Melbourne water model validation - identifying and reducing system uncertainty

In 2010 Melbourne Water and Opus International Consultants undertook the validation of the Melbourne Water bulk water transfer model. This paper presents a case study of the project, focusing on the development of a robust methodology for validating a large and complex system that included setting realistic expectations for the validation and development of a field test plan.

This project involved the systematic recording of network information to identify, quantify and reduce data gaps to improve the model accuracy. Key to this process was the development of a system schematic. The use of this schematic enhanced the recording of data gaps and validation results whilst viewing the network on a single sheet of paper. This enabled an appreciation of the interaction of the various parts of the network, and allowed potential solutions for issues to be identified.

The systematic development of reservoir and trunk main balances gave a thorough audit of over 300 permanent flow meters that were used for the validation. Data gaps and inconsistencies identified through these balances provided guidance to the expected level of validation possible for the model.

A field test plan was developed based on the validation results. The anomalies and uncertainties identified in the validation directed the field test planning. The field test data will be used for calibration.

This case study presents the methodology used to achieve the model validation and the development of current and future peak day scenarios, and the field test plan. The role of the schematic and novel approaches to developing simple model representations of real life operations are presented. This case study will be of interest to modellers dealing with large networks and their managers, and will show a realistic pragmatic approach to model validation, field test planning and the simulation of a large and complex network.