Docstoc

Dispersion Modeling

Document Sample
Dispersion Modeling Powered By Docstoc
					Dispersion Modeling
   Marti Blad, Ph.D., P.E.




                       A Brief Introduction


 smoke stacks image from Univ. of Waterloo Environmental Sciences
          Introduction
 Many different types of models
 Limitations & assumptions
 Math and science behind models
   Transport phenomena
   Computers do Math for you
 Gaussian dispersion models
 Screen3 model information
 Why use mathematical models

                                   2
              Types of Models
 Gaussian Plume
   Mathematical approximation of dispersion
 Numerical Grid Models
   Transport & diffusional flow fields
 Stoichastic
   Statistical or probability based
 Empirical
   Based on experimental or field data
 Physical
   Flow visualization in wind tunnels, scale models,etc.
                                                            3
 Limitations & Assumptions
 Useful tools: right model for your needs
 Allows quantification of air quality problem
   Space – different distances, scale
   Time – different time scales
      Steady state conditions?
 Understand limitations
   Mathematics-different types
   Chemistry-reactive or non-reactive
   Meteorology-Climatology

                                                 4
Momentum, Heat & Mass Transport
 Advection
   Movement by flow (wind)
 Convection
   Movement by heat
      Heat island
 Radiation
 Diffusion
   Movement from high to low concentration
      Molecule Dance
 Dispersion
   Tortuous path, spreading out because goes around
    obstacles                                          5
Diffusion & dispersion




                         6
Transport of Air Pollution
              Plumes tell story
                Ambient vs DALR
              Models predict air
               pollution
               concentrations
              Input knowledge of
               sources and
               meteorology
              Chemical reactions
               may need to be
               addressed
                                    7
Models allow multiple mechanisms




                             8
Buoyancy =Plume rise




                       9
         Gaussian Dispersion
         z


                                           Dh = plume rise

                                             h = stack height
    Dh
                                            H = effective stack
             H = h + Dh                         height
H
    h                                   x

                              C(x,y,z) Downwind at (x,y,z)?

                          y
     Gaussian Dispersion
    Concentration Solution


                                                 z  H 2  
                                             exp               
                                               2 z
                                                            2
                                                                 
                   Q              y 2 
                                                                  
C x , y ,z              exp   2                          
                2 u y z 
                                  2 y   
                                           
                                                     z  H 2  
                                             exp               
                                               2 z
                                             
                                                            2
                                                                 
 The Gaussian Plume Model
 The mathematical
 shape of the curve
 is similar to that of
 Gaussian curve
 hence the model is
 called by that
 name.




                            12
          Gaussian-Based
         Dispersion Models
 Plume dispersion in lateral & horizontal planes
  characterized by a Gaussian distribution
   Picture
 Pollutant concentrations predicted are
  estimations
 Uncertainty of input data values
   approximations used in the mathematics
   intrinsic variability of dispersion process


                                                    13
         Simple Gaussian
        Model Assumptions
 Continuous constant pollutant emissions
 Conservation of mass in atmosphere
   No reactions occurring between pollutants
   When pollutants hit ground: reflected, or absorbed
 Steady-state meteorological conditions
   Short term assumption
 Concentration profiles are represented by
  Gaussian distribution—bell curve shape


                                                         14
Gaussian Plume Dispersion
 One approach: assume each individual plume behaves
  in Gaussian manner
    Results in concentration profile with bell-shaped curve




                                                               15
                Is this clear?
 Time averaged concentration profiles about
  plume centerline
    Recall limitations
 Normal Distribution is used to describe random
  processes
    Recall bell shaped curves in 3-D
 Maximum concentration occurs at the center of
  the plume
    See up coming model pictures
 Dispersion is in 3 directions
                                                   16
Gaussian Plume




                 17
Graphic Gaussian Dispersion
 Gaussian behavior extends in 3 dimensions




                                              18
What is a Dispersion Model?
 Repetitious solution of dispersion equations
   Computer solves over and over again
   Compare and contrast different conditions
 Based on principles of transport
   Complex mathematical equations
   Previously discussed meteorological conditions
 Computer-aided simulation of atmosphere based
  on inputs
   Best models need good quality and site specific data

                                                       19
Computer Model Structure

      INPUT DATA: Operator experience

                  EMISSIONS
     METEROLOGY               RECEPTORS




          Model does calculations



       Model Output: Estimates of
       Concentrations at Receptors


                                          20
            Screen 3 model
 Understand spatial and temporal relationships
 One hour concentration estimates
   Caveat in program
 Meteorology
 Source type and specific information
   Point, flare, area and volume
 Receptor distance
   Discrete vs automated
 Receptor height

                                                  21
     Meteorological Inputs
 Actual pattern of dispersion depends on
  atmospheric conditions prevailing during
  the release
 Appropriate meteorological conditions
   Wind rose
      Speed and direction
   Stability class
   Mixing Height
   Appropriate time period
                                             22
                 Point Source
 Source emission data
     Pollutant emission data
        Rate or emission factors
     Stack or source specific data
        Temperature in stack
        Velocity out of stack
   Building dimensions
   Building location
   Release Height
   Terrain
     More complex scenarios
                                      23
Different stack scenarios




                            24
 Model inputs effect outputs
 Height of plume rise calculated
   Momentum and buoyancy
   Can significantly alter dispersion & location of
    downwind maximum ground-level concentration
 Effects of nearby buildings estimated
   Downwash wake effects
   Can significantly alter dispersion & location of
    downwind max. ground-level concentration



                                                       25
Conceptual effect of
     buildings




                       26
Spatial relationships




                        27
Gaussian Plume




                 28
29
Screen3 Area Source
 Emission rate
 Area
   Longest side, shortest side
   Release height
 Terrain
   Simple Flat
   Reflection and absorption
 Distances
   Discrete vs automated
 Receptor height
                                  30
Why Use Dispersion Models?
 Predict impact from proposed and/or existing
  development
   NSR- new source review
   PSD- prevention of significant deterioration
 Assess air quality monitoring data
   Monitor location
 Assess air quality standards or guidelines
   Compliance and regulatory
 Evaluate AP control strategies
   Look for change after implementation
                                                   31
       Why Use Dispersion Models?

 Evaluate receptor
    exposure
 Monitoring network
    design
       Review data
       Peak locations
       Spatial patterns
 Model Verification


                                                                                           32
image from collection of Pittsburgh Photographic Library, Carnegie Library of Pittsburgh
                     Review
 Transport Phenomena
    Meteorology and climatology
    Add convection, pressure changes
 Gaussian = even spreading directions
    Highest along axis
    Not as scary as sounds
 Input data quality critical to model quality
 Screen 3 limitation for reactive chemicals
    No reactions assumed to create or destroy
 Create picture for Screen3 word problems
                                                 33

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:23
posted:4/16/2012
language:English
pages:33