While access to improved water sources has steadily increased across sub-Saharan Africa in recent years, a number of the water sources installed have been found to be dry/low-yielding, unsafe for human consumption, and marred by water colour, taste, and odour problems. A number of factors could be causing these yield and water quality problems. Understanding the interrelated nature of these factors and their inherent complexity is essential if these yield and water quality problems are to be avoided in the future.

Focusing on rural handpump-boreholes (HPBs), this research uses systems thinking and the systems-based analytical tools, Bayesian Network (BN) and Causal Loop Diagram (CLD) analysis, to systemically assesses the reasons for these yield and water quality failures. To begin, BN analysis is used to quantitatively identify the on-site factors that most greatly reduce the probability of HPB yield and water quality failure (for example, rock type and handpump and borehole materials). Data from Ethiopia, Malawi, and Uganda are in the BN. Key findings from the BN analysis highlight the importance of borehole and cylinder depth, rock type, handpump parts, and borehole design when considering HPB yield and water quality.

CLDs are then used to identify the underlying factors and drivers that affect the state of the BN’s on-site factors. Specific focus is placed on the implementation work (the siting and drilling/installation work) when assessing the underlying factors and drivers. The CLDs highlight the importance of having balancing loops within the system (for example, those of monitoring and evaluation) and appropriate systemic goals (for example, focusing on the quality of the implementation work, rather than the number of new water sources when constructing new water sources), given the self-correcting power these aspects possess. As an extension of the CLDs, a case study analysis of the Ugandan HPB implementation process is presented, in which the state of the key factors and drivers emerging from the CLD analysis are assessed and evaluated for the Ugandan context. A series of intervention points, based on the findings from the CLD analysis are recommended for Uganda.

The key assumption underlying this research is that HPB failure is the result of the interplay between a wide array of factors, deriving, for example, from the geology, proximity to pollutants, drilling/installation contract price and payment terms, among others. By analysing how these various factors interact, this research increases our understanding of the interconnected nature of the factors that lead to HPB yield and water quality failure and highlights several intervention points for alleviating these failures in the future.