Predict Performance to Keep Water Flowing Reliably

How do you operate a water network effectively and efficiently? You must maintain water pressure, reduce water loss, respond quickly to incidents, and maintain safe water quality – all within budget, and under the threat of fines if you fail.

The Challenge

Understanding What’s Happening and About to Happen

To achieve this, you need to know what’s going on, and what will likely happen next, across your networks – in some detail.

Of course, demand and network conditions change continuously. Operators need live information that they can act on.

Arial view of a city
dripping outside faucet

One option would be to install lots of sensors, but these are costly to buy and maintain (with the need for frequent battery replacement), and they’re only as useful as the operators’ ability to understand what the data is telling them, in a timely way.

The alternative is live modeling. It requires a hydraulic model, data feeds of the boundary conditions, and live modeling software.


Real-time modeling allows us to quickly figure out if we’re looking at routine operations, stressed system conditions, or other operational anomalies.

Water Quality ManagerCity of Boulder, Colorado

The Solution

  • The Power of Predicting Accurate Consumption Data

    Water network operators know the usage at district metered area (DMA) level (water out vs water in, according to the meters at the entry and exit points of the DMA). But this doesn’t provide detailed enough information to run the network effectively. The operator needs to know about usage within it, ideally at the level of individual properties. This would be easy if every property had a smart meter, reporting consumption every 15 minutes. But few network operators have the luxury of that data, and so need some demand prediction tools to aid their modeling. (Get in touch if you’d like more detail on this.)

    Screenshot showing water supply network software
  • Empowered to Optimize the Network

    Now you have a live hydraulic model of the network that will predict its behavior according to the expected demand. The model can bring in a prediction of demand (such as with DemandWatch Pro) and live telemetry/SCADA data, such as how full a reservoir is, to set the initial conditions.

    It can also predict what’s likely to happen in the gaps between sensors, giving a fuller picture of the network without further investment in field monitoring equipment.

    This means that water utilities are now empowered to optimize operations.

    Arial view of a town
  • Be Warned

    Live modeling can generate two types of warning: simulation warnings and verification warnings.

    Simulation Warnings

    Take a simple situation where there are a pump and reservoir that supply a DMA. The network operator would normally top up the reservoir at night when it’s more cost-effective to run the pump thanks to cheaper overnight electricity. The model says that on some mornings – like high-use holidays – the reservoir alone won’t meet demand. Then the operator can adjust the pump schedule for the day, to keep the pump running during the morning. Customers will experience their normal water pressure. Live modeling means the operator can anticipate demand and be ready to meet it.

    Verification Warnings

    If the network operator uses live telemetry to verify the model, and the two disagree, the operator can investigate. There could be a burst, or an issue with a pump, for example. Live modeling means the operator gets a warning (beyond typical SCADA), and data that they can compare with the ‘normal’ situation in the model, to get them started with identifying, diagnosing, and fixing any problems.

    Corroded water pipe
  • Fixing Issues

    If a warning shows that there could be a problem and a field crew needs to investigate, they’ll need a work order (WO) to work from. The WO might tell them that a pump needs to be isolated by closing a valve or a crew might need to be dispatched to restart a pump.

    One efficient way to do this is to have the live model connected to an asset management program.

    Asset management software collects together, and runs analytics on, all the information about pipes, pumps, valves, bearings, and other utility assets, enabling effective daily management as well as prioritized capital improvement planning.

    Using a program such as InfoAsset to set up the WO is highly efficient. It can provide the relevant infrastructure data and has WO generation capabilities built in. InfoAsset lets crews access WOs and file their reports from mobile devices at the work site. The work they carry out gets recorded in the system immediately and with minimum effort. If the network operator is also using IWLive Pro for their live model (with either InfoWorks WS Pro, InfoWater or InfoWater Pro as their static model), the updates in InfoAsset feed directly into it – updating both the asset records and the model, immediately and effortlessly.

  • Incident Management

    When something unexpected happens, live modeling will also help manage the incident. Operators can model 'what if' scenarios, such as taking a pump offline for maintenance, or closing a valve to isolate a burst pipe, and seeing the impact on different parts of the network. They can choose the best course of action, manage customers’ expectations, and give them the greatest possible time to plan for any disruption.

    Water pipes in the ground
  • Energy Savings - Without Lifting a Finger

    IWLive can recommend energy savings in the form of scheduling pump(s) to run off-peak tariff hours to minimize electricity costs. The operator can switch the pump remotely or create a Work Order to dispatch a crew to the pumping station. Then, once the pump status has changed, the telemetry reports back to IWLive Pro – which automatically updates the model.

When you evaluate the decision for your organization, it’s worth comparing the costs of:

  • Sticking with a static model only used for planning 
  • Lack of confidence in results when using an outdated hydraulic model when performing utility-critical functions like event management
  • Installing more sensors and making the most of the data they generate
  • Traditional model calibration – with live modeling, you get live calibration too; there’s no need to rely on an uncalibrated or out-of-date model as the basis for decision-making

If you’d like to talk over the options, get in touch.

Discover more about the software