towar rain garden drains: a michigan urban retrofit

8/3/2019 Towar Rain Garden Drains: A Michigan Urban Retrofit

http://slidepdf.com/reader/full/towar-rain-garden-drains-a-michigan-urban-retrofit 1/17

1

Towar Rain Garden Drains: A Michigan Urban Retrofit Low Impact Stormwater Management System

Patrick E. Lindemann 1, Carla Florence Clos 2, Brian Cenci 3, P.E., Larry Protasiewicz 4, P.E., and

Tim Inman5

, P.E., P.S. 1 Ingham County Drain Commissioner, 707 Buhl Avenue, Mason, MI 48854; [email protected] Ingham County Deputy Drain Commissioner, 707 Buhl Avenue, Mason, MI 48854; [email protected] 3Senior Project Engineer, Fitzgerald Henne & Associates, Inc., 4063 Grand Oak Dr., Lansing, MI 48911;cencib@ fitzhenne.com 4 Project Manager, Spicer Group, Inc., 110 W. Michigan Avenue, Suite 725, Lansing, MI 48933;[email protected] 5 Project Engineer, Spicer Group, Inc., 110 W. Michigan Avenue, Suite 725, Lansing, MI 48933;[email protected]

ABSTRACT: Near Lansing, Michigan, long-standing flooding problems persisted in Towar Gardens, a 200-acre residential neighborhood with over 400 homes. After minor rainfall events,

backyard and basement flooding with sewage back-ups would occur. The Ingham County DrainCommissioner (ICDC) was petitioned to bring drainage relief. A retrofit Low ImpactStormwater Management System (LISMS) was found ideal for this nearly built-out, older urbanneighborhood. The Towar Rain Garden Drains (TRGD) utilized rain gardens and bioswales tocollect and filter stormwater that was conveyed downstream through concrete road pipes. Intotal, 8.25 miles of new drains were constructed, including 5.62 acres in 150 separate raingardens and bioswales. With infiltration into the engineered bioretention soil profiles andevapotranspiration from native vegetation, 100-year storms can be drained by outlet pipes nolarger than 24-inches. The TRGD is commonly referenced as the largest LISMS retrofit projectconstructed in Michigan and its cost, just over $9.8 million, was half the cost of a traditionalstormwater system. Ongoing monitoring and testing have suggested that flooding has beeneliminated and that pollutant loads, peak discharge, and volume have been reduced. Despite theeffectiveness of LISMS, like TRGD, its many construction and maintenance challenges can bedaunting for public projects.

INTRODUCTION

Towar Gardens is an area in Ingham County, Michigan, near the state capital of Lansing, thatwas originally settled in the 1800s at a time in the history of the State when settlers were beingwarned “Don’t go to Michigan, that land of ills, the word means ague, fever and chills.” As itturned out, Towar Gardens was aptly named, for it once was an area of low-lying, richagricultural land made productive in part by an historic and ambitious extensive network of drainage canals that drained a 7000-acre marsh known as the Chandler Marsh--“the BigMarsh.” This network of canals was to become the basis of the Remy Chandler IntercountyDrain system that still outlets water, including that from the Towar Gardens area, to the Looking



2

Glass River, and eventually to Lake Michigan. The history of Michigan settlement is repletewith such stories of drain infrastructure making land suitable for habitation and economicuse. Today, the 200-acre area of Towar Gardens is almost exclusively residential, with over 400

platted single-family homes, and several multi-family and commercial developments. The areais nearly built-out now, with the majority of dwellings having been constructed either during the

1920s -1940s or in the 1980s. Due to the modest size of the homes and lots within thisneighborhood, it is considered to be one of the most affordable housing opportunities in the twoaffluent communities in which it is located, Meridian Township and the City of East Lansing.

Despite the formal platting of most of the land and its unique history connected to a massiveturn-of-the-century drainage project, the drainage system in Towar Gardens remained inadequatein size, type, and condition to accommodate the changes in runoff that accompanied the shiftover time from agricultural to residential land-use. The lack of an adequate collector system aswell as the lack of an organized connection to the outlet in the Remy Chandler IntercountyDrain, combined with site constraints such as low-lying, level topography and poorly-drainedsoils, resulted in a decades long history of widespread and frequent flooding even after minor

rainfall events of less than one-year storms. Localized ponding could remain in yards for dayseven after 30-minute rainfall events. Roads were in poor repair because of saturated sub- bases. It was not uncommon for sump pumps to run almost constantly, simply re-circulatingwater from basements to yards and then back into foundation drains. Illegal connections of sump

pumps to the sanitary system (the only real available outlet) and inflow and infiltration resultedin a history of sewage back-ups. Resident complaints over the loss of private property and theunhealthy conditions from stagnant water were frequent and long-standing. Finally, with

problems exacerbated by a series of wet years from 2000-2003 heightening the concerns of public officials, one of the municipalities petitioned the Ingham County Drain Commissioner for drainage improvements.

In Michigan, drainage is administered under the Drain Code, Public Act 40 of 1956, whichvests in an elected county official, the County Drain Commissioner, jurisdiction over publicdrains, open and closed, and over improvements of those drains. Although the County DrainCommissioner must be petitioned by landowners or municipalities to initiate a new drain project,the County Drain Commissioner is then given great power and discretion to act to improvedrainage and solve problems when found necessary. This power reflects the historic importanceof drainage in Michigan. Importantly, under the Drain Code, County Drain Commissioners areallowed to assess for the drainage that is provided based on a principle of benefits derived. This

power to assess, often absent from other public works laws, gives the County DrainCommissioner a clear and direct funding source for both the construction of the infrastructure, aswell as importantly, for its continued maintenance.

Early on, it became clear to the County Drain Commissioner why nothing had been done tocorrect the drainage problems in Towar Gardens despite decades of complaints. Retrofitting anadequate drainage system that was affordable and practical into such a largely built-out, older neighborhood with so many site constraints such as low-lying, flat topography, poorly-drainedsoils, downstream outlet elevation limitations for gravity flow, lack of land for detentionfacilities without removal of affordable housing stock, and multiple utility conflicts in narrowrights-of-way; were indeed challenges for a traditional stormwater management system



3

approach. Overcoming these engineering challenges made estimates of the cost of constructionof a traditional drain project (large gravity flow pipes and ponds) along with other alternatives,such as pumping, well beyond what would be acceptably affordable for this modestneighborhood. Estimates of construction of a traditional stormwater system were close to $20million. Moreover, there was even some question as to whether it would be feasible to construct

a traditional drain project given the environmental permitting necessary to alter downstreamopen channel elevations in the Remy Chandler Intercounty Drain, required for a gravity systemto function given the size of pipes needed for this drainage area. In addition, the neighborhoodwas located in a Federal Clean Water Act Phase II area and any construction would have to meetall Federal water quality discharge requirements along with the requirements of the IntercountyDrainage Board for discharge to the Remy Chandler Intercounty Drain.

The very issues that made a traditional drainage approach untenable and cost prohibitivethough, made a Low Impact Stormwater Management System (LISMS) approach ideal. Relyingon an approach that more mimicked pre-development hydrological conditions could reduce thehydraulic gradient throughout the entire drain system by nearly three feet because the peak

discharge and time of concentration of individual catchment areas within the watershed could besignificantly reduced. Detention could be dispersed in bioretention areas such as rain gardensand bioswales throughout the system within the narrow, 50-foot road rights-of-way or onindividual residential properties, thus avoiding the need for removal of homes or use of vacantlots for constructing large ponds, something that would have been negatively received by thecohesive Towar Gardens neighborhood. Moreover, rain gardens and bioswales could bedesigned around existing infrastructure and land-uses. Increasing the residence time usingsource control could also allow conveyance pipes to be downsized so that the existing outletelevations at the Remy Chandler Intercounty Drain could still accommodate drainage by gravityfor this flat neighborhood (approximately 6-feet of fall over 3500 ft). Furthermore, prolongingresidence times at the surface of the rain gardens with infiltration through an engineered soil

profile could provide removal of the typical urban pollutants to meet water quality standardswithout expensive additional mechanical equipment. Early cost estimates for construction of aLISMS were half that of the estimates for construction of a traditional stormwater system.

OVERVIEW TOWAR RAIN GARDEN DRAINS

The Towar Rain Garden Drains (TRGD) were designed to function primarily as a LISMS thatdetains and releases stormwater at a lower rate downstream, reducing peak discharge rate, while

providing for volume control at the source, and removing typical urban pollutants such as totalsuspended solids, phosphorus, and nitrogen; while still accommodating the typical storm andrunoff events accommodated by more traditional public drain systems (10-year, 24-hour firstflush rain events and 100-year, 24-hour storm detention). Furthermore, far less land, 4.6% of theland drained, was needed to meet these storm and runoff standards using bioretention areas (raingardens and bioswales) with source control than would have been necessary with a moretraditional system detention ponds (15-20% typically). The TRGD system consisted of 8.25miles of drain infrastructure, including 5.62 acres in 150 individual underdrained rain gardensand non-underdrained bioswales, which, along with regular roadside ditches, serve as the



4

collection system that convey stormwater to road pipes (24-inch or less) that ultimately connectto an outlet in the Remy Chandler Intercounty Drain. Each rain garden was engineered with asegregated soil and aggregate medium of a minimum of 6-inches of topsoil and compost mix, 6-inches of Michigan Department of Transportation (MDOT) 2NS sand with a layer of filter fabricseparating that layer from 12-inches of MDOT 2G stone above and at least 12-inches of MDOT

2G stone either side of a 12-inch perforated pipe which connects to the road pipes (see FIG. 1 below). Overflow structures connected to the underdrains were incorporated into every raingarden for when the 10-year, 24-hour duration design rain event was exceeded. Every lot was

provided a sump pump lead. Plant selection for the bioretention areas was a criticalelement. Plants were selected for development of roots for infiltration, hardiness, lowmaintenance, compatibility with location adjacent to roads (salt tolerance and height limitations),tolerance for fluctuating hydrological conditions, and aesthetics. Generally native plant specieswere selected because they met most of the criteria the best. A few bioretention areas were

planted with turf grass due to landowner allergies. Bioretention areas were initially planted with111 pounds of 95 different species of native wildflower seeds and 52,000 plugs of 125 differentspecies of native wildflower forbs. Construction was a massive undertaking with every road in

the neighborhood torn up at some point and, because of the source control, nearly every propertyimpacted by construction of at least one rain garden, bioswale, ditch, and/or pipe. The projectcommenced in March 2006 and finished late 2007. The cost of the TRGD project was just over $9.8 million.

FIG. 1. Typical engineered rain garden cross section.



5

PERFORMANCE POST CONSTRUCTION

There have now been over three and half years of experience post construction with theTRGD. Some 420 rainfall events have occurred during this time, producing approximately 104

inches of precipitation (see Table 1 below). No flooding has been observed and no drainage or flooding complaints have been received during this time. The project has clearly addressedflooding issues as intended for the one-year through the twenty-five year events that haveoccurred during this period of time. This is in contrast with prior to project construction, when alandowner survey revealed that 64% had experienced drainage problems and 13.5% experienced

basement flooding, along with the many complaints of flooding of structures, streets and/or property submitted to the Drain Office after even minor rainfall events.

Table 1. Frequency Analysis of Precipitation Events Post TRGD Construction

Beginning date of analysis January 1, 2008

Ending date of analysis August 10, 2011Total number of events analyzed 420Total precipitation during period 104 inchesTotal number of events not greater than 1 year return 416Total precipitation of events not greater than 1 year return 90.6 inchesTotal number of events greater than 1 year return 4Total precipitation of events greater than 1 year return 13.4 inches

Number of 1 year return events (100% Chance Recurrence) 5 Number of 5 year return events (20% Chance Recurrence) 1 Number of 10 year return events (10% Chance Recurrence) 2 Number of 25 year return events 4% Chance Recurrence) 1

Number of 50 year return events 2% Chance Recurrence) 0 Number of 100 year return events (1% Chance Recurrence) 0

Prior to construction, landowners often reported to the Drain Office that sump pumps ran “allthe time”. Excessive sump pump demand (and the costs associated with that demand) was onlyone of the negative consequences of the high groundwater table prior to the TRGD. Other negative consequences included flood damage to basements and their contents, dehumidifier operation costs, foundation damage and propagation of hazardous molds and mildews withinliving spaces. Post construction anecdotal reports from homeowners indicated that sump pumpusage had greatly decreased, although few could quantify the change. However, twohomeowners who had rain gardens constructed on their properties actually tracked sump pump

usage before and after the TRGD project, keeping careful detailed records. Their sump pumprecords allowed for the projection of annual operating and replacement cost savings from lower power consumption and extended sump pump life expectancy (see Table 2 below). Analysis wasalso done for dehumidification costs on an annual basis (see Table 3 below). Each of thesehomeowners could be expected to experience total savings of approximately $350 annually with

just the savings from the sump pump and dehumidifier. The annual assessment for theconstruction of the TRGD for each of these homes is approximately $255.



6

Table 2. Annual Sump Pump Operating and Replacement Cost Savings

BeforeProject

After Project Savings

Cycles per hour

(30 second duration, ½ horsepower pump) 30 6 24Annual energy cost(assuming $0.10/kWh) $82 $16 $65Annualized life cycle replacement cost $165 $13 $153Total per year (one home) $247 $29 $218

Table 3. Annual Dehumidifier Operating and Replacement Cost Savings

BeforeProject

After Project Savings

Dehumidifier operation (min./hr.) 45 15 30Days of operation (assuming 300 sq.ft. basement) 180 180 0Energy cost (assuming $0.10/kWh) $162 $54 $108Annualized life cycle replacement cost $30 $6 $24Total per year (one home) $192 $60 $132

FLOW AND WATER QUALITY PERFORMANCE TESTING – NATURAL EVENTS

In 2011, flow and water quality data were collected for twelve monitored rainfall events (seeTable 4 below) in three different rain gardens (20 different sample sets) using ISCO 3700 andISCO 6700 auto samplers and ISCO 2150 flow meters. Influent and effluent flows weremonitored with ISCO sensors on a 5-minute interval. Water quality samples were collected for

both the surface water influent and the tile water discharge from each location on a 4-hour timeinterval. The water samples collected were analyzed for Total Suspended Solids 6 (TSS), TotalPhosphorous 7 (TP), and Total Nitrogen 8 (TN).



7

Table 4. Monitored Natural Events

Date FrequencyRainfall

(in)Pollard

AveGibson

AveHartAve

April 20, 2011 6 month, 2 hour 0.98 x

April 26, 2011 <2 month, 3 hour 0.07 xApril 27, 2011 <2 month, 8 hour 0.34 xMay 10, 2011 <2 month, 2 hour 0.02 xMay 12, 2011 <2 month, 2 hour 0.25 xMay 13, 2011 10 year, 4 hour 2.32 xMay 25, 2011 <2 month, 2 hour 0.51 xMay 26, 2011 <2 month, 4 hour 0.20 xMay 29, 2011 2 month, 2 hour 0.68 x x xJune 16, 2011 2 month, 30 minute 0.45 x x xJune 21, 2011 <2 month, 2 hour 0.23 x x xJune 22, 2011 <2 month, 2 hour 0.26 x x x

Flow Data Results and Discussion

All of the twelve monitored rainfall events produced runoff into the rain gardens. Eight of thetwenty sample sets, however, produced no effluent from the rain gardens, in other words, the raingardens completely held the runoff from the event. All of the rain gardens proved to be effectivein reducing the peak discharge, ranging from 100% (where the event was completely held) to34% reduction (the event that exceeded the capacity of the rain garden and flowed out the weir was not included). The average of the peak discharge reduction for the events that dischargedthrough the rain garden was 81%. Hydrographs taken at the Pollard Avenue rain garden for theApril 20 th and June 16 th events are shown in FIG. 2 and FIG. 3. These hydrographs were fairlytypical of the flow data testing.

Peak discharge on the April 20 th event was reduced from 0.5 cfs to 0.15 cfs. The total volumedischarged from the rain garden was reduced by as much as 80%. During the June 16 th event,there was a peak discharge reduction of 50%, although there was not a significant reduction inthe overall volume of water discharged. It appears that peak discharge reduction is rather consistent among the events, however the total volume of discharge is more a function of theantecedent conditions and other site factors at the time of the event, the intensity of the storm,and the elevation of the overflow weir rather than the overall volume of rainfall.



8

FIG. 2. Hydrographs of April 20th rainfall event at the Pollard Avenue rain garden.

FIG. 3. Hydrographs of June 16th rainfall event at the Pollard Avenue rain garden.



9

Water Quality Results and Discussion

The Pollard Avenue rain garden was the largest garden equipped with monitoring equipmentand it proved to have the highest efficiency in removing TSS and TP. The first flush removal of

TSS for the eight events averaged 94%, with a range from 80 to 100% (see FIG. 4.). The firstflush removal of TP for the eight events averaged 71%, with a range from 29 to 94% (see FIG.5.). The first flush removal of TN for the three events that were measured (and that were notconfounded by antecedent conditions such as June 22nd) averaged 27%, with a range from 15 to38% (see FIG.6.). TSS and TP removal efficiencies over time are shown in FIG. 7 and FIG. 8.It appears that a nutrient that is attached to the sediment (TP) is removed to a greater extent thana water soluble nutrient (TN).

FIG. 4. TSS first flush removal efficiency of the Pollard Avenue rain garden.

0%10%20%30%40%50%60%70%80%90%

100%

A p r i l 1 9 t h

, 2 0 1 1

M a y 1 3 t h

, 2 0 1 1

M a y 2 5 t h

, 2 0 1 1

M a y 2 6 t h

, 2 0 1 1

M a y 2 9 t h

, 2 0 1 1

J u n e 1 6 t h

, 2 0 1 1

J u n e 2 1 s t , 2 0 1 1

J u n e 2 2 n d

, 2 0 1 1

T S S R e m o v a l

Pollard Avenue TSS Removal

0%10%20%30%40%50%60%70%80%90%

100%

A p r i l 1 9 t h

, 2 0 1 1

M a y 1 3 t h

, 2 0 1 1

M a y 2 5 t h

, 2 0 1 1

M a y 2 6 t h

, 2 0 1 1

M a y 2 9 t h

, 2 0 1 1

J u n e 1 6 t h

, 2 0 1 1

J u n e 2 1 s t , 2 0 1 1

J u n e 2 2 n d

, 2 0 1 1

T P R e m o v a l

Pollard Avenue TP Removal



10

FIG. 5. TP first flush removal efficiency of the Pollard Avenue rain garden .

FIG. 6. TN first flush removal efficiency of the Pollard Avenue rain garden.

FIG. 7. TSS removal efficiency of the Pollard Ave. rain garden over time-April 20 th event.

-20%

0%

20%

40%

60%80%

100%

A p r i l 1 9 t h

, 2 0 1 1

M

a y 1 3 t h

, 2 0 1 1

M

a y 2 5 t h

, 2 0 1 1

M

a y 2 6 t h

, 2 0 1 1

M

a y 2 9 t h

, 2 0 1 1

J u n e 1 6 t h

, 2 0 1 1

J u n e 2 1 s t , 2 0 1 1

J u

n e 2 2 n d

, 2 0 1 1

T S S R e m o v a l

Pollard Avenue TN Removal

1570

483

148 144 11620 6 7 6 1

99% 99%

95%96%

99%

75%

80%

85%

90%

95%

100%

0

200

400

600

800

1000

1200

1400

1600

1800

T S S R e m o v a l E f 5 i c i e n c y ( % r e m o v e d )

T S S C o n c e n t r a t i o n m g / L

In@luent TSS

Ef@luent TSS

TSS Reduction



11

FIG. 8. TP removal efficiency of the Pollard Ave. rain garden over time - April 20 th event.

FLOW AND WATER QUALITY PERFORMANCE TESTING - SYNTHETIC EVENTS

Synthetic flow events were designed to determine the hydraulic loading capacity of four raingardens (three planted with perennials and one planted with turf grass) and a ditch planted withturf grass (see Table 5). Once the loading capacity was determined, nutrients were added to theinfluent water to determine the removal efficiency for TP and TN at its hydraulic loading

capacity.

Table 5. Synthetic Event Rain Garden and Ditch Types and Locations

Type Location DateRain Garden - Perennial Gibson Avenue June 15, 2011Rain Garden - Perennial Hart Avenue June 27, 2011Rain Garden – Perennial Pollard Avenue July 12, 2011

Rain Garden – Turf Biber Street July 8, 2011Ditch - Turf Gibson Avenue July 14, 2011

Municipal water was used for each synthetic event. The water was free from total suspendedsolids. ISCO 6700 series flow samplers and ISCO 2150 flow meters were used to determine theinfluent and effluent flow rates and to take water samples during the test. First, each of the fivetest locations were loaded with clean water to determine the influent rate at which water could beeffectively treated prior to discharge over the overflow catchbasin or out the downstream end of the system without treatment. Then phosphorus and nitrogen were added to the influent waters.Each site was then loaded with the nutrient rich water and its concentration was measured at the

7.8

2.69

1.03 0.972 0.8580.535 0.501 0.372 0.267 0.229

93%

81%

64%73%

73%

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

0

1

2

3

4

5

6

7

8

9

10

T P R e m o v a l E f 5 i c i e n c y ( % r e m o v e d )

T P C o n c e n t r a t i o n m g / L

In@luent TP

Ef@luent TP

TP Reduction



12

influent and effluent locations. The following is a more detailed description of the test at eachlocation.

Rain Garden – Perennial Gibson Avenue : Metered water was added to determine thehydraulic loading capacity. Then 58 mg/L of phosphorus and 38 mg/L of nitrogen were added to

the influent water in the form of Mono-ammonium Phosphate (MAP), 11% nitrogen and 52% phosphorus. A total of 1100 gallons of nutrient loaded water were added to the garden for a period of 20 minutes. Both influent and effluent waters were monitored for flow rate andnutrient concentrations. A total of 5 influent samples and 7 effluent samples were taken (seeFIG. 9).

Rain Garden – Perennial Hart Avenue: Metered water was added to determine thehydraulic loading capacity. Based upon the rather high concentrations of nutrients experiencedat the Gibson Ave. basin it was determined to lower the concentrations of nutrients in theinfluent water by using a chemical feed metering pump and adding the nutrients at a lower metered rate. Then 5 mg/L of phosphorus and 1.1 mg/L of nitrogen were metered into the

influent water in the form of MAP, 11% nitrogen and 52% phosphorus. A total of 1800 gallonsof nutrient loaded water were added to the garden for a period of 46 minutes. Both influent andeffluent waters were monitored for flow rate and nutrient concentrations. A total of 9 influentsamples and 11 effluent samples were taken (see FIG. 10).

Rain Garden – Perennial Pollard Avenue: Metered water was added to determine thehydraulic loading capacity. Based upon the rather high concentrations of nutrients experiencedat the Gibson Ave. basin it was determined to lower the concentrations of nutrients in theinfluent water by using a metering pump and adding the nutrients at a lowered metered rate.Then 34 mg/L phosphorous and 35 mg/L nitrogen were added to the influent water in the form50% MAP (11% nitrogen, 52% phosphorus) and 50% Urea (46% nitrogen). A total of 5100gallons of nutrient loaded water were added to the garden for a period of 41 minutes. Bothinfluent and effluent waters were monitored for flow rate and nutrient concentrations. A total of 10 influent samples and 11 effluent samples were taken (see FIG. 11).

Rain Garden – Turf Biber Street: Metered water was added to determine the hydraulicloading capacity. Based upon the rather high concentrations of nutrients experienced at theGibson Ave. basin it was determined to lower the concentrations of nutrients in the influentwater by using a metering pump and adding the nutrients at a lowered metered rate. Then 5.0mg/L phosphorous and 5.5 mg/L nitrogen were added to the influent water in the form 50%MAP (11% nitrogen, 52% phosphorus) and 50% Urea (46% nitrogen). A total of 960 gallons of nutrient loaded water were added to the garden for a period of 41 minutes. Both influent andeffluent waters were monitored for flow rate and nutrient concentrations. A total of 10 influentsamples and 14 effluent samples were taken (see FIG. 12).

Ditch – Turf Gibson Avenue: Metered water was added to determine the hydraulicloading capacity. Based upon the rather high concentrations of nutrients experienced atthe Gibson Ave. basin it was determined to lower the concentrations of nutrients in theinfluent water by using a metering pump and adding the nutrients at a lowered meteredrate. Then 2.5 mg/L phosphorous and 7.0 mg/L nitrogen were added to the influent water



13

in the form 50% MAP (11% nitrogen, 52% phosphorus) and 50% Urea (46% nitrogen). Atotal of 2600 gallons of nutrient loaded water were added to the garden for a period of 38minutes. Both influent and effluent waters were monitored for flow rate and nutrientconcentrations. A total of 8 influent samples and 8 effluent samples were taken (see FIG.13).

Based upon the results obtained from the synthetic events, it appears that several generalstatements can be made. First, rain gardens planted with perennials have reduced peak flow,increased duration of time of storage and effective treatment of TP and TN. Second, raingardens planted with turf grass have limited capacity to reduce peak flow, are not effective for altering time of storage and limited effectiveness for treatment of TP and TN. Third, roadsideditches planted with turf grass did not reduce peak flow, did not increase duration of time of storage and had no effectiveness at treatment of TP and TN.

FIG. 9. Rain Garden – Perennial Gibson Avenue.

57.80

21.00

44.6

4.2

0.00

10.00

20.00

30.00

40.00

50.00

60.00

70.00

0 10 20 30 40 50 60 70 80

N u t r i e n t C o n c e n t r a t i o n ( m g / L )

Time (minutes)

Rain Garden – Perennial Gibson Avenue

Inf. TP Conc.

Eff. TP Conc.

Inf. TN Conc.

Eff. TN Conc.



14

FIG. 10. Rain Garden – Perennial Hart Avenue.

FIG 11. Rain Garden – Perennial Pollard Avenue.

2.08

0.59

0.00

0.50

1.00

1.50

2.00

2.50

0 10 20 30 40 50 60


Time (minutes)

Rain Garden – Perennial Hart Avenue

Inf. TP Conc.

Eff. TP Conc.

3.47

2.15

8.30

4.97

0.00

1.00

2.00

3.00

4.00

5.00

6.00

7.00

8.00

9.00

0 10 20 30 40 50 60


Time (minutes)

Rain Garden – Perennial Pollard Avenue

Inf. TP Conc.

Eff. TP Conc.

Inf. TN Conc.

Eff. TN Conc.



15

FIG 12. Rain Garden – Turf Biber Street

FIG. 13. Ditch – Turf Gibson Avenue.

3.64

1.12

5.68

6.31

0

2

4

6

8

10

12

0 10 20 30 40 50 60 70 80


Time (minutes)

Rain Garden – Turf Biber Street

Inf. TP Conc.

Eff. TP Conc.

Inf. TN Conc.

Series4

0.00

1.00

2.00

3.00

4.00

5.00

6.00

7.00

8.00

0 5 10 15 20 25 30 35 40 45


Time (minutes)

Ditch - Turf Gibson Avenue

Inf. TP Conc.

Eff. TP Conc.

Inf. TN Conc.

Eff. TN Conc.



16

GENERAL DISCUSSION AND CONCLUSION

Clearly, the LISMS retrofit in Towar Gardens is successfully managing the stormwater,exceeding all project goals for cost-effectively eliminating flooding, removing pollutants, andreducing volume and peak discharge. The low impact design of the Towar Rain Garden Drainsoffered an opportunity to overcome the complex problems of retrofitting new infrastructure intoan older, urban neighborhood and still meet modern standards for both water quantity and water quality management. Indeed, the experience in Towar Gardens is not unlike that of other LISMS

projects. So then, given the success of projects like the TRGD, why aren’t there more such public drain projects? This is the real conundrum of LISMS. Or is it? This project gave insightinto a number of “hidden” challenges (besides the engineering related challenges discussedearlier) that must be dealt with for a large-scale LISMS project to come to fruition.

The greatest challenge is the relative “newness” of LID. The lack of knowledge andexperience with LID still contributes to a neophobia on behalf of those practitioners in the

business of development, those public officials making decisions about infrastructure, and thoseregulatory agencies that must approve designs, not to mention the public who will live with the

project. There is a fear among engineers that a rain garden may not function like a catch basin or that they don’t understand design elements involved and have no experience with it, so theyreject it. Regulatory agencies often have rules that constrain the use of LID. For example,initially when the TRGD were reviewed by the Ingham County Road Commission, their standards for 22-foot wide roads with 3-foot gravel shoulders on each side would have made theuse of rain gardens along the roadways in the 50-foot rights-of-way completely impossible.Also, anytime there is something new or different, there are questions regarding increased costsor complexity. Public officials making the decisions about infrastructure must have the politicalwill to stick with a design that has not had as much history as a more traditional design. At timesit seems easier to just go with the old tried and true approaches rather than fight against theresistance to something new. Furthermore, LID projects are still not seen as “mainstream”.Such a perspective can contribute to viewing these sorts of projects as somehow still“experimental” and something that should be funded with grants so that the public’s money isnot put at risk. Indeed, risk aversion by those charged with making public infrastructuredecisions is a real factor in these sorts of projects.

Also challenging for LID projects is having the drain infrastructure, in part, above ground andvisible, and moreover, as in the case of the TRGD, literally in people’s front, side, and back yards. This brings a whole new set of challenges to a project. When a catch basin is installed allthat is seen is a 2-foot grate on the ground or in a curb line. What isn’t seen is all the connected

pipes under the surface and the large concrete structure that is underneath the grate. Installing arain garden that is above ground, puts it in the public’s face. Aesthetics become very importantwhen property owners are faced with looking at and living with a rain garden everyday. Indeed,after construction of the TRGD, the only complaints that are received from the neighborhood areabout the plants and how the rain gardens look, not about the functioning of the drain.Maintenance costs with LISMS can indeed be much higher than with a traditional drain project,not only because of public demand for aesthetics, but also because rain gardens are livingsystems that need continued care. From a life cycle perspective, these costs are offset by the far



17

lower initial construction cost, but still, if there is not adequate money for maintenance under whatever public works laws the drain infrastructure is being built and must operate, then LISMSmay not be a good choice. In any case, dealing with such issues is a new world for those chargedwith building public drain infrastructure and this newness is daunting for some.

Public involvement is going to be a major factor in the future acceptance of LISMS. Indeed, agreat deal of time was spent during the construction of the TRGD as well as during ongoingmaintenance to both educate and involve the public. Success of these projects has to bemeasured not just in whether the project has managed both the water quantity and water qualityissues, but also in the acceptance and understanding by the public regarding the project. Use of LISMS offer a great opportunity for public officials to address the goals of the Federal CleanWater Act Phase II by involving the public and by having the infrastructure not out-of-sight andout-of-mind, but in their face, so that there is a greater appreciation of the connectivity betweenwhat is done on the land and water resources. Ultimately, the future of LID in Michigan, aselsewhere, will no doubt hinge on such public involvement and eventual public acceptance.

Although there are great challenges using LISMS, these are minor compared to the great benefits that are achieved with this approach. There are approximately two million miles of pipeinfrastructure in the forty-eight continental states, with about 80% of it 100-years old, or older,and in need of repair. There will no doubt be an important role for LISMS approaches incost-effectively rebuilding our nation’s crumbling drain infrastructure to modern standards.

Footnotes

6. TSS was tested using the USEPA Gravimetric Method.7. TP was tested using the EPA Compliant Ascorbic Acid Method (Method 10210) using HACHTNT+ test-in-tube platform on a DREL 2800 spectrophotometer. The lowest detection limit for these tests is 0.05mg/L P. This test requires an Acid Digestion Step.8. TN was tested using the Persulfate Digestion Method (Method 10208) using the HACHTNT+ test-in-tube platform on a DREL 2800 spectrophotometer. The lowest detection limit for these tests is 1mg/L N. This test requires an Acid Digestion Step.

towar rain garden drains: a michigan urban retrofit

Documents