Ward.

Abstract

Livingstone Town has a water supply service coverage of 81% and has been experiencing a high Non-Revenue Water(NRW) which stands at 43%compared to the recommended value of 23% suggested for developing countries. The high NRW poses a threat to the sustainability of service provision as more resources are used compared to the revenue generated.Despite the town having a water audit team in place, high non-revenue water is still a challenge. It is from this background that a study was carried out in Livingstone Town in the periodfrom December 2015 to March 2016. The objective was to determine the trends of NRW from 2008 to 2015, establish factors contributing to NRW and estimate real losses using a hydraulic model (EPANET).The study focused on Lizuma Ward. Historical data maintained by the water utility on volume of water supplied and billed from 2008 to 2015 was analysed to determine the NRW trends for Livingstone Town.Determination of billing and meter errors was done through an independent meter reading exercise and meter testing respectively in Lizuma Ward. The results showed that yearly minimum and maximum NRW for Livingstone Town from 2008 to 2015 was 43% and 49% respectively with an average of 45%. The minimum and maximum NRW for Lizuma Ward was 62 % and 68% respectively with an average of 65% of which 85%were real losses and 15 % were apparent losses.Statistical analysis performed on the NRW for the eight year period gave a coefficient of variation of 0.04 which showed low variability of NRW from year to year.The Mann-Kendall trend tests indicated that there was no distinctive trend observed for NRW during the period under investigation as the computed p-value (0.898) was greater than significant level (0.05). The study also established that apparent losses included under billing (18 %), over registering of meters by 3.1 % and unauthorized consumption (85.1 %). Unauthorized consumption was deduced from total apparent losses, billing and meter errors. The accuracy ofEPANET for estimating real losses was determined by performing a t-test on simulated and measured inflow to the ward. The test gave a p-value of 0.852 suggesting that there was no significant difference in the means. The t-tests was also performed on the simulated and measured pressure on five different location in Lizuma Ward distribution network which gave p-values of 0.842, 0.18, 0.131, 0.247 and 0.66 which were higher than significant level (0.05). Therefore there was no difference in the means of simulated and measured pressure hence suggesting that EPANET could be used to estimate real losses. The study concluded that there was no distinctive pattern observed for NRW trend.The main contributing factor to NRW were the real losses (85 %) while the unauthorized consumption (85%) was the major contributing factor to apparent water losses. EPANET can be used to estimate real losses. The study recommends that Southern Water and Sewerage Companyshould consider using EPANET in estimating the real losses, replace old pipes and water audit team should investigate unauthorized consumption.