This paper was written in the Botswana village of Romotswa in 1987, the seventh year of drought in Southern Africa. Romotswa has a history of rainwater catchment. Many of the private buildings as well as public buildings are equipped with reservoirs, most of them dating from the time before municipal water supply existed. In 1983 a large underground water source was detected here by geologists. The yield of the drilled wells is of such magnitude that for some time in 1984, the towns of Gaborone, the capital of Botswana, and Lobatse were both supplied with water from Romotswa. During my residence in this little border village in SouthEast Botswana, 30 km away from the capital, I noticed that rainwater catchment had lost momentum as a result of the centralized water supply. Many of the catchment facilities were poorly maintained so that much of the rainwater during the rare but heavy downpours was lost due to gutter leaks and/ or dislocated downpipes. At the same time newly built schools and other public buildings had been equipped with reservoirs which were much too small for the huge catchment areas. It became obvious that no calculations and no design for the systems had been made. Unprofessionally fixed gutters and downpipes were another problem. It should be mentioned that drought in Botswana, as in other parts of the world's desert belt, does not mean that no rain occurs at all. It simply means that the mean annual rainfall has not been reached and sometimes -just as serious - that the distribution of precipitation is extremely unequal. In such unfavourable climatic conditions heavy downpours of up to 100 mm in one hour do more harm than good to the arable land. The water does not penetrate the very dry and hard soil and the massive runoff results in severe soil erosion. Together with my experience gained as Town Architect in Lobatse, all these observations have influenced the structure of this paper. Originally planned as a technical manual for the construction of reinforced bricktanks, it became clear that the entire issue of rainwater catchment and storage should be put into a broader context. At the same time UNICEF, Kenya, had approached the Editor with the Laurie F. Childers manual for a ferro-cement tank. It was agreed that the accumulated knowledge in this paper should be used to provide an introduction of different techniques for building rainwater reservoirs. Since the possibilities of building rainwater reservoirs depend entirely on the availability of the required materials, it is the intention of this publication to cater for a variety of situations. N.J. Wilkinson of the Botswana Technology Centre has published a manual showing another technique of building a ferro-cement tank. This technique has advantages under certain circumstances. E.H. Robinson has erected structures for "Christian Action for Development in the Caribbean" using another technique. Each of the different structures has its advantages. Which one should be chosen depends on the situation at the specific site. Much emphasis is placed on material testing and mixing. Here the advice given by Childers matches the experience I have made with waterproof cement plaster and which has not been mentioned in any other publication. Only a careful screening of river sand and uniform mixing will result in waterproof structures. The nit is an equally important coat often forgotten, and the necessity of curing cannot be overemphasized. An article by Kiran Mukerji in "Trialog" describes his experiences in India with the construction of a rainwater reservoir. He had chosen an interesting two leaf structure combining the ferro-cement technique with air-dried clay bricks, using the latter as permanent shuttering. He stresses the need for professional artisans and limits the possibilities for self-help construction. Having worked for seven years in different African countries, I can confirm that this reflects my own experience, and leads me to comment on appropriate technology. This term is often misunderstood for different reasons. The first is that too many academics from developing countries have been educated in the western industrialized world. Because of their academic orientation, they have little practical technical knowledge. They consider appropriate technology to be a second-class technology invented by the industrialized north, good enough for the poor south. More often than not they prefer to discard traditional technical knowledge. A second reason is that some aid organizations have indirectly supported these arguments by allowing their engineers sent to developing countries to use techniques which are not appropriate to the economical, technical and social background of these countries. The windmill designed and manufactured for borehole pumps will not replace the diesel pump if it costs 20 times more than the latter as long as cheap fuel is available at all times. The plough which is manufactured from scrap has the advantages of being labour-intensive and recycling material, but its success is only assured if a comparable industrial product is not available. This means that appropriate technology must first of all be a technology which is based on a cost-benefit analysis. Secondly it is difficult to create a motivation, which is needed to achieve the acceptance of any new technique. It is therefore of fundamental importance that the manufactured product works properly. Techniques which function poorly or not at all result in frustration, demotivation and resignation, and for many years people will be reluctant to accept any change. The consequence of this must be that appropriate technology also requires professional engineering. According to my experience it is more appropriate to train and instruct a contractor and his already qualified craftsmen than to use unskilled labour in self-help construction projects without professional supervision. In construction work there are many jobs which can be done by unskilled labour, but substituting casual labour for trained plasterers should not be tried.

This manual offers advice on a more professional approach towards rainwater catchment and the construction of different types of reservoirs. It also offers a selection of the most appropriate reservoir types and gives technical advice for the construction work. As far as possible it has been kept on a level which would allow an experienced bricklayer to use the information or a building technician acting as Clerk of Works or Supervisor to advise bricklayers and plasterers on the site. It is not suitable for laymen in the construction field. Just as it is not possible to learn the technique of bricklaying by reading a book, it is not possible to write a construction manual imparting all the knowledge needed for people without the practical experience in the construction field.

Finally I wish to thank Karin Bell who did the illustrations and Siggi Gross who took the photographs.

Special thanks go to my colleague Frederick McKelton who corrected the manuscipt.

Rolf Hasse