Undergraduate Thesis - Feasibility of Rainwater Harvesting in Paarden Island

Assumptions

Assumptions that were made concerning the roof surface in Paarden Island included:

• It was assumed that the GIS Data that was given was as accurate as was possible. The cadastral data that was received was dated August 2011 and since that time there have been a few developments that have occurred which have modified the layout of the buildings. Thus for this thesis the volume of potential rainwater that could be collected was thus estimated for August 2011;

• It was assumed that all roof surfaces that were delineated where able to collect usable rainwater. This is not entirely true however since it was observed that there were a lot of roof surfaces that were made from asbestos which is not a suitable roof material. It was also observed that most of the roof surface in the area were very badly maintained and starting to fall apart. The growth of mould was also observed on many of the roof surfaces which would result in rainwater having a high volume of biological contaminants as seen in figure 5-1. Thus for the calculation of rainwater volume from roof surfaces, the quality of the roof surfaces were not considered;

• It was assumed that all roofs in Paarden Island had gutters and down pipes install so that all the rainwater could be collected. This was not entirely true since some of the buildings had roofs that were designed without gutters and down pipes or the gutters had been broken and were no longer usable as seen in figure 5-2;

• It was assumed that all roof surfaces that were less than 50m2 would not contribute to the rainwater collection. This would be partly true since the additional cost of connecting a roof smaller than 50 m2 could incur extra costs especially if the structure was located more than 40 m from the main structure as would be the case with guard houses at the gate entrances;

• It was assumed that erven that did not have a registered paying municipal water account but did share a roof with a erf that did have a registered paying municipal account would all collect the same rainwater and then share this amount;

• It was also assumed that multiple business share a single erf would also share the rainwater that was collected from the roof surfaces;

• It was assumed that all the rainwater that comes from the roof surfaces will be used. This assumption is valid for small roof surfaces in which the rainwater that has accumulated is less than 50kL. Any volumes of rainwater above 50kL may not be collected due to the cost and size of the rainwater tanks that would be required. These restriction could result in the loss of rainwater;

• It was assumed that the amount of rainwater that could be collected within a month was collected over the whole month and not just on the few days that it rained. The number of days in which rained occurred during certain months varied from between 1 to 10 days; and

• It was assumed that the amount of rainwater that was collected for each month would be a reliable source of freshwater. This assumption would not be accurate since annual rainfall varied considerably from year to year and that some years received above average rainfall and other below average rainfall.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 5-1: Organic matter growth that covers most roof surfaces in Paarden Island

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 5-2: Broken down pipes and gutters due to lack of repair

 

Delineation of roofs in Paarden Island

 

 

Figure 5-3: Roof surfaces in Paarden Island

 

Figure 5-3 shows the delineation of roof surfaces in Paarden Island. The delineation of the roof surfaces were done in ArcGIS using the cadastral GIS data that was sourced from the UCT GIS Department. The area of each roof surface was then calculated using the ArcGIS geometry calculator.  All roof surfaces less than 50m² were then removed from the data set. Each roof surface was then assigned to an ERF number which corresponded to a registered paying municipal water account.

 

Estimation of roof surface runoff coefficient

The roof surface runoff was estimated to be 0.8 based on the recommendations in table 2-1. Most of the new roof surfaces in Paarden Island are made from galvanised zinc sheet metal that has a slope of about 10° to 30°. The sloped roofs help rainwater to runoff the roof surfaces and not collect rainwater on the roof surface which will provide a higher efficiency of runoff. Since most of the other roof surface types that are being used are either inappropriate or aging and require some form of replacement it is recommended that these roof surfaces are replaced with galvanised zinc sheet metal roofing. The median rainfall that has been recorded is also above 5mm and if the first flush diverter uses the first 1mm then a coefficient of 0.8 would be required.

 

Calculation of potential rainwater that can be harvested from roof surfaces

The rainwater that can be harvested is calculated using equation 3.1. The value of the runoff coefficient for all erven was taken to be 0.8. The value of the roof surface area was calculated in ArcGIS and the monthly average rainfall was calculated in Appendix A. The result of these calculations can be seen in Appendix B table B-1.

A total of 254 different erven were considered that could be assigned to a paying municipal water account. The total roof area in Paarden Island was estimated at 528 000 m² and would be able to provide up to an estimated 198 000 kL of rainwater supply per annum. The estimated rainwater supply would not necessarily all be captured due to wastage from insufficient storage capacity, which would be a result of available finance or yard and factory floor space. The months between October and March that receive low monthly rainfall could result in inefficiencies that are greater than the runoff coefficient that is used. The inefficiencies would be due to smaller rainfall events occurring.

Based on the standard deviation of annual rainfall in Cape Town in Appendix A the yearly collected amount of rainwater can vary by as much as 28% from the average. The estimated collectable rainwater per annum varies from 142 500 kL to 253 000 kL. If all the rainwater could be collected and the runoff coefficient is correct then the 198 000 kL rainwater supply represents a massive source of freshwater for the area which currently only uses an estimated 490 000kL of municipal water supply. The roof surface rainwater supply could on average account for almost 40% of the required freshwater supply.