top of page




Inflow and infiltration (“I&I”) is a common problem in sewer

systems.  Inflow occurs when rain or snow melt enters the

sewer system, typically from manholes or illegal connections

such as storm drains or roof drains. Inflow typically occurs rapidly

and produces a level/flow signature that can be correlated to

external rain events, for example. Infiltration is the entrance of

ground water into a sewer system through cracks or gaps in pipes or manholes. The function of an I&I detection system is to (a) locate the sources of I&I; (b) quantify the effects of I&I in order to prioritise remediation based on relative amounts of I&I; and (c) quantify the results of the repairs.


An I&I monitoring and detection system typically consists of the following components:


  • Sensors

  • Communications

  • Data analysis

  • Recommended actions and repair evaluation


The first step in implementing an effective I&I system is locating the sensors. Sensors should be located in positions in the collection system that can distinguish one branch of the collection system from another, enabling the comparison of I&I from one branch to another. Sensors should also be easily relocated, to enable isolation of smaller and smaller parts of the collection system to narrow down I&I sources, as a complete simultaneous system monitoring network is not generally feasible. Select portions of the system can be densely monitored.


Sensors may be flow sensors or water level sensors. Flow sensors (e.g. area velocity meters) can directly measure flow at a given location, and can directly compare wet versus dry weather flows. The disadvantages for flow sensors are that they tend to be expensive, require significant maintenance and are generally deployed only for short periods during which there may or may not be rain events typical for the region. Level sensors can be used to qualitatively compare wet versus dry flows based on level, or otherwise be linked to hydraulic models that can quantitatively compare wet versus dry weather flows. Level sensors tend to be lower cost than flow sensors, have lower maintenance, and because of the lower costs, can be installed permanently and more pervasively for broader and longer term I&I monitoring. The disadvantage of level sensors is that flows are not directly measured but must be inferred through other means, such as Manning’s equation or hydraulic models.


The other sensors required for I&I monitoring are rain gauges. Physical rain gauges are discussed in other sections of the SWAN tool, but alternatives are now used based on radar data with ground truth calibration. Physical rain gauges, such as tipping bucket rain gauges (“TBR”), are considered the “gold standard” for rain measurements, but they also suffer from high maintenance. Care must be taken to properly locate TBR to avoid shadowing effects. Because a tightly packed array of TBR is expensive to deploy and operate, radar based rain data can be an effective alternative to understanding the input to an I&I monitoring system. Radar rain gauges (“RRG”) use no labour, are moved by the click of a mouse, and can be placed in high densities. RRG can have low measurement bias when rain intensities are high, but can be calibrated by local TBR or other physical gauges to correct for this bias.


Communications to and from the flow/level sensors and the physical rain gauges can take place via a variety of means:


  • Wired communication

  • Local wireless, such as Bluetooth, WLAN (IEEE 802.11)

  • Wide area wireless, such as private radio networks

  • Commercial wireless, such as cell phone providers or satellite data providers


Two-way communication is generally preferred, as it allows control of remote sensors and remote operational parameter modification.


Data analysis can take many forms for I&I, but generally includes the integration of two or more of the following components:


  • Rain/runoff data

  • Hydraulic models

  • Real-time and historical level/flow data

  • Ground water levels

  • Local stream or tidal data

  • Risk analysis for spills


The output of data analysis using the input streams above are: a prioritised list of repairs and the financial impact of these repairs. Based on the severity of the requirements and costs, the utility can make decisions such as: make the repairs now, make the repairs later, or continue to monitor the locations to avoid any catastrophic event (such as a pipe collapse). Even after the repairs are made, monitors should be left in place to quantify the benefit of the repair and provide feedback for future repairs.





Two fundamental goals of a sewer system operator include:


  1. Keeping the sewage collected from entering the environment without sufficient treatment;

  2. Keeping the environment out of the sewer.


I&I detection and monitoring exactly keeps the environment out of the sewer.  While there are times when this may be beneficial – for example capturing the initial runoff of a storm event into a sewer is often better than having this runoff end up uncontrolled into streams and rivers – for the most part, not capturing clean storm water and runoff has two significant benefits. First, clean water is being treated at the treatment plant as sewage, and this is expensive, approximately $2 - $3/kGal, so this cost is avoided if this water does not enter the collection system, and second, unintended water entering the collection system invariably leads to sewer overflows, which violates #1 above.





The installation and operation of an I&I detection and monitoring system requires the following elements:


  1. A plan of deployment of sensors, based on the budget available and the sensors chosen for the system;

  2. The physical deployment and calibration of the sensors;

  3. Testing and verification of the communication system;

  4. Verification that data received is reliable, repeatable, and robust.

bottom of page