Hong Kong Water Supplies Department (WSD) commissioned the Centre for Water Systems in July 2010 to complete a study titled "Investigation on Development and Implementation of Pump Scheduling Optimisation Methodology and Real-time Optimal Pumping and System Control" to demonstrate the benefits of optimised pump scheduling to the operation of their water supply systems, to develop the methodologies and software elements that could be integrated with Department’s IT and SCADA systems and implemented in real time.

The work builds on the workshop delivered by Professors Dragan Savić and Zoran Kapelan during their visit to WSD in May 2010.

The objectives of the study included:

1. Complete an off-line investigation of the benefits of optimised pump/valve scheduling for two pilot systems.
2. Develop a methodology for both off-line and on-line optimisation of pump/valve schedules.
3. Transfer the knowledge and technology and provide advisory services on conducting the investigation to complete the other 8 water supply systems.

AQUATOR® is a commercial software for developing and running simulation models of natural rivers, water resources and water supply systems, using different operational rules, constraints and priorities. Developed by Oxford Scientific Software, it is being used by several water companies in the UK. Dragan has led the project aimed at linking AQUATOR to a Multiobjective Genetic Algorithms optimisation module.

Initially GANetXL, an add-in for Microsoft Excel®, developed by Dragan and colleagues was linked to AQUATOR, for the optimization of reservoir operation. However due to the excessive computational time required, AQUATOR-GA, a new GA application was developed, using distributed computing, which has been integrated within the AQUATOR environment.

The new tool has successfully been applied to two case studies in United Utilities and several others including Scottish Water and Environment Agency (South West Region).

More information can be fouND here.

The water resources planning process in the UK requires water companies to carry out an appraisal of options to determine the least cost solution to implement in order to maintain their respective supply-demand balances.

The Environment Agency’s Water Resources Planning Guideline (November 2008), applicable in England and Wales, requires that companies consider the financial, environmental and social costs of proposed schemes, as well as wider issues of risk, uncertainty, sustainability, climate change, energy and environmental concerns. Currently the least-cost modelling approach is adopted, approximated by calculating the Average Incremental Social Cost (AISC) of each option and comparing these, usually through an iterative process.

Dragan has been working with the UK consultant ARUP to develop and apply a modelling tool to optimise water resources planning process according to a defined set of agreed sustainability criteria and use of multiobjective optimisation methods to provide support for the decision-making process by providing information on the potential solutions and trade-offs.

Data-Driven modelling (DDM), which borrows heavily from Artificial Intelligence techniques, aims to develop mathematical models from observed data. These models are particularly useful when it is not possible or is too difficult to develop physically-based models (e.g., hydrodinamic flow models, etc). Prof. Savic has pioneered the use of DDM in water applications. His work has evolved from studies of genetic programming and artificial neural networks to the development of Evolutionary Polynomial Regression (EPR), a novel DDM tools that has been used extensively in the UK to model deterioration of water distribution and sewer pipes. The recent clients include Anglian Water, South West Water and Wessex Water.

A Genetic Algorithm (GA) is a search/optimization technique capable of solving many problems that have been unsolvable by "classical" optimization techniques. Prof. Savic's research on GAs has resulted in a number of tools that have been and still are used by the UK Water Industry and their consultants. For example, he has initiated the development of a GA tool that is an add-on to Excel. GAnetXL can optimize models developed in Excel or any other model that can be linked to an Excel spreadsheet. This single and multipleobjective optimization tool is available freely for research purposes, but has also been used on a number of consulting projects. GAnetXL can be downloaded here.

At an early stage Mouchel (formerly Ewan Associated Ltd, EAL) recognised the potential of the novel optimal design methods developed by Prof. Savic and colleagues for applying this leading edge technology within their consultancy practice. For example, together with EAL and United Utilities as partners, his methodology was used to plan a water supply scheme in the York Region (Canada), saving $54million or 35% of the expected costs. A similar study for Essex and Suffolk Water demonstrated a 49% saving on a network strengthening project.

As a result of an EPSRC funded research project led by Prof. Savic titled “Whole Life Cost Management of Water Distribution Systems”, which was done in collaboration with the University of Sheffield and six water companies, a spin-off consultancy company, SEAMS Ltd., was founded in 2002.