Flood Improvement and LID Modeling Using XP SWMM

Transcription

1 Flood Improvement and LID Modeling Using XP SWMM Andrew Juan 1, Nick Fang 2, and Philip Bedient 3 Rice University August, 2013 Introduction First developed in 1971, the EPA's Storm Water Management Model (SWMM) has been widely used in urban areas for planning, analysis, and design related to stormwater runoff, combined sewers, sanitary sewers, and other drainage systems. XP SWMM, published by XP Solutions Inc., uses the EPA SWMM model as the primary runoff hydrograph simulator for runoff quantity and quality. The major advantage of XP SWMM over other software packages is its ability to combine a 1D river hydraulic model with a 2D rainfall runoff model to generate floodplain maps. In addition, XP SWMM is approved by FEMA for the hydrologic and hydraulic applications detailed in this study. There are three separate modules present in XP SWMM: runoff, sanitary, and hydraulic. The runoff module includes subcatchment areas for hydrologic modeling. The sanitary module allows for pollutant loading and water quality analyses, and the hydraulic module is capable of simulating the hydraulics of channels and storm sewers. Most of the work for Rice University s flood improvement project was accomplished in XP SWMM s hydraulic module, while the LID modeling was conducted using the runoff module. Harris Gully Watershed Rice University is located within the Harris Gully watershed, which covers a drainage area of approximately 5 square miles, consisting mostly of fully developed residential and light commercial land 1 Graduate student, Civil and Environmental Engineering Department, Rice University, PH (713) , Andrew.juan@rice.edu 2 Project Manager, Civil and Environmental Engineering Department, Rice University, PH (713) , nickfang@rice.edu 3 Professor, Civil and Environmental Engineering Department, Rice University, PH (713) , bedient@rice.edu 1

2 use. The area is also home to Hermann Park and the Texas Medical Center, the largest medical center in the world. A number of severe storms have impacted the Harris Gully watershed, the most notable of which being Tropical Storm Allison in TS Allison caused severe flooding problems throughout the entire watershed, resulting in roughly $5 billion in damages. Inundation was especially severe at the downstream portions of Harris Gully, where Rice University and the Texass Medical Center are located. After TS Allison and as part of the Tropical Storm Allison Recovery Projectt (TSARP), a number of drainage improvement efforts were conducted to mitigate future flooding problems in the watershed (see Figure 1). Located at Kirby Dr., McGregorr Dr. (now called Cambridge Dr.), and Hermann Dr., these structural improvements were designed to divert stormwater away from the primary Harris Gully culvert and distribute it along Brays Bayou. These improvements and other changes within the Harris Gully watershed have been incorporated into this study. Rice TMC Figure 1: The Harris Gully Watershed with drainage improvements depicted in blue. 2

3 Rice University Flood Improvement Project The main purpose of the project was to represent Rice University s existing storm water network using a 1D/2D hydraulic model (XP SWMM) and to evaluate the effects of a proposed drainage alternative on Rice s 100yr floodplain, focusing specifically on the area adjacent to Alumni Drive. The pipe network of Rice University was represented as a system of links and nodes in XP SWMM. In the proposed alternative, new 48 pipes would be connected to existing 48 sewers under the Center Quadrangle. The pipes would then be run along Alumni Drive, underneath the intramural fields and stadium parking lots, and into Harris Gully. A comparison between the existing and proposed alternatives is depicted in Figure 2. 3

4 Figures 2(a) and (b): Existing (top) and proposed (bottom) storm water network on Rice Campus. Changes to the existing network are highlighted in blue in 2b. To create the base XP SWMM model, Tropical Storm Erin (August 16, 2007) was used. The primary reason for the selection of this storm was the availability of high water marks at several locations throughout the Rice campus. These records later proved to be instrumental in model calibration. The rainfall data for this event were obtained from a HCOEM rain gauge at Holcombe (OEM400) that recorded a total rainfall depth of 6.2 in. The rainfall data are shown in Figure 3 below. 4

5 Figure 3: Cumulative Rainfall from HCOEM Gauges. Land use was categorized within XP SWMM as either HiFriction or LoFriction. HiFriction represented highly impervious areas, mainly the buildings on Rice campus. The rest of the Rice the areas of lower imperviousness was categorized as LoFriction. Figure 4 shows Rice s land use characteristics as represented in XP SWMM. 5

6 Figure 4. Land use characteristics of Rice University as represented by XP SWMM. After running the base model with TS Erin, the simulated water surface elevations from XP SWMM were compared with observed high water marks at various locations. The simulated inundation depths generally corresponded well with the observed high water marks along Alumni Drive. Figure 5 shows the comparison between the simulated and observed water surface elevations at Alumni Drive for TS Erin. Figure 5: Comparison between XP SWMM modeled high water marks and those observed during TS Erin. 6

7 To compare the performances of existing and proposed conditions of the Rice campus, the 100yr24hr design storm was simulated. The proposed alternative was found to be successful in reducing the overall inundation levels along Alumni Dr. for the design storm. In general, approximately 0.5 1ft of inundation reduction was observed as illustrated in Figures 6 and 7. This flooding reduction would be beneficial for several of the existing buildings adjacent to Alumni Drive, especially those with basement laboratories and/or valuable equipment and research material, such ass the Keith Weiss Geological and Anderson Biological Laboratories as well as the Space Science and Technology Building. Figure 6: 100 yr flood levels from Existing (left) and Proposed (right) Conditions. 7

8 Figures 7( (a) and (b): Stage Comparisons at Selected Nodes for 100yr24hr Storm. The existing stormwater network performance is symbolized by the orange curve. Performance with the proposed improvements is depicted by the blue curve. 8

9 LID Modeling In addition to increasing drainage capacities through the installation of new pipes, another possible strategy for flood mitigation on Rice s campus is the implementation of Low Impact Development (LID) practices. Since the late 1990s, LID has emerged as an alternative to traditional urban development strategies. However, unlike most traditional methodologies, LID features are designedd to manage storm runoff as close as possible to its origin. This is accomplished through the use of distributed technologies such as green roofs, bio swales, and rain gardens, with the end goal of reducing the negative impacts of urbanization on natural resources by maintaining pre developmenit quite suitable for quantifying the potential benefits of hydrologic conditions. The capabilities and ease of use of XP SWMM make LID adoption. In this preliminary study, a hypothetical greenn roof scenario was simulated in the runoff module of XP SWMM. Analysis was focused on the four buildings that drained to node 335 on Alumni Drive. The green roofs, as modeled in XP SWMM, are shown in Figure 8 below. Figure 8: Green roof modeling using XP SWMM. Comparisons were made between the base scenario (traditional concrete roofs) and the LID scenario (green roofs) by using the Soil Conservation Service (SCS) Curve Number (CN) method to dictate the 9

10 hydrologic performance of each roof type. The SCS CN method is commonly used in many hydrologic models to simulate watershed runoff. The method appliess an empirical approach to determine the appropriate volume of stormwater runoff for a given area byy characterizing its hydrologic responsee with a Curve Number or CN, a parameter that represents the combined impacts of soil type, vegetative cover, and moisture conditions on runoff generation. Figure 9 presents some preliminary results from the scenarios using Green Roofs and traditional hydrologic modeling schemess in XP SWMM. For this particular study, the green roofs in the LID scenario were represented by SCS Curve Number 86, while the concrete roofs were represented by SCS Curve Number 77. Comparisonss were also made between two different runoff methods: the SWMM runoff f method and the kinematic wave method. After simulating several design storms (1in rainfall, 2yr 3hr, 10yr 12hr), initial results indicated that the green roofs were able to reduce peak flows as well as total runoff volume by approximately 20%. In addition, it was observed different routing methods resulted in discrepancies in the magnitude and timing of peak flows. Specifically, it was found that the SWMM runoff method resulted in delayed and decreasedd peak flows when compared to the kinematic wave method. These resultss are by no means conclusive, indicating that more studies need to be conducted to validate these findings. Figure 9: Preliminary Results from Green Roofs vs. Traditional Models in XP SWMM. 10

11 Conclusions and Work in Progress Overall, XP SWMM was found to be instrumental in simulating and depicting the drainage conditions for the Rice flood improvement project; and based on the preliminary results, it also has potential for application in the design and evaluation of different LID features. Currently, Rice is in the process of constructing several new buildings throughout campus, and it is likely that this activity will impact existing hydrologic conditions. Therefore, it is paramount that more studies are conducted to further improve our understanding of drainage performance at Rice. Future analyses may include performing a more detailed hydrology/hydraulics study in the areas beyond Alumni Drive as well as the evaluation of other LID features, including porous pavements, swales, and infiltration strips. 11