Browsing by Author "Mulugeta, Yilma"

Now showing 1 - 1 of 1
  • No Thumbnail Available
    Item
    Water Quality Assessment Using Optimal Multiobjective Waste-Load Allocation Approach: The Case Of Little Akaki River
    (Addis Ababa University, 2019-04) Mulugeta, Yilma; Zebene, Kiflie (Assoc. Prof.); Andreas, Windsperger *(PhD); Nebiyeleul, Gessese (Phd)
    In Addis Ababa, indiscriminant waste disposal from domestic, industrial and commercial sources affect the water quality of the Little Akaki River. Moreover, this situation is limiting the usability of the River. Consequently, in many studies, the River is regarded as one of the most polluted Rivers in Ethiopia. On the other hand, there are inadequate comprehensive studies on the river principally due to insufficient research fund. In addition, existing studies on the River mainly focus on concentration measurement of certain constituents and their comparison against local and international standards. This approach has limited the public and policy makers from knowing the exact pollution status of the River. Furthermore, a number of studies conducted on the pollution problem and ongoing efforts to enforce existing environmental regulations have not effectively restored the river water quality. As a result, its pollution problem has increasingly worsened. Therefore, in this study, the application of statistical multivariate analysis for regular and economical water quality assessment, water quality index analysis for summarizing the water quality situation and optimal waste-load allocation modeling as a tool for decision-making are sought. For statistical multivariate analysis, twenty-seven locations from the River and tributaries were sampled and analyzed in October/November 2015. Afterwards, multivariate statistical tools were used to investigate data from measurements and laboratory analysis. Accordingly, the cluster analysis divided the sampled sites into three according to level of their pollution. This result indicates that water quality variation was caused because of the difference in land-use conditions. In addition, for the spatial analysis of the three pollution groups, backward stepwise approach of discriminant analysis was identified to provide data reduction (87.5%) to two parameters resulting in 85.2% correct assignment. The principal component analysis/factor analysis identified ten parameters accounting for 81.9% of total variation. However, data reduction was not significant. The factors that were latent and identified from the principal components’ varimax rotation suggest that variation in water quality was caused mainly by domestic sewage. The outcomes show that the methods can be applied to evaluate the river water quality variation using three monitoring sites and ten parameters: total nitrogen, total suspended solids, total ammonia, chemical oxygen demand, nitrite, total phosphorus, phosphate, nitrate, biological oxygen demand and electrical conductivity. This, in consequence, requires lesser cost and effort and hence paves way for more affordable, regular water quality evaluation of Little Akaki River. For index analysis, twelve water quality parameters from twenty-seven sampling sites in the Dry season (January/February, 2017) and Wet season (October/November, 2015) were used for index determination. Results show that, all sampling sites except one site in the upstream were under poor water quality category. Afterwards, the neural network model was trained and validated, for twelve inputs and one output, using several combinations of hidden layers (2-20), number of neurons in the hidden layers (5, 10, 15, 20, 25), transfer, training and learning functions. The most optimal model architecture was obtained with eight hidden layers, fifteen hidden neurons that resulted in R2 value of 0.93. This shows a good agreement between the calculated and predicted index values suggesting that artificial neural network can be successfully applied for modeling Little Akaki River’s water quality index. One of the ongoing efforts by the Addis Ababa Water and Sewerage Authority to control and minimize impact from pollution source is channeling a portion of domestic discharges to central treatment plants. However, the nature of these discharges is variable in time and space. Accordingly, impact on the river varies. Under limited treatment plant capacity, the river’s health can be maximized either by putting in place strict environmental control or by preferentially channeling the streams having significant impact on the river. Identification of these streams can be done through routine field sampling and laboratory analysis and decision making afterwards. However, this requires high financial, time and human resource. In this regard, water quality simulation can help to understand the interaction between pollution sources and the river. For this, monitoring data can be used to predict pollution contribution of the various sources on the river and agencies can apply this approach for environmental decision-making. In this study,QUAL2Kw was used to predict the river water quality. The model was calibrated and validated using data collected during dry (January/February, 2017) and wet seasons (October/November, 2015), respectively. The results from the calibrated model indicate that the model was able to reasonably predict the pollution of the river with R2 values of 0.91, 0.90, 0.81 and 0.89, respectively for dissolved oxygen, biological oxygen demand, total nitrogen and total phosphorus. Moreover, sensitivity analysis showed that dissolved oxygen, biological oxygen demand, total nitrogen and total nitrogen predictions are highly sensitive to point source flow and Manning’s n. Therefore, this model may be applied as an option for water quality management of the Little Akaki River. For optimal waste-load allocation, a simulation-optimization model was developed through integration of a water quality model - QUAL2Kw and genetic algorithm - PIKAIA. Afterwards, cost-performance and cost-performance-equity models were applied on water quality data set. The model resulted in pareto-optimal curves for conflicting objectives such as treatment cost and equity versus water quality performance. These curves offer convenient means for informed decision-making during environmental planning and implementation. Especially, the strategy is helpful in finding compromised solutions for pollution problems with conflicting objectives. In general, the study results indicate that significant waste load reduction is required for an improved water quality condition of the Little Akaki River.