Design parameters for rainwater catchment systems in arid and semi-arid lands of Kenya
Abstract
The study entails the analysis of design parameters for rainwater catchment systems CRWCS) as a
source of water supp ly, especially in the semi-arid and arid lands CASAL) where other sources of
water are limited. The design parameters in question are the design rainfall occurrence, amount,
distribution, recording interval and reliability, length of the wet and dry spells, water consumption
rates, number of consumers, catchment characteristics, storage capacities, etc.
In the thesis, rainfall analysis was accomplished through rainfall time series analysis, rainfall
occurrence, rainfall amount and distribution reliability and the occurrence of length of wet and dry
spells based on the 77 years of rainfall records at Kibwezi. The aspect of rainfall distribution was
introduced by frequency analysis of the standardized proportions of the monthly rainfall contributing
to the annual rainfall at various reliability levels.
The time series analysis entailed homogeneity, trend and periodicity tests. The homogeneity and trend
tests were performed through graphical and statistical methods. The periodicity test were carried out
through spectrum and correlogram analysis. The time series analysis results showed that the data
were homogeneous and free of any significant trend. But there were evidence of the following
periodic cycles, 3.3-3.5 years, 5-5.5 years and 10-14 years.
The relationships between rainfall occurrence transitional probabilities and the probability of any day
being wet; rainfall amount and the number of rainy days: length of wet spells and the rainsum and
rainfall amount and the rainsum were developed. The length of the wet and dry spells and the
occurrence of wet days were found to follow the Markov model. The drought chronology was also
evaluated and the occurrence of drought of various magnitude determined.
The design monthly rainfall at various distribution reliability levels were obtained from the ratio of
each month's rainfall contributing to the annual rainfall multiplied by the expected annual rainfall at
the same reliability level. The monthly rainfall interval was adopted for Kibwezi rainfall since the
daily rainfall were poorly distributed and the annual rainfall does not consider the rainfall distribution
within the year.
The mass curve technique was found supenor 111 the determination of the storage capacities and
catchment areas at different water demands and rainfall reliability levels. The storage cap-city and
catchment area varies respectively with rainfall distribution and rainfall amounts at various reliability
levels.
Selection of the design parameters were attached to their probability of occurrence and hence their
reliability levels, which gives an insight of the overall reliability of the RWCS. Provision of a 100%
reliability for RWCS may not be practical due to the stochastic nature of rainfall and also socioeconomic
considerations. Therefore. optimal RWCS can be designed at various reliability levels
depending on the individual economic status and the use of the system. In the thesis, RWCS design
procedures were considered in general and specific type of the catchments or the storage facilities
were not the basic concern as only their design parameters were emphasised. The uses of the RWCS
were not also a major concern, although RWCS for domestic water supply has been given priority.
Design curves based on the Kibwezi environment were developed, which gives the optimal catchment
area and the storage capacity at various reliability levels. The consumption rates were varied between
10-60 litres/person/day and the family size or the number of consumers were varied between 6-108
persons.
Water resources field survey was also conducted in some parts of Machakos and Makueni districts.
The survey assessed the water supply systems. extent of water shortages, types of RWCS, government
and/or NGOs interventions, family sizes, water consumption rates. socio-economic status and future
development of water supply systems. This survey revealed that this part of the country and other
regions with similar climatic conditions suffer from severe water shortages. Severity was found to
increase with aridity.
It was also encouraging to learn the willingness of the inhabitants to adopt rainwater harvesting as a
future source of good quality water. This is partly due to lack of alternatives and also due to the
ongoing promotion of utilization of rainwater by various NGOs working in this region. Therefore,
the developed design procedures will go along way in assisting the rural community meet their water
demands at minimum cost because most of the existing designs were inadequate and unreliable. The
RWCS promoted and funded by some of the NGOs were 'standard' for different families and regions.
Thus, the promotion of rainwater utilization should go hand in hand with the promotion of appropriate
design procedures.
Citation
Master of Science in Soil and Water EngineeringPublisher
University of Nairobi Department of Agricul tural Engineering