Environmental Engineering Reference
In-Depth Information
2
.
M
ethodology
2.1. Experimental campaign
An experimental campaign will be carried out with at least
two high time resolution
monitors located downwind the solid bulk handling areas and a meteorological tower.
For planning the experimental measurement campaigns (time of the year and
sampling locations), the meteorological situation and associated wind field in the
study area will be
previously evaluated. During the campaign, the emission-
generating activities will be annotated closely, in order to enable a further detailed
study of the operations involved in every time period.
2.2. Emission estimate method
A high resolution Gaussian dispersion model called the MELPUFF-microscale
(Martín et al., 2007; Aceña et al., 2002) was used. This is a Lagrangian Puff model
that can estimate the contribution of every emitted puff at selected locations where
monitors are continuously recording particle concentrations. A first run is done
assuming a prescribed and identical and constant emission rate (1 g/puff and
1 puff/min) during the selected period. With this run, the relative contribution of
each puff “j” at point “i” (Monitor location) and time “k”. (R
i,j,k
) is computed.
The measured concentration of a pollutant (Co
i,k
) at a given point “i” and time
“k” may be expressed as a linear combination of the contributions of each
theoretical emitted puff “j”, by the expression: C
i,j
= ∑a
j
R
i,j,k
Where
a
j
is amount of pollutant emitted in each puff “j” from the source area.
This is an equation system being
a
j
the unknowns. The equation system is usually
over determined (more equations than unknowns variables). The next scheme
shows the structure of this equation system. A multi-regression technique is applied
to solve the over determined system of linear equations.
Monitor Puff Time Concentration.
i j k
1
1
1
C
11
a
1
R
111
0
0
……
0
1
2
2
C
12
1
R
112
a
2
R
122
0
…….
0
1
3
3
C
13
1
R
113
a
2
R
123
a
3
R
133
…….
0
…….
…….
…….
…………
……. ……. ……. …….
…….
1
n
n
C
1n
1
R
11n
a
2
R
12n
a
3
R
13n
…….
a
n
R
1nn
1
…….
…….
…………
……. ……. ……. …….
…….
1
m
C
1m
0
0
……
…….
a
n
R
1nm
