Grantee Research Project Results

Final Report: Developing and Applying a Rooftop Rainwater Harvesting Decision Tool for Impoverished Communities in South Africa

EPA Grant Number: SU834339
Title: Developing and Applying a Rooftop Rainwater Harvesting Decision Tool for Impoverished Communities in South Africa
Investigators: Ward, Andy
Institution: The Ohio State University
EPA Project Officer: Page, Angela
Phase: I
Project Period: August 15, 2009 through August 14, 2010
Project Amount: $9,992
RFA: P3 Awards: A National Student Design Competition for Sustainability Focusing on People, Prosperity and the Planet (2009) RFA Text |  Recipients Lists
Research Category: Pollution Prevention/Sustainable Development , P3 Challenge Area - Sustainable and Healthy Communities , P3 Challenge Area - Safe and Sustainable Water Resources , P3 Awards , Sustainable and Healthy Communities

Objective:

Several logistical problems prevented the team from traveling to South Africa in the autumn of 2009. The trip is being taken, between March 13-March 31, 2010, which is simultaneous with the submission of this report. We have therefore requested and been granted a one-year no-cost extension to the project. This Executive Summary and Annual Report presents progress to date and a summary of activities for the next year. 

The primary goal of the project is to develop a Decision Support Tool that can be used to determine the economic benefits of and to correctly size a rooftop rainwater harvesting system for impoverished communities in South Africa that is in support of U.S. EPA’s Clean Water Act, Section 104.  Currently, about 20% of the world’s population is without a secure supply of potable water; this number exceeds 40% in Africa. In impoverished communities, such as those found throughout South Africa, unemployment is high, the standard of living is low, and there are numerous waterborne diseases and health concerns from standing water and fly infestations.  Often, these impoverished people do not have the resources to satisfy basic bathing and sanitation needs.  By developing this Decision Support Tool and subsequent rainwater harvesting systems that can be adapted all over the world, we hope to ameliorate some of these issues.  At the individual household and community scales a range of rooftop harvesting systems such as rain barrels and cisterns have been developed. However, rain barrels often are not sized according to the available roof surface area, collect only a small fraction of the available precipitation, and an appropriate cost-benefit analysis normally is not performed. Therefore, while rooftop rainwater harvesting systems might be considered by some as off-the-shelf technologies the Decision Support Tool we are developing is very innovative as it will: (1) evaluate design thresholds associated with alternative end uses, climatic factors, and infrastructure constraints (i.e., rooftop size and height above the ground, topography, space availability, etc.); and (2) consider socio-economic factors and government policy (i.e., security, economic benefits, health and living benefits, needs to conserve water, availability of subsidies, social acceptance, etc.).  

The project builds on a prototype tool that was developed in the 2008-2009 academic year by a undergraduate team as part of a capstone design experience. One member of that team is now a graduate student and is providing leadership to the P3 Project. The P3 project activities will include: (1) using the Decision Support Tool to design a rooftop rainwater harvesting system for a housing unit in the iEEECO™ Village located at Witsand, Atlantis, South Africa; (2) incorporate a visit to Cape Town, South Africa to consult with the Witsand iEEECO™ Village project team and other local stakeholders and to then implement the installation of the rooftop harvesting system and a vegetable garden; (3) establish a mechanism for evaluating the performance of the rooftop rainwater harvesting system; (3) have discussions with stakeholders on factors that should be incorporated in the Decision Support Tool; (4) identify local experts and databases that will be useful in refining the development of the Decision Support Tool that will include an Expert System; and (5) as an outcome of this work, and the implementation activities, enhance the Decision Support Tool. In addition, data will be obtained on a rainwater harvesting system in Columbus, Ohio. Anticipated enhancements include the inclusion of algorithms to account for deficit and excess soil water influences on crop growth and further work on currency and unit conversions to make the software more global. Efforts will also be made to evaluate water quantity and quality benefits. 

Work during the first 6 months of the project has focused on enhancing the Decision Support Tool and in organizing the visit to South Africa. We decided to restructure the prototype tool that had been developed prior to the award in order to make it more user friendly and intuitive, The tool is now divided into the following three essential sections: system, end use, and results.  In the system section the user defines the basic components of the harvesting system, regardless of what the stored rainwater will be used for.  These parameters include precipitation, temperature, roof area, gutters, and storage tank components.  In the end use section, currently only developed for vegetable garden irrigation, the used defines the growing season, garden area, crops, soil type, and system costing variables.  Finally after economic inputs such as interest rate, years of operation, and replacement water cost, the results are displayed.  In addition to economic analysis, the assessment tool displays mass balances for the system (Figure 1).

Preliminary development of the assessment tool involved reviewing literature to construct a model of the soil water balance of a vegetable garden.  The starting point was to use a soil water balance equation describing changes in storage throughout the year. Precipitation is a user defined input that should be determined from reliable sources for a certain geographic location.  If daily values are available, a representative year of data are input into the tool.  If only monthly averages are available, we have developed a synthetic distribution to randomly distribute the monthly rainfall into daily events.  These methods allowed for construction of daily precipitation tables for various locations.  Overflow is built into the tool by setting a maximum soil moisture capacity of typical saturation values.  Percolation water is assumed to be negligible in the preliminary tool.  Irrigation occurs when the soil moisture falls below the critical value for the given vegetable.  If crops are grown in soil moisture conditions for extended time below the critical value, yields drop more rapidly.  Irrigation is also a function of the storage tank water balance.  If the storage tank is empty, irrigation will not occur.  

Though many empirical and meteorological methods have been developed in the past several decades, measuring or estimating ET remains a difficult task.  This tool uses the Thornthwaite Method to generate mean monthly ET values.  The Thornthwaite method was chosen because only temperature data are needed, and the data available for the study sites are limited and we anticipate that this will be a common problem.  The formula in the water balance for ET was further enhanced using a quadratic formula derived from the results of Imtiyaz et al. (2000a and 2000b).  This formula assumes that ET varies as a function of the actual soil moisture over the field capacity soil moisture, with the maximum value being the mean monthly ET determined from the Thornthwaite Method.

Figure 1: Storage tank water balance output from the assessment tool for a system with a 35 square meter roof, 1500 liter storage tank, and 10 square meter vegetable garden of carrots and broccoli with sandy soils

Imtiyaz et al. (2000a and 2000b) conducted experiments in Botswana of responses to vegetable crops grown under varying irrigation schedules.  Irrigation schedules were based on cumulative pan evaporation.  The results from these experiments were quantified as the crop yield as a function of mean water applied over the course of the growing season.  Regression equations were then determined for each crop and each year of experimentation (Imtiyaz et al. 2000a and 2000b).  We then used these regression equations to generate curves of the percentage of maximum expected yield as a function of the fraction of water applied to water required for each crop (Figure 2).  This was then used in the assessment tool to obtain yield results for determining the benefits of a rooftop rainwater harvesting system with a vegetable garden end use. 

We used the experiments in Botswana to validate our soil water balance in the assessment tool by generating soil moisture curves over the course of the growing season.  The results showed an expected fluctuation in soil moisture for the five varying irrigation schedules and average soil water contents and yields similar to those reported ( Imtiyaz et al. 2000a and 2000b).

Figure 2: Example of a crop yield curve generated from the regression equations in Imtiyaz et al. (2000a and 2000b)

Summary/Accomplishments (Outputs/Outcomes):

The main outputs will be the Decision Support Tool and the design, installation and evaluation of a demonstration harvesting system that will be installed at the Witsand iEEECO™ Village Project Site in Atlantis, South Africa. The iEEECO™ Village project is underway and we have an agreement with the City Project Manager, The Community Organization and the Lead Consult (PEER Africa (Pty) Ltd.) to implement the project. The main outcomes will include practical knowledge on how to establish sustainable rainwater harvesting systems that enhance water resources and the quality of life for impoverished communities in South Africa and, in the long term, other regions of the world.

Installation of systems that have been sized using the Decision Tool will occur at the end of March, 2010. Future work relating to data collection as Part of the Phase I project is discussed in the next section together with possible activities that will be presented  in the Phase II proposal and Final Report that will be submitted in 2011. At all locations in South Africa, data will be collected on rainfall, water collected in the tank and water usage. Data will also be obtained on crop yields. Qualitative information will be obtained on how well the system worked and whether there was a noticeable reduction in water ponding in the vicinity of dwelling.

Conclusions:

In the next year we plan to install rooftop rainwater harvesting systems at Witsand in the Cape Province and to obtain data at that location and two other locations in KwaZulu Natal. We will also obtain data on one system in Columbus Ohio and will explore with a local watershed group the possibility of obtaining data as part of a community-wide project they have just initiated in a part of Franklin County, Ohio.

We have initiated a collaboration with Zahke Agricultural College in KwaZulu Natal and anticipate partnering with Professor Bill Ball, at Johns Hopkins University, who is already conducting projects at the College. Both Richard Dadla, the director of Zahke College, and Bill Ball participated in an NSF funded workshop on Globalizing Engineering Education that I organized in Cape Town in October 2008. Last summer we installed a drip irrigation system at the college and discussed obtaining data on a rooftop rainwater harvesting project they had just initiated. Data collection on the existing system will be undertaken as part of the Phase I Project. There is also much potential to expand that project as they have a lot of roof space which is not currently used to harvest rainwater. The college also runs a food security project in the Richmond Municipality where they are assisting communities to produce their own vegetables so that they are food secure.  They have established 75 homesteads gardens to produce vegetables on a continuous basis. However water availability in winter and also in summer during the dry period is a major challenge. They have introduce moisture trench technology as source of water that would allow production to continue even during the dry period but there are some concerns as the source might not be enough for the whole winter period. Most of the homesteads have houses with good roofs to collect rainwater which can be used for irrigation for vegetable production. During rain events runoff poses serious problems as stagnant water ponds adjacent to some homes, runoff causes severe erosion, and water entering stream systems is often polluted. 

We anticipate that data collection will also be undertaken at a school in a rural Zulu community at Potshini, in KwaZulu Natal.  In 2008, we installed a collection system and an irrigation system at the school. They have had some problems with the system but we understand that it is now operational. We will visit the site on March 22 and will ascertain if data collection is viable.

A fourth opportunity that will be considered for a Phase II study will be to size and install rooftop rainwater harvesting systems at the Muzi Thuzi Elementary School in Edendale which is located just outside Pietermaritzburg. During the past few years we have conducted several community garden projects at the school and have explored opportunities for recycling greywater. However, a major issue at the school is that the ground becomes saturated during rainfall events and there is often ponded water in the limited playground areas for the children at the school.  This water often mixes with overflows from the toilets. Polluted runoff and subsurface flow from the school then discharges into a small wetland at the edge of the school that outlets into a channelized stream system. While the wetland area is fenced, parts of the channelized system are open to anyone who wishes to use or play in the water and poses a health hazard.

References:

• Imtiyaz, M., N.P. Mgadla, B. Chepete, and S.K. Manase. 2000a. Response of six vegetable crops to irrigation schedules. Agricultural Water Management 45: 331-342.
• Imtiyaz, M., N.P. Mgadla, S.K. Manase, K. Chendo, and E.O. Mothobi. 2000b. Yield and economic return of vegetable crops under variable irrigation. Irrigation Science 19: 87-93.

Supplemental Keywords:

Economic model, water harvesting, rain barrel, developing country, water pollution