Combined sewer overflow (CSO) modeling
Combined sewer systems were designed for a different era. Built across many European and North American cities during the 19th and early 20th centuries, they carry both sanitary wastewater and stormwater runoff in a single pipe network. Under normal conditions, this works: flows reach the treatment plant and are processed before discharge. During heavy rainfall, however, the system can exceed its capacity, and when that happens, a mixture of untreated sewage and stormwater is discharged directly into receiving water bodies. That discharge event is a combined sewer overflow, or CSO.
Understanding how and why CSO events occur, and how to predict and manage them, is one of the central challenges in urban drainage engineering today. This article builds that understanding progressively, starting with the physical mechanics of CSO, moving through the principles of hydraulic modeling that make simulation possible, and arriving at the practical applications that connect modeling to regulatory compliance and infrastructure investment decisions.
What is combined sewer overflow (CSO) and how does it occur?
A combined sewer overflow is the discharge of diluted but untreated wastewater from a combined sewer system at a designated relief point, triggered when the hydraulic capacity of the network is exceeded. CSO structures, also called overflow chambers or regulators, are intentional design features, not failures. They exist to prevent sewage from backing up into streets and buildings. The problem is not the structure itself, but the frequency, volume, and environmental impact of the discharges it releases.
The trigger is almost always rainfall. When rain falls on an urban catchment, runoff drains into the combined sewer network through gullies, road drains, and surface connections. This additional flow accumulates rapidly, and if the downstream conveyance capacity, pipes, pumping stations, and the treatment works inlet, cannot absorb it, water levels in the network rise until they reach the overflow weir. At that point, flow spills over the weir and is discharged, typically to a river, estuary, or coastal water.
Several factors determine how quickly and how severely a combined sewer system reaches overflow. The key variables include:
- The imperviousness of the catchment, how much of the surface area is covered by roads, rooftops, and paved ground that generates rapid runoff
- The intensity and duration of the rainfall event, short, intense storms often cause more acute surcharging than longer, moderate rainfall
- Antecedent conditions, whether the ground is already saturated or the network is already carrying elevated dry-weather flows
- The hydraulic capacity of pipes and structures downstream of the overflow point
- The design and setting of the overflow weir itself, its crest level relative to normal operating water levels
A common misconception is that CSO events are rare or exceptional. In reality, combined sewer systems in older urban areas can overflow dozens of times per year, particularly as climate change intensifies rainfall patterns and urban development increases impervious surface coverage. This frequency is precisely what makes CSO modeling such a critical engineering discipline.
How CSO modeling works: core hydraulic principles
CSO modeling uses physics-based simulation to replicate the hydraulic behavior of a combined sewer system under varying rainfall and flow conditions. The goal is to predict when, where, and at what volume overflow events will occur, and to understand how changes to the network or catchment will affect those outcomes.
The hydraulic foundation of CSO simulation is the Saint-Venant equations, a set of partial differential equations that describe the conservation of mass and momentum in open-channel and pressurized pipe flow. In practice, most CSO models use a one-dimensional (1D) representation of the pipe network, solving these equations along each pipe link to compute water depths, velocities, and flow rates at each time step. This approach, implemented in standards-based engines such as SWMM (the US EPA’s Storm Water Management Model), has been validated extensively and is accepted by regulatory bodies worldwide.
Building on that pipe network representation, a complete CSO model also needs to account for how rainfall becomes runoff before it enters the sewer. This is handled through a hydrological sub-model that transforms a rainfall time series into a runoff hydrograph for each sub-catchment draining to the network. The key parameters here are:
- Sub-catchment area and imperviousness, which together determine how much of the rainfall contributes to runoff
- Surface routing parameters, which control how quickly runoff reaches the sewer inlet
- Infiltration parameters, which describe how much rainfall soaks into permeable surfaces rather than running off
The hydraulic and hydrological components are coupled: at each time step, the runoff computed for each sub-catchment is injected into the pipe network as an inflow, and the network model routes that flow through the system. When water levels at an overflow structure reach the weir crest, the model computes the overflow discharge using a standard weir equation, diverting flow out of the network to the receiving water body. This coupling between surface hydrology and pipe hydraulics is what makes CSO simulation genuinely predictive rather than simply descriptive.
Building a CSO model: data inputs and network representation
A CSO model is only as reliable as the data it is built from. Assembling that data is often the most time-consuming phase of a modeling project, and understanding what is needed, and why, helps engineers prioritize data collection efforts and make informed decisions about where simplification is acceptable.
Network geometry and topology
The pipe network forms the structural backbone of the model. For each pipe link, the model requires the diameter or cross-sectional dimensions, the pipe material (which determines roughness), the invert levels at each end, and the length. Manholes and inspection chambers are represented as nodes, with their invert and cover levels defining the storage volume available before surcharging occurs. Overflow structures are modeled as weirs or orifices, with their geometry and crest levels determining the threshold at which discharge begins.
In practice, this data comes from asset management systems, GIS databases, and field surveys. Data quality varies significantly across utilities, older networks may have incomplete records, inaccurate invert levels, or undocumented connections. Identifying and resolving these gaps before model build begins saves considerable effort during calibration.
Catchment delineation and rainfall input
Each inlet point in the network drains a defined area of the urban surface. Delineating these sub-catchments, assigning each parcel of land to the sewer inlet it drains to, is a spatial analysis task that typically draws on GIS data including topography, land use, and road network layout. The imperviousness of each sub-catchment is estimated from land use data or remotely sensed surface classification.
Rainfall input can take two forms depending on the modeling objective. Design event modeling uses synthetic rainfall profiles, for example, a 1-in-10-year storm of defined duration, to assess system performance against a specific return period. Continuous simulation uses observed rainfall records (often from rain gauges or radar data) over days, months, or years to build a statistical picture of CSO frequency and volume across a range of real events. Continuous simulation is increasingly preferred by regulators because it captures the full range of conditions a system experiences, including the cumulative effect of antecedent wetness.
Dry-weather flow estimation
A combined sewer carries sanitary wastewater at all times, not just during storms. Dry-weather flow (DWF) must be represented in the model because it occupies pipe capacity that would otherwise be available to absorb storm inflow. DWF is typically estimated from population data and per capita water consumption rates, adjusted for trade effluent contributions and infiltration into the sewer from groundwater. Diurnal patterns, the variation in flow through the day as domestic and commercial activity changes, are applied to produce a realistic time-varying DWF profile.
Calibrating and validating your CSO model
Model calibration is the process of adjusting model parameters so that simulated outputs match observed measurements from the real system. In CSO modeling, calibration is not a one-time exercise, it is an iterative process that builds confidence in the model’s predictive capability across a range of conditions.
The primary calibration targets in a CSO model are flow and water level. Flow monitors installed in key pipes and at overflow structures record the actual flow rates during storm events. Water level loggers at manholes and overflow chambers record surcharge depths and overflow onset times. The model is run against the same rainfall events and the simulated outputs are compared to these observed records. Where discrepancies exist, parameters are adjusted within physically defensible ranges to improve the match.
The parameters most commonly adjusted during calibration include:
- Sub-catchment imperviousness, even small changes can significantly affect peak inflow to the network
- Surface routing parameters, which control the timing and shape of the runoff hydrograph
- Pipe roughness values, which affect flow velocities and head losses through the network
- Infiltration parameters, which determine how much of the rainfall on permeable surfaces enters the sewer as slow, sustained inflow
Validation is a separate and equally important step. Once the model has been calibrated against one set of observed events, it is tested against a different set of events that were not used in calibration. A model that performs well in calibration but poorly in validation is likely overfitted, its parameters have been tuned to match specific events rather than to represent the underlying physics of the system. A validated model, by contrast, demonstrates that it can reliably predict system behavior across conditions it has not been directly calibrated against.
A practical example illustrates why this matters: a model calibrated only against moderate rainfall events may underestimate peak surcharge during extreme storms if the infiltration parameters have been set too generously. Testing the model against a high-intensity event during validation would reveal this weakness before the model is used to inform flood risk assessments or infrastructure investment decisions.
Applying CSO models to regulatory compliance and infrastructure planning
A calibrated and validated CSO model is a decision-support tool. Its value lies not in producing a single answer but in enabling engineers and planners to explore a range of scenarios, quantify risks, and evaluate the consequences of interventions before committing capital expenditure.
Regulatory compliance assessment
Regulators in many jurisdictions are tightening the standards applied to combined sewer systems. In the European Union, the Urban Wastewater Treatment Directive and its ongoing revisions require utilities to demonstrate that CSO frequency and volume meet defined environmental standards. In the United States, EPA consent decrees and long-term control plans require utilities to characterize their CSO discharges and implement programs to reduce them. CSO modeling provides the quantitative evidence base for these compliance processes, demonstrating current performance, projecting future conditions, and showing the effect of proposed control measures.
Continuous simulation is particularly valuable in this context. By running the model against a long historical rainfall record, engineers can estimate the annual frequency and volume of CSO discharges at each overflow structure and compare those estimates against regulatory thresholds. This statistical characterization of CSO behavior is more informative than a single design event result and more defensible in regulatory discussions.
Infrastructure planning and intervention assessment
CSO models support infrastructure planning by allowing engineers to test the hydraulic effect of proposed interventions before construction begins. Common interventions evaluated through CSO simulation include:
- Sewer upsizing, replacing undersized pipes to increase conveyance capacity
- Storage tank construction, offline or inline tanks that capture peak flows and release them gradually after the storm passes
- Sewer separation, converting combined sewers to separate foul and surface water systems, which eliminates CSO risk entirely but at significant cost
- Sustainable drainage systems (SuDS), green infrastructure measures such as permeable paving, green roofs, and detention basins that reduce runoff entering the combined sewer
- Real-time control, using sensors and actuated structures to optimize flow distribution across the network during storm events
By simulating each option against the same rainfall scenarios, engineers can compare their hydraulic effectiveness directly and rank them against cost, deliverability, and co-benefits such as urban greening or groundwater recharge. This scenario-based approach is far more rigorous than rule-of-thumb sizing and provides a transparent audit trail for investment decisions.
Tools like Fluidit Storm are built specifically for this kind of combined sewer overflow simulation and urban drainage analysis, enabling engineers to run these scenario comparisons efficiently within a single modeling environment. If you are evaluating how hydraulic simulation can support your CSO compliance program or capital planning process, speak with our engineering team to understand how modern CSO modeling fits your specific network and regulatory context.
