This study presents trend analysis and forecasting of water level in Mtera dam. Data for water level were obtained from Rufiji Basin Development Authority (RUBADA). The study analyzed trend of water level using time series regression while forecasting of water level in Mtera dam was done using Exponential smoothing. Results revealed that both maximum and minimum water level trends were decreasing. Forecasted values show that daily water level will be below 690 (m.a.s.l) which is the minimum level required for electricity generation on 2023. It was recommended that proper strategies should be taken by responsible authorities to reduce effects that may arise. Strategies my include constructing small dams on upper side of Mtera dam to harvest rain water during rainy season as reserves to be used on dry season. In long run Tanzania Electric Supply Company (TANESCO) should invest into alternative sources of energy.

Water loss in the water distribution systems (WDS) is a challenge to many water authorities in the world but the problem is crucial in the less developed countries. The effect of water losses in the WDS includes the reduction in the revenue and availability of water, interruption in the quality of water, and inflation of the operation and maintenance cost of the water authorities. Using data from the Moshi Urban Water Supply and Sanitation Authority (MUWSA) Tanzania, an assessment of strategies used for water loss management (WLM) was carried out through an integrated model of Multi-Criteria Decision Making (MCDM) and Integer Linear Programming (ILP) which is an optimisation technique. The family of MCDM methods, Multi-Attribute Value Theory (MAVT), Simple Multi-Attribute Rating Technique Exploiting Ranks (SMARTER), and Simple Additive Weighting (SAW) were employed to assess and prioritise the strategies while the ILP was used to formulate a decision model. The model was used to select a portfolio of the best strategies. Sixteen strategies were identified. The results show that the comparison between the bulk meter and customers' meter on detecting the physical or apparent losses was ranked as the best strategy in managing the loss while the network zoning was ranked as the worst strategy. The model selected thirteen out of sixteen strategies to form the portfolio of the best strategies to be employed by the MUWSA for water loss management. Furthermore, the model was found to be robust as the selected portfolio of strategies remained the same even when the weights of the criteria were changed. The developed model in this study will assist the decision-makers to assess, prioritise and choose the best strategies for reducing or controlling water loss in the distribution system.

Water loss in water distribution systems (WDS) is a serious problem in Tanzania and the third world countries at large. A lot of water is lost on its way before reaching the consumers. This causes a shortage of water supply which leads to loss of revenues of the concerned water authorities. The control or reduction of water loss in the WDS is closely dependent on the commitment of the decision-makers and on the strategies and budget, they set for that purpose. This paper presents a combined model of Multi-Criteria Decision Making (MCDM) and Numerical optimization techniques which may help decision-makers to prioritize and select the best strategies to be used in the management of water loss in the WDS at Moshi Urban Water Supply and Sanitation Authority (MUWSA), Tanzania. The Multi-Criteria Decision Making family methods namely the Multi-Attribute Value Theory (MAVT), Simple Multi-Attribute Rating Technique Exploiting Ranks (SMARTER), and Complex Proportional Assessment (COPRAS) were used to evaluate and prioritize the strategies, whereas the Integer Linear Programming (ILP) technique a numerical optimization technique was used to select the best strategies or alternatives to be employed in water loss management. The results show that the most preferable alternative is replacement of dilapidated pipes while the least preferable alternative is network zoning. The model selects thirteen out of sixteen alternatives, which cost 97% (TZS 235.71 million) of the total budgets set by the water authority to form a portfolio of the best alternatives for water loss management. Furthermore, the model showed robustness as the selected portfolio of alternatives remained the same even when the weights of the evaluation criteria changed.

This study presents the modelling of impacts of climate change on water resources. Mtera dam in Tanzania was taken as a case study. Data for climate variables on four stations were obtained from Tanzania Meteorological Agency (TMA) while data for water level were obtained from Rufiji Basin Development Authority (RUBADA). The study aimed at doing regression analysis on all stations to analyze the impacts of change in climate variables on water level. Results show that rainfall was significant predictor of water level at Iringa and Dodoma while temperature and sunshine were significant at Mbeya station. Change in climate variables accounted for 37% of the fluctuations of water level in the dam. It was recommended that TANESCO should construct small dams on upper side of Mtera dam to harvest rain water during rainy season. In long run TANESCO should invest into alternative sources of energy.

In this study, an optimal clean water distribution system cost model has been developed to find the minimum cost to distribute clean water. The model was then tested with real data collected from Ihumwa water distribution network of Dodoma city and other treatment cost data from literature to test the workability of the model. Hydraulic parameters such as head losses of the pipes, flow velocity and pipe pressure are calculated using water flow software. The resulted model was solved using LINGO software and the optimal cost of clean water distribution system was found by testing the different maximum and minimum velocity and pressure that give an optimal cost.