Melbourne Water achieves sustainability goals with MPC

Melbourne Water manages the major water, sewerage and drainage networks across Greater Melbourne, and cares for 25,000 km of rivers and creeks that sustain healthy communities and the environment.

Melbourne Water’s largest water reservoir, the Thomson Reservoir, makes up about 60% of the city’s total water storage capacity, and feeds into one of the largest water supply networks in Australia. It is at the core of water supply for the city of Melbourne and the surrounding suburbs.

To ensure the Thomson River environment downstream of the dam is protected, the flows in the river must be controlled and optimised so that the correct amount of water is released at all times.

Thomson Reservoir is filled by rainfall received by the surrounding catchment. It can transfer over 1 billion litres of water per day to Upper Yarra Reservoir, which in turn supplies Silvan Reservoir. From Silvan Reservoir, water can be distributed throughout metropolitan Melbourne.

Finding the optimal amount of water to release is complex. If too much water is released, the source water is wasted. If too little is released, the environmental conditions downstream of the dam are impacted and Melbourne Water fails to meet its minimum compliance requirements.

Melbourne Water must take into consideration the existing river flow and levels, which are impacted by crop irrigation, rainfall runoff, tributary flow and evaporation. It must also consider fluctuating customer demand. Normally there is a time lag of up to 10 hours from when changes are made to release flows at the dam to when those changes are seen at the furthest compliance point.

The complexity of the network, paired with factors like lag time and multiple environmental variables, made this system a perfect candidate for model predictive control.

Digitalisation consultant Kalypso — a Rockwell Automation business — worked with Melbourne Water to help develop ways to control the water leaving the Thomson Reservoir with increased efficiency and precision by implementing FactoryTalk Analytics Pavilion8 model predictive control technology from Rockwell Automation.

While the water industry is relatively mature when it comes to automation and system optimisation, quality, process variability, outages and downtime are still persistent challenges. The complexity of calculations, the duration of focus, and the speed and precision required are poorly suited to human capabilities. MPC applies a data science layer on top of existing regulatory controls to continuously monitor and predictively optimise process behaviour.

FactoryTalk Analytics Pavilion8 Model Predictive Control (MPC) provides an intelligence layer on top of automation systems that can continuously optimise plant operations in real time. Its flexible hybrid modelling capabilities incorporate available process knowledge to deliver accurate, high-fidelity models, and can handle both nonlinear and linear processes simultaneously.

MPC technology continuously assesses current and predicted operational data, compares them to desired results, and adjusts control targets to reduce process variability, operate within equipment constraints and improve performance.

For Melbourne Water, the MPC system analyses a set of variables (flows from tributaries, rainfall, evaporation, irrigation consumption, etc) in real time and continuously, predicting and driving the set-point adjustments to maintain optimal system function, removing the strain of extended observation from conventional operation and eliminating the root cause of variability, outages and downtime.

Using the new system, Melbourne Water was able to save 2 billion litres of water annually, conserve energy and reduce strain on the workforce while improving compliance. Melbourne Water has thus set the bar for autonomous operations and responsible stewardship over a critical natural resource.

Image: Thomson Reservoir. Image courtesy Melburnian (via Wikimedia Commons) under CC BY 3.0