&EPA
United States
Environmental Protection
Agency
2012 GREEN INFRASTRUCTURE TECHNICAL ASSISTANCE PROGRAM
City of La Crosse
La Crosse, Wl
Using Green Infrastructure to Mitigate Flooding in
La Crosse, Wl
Assessment of Climate Change Impacts and System-Wide Benefits
Photo credit: Jennifer Olson, Tetra Tech
December 2014
EPA832-R-15-002
-------
This page is intentionally blank.
-------
About Green Infrastructure and the 2012
Community Partner Program
Stormwater runoff is a major cause of water pollution in urban areas. When rain falls in undeveloped
areas, the water is absorbed and filtered by soil and plants. When rain falls on our roofs, streets, and
parking lots, however, the water cannot soak into the ground. In most urban areas, Stormwater is drained
through engineered collection systems and discharged into nearby waterbodies. The Stormwater carries
trash, bacteria, heavy metals, and other pollutants from the urban landscape, degrading the quality of the
receiving waters. Higher flows can also cause erosion and flooding in urban streams, damaging habitat,
property, and infrastructure.
Green infrastructure uses vegetation, soils, and natural processes to manage water and create healthier
urban environments. At the scale of a city or county, green infrastructure refers to the patchwork of
natural areas that provides habitat, flood protection, cleaner air, and cleaner water. At the scale of a
neighborhood or site, green infrastructure refers to Stormwater management systems that mimic nature by
soaking up and storing water, such as rain gardens, permeable pavement, and green roofs. These
neighborhood or site-scale green infrastructure approaches are often referred to as "low impact
development."
EPA encourages the use of green infrastructure to help manage Stormwater runoff. In April 2011, EPA
renewed its commitment to green infrastructure with the release of the "Strategic Agenda to Protect
Waters and Build More Livable Communities through Green Infrastructure." The agenda identifies
community partnerships as one of five key activities that EPA will pursue to accelerate the
implementation of green infrastructure.
EPA announced partnerships with 10 "model communities" in April 2011. These communities have
demonstrated how green infrastructure can supplement or substitute for single-purpose "gray"
infrastructure investments such as storm sewers and detention ponds.
In February 2012, EPA announced the availability of $950,000 in technical assistance to a second set of
partner communities to help overcome some of the most common barriers to green infrastructure. EPA
received letters of interest from over 150 communities across the country. EPA selected 17 of these
communities to receive assistance with code review, green infrastructure design, and cost-benefit
assessments.
Through the assistance provided to the City of La Crosse, Wisconsin, EPA developed a detailed Storm
Water Management Model of the Johnson Street Basin. The model was used to evaluate the potential for
green infrastructure to mitigate peak flows and reduce associated flooding and pollutant loads. Porous
pavement and bioretention were modeled in the street right of way at various levels of implementation. A
series of performance curves was developed which summarizes the flood reduction effectiveness of the
modeled green infrastructure and the cost-effectiveness of each scenario.
For more information, visit http://water.epa.gov/infrastructure/greeninfrastructure/gi_support.cfm
-------
Acknowledgements
The City of La Crosse would like to recognize the following project partners for their contributions to this
project:
Bernard Lenz, City of La Crosse
Yuri Nasonovs, City of La Crosse
Amy Peterson, City of La Crosse
Sally Kefer, Wisconsin DNR
David Liebl, University of Wisconsin
Bob Newport, EPA Region V
Jonathan Moody, EPA Region V
Tamara Mittman, U.S. EPA Headquarters
Christopher Kloss, U.S. EPA Headquarters
Jennifer Olson, Tetra Tech
George Remias, Tetra Tech
John Stein, Tetra Tech
Garrett Budd, Tetra Tech
This report was developed under EPA Contract No. EPA-C-11-009 as one of the 2012 EPA Green
Infrastructure Community Partner Projects.
-------
Contents
Executive Summary 1
1. Project Goals 2
2. Study Area 3
3. Model Development and Existing System Evaluation 6
4. Green Street Scenarios 8
4.1 Green Infrastructure Selection 8
4.2 Green Infrastructure Extent 10
4.3 Green Infrastructure Costs 13
5. Green Street Evaluation 14
5.1 Flood Mitigation 14
5.2 Cost-Effectiveness Curves 18
5.3 Pollutant Load Reduction 20
5.4 Recommended Implementation Scenario 20
6. Findings and Conclusions 22
7. References 23
Appendix A: Model Construction, Parameterization, and Calibration
Appendix B: Additional Green Infrastructure Design Parameters
Appendix C: Performance Curves Supporting Selection of the Most Effective Practice
Figures
Figure 1. Example of manhole that typically floods, note open grate which addresses safety
concerns related to solid manhole covers being lifted due to force of water 2
Figure 2. Johnson Street Basin watershed 3
Figure 3. Johnson Street Basin topography 4
Figure 4. Typical residential areas in the Johnson Street Basin 5
Figure 5. Johnson Street Basin land use 5
Figure 6. City reported flooding locations 6
Figure 7. Modeled results showing surface flooding for the 10-year 2-hour event 8
Figure 8. Example permeable pavement (concrete pavers) application in parking lane on residential
street 9
Figure 9. Example bioretention areas: part of a complete street, linear feature between curb and
sidewalk, and as a bump out 9
-------
Figure 10. Example of alley (top) and local road with two parking lanes (bottom) in Johnson Street
Basin 10
Figure 11. La Crosse road types 11
Figure 12. Conceptual placement of green infrastructure on a local road 12
Figure 13. System flooded versus green infrastructure scenario, 3 month, 2-hour event 15
Figure 14. System flooded versus green infrastructure scenario, 10-year, 2-hour event 16
Figure 15. Volume flooded versus green infrastructure scenario, 10-year, 2-hour event 17
Figure 16. Example hydrograph for representative catchment, 10-year, 2 hour storm event 17
Figure 17. Cost-effectiveness curve, 3-month, 2 hour event 18
Figure 18. Cost-effectiveness curve, 10-year 2 hour event 19
Figure 19. Cost envelope for implementing permeable pavement with four feet of grave storage in
75 percent of the watershed 19
Figure 20. Pollutant and volume reductions for permeable pavement with four feet of storage, 75
percent scenario 20
Figure 21. Recommended implementation scenario for green infrastructure, model results showing
surface flooding for the 10-year, 2-hour event 21
Figure 22. Modeled existing conditions showing surface flooding for the 10-year 2-hour event 21
Tables
Table 1. SWMM predicted street flooding results 7
Table 2. Street assumptions and green infrastructure applicability 12
Table 3. Extent of green infrastructure for modeled scenarios 13
Table 4. Green infrastructure lifecycle cost assumptions (2012$) 14
Table 5. System flooded versus green infrastructure scenario, 3 -month, 2-hour event 15
Table 6. System flooded versus green infrastructure scenario, 10-year, 2-hour event 16
Table 7. Permeable pavement with 4 foot gravel storage cost range for each implementation
scenario 20
-------
Executive Summary
As climate change increases the frequency of intense rain events in the Midwest (USGCRP 2009),
appropriate adaptation strategies are needed to manage the impacts of increased runoff volumes and rates.
The City of La Crosse, Wisconsin offers an illustrative example. With its flat topography and long pipe
network, the City's drainage infrastructure is very sensitive to increases in rainfall amounts and intensity.
When intense storms occur, water either can't enter the storm sewer system or flows backward in
topographically low areas, lifting storm sewer covers and forming geysers in the streets. The City has
prioritized the Johnson Street Basin on the south side of the City for adaptation strategies, but lacks a plan
beyond the standard practice of upsizing pipes.
At the same time, the City and County of La Crosse have identified several sustainability goals related to
the City's transportation infrastructure. By providing greater access to bicycles and pedestrians, the City
hopes to expand transportation choices, reduce fossil fuel consumption, and improve public health.
In seeking strategies to adapt to a changing climate and advance its sustainability goals, the City
identified green streets as a promising alternative. Green streets integrate green infrastructure into the
street and rights-of-way to intercept runoff from the street and adjacent parcels and reduce the burden on
the storm sewer system. By reducing the urban heat island effect and improving aesthetics, green streets
also enhance the bicycle and pedestrian environment.
This project evaluates the potential of green streets to mitigate flooding in the Johnson Street Basin. The
project consisted of five steps. First, a detailed EPA Storm Water Management Model (SWMM) was
developed to represent runoff generation and conveyance in the Johnson Street Basin. Second, three green
street designs were selected for further analysis. One of the selected designs added bioretention to the
right-of-way, while two of the selected designs added permeable pavement. Third, the applicability of
each design within the basin was determined. While there was sufficient space to place bioretention along
30 percent of local roads, permeable pavement could be installed along 80 percent of major roads and 90
percent of local roads. Fourth, each design was simulated in the SWMM model across a range
implementation levels. Finally, cost data were integrated to yield cost effectiveness curves.
The model results demonstrate the potential for green streets to significantly reduce localized flooding
and mitigate the impacts of climate change. The three modeled systems had the capacity to nearly
eliminate flooding from the 3-month, 24-hour event. Currently, 17% of manholes are predicted to flood
during the 3-month, 24-hour event under existing conditions. Installing bioretention at all possible
locations reduces flooding by 88%; installing permeable pavement at all possible locations eliminates
flooding. Permeable pavement was also a lower cost option. The model indicated that permeable
pavement was the most effective system for reducing the extent and duration of flooding associated with a
large storm event (2.86 inches). Under current conditions, 63% of manholes are predicted to flood during
a 2.86 inch event. Full implementation of permeable pavement with a storage depth of 4 feet is predicted
to reduce flooding by 87%, resulting in fewer than 10% of manholes flooding during a large storm event.
Complete green infrastructure build-out is not required to positively impact local flooding. Modeling
projects that install permeable pavement in 75% of appropriate locations showed that system flooding was
reduced by 68% (with 20% of manholes flooding as opposed to current conditions of 63% of
manholes).While this study evaluated basin-wide green infrastructure implementation, these results also
suggest that prioritizing problem areas and optimizing implementation activities in priority areas would
likely result in less costly solutions.
-------
1. Project Goals
The City of La Crosse has identified climate change adaptation as a key priority. In 2012, the
Sustainable La Crosse Commission conducted a workshop on climate change for the La Crosse area
governments and community leaders. Attendees discussed the risks that climate change might pose to La
Crosse, and identified a range of short- and long-term measures that could enhance the resiliency of
human and natural systems. Building on the workshop, La Crosse worked with the Wisconsin Department
of Natural Resources to complete a Climate Adaptation Study (Kefer 2013).
Among the highest adaptation priorities identified by La Crosse is reducing flood hazards. In the next half
century, precipitation is expected to increase in the region, and storms are expected to become more
severe (Sustainable La Crosse Commission 2012). With the City's flat topography and long pipe network,
many areas of the City are very sensitive to increases in rainfall amounts or intensity. When intense
storms occur, water either can't enter the storm sewer system or flows backward in topographically low
areas, lifting storm sewer covers and
forming geysers in the streets (Figure
1). In recent years, homes have
collapsed due to saturated soils and
standing water during flood events,
and residents are filing more claims
for property damage. The City has
prioritized the Johnson Street Basin on
the south side of La Crosse for
mitigation, but currently lacks a plan
beyond the standard practice of
upsizing pipes.
La Crosse and La Crosse County have
also set a series of ambitious
sustainability and public health goals.
In 2009 the City and County adopted a
Strategic Plan for Sustainability with
goals and actions in the areas of
energy consumption, transportation,
purchasing decisions, waste
generation, natural resource
preservation, and community
livability. Several of the actions address the City's transportation infrastructure, calling for more bicycle
and pedestrian friendly streets. By providing greater access to bicycles and pedestrians, the City can
expand transportation choices, reduce fossil fuel consumption, and improve public health.
In seeking strategies to adapt to a changing climate and advance its sustainability goals, the City
identified green streets as a promising alternative. By absorbing and slowing the flow of water, green
streets can reduce the burden on the storm sewer system and mitigate localized flooding. Also, by adding
vegetation and permeable surfaces to the built environment, green streets can improve aesthetics, reduce
the urban heat island effect, and create a more bicycle and pedestrian-friendly environment. Recognizing
the promise of green streets and the importance of effective stormwater management, La Crosse adopted
the first "Green Complete Streets" ordinance in Wisconsin, and is generating funds for infrastructure
upgrades through a new stormwater utility. The City has also started implementing green infrastructure as
part of road reconstruction projects.
Photo credit: Tetra Tech
Figure 1. Example of manhole that typically floods, note open grate
which addresses safety concerns related to solid manhole covers
being lifted due to force of water.
-------
The goal of this project is to assess the potential for green infrastructure in green streets to mitigate
flooding in La Crosse. The study area selected for this project is the Johnson Street Basin, a high- density,
residential area of the City that is experiencing flooding from more frequent and intense storms. The City
plans to apply the project findings to install recommended facilities in the Johnson Street Basin and to
build capacity for green infrastructure planning.
2. Study Area
La Crosse is a city of 51,000 located in western Wisconsin. Bordered to the west by the Mississippi River
and to the east by 500 foot bluffs, the City is built on a broad alluvial plain formed at the confluence of
the Black and La Crosse Rivers. Runoff from the bluffs and stormwater from the City are conveyed
toward the rivers across this broad plain. Conventional stormwater facilities and best management
practices (BMPs) are especially challenging given these topographical constraints.
The Johnson Street Basin is located entirely within La Crosse and drains to the Mississippi River through
La Plume Slough (Figure 2). The 769 acre basin is fairly flat, ranging from a high point of 684.5 feet near
the intersection of 14th Street and King Street to 663.0 feet near U.S. Route 14 and Johnson Street.
Topography includes slight depressions near the southeast and northeastern corners (Figure 3), with an
average slope of one percent.
Johnson Street Basin
1,000 2,000 3,000 4,000
H= —^^^™— =iFeet
Figure 2. Johnson Street Basin watershed.
-------
Elevation
High
N
+
750 1,500 2,250 3,000
•— =^^^«= =]Feet
Figure 3. Johnson Street Basin topography.
Soils are formed on sand and gravel outwash deposits, and therefore exhibit high infiltration rates. The
majority of the basin consists of residential land uses constructed between 1930 and 1950, encompassing
72 percent of the area (Figure 4 and Figure 5). Other land uses include industrial and commercial
properties, churches, hospitals/clinics, and schools. Approximately two thirds of the basin (67.2 percent
or 517 acres) is impervious, which includes approximately 38 acres of parking lots and 248 acres of road
rights-of-way.
Field reconnaissance of the study area was conducted in October 2012. New development and re-
development over the past couple of decades has resulted in significant hydrologic changes within the
watershed, including increased runoff volumes and peak flows. For example, most of the City's gravel
alleys have been paved. These alleys are now concrete-lined and sloped such that runoff leaves the site
very quickly. While the storm sewer system was extended several times to accommodate the additional
development, City design criteria (i.e., design of storm sewers to convey a 10-year, 24-hour storm) were
not consistently followed.
Consequently, several locations within the study area frequently flood (Figure 6), and the time to drain the
flooded area continues to increase as the runoff volume increases. To mitigate this flooding, the City has
installed bioretention areas, disconnected downspouts, and installed underground storage in a few
locations. The City has also intentionally undersized many storm sewers to allow surface flooding
upstream and help minimize severe flooding downstream such as along Adams Street in the Johnson
Street Basin.
-------
I
Photo credit: Tetra Tech
Figure 4. Typical residential areas in the Johnson Street Basin.
Commercial/Industrial/lnstitutional Zone
Residential Zone
1,000 2,000 3,000 4,000
^•= __^^^^™_ => Feet
Figure 5. Johnson Street Basin land use.
-------
^iv«aoffi
Storm Sewer Problem Spots
Street Flooding Problems
1,000
2,000
3,000
Figure 6. City reported flooding locations.
3. Model Development and Existing System
Evaluation
EPA's Storm Water Management Model (SWMM; EPA 2008) was used to develop a hydrologic and
hydraulic model of the Johnson Street Basin. SWMM is a dynamic hydrology-hydraulic-water quality
model that simulates runoff quantity and quality from urban areas. SWMM can track the quantity and
quality of runoff generated within each sub catchment, as well as the flow rate, flow depth, and quality of
water in each pipe and channel. SWMM 5 can also model the hydrologic performance of several green
infrastructure controls, including permeable pavement, rain gardens, green roofs, street planters, rain
barrels, infiltration trenches, and vegetated swales.
Data required for this SWMM analysis include design storms to drive the model; catchment physical
characteristics (topography, soils, land cover) to construct and parameterize the model; and observed
flooding response to calibrate the model. The model construction, parameterization, and calibration are
described in Appendix A, including the data sources for the design storms and catchment physical
characteristics.
-------
Once a SWMM model was developed for the Johnson Street Basin, the extent and duration of flooding in
the existing system was evaluated. This evaluation was conducted to establish a baseline for comparison
to the green street scenarios. As described in Appendix A, four design storm events were used to drive the
SWMM model: a 3-month, 2-hour storm event to represent existing flooding conditions; a 1-year, 24-
hour storm event to simulate a small, frequent storm event; a 10-year, 24-hour storm to represent design
criteria listed in the La Crosse ordinance; and a 10-year, 2-hour climate change storm event to represent a
potential short duration, high intensity climate change scenario. The 10-year, 2-hour climate change storm
event was also identified by City staff as the targeted level of service, or the design storm event that
should not cause flooding.
Significant flooding was predicted for each of the design storm events, with the duration and frequency of
flooding generally increasing as the peak 1-hour depth and the total rainfall depth increase. Note that the
duration of manholes flooded represents the accumulated duration of flooding over all the manholes in the
system. As shown in Table 1, the percent of manholes flooded in the existing system ranged from 17%
for the 3-month, 2-hour storm to 63% for the 10-year, 2 hour climate change storm. Similarly, the
duration of manholes flooded ranged from 45 hours for the 3-month, 24-hour storm to 505 hours for the
10-year, 2 hour climate change storm. Figure 7 illustrates the widespread flooding that occurs in the
existing system for the 10-year, 2-hour climate change storm event.
Table 1. SWMM predicted street flooding results
Design storm
event
3-month, 2 hour3
1-year, 24 hour3
10-year, 24 hour3
10-year, 2 hour
climate change
Total
rainfall
depth
(inch)
0.83
2.23
4.40
2.86
Peak
1 hour depth
(inch)
0.67
1.05
2.07
2.32
Manholes
flooded
17
36
57
63
Duration of
manholes
flooded
(hours)
45
130
481
505
Flooded
volume
(million
gallons)
0.8
3.0
11.9
14.7
Average peak
flooding depth
(inch)
2
3
8
9
a. Values are based Huff and Angel 1992
-------
Manholes Flooded: 249
Total Duration of Manholes Flooded: 505-hours
T Flooded Volume: 14.73 MG
Average Peak Flooding Depth: 9-inches
.^Problem
£ Area "A"
o-
Market S.
Storm Sewers
10-Year, 2-Hour Climate Change Event
Existing Conditions Flooding (ft3)
o up to 2,500
O 2,500 to 5,000
O 5,000 to 10,000
O 10,000 to 20,000
Problem
$L Area "B"
Problem
Area "C"
oWinnebagoSt O
-—Mississippi StQ
| • (Lksofst
1-Hjr
O0/0001
I -i
•x™\ *~* j^K^ __/~\
rnam St
O
greater than 20,000
1,000
2,000
3,000
4,000
=]Feet
Greer^Bay^it
Barlow St
Figure 7. Modeled results showing surface flooding for the 10-year 2-hour event.
4. Green Street Scenarios
La Crosse is committed to implementing a green streets program to help mitigate flooding, promote water
quality treatment, and advance its sustainability goals. To evaluate the impact of green streets on localized
flooding in the Johnson Street Basin, several green street scenarios were simulated in the SWMM model
described above. This section describes how key modeling inputs for each scenario were defined,
including the green infrastructure practices selected, the assumed design parameters for each practice, and
the extent of each practice within the Johnson Street Basin. The costs associated with these design
parameters are also discussed.
4.1 Green Infrastructure Selection
Green streets implement green infrastructure within the street right-of-way to manage runoff from both
the street and adjacent parcels. Green street features can include permeable paving (Figure 8),
bioretention areas (Figure 9), sidewalk planters, landscaped medians, vegetated swales, and street trees.
The most common approaches include bioretention areas located between the edge of the pavement and
the edge of the right-of-way and permeable pavement installed in the parking lanes. Permeable pavement
and bioretention were therefore selected for further analysis.
-------
Permeable pavements work by allowing
streets, parking lots, sidewalks, and other
impervious covers to retain their natural
infiltration capacity while maintaining the
structural and functional features of the
materials they replace. Permeable pavements
contain small voids that allow water to drain
through the pavement to an aggregate
reservoir and then infiltrate into the soil. The
depth of the gravel storage layer is an
important design parameter.
Bioretention typically consists of a shallow,
vegetated basin that collects and absorbs
runoff from impervious areas. These practices
usually consist of a grass buffer strip, ponding
area, mulch layer, and planting soil media.
Similarly to permeable pavement, the depth of
the soil media layer is an important design parameter.
Photo credit: Tetra Tech
Figure 8. Example permeable pavement (concrete
pavers) application in parking lane on residential street
Based on input from city staff, three green street systems were modeled for this study: a permeable
pavement system with a storage depth of two feet, a permeable pavement system with a storage depth of
four feet, and a bioretention system with a soil media depth of three feet. Additional design parameters for
each practice are described and summarized in Appendix B.
Photo credits: Tetra Tech
Figure 9. Example bioretention areas: part of a complete street, linear feature between
curb and sidewalk, and as a bump out.
-------
4.2 Green Infrastructure Extent
To develop a series of green street scenarios
representing different levels of implementation, the
maximum extent of each practice was estimated along
three categories of roads: 1) major roads, 2) local roads,
and 3) alleys (Figure 10). The various road types in La
Crosse are illustrated in Figure 11.
Based on input from city staff, review of aerial photos,
and field reconnaissance, the maximum extent of
permeable pavement was determined to be 80 percent
along major roads, 90 percent along local roads, and
100 percent along alleys—corresponding to 286,000
linear feet of permeable pavement. This is because
permeable pavement was determined to be feasible
along the entire length of each road type with the
exception of intersections.
In contrast, the maximum extent of bioretention was
determined to be 30 percent along local roads only -
corresponding to 85,000 linear feet of bioretention.
This is because city staff considered bioretention to be
infeasible along driveways and intersections (where
bioretention could obstruct views or access) and along
major roads and alleys (which consist of driving lanes
only and lack space for bioretention).
Table 2 summarizes the road properties and maximum
extent of green infrastructure for each road type, while
Figure 12 illustrates atypical local road showing
placement of green infrastructure.
Photo credit: Tetra Tech
Photo credit: Tetra Tech
Figure 10. Example of alley (top) and local road
with two parking lanes (bottom) in Johnson
Street Basin.
Once the maximum extent of permeable pavement and
bioretention was known, green street scenarios were
developed spanning 25%, 50%, 75%, and 100% of the
maximum extent of implementation (Table 3). For the permeable pavement systems, two additional
scenarios were developed - one in which permeable pavement was implemented in 100% of the feasible
area as well as the entire width of the allies, and a second in which permeable pavement was implemented
in 100% of the feasible areas, the entire width of the allies, and a portion of existing parking lots. Parking
lots account for approximately 38 acres of impervious area in the watershed and are often located
upstream of flood prone areas. Sixty percent of the parking lots were assumed to be converted to
permeable pavement.
10
-------
•••mil ituE I
Figure 11. La Crosse road types.
11
-------
Table 2. Street assumptions and green infrastructure applicability
Summary of Road Properties
Category
Percent of street
Average width
Number of driving lanes
Is there on-street parking?
Where is parking?
Parking on 1 or 2 sides?
Unit
(%)
(feet)
(#)
(Yes or No)
(#)
Major Roads
8
49
4
Y
Within driving lanes
2
Local Roads
63
37
2
Y
Dedicated parking
lanes
2
Alley
29
17
1
N
NA
NA
Maximum Extent of Bioretention
Category
Percent street length converted
to bioretention
Width of bioretention
Unit
(%)
(feet)
Major Roads
None
NA
Local Roads
30
8 each side of road
Alley
0
NA
Maximum Extent of Permeable Pavement
Category
Percent street length converted
to permeable pavement
Width of permeable pavement
Unit
(%)
(feet)
Major Roads
80
8-10
Local Roads
90
16
Alley
100
4
Existing
Conditions!
Permeabl
Pavement
Figure 12. Conceptual placement of green infrastructure on a local road.
12
-------
Table 3. Extent of green infrastructure for modeled scenarios
Implementation Scenario
Existing Conditions
25%
50%
75%
100%
100+Alleys
100+Alleys+Parking Lots
Extent of Modeled Porous Pavement3
(square feet)
0
571,000
1,142,000
1,713,000
2,284,000
2,532,000
3,410,000
Extent of Modeled Bioretention3
(square feet)
0
170,000
340,000
510,000
680,000
—
—
a. Assumes applicability based on Table 2.
4.3 Green Infrastructure Costs
Resource constraints may limit the type and number of practices that can be used to achieve program
goals. Costs are evaluated with estimated reductions to select the final set of practices that are most cost-
effective. There are three types of costs to consider over the life cycle of a green infrastructure
intervention:
• Probable Construction Costs - the initial cost to construct
• Annual Operation and Maintenance - the annual costs to maintain
• Repair and Replacement Costs - the additional costs to repair or replace
The lifecycle period was defined as 20-years to account for costs for replacing some practices. No land,
capital, administration, demolition, or legal cost factors were defined for any of the probable construction
costs. Three unit costs were defined for each practice to represent an envelope of possible costs. A range
of probable construction costs were used to represent low, median, and high construction costs.
Construction costs have a high level of variability from project to project. The variability in costs can be
attributed to the level of experience of the designers and contractors, the number of practices constructed
in a given area, quality of the construction documents, and availability of special equipment or specific
supplies (e.g., soil amendment or aggregate).
Operation and maintenance costs were held constant. Each unit cost was converted to 2012 dollars by
applying a three percent inflation rate by the number of years from the published year of the cost data to
2012. A discount rate of 3 percent was used for converting annual operation and maintenance costs to
present value. There were no repair and replacement costs included in this analysis since the life span of
these practices is expected to exceed 20-years.
The following references were evaluated when determining appropriate costs for the Johnson Street
Basin:
• BMP and Low Impact Development Whole Life Cost Models Version 2.0. Water Environment
Research Foundation (WERF 2009)
• National Green Values Calculator, Center for Neighborhood Technology (Center for
Neighborhood Technology 2009)
• The Cost and Effectiveness of Stormwater Management Practices, University of Minnesota
(Weiss et al. 2005)
13
-------
Long-Term Hydrologic Impact Analysis Low Impact Development Version - 2.0
• Low Impact Development for Big Box Retailers. Prepared for U.S. Environmental Protection
Agency (Low Impact Development Center 2005)
• Low Impact Development Manual for Michigan, Southeast Michigan Council of Governments
Table 4 presents the assumed lifecycle costs for each green street design. Note that these costs represent
retrofit costs. The marginal costs of adding bioretention or permeable pavement to a planned street
improvement project would be lower.
Table 4. Green infrastructure lifecycle cost assumptions (2012$)
Cost Parameter
Net Present Value (NPV) = (A) + (B)
(A) Probable Construction Costs NPV
(B) O and M Present Value
O and M Annual Costs
Unit
$/SF
$/SF
$/SF
$/SF
Bioretention
Low
$15
$8
$7
$0.72
Median
$23
$16
$7
$0.72
High
$31
$24
$7
$0.72
Permeable Pavement
Low
$12
$8
$4
$0.28
Median
$16
$12
$4
$0.28
High
$20
$16
$4
$0.28
5. Green Street Evaluation
As described above, three sets of green street scenarios were developed for the Johnson Street Basin, one
representing permeable pavement with a storage depth of 2 feet, one representing permeable pavement
with a storage depth of four feet, and one representing bioretention with a soil media depth of three feet.
For each scenario, SWMM simulations were conducted for two design storms: the 3-month, 2-hour storm,
and the 10-year, 2-hour climate change storm. These design storms were selected because the city was
primarily interested in flooding associated with intense, short duration events.
5.1 Flood Mitigation
Modeling results indicate that all three green street designs significantly reduce the extent of flooding for
the 3-month, 2-hour storm event (Figure 13 and Table 5). At 100% implementation, the percent of
manholes flooded declines from 17% without green infrastructure to 2% with bioretention, and 0% for
both permeable pavement systems. Flood reduction for the 10-year, 2-hour event, however, differs
significantly between the bioretention and permeable pavement systems (Figure 14 and Table 6). At
100% implementation, the percent of the drainage system flooded declines 7.9% with bioretention, from
63% of manholes flooded to 58% of manholes flooded. In contrast, the percent of the system flooded
declines 39% for permeable pavement with a storage depth of 2 feet, and 87% for permeable pavement
with a storage depth of 4 feet. This is due to the limited extent of bioretention in this watershed (30
percent of street length for local roads only) when compared to the extent of permeable pavement (80 - 90
percent of all roads).
Figure 15 presents the volume of flooding for each scenario for the 10-year, 2-hour event and illustrates
that although bioretention does not substantially reduce the number of manholes flooded (Figure 14 and
Table 6); the volume of flooding is reduced. Figure 16 presents ahydrograph for the 10-year, 2-hour
storm event showing existing conditions and three scenarios for a representative catchment consisting
primarily of residential land uses. The hydrograph shows a small decrease in peak flows for bioretention
with three feet of amended soil and permeable pavement with two feet of gravel storage although the
14
-------
duration and volume are both reduced. The permeable pavement design with 4 feet of gravel storage is the
most effective green infrastructure system to reduce flooding for both rainfall events.
Flooding is further mitigated in the additional permeable pavement scenarios including greater
implementation in alleys and parking areas. For the scenario in which permeable pavement with a storage
depth of 4 feet is applied to all feasible areas, the entire width of alleys, and a portion of parking areas, the
percent of manholes flooded is reduced from 63% without green infrastructure to 1% with green
infrastructure. This demonstrates the capacity of intensive green infrastructure to achieve the city's
targeted level of service.
Appendix C includes additional performance curves that were used to support the selection of the most
effective practice including combinations of the system (manholes) flooded, volume flooded, duration of
street flooding, average depth of flooding, costs, scenario, and green infrastructure area.
System Flooded vs Green Infrastructure Implemented
P=3mo2hr{0.83in)
o
_o
u_
E
-------
System Flooded vs Green Infrastructure Implemented
P=10yr2hr{2.86in}
T3
-------
Volume Flooded vs Green Infrastructure Implemented
P=10yr2hr{2.86in)
T3
0)
T3
o
o
-------
5.2 Cost-Effectiveness Curves
Integrating cost information with effectiveness yields a cost-effectiveness curve (Figure 17 and Figure
18). Results for the small storm event (0.83 inches) indicate that all of the practices are similarly cost-
effective; however the permeable pavement design with 4 feet of storage is slightly more cost-effective.
In analyzing cost effectiveness curves it is often helpful to identify the "knee of the curve", or the point on
the curve where additional costs to do not result in significant benefits. The knee of the curve solution for
the permeable pavement design with 4 feet of storage occurs at a cost of $9 million and results in a 100%
reduction in the extent of system flooding. This solution is the result of implementing permeable
pavement in 25 percent of the applicable area.
Results for the large storm event differ in that permeable pavement is shown to be a much more cost-
effective solution to significantly reduce flooding in this watershed. This is due to both the limited
applicability of bioretention in the Johnson Street Basin, and the high intensity of the simulated event.
The knee of the curve solution is the result of implementing permeable pavement in 75 percent of the
applicable area. This solution occurs at a cost of $27 million and results in a 68% reduction in the extent
of system flooding (from 63% of manholes flooded to 20% of manholes flooded). Increasing the level of
implementation to 100 percent permeable pavement results in an 87% reduction in system flooding, but
has a higher cost per unit benefit.
Volume Flooded vs Green Infrastructure Cost
P=3mo2hr{0.83in)
•Bioretention (3-ft)
•Permeable Pvmt(2-ft)
•Permeable Pvmt (4-ft)
Knee-of-the-curve solution
10 20 30 40 50
Estimated 20yr Lifecycle Cost ($Million)
60
Figure 17. Cost-effectiveness curve, 3-month, 2 hour event.
18
-------
System Flooded vs Green Infrastructure Cost
P=10yr2hr(2.86in}
70%
-Bioretention(3-ft)
-Permeable Pvmt (2-ft)
-Permeable Pvmt (4-ft)
^^
Knee-of-the-curve solution,
—75 percent scenario— —<~
0 10 20 30 40
Estimated 20yr Lifecycle Cost ($Million)
50
60
Figure 18. Cost-effectiveness curve, 10-year 2 hour event.
The cost data presented above represent an average life cycle cost. Figure 19 presents an example cost
envelope which more realistically represents the range of potential costs. The cost envelope provides a
low and high cost which takes into consideration variability and uncertainty of available cost data. Table
7 summarizes the life cycle cost range for each implementation scenario.
System Flooded vs Green Infrastructure Cost Envelope
P=10yr2hr(2.86in)
70%
• Permeable Pvmt (4-ft)
Low Cost
High Cost
0%
10 20 30 40 50
Estimated 20yr Lifecycle Cost ($Million)
Figure 19. Cost envelope for implementing permeable pavement with four feet
of grave storage in 75 percent of the watershed.
19
-------
Table 7. Permeable pavement with 4 foot gravel storage cost range for each implementation scenario
Green infrastructure scenario
25 percent
50 percent
75 percent
100 percent
100 percent + Alleys
100 percent + Alleys + Parking Lots
Estimated costs
(Million $) a
$6.9-$11.4
$13.7 -$22.8
$20.6 -$34.3
$27.4 -$45.7
$30.5 -$50.8
$40.9 -$68.2
a. Costs are the same for both rainfall events, effectiveness varies
5.3 Pollutant Load Reduction
A summary of the pollutant load reductions for permeable pavement with four feet of gravel storage and
75 percent implementation is presented in Figure 20. All pollutants are attenuated for the small storm
event (0.83 inches) and 100 percent of the runoff volume is captured and infiltrated. For the larger storm
event (2.86 inches), 77 percent of the pollutants are removed and 76 percent of the runoff volume is
infiltrated. All of the pollutant removal it attributed to volume control only; removal associated with
detention prior to being routed downstream in bioretention areas is insignificant based on the model
results.
13-month, 2 hour (0.83 inches)
10-year, 2 hour (2.86 inches)
TSS
TP TN
Pollutant
Volume
Figure 20. Pollutant and volume reductions for permeable pavement with four feet of
storage, 75 percent scenario.
5.4 Recommended Implementation Scenario
The goal of this project was to evaluate feasible scenarios to provide flood control using permeable
pavement and bioretention. Figure 21 presents the knee-of-the-curve solution for controlling flooding for
the 10-year, 2-hour event. This represents the level of flooding present after implementation of permeable
pavement with a storage depth of 4 feet in 75% of applicable areas. For comparison purposes, the existing
conditions for the same rainfall event are presented in Figure 22.
20
-------
Manholes Flooded: 78
Total Duration of Manholes Flooded: 76-hours
T Flooded Volume: 1.38 MG
Average Peak Flooding Depth: 3-inthes
_ — ., Problem
£ Area/'A"
Kin,
Problem
Area"B"
I?
/\/ Storm Sewers
Porous Pavement (75% of Road) (ft*) ^>
10-Year, 2-Hour Climate Change Event
° up to 2,500
O 2,500 to 5,000
O 5,000 to 10,000
O 10.000 to 20,000
greater than 20,000
+
1,000
2,000
3,000
Figure 21. Recommended implementation scenario for green infrastructure, model results
showing surface flooding for the 10-year, 2-hour event.
Manholes Flooded: 249
Total Duration of Manholes Flooded: 505-hours
T Flooded Volume: 14.73 MS
Average Peak Flooding Depth: 9-inches
„ _ „ Problem
Area "A"
iCf* Sfcj—P-
. @'Ma^iso,
<5
Market St
Storm Sewers
•*.
Existing Conditions Flooding (ft )
10-Year, 2-Hour Climate Change Event
o up to 2,500
O 2,500 to 5,000
O 5,000 to 10,000
O 10,000 to 20,000
-« +&
• -4{— O—i-OCCFa
Problem
0^feDQ^trea"B"
oO ' O QsLf
ff- ^M^ronl^xl^-o^o
/inv ^
'OQ^0^'* **-S>-o
eWmnebagoSt p
Q -MiJissipp7sO
T
Problem
Area"C"
rnam St
6
greater than 20,000
0 1,000
2,000
Green
Barlow St
Figure 22. Modeled existing conditions showing surface flooding for the 10-year 2-hour event.
21
-------
6. Findings and Conclusions
A detailed SWMM model was developed to help La Crosse select green street options that meet both
flood mitigation and livability goals. Within the Johnson Street Basin, bioretention and permeable
pavement were evaluated to determine their potential to mitigate existing flooding conditions. Findings
are presented below:
• The three modeled systems had the capacity to nearly eliminate flooding from the 3-month,
24-hour event. Currently 17% of manholes are predicted to flood during the 3-month, 24-hour
event under existing conditions. Installing bioretention at all possible locations reduces flooding
by 88%; installing permeable pavement at all possible locations eliminates flooding. Permeable
pavement was also a lower cost option.
• The model indicated that permeable pavement was the most effective system for reducing the
extent and duration of flooding associated with a large storm event (2.86 inches). Under current
conditions, 63% of manholes are predicted to flood during a 2.86 inch event. Full implementation
of permeable pavement with a storage depth of 4 feet is predicted to reduce flooding by 87%,
resulting in fewer than 10% of manholes flooding during a large storm event.
• The greater effectiveness of the permeable pavement systems is largely attributable to the greater
area available for implementation of permeable pavement compared to bioretention. While
permeable pavement can be installed along most of the roadway (or 286,000 linear feet of
roadway), bioretention can only be installed along portions of the roadway where it will not
obstruct views or access (or 85,000 linear feet of roadway). Given the ample space available for
permeable pavement and the flood control objective for this basin, permeable pavement
represents the most effective green street approach for the Johnson Street Basin.
• Permeable pavement with a 4 foot gravel storage bed was determined to be the highest
performing system in the basin that can be used to meet the City's objectives for flood control.
• Implementing permeable pavement on 25 percent of the potential street area represents the knee-
of-the-curve solution to mitigate flooding from the 3-month, 2-hour storm event.
• Implementing permeable pavement on 75 percent of the potential street represents the knee-of-
the-curve solution to mitigate flooding from the 10-year, 2-hour storm event.
Focused implementation is recommended to address wide-spread flooding issues and problem areas
identified in this report. At least one of the identified problem areas may require public-private
partnerships to implement green infrastructure that addresses parking lot runoff in order to achieve flood
reduction goals. Parking lots in problem areas should be further evaluated to determine potential for
retrofits and alleys should be considered for permeable pavement retrofits since these areas are
contributing to current flooding problems.
Implementation of green infrastructure should be focused on problem areas first, then address important low
lying areas, and finally include wide-spread adoption throughout the watershed as needed. Adaptive
management is needed, as is monitoring of results that can be used to refine modeling work and adjust
implementation planning. The reality of focused implementation is dependent on capital improvement plans
and road reconstruction activities that are underway or being planned. Intersecting these planning activities
with high priority implementation activities is critical to ensuring a cost-effective management strategy.
Additional analysis could be completed to further optimize implementation activities based on priority
areas and evaluate the potential impact of focused implementation activities, which would likely result in
significantly lower costs.
22
-------
7. References
Butcher, J., A. Parker, S. Sarkar, S. Job, M. Faizullabhoy, P. Cada, J. Wyss, R. Srinivasan, P. Tuppad, D.
Debjani, A. Donigian, J. Imhoff, J. Kittle, B. Bicknell, P. Hummel, P. Duda, T. Johnson, C.
Weaver, M. Warren, and D. Nover. 2013. Watershed Modeling to Assess the Sensitivity of
Streamflow, Nutrient, and Sediment Loads to Potential Climate Change and Urban Development
in 20 U.S. Watersheds. EPA/600/R-12/058A. National Center for Environmental Assessment,
Office of Research and Development, U.S. Environmental Protection Agency, Washington,
DC. Public review draft, March 5, 2013. Available online at:
http://cfpub.epa.gov/ncea/global/recordisplav.cfm?deid=247495#Download.
Center for Neighborhood Technology. 2009. Green Values National Stormwater Management Calculator.
Available online at: http ://greenvalues .cnt.org/national/calculator.php.
EPA (U.S. Environmental Protection Agency). 1983. Results of the Nationwide Urban Runoff Program,
Volume 1 - Final Report. U.S. Environmental Protection Agency, PB84-185552, December 1983.
EPA (U.S. Environmental Protection Agency). 2008. Stormwater Management Model. User's Manual
Version 5. Revised in March 2008.
FDEP (Florida Department of Environmental Protection). 2006. TMDL Protocol.
Huff, F. and J. Angel. 1992. Rainfall Frequency Atlas of the Midwest. Prepared by the Midwestern
Climate Center and Illinois State Water Survey. Bulletin 71.
Kefer, S.J. 2013. La Crosse Area Climate Adaptation Study. Miscellaneous Publication PUB-SS-1119
2013. Bureau of Science Services, Wisconsin Department of Natural Resources Madison,
WI. Available online at: http://dnr.wi.gov/files/PDF/pubs/ss/SSl 119.pdf
Long-Term Hydrologic Impact Analysis Low Impact Development Version - 2.0. Developed by Purdue
University. Available online at: https://engineering.purdue.edu/~lthia/.
Low Impact Development Center. 2005. Low Impact Development for Big Box Retailers. Prepared for
U.S. Environmental Protection Agency
Southeast Michigan Council of Governments. 2008. Low Impact Development Manual for Michigan, A
Design Guide for Implementers and Reviewers. Available online at:
http://library.semcog.org/InmagicGenie/DocumentFolder/LIDManualWeb.pdf
Sustainable La Crosse Commission. 2012. La Crosse Climate Change Adaptation Workshop Report
Synopsis. February 28, 2012. Available online at: http://sustainablelacrosse.com/.
Tetra Tech, Inc. 2001. Low-Impact Development Management Practices Evaluation Computer Module
User's Guide. Prepared for Prince George's County, Department of Environmental Resources, by
Tetra Tech, Inc., Fairfax, VA.
USGCRP (U.S. Global Change Research Program). 2009. Global Climate Change Impacts in the Unites
States. Karl, T.R., J. M. Melillo, and T. C. Peterson (eds.). United States Global Change Research
Program. Cambridge University Press, New York, NY, USA. Available online at:
http: //www. globalchange. gov/.
23
-------
Washington State University. 2009. Bioretention Soil Mix Review and recommendations for Western
Washington. Prepared for Puget Sound Partnership. Prepared by Washington State University,
Pierce County Extension.
Weiss, P., J. Gulliver, and A. Erickson. 2005. The Cost and Effectiveness of Stormwater Management
Practices. University of Minnesota.
WERF (Water Environment Research Foundation). 2009. BMP and Low Impact Development Whole
Life Cost Models Version 2.0.
24
-------
Appendix A: Model Construction, Parameterization,
and Calibration
EPA's Storm Water Management Model (SWMM; EPA 2008) was used to develop a hydrologic and
hydraulic model of the Johnson Street Basin. SWMM is a dynamic hydrology-hydraulic-water quality
model that simulates runoff quantity and quality from urban areas. The runoff component operates on a
collection of sub catchment areas that receive precipitation and generate runoff and pollutant loads. The
routing component transports this runoff through a system of pipes, channels, storage/treatment devices,
pumps, and regulators. SWMM can track the quantity and quality of runoff generated within each sub
catchment, as well as the flow rate, flow depth, and quality of water in each pipe and channel. SWMM 5
can also model the hydrologic performance of several green infrastructure controls, including permeable
pavement, rain gardens, green roofs, street planters, rain barrels, infiltration trenches, and vegetated
swales.
Data required for this SWMM analysis include design storms to drive the model; catchment physical
characteristics (topography, soils, land cover) to construct and parameterize the model; and observed
flooding response to calibrate the model. This appendix describes the model construction,
parameterization, and calibration, including the data sources and key modelling assumptions.
1. Climate Data
Four design storm events were used to drive the Johnson Street Basin SWMM model:
• A 3-month, 2-hour Atlas storm event (Precipitation = 0.83 inches; Peak 1-hour depth = 0.67
inches) to represent existing flooding conditions
• A 1-year, 24-hour Atlas storm event (Precipitation = 2.23 inches; Peak 1-hour depth = 1.05
inches) to simulate a small, frequent storm event
• A 10-year, 24-hour Atlas storm event (Precipitation = 4.40 inches; Peak 1-hour depth = 2.07
inches) to represent design criteria listed in the La Crosse ordinance
• A 10-year, 2-hour climate change storm event (Precipitation = 2.86 inches; Peak 1-hour depth =
2.32 inches) to represent a potential short duration, high intensity climate change scenario.
The first three storm events were derived from the Rainfall Frequency Atlas of the Midwest (Atlas; Huff
and Angel 1992). The City currently references the Atlas as part of the City ordinance, and the 10-year,
24-hour design storm is used to size storm sewers. Each storm event was represented in the model with an
SCS type II distribution, which is the standard distribution for this region.
The City was concerned, however, that the current Atlas did not account for the impact of climate change
on recent storm events. Their observations suggest that intense short-duration storm events are currently
occurring at a higher frequency than expected based on historic Atlas data. The City therefore asked that a
10-year, 2-hour climate change storm event be developed to evaluate street flooding. The 10-year
frequency was chosen to match the frequency that the City currently uses for sizing storm sewers.
To develop a climate change storm event, this study consulted a recent EPA report on the potential
impacts of climate change and urban development on watershed response (Butcher et al. 2013). As part
of the report, a range of climate change scenarios was analyzed for rain gauges located across the country.
A-l
-------
Since the report did not include a rain gauge located in La Crosse, rain gauges near La Crosse were
selected for further evaluation. Based on comparisons of rainfall depths for a series of design storms, the
rain gauge located in Titonka, Iowa was identified as most representative of rainfall patterns in La Crosse.
Figure 1 compares the design storm depths for La Crosse derived from the Atlas to the design storm
depths for Titonka. At all recurrence intervals, rainfall depths are very similar for the shorter duration
events. Larger differences were identified for the 24-hour duration events, but since the goal was to derive
a 10-year, 2-hour rainfall event, this difference was determined to be acceptable and the Titonka rain
gauge was selected for use as a surrogate rain gauge for La Crosse.
1-hour 2-hour 3-hour 6-hour
Recurrence Interval
24-hour
La Crosse 5-year
I Titonka 5-year
La Crosse 10-year
I Titonka 10-year
La Crosse 25-year
iTitonka 25-year
Figure 1. La Crosse Atlas rainfall values compared to Titonka rain gauge.
The EPA report evaluated six potential climate change scenarios for the period 2040-2070 developed
from the North American Regional Climate Change Assessment Program's archive of dynamically
downscaled climate products (Butcher et al. 2013). Once a surrogate gauge was selected for La Crosse,
the 10-year, 2-hour storm event for each climate change scenario was compared to the 10-year, 2-hour
storm event under existing conditions (Table 1). To produce a conservative estimate of existing system
capacity, this study considered only the three climate change scenarios with the greatest percent increase
in rainfall depth for the 10-year, 2-hour storm event. The average percent change for these three scenarios
(12 percent) was then applied to the published Atlas 10-year, 2-hour rainfall depth (2.55 inches) to
develop the climate change scenario (2.86 inches).
Table 1. Comparing climate change scenarios for 10-year, 2 hour discrete storm events to existing conditions
Climate change
scenario
Existing conditions
crmcgcmS
hrm3_hadcm3
rcm3_gfdl_slice
gfdl_slice
rcm3_cgcm3
WRFG_ccsm
10-year, 2-hour storm event P (in)
(based upon 2-hour discrete storm events)
(inch)
2.08
2.36
2.21
2.25
1.92
2.37
2.28
Percent change from existing
conditions
(%)
0%
13%
6%
8%
-8%
14%
10%
Scenarios derived from North American Regional Climate Change Assessment Program
A-2
-------
A SCS type II distribution was applied to this rainfall depth similar to the existing condition rainfall. A 12
percent increase was applied to each time step. This assumption is reasonable due to the short duration of
the rainfall event. Table 2 presents the published values from the Atlas and how they were revised for the
10-year, 2-hour climate change rainfall event.
Table 2.10-year, 2 hour rainfall event depths for existing conditions, climate change scenario
Duration
15 minute
30 minute
1 hour
2 hour
10-year, 2-hour rainfall
(provided in the Atlas)
(inch)
1.19
1.63
2.07
2.55
10-year, 2-hour climate change scenario
(increase by 12-percent)
(inch)
1.33
1.83
2.32
2.86
2. Hydrologic Model Development
To estimate the amount of precipitation infiltrated, stored in depressions, and converted to runoff,
SWMM requires the user to define a set of hydrologic model parameters, including subcatchment
drainage area, overland flow paths, percent imperviousness, slope, Manning's n for pervious and
impervious areas, depression storage, and soil infiltration characteristics (e.g., curve number). To define
these hydrologic model parameters for the Johnson Street Basin, the following data sources were
consulted:
• Subcatchment delineation as provided by the City of La Crosse
• SSURGO (for soil data)
• LiDAR data (for topography and slopes)
• Aerial photography, land use/ zoning maps, and impervious area maps (for land cover)
Based on these data sources, a SWMM model was developed to represent the hydrologic response within
the Johnson Street Basin under existing conditions. The watershed is 769 acres in size, and the
downstream model extent is located at the confluence between the Mississippi River and the Johnson
Street Basin. The average subcatchment size is 3.4-acres.
Figure 2 presents the impervious area within the study area (67.2 percent impervious). The
imperviousness that was considered directly connected was defined during a site visit and adjusted as
needed during model verification.
The soil parameters were listed for the entire study as hydrologic soil group type B soils by reviewing the
available soil (SSURGO) data. The Horton infiltration equation was used in SWMM to represent soil
infiltration. Table 3 presents the model soil parameters used in SWMM. These values were derived from
discussions with city staff and by reviewing literature including FDEP 2006 and Tetra Tech 2001, for
applicable values.
A-3
-------
Impervious Surfaces (67.2%)
Figure 2. Johnson Street Basin watershed imperviousness.
Table 3. Soil parameters
Soil type
(hydrologic soil
group)
B
Maximum infiltration rate
(inch/hour)
9
Minimum infiltration rate
(inch/hour)
0.5
Maximum soil storage
volume (inch)
5
Uncalibrated water quality pollutant loads for total suspended solids, total phosphorus, and total
nitrogen were calculated in SWMM. Wash off pollutant loads were calculated by defining an event
mean concentration (EMC) for total suspended solids, total nitrogen, and total phosphorus for each land
use category, which was multiplied by the modeled runoff to estimate pollutant loads. Dynamic buildup
and wash-off for these pollutants were not modeled. Land use categories were assigned to each
modeled subcatchment using the City zoning map. Table 4 presents the land use categories and EMC
values utilized for this study which were selected from the Nationwide Urban Runoff Program summary
report (EPA 1983). The EMCs represent median concentration and are applied as constants for all storm
sizes.
A-4
-------
Table 4. Land use event mean concentrations
Land use
Residential
Non-residential
TSS (mg/l)
101
69
TP (mg/l)
0.383
0.201
TN (mg/l)
2.636
1.751
Source: EPA 1983
3. Hydraulic Model Development
A dynamic wave SWMM model was developed to represent the hydraulic routing of the hydrographs
generated from the SWMM hydrologic model within the Johnson Street Basin storm sewer network under
existing conditions (Figure 3). The model extents include almost every 12-inch diameter pipe in the
watershed as provided by the City and extends to the confluence of the Mississippi River and Johnson
Street Basin. The SWMM hydraulic model includes 424 links and nodes. Discussions with City staff
indicated that backwater impacts from the Mississippi River are considered minimal to negligible.
Therefore, the SWMM outfall boundary is modeled as a free flow boundary condition.
o Loading Points
o Modeled Nodes
12 inch pipes
15 to 21 inch pipes
— 24 to 42 inch pipes
— 48 to 72 inch pipes
£3 Modeled Basins
N
0 1,000
2,000
3,000
4,000
=lFeet
Figure 3. Modeled storm sewer network for the Johnson Street Basin.
A-5
-------
4. Model Debug and Verification
Model debugging and verification was performed to improve model credibility and confidence in results.
A series of quality control procedures were conducted to verify that the model is a reasonable
representation of the Johnson Street Basin and its stormwater system, including flow and volume checks.
Similarly, the SWMM model was checked to be free of various types of instabilities inherent to the
numerical calculation schemes the model uses to approximate real-world conditions. Due to the detail and
complexity of this SWMM model, flooded nodes were not modeled using overland flow paths that are
routed to other modeled nodes. Instead, each flooded node was modeled using a general ponded area
assumption (10,000 square feet per foot) that generally represents the area within the public right-of-way
between manholes, and the flooded runoff volume was routed into the modeled storm sewer network
when capacity was available.
City staff provided input on three separate occasions to ensure the model predictions matched historical
flooding conditions prior to completing model verification. City staff also indicated that short, intense
storm events were typically what led to observed flooding. The City proposed utilizing a 2-inch, 2-hour
storm event to compare to observed flooding problems. To test model response and sensitivity, a series of
short duration, high intensity storm events were identified by referencing the Atlas (Huff and Angel
1992). Storm events ranged from less than 1 inch (3-month, 2-hour storm) to greater than 2 inches (10-
year, 2-hour storm). Model results were compared to City reported flooding, and model parameters were
adjusted until there was reasonable agreement between the observed flooding conditions and model
results.
Reviewing the model results during the sensitivity analysis, the common flooded areas reported by the
City occurred when the peak intensity exceeded 0.50 inches over a 30-minute period, which is essentially
the same as the 2-inch, 2-hour storm the City suggested using for the model verification purposes. City
staff confirmed that the frequency, severity, and location of modeled flooding problems generally
corresponded to observed flooding problems, and concluded that the model provides a sufficiently
accurate approximation of observed conditions. Figure 4 presents a summary of the areas where the
existing storm sewer network floods during the 3-month, 2-hour (0.83 inch) storm event with a SCS type
II distribution and presents the cubic feet of flooding occurring at each location. Results for the 3-month,
2-hour rainfall event include 67 flooded manholes for a duration of 45 hours. The average peak flooding
depth is 2 inches for a total flood volume of 0.76 million gallons.
A-6
-------
Manholes Flooded: 67
Total Duration of Manholes Flooded: 45-hours
T Flooded Volume: 0.76 MG
Average Peak Flooding Depth: 2-inches
— _ ^Problem
Problem
Area"B"
Storm Sewers
3-Month, 2-Hour Event
Existing Conditions Flooding (ft3)
o up to 2,500
O 2,500 to 5,000
O 5,000 to 10,000
O 10,000 to 20,000
^ ' I ;jf
/ — n/iVj;, ft
, Winnebago St .
-Mistissippi-St-
O
N
4-
greater than 20,000
1,000
2,000
3,000
4,000
Barlow St
Figure 4. Existing conditions flooding, 3-month 2-hour event.
The most extensive flooding is concentrated in a few areas (Problem Areas A, B, and C in Figure 4).
Reviewing some of these problems led to the following observations regarding the hydrology and storm
sewer network:
• Problem Area A (King Street and Cass Street between 6th and 8th) has a very high
imperviousness (70 percent) and is mostly non-residential with several parking lots that do not
have stormwater control measures. The storm sewer network consists primarily of a 12-inch
diameter pipe, which appears to be significantly less than needed.
• Problem Area B (King Street and Cass Street between 13th Street and 15th Street) has relatively
high impervious area (greater than 50 percent) in a primarily residential area with only a modest
amount of parking lots. When visiting the study area, most of the residential houses appeared to
have disconnected rooftops, which helps reduce the peak flow. At the intersection of 15th Street
and Cass Street is a constricting 1.75-ft diameter pipe, which leads to frequent surcharging and
backup. Similarly, further downstream along 15th Street between Winnebago Street and State
Street are a couple of pipe segments where the pipe slope appears to be zero percent, which
results in pipe surcharging upstream.
• Problem Area C (Adams Street and Farnum Street between 18th Street and 20thStreet) is located
in a predominantly residential area with an average imperviousness of 50 percent. However, the
A-7
-------
City has intentionally undersized pipes along Adams Street and Farnum Street to help reduce the
flooding elsewhere. Along Farnum Street near the intersection with 16th Street, a 3-foot diameter
pipe along Farnum Street and a 3-foot diameter pipe along 16th Street intersect and discharge into
a single 3-foot diameter pipe, resulting in street flooding upstream.
A-8
-------
Appendix B: Additional Green Infrastructure Design
Parameters
Green infrastructure is modeled in SWMM using a combination of vertical layers whose properties are
defined on a per-unit-area basis. The entire catchment impervious area was routed to the green
infrastructure. The vertical layers can include a surface layer, pavement layer, soil layer, storage layer,
and underdrain layer. Depending on the physical composition of each green infrastructure practice,
various combinations of layers were applied. Figure 1 shows the vertical layers that can be used in
SWMM to represent green infrastructure practices. Table 1 shows the vertical layers that were used in this
analysis to represent bioretention and permeable pavement.
Each vertical layer within a bioretention or permeable pavement area is characterized by several design
parameters that determine the hydrologic function of the practice. These design parameters include the
depth of the layer, the porosity or void ratio, and other parameters that define the rate at which water can
flow through the layer. The green infrastructure design parameters assumed for this analysis are presented
in Table 2. Soil infiltration parameters required for the soil layer and storage layer were determined on the
basis of the assumed soil substrate. The infiltration rate into the bioretention media (6 inches per hour) is
derived from Washington State Department of Ecology for bioretention soil mixes with an assumed safety
factor of 2 (Washington State University 2009).
The bioretention soil layer shown in Figure 1 is modeled in SWMM using the Green Ampt equation.
While bioretention uses the soil layer for detaining stormwater, permeable pavement detains stormwater
in the storage layer. Referring to Table 1, neither bioretention nor the permeable pavement was modeled
in SWMM with an underdrain, since La Crosse soils have a high enough infiltration rate not to warrant an
underdrain.
Rainfall ET
Runon
r=*
^
f t
Surface Layer
t
Soil Layer '
Storage Layer
Infiltratk
n
it
Percolat
n
<>
on
Underdrain
Infiltration
Figure 1. General schematic of SWMM BMP layers (pavement layer is substituted for soil layer for permeable
pavement).
B-l
-------
Table 1. SWMM Vertical layers
BMP type
Bioretention
Permeable
Pavement
Surface
Yes
Yes
Pavement
NA
Yes
Soil
Yes
NA
Storage
Noa
Yes
Underdrain
No
No
a. Bioretention storage is included in soil layer
Table 2. BMP layer parameters
SWMM parameter
Unit
Bioretention
Permeable pavement
Surface layer parameters
Storage Depth
Vegetation Volume Fraction
(storage volume removed
due to vegetation volume)
Surface Roughness
Surface Slope
Swale Side Slope
inch
dimensionless
dimensionless
%
horizontal : vertical
12
0.1
0.35
1
2
0
0
0.011
1
NA
So/7 layer parameters
Thickness
Porosity
Field Capacity
Wilting Point
Conductivity
Conductivity Slope
Suction Head
inch
fraction
fraction
fraction
inch/hour
dimensionless
inches
36
0.4
0.25
0.1
6
7.5
2.4
NA
NA
NA
NA
NA
NA
NA
Pavement layer parameters
Thickness
Void Ratio (not porosity)
Impervious Surface Fraction
(impervious vs porous
surface)
Permeability
Clogging Factor
inches
fraction
ratio
inch/hour
NA
NA
NA
NA
NA
NA
6
0.2
0 (continuous)
100
NA
Storage layer parameters
Height
Void Ratio
Infiltration Rate
Clogging Factor
inches
fraction
inch/hour
NA
NA
0.67
1
NA
24 and 48
0.67
1
NA
B-2
-------
Appendix C: Performance Curves Supporting
Selection of the Most Effective Practice
3-month, 2-hour Performance Curves
System Flooded vs Green Infrastructure
P=3mo2hr(0.83in)
Bioretention (3-ft)
Permeable Pvmt (2-ft)
Permeable Pvmt (4-ft)
500 1,000 1,500 2,000 2,500 3,000 3,500
Green Infrastructure Area (1,000 SF)
System Flooded vs Green Infrastructure Implemented
P=3mo2hr(0.83in)
•Bioretention (3-ft)
'Permeable Pvmt (2-ft)
•Permeable Pvmt (4-ft)
Existing
18%
Q_
2
S- 8%
1/1
c:
0)
(
System Flooded vs Green Infrastructure Cost
P=3mo2hr(0.83in)
I — «— Bioretention (3-ft)
Y«L — »— Permeable Pvmt (4-ft)
\\
\*
\ V,
\\
^- V X LA A
) 10 20 30 40 50 60
Estimated 20yr Lifecycie Cost (SMillion)
Figure C-1. System flooded performance curves, 3-month, 2-hour storm event.
C-l
-------
Volume Flooded vs Green Infrastructure
P=3mo2hr(0.83in)
Bioretention (3-ft)
Permeable Pvmt (2-ft)
Permeable Pvmt (4-ft)
500 1,000 1,500 2,000 2,500 3,000 3,500
Green Infrastructure Area (l.OOOSF)
Volume Flooded vs Green Infrastructure Implemented
P=3mo2hr{0.83in)
Bioretention (3-ft)
Permeable Pvmt (2-ft)
Permeable Pvmt (4-ft)
0.0
Existing 25% 50% 75% 100% 100%+ 100%
Alley AlleY +
Scenario pki
Volume Flooded vs Green Infrastructure Cost
P=3mo2hr(0.83in)
0.7
t? 05
"g 0.5 -
LL
3J
D
[ —*— Bioretention (3-ft)
\ —*— Permeable Pvmt (2-ft)
\
ft — •— Permeable Pvmt [4-ft)
JL
I
\\
\\
\\
r
V ^*»SA *. ^ *. *.
0 10 20 30 40 50 60
Estimated 20yr Lifecycle Cost ($Million)
Figure C-2. Volume flooded performance curves, 3-month, 2-hour storm event.
C-2
-------
E
X
1
_o
1
1
Total Duration of Street Flooding
vs Green Infrastructure
P=3mo2hr{0.83in)
50 -T-
40
25
5 -
V -*- Permeable Pvmt (2-ft)
1\ — •— Permeable Pvmt (4-ft)
\\
\\\
4\
K
\x
Y> X . . .
0 500 1,000 1,500 2,000 2,500 3,000 3,500
Green Infrastructure Area (1,000 SF)
Street Flooded (Hours)
Total Duration of Street Flooding
vs Green Infrastructure Implemented
P=3mo2h(0.83in)
10
V\
flk\. —*— Permeable Pvmt (2-ft)
\\ \
\\ -S
\\ -S
\
V x .^~^
Existing 25% 50% 75% 100% 100% 100% +
+ Alley AlleV-
Scenario ™S
Total Duration of Street Flooding
vs Green Infrastructure Cost
P=3mo2hr(0.83in)
12
^
•o
1
^
OJ
»
Vw
V\
Vu
V\V
™ \\\
V ^v»
0 10
1^
_^
Permeable Pumt (2-ft)
Permeable Pvmt (4-ft)
20 30
40 50 60
Estimated 20yr Lifecycle Cost ($Million)
Figure C-3. Total duration of street flooding performance curves, 3-month, 2-hour storm event.
C-3
-------
Average Peak Depth of Street Flooding
vs Green Infrastructure
P=3mo2hr(0.83in)
C
" 1 fi*
c
Iu
^ 0.8 -
.c
Ol
Q nd
o.o
1 — «— Bioretention (3-fH
'K — »— Permeable Pvmt (4-ft)
\
Vv
vv
XX
^v ^^^^^
S 0 10 20 30 40 50 60
3.
Estimated 20yr Lifecycle Cost ($)
Figure C-4. Average peak depth of street flooding performance curves, 3-month, 2-hour storm event.
C-4
-------
10-year, 2-hour Performance Curves
System Flooded vs Green Infrastructure
P=10yr2hr(2.86in)
70%
"g 50% -
X \
-Bioretention (3-ft)
-Permeable Pvmt (2-ft)
-Permeable Pvmt (4-ft)
0 500 1,000 1,500 2,000 2,500 3,000 3,500
Green Infrastructure Area (1,OOOSF)
System Flooded vs Green Infrastructure Implemented
P=10yr2hr(2.86in)
Existing 25% 50% 75%
Scena
trio
100% 100% + 100% *
Alley AlleV+
Pkg
ystem Flooded (%)
L) .£» LH O1 -J
•s
a)
e
£ in0/
System Flooded vs Green Infras
P=10yr2hr(2.86in)
>-* l>it A
^^r^v
\ \
\
\
V
tructure Cost
—»— Bioretention (3-ft)
-*- Permeable Pvmt (2-ft)
— * — Permeable Pvmt (4-ft)
y^
\
*\
\
'^^^—Oi
0 10 20 30 40 50 60
Estimated 20yr Lifecycle Cost (SMillion)
Figure C-5. System flooded performance curves, 10-year, 2-hour storm event.
C-5
-------
Volume Flooded vs Green Infrastructure
P=10yr2hr(2.86in)
5" 12
"g 10
-a
§ 8
u_
O)
3 4
0 -
(
i^ — «— Bioretention (3-ft)
^V ^X. —*— Permeable Pvmt (2-ft)
V \
^S, Tk —•— Permeable Pvmt (4-ft)
». ^v
A \
\ X
-S X
N. *v^-
— ^-*— t-— ^^4
) 500 1,000 1,500 2,000 2,500 3,000 3,500
Green Infrastructure Area (1,OOOSF)
Volume Flooded vs Green Infrastructure Implemented
P=10yr2hr(2.86in)
14
1. "
T3
§ 8
u_
p ft
1 4
0 -
EM
^^^^z^^^ — *~ f ermeable Pvmt (2-ft)
x >r~^-4^,_
N^ "X^ ^**-^^^ — •— Permeable Pvmt (4-ft)
\ V ^
\ V
V x__^
^~~^_ t^^^t
sting 25% 50% 75% 100% 100%+ 100% +
Alley AlleV +
Scenario pk«
Volume Flooded vs Green Infrastructure Cost
P=luyr2hr(2.86in)
14 ^
(3 -tj
1
oJ 1U
-o
o 8
U-
QJ
^
-§ 4 -
»^ t Bioretention (3-ft)
^^J~S. -*- Permeable Pvmt (2-ft)
A/^. Ti. —*— Permeable Pvmt (4-ft)
\ V ^v
\ V
\ ^
5 s:
v ^s^
^-^_\.
0 10 20 30 40 50 60
Estimated 20yr Lifecycle Cost ($Million)
Figure C-6. Volume flooded performance curves, 10-year, 2-hour storm event.
C-6
-------
Total Duration of Street Flooding
vs Green Infrastructure
P=10yr2hr(2.86in)
¥
"O
OJ
o
LJL
V,
\^
—4 — Bioretention (3-ft)
-*- Permeable Pvmt (2-ft)
— «— Permeable Pvmt (4-ft)
\ ^
V ^
>^
^^^.
0 500 1,000 1,500 2,000 2,500 3,000 3,500
Green Infrastructure Area (l.OOOSF)
Total Duration of Street Flo
vs Green Infrastructure Imple
P=10yr2hr(2.86in)
500 i
i
1 400
•O
~g 300
LJ_
3d ing
mented
— »— Bioretention (3-ft)
—*- Permeable Pvmt (2-ft)
— •— Permeable Pvmt (4-ft)
X^^T"""^*
\ ^
*\ ^*— -^
^^— ^\,
Existing 25% 50% 75%
Scenario
iocm 100%+ 100%+
Alley Alley +
Pkg
Total Duration of Street Flooding
vs Green Infrastructure Cost
P=10yr2hr{2.86in)
600 -i-
3
T3
OJ
O
LJ_
|200
t/1
— »— Bioretention (3-ft)
^^^v^. — *— Permeable Pvmt (2-ft)
x<\
^>^,X. — •— Permeable Pvmt (4-ft)
x^^V
\ ^
^V ^^^
^^^
0 10 20 30 40 50 60
Estimated 20yr Lifecycle Cost (SMillion)
Figure C-7. Total duration of street flooding performance curves, 10-year, 2-hour storm event.
C-7
-------
Average Peak Depth of Street Flooding
vs Green Infrastructure
P=10yr2hr(2.86in)
I 91
DD
"O
O
u_
OJ
^ s
•s 4
a 3 .
o
O)
QJ
^^ "S^ -*- Permeable Pvmt (2-ft)
^^ *v —•— Permeable Pvmt (4-ft)
*S. >-
\ >v
"V^ "^-^^^
^^V_ ^^^"^^^-^
^^^ ^^^^
^"~*'-J*" «.
ra 0
> 0 500 1,000 1,500 2,000 2,500 3,000 3,500
Green Infrastructure Area (l.OOOSF)
O
o 7
LL '
I" 3
Q
-^ 2
01
QJ
00
Average Peak Depth of Street Flodding
vs. Green Infrastructure Implemented
P-10yr2hr(2 86in)
—•—Bioretention (3- ft)
"^ ~^^-~^ ~*~ Permeable Pvmt (2-11)
^^ ^S^ ^^^^ < Permeable Pvmt (4-ft)
>. -^ ^"""^^.^
\ \ *
^V "5^^^^
^^'v^ ^* ^v^
^*»^^^^ ^^
^ • ' •
HI
^ Existing 25% 50% 75% 100% 100%+ 100% +
Alley Alley +
Scenario ^
Average Peak Depth of Street Flooding
vs Green Infrastructure Cost
P=10yr2hr(2.86in)
10
~ 9 i
5B
u_
0 4 -
Q. :>
Q
n
S. ! .
&
^*^\**\. A Permeable Pvmt (2-ft)
^O*w *V —•—Permeable Pvmt (4-ft)
X. T ^^.
V \.
V^ ^*>v^^
*\^^ ^^T»r-^^^^
^v^^ ^^^A
S^-^1 ^
1 10 20 30 40 50 60
Estimated 20yr Lifecycle Cost ($)
Figure C-8. Average peak depth of street flooding performance curves, 10-year, 2-hour storm event.
C-8
------- |