The Development of a Computerized System for Auditing Real Time or Historical Water Use from Large Reservoirs in Order to Promote the Efficiency of Water Use
Report No. 1300/1/05
July 2005


South Africa is a water scarce country.  The water situation reports contained in the National Water Resources Strategy (NWRS, First Edition 2004) indicate that an alarming number (more than 50%) of South Africa’s water resources are considered to be either fully-allocated or over-allocated.  The statistic is concerning given the fact that few economically viable water augmentation options (e.g. the building of dams and inter-basin transfers) exist, as most viable sites have already been developed.  Given the limited potential to address the situation with a supply-side solution, focus has shifted to the improvement of water use efficiencies and water resource management.  Improved levels of water use efficiency may liberate water with little to no reductions in the production levels associated with the use of the water.  However, it has been documented that improved levels of water use efficiency may in fact reduce the water available to downstream water users as a result of reduced return flows associated with gains in water use efficiency (Grové, 1997).

The efficient operation of large dams is particularly important as dams increase the yield of catchments by retaining water during periods of water abundance for use during periods of scarcity.  The underlying objective of this research project was to assess options with which efficiencies related to the use of water could be improved, with particular reference to the efficiency of water use from large dams.   The original thinking was that water use efficiencies would be improved by the introduction of the water audit system for the following reasons:
Clearly, improved monitoring networks and a system (auditing procedures) of reconciling actual water use against the entitlements of the water users to the use the water were central to the originally envisaged water audit system.  In the course of the project however, it was discovered that a “use-it-or-lose-it” principle is generally adopted for the management of water use from large dams in South Africa. The implication of the principle is that if a given water user does not use his water use entitlement in a certain time-period (generally a year), the entitlement is lost in that time-period.  An investigation of the potential impact of this principle on water use efficiency revealed that the principle does not promote improved water use efficiency.  
In contrast, the principle may in fact induce the inefficient use of water.  The reasoning relates to the fact that different categories of water users are identified in South Africa (e.g. high, medium and low assurance users).  Irrigators, who are responsible for over 50% of South Africa’s water use (NWRS, First Edition 2004), are generally awarded lower levels of assurance than industrial and domestic water users.  Low assurance water users pay lower water use charges than users with higher water use charges.  The assurance levels influence (i) priority levels with which water users are to receive water in times of water scarcity, and (ii) the level of water restrictions faced by the respective categories of water users during periods of water restrictions.  

The inefficiency results from the fact that water users do not have incentive to use water more efficiently.  The low assurance users in particular have every incentive to use all the water they can, particularly when dams levels start dropping and the impostion of restrictions starts being discussed by water resource managers and stakeholders.  These water users will try to use as much water as they can before restrictions are imposed.  This action paradoxically leads to dam levels which draw down quicker than they would have if no restrictions were imminent.  The quick draw down in dam levels as restriction levels are neared then results in the imposition of restriction levels.  Bear in mind that there may be several dam restriction levels, with restrictions becoming more severe as the dam levels lower.   

The realisation of the potential negative impact of the use-it-or-lose-it principle of water use from dams on water use efficiency changed the course of the project to a large extent, as an alternative “use-it-or-bank-it” principle was explored, as this was believed have a number of benefits over the existing system, including:
The focus of the project thus shifted away from a purely “operational water audit system”, to an understanding of what would be required to implement the new water management system for large dams based on the “use-it-or-bank-it-principle”.  The management system based on the “use-it-or-bank-it principle” has been referred to as a Fractional Water Allocation and Capacity Sharing (FWA-CS) water management system in this document, whereas the management system based on the “use-it-or-bank it” principle is referred to as a Priority-based River and Reservoir Operating Rule (PRROR) system.  The assessment of the FWA-CS arrangement included an assessment of the National Water Resources Strategy (NWRS, First Edition 2004) and 1998 National Water Act to ensure that it would be legally possible to adopt this new water resource management approach.  It was found that nothing in the 1998 National Water Act or National Water Resources Strategy (NWRS, First Edition 2004) disallowed the adoption of the new FWA-CS water management approach.  Secondly, it was necessary to assess if the computer models currently used by the Department of Water Affairs and Forestry (DWAF) could in fact support the new water resource management approach.  It was found that the Water Resources Yield and Water Resources Planning models used by the DWAF could not in their current form support the FWA-CS water management approach.  A model was sourced which could support the FWA-CS water management report.  The model, the Mike Basin model, was developed by the Danish Hydraulic Institute.  The authors tailored the Mike Basin model to include some functionality required for South African conditions, such as the development of an In-stream Flow Requirement (IFR) module, as well as a module to derive yield curves for water resources.

The project research area
The Mhlathuze Catchment was chosen as the research project area, as it is one of the first catchments in the country in which the Compulsory Licensing process has been initiated, as the catchment is currently deemed to be over-allocated.  Furthermore, the catchment is heavily dependent on water stored in a large dam located in the upper reaches of the catchment, i.e. the Goedertrouw Dam.  The Mhlathuze Catchment was a suitable research area given the reliance on the Goedertrouw Dam, as well as the fact that the Mhlathuze Catchment is one of the first catchments to undergo the compulsory licensing process in South Africa, as a key objective of the project was to share with the stakeholders and water resource managers involved in the Compulsory Licensing process any relevant findings from this research.

Project objectives and activities undertaken

Objective 1: To develop a water audit system for large dams.

The scope of the deliverable was increased to include the development of an audit system for a catchment.  The option to use the WRYM supporting the PRROR institutional arrangement for the further development of the audit system was not pursued as the WRYM was unable to support the FWA-CS institutional arrangement, and was unable to operate on a near-real-time basis (which would be required to calculate water use entitlements during times of water scarcity).  Instead, the decision was taken to purchase the Mike Basin planning model (for approximately R40,000), as the model can support the FWA-CS as well as the PRROR institutional arrangements, and can operate on a daily, weekly or monthly time step.  The model was tailored in order to meet the requirements of water resource planners, as the IFR and Yield modules were developed to interact with the Mike Basin model via the COM interface.

Objective 2: To increase the understanding and knowledge of the practicalities, strengths and weaknesses, and potential costs and benefits of developing and implementing a “water audit system” for use by CMAs.  Although a fully operational water auditing system was not developed in the course of this project, largely as a result of the increased project scope and assessment and accommodation of the FWA-CS institutional arrangement, the requirements of this aim are discussed within this document.  Furthermore, persons in DWAF were consulted on a few occasions to assess if and how the FWA-CS institutional arrangement could be integrated into the operational management of water resources.  Stakeholders were also consulted to assess their interest in the institutional arrangement, and the benefits, strengths and weaknesses of the FWA-CS institutional arrangement.  

Objective 3: To transfer knowledge/technology to decision makers and stakeholders.
A number of presentations of the FWA-CS institutional arrangement were held with DWAF: Head Office, DWAF: Regional Office as well as stakeholders in the Mhlathuze Catchment.  The Mike Basin software has been demonstrated within DWAF.  

Objective 4: To assess the feasibility of implementing the water audit system for the management of South Africa’s water resources.
Consideration has been given to the 1998 National Water Act, NWRS (First Edition, 2004) and discussions have been held with DWAF:HO, DWAF:RO and Mhlathuze stakeholders in order to report on the feasibility of (i) introducing the FWA-CS institutional arrangement in order to promote water use efficiency from large dams, and (ii) the feasibility of developing and implementing a water audit system.  

Project outcomes

The following are notable outcomes of the project:

Capacity building

During the course of the project the capacity of a number of individuals and organisations was developed.  Details of these are provided in Appendix A of this report.

The following conclusions can be drawn from this research

The following recommendations, resulting form this project, are made: