Source impact assessment methods

1
What is receptor model ?
The Receptor Sampling site
Receptor model is the method to identify the
sources and the contribution rate of SPM based on
the data on chemical composition, particle size,
concentration variation, particle form, etc.
which are obtained at a certain point and a
certain period
Estimate the contribution of each type of
sources but not of individual release source
2
Receptor Model
Microscopic methods
Chemical Methods
optical
Enrichment Factor
Qualitative analysis
Time series analysis
S.E.M.
Special series analysis
Chemical Mass Balance
Quantitative analysis
Multivariate data analysis methods
3
Type of Receptor Model Microscopic methods
Morphological observation

• Focus on particle size, color, form, surface
• characteristics and optical nature
• advantage most effective to identify the source
  from trees tissue and pollen
• disadvantage take time, a large number of
  particles must be observed, costly, not
  sensitive enough to identify the org. particles
  and non-crystalline particles.

4
Fundamental of Receptor Model
Based on measured mass concentration and the
use of appropriate mass balance
5
Fundamental of Receptor Model
p
C
Sj

j
C total mass measured at the receptor siteSj
contribution of the source j to the receptor site
Assumption no interaction between aerosols that
causes mass removal or formation Fe total
Fesoil Feauto Fecoal .
6
Particle emitted by soil 200 mg(Si) /g(20) and
32 mg(Fe)/g(3.2) power plant (Si) /g(1) and
150 mg(Fe)/g(15) PM10 concentration 32 ug/m3
EXAMPLE

• Sitotal Sisoil Sipower
• Fetotal Fesoil Fepower

If S and P total aerosol contributions(in
ug/m3) from soil dust and power plant to the PM10
concentration in the receptor, then
PM10 S P E
Where E contribution from any additional
sources
Sitotal 0.2S 0.01PFetotal 0.032S 0.15P
S 12 ug/m3(37.5) ,P18 ug/m3(56.2), E 2
ug/m3(6.3)
7
Fundamental of CMB Model

• -concentration of chemical species are used to
  apportion mass of airborne particles
• - composition patterns of emissions from various
  classes of sources are different enough that one
  can identify their contributions by measuring
  concentrations of many species in sample
  collected at a receptor site

8
Fundamental of CMB Model
p
Ci
aijSj

j
Ci concentration of compound i at the receptor
site( 1lt i ltn)aij concentration of compounds
i in source jSj contribution of the source j
to the receptor site ( 1lt j ltp)
In a matrix form C ASFor ngtp, this equation
can be solved
9
Major assumptions used by CMB Model

• -Compositions of source emissions are constant.
• -Species included are not reactive.
• -All sources contributing significantly to the
  receptor have been included in the
  calculation.
• -No relationship among the source uncertainties
• -Number of sources is less than or equal to the
  number
• of species
• -Measurement uncertainties are random,
  uncorrelated,
• and normally distributed.

10
CMB method limitation

• - information is required about the number of
• sources types and their chemical composition
• - the source of secondary formed PM
• can not be analyzed
• - particle sources having similar chemical
• composition can not be classified

11
Fundamental of Multivariate Model
The chemical concentration at one receptor site
The concentrations contributed by each emission
source type
Elemental composition for each source category
x

Target to be estimated
Receptor data (given)
Source profile (given)
p
Cik

aijSjk
j
12
Major assumptions used by Multivariate Model
-Compositions of emission sources is
constant. -Chemical species used do not interact
with each other and their concentration are
linearly additive. - Measurement errors are
random and uncorrelated - Minimum no. of samples
needed is
N gt 30 V3
2
- The effect of processes that affect all sources
equally is much smaller than the effect of
processes that influent individual sources -
Eigenvector rotations (if used) are physically
meaningful
13
Relation between the emission source and SPM in
the air.
Elemental concentration
contribution
SPM in the air
V Ni Zn
Si Al Fe
Na Cl
industry
O V o o Ni o Zn
Si Al Fe
Soil
Na Cl
Sea salt
14
Al 1.0 ug/m3Zn 0.1 ug/m3Na
0.6 ug/m3Ca 0.4 ug/m3
CMB calculation
Elemental composition of SPM at receptor site
15
Elemental mass fraction for each emission source
category (source profile), ()
Elements Soil Industry Sea
salt Al 10.0 1.0 0.0 Zn 0.5
5.0 0.0 Na 1.0 1.0 30.0 Ca 2.0
5.0 1.0 Others 86.5 88.0 69.0
16
Equations for the least square CMB model

• Al 0.10x 0.01y 1.0.(1)
• Zn 0.005x 0.05y 0.1.(2)
• Na 0.01x 0.01y 0.30z 0.6.(3)
• Ca 0.02x 0.05y 0.01z 0.4.(4)

Result Soil 10.94 Industry 1.04 Sea
salt 1.63
17
Equation for the marker elements

• Al 0.10x 0.01y 1.0.(1)
• Zn 0.005x 0.05y 0.1.(2)
• Na 0.01x 0.01y 0.30Z 0.6.(3)
• Where x contribution of soil
• y contribution of industry
• z contribution of sea salt
• Result x 9.90
• y 1.64
• z 1.01

18
Only marker elements
soil Al (0.1) industry Zn (0.05) sea salt Na
(0.30)
Marker and contents

• Soil 1.0/0.10 10.0 ug/m3
• industry 0.1/0.05 2.0 ug/m3
• sea salt 0.6/0.30 2.0 ug/m3

Calculation
19
SPM analytical result
Apportionment by CMB model
SPM concentration 20 ug/m3
Other 6.5 ug/m3
32
Sea salt 1.6 ug/m3
8
Na 0.6 ug/m3
Industry 1.0 ug/m3
5
Zn 0.1 ug/m3
55
Al 1.0 ug/m3
Soil 10.9 ug/m3
Interpretation by chemical mass balance(CMB) model
20
SPM analytical result
Apportionment by CMB model
SPM concentration 20 ug/m3
Other 6.5 ug/m3
32
Sea salt 1.6 ug/m3
8
Na 0.6 ug/m3
Industry 1.0 ug/m3
5
Zn 0.1 ug/m3
55
Al 1.0 ug/m3
Soil 10.9 ug/m3
Interpretation by chemical mass balance(CMB) model
21
Factor analysis-Multiple Regression ( FA-MR) model

• FA find a minimum number of factor that explain
  most of the variance of the system (large for the
  several factor associated with important sources)

MR mass concentrations of several marker
elements of sources coefficient average masses
contributed by sources
22
Concentration data
Factor analysis
Factor Loading
Factor score
Calculation of Absolute Factor Score
Estimate probably major sources contributed to
the receptor site
Multiple Regression
FA-MR model
Source Contribution
23
Data needs to confirm source apportionment

• - meteorological data/ wind rose/wind field
• - concentration of gaseous pollutants
• - land use/ emission sources/ air vivro model
• - activities of coal mine
• - variation and quality of coal

24
Comparison between CMB and FA-MR model

• Advantage

CMB -Model Everyone can calculate the
contribution rate according to the model
instruction. We can technically calculate the
contribution rate based on only one ambient data.
(However, it is undesirable since we can not
verify the result based on only one data.
25
Comparison between CMB and FA-MR model

• Advantage

FA-MR model In Case that we do not know the
important source categories. FA-MR can be
applied. According to the results, we can
estimate the distinguished emission source and
their source profile.
26
Comparison between CMB and FA-MR model

• Disadvantage

CMB -Model All of the source categories should be
included for the CMB calculation. Some
preliminary information is required at the stage
of the set-up the source profile. The source
profile for distinguished emission source should
be determined for the CMB application.
27
Comparison between CMB and FA-MR model

• Disadvantage

FA-MR model The high level of skills for the
statistics and the knowledge for the emission
source profile are required. A large number of
data (ambient data) is required, usually more
than 40 samples.
28
(No Transcript)