The water industry has talked a great deal about moving away from the ‘sawtooth’ model to leakage that is driven by a reactive approach to Active Leakage Control (ALC), with the help of permanent monitoring. But why is this limited to acoustic monitoring? The industry must also shift away from the sawtooth approach to network hydraulic models, which are used to measure, understand and manage network and asset performance.
How often does your utility update hydraulic network models?
The general number is every 5 years. How useful is that static model that doesn’t adapt as your network changes? In reality, model calibration often happens far less frequently. Outdated models, whilst better than nothing, are not as useful as they could be. They tell us what ‘used to’ happen in the water network, not what ‘is’ happening in the network.
So, how soon does a static model become outdated and need some TLC?
The answer is… whenever something changes. Valves are operated more than you are aware and probably more than you’d like! They are also often not put back the way you modelled them. Pressure control assets are sometimes changed without updating the model and control assets can start to behave unexpectedly when they require maintenance.
Over a 5 year period, these and many other variables change; new developments, large users coming and going, commissioning or decommissioning control assets, pipe replacement (if you’re not in the UK), fixing long-running leaks and more. These all affect the accuracy of your network model and your ability to make accurate and timely tactical and strategic decisions
As a modeller, being asked questions that require checking and updating the model can be a time consuming and manual task to figure out what’s changed and to make the updates needed.
Wouldn’t using models to make active decisions, plan and make changes to networks be more efficient and effective if the models were calibrated and maintained for when you need them? Let us know your thoughts.
What can be used to look after your model?
With this in mind, Inflowmatix have developed a data agnostic network modelling ‘service’ that automates and maintains the calibration of network models in near real-time, reducing cost and time, whilst making outcomes a deterministic process, moving away from the trial-and-error approach to calibration. The service maintains the accuracy of network models using live data and lets you know as changes arise over time, so what’s actually going on in your network based on the data. The ‘sawtooth’ effect becomes a smooth line, with consistently useful network models.
Furthermore, arboricity™ provides a richer insight into your network operation through data-driven algorithms and mathematical solvers to identify anomalous network behaviour such as leaks, bursts, throttled or closed valves in order to highlight changes or problems for faster resolution by network operations and field teams. When you’re confident of how your network is operating from the near real-time data, arboricity™ can update the model using the latest pressure, flow and elevation data points from sensors in the network.
With arboricity™, you can look after your model and your model will look after you!