November 2019: Project Update

Our team is excited to share an update along with some relevant data findings that may be of interest for all those participating in the project. We think this will help further explain our workplan tasks to-date, while also emphasizing the overall objectives of our project.

Please see the sections below that include a partial summary of the data findings.


The RainGrid cistern controller captures and analyzes predictive precipitation data using a 5-day forecast. It can utilize this data by communicating storage capacity data back to the server to automatically empty in anticipation of upcoming precipitation events. The homeowner is capable of viewing data in real-time through a cloud platform to monitor the cistern analytics; such as diverted water and storage capacity.

This is demonstrated using water level data from an installed RainGrid cistern for two rain events that occurred the week of May 24th, 2019: May 25th and May 28th (see figure below). On May 24th, the cistern was partially full from a previous rain event. Using predictive weather data, the system sensed an incoming event and emptied to ensure full capacity in order to capture the rain event on May 25th. A small rain event later occurred, partially filling the cistern again, where it remained for the next couple of days. In the evening of the 27th, the cistern emptied again in anticipation of a large incoming event on the 28th. The cistern then filled quickly during a large storm event on the 28th, where it remained full for multiple days, as the water did not manually empty and no further rain events occurred through the rest of the week. This harvested rain water was then available for homeowner use until it emptied prior to the next predicted rain event.

Example of RainGrid Controller Automatically Emptying Cistern Before Rainfall Events

One of the primary objectives of the project was to demonstrate the ‘smart’ stormwater control capabilities of the RainGrid cisterns with conventional Low Impact Development (LID) practices. This enabled capturing at least 50 percent of all 90th percentile storm events. The 90th percentile storm is defined by the rainfall depth that is equal to or greater than 90% of all rainfall events. In Collingwood, this event is defined as a 30mm rainfall depth. For modelling purposes, the duration of the storm event was set to four (4) hours.

This reduction has been achieved with the RainGrid cistern through harvesting of roof run-off. The tabulated results below using a predictive modelling tool, demonstrates the effectiveness of rainfall reduction for an existing (medium density) residential property that had a RainGrid cistern installed. Rainfall reduction was based on multiple parameters such as: land use, soil type, storage capacity etc.

Water Balance Comparison (Pre to Post Analysis of a Medium Density Residential Property with a RainGrid Cistern)

Prior to installation of the RainGrid cistern (Pre Total) 24.7m3 (approximately 70%) of rainfall was running off the property into the storm sewer system throughout the duration of the storm. After the RainGrid cistern installation (Post Total), the runoff was reduced to 17.2 m3 (approximately 49% of total rainfall) for the same storm event – a reduction of more than 30% of runoff. Total rainfall reduction (runoff reduction + infiltration) increased from 30% to 51% with the introduction of a RainGrid cistern. Therefore, from an individual residential lot perspective, reduction of the 90th percentile storm runoff volume was achieved. It should also be noted that if an additional RainGrid cistern was installed to capture the balance of the roof area, even less runoff would leave the site and exceed the required metric related to the 90th percentile storm.

Similar results were observed for low density residential sites that had an installed RainGrid cistern. Infiltration for low density lots is much greater, due to an increase in pervious area when compared to the medium density property. Findings from an example site with an installed RainGrid cistern is shown below. Initial rainfall reduction was already greater than 50% for the 90th percentile storm; however, with the addition of a RainGrid cistern nearly 75% of rainfall is being retained on site, a relative increase of more than 30% of rainfall reduction.

Water Balance Comparison (Pre to Post Analysis of a Low-Density Residential Property with a RainGrid Cistern)

The capabilities of the RainGrid cistern in combination with other LID features were also monitored for potential to capture 100% of rainfall during the 90th percentile storm. Permeable parking lot systems were installed at three (3) commercial/institutional lots in the Town of Collingwood, and monitored as part of the project. At two (2) of the properties, a RainGrid cistern was also installed to create a “treatment train” with these other LID features. One of the properties included rain gardens as well.

The RainGrid cistern was modelled as part of a treatment train with the addition of the permeable parking lot system and rain gardens. The table below presents the modelled scenario at the commercial lot with the RainGrid cistern, permeable parking lot system, and rain gardens. Prior to installing the LID features, more that 70% of rainfall was running off the property into the adjacent storm sewer. After the installation of the RainGrid cistern, 100% of rainfall was captured and retained on site for the 90th percentile storm. This property was profiled in the Water Canada magazine (September/October Issue). Click here to view the article.

Water Balance Comparison (Pre to Post Analysis of a Commercial Property with a RainGrid Cistern and other LID Features)

Finally, continued efforts are being made into using collected data and modelling of these individual lot-level controls across an entire subdivision or community to demonstrate the benefits to stormwater mitigation. In addition, future analysis will be completed to determine maximum return on investment for increasing the number of rain cisterns for individual lots so communities can better manage large storm events. When scaled to a subdivision wide effort, this could potentially reduce size and cost of traditional stormwater management features (e.g. stormwater ponds).


Real-time flow rate and water level data from a SafeSump system can be uploaded to a cloud control panel so the property owner has continuous access via their preferred communication device. This allows the property owner to stay informed on the status of their SafeSump as the system will alert them if there is a water level issue in their sump pit. This data can also be used by the homeowner to identify possible abnormal factors contributing to their sump pump activity.

Example Flow Rate vs. Precipitation Data for a Collingwood Home with a Safe Sump System

The above figure demonstrates flow rate data collected from a SafeSump (sump pump) installed in Collingwood. This pump was determined to not typically respond strongly to precipitation events. Instead, the system would continually pump at a low rate throughout a 24-hour timeframe. In addition, it was also found that the pump had a strong response to snowmelt periods through the winter, when the ground’s infiltration capabilities were limited, due to the subsurface being frozen. Data from this SafeSump system was then analyzed in real-time with other (nearby) installed SafeSump systems to determine whether the surrounding area is one of high risk.

Collected data was also useful for the Environment Network (a core project partner) as part of stakeholder engagement and consultation activities. We believe that this information will be helpful for the municipality to initiate a basement flood awareness program for all residents.

Efforts have also been made in using the data to identify additional factors impacting proper sump pump operations. This includes grading of lots, soil properties, discharge pipe location, and shallow groundwater depth.

The SafeSump system includes a proprietary variable speed capability for the sump pump. This unique attribute enabled our team to assess how variable speed pump rates can reduce peak flows and better manage stormwater at the property level. For example, this variable speed pump capability has the potential to reduce peak sump pump flows from entering municipal stormwater infrastructure.

Additionally, the data can be analyzed to assess potential non-compliant connections to the sanitary sewer system as inflow (water entering the wastewater system that does not need to be treated). Inflow can reduce efficiency and increase cost of sewage treatment plants, and has the potential to cause sewer overflow and overwhelm the system, leading to health risks and property damage. By analyzing timing and magnitude of peak flow data from SafeSump system data and comparing it with flow records within the nearby sanitary sewer system, potential impacts from non-compliant connections can be identified. Therefore, this will allow municipalities to focus efforts in better regulating these connections.

Summary & Next Steps

Information collected to-date is supportive of our original objective to examine solutions that allow for the management of sump water level at residential properties. For example, the possibility of pumps being connected to downstream lot level controls (such as rain gardens or designated infiltration areas), could prevent this water from entering the storm sewer system altogether, or at a minimum, detain the water until after the storm event has passed. Similar to the RainGrid systems, the collected data can be scaled for large scale implementation using a stormwater model and identifying optimal solutions to better manage lot-level stormwater drainage.

Lastly, our team will post another project update early in the new year as our activities conclude for this specific initiative. We look forward to communicating our overall project results in a comprehensive summary on our project site. 

– The Collingwood Stormwater Project Team