Real-time sewer monitoring & simulation
Sewer networks are among the most complex and consequential pieces of urban infrastructure, yet for most of their history, they have been managed with surprisingly limited visibility into what is actually happening inside them. Operators have relied on periodic inspections, manual calculations, and reactive maintenance to keep systems functioning. That approach is under growing pressure. As cities expand, rainfall patterns intensify, and regulatory scrutiny of combined sewer overflows tightens, the gap between what utilities need to know and what they can currently see is becoming harder to ignore.
This article builds a progressive understanding of real-time sewer monitoring and simulation, starting with what the technology actually is, moving through how data and hydraulic models work together, exploring where these capabilities are applied in practice, and arriving at the concept of a sewer digital twin. Each section builds on the one before, so by the end you will have a clear picture of both the technical foundations and the operational value of this approach.
What is real-time sewer monitoring?
Real-time sewer monitoring is the continuous measurement of conditions inside a sewer network using physical sensors, with data transmitted automatically to a central system as events unfold. Rather than capturing a snapshot of system state during a scheduled inspection, real-time monitoring provides a continuous stream of information that reflects what the network is doing right now.
The sensors used in urban drainage monitoring typically measure flow velocity and depth, water level, rainfall intensity, and water quality indicators such as conductivity or turbidity. These instruments are installed at strategically selected points across the network, at key junctions, upstream of overflow structures, at pump stations, and within combined sewer systems where overflow risk is highest. The data they produce is transmitted via telemetry, often using cellular or radio networks, to a supervisory control and data acquisition (SCADA) system or a dedicated data management platform.
It is worth distinguishing real-time monitoring from traditional condition assessment. A CCTV inspection or a manual flow survey produces valuable information, but it describes the network at a single point in time. Real-time monitoring, by contrast, tracks how the network responds dynamically to changing inputs, particularly rainfall events, which can transform a system’s hydraulic state within minutes. For utilities managing combined sewer systems, where the risk of combined sewer overflow (CSO) escalates rapidly during storm events, this continuous visibility is operationally critical.
How real-time data and simulation work together
Sensor data alone tells you what is happening at specific measurement points. Simulation tells you why it is happening and what is likely to happen next. The real power of real-time sewer monitoring emerges when live data is connected to a physics-based hydraulic model of the network, a combination that transforms isolated measurements into a system-wide understanding of network behavior.
The role of hydraulic simulation
A hydraulic model of a sewer network uses the physical laws governing fluid flow, conservation of mass, momentum, and energy, to calculate how water moves through pipes, manholes, storage structures, and overflow points under any given set of conditions. Sewer simulation platforms built on the SWMM (Storm Water Management Model) standard, which is maintained by the US Environmental Protection Agency and approved by governments worldwide, provide the computational foundation for this kind of analysis. The model represents the network’s geometry, pipe characteristics, and boundary conditions, and can simulate how the system responds to a rainfall event or a change in inflow.
Connecting live data to the model
When real-time sensor data is fed into a hydraulic model, the model can be continuously updated to reflect current network conditions. This process, often called model state estimation or data assimilation, adjusts the model’s boundary conditions and initial states to match what the sensors are reporting. The result is a simulation that is not just theoretically accurate but empirically grounded in what the network is actually doing at that moment.
For example, if a flow meter at a trunk sewer junction reports an unexpected surge in depth during a rainfall event, a connected simulation can immediately calculate how that surge propagates downstream, which overflow structures are approaching their activation thresholds, and how much storage capacity remains in the system. This is information that no sensor network alone can provide, because sensors only see what is happening at their specific locations, the simulation fills in the picture across the entire network.
Predictive capability
Connecting real-time data to simulation also enables forward-looking analysis. By combining live network state data with rainfall forecast inputs, operators can run near-future simulations that project system behavior over the coming hours. This predictive capability is what separates real-time hydraulic modeling from simple monitoring: instead of reacting to events after they occur, operators can anticipate them and take preventive action before thresholds are breached.
Key applications in smart city sewer management
Building on the understanding that real-time data and simulation together create a continuously updated, predictive picture of sewer network behavior, it becomes clear why this combination is central to smart city sewer management. The applications span operational decision-making, regulatory compliance, and long-term planning.
Combined sewer overflow monitoring and prediction
Combined sewer overflow monitoring is one of the most immediate and consequential applications. CSO events occur when combined sewer systems, carrying both wastewater and stormwater, exceed their capacity during rainfall, causing untreated sewage to discharge into receiving water bodies. Regulators in many countries require utilities to monitor, report, and progressively reduce CSO frequency and volume.
Real-time simulation allows operators to move beyond passive recording of overflow events toward active CSO prediction. By tracking system state continuously and running short-horizon simulations during rainfall events, utilities can identify which overflow structures are most likely to activate, estimate the volume and duration of potential discharges, and take operational steps, such as adjusting pump rates or activating storage capacity, to reduce or prevent the overflow. HSY, the Helsinki Metropolitan Area water utility in Finland, has used this approach to move from manual CSO calculations to automated, simulation-driven insights across their combined sewer network.
Operational control optimization
Real-time sewer simulation also supports the optimization of pump stations and control structures within the network. Pumping schedules that were set based on average conditions may perform poorly during extreme events or periods of unusual inflow. With a live hydraulic model, operators can simulate the effect of a proposed pump setting change before implementing it in the real network, reducing the risk of unintended consequences such as surcharging upstream or creating pressure transients that stress aging infrastructure.
Early warning and incident response
Automated alerts generated by real-time monitoring systems can flag anomalies, unexpected level rises, flow reversals, or sensor readings that deviate from model predictions, that may indicate blockages, structural failures, or illegal discharges. When these alerts are connected to a simulation environment, operators can quickly assess the likely cause and spatial extent of the anomaly, prioritizing field response more effectively than a simple threshold alarm system would allow.
Building a sewer digital twin with live data
A sewer digital twin is the most complete expression of the concepts covered in this article. It combines a physics-based hydraulic model of the sewer network with continuous real-time data integration, creating a living digital representation of the physical system that evolves as the network itself evolves.
The key distinction between a well-calibrated static model and a digital twin is the live data connection. A static model is built, calibrated against historical observations, and then used for scenario analysis and planning, but it does not automatically update when conditions in the real network change. A digital twin, by contrast, is continuously synchronized with sensor data from the physical system. Its state at any given moment reflects the actual state of the network, not a historical snapshot or a design assumption.
What a sewer digital twin enables
A fully operational sewer digital twin supports several capabilities that a static model cannot:
- Continuous state awareness: operators always have an accurate, model-based picture of system conditions across the entire network, not just at sensor locations
- Predictive simulation: near-future forecasts based on current network state and rainfall predictions, enabling proactive operational decisions
- Scenario testing: proposed operational changes or infrastructure interventions can be tested against the live model before being implemented in the real network
- Automated CSO monitoring: overflow events are detected, quantified, and reported automatically, reducing manual effort and improving regulatory compliance documentation
- Long-term performance tracking: the digital twin accumulates a continuous record of system behavior, providing the data foundation for capacity assessments and climate resilience planning
The path from static model to digital twin
Building a sewer digital twin is not a single step, it is a progression. Most utilities begin with a calibrated static hydraulic model, often built on SWMM. The next step is establishing data integrations that connect sensor feeds and SCADA systems to the model environment. As those integrations mature and the model is continuously validated against real observations, the system transitions from a planning tool into an operational digital twin.
Platforms like Fluidit Storm are designed to support this progression, enabling utilities to build and maintain physics-based sewer models that can be connected to real-time data sources as their monitoring infrastructure develops. The goal is not to replace the hydraulic engineering expertise that underpins a good model, but to give that expertise a continuously updated, data-rich environment in which to operate.
For utilities and infrastructure operators ready to explore what real-time sewer simulation looks like in practice for their network, get in touch with our engineering team to discuss your specific context and where to start.
