Fast Response Chemical, Biological, & Radiological Dispersion … · USEPA wind tunnel experiment....
Transcript of Fast Response Chemical, Biological, & Radiological Dispersion … · USEPA wind tunnel experiment....
Fast Response Chemical, Biological, & RadiologicalDispersion Modeling in Cities
QUIC Sensor Siting Tool
Hit probability map for first collector Hit probability map for 2nd collector
The QUIC Collector Siting GUI is an automatedtool for determining the best (and worst) places toput bio collectors around a building complex.
QUIC is a fast response urban dispersion modelthat runs on a laptop. CBR agent dispersion canbe computed on building to neighborhood scalesin tens of seconds to minutes.
This work is primarily supported by the DHSBiological Countermeasures Program.
releaselocation
QUIC is comprised of:
* QUIC-URB
produces 3D wind field around buildingsusing empirical/ diagnostic model
* QUIC-PLUME
accounts for building-induced turbulencethrough Lagrangian random-walkdispersion model
* QUIC-GUI
graphical user interface for set-up,running, and visualization (QUIC-GUI)
What is QUIC? The Quick Urban & Industrial Complex(QUIC) dispersion modeling system
Boston DNC
DC Mall tracer dispersion simulation
The QUIC Graphical User Interface
The QUIC-GUI allows for the user to easilyset-up building layouts, specify the winds,choose a CBR agent type, and pick arelease location.
2D and 3D visualization tools allow the userto rapidly display wind flow and plumedispersion patterns.
Why Account for Buildings?
How has QUIC been used?
How accurate is QUIC?
Vulnerability & Threat Assessments
A chemical spill vulnerability assessment fora rail car accident.
Table-top Training Exercises
Urban Field Experiment Design
Clean-up & Restoration Studies
CBR Sensor Siting for Special Events
Outdoor-to-Indoor Infiltration
Real-time Response for Specific Sites
RDD Explosive-Driven Events
Population Exposure Calculations
Simulations can be performed to determine howmuch tracer needs to be released in order to bedetected by the samplers, but not over saturatethem. Midtown Manhattan simulation.
Sampler Over-Saturated
BelowDetectionThreshold
Pressures on building surfaces can be computed byQUIC as in this Salt Lake City Simulation. Thepressures can then be provided as boundaryconditions to indoor models to simulate indoordispersion.
Pressure distribution & flow streamlines
Deposition can be calculated on the groundas well on building walls and roofs as in thisWashington DC Mall example. The mass ofagent on individual faces or buildings can beprovided in order to help estimate themagnitude of clean-up efforts.
Davidson et al. (1995)
Buildings enhance lateral spread. Plan view ofsmoke dispersing over flat terrain (right) andgoing through an array of buildings (left).USEPA wind tunnel experiment.
Buildings enhance vertical spread. Side view ofsmoke being lofted high into the air by tall buildings.USEPA experiment from Heist et al. (2004).
Lethal concentration threshold patterns are differentwhen buildings are included (left). There is no cross-wind channeling in the non-building aware simulation(right).
An explosive release of radioactive particles in ahigh-rise area showing lofting of lighter particles(red) and trapping of heavier particles (blue).
An automated ArcGIS mapping tool can importa QUIC-generated plume (left) and computeexposures to the population (right).
University of Utah
Measurements Model Results
xsource
QUIC has been tested against both field andwind-tunnel data (see below), but due to thecomplexities of urban dispersion more testingis needed and is ongoing.
QUIC will never give perfect answers, but theit will account for the effects of buildings in anapproximate way and provide more realismthan non-building aware dispersion models.
Plume dispersion comparisons with Salt LakeCity Urban 2000 field data (filled circles).
Wind field comparisons using L. Manhattanwind-tunnel data (Bowker et al., 2004).
Wind-tunnel comparison of the concentrationfield behind a “high-rise” building.
Comparison of vortex circulation in a street canyon wind-tunnel experiment.
ModelExperimental
Wind vectors not scaled to match