8th ASCE Specialty Conference on Probabilistic Mechanics and Structural Reliability
PMC2000-084
RAINWATER HARVESTING AND THE RELIABILITY CONCEPT M. Mafizur Rahman, M. ASCE Bangladesh University of Engineering and Technology,Dhaka-1000
[email protected] Fateh-Ul-Anam Muhammad Shafee Yusuf Bangladesh University of Engineering and Technology,Dhaka-1000
[email protected]
Introduction Rural water supply in Bangladesh is based on groundwater, as it is free from pathogenic microorganisms and available in adequate quantity in shallow aquifers. In Bangladesh, except in coastal and hilly areas, a remarkable success has been achieved by providing 97 percent of the rural population with tubewell water. In the coastal belt, high salinity in surface and ground water and in the hilly areas, absence of good ground water aquifers as well as difficulties in tubewell construction in stony layers are the main constraints for the development of a dependable water supply system. At present, the success achieved in hand tubewell based rural water supply is on the verge of collapse due to the presence of arsenic in ground water in access of acceptable levels in the shallow aquifers. Provision of arsenic contamination free water is urgently needed to mitigate arsenic toxicity and to protect the health and well being of the rural population living in acute arsenic problem areas. The people, particularly the women, living in the problem areas have to walk long distances to fetch water from an available source (Ahmed, 1993). A rainwater based water supply system requires determination of the capacity of the storage tank and catchment area for rainwater collection in relation to the water requirement, rainwater intensity and distribution. The main advantages of a rainwater system are that the quality of rainwater is comparatively good, it is independent and therefore suitable for scattered settlement and the owners/users can construct and maintain the system. On the other hand, the availability of rainwater is limited by the rainfall intensity and availability of a suitable catchment area. The mineral free rainwater may not be liked by many and the poorer section of the people may not have a roof/catchment area suitable for rainwater harvesting. The design of rainwater roof catchment system is very simple if the standard deterministic method such as mass curve analysis is adopted. The most important aspect of the design is the uncertainty of the parameters. The design of a rainwater roof catchment system deals with a number of uncertain factors, which the design should incorporate. Rainfall, per capita water consumption, available roof area, economic capability of the household to construct the storage tank are uncertain parameters parameters and are the design parameters for design of a roof catchment system. This emphasizes the need of reliability based design approach be adopted instead of the deterministic approach. The reliability concept is important for the rainwater catchment system before it is constructed
Mafizur and Yusuf
1
as well as for the existing system after the construction. The objective of the study is to highlight the aspects of reliability based design and presenting the utility of the results.
Objectives of the Study The objectives of the study include: 1. Determination of the storage tank size for available roof area to collect water from rain for drinking purpose. 2. Estimation of reliability that the demand will be met for an available roof area for the local rainfall. 3 . Determining the frequency of shortage of water from roof catchment system for different sizes of roofs, tanks and consumption.
Data Collection The major data for design of a rainwater storage system is the rainfall information of Barisal in the southern part of Bangladesh. In this study, monthly rainfall data are used for analysis. Shorter time intervals (i.e. daily, weekly data) increase the accuracy of the results only slightly, but do increase the amount of calculations considerably. Monthly average rainfall data of 21 years (1975 to 1995) of these locations were collected from Bangladesh Meteorological Department (BMD).
No. of people
Rainfall data
Per capita daily
Roof material characteristic
Mass Curve Analysis
Storage Volume for different demands and roof areas
Reliability of a tank size for various demand and roof area
Degree of security of different tank sizes for varying demands and roof sizes
Figure 1. Overall design methodology
Mafizur and Yusuf
2
Methodology Overall design procedure is explained in Fig. 1. Mass curve analyses was the basic method applied for the design of storage device. Input variables include monthly average rainfall data, roof area, roof runoff coefficient, per capita water consumption per day and the size of the population using the storage system. These data were used to calculate the water available as supply from the roof catchment system and the demand. Supply of water was calculated following the basic equation: Q=A*f*R
(1)
where, A is the roof area (m 2), f is the roof runoff coefficient (a character of the roofing material and slope) and R is the monthly average rainfall depth (m) and Q is the amount of water (m3) obtained as supply for use. Hofkes (1981) showed typical values of the roof runoff coefficients . The tank size necessary for storage of rainwater for any year is the difference between the maximum surplus of water in the rainy season and the maximum deficit in the dry season as determined by the mass curve method for that year. The statistical distribution for demand and supply of rainwater follow normal distribution (Yususf, 1999). The reliability (Rahman, 1997) that water of any given demand area will be available from available roof was determined by the load resistance concept by mean first order second moment method (MFOSM) method. Forecast of situation arising from the construction of a tank smaller than the design size is important. The constructed tank of sufficient volume provides the security that rainwater will not spill over the tank for a span of time (few years). Tank volumes for different security levels were calculated depending on the demand and the roof area.
Design Consideration And Calculations The design of rainwater storage system depends on a number of uncertain factors. These are, the rainfall itself, the per capita water consumption in a day, the number of people using the storage system and the size of the roof area available to supply water. It is necessary to provide the explanations in a stochastic manner rather than only the deterministic approach. The storage tank volume (liter) was determined for varying demand (litres/ day) and available roof area (m 2). Since many uncertain factors are involved, the “Reliability” concept was introduced with rainwater harvesting. Reliability means the probability that a given size of tank will be sufficient to supply necessary amount of water. The probability that a tank of a given size will be sufficient to store all the water over a certain span of years are also important. It was determined by introducing a term “Degree of security”.
Mafizur and Yusuf
3
Roof runoff coefficient ( f ) was assumed to be 0.8 (applicable for corrugated sheets) throughout the study. Yusuf (1999) showed insignificant difference among the end results for variation of this parameter in the range of 0.75 – 0.85. Mean roof area of the individual households was approximated to be around 40 m 2.This approximation is important in the implementation stage while utilizing the results of the study but not during the analyses, since the study was performed for a range of situations where the individual households as well as the community level uses were considered. Per capita water consumption was assumed to be 6 liters/day and the average size of the household were assumed to be 5. The results were presented based on the total water demand in a day and not on the per capita demand demand or the population size. Demand and supply of rainwater were assumed to be normal variates (Yusuf, 1999) for the statistical calculations.
Results And Discussions Fig.2 shows the tank sizes for Barisal. From the number of users and the per capita consumption consumption in a day, the total demand per day can be calculated in liters. This gives the value in abscissa in Fig. 2. The roof area available in a particular situation should be a given data. A particular curve should be selected based on this data. Thus the tank size required can be calculated for a case from the ordinate. Curves are provided for roofs having sizes of 20, 40,60,80 and 100 m 2. Roof area for which curves are not given but falls within these values, proper interpolation should be done. Tank size reduces as the roof area increases for a fixed demand. Bigger roofs provide more water even as a result of small rainfall in the dry seasons reducing the need of stored water in these days.
20000 18000
16000 60
) s r 14000 e t i l ( 12000 e z i S 10000 k n a 8000 T
Catchment Area (m
2)
80 20
40
100
6000 4000 2000 0 0
10
20
30
40
50
60
70
80
90
10 0
Demand (liter/day)
Figure 2. Tank sizes for Barisal
Mafizur and Yusuf
4
100 90 80 70
) % ( y t i l i b a i l e R
60 50 40 30 20 10 0 0
1
2
3
4
5
6
7
8
9
10 10
2
Demand / Area (liter/day/m )
Fig. 3 Reliability of supply of demand at Barisal
Fig. 3 shows the reliability of supply of water for Barisal. The total demand in a day should be divided by the available roof area and the abscissa is fixed for a location. As the roof area increases for any fixed demand, the demand/ area value decreases and a shift in the abscissa occurs to the left. This increases reliability. Storage tank size for varying demand, roof area and degree of security for Barisal is provided (Fig.4). As the roof area and demand are fixed the abscissa is determined. A line for a particular degree of security should be chosen and the tank size for unit roof area can be determined. The actual tank size can be found by multiplying this value with the roof area which was used for determining the abscissa. 1200 ) 2 m / r e t i l ( a e r A / e m u l o V e g a r o t S
Degree of security 99% 1000
Degree of security 90% Degree of security 80%
800
600
400
200
0 0.0
0.5
1.0
1.5
2.0
2.5
3.0
Demand / Area (liter/day/m
3.5 2
4.0
4.5
5.0
)
Fig. 4 Storage tank size for varying demand, roof area and degree of security for Barisal
Mafizur and Yusuf
5
Conclusion The results can be utilized for design of a new roof rainwater catchment system as well as for assessing the reliability index of the existing rainwater roof catchment systems. Fig. 2 can be utilized prior to construction of designing the storage tank system for given demand and roof size. The roof size and demand can be estimated from the field data prior to construction. For a specific location demand and rainfall can be estimated probabilistically with a mean. These values will provide the reliability index (Fig. 3) indicating the overall reliability of the system prior to installation. This index physically means the reliability that the rainfall is enough to provide the demand under the given values of roof size and rainfall whatever the size of the storage tank is. Storage tank size can be determined (Fig. 4). For given roof catchment area and demand so that the tank is filled for some given period of time to provide the supply. The results, thus provide guidelines for design of roof catchment system with known reliability and functionality. References Ahmed, A. (1993), Prospects for Rainwater Catchment in Bangladesh and Its Utilization, Proc. 6th Int. Conf. on Rainwater Catchment System, Nairobi, 1-6 August. Hofkes, E.H. (1981), Rainwater Harvesting for Drinking Water Supply. International Reference Center for Community Water Supply and Sanitation, The Hague. Rahman, M.M. (1997), Reliability analysis for the planning of flood control and environmental conservation, Ph.D. Dissertation, Department of Civil Engineering, University of Tokyo Yusuf, F. M.S. (1999), Rainwater Harvesting Potential in Bangladesh, M.Engg thesis, Department of Civil Engineering, Bangladesh University of Engineering and Technology, Dhaka
Mafizur and Yusuf
6