Page 1
Two-Dimensional Flood Modeling using RiverFlow2D SMS
Workshop
Reinaldo Garcia, PhDReinaldo Garcia,Hydronia LLCHydronia LLCPembroke Pines, FL
Cameron Jenkins, PECameron Jenkins, PEWEST Consultants Inc.WEST ConsultanSan Diego, CA
September 8th 2015
2015 Floodplain Management Association Annual Conference
Page 2
Workshop Agenda08:30 – 08:45 Check-in, introduction, review agenda.
08:45 – 09:30 Overview of the RiverFlow2D model.
09:30 – 10:15 Introduction to SMS RiverFlow2D GUI.
10:15 – 10:30 Break.
10:30 – 12:00 Hands-on session: Setting up a RiverFlow2D SMS project.
12:00 –12:30 Comparing RiverFlow2D with other 2D models?
Page 3
Flood Simulation with Autodesk®
Infraworks® 360 Project Boulder
• Autodesk-Hydronia collaboration agreement to integrate RiverFlow2D model in Autodesk® InfraWorks® 360 software
• Available as a free trial preview until Nov. 2015• Autodesk Labs site (labs.autodesk.com).
Page 4
Presentation
Does it Matter Doing 2D Flood Modeling 150 Times Faster?
• Thursday 11:00 a.m. - 12:30 p.m.• Location: Celebrity H
Page 5
Overview of the RiverFlow2D model
Page 6
Outline
• Two-dimensional models• RiverFlow2D hydrodynamics• Hydraulic and hydrologic components• Add-on modules• Equations, boundary conditions and numerical
methods • Assumptions and limitations• Tests
Page 7
One-dimensional Modeling
Q, A, V
Page 8
1D Models
• WSEL is constant on each cross section• Average velocity • Velocity perpendicular to cross section• Modeler predetermines flow direction!
Page 10
2D ModelingS
yxtGFU
Page 12
2D Model Limitations
• Neglect vertical Accelerations
• Neglect vertical velocity• Don’t consider
secondary currents
Page 13
2D Models• Structured mesh-Grid
– Velocity vectors can have any direction, or
– Velocity has predetermined directions
• Unstructured Flexible mesh (Triangles-quadrilaterals)– Velocity vectors can have any
direction
Page 15
Square vs Triangle
400,000 Triangles2,200,000 Squares
5.5 times more cells
Page 18
RiverFlow2D Model
• Combined Hydrologic-Hydraulic model
• Unstructured flexible mesh• Robust dry-wet bed algorithm• Finite-Volume numerical engine• Ensures stability supercritical and
subcritical regimes• Extensively validated in a broad
range of real-world projects• Add-on Modules: ST, PL, MD• Hydraulic structures• Aquaveo SMS GUI • GPU version.
Page 19
RiverFlow2D Equations
SyxtGFU
Page 21
Hydraulic Components
• Culvert calculations using FHWAformulation
• Weirs (levees and sound walls)• Internal rating tables• Sources and sinks • Bridges • Gates• Wind stress
Page 22
RiverFlow2D BRIDGES Component CapabilitiesArbitrary plan alignment
Complex bridge geometry
Free surface flow
Pressure flow
Overtopping
Combined pressure flow and overtopping
Page 23
Implementation of Bridges in 2D
• Friction term
• Pressure variation due to bed slope
• Bed shear stress
• Local energy loses
= + += 0, ,= 0, , , ,= 0, , , ,
Page 24
Testing• RiverFlow2D vs Experimental data• and HEC-RAS
Page 26
207 experiments
• RiverFlow2D
• HEC-RAS
Page 27
Testing
• Free-surface flow
Q(m3/h) DATA RiverFlow2D RiverFlow2D HEC-RAS HEC-RASBridge Depth Depth Error Depth ErrorRect 1 9.20 5.70 5.20 8.77 6.06 6.32Rect 2 9.20 4.20 4.36 3.81 4.78 13.81Rect 2 14.80 6.00 5.91 1.50 6.52 8.67T 5.80 3.20 3.27 2.19 3.49 9.06T 8.20 4.00 3.98 0.50 4.29 7.25Arch 5.80 5.00 5.03 0.60 5.82 16.40
Page 28
Testing
• Bridges under pressure flow
Q(m3/h) DATA RiverFlow2D RiverFlow2D HEC-RAS HEC-RASBridge Depth Depth Error Depth ErrorRect1 12.8 7.5 6.54 12.80 8.39 11.87T 15.9 6.5 6.09 6.31 6.87 5.69Rect3 5.9 4.3 4.47 3.95 5.29 23.02Arch 8.26 5 5.65 13.00 8.38 67.60
Page 31
RiverFlow2D GATES Component
Page 32
RiverFlow2D GATES ComponentCapabilities
DATA• Crest elevation Zc• Hgate• Ha time series
• Arbitrary plan alignment• Free surface flow• Overtopping
Page 33
Hydrologic Component
• Distributed model• Rainfall/Evaporation
– Spatially distributed, varied in time
• Infiltration– Spatially distributed
parameters– Infiltration losses
Page 34
Distributed Hydrologic Modeling• Highly dependent on mesh structure
• Square element grids are the most numerically viscous and show flow directionality
• Triangular-cell flexible meshes outperform gridded models.
Page 35
Hydrologic Component• Rainfall/Evaporation
– Uniform in space, varied in time
– Spatially distributed, varied in time
– Allows for virtually any rainfall input: Radar, gauges, etc.
Page 36
Hydrologic Component• Infiltration
– Spatially distributed parameters
– Infiltration rates• Horton• Green-Ampt• SCS-CN
Page 37
Required Data• Rainfall/Evaporation
– Precipitation time series for each gauge– Polygons of influence of each gauge
• Infiltration– Polygon data set for infiltration parameters– Infiltration method calculation data for each polygon
• Horton: k, fc, fo• Green-Ampt: K, , • SCS: CN,
Page 38
On-going developments: SAC-SMA• SAC-SMA (Sacramento Soil
Moisture Accounting Model)• Conceptually based rainfall
runoff model with spatially lumped parameters
• Widely used by NOAA-NWS river forecast centers in the US
• Developed to generate river forecasts on watersheds with a response time > 12 hours
Burnash, R. J. C., (1995). The NWS River Forecast System - catchment modeling. In: Singh, V. P. (Ed.). Computer Models of Watershed Hydrology, 311-366. Burnash, R.J.C., R.L. Ferral, and R.A. McGuire, (1973),’ A Generalized Streamflow Simulation System - Conceptual Modeling for Digital Computers’, U.S. Department of Commerce, National Weather Service and State of California, Dept. of Water Resources, March.
Page 39
High Resolution 2D Hydraulics
Page 40
Detailed hydrodynamics around Structures
Green River,USA.
Page 41
Urban Flooding from Rainfall Events. Haiti.
Page 42
RiverFlow2D Modules
• MD: Mud and Debris Flows• ST: Sediment Transport• PL: Pollutant Transport • OilFlow2D
Page 43
MD: Mud and Debris Flow
• Non-Newtonian Fluids• Hyperconcentrated flows• 8 Rheological Formulations• Granular flow• Mud/debris floods• Tailings dams
Page 44
Rudd Creek, Utah, USA.
Page 45
Dam-Break Flooding, Mexico.
Page 46
ST: Sediment Transport
• Bed load• Suspended load• Multiple size fractions• Erosion / deposition
Page 47
High-resolution Erosion and Deposition
Page 48
PL: Pollutant Transport
• Advection-Dispersion-Reaction• Multiple – simultaneous pollutants• Inter-pollutant first order reactions
Page 49
Nitrogen ConcentrationsTrou-du-Nord River, Haiti
Page 50
OilFlow2D
• Overland oil spills• Any viscous fluid• Viscosity-density f(Temperature)• Temperature f(time)
Page 51
Overland Oil Spills. Pipeline Bursts, California, USA,
Page 52
RiverFlow2D Verification & Validation
Page 53
RiverFlow2DVerification and Validation
Page 55
Questions?
[email protected] www.hydronia.com