On average approximately 350 million litres of water moves through the plant every day before being distributed around the city
Melbourne Water is using artificial intelligence (AI) and machine learning in a unique approach to driving down electricity use in its Winneke water treatment plant. The project is expected to reduce pump station energy costs by around 20 per cent per year and is being tested at other locations.
The system was developed in-house by one of Melbourne Water’s experienced data analysts so it is able to understand a range of factors which are unique to the system, including reservoir level, available pumps and past performance.
Winneke is one of the major water treatment sites for Melbourne’s potable drinking water. On average approximately 350 megalitres/million litres of water moves through the plant every day before being distributed to millions of homes and business around the city.
The plant has a daily targeted flow rate for water production to ensure Melbourne has the right amount of drinking water at all times. The target is different every day, meaning different pumps running at different speeds.
To ensure these pumps are operated at maximum efficiency while still achieving the required flow rate, Melbourne Water is using the AI program which mines historical pump operational data to ‘learn’ the most efficient pump configuration at any given time. The system was developed using the Python platform.
“The Python program is able to utilise our historical data to determine the most energy efficient combinations of pumps and, the associated speeds to run them at, in order to achieve the necessary flow rate,” said the automation team leader, Russell Riding.
He added: “Because it was developed by one of our experienced data analysts, it is able to understand a range of factors which are very unique to our system, including reservoir level, available pumps and past performance”.
“The program even gives us the ability to switch to a special training mode where our operations team can test a wide range of pump combinations which may not normally be utilised so the program can learn these for future reference.
“When in operational mode, the Python program determines optimal pump calibrations and sends them directly to the pump system without any human intervention; The AI determines the best settings and then applies them in real time.”
Cyber-security was an important consideration when trialling the system: the AI is stored on a computer which is not connected to the broader Melbourne Water network, or the internet.
“The local control system also has rules built in to its code to ensure the AI system can only optimise the pump operations within set parameters,” said Riding. “This is an important failsafe feature to ensure production can continue if the AI system fails.”
If you like this, you might be interested in the following:
Partnership aims to tackle Australian water challenges
Smart water meter has a fully integrated Internet of Things-type network communications module
Itron to upgrade Barbados’ water infrastructure
Utility chose to collaborate with Itron to improve the monitoring of its water supply network
Melbourne citizens explore the future
The Australian city is hosting Knowledge Week 2018 where citizens can sample future technologies such as a heat-resistant pavement and chatbot nurses