LOW IMPACT DEVELOPMENT MODELING APPROACHES

Transcription

1 LOW IMPACT DEVELOPMENT MODELING APPROACHES A Presentation to the Lake Tahoe Interagency Task Force May 4, 211 William C. Lucas, Partner, Chief Scientific Officer Terrasolve LLC

2 Schematic of 67% Impervious Site, Fully Disconnected Lucas (211) 4.92 acre commercial/office complex with 67% impervious cover. Perimeter parking is graded to drain into perimeter bioswales with no curb. Bioswales with 1:1 inflow side slopes into bottom width of 4-1 feet. Roof runoff from office dispersed over lawn to inlets between buildings. Commercial roof runoff conveyed to top of south bioswale. Central parking conveyed to head of southeast bioswale. Swales conveyance shown as yellow lines, originating halfway up source bioswale. Schematic Layout of Disconnected Site N

3 Input Parameters for in./hr. Infiltration Rate Scenario SUBAREAS No. Name Outlet Area Imperv % Width Slope % N-Imperv N-Perv S-Imperv S-Perv Suction HydCon S8 North Parking S S9 North Swale J S13 Office Roofs S S14 Office Walks S S11 Office Walks S S1 Office Lawn S CONDUITS No. Name Inlet Outlet Length Manning Inlet Outlet Shape Depth Width Sides Sides C1 North Swale J1 J TRAPEZOIDAL C4 Trench Drain J2 J RECT_OPEN 2.5 C5 West Swale J5 J TRAPEZOIDAL C2 East Swale J3 J TRAPEZOIDAL C3 South Swale J4 J TRAPEZOIDAL C6 Pipe J6 POI CIRCULAR 1.5 Impervious areas are routed onto adjacent pervious bioswale subareas. Hydcon = K sat, Suction = Ψ. K sat is in./hr. in this case. Swales Manning s n fixed at.1. Low end of likely values. Three scenarios used: in./hr. with all impervious as if swales are pipes,.25 in./hr. with bioswale flow, and in./hr. with bioswale flow. Lucas (211)

4 Subarea Results from in./hr. Infiltration Rate Scenario SUBAREAS Total Total Total Total Total Peak Runoff No. Name Runon Evap Infil Runoff Runoff Runoff Coeff in in in in 1^6 gal CFS S8 North Parking S9 North Swale S13 Office Roofs S14 Office Walks S11 Office Walks S1 Office Lawn S7 Landscape Islands S6 East Parking S2 East Swale S12 Comm. Walks S4 Comm. Roof S5 Comm. Walks S1 Central Parking S3 South Swale System SWMM model for 25 Design Year used in the Philadelphia LTCPU. SWMM computes runon as well as runoff. Runon onto north bioswale from north parking area is almost 82 inches. However that this bioswale infiltrates 19 inches, so total runoff is less than 15 inches. Similar relationships seen for other source/wetted area subareas. Lucas (211)

5 Conduit Results from in./hr. Infiltration Rate Scenario CONDUITS Maximum Maximum Max Max/ Max/ Time of Max No. Name Flow Velocity Depth Full Full Occurrence CFS ft/sec ft. Flow Depth days hr:min C1 North Swale :3 C4 Trench Drain :6 C5 West Swale :2 C2 East Swale :2 C3 South Swale :29 C6 Pipe :2 Conveyance by bioswales reduces runoff flow rates and velocities. Note low velocity in most of the swales except west, which has a much higher hydraulic load. This conveys runoff from the north & central parking subareas, as well as commercial roof. Note the maximum flow depths were all below 6 inches. Average flow depths were much lower (1-3 ). Most flows were well below grass height, so Manning s n would be higher than.2. Lucas (211)

6 Event From: 8/15/25 12:17:57 AM To: 8/15/25 5:13:13 AM (4.92 hours) This is a very intense summer thunderstorm. No runoff for first 2 minutes. The.25 in./hr. scenario had 23% reduction in peak flow. But runoff volume reduced by only 7%. in./hr. scenario had peak reduction of 43%, runoff reduced by 3%. This is the event with the maximum flow depths and velocities. Rainfall (in/hr) Outflow (cfs) Mon Aug 25 S1 FullDisconnection System NoDisconnection-25 System FullDisconnection System System FullDisconnection-25 1AM 2AM 3AM 4AM 5AM Date/Time Maximum Rainfall(in/hr):3.444 Rainfall(in):1.175 Scenario: None Full.25 Full Maximum Outflow (cfs): Mean Outflow (cfs): Total Outflow (ft³): 13,92 12,92 9,716 Lucas (211) Exceedance

7 Event From: 8/15/25 12:17:57 AM To: 8/15/25 5:13:13 AM (4.92 hours) Lucas (211) These are the flow depths and velocities in this event. East bioswale has greatest flow for cross section, so highest flow rate & depth. North and West bioswales have slightly less flow rates and depths, but the peak depth remains within the grass height. South bioswale much wider, so very low flow depths and velocities. This demonstrates how bioswale conveyance can be designed to reduce peak flows. Depth (ft) Velocity (ft/s) Mon Aug 25 C5 C2 C1 C3 1AM Piscataway Creek Flow 4.5 depth 1.5 depth East Bioswale.9 fps North Bioswale.7 fps West Bioswale.6 fps South Bioswale.3 fps 2AM 3AM 4AM 5AM Date/Time Maximum Flow rates and depths, event of 8/15/25

8 Event From: 1/14/25 4:54:19 AM To: 1/14/25 6:6:33 PM (13.2 hours) Less intense event, but with quite high peak intensity. Disconnection eliminates runoff for first hour. The.25 inch/hour scenario had 57% reduction in peak flow. Runoff volume was reduced by 28%. inch/hour scenario had peak reduction of 94%, and runoff reduced by 95%. This demonstrates how disconnection can very substantially reduce peaks and volumes. Rainfall (in/hr) Outflow (cfs) Fri Jan 25 S1 FullDisconnection S1 FullDisconnection-25 System System FullDisconnection System FullDisconnection-25 6AM Lucas (211) Maximum Rainfall(in/hr):.768 Total Rainfall(in):1.679 Scenario: None Full.25 Full Maximum Outflow (cfs): Mean Outflow (cfs): Total Outflow (ft³): 19,44 14,5 961 Exceedance 9AM 12PM 3PM 6PM Date/Time

9 Event From: 1/7/25 9:52:28 PM To: 1/9/25 2:19:1 AM (28.44 hours) This is largest event of design year, equivalent to perhaps a 5-year event. Disconnection eliminates runoff for first 2 hours. However, the.25 inch/hour scenario actually had higher peak flows, due to swale timing from source areas. Runoff volume was still reduced by 25%. inch/hour scenario had 11% reduction in peak flow. Runoff volume was reduced by 59%. Rainfall (in/hr) Outflow (cfs) Oct 25 S1 FullDisconnection-25 S1 FullDisconnection System FullDisconnection-25 System System FullDisconnection Sat Lucas (211) 3AM 6AM 9AM 12PM 3PM 6PM 9PM 9 Sun Date/Time Maximum Rainfall(in/hr):1.48 Total Rainfall(in):3.674 Scenario: None Full.25 Full Maximum Outflow (cfs): Mean Outflow (cfs): Total Outflow (ft³): 41,9 1,16 17,25 Exceedance

10 25 Design Year Results Results were also evaluated in terms of CSO exceedances, based on a design rate of 5 cfs/acre. Flows below this value would not trigger an overflow. Percentage reductions shown in parentheses after parameter value. Disconnection Scenario (in./hr.): Maximum Outflow (cfs): Mean Outflow (cfs): Duration of Exceedances (hrs): Number of Exceedances: Volume of Exceedances (ft³): Total Outflow (ft³): None () ,5 438,7 Full (.25) (22.2%) 55 (6.4%) (7.9%) 31 (76.3%) 13, (51.1%) 173,6 (6.4%) Full () 6.89 (43.2%) 2 (85.6%) 1.74 (91.8%) 11 (91.7%) 39,79 (8.7%) 63,3 (85.6% 85.6%) Lucas (211) This modeling exercise indicates that through effective disconnection alone, you can theoretically obtain remarkable hydrological performance even at 67.2% impervious. The bioswales can be landscaped as part of the required perimeter buffer that would be placed on this site anyway, without the need for any specific BMP.

11 Ahearn, D., and R. Tveten. 28. Legacy LID: Stormwater Treatment in Unimproved Embankments Along Highway Shoulders in Western Washington. In: International Low Impact Development Conference, November 16-19, 28, Seattle, Washington. Barrett, M.E. 24. Performance and Design of Vegetated BMPs in the Highway Environment. In: Critical Transitions In Water And Environmental Resources Management, Proceedings of The 24 World Water and Environmental Resources Congress, June 27-July 1, 24, Salt Lake City, Utah. Caltrans. 23. Final Report: Roadside Vegetated Treatment Sites (RVTS) Study. CTSW-RT-3-28, Caltrans Division of Environmental Analysis, Sacramento, California. Herrera. 29. Final Project Report: Compost-Amended Vegetated Filter Strip Performance Monitoring Study. Prepared for Washington State Department of Transportation, by Herrera Environmental Consultants, Inc., Seattle, Washington. Kaighn, R.J., and S.L. Yu Testing of Roadside Vegetation for Highway Runoff Pollutant Removal. Transportation Research Record 1523: Kearfott, P.J., M.P. Aff, M.E. Barrett, and J.F. Malina. 25. Stormwater Quality Documentation of Roadside Shoulders Borrow Ditches. IDS-Water White Paper 179, Center for Research in Water Resources, University of Texas, Austin, Texas. Lancaster, C.D. 25. A Low Impact Development Method for Mitigating Highway Stormwater Runoff - Using Natural Roadside Environments for Metals Retention and Infiltration. Masters Thesis, Washington State University, Department of Civil and Environmental Engineering, Pullman, Washington, 157 pp. Lantin, A., and M. Barrett. 25. Design and Pollutant Reduction of Vegetated Strips and Swales. In: World Water Congress 25, May 15, 25, Anchorage, Alaska. Reister, M., and D.R. Yonge. 25. Application of a Simplified Analysis Method for Natural Dispersion of Highway Stormwater Runoff. Prepared for Washington State Department of Transportation, by Washington State Transportation Center - Washington State University, Pullman, Washington. Yonge, D.R. 2. Contaminant Detention in Highway Grass Filter Strips. Report No. WA-RD 474.1, Washington State Department of Transportation, Olympia, Washington.

12 Bioretention Planter/Trench (Graphics courtesy of CDM and PWD) MEDIA INFLOW Inflows captured by a curb cut inlet and dispersed over surface of planter. Flows then enter media with high flows directly into trench via dome inlet drain. Drain raised by 2 to direct flows into media in first flush and small events. Trench is a 5 wide stone cell linking each planter to the next one up the street. Trench has separate underdrain to convey flows to outlet control structure. Outlet controls comprise orifice for CSO criteria, plus weir for high flows.

13 Bioretention DS Routing Separate nodes for the curb cut inlet, planter pond, planter media and trench. Surrounding soil node mimics suction head underlying infiltration responses, where initial rates into dry soils are faster. Planter Pond discharges both into the media, and directly into the trench once ponded >2. Flows into the media based on an orifice designed to mimic the flow into the media. Flows from the media based on Darcy s law applied to differential head. Infiltration from trench routed according to saturated conductivity applied to wetted area. Surrounding soil node modeled as narrow column filled by a small orifice from trench. This node has a constant outflow to empty between events, representing soil recovery. 7 Surface Flow 8 Controlled Inlet 3 Street Lower 4B Planter Pond 4 4A Inlet 4 4C Media 4 Lucas (29) 5 5' Trench 2A Inlet 3 2C Media 3 1 Street Upper 2B Planter Pond 3 6 Surrounding Soil HydroCAD Routing Diagram, Two Planters and one Trench C

14 Bioretention CS Routing In addition to DS modeling in HydroCAD, SWMM5 model used to project annual response. Using same input parameters as entered as the DS model, SWMM5 then run with same design events to ensure that same response occurs. Once model equivalency confirmed, SWMM5 then run with Design Year rainfall. Design Year based on 25 distribution, with events modified to mimic annual distribution of events observed over 5 years. SampleD StreetLower StreetUpper Gage16 OrificeLower InletLower CurbUpper OrificeUpper InletUpper PondLower PondUpper MediaInLower CurbLower MediaLower MediaInUpper MediaUpper TrenchInUpper TrenchInLower MediaOutUpper MediaOutLower StreetUpNone InletDown OrificeFlow Trench OverflowWeir SuctionFlow SoilSuction 1 SaturatedFlow StreetLowerNone SuctionRecovery InletNone Surface Infiltrated 2 Infiltration SWMM Routing Diagram Lucas (29)

15 Bioretention Routing Lucas (29) Preprocessing in Excel used to develop stage area/perimeter relationships and other rating curve relationships. It is also used to develop equivalent orifice sizing for the media inflow and the surrounding soil Hortonian response. Note that design soil infiltration rate is.1 per hour. This is used to project stage/outflow rates for saturated hydraulic conductivity. This utilizes the full potential of a design tool such as HydroCAD or SWMM5 which allows for backwater routines and rating curves. Suction Design uses SWMM factors for recovery. Media Void% 3% Stone Void% 4% MEDIA DATA 1.16 Ksat (cm/hr) Length 5.81 Width.91 Elevation (m) Area (sq.m) Head (rel.) Volume Pore Vol. Q (lps) SWMM Area Interval INTO MEDIA No. Orifices 1 Diameter (cm) Or.Area (sq.cm.) Head (cm.) 5.18 Design Q (lps) FROM MEDIA Length (cm) Darcy (lps/m) INFLOW UNDERDRAIN DATA Or. (sq.cm.).7432 Dia (cm.) 3.48 Area/m.958 No. per m 39 elev Length No. Orifice Or. Area Or. Rate (lps) Q (lps) TRENCH DATA.254 Ksat (cm/hr) Underdrain Length Width 1.52 Surround/Trench Length 324 Width 1.52 Elevation Bottom Area Side Area Sides (lps) Q (lps) Ksat (lps) SWMM Area Pore Vol SWMM Area 17.3 Pore Vol SUCTION DESIGN Suction (cm) 17.8 Diameter (cm) 3.81 Peak Q (lps) 6.8 Suction Ratio 14. Or.Area (sq.cm) Design Q (lps) Recovery (hr) Depth (m) 4.13 Rise Vol.(cu.m) Recovery (lps) 382 Area (sq.m) 7.9 Node Vol.(cu.m) OUTFLOW UNDERDRAIN DATA 381 (cm/hr) elev Area (sq.m) Flow (lps) Perfs (sq.cm) Rate (lps) Q (cfs) Diameter (cm) 4.5 Orifice (sq.cm) Design Q (lps) 4.82 Excel entries for both SWMM and HydroCAD Models

16 8 /1 5 /5 : 1 /7 /5 1 6 : 1 /8 /5 : 1 /8 /5 8 : 1 /8 /5 1 6 : Hydrographs and Storage Depths, a) April 2, (b) August 15 and (c) October 8, 25 Events. Note how the difference in elevation between the trench and the suction node results in corresponding suction infiltration response. Bioretention CS Routing R ainfall (m m /hr) Elevation (m ) SoilSuction Trench SuctionFlow SaturatedFlow /1/5 22: 4/2/5 6: 4/2/5 14: 4/2/5 8/15/5 22: : 8/15/5 8: 1/7/5 16: 1/8/5 : 1/8/5 8: 1/8/5 16: Overflow Weir El..555m Lucas (29) Infiltration Flow Rates and Node Depths, (a) April 2, (b) August 15 and (c) October 8, 25. Top Row displays Rainfall. Middle Row displays Trench and Suction Node Elevations. Bottom Row displays Trench Outflow into Surrounding Soil via Suction and Saturated Flow. Note Differing Scales for Rainfall and Flow.

17 Bioretention CS-DS Routing Lucas (29) Flow (L/sec) a) Media Inflow SWMM Media Upper In HydroCAD Upper In Constant Media Upper In Time (hr.) Depth (m) c) Media Depth SWMM Media Upper HydroCAD Media Upper Constant Media Upper Time (hr.) Depth (m) e) Trench and Soil Depth Constant Trench HydroCAD Trench SWMM Trench HydroCAD Soil SWMM Soil Time (hr.) Flow (L/sec) b) Media Outflow SWMM Media Upper Out HydroCAD Upper Out Constant Media Upper Out Time (hr.) Flow (L/sec) d) Trench Inflow SWMM Trench In HydroCAD Trench In Constant Trench In Time (hr.) Flow (L/sec) 2. f) Trench and Soil Exfiltration Constant Trench Out SWMM Trench Out HydroCAD Trench Out SWMM Soil Out HydroCAD Soil out Time (hr.) Hydrographs and Storage Depths, 1.5 Inch (~.6 Year) Sample D Event. E Note characteristic double hump after peak of inflow hydrograph. Outflows in constant head media lower due to disproportionate effect fect of head.

18 5 Bioretention CS Routing Link Flow Link MediaInUpper Link MediaInLower Link MediaOutUpper Link MediaOutLower Lucas (29) 4 Flow (CFS) : 1/4/5.2 12: 1/4/5 : 1/5/5 12: 1/5/5 : 1/6/5 12: : 12: : Link Flow 12: : 12: : 12: 1/6/5 1/7/5 1/7/5 1/8/5 1/8/5 1/9/5 1/9/5 1/1/5 1/1/5 Link OverFlowWeir Link OutletOrifice Link TrenchOut Link InfiltrationSoil : 1/11/5 12: 1/11/5 : 1/12/5 12: 1/12/5 : 1/13/5 Flow (CFS) lps/ha. (5 cfs/ac.) CSO threshold 5 Depth (ft) : 1/4/ : 1/4/5 12: 1/4/5 12: 1/4/5 : 1/5/5 : 1/5/5 12: 1/5/5 12: 1/5/5 : 1/6/5 : 1/6/5 12: 1/6/5 12: 1/6/5 : 12: : Node Depth 12: : 12: 1/7/5 1/7/5 1/8/5 1/8/5 1/9/5 1/9/5 Node MediaUpper Node MediaLower Node Trench : 1/7/5 12: 1/7/5 : 1/8/5 12: 1/8/5 : 1/9/5 12: 1/9/5 : 1/1/5 : 1/1/5 12: 1/1/5 12: 1/1/5 : 1/11/5 : 1/11/5 12: 1/11/5 12: 1/11/5 : 1/12/5 : 1/12/5 12: 1/12/5 12: 1/12/5 : 1/13/5 : 1/13/5 Hydrographs and Depths, 1.5-inch, 2. Inch, and 2.69-inch Sample D Events. Double hump becomes even more pronounced due to preferential bypass flow into stone.

19 5 Bioretention CS Routing Link Flow Link MediaInUpper Link MediaOutUpper Link MediaInLower Link MediaOutLower Lucas (29) 4 Flow (CFS) Flow (CFS) 18: 1/11/ : 1/12/5 6: 1/12/5 12: 1/12/5 3.5 lps/ha. (5 cfs/ac.) CSO threshold 18: 1/12/5 : 6: 12: 18: : 6: 12: 18: : 6: 12: Link Flow 1/13/5 1/13/5 1/13/5 1/13/5 1/14/5 1/14/5 1/14/5 1/14/5 1/15/5 1/15/5 1/15/5 Link OverFlowWeir Link OutletOrifice Link TrenchOut Link InfiltrationSoil 18: 1/15/5 : 1/16/5 6: 1/16/5 12: 1/16/5 18: 1/16/5 : 1/17/5 Depth (ft) 18: 1/11/ : 1/11/5 : 1/12/5 : 1/12/5 6: 1/12/5 6: 1/12/5 12: 1/12/5 12: 1/12/5 18: 1/12/5 18: 1/12/5 : 1/13/5 : 1/13/5 6: 1/13/5 6: 1/13/5 12: 18: : 6: 12: 18: : 1/13/5 1/13/5 1/14/5 Node 1/14/5 Depth 1/14/5 1/14/5 1/15/5 Node MediaUpper Node MediaLower Node Trench 12: 1/13/5 18: 1/13/5 : 1/14/5 6: 1/14/5 12: 1/14/5 18: 1/14/5 : 1/15/5 6: 1/15/5 6: 1/15/5 12: 1/15/5 12: 1/15/5 18: 1/15/5 18: 1/15/5 : 1/16/5 : 1/16/5 6: 1/16/5 6: 1/16/5 12: 1/16/5 12: 1/16/5 18: 1/16/5 18: 1/16/5 : 1/17/5 : 1/17/5 Hydrographs and Storage Depths, January 12-14, 25 Events. Still see characteristic double hump after peak of inflow hydrograph. (Lucas, 29)

20 Bioretention CS Routing Lucas (29) : Flow (CFS) Link Flow Link OverFlowWeir Link OutletOrifice Link TrenchOut Link InfiltrationSoil 6: 12: 18: : 6: 12: 3.5 lps/ha. (5 cfs/ac.) CSO threshold 18: : Flow (CFS) Link Flow Link MediaInUpper Link MediaOutUpper Link MediaInLower L : 6: 12: 18: : 6: 12: Depth (ft) : 6: Node Depth Node MediaUpper Node MediaLower Node Trench 12: 18: : 6: 12: 18: : Total Inflow (CFS) : 12: Node Total Inflow Node InletNone Node InletDown 18: : 6: 12: Hydrographs and Storage Depths, 9.83 cm (3.87 in.) event of October 8, 25 Note how media outflow eliminated during peak inflow- actually negative! Even in this large event, substantial volume and peak flow reductions.

21 Bioretention CS Routing Lucas (29) 1 Existing Conditions Exceedance Frequency lps/ha. (5 cfs/ac.) 3.5 lps/ha. (5 cfs/ac.) 3.5 lps/ha. (5 cfs/ac.) Event Peak Total Inflow (CFS) Exceedance Frequency mm/hr. Infiltration Rate 3.5 lps/ha. (5 cfs/ac.) 1 2 Event Peak Total Inflow (CFS) 3 Exceedance Frequency mm/hr. Infiltration Rate 3.5 lps/ha. (5 cfs/ac.) 1 2 Event Peak Total Inflow (CFS) 3 Exceedance Frequency mm/hr. Infiltration Rate 3.5 lps/ha. (5 cfs/ac.) 1 2 Event Peak Total Inflow (CFS) 3 Flow Exceedance Frequencies, Comparison of Infiltration Rates at 5 in-h -1,.1 in-h -1 and.2 in-h -1 to Uncontrolled Inlet Flows. Note how peak exceedances decrease to only several events in the controlled systems.

22 Bioretention CS Routing 1 1 Percent Less Than 1%.1% % 1% 1% NO LID Interval Flow LID Interval Flow NO LID Percent Less LID Percent Less Lucas (29) Flow (lps) Lps-ha -1 (5 cfs-ac -1 ) Minute Flow Intervals Annual Flow Exceedance Frequencies, Uncontrolled Runoff Compared to Planter-Trench System at 2.54 mm-h -1 (.1 in.-h -1 ) Infiltration Rate. Note far more flows discharged below threshold with Planter Trench, with a long period of very low flows. This drains the system so even large events are substantially attenuated.

23 Bioretention CS Results Scenario Uncontrolled Runoff Controlled Inlet at 1.27 mm-h -1 Controlled Inlet at 2.54 mm-h -1 Controlled Inlet at 5.8 mm-h -1 Runoff Volume (cu.m.) Infiltrated Volume (cu.m.) Lucas (29) Infiltration Depth (m.) Percentage Infiltrated % 36.9% 49.2% 63.1% Number of Exceedances Exceedance Volume (cu.m.) System intercepts and infiltrates 36.9% to 63.1% of the entire annual runoff, even given very conservative infiltration rates. This is accomplished with a planter area of only 1.2% of the impervious source area, representing a loading ratio of over 8:1. This pretreats runoff into the trench. It will need to be maintained! The underlying trench area is 5.6% of the impervious source area, resulting in an effective infiltration loading ratio of 18:1.

24 Bioretention Modeling Conclusions DS Conclusions: System set up so that bypass begins just beyond the 2. event. System overflows only in one-year events and larger. Still treats a lot of runoff. The system is thus seems capable of treating the vast majority of annual runoff. Using a more realistic design storm and discrete routing, it seems possible to mimic the behavior of how a planter/trench system will operate over a variety of such events. CS Conclusions: Still need to run continuous simulation modeling to really get at the extent of CSO reductions. This means using a SWMM model set up and calibrated for Philadelphia. Using a realistic design year it is possible to project the behavior of a planter/trench system over annual rainfall distribution.

25 Acknowledgement This work was performed under contract to CDM and funded by the Philadelphia Water Department. William Lucas, Partner, TerraSolve LLC. 47 Highway 79, Morganville, NJ p: