Defining reliability for rainwater harvesting systems

Abstract

Rainwater harvesting systems such as roof-raintank water harvesting systems for internal domestic use, and artificial catchment rainfall-runoff harvesting systems connected to farm dams for agricultural purposes, have many advantages as sustainable water resources. Rainwater harvesting systems can provide adequate high quality water across significant areas of Western Australia (WA) that do not have access to the comprehensive water supply scheme. Declining rainfall trends in south-western Australia due to climate variability make rainwater harvesting systems more significant as a water resource for isolated farms and communities in arid and semi-arid areas (wheat-belt and pastoral areas) in WA. The design of catchment-dam systems, especially size of dam and catchment, will impact the efficiency and cost of construction of the system, in arid and semi-arid areas of WA and they are generally determined to satisfy the targeted demand reliability of catchment-dam systems. The reliability of catchment-dam systems can be defined as the probability that the system will supply a required demand of water during a specified time. The result of water balance simulations for the design of harvesting system can be affected by time interval for the water balance simulation such as daily, weekly or monthly calculation. The practical reliability of catchment-dam systems can be defined using available water supply demand or period with water supply failure. Determination of the modelling time interval and the definition of reliability will therefore be a critical design factor for water supplies. Evaporation from a farm dam will be affected not only by weather conditions (e.g. temperature, wind) but also by dam design characteristics such as surface area. Therefore, catchment-dam designs will return different water balance simulation results than those for roof-raintank systems, in arid and semi-arid climates. Farm dams undergo significant water loss through evaporation while roof-raintank systems (if capped) are unaffected by evaporative losses. The definition of the reliability of a rainwater harvesting system is a significant factor in determining the demand reliability that may result in the under-estimation or over-estimation of supply efficacy and reliability, particularly in dry climates. Simulation process for the reliability of rainwater harvesting system has no great difference depending on which modeling time interval and the definition of reliability are applied. However, there is the variability in the calculated reliability considerably depending on regional conditions. This paper assesses this variability in the reliability of catchment-dam systems through an evaluation of the influence of the modelling time interval for the water balance simulations; and the definition of reliability. The reliability of catchment-dam systems for ten sites located in dryland agricultural areas of WA are evaluated using daily and weekly modelling time intervals and five definitions of reliability (volume-based estimation, daily, weekly, monthly, and annual period-based estimations). Evaluation results indicate that using annual period-based estimation for reliability is not suitable for arid and semi-arid areas in WA due to the risk of under-estimation. When the cycle of agricultural activity and water demand in WA is considered, volume-based estimation, and daily and weekly period-based estimations have the risk of over-estimation. Therefore, the use of monthly period-based estimation is recommended for the design of artificial water harvesting systems in the dryland agricultural areas in south west WA.