Utility Takes Innovative Approach to Reducing Urban Flooding

Urban flooding is an issue faced by many cities around the country. Due to the fully developed nature of older cities, finding solutions to these flooding problems is not an easy task.

Jul 1st, 2003
Th 128619

By Suresh L Hettiarachchi, Anthony J. Luft, and Jane Onorati

Urban flooding is an issue faced by many cities around the country. Due to the fully developed nature of older cities, finding solutions to these flooding problems is not an easy task. The City of Minneapolis, MN, belongs to this category. Flooding became a prominent topic of discussion in 1997 during a particularly wet month of July. Consecutive severe rainstorms, the first of which was equal to a 200-year rainfall event, caused widespread flooding resulting in significant economic impact to the city.


Typical cross section of the storage chamber design and connection to main trunk sewer in zone 2.
Click here to enlarge image

Subsequently, the city developed a report, Flood '97, using complaint reports by residents, field inspections, and preliminary hydraulic and hydrologic analysis, which led to the Flood Mitigation Program for the city. Flood Area 27 (FA 27), which is a residential neighborhood approximately four miles southeast of the city center is one of the several watersheds identified in the Flood '97 report. Following is a brief discussion of the development of a Flood Mitigation Project in FA 27.

Project Area

Flood Area 27 is a 225-acre watershed with predominantly residential land use. The storm sewer network culminates at the southern end with a 60-inch diameter pipe that outlets into Minnehaha Creek. FA 27 falls within the jurisdiction of the Minnehaha Creek Watershed District (MCWD). MCWD is tasked with managing and protecting the creek, which is considered a national natural resources asset. Lake Hiawatha, which is part of the creek drainage system, lies immediately to the west of FA 27.


Flooding in Zone 3 during an extreme rain event in the Summer of 1997.
Click here to enlarge image

The Flood '97 report identified three locations within the FA 27 watershed for which flood mitigation should be implemented. These three flood zones are local low areas. Surface runoff from surrounding areas and storm water from surcharging manholes collect in these zones to cause flooding. The flooding in Zones 2 and 3 was quite severe on July 1, 1997. Residents in the area reported that the intersection of 39th Street E and 30th Avenue S, in Zone 2, was under three feet of water during that storm. Residents of both Zone 2 and Zone 3 areas have complained of frequent flooding, but to a lesser extent, during less severe storms.

Preliminary Analysis

The city hired HDR to provide consulting engineering services to help analyze and develop solutions to the problems faced by residents in these flood zones. A model of the storm sewer network was developed using XP-SWMM modeling software to analyze the existing system and develop possible solutions. Many alternatives were evaluated, including:

• Relief pipes that would outlet to Lake Hiawatha.
• Options to install an additional storm sewer trunk line alongside the existing main sewer draining to Minnehaha Creek, or
• Above-ground storage, which would involve taking of homes to build a detention pond.

The alternatives were submitted to the Minneapolis Parks and Recreation Board (MPRB) and MCWD for consideration.

The MPRB expressed concern over discharging additional storm water to the lake. The water quality in Lake Hiawatha has degraded over time. The MPRB was concerned that fecal coliform levels might increase due to discharge of additional storm runoff. Increased coliform levels come predominantly from waterfowl and pet wastes in yards, streets and parks that wash into lakes either directly or through storm sewers as the result of a heavy rain. In addition, the proposed outlets would be installed in close proximity to a high-use beach area. The option to route the new sewer to Lake Hiawatha was discarded due to possible environmental issues resulting from potential impact to the beach.

The MCWD was concerned about how adding an additional trunk line outlet would impact the creek, which is already heavily burdened by numerous storm sewers from city neighborhoods that outlet to the creek, resulting in 'flashy' peak flows. The flashy nature of these flows has caused stream bank erosion and overbank degradation, and the MCWD has instituted many rules to control increased discharge rates and volumes to protect the creek.

Other options considered included above ground storage, which was rejected due to cost and impact to the residents.

This left the option of the additional trunk sewer outleting to the creek, initiating a series of debates dealing with the timing of the peaks from FA 27 to the creek, and options to improve the quality of runoff to the creek.

Sewer Model

To evaluate the timing impacts on the creek, the storm sewer model was refined. The watershed discretization was increased from nine to 30 sub watersheds. Streets above the main sewer lines that convey excess runoff were added to the model to accurately simulate the flooding at low areas. Field observations were done during storm events in the summer of 2002 to verify the sub-watershed divides and street flow patterns.

The outlflow hydrograph for FA 27, developed from the refined model, was then compared to the hydrograph for the creek. The hydrograph for the creek was developed from gauge data collected by the MPRB from a gauge located approximately 200 feet downstream from the FA 27 outlet. The creek hydrograph indicated a 'double hump' shape with two peaks occurring about a day and a half apart. The leading peak in the creek hydrograph coincided with the modeled peak discharge from FA 27. Based on this analysis, it was very clear that the outflow from FA 27 has a direct impact on the peak flows in the creek. These findings forced a change in design philosophy, from adding sewers to flush floodwaters to the creek ahead of its natural crest, to finding a design that would not increase the peak discharge rate to the creek from FA 27.

Design Concept

The design concept changed to focus on the three zones individually rather than one system that would ease flooding at all three locations. The criteria of the Minneapolis Flood Mitigation Program required a solution that would protect homes during an extreme storm, which is defined as a 100-year storm for this design. This approach allows for water to pond in the street to a level that will not cause damage to homes. The idea of protecting homes added 'flood proofing' to the mix of elements available to devise a solution. Detailed model analysis, site surveys that included walks in the rain, and many hours of discussion and deliberation led to a design package consisting of the following elements to form an innovative approach to ease flooding problems in FA 27.

• Underground linear storage chambers
• Re-routing and upgrading storm sewers
• Street grading to re-direct surface runoff
• Flood proofing

Underground Storage

Five 600-foot-long underground-storage chambers were designed in Zone 2 with a combined storage capacity of 3.2 acre-feet. The chambers consist primarily of conventional 10-foot-wide pre-cast box culvert sections. Flap gates connect the chambers to the existing storm drain system.

The design of the storage chambers required balancing several criteria. The goal was to maximize storage capacity while simultaneously maintaining a positive slope to facilitate cleaning sediments without violating the minimum cover requirement, avoid as many in-place utilities as possible, and protect boulevard trees. To do this, the height of the chambers increased in an upstream direction from 4 feet to 6 feet to approximate the street profile. The chambers are of uniform height throughout where cover is not a problem.

Each storage chamber is fed directly by at least six curb-opening type catch basins installed on the streets plus two out of the eight intersection inlets. The remaining inlets in the intersections are connected to the existing storm drain system. In some cases, additional inlets were added to the intersections that are connected to the storage chambers via an 18-inch equalization pipe.

Since the area that receives the most flooding is the area with the least cover and therefore the smallest chamber, the equalization pipe was designed to connect this area to the other chambers that have greater capacity. During a flood, the depth of water in the most flooded intersection would drive the water up the equalization pipe.

To assist in cleaning the chambers, three 2-foot deep sumps were designed in each chamber. The first sump is located at mid point and the second sump is directly behind the flap gate. There is also a third sump under the flap gate to prevent debris from blocking the gate. Each sump has weep holes to remove standing water after the chamber drains.

The storage chambers are distributed under the streets that form the sub-watershed that eventually drain to the intersection of 39th Street E and 30th Avenue S, which is the focus point of zone 2. The box culverts that form the storage chambers will initially function as conveyance for the immediate runoff from properties along the street. Inlets are provided along the length of the street to capture the direct runoff and prevent run-by from collecting at Zone 2. During this period, the main trunk sewer along 39th Street E still has capacity to drain the area and the flap gates at the end of each of the box culverts remain open.

During the peak period of the storm, the function of the box culverts changes from conveyance to storage. The main trunk sewer is now surcharged and the flap gates close. Normally, the runoff during this period cannot drain out through the sewer and would flow along the streets to the low point at Zone 2. The underground storage chambers and connected inlets prevent this by continuing to capture and storing the immediate runoff during the peak of the storm. Once the peak has passed, the flap gates gradually open and the storage chambers return to function as conveyance systems.

The required storage capacity is greatly reduced by this approach of lateral storage as opposed to in-line storage. Storage capacity requirements are further reduced by maintaining a direct connection to the trunk sewer system to allow drainage during the initial and latter period of the storm.

Re-routing, Upgrading Sewers in Zone 3

Similar to Zone 2, a majority of the flooding at Zone 3 is caused to run-by from surrounding areas. Because of conflicts with sanitary sewer lines and lack of clearance, underground storage was not feasible at this location. Instead, a more conventional approach of storm sewer upgrades and re-routes is employed to flush the water from the low area. The existing storm sewer draining the low spot is increased in size to add capacity. A new storm sewer is added taking a different route and connecting to the main trunk line at a location further downstream from the existing pipe connection. The thought is to allow the direct runoff at zone 3 to drain faster out of the area before the main trunk line surcharges.

The upgraded storm drain is isolated from the existing system by a flap gate and the most flood prone area is removed from the current drainage pattern and rerouted southward to connect back to the main trunk line. The upgrading consists of replacing 9-, 12-, and 15-inch pipes with 18-inch and 24-inch Reinforced Concrete Pipe (RCP). The rerouted line consists of approximately 1,100 feet of 36-inch RCP.

Street crowns or 'speed bumps' are used to re-direct or delay street flow that would eventually collect in Zone 3. The modeling results indicate that an appreciable volume of water flows into Zone 3 due to back flow in the storm sewer connecting to the main trunk line during the peak flow period. Flap gates at critical connections stop this back flow and prevent this volume of water flooding into Zone 3.

Flood Proofing in Zone 1

After careful examination of the complaint reports at Zone 1, it was determined that the damage to homes was caused by sanitary sewer back ups rather than surface flooding. The model analysis indicates that the streets do flood in Zone 1, but the level and extent of flooding is within the design concept of protecting homes. Flood proofing measures are employed in Zone 1 to add further protection to houses observed to be at 'high risk'.

Current Status

Currently, the plans are being completed and phased construction is scheduled to commence this summer with completion by the summer of 2005.


About the Authors: Suresh Hettiarachchi is a water resources engineer in HDR's Minneapolis office. Anthony Luft is a water resources and civil engineer in HDR's Minneapolis office. Jane Onorati is a project manager for the City of Minneapolis Public Works Department.

More in Home