Proposed HR Milner Expansion Project Environmental Impact Assessment Report
Appendix 7C: CALPUFF and CALMET Methods and Assumptions
Maxim Power Corp. Page 7C-1 January 2009
Appendix 7C CALPUFF and CALMET Methods andAssumptions
Proposed HR Milner Expansion Project Environmental Impact Assessment Report
Appendix 7C: CALPUFF and CALMET Methods and Assumptions
Maxim Power Corp. Page 7C-2 January 2009
7C.1 CALMET Modelling
7C.1.1 Introduction
Meteorology determines the transport and dispersion of industrial emissions, and hence plays a
significant role in determining air quality downwind of emission sources. For the air quality assessment,
meteorological data for the year 2007 were used to define transport and dispersion parameters.
Meteorological characteristics vary with time (e.g., season and time of day) and location (e.g., height,
terrain and land use). The CALMET meteorological pre-processing program was used to provide
temporally and spatially varying meteorological parameters for the CALPUFF model. This appendix
provides an overview of the meteorology for the region as well as the technical details and options that
were used for the application of the CALMET meteorological preprocessor to the Project study area.
7C.1.2 Model Description
The following description of the CALMET model’s major model algorithms and options are all excerpts
from the CALMET model’s user manual (Scire et al. 2000a).
The CALMET meteorological model consists of a diagnostic wind field module and micrometeorological
modules for overwater and overland boundary layers. The diagnostic wind field module uses a two-step
approach to the computation of the wind fields (Douglas and Kessler 1988), as illustrated in Figure 7C-1.
SOURCE: Scire et al. 2000a
Figure 7C-1 Flow Diagram of Diagnostic Wind Module in CALMET
Proposed HR Milner Expansion Project Environmental Impact Assessment Report
Appendix 7C: CALPUFF and CALMET Methods and Assumptions
Maxim Power Corp. Page 7C-3 January 2009
In the first step, an initial guess wind field is adjusted for kinematic effects of terrain, slope flows, and
terrain blocking effects to produce a Step 1 wind field. The initial guess field is either a uniform field
based on available observational data or the output from the NCAR/PSU Mesoscale Modelling System
(MM4/MM5). The second step consists of an objective analysis procedure to introduce observational
data into the Step 1 wind field to produce a final wind field. Wind fields generated by the prognostic wind
field module can be input to CALMET as either the initial guess field or the Step 1 wind field.
7C.1.2.1 Diagnostic Wind Field Module – Initial Guess Field
Options exist within CALMET to create an initial guess field either by interpolating observation data or by
using output from a prognostic meteorological model, such as the NCAR/PSU Mesoscale Modelling
System (MM4/MM5). The prognostic model data is usually run over a very large domain with much
coarser resolution than that applied with CALMET. CALMET will interpolate the prognostic data to
develop a 3-D fine scale first guess field of wind speeds and directions.
Step 1 Wind Field
The step one wind field is adjusted for kinematic effects of terrain, slope flows, and blocking effects as
follows:
Kinematic Effects of Terrain: The approach of Liu and Yocke (1980) is used to evaluate kinematicterrain effects. The domain scale winds are used to compute a terrain forced vertical velocity, subjectto an exponential, stability dependent decay function. The kinematic effects of terrain on thehorizontal wind components are evaluated by applying a divergence minimisation scheme to theinitial guess wind field. The divergence minimisation scheme is applied iteratively until the threedimensional divergence is less than a threshold value.
Slope Flows: An empirical scheme based on Allwine and Whiteman (1985) is used to estimate themagnitude of slope flows in complex terrain. The slope flow is parameterised in terms of the terrainslope, terrain height, domain scale lapse rate, and time of day. The slope flow wind components areadded to the wind field adjusted for kinematic effects.
Blocking Effects: The thermodynamic blocking effects of terrain on the wind flow are parameterisedin terms of the local Froude number (Allwine and Whiteman 1985). If the Froude number at aparticular grid point is less than a critical value and the wind has an uphill component, the winddirection is adjusted to be tangent to the terrain.
Step 2 Final Wind Field
The wind field resulting from the adjustments of the initial guess wind described above is the Step 1 wind
field. The second step of the procedure involves the introduction of observational data into the Step 1
wind field through an objective analysis procedure. An inverse distance squared interpolation scheme is
used which weighs observational data heavily in the vicinity of the observational station, while the Step 1
wind field dominates the interpolated wind field in regions with no observational data. The resulting wind
field is subject to smoothing, an optional adjustment of vertical velocities based on the O'Brien (1970)
method, and divergence minimisation to produce a final Step 2 wind fields.
7C.1.3 Study Period
The CALMET meteorological model was run for one full year from January 1, 2007 to January 1, 2008.
7C.1.4 Meteorological Domain
The CALMET meteorological domain adopted for this project is summarized below in Table 7C-1. For a
graphical representation of the 2,500 km2area, refer to Figures 7C-2 and 7C-3.
Proposed HR Milner Expansion Project Environmental Impact Assessment Report
Appendix 7C: CALPUFF and CALMET Methods and Assumptions
Maxim Power Corp. Page 7C-4 January 2009
Table 7C-1 Map Projections and Horizontal Grid Parameters
Parameter Value
Map Projection UTM
UTM Zone 11
Datum WGS-84
Number of Grid Cells (nx, ny) 100, 100
SW Corner (Easting, Northing) 337.0km, 5961.0 km
Grid Spacing 0.5 km
The meteorological domain was selected to cover the region surrounding the proposed site location,
centered on the Project. Included in the southern part of the domain is the Town of Grande Cache,
Alberta.
A horizontal grid spacing of 0.5 km was selected for the CALMET simulation; the CALMET domain
therefore corresponds to 100 rows by 100 columns. With this grid spacing, it was possible to maximize
run time and file size efficiencies while still capturing large-scale terrain feature influences on wind flow
patterns.
To properly simulate pollution transport and dispersion, it is also important to simulate the representative
vertical profiles of wind direction, wind speed, temperature, and turbulence intensity within the
atmospheric boundary layer (i.e., the layer within about 2,000 metres above the Earth’s surface). To
capture this vertical structure, eight vertical layers were selected. CALMET defines a vertical layer as the
midpoint between two faces (i.e., nine faces corresponds to eight layers, with the lowest layer always
being ground level or 10 m). The vertical faces used in this study are 0, 20, 40, 80, 160, 320, 600, 1,400
and 2600 m.
Grande Cache
BRITISHCOLUMBIA
ALBERTA
Smoky River
Kakwa River
Cutbank River
Muske
g Rive
r Berland River
40
40
118°30'0"W
118°30'0"W
119°0'0"W
119°0'0"W
119°30'0"W
119°30'0"W120°0'0"W54
°30'0
"N
54°3
0'0"N
54°0
'0"N
54°0
'0"N
53°3
0'0"N
53°3
0'0"N
HR Milner Expansion Project Environmental Assessment
CALMET DomainFIGURE NO.
7C-2
PREPARED BY
Data Provided By: AltaLIS (2006), Government of Canada (2007).
Last
Modif
ied: J
uly 7,
2008
By:
mdes
ilets
PREPARED FORAB SKBC
USA
Areaof
Interest
JW-1033372-035
Proposed Plant SiteHighway 40Major RoadRailwayTown BoundaryCALMET DomainEnterprise/ CO-OP ZoneProtected AreaWilmore Wilderness ParkWeyerhaeuser FMA
0 5 10 15
Kilometres - 1:400,000
BC
AB
SK
NWT
Proposed HR Milner Expansion Project Environmental Impact Assessment Report
Appendix 7C: CALPUFF and CALMET Methods and Assumptions
Maxim Power Corp. Page 7C-6 January 2009
7C.1.4.1 Regional Topography
A dispersion model requires terrain information since terrain can influence the airflow and the dispersion.
The terrain information in Figure 7C-3 is for a 50 by 50 km area (i.e., study area) centered on the project
site and is based on the digital elevation model (DEM) obtained from the U.S. Geological Survey SRTM
(Shuttle Radar Topography Mission) (http://srtm.usgs.gov/). This data has a horizontal resolution of 90 m;
which is sufficient for air quality assessments. The following are noted relative to the terrain:
The center of CALMET domain is located at 5986000 m N and 362000 m E (NAD83 UTM) and thebase elevation of the well is 917 m above sea level (m ASL).
The lowest terrain (800-950 m ASL) occurs along the valley and one the northeast boundaries of thestudy area.
The highest terrain (>2450 m ASL) occurs on the peaks located southwest of the study area.
The valley is an important feature of the domain. The lowest elevations in the domain occur along the
alley and in the northeast corner of the domain. While valleys and elevated terrain features can affect
surface wind flow patterns, the selected grid spacing (0.5 km) is sufficient to resolve location terrains
influences.
7C.1.4.2 Land-use Data
Land-use in the CALMET domain is consisting primarily of forests. The domain is characterized by
coniferous forest (63.24 percent), mixed forest (31.94 percent), deciduous forest (2.79 percent), and
barren land (2.03 percent). Values for surface roughness (z0), leaf area index (LAI), albedo, Bowen ratio,
anthropogenic heat flux, and soil heat flux are given in Table 7C-2. The year was divided into four
seasons as follows: winter (December, January, and February), spring (March, April and May), summer
(June, July, and August) and fall (September, October and November). Figure 7C-4 shows the land use
on a 0.5 km resolution basis for the CALMET domain.
PREPARED BY
PREPARED FOR
FIGURE NO.
340000 345000 350000 355000 360000 365000 370000 375000 380000 385000
340000 345000 350000 355000 360000 365000 370000 375000 380000 385000
59
65
000
59
70
000
59
75
000
59
800
00
59
850
00
599
00
00
599
50
00
600
00
00
600
50
00
601
000
0
59
65
000
59
70
000
59
75
000
59
800
00
59
850
00
599
00
00
599
50
00
600
00
00
600
50
00
601
000
0
Distance in Metres
NAD 83 UTM Zone 11
0 5000 10000 15000
HR MILNER EXPANSION PROJECT ENVIRONMENTAL ASSESSMENT
P:/
JW
_P
roje
cts
/Post C
MIC
/10
33
37
2/A
ir Q
ua
lity/R
EP
OR
TS
/Te
chn
ical A
ppe
nd
ice
s/A
pp
en
dix
C/A
ppe
nd
ix C
Fig
ure
s/F
igu
re C
-3.s
rf
Terrain Elevations for the Modelled CALMET Domain
7C-3
Project Site
Map Features
950
110
0
125
0
140
0
155
0
170
0
185
0
200
0
215
0
230
0
245
0
Terrain Elevation (m asl)
Proposed HR Milner Expansion Project Environmental Impact Assessment Report
Appendix 7C: CALPUFF and CALMET Methods and Assumptions
Maxim Power Corp. Page 7C-8 January 2009
Table 7C-2 CALMET Domain Land-use Characterization and Associated Geophysical Parameters
Land Use ClassSurface Roughness (m) Albedo Bowen Ratio
Winter Spring Summer Fall Winter Spring Summer Fall Winter Spring Summer Fall
Mixed Forest 1.0 1.0 1.0 1.0 0.1 0.1 0.1 0.1 1.0 1.0 1.0 1.0
Deciduous Forest 1.0 1.0 1.0 1.0 0.1 0.1 0.1 0.1 1.0 1.0 1.0 1.0
Coniferous Forest 1.0 1.0 1.0 1.0 0.1 0.1 0.1 0.1 1.0 1.0 1.0 1.0
Barren Land 0.05 0.05 0.05 0.05 0.3 0.3 0.3 0.3 1.0 1.0 1.0 1.0
Land Use ClassSoil Heat Flux (fraction) Anthropogenic Heat Flux (W/m
2) Leaf Area Index
Winter Spring Summer Fall Winter Spring Summer Fall Winter Spring Summer Fall
Mixed Forest 0.15 0.15 0.15 0.15 0.0 0.0 0.0 0.0 7.0 7.0 7.0 7.0
Deciduous Forest 0.15 0.15 0.15 0.15 0.0 0.0 0.0 0.0 7.0 7.0 7.0 7.0
Coniferous Forest 0.15 0.15 0.15 0.15 0.0 0.0 0.0 0.0 7.0 7.0 7.0 7.0
Barren Land 0.15 0.15 0.15 0.15 0.0 0.0 0.0 0.0 0.0 0.0 0.05 0.05
PREPARED BY
PREPARED FOR
FIGURE NO.
340000 345000 350000 355000 360000 365000 370000 375000 380000 385000
340000 345000 350000 355000 360000 365000 370000 375000 380000 385000
596
50
00
597
00
00
597
500
05
98
000
05
985
00
05
990
00
05
995
00
060
00
00
060
05
000
60
10
000
596
50
00
597
00
00
597
500
05
98
000
05
985
00
05
990
00
05
995
00
060
00
00
060
05
000
60
10
000
Distance in Metres
NAD 83 UTM Zone 11
0 5000 10000 15000
P:/JW
_P
roje
cts
/Post C
MIC
/10
33
37
2/A
ir Q
ua
lity/R
EP
OR
TS
/Techn
ical A
ppe
ndic
es/A
pp
en
dix
C/A
pp
en
dix
C F
igu
res/F
igure
C-4
.srf
HR MILNER EXPANSION PROJECT ENVIRONMENTAL ASSESSMENT
CALMET Land-use Categories Specified for the Modelled Domain
7C-4
Project Site
Map Features
Barren Land
Coniferous Forest
Deciduous Forest
Mixed Forest
Proposed HR Milner Expansion Project Environmental Impact Assessment Report
Appendix 7C: CALPUFF and CALMET Methods and Assumptions
Maxim Power Corp. Page 7C-10 January 2009
7C.1.5 Meteorological Inputs
The CALMET model requires the input of surface and upper air meteorological fields. Meteorological data
from site surrounding the Project were reviewed, and the results were used for the CALMET
meteorological model. The meteorological model was applied to the period January 1, 2007 to January 1,
2008. For this assessment, North American Mesoscale (NAM) gridded meteorological data on a 12 km
grid were obtained from the NOAA Operational Model Archive and Distribution System (NOMADS) for the
full 2007 year period (January 1, 2007 to January 1, 2008). These data were used as input to the MM5
model to produce finer scale meteorological data on a 4 km grid. The meteorological data produced by
MM5 model (a mesoscale meteorological model produced by Penn State/NCAR) were used as an initial
guess field in CALMET model.
The relative location of MM5 4km grid points used as inputs into CALMET is shown with respect to the
computational domain in Figure 7C-5.
PREPARED BY
PREPARED FOR
FIGURE NO.
340000 345000 350000 355000 360000 365000 370000 375000 380000 385000
340000 345000 350000 355000 360000 365000 370000 375000 380000 385000
596
50
00
597
00
00
597
500
05
98
000
05
985
00
05
990
00
05
995
00
060
00
00
060
05
000
60
10
000
596
50
00
597
00
00
597
500
05
98
000
05
985
00
05
990
00
05
995
00
060
00
00
060
05
000
60
10
000
Distance in Metres
NAD 83 UTM Zone 11
0 5000 10000 15000
Project Site
P:/
JW
_P
roje
cts
/Po
st
CM
IC/1
03
337
2/A
ir Q
ualit
y/R
EP
OR
TS
/Te
ch
nic
al A
ppe
nd
ices/A
pp
en
dix
C/A
ppe
nd
ix C
Fig
ure
s/F
igu
re C
-5.s
rf
HR MILNER EXPANSION PROJECT ENVIRONMENTAL ASSESSMENT
Location of MM5 4km Grid Points used as Meteorological Input
7C-5
Map Features
95
0
11
00
12
50
14
00
15
50
17
00
18
50
20
00
21
50
23
00
24
50
Terrain Elevation (m asl)
MM5 Grid Point
Proposed HR Milner Expansion Project Environmental Impact Assessment Report
Appendix 7C: CALPUFF and CALMET Methods and Assumptions
Maxim Power Corp. Page 7C-12 January 2009
7C.1.5.1 Wind
CALMET has traditionally been initialized with meteorological inputs from surface stations within the
region of interest as well as information from nearby twice-daily radiosonde stations. However, the use of
prognostic meteorological fields output from models such as MM5 is increasingly being used as input for
the CALMET model. The primary advantages of using prognostic data to help initialize CALMET are as
follows:
Prognostic model output can provide input data at higher spatial resolution than can radiosonde data,and, as such, is potentially better able to represent mesoscale meteorological circulations;
In remote locations, without nearby surface stations, prognostic data can provide reasonableestimates of local surface meteorological conditions;
While radiosonde data is only available twice daily, prognostic models can provide CALMET withinitialization data at hourly increments.
Surface Winds
Wind roses summarizing hourly winds predicted by CALMET (Jan 01 to Dec 31, 2007) and hourly wind
measurements (Dec 06 to 31, 2007) are shown in Figure 7C-6 for the project site. Wind roses are an
efficient and convenient means of presenting wind data. The length of the radial barbs gives the total
percent frequency of winds from the indicated direction while portions of the barbs of different widths
indicate the frequency of associated wind speed categories.
Wind roses in Figure 7C-6 show local winds are largely affected by the variation in terrain surrounding the
projects site. The wind flow in the vicinity of the proposed Project site location is probably dominated by
the westerly winds. Note that due to lack of real observations, we have relied on the numerical
meteorological model MM5 to estimate site specific wind patterns
Proposed HR Milner Expansion Project Environmental Impact Assessment Report
Appendix 7C: CALPUFF and CALMET Methods and Assumptions
Maxim Power Corp. Page 7C-13 January 2009
Measured Winds at Project site
(Dec 06 - 31, 2007)
Modelled Winds at Project Site
(Dec 06 - 31, 2007)
NORTH
SOUTH
WEST EAST
13%
26%
39%
52%
65%
WIND SPEED
(m/s)
>= 10.0
8.0 - 10.0
6.0 - 8.0
4.0 - 6.0
2.0 - 4.0
0.1 - 2.0
Calms: 6.41%
NORTH
SOUTH
WEST EAST
13%
26%
39%
52%
65%
WIND SPEED
(m/s)
>= 10.0
8.0 - 10.0
6.0 - 8.0
4.0 - 6.0
2.0 - 4.0
0.1 - 2.0
Calms: 0.00%
Modelled Winds at Project site
(Jan 01 to Dec 31, 2007)
NORTH
SOUTH
WEST EAST
13%
26%
39%
52%
65%
WIND SPEED
(m/s)
>= 10.0
8.0 - 10.0
6.0 - 8.0
4.0 - 6.0
2.0 - 4.0
0.1 - 2.0
Calms: 0.02%
Figure 7C-6 Wind Roses Depicting Hourly Surface Winds at project site (2007)
Proposed HR Milner Expansion Project Environmental Impact Assessment Report
Appendix 7C: CALPUFF and CALMET Methods and Assumptions
Maxim Power Corp. Page 7C-14 January 2009
Surface Wind Vector Plots
Wind vector plots are a useful means of evaluating model performance by assessing the relative realism
of wind-flow patterns. In the diagram, an arrow is shown to represent the direction and velocity of the
wind for each meteorological grid cell. The direction of the arrow indicates the direction that the wind is
blowing towards and the size of the arrow indicates the relative wind speed. Figures 7C-7, 7C-8 and 7C-9
present sample wind vector diagrams depicting surface wind flow at three meteorologically-different
simulation hours over the study area.
Figure 7C-7 shows the wind field as a vector plot at 15:00 LST on July 17, 2007 under convective
conditions (Pasquill-Gifford (PG) class B). The general airflow in the west and southwest part of the
domain appears to be from the southwest and south directions while easterly winds dominate the
northeast part of domain. Winds at the site are from the southwest. Low wind speeds are predicted in the
northeast of valley, with higher wind speeds occurring in the west part of the domain.
Figure 7C-8 shows the wind field as a vector plot at 02:00 LST on January 28, 2007 under very stable
conditions (PG class F). North-westerly flow dominates the domain for this hour. Winds around the project
site shows strong local terrain effects.
Figure 7C-9 shows the wind field as a vector plot at 13:00 LST on July 1, 2007 under neutral conditions
(PG class D). A west-southwest flow dominates the domain for this hour.
The vector plots presented in Figures 7C-7 to 7C-9 were not selected to be representative of any select
meteorological condition. The vector plots are shown as examples of the variability of the airflow that can
occur over the 50 by 50 km model domain in any given hour.
Predicted Upper Wind Plots
Figure C–10 shows the wind roses predicted by CALMET for the Project site for varying elevations above
ground level. Model-output values were extracted from the CALMET grid point nearest to the site location.
The results indicate that south-westerly winds dominate all five levels while there is a tendency for more
westerly winds with increasing height above the ground.
PREPARED BY
PREPARED FOR
FIGURE NO.
340000 345000 350000 355000 360000 365000 370000 375000 380000 385000
340000 345000 350000 355000 360000 365000 370000 375000 380000 385000
596
50
00
597
000
05
97
500
05
980
00
05
985
00
05
990
00
059
95
00
060
00
000
60
05
000
60
10
000
596
50
00
597
000
05
97
500
05
980
00
05
985
00
05
990
00
059
95
00
060
00
000
60
05
000
60
10
000
Distance in Metres
NAD 83 UTM Zone 11
0 5000 10000 15000
Project Site
Map FeaturesP
:/JW
_P
roje
cts
/Post C
MIC
/10
33
37
2/A
ir Q
ua
lity/R
EP
OR
TS
/Techn
ical A
ppe
ndic
es/A
pp
en
dix
C/A
pp
en
dix
C F
igu
res/F
igure
C-7
.srf
HR MILNER EXPANSION PROJECT ENVIRONMENTAL ASSESSMENT
Predicted Surface Wind Field for Unstable Conditions: July 17, 2007 at 15:00 LST
C-7
95
0
11
00
12
50
14
00
15
50
17
00
18
50
20
00
21
50
23
00
24
50
Terrain Elevation (m asl)
Wind Vector
PREPARED BY
PREPARED FOR
FIGURE NO.
340000 345000 350000 355000 360000 365000 370000 375000 380000 385000
340000 345000 350000 355000 360000 365000 370000 375000 380000 385000
596
50
00
597
000
05
97
500
05
980
00
05
985
00
05
990
00
059
95
00
060
00
000
60
05
000
60
10
000
596
50
00
597
000
05
97
500
05
980
00
05
985
00
05
990
00
059
95
00
060
00
000
60
05
000
60
10
000
Distance in Metres
NAD 83 UTM Zone 11
0 5000 10000 15000
Project Site
Map FeaturesP
:/JW
_P
roje
cts
/Post C
MIC
/10
33
37
2/A
ir Q
ua
lity/R
EP
OR
TS
/Techn
ical A
ppe
ndic
es/A
pp
en
dix
C/A
pp
en
dix
C F
igu
res/F
igure
C-8
.srf
HR MILNER EXPANSION PROJECT ENVIRONMENTAL ASSESSMENT
Predicted Surface Wind Field for Stable Conditions: January 28, 2007 at 02:00 LST
7C-8
95
0
11
00
12
50
14
00
15
50
17
00
18
50
20
00
21
50
23
00
24
50
Terrain Elevation (m asl)
Wind Vector
PREPARED BY
PREPARED FOR
FIGURE NO.
340000 345000 350000 355000 360000 365000 370000 375000 380000 385000
340000 345000 350000 355000 360000 365000 370000 375000 380000 385000
596
50
00
597
000
05
97
500
05
980
00
05
985
00
05
990
00
059
95
00
060
00
000
60
05
000
60
10
000
596
50
00
597
000
05
97
500
05
980
00
05
985
00
05
990
00
059
95
00
060
00
000
60
05
000
60
10
000
Distance in Metres
NAD 83 UTM Zone 11
0 5000 10000 15000
P:/
JW
_P
roje
cts
/Post C
MIC
/10
33
37
2/A
ir Q
ua
lity/R
EP
OR
TS
/Te
chn
ical A
ppe
nd
ice
s/A
pp
en
dix
C/A
ppe
nd
ix C
Fig
ure
s/F
igu
re C
-9.s
rf
HR MILNER EXPANSION PROJECT ENVIRONMENTAL ASSESSMENT
Predicted Surface Wind Field for Neutral Conditions: July 1, 2007 at 13:00 LST
7C-9
Project Site
Map Features
95
0
11
00
12
50
14
00
15
50
17
00
18
50
20
00
21
50
23
00
24
50
Terrain Elevation (m asl)
Wind Vector
Proposed HR Milner Expansion Project Environmental Impact Assessment Report
Appendix 7C: CALPUFF and CALMET Methods and Assumptions
Maxim Power Corp. Page 7C-18 January 2009
NORTH
SOUTH
WEST EAST
6%
12%
18%
24%
30%
WIND SPEED(m/s)
>= 10.0
8.0 - 10.0
6.0 - 8.0
4.0 - 6.0
2.0 - 4.0
0.1 - 2.0
Calms: 0.08%
240 m
NORTH
SOUTH
WEST EAST
6%
12%
18%
24%
30%
WIND SPEED(m/s)
>= 10.0
8.0 - 10.0
6.0 - 8.0
4.0 - 6.0
2.0 - 4.0
0.1 - 2.0
Calms: 0.07%
120 m
NORTH
SOUTH
WEST EAST
7%
14%
21%
28%
35%
WIND SPEED(m/s)
>= 10.0
8.0 - 10.0
6.0 - 8.0
4.0 - 6.0
2.0 - 4.0
0.1 - 2.0
Calms: 0.03%
60 m
NORTH
SOUTH
WEST EAST
7%
14%
21%
28%
35%
WIND SPEED
(m/s)
>= 10.0
8.0 - 10.0
6.0 - 8.0
4.0 - 6.0
2.0 - 4.0
0.1 - 2.0
Calms: 0.07%
30 m
NORTH
SOUTH
WEST EAST
8%
16%
24%
32%
40%
WIND SPEED
(m/s)
>= 10.0
8.0 - 10.0
6.0 - 8.0
4.0 - 6.0
2.0 - 4.0
0.1 - 2.0
Calms: 0.14%
10 m
Figure 7C-10 Predicted Winds at Various Elevations above the Project Site
Proposed HR Milner Expansion Project Environmental Impact Assessment Report
Appendix 7C: CALPUFF and CALMET Methods and Assumptions
Maxim Power Corp. Page 7C-19 January 2009
7C.1.5.2 Stability and Mixing Heights
Atmospheric Stability
Atmospheric turbulence near the earth’s surface is often described in terms of atmospheric stability, which
is governed by both thermal and mechanical factors. Meteorologists define six stability classes (referred
to as the Pasquill Gifford [PG] classes):
Stability classes A, B and C occur during the day, when the earth is heated by solar radiation. The airnext to the earth is heated and tends to rise, enhancing vertical motions. This is referred to as anunstable atmosphere.
Stability class D is associated with completely overcast conditions (day or night) when there is no netheating or cooling of the earth, transitional periods between stable and unstable conditions, or duringhigh wind speed periods (winds greater than 6 m/s [or 22 km/h]). This is referred to as a neutralatmosphere.
Stability classes E and F occur during the night, when the earth cools due to long-wave radiationlosses. The air next to the earth cools, suppressing vertical motions. This is referred to as a stableatmosphere.
Stability classes undergo a significant daily variation, and they have a seasonal dependence. Stability
classes can be determined from routine airport observations using the method devised by Turner (1963).
A stability classification calculation algorithm is also included in the CALMET model. Table 7C-4 presents
the frequency of predicted seasonal PG stability classes at the Project site on a seasonal and annual
basis.
Atmospheric conditions at the proposed site location are neutral and stable at most times during the year.
Stable conditions occur less frequently in spring and summer than in fall and winter. Unstable conditions
occur more frequently during the spring and summer months than during fall and winter as convective
conditions are more prominent during this time of year.
Table 7C-4 Frequency of Predicted PG Stability Classes at the Project Site
CaseNumber
ofHours
A B C D E F
VeryUnstable
ModeratelyUnstable
SlightlyUnstable
NeutralModerately
StableVery
Stable
Winter 2160 0.0 2.5 10.7 30.1 26.9 29.7
Spring 2208 1.0 17.3 20.6 20.6 9.4 31.1
Summer 2208 2.2 24.5 20.2 21.8 8.5 22.8
Fall 2184 0.1 5.7 18.2 26.2 16.6 33.2
Year 8760 0.8 12.6 17.5 24.7 15.3 29.2
Mixing Heights
The mixing height is the depth of the unstable air in the atmospheric boundary layer and is influenced by
mechanical and buoyant forces. The height of the mixing layer is an extremely important factor in
determining the dispersion of pollution in the atmosphere. Under low mixing heights, a relatively small
emission amount can have a marked effect on local air quality.
The CALMET model calculates a maximum mixing height, as determined by either convective or
mechanical forces. The convective mixing height is the height to which an air package will rise under the
Proposed HR Milner Expansion Project Environmental Impact Assessment Report
Appendix 7C: CALPUFF and CALMET Methods and Assumptions
Maxim Power Corp. Page 7C-20 January 2009
buoyant forces created by the heating of the earth’s surface. The convective mixing height is dependent
on solar radiation amount, wind speed, as well as the vertical temperature structure of the atmosphere.
Mechanical mixing heights are, similarly, the height to which an air package will rise under the influence
of mechanical-invoked turbulence. The mechanical mixing height is proportional to low-level wind speeds
and surface roughness.
For this assessment, the CALMET post-processor was used to extract the mixing heights from CALMET
output files, and the mixing height predictions for the Project site are provided in Figure 7C-12. The mean
maximum afternoon values during winter, spring, summer and fall are approximately 800, 1500, 1600,
and 1100 m, respectively. The minimum values for each season are predicted to occur during the night.
Figure 7C-12 also shows that the mixing heights during the winter and fall months exhibit less of a diurnal
fluctuation that the mixing heights in spring and summer.
Proposed HR Milner Expansion Project Environmental Impact Assessment Report
Appendix 7C: CALPUFF and CALMET Methods and Assumptions
Maxim Power Corp. Page 7C-21 January 2009
NOTE:
Winter: December, January, February
Summer: June, July, August
Spring: March, April, May
Fall: September, October November
Figure 7C-11 Predicted Mixing Heights for Different Seasons and Times of Day
Proposed HR Milner Expansion Project Environmental Impact Assessment Report
Appendix 7C: CALPUFF and CALMET Methods and Assumptions
Maxim Power Corp. Page 7C-22 January 2009
7C.1.6 CALMET Technical Options
The technical options used in running CALMET are entered through a CALMET control file. The input
parameters for the CALMET control file used in the Project modelling assessment are provided in Tables
7C-5 to 7C-12. Model default values, as recommended by the United States Environmental Protection
Agency (U.S. EPA 1998), are presented for comparative purposes. In most cases, these default values
were used.
Table 7C-5 CALMET Model Options Groups 0: Input and Output File Names
Parameter Default Project Comment
NUSTA - 0 Number of upper air stations
NOWSTA - 0 Number of overwater met stations
MM3D - 1 Number of MM4/MM5/3D.DAT files
NIGF - 0 Number of IGF-CALMET.DAT files
Table 7C-6 CALMET Model Options Groups 1: General Run ControlParameters
Parameter Default Project Comment
IBYR - 2007 Starting year
IBMO - 1 Starting month
IBDY - 1 Starting day
IBHR - 0 Starting hour
IBSEC - 0 Starting second
IEYR - 2008 Ending year
IEMO - 1 Ending month
IEDY - 1 Ending day
IEHR - 0 Ending hour
IESEC - 0 Ending second
ABTZ - UTC-0700 UTC time zone
NSECDT 3600 3600 Length of modeling time-step (seconds)
IRTYPE 1 1 Run type
LCALGRD T T Special data fields
ITEST 2 2 Flag to stop run after SETUP phase
Proposed HR Milner Expansion Project Environmental Impact Assessment Report
Appendix 7C: CALPUFF and CALMET Methods and Assumptions
Maxim Power Corp. Page 7C-23 January 2009
Table 7C-7 CALMET Model Options Group 2: Grid Control Parameters
Parameter Default Project Comment
PMAP UTM UTM Map projection
IUTMZN - 11 UTM Zone
UTMHEM N N Hemisphere for UTM projection
DATUM WGS-84 WGS-84 Datum-region for output coordinate
NX - 100 Number of X grid cells
NY - 100 Number of Y grid cells
DGRIDKM - 0.5 Grid spacing (km)
XORIGKM - 337.0 Reference coordinate of SW corner of grid cell (1,1) -Xcoordinate (km)
YORIGKM - 5961.0 Reference coordinate of SW corner of grid cell (1,1) -Ycoordinate (km)
NZ - 8 Number of vertical grid cells
ZFACE - 0,20,40,80,
160,320,600,
1400,2600
Vertical grid definition: Cell face heights in arbitrary verticalgrid (m)
Table 7C-8 CALMET Model Options Group 3: Output Options
Parameter Default Project Comment
Disk Output:
LSAVE T T Save met data in unformatted output files
IFORMO 1 1 Type of unformatted output file
Line Printer Output:
LPRINT F F Print meteorological fields
IPRINF 1 12 Print interval (hrs)
IUVOUT (NZ) 0 1,0,0,0,0,
0,0,0
Specify which layers of U,V wind component to print
IWOUT (NZ) 0 0,0,0,0,0,
0,0,0
Specify which level of the w wind component to print
ITOUT (NZ) 0 1,0,0,0,0,
0,0,0
Specify which levels of the 3-D temperature field to print
Meteorological fields to print:
VariablePrint?
0 = no print1 = print
Comment
STABILITY 1 PGT stability
USTAR 0 Friction velocity
MONIN 0 Monin-Obukhov length
MIXHT 1 Mixing height
WSTAR 0 Convective velocity scale
PRECIP 1 Precipitation rate
SENSHEAT 0 Sensible heat flux
CONVZI 0 Convective mixing height
Proposed HR Milner Expansion Project Environmental Impact Assessment Report
Appendix 7C: CALPUFF and CALMET Methods and Assumptions
Maxim Power Corp. Page 7C-24 January 2009
Table 7C-8 CALMET Model Options Group 3: Output Options (cont’d)
Parameter Default Project Comment
Testing and debug print options for micrometeorological module:
LDB F F Print input meteorological data and internal variables
NN1 1 1 First time step for which debug data are printed
NN2 1 1 Last time step for which debug data are printed
LDBCST F F Print distance to land internal variables
Testing and debug print options for wind field module:
IOUTD 0 0 Control variable for writing the test/debug wind fieldsto disk files
NZPRN2 1 0 Number of levels, starting at surface, to print
IPR0 0 0 Interpolated wind components
IPR1 0 0 Terrain adjusted surface wind components
IPR2 0 0 Smoothed wind components and the initialdivergence fields
IPR3 0 0 Final wind speed and direction
IPR4 0 0 Final divergence fields
IPR5 0 0 Winds after kinematic effects are added
IPR6 0 0 Winds after the Froude number adjustment is made
IPR7 0 0 Winds after slope flows are added
IPR8 0 0 Final wind field components
Table 7C-9 CALMET Model Options Group 4: Meteorological Data Options
Parameter Default Project Comment
NOOBS 0 2 No surface, overwater, or upper air observations
Number of Surface & Precipitation Meteorological Stations:
NSSTA - 0 Number of surface stations
NPSTA - -1 Number of precipitation stations
Cloud Data Options:
ICLOUD 0 3 Gridded cloud fields
File Formats:
IFORMS 2 2 Surface meteorological data file format
IFORMP 2 2 Precipitation data file format
IFORMC 2 2 Cloud data file format
Proposed HR Milner Expansion Project Environmental Impact Assessment Report
Appendix 7C: CALPUFF and CALMET Methods and Assumptions
Maxim Power Corp. Page 7C-25 January 2009
Table 7C-10 CALMET Model Option Group 5: Wind Field Options andParameters
Parameter Default Project Comment
Wind Field Model Options:
IWFCOD 1 1 Model selection variables
IFRADJ 1 1 Compute Froude number adjustment
IKINE 0 0 Compute kinematic effects
IOBR 0 0 Use O’Brien procedure for adjustment of the vertical velocity
ISLOPE 1 1 Compute slope flow effects
IEXTRP -4 -1 Extrapolate surface wind observations to upper layers(similarity theory used with layer 1 data at upper air stationsignored)
ICALM 0 0 Extrapolate surface winds even if calm
BIAS 8*0 8*0 Layer-dependent biases modifying the weights of surface andupper air stations
RMIN2 4 -1 Minimum distance from nearest upper air station to surfacestation for which extrapolation of surface winds at surfacestation will be allowed
IPROG 0 14 Use gridded prognostic wind field model output fields as inputto the diagnostic wind field model (from MM5.DAT)
ISTEPPG 1 1 Timestep (hours) of the prognostic model input data
IGFMET 0 0 Use coarse CALMET fields as initial guess fields
Radius of Influence Parameters:
LVARY F F Use varying radius of influence
RMAX1 - 12 Maximum radius of influence over land in the surface layer (km)
RMAX2 - 12 Maximum radius of influence over land aloft (km)
RMAX3 - 5 Maximum radius of influence over water
Other Wind Field Input Parameters:
RMIN 0.1 0.1 Minimum radius of influence used in the wind field interpolation(km)
TERRAD - 15 Radius of influence of terrain features (km)
R1 - 3 Relative weighting of the first guess field and observations inthe surface layer (km)
R2 - 3 Relative weighting of the first guess field and observations inthe layers aloft (km)
RPROG - 0 Relative weighting parameter of the prognostic wind field data(km)
DIVLIM 5.0E-6 5.0E-6 Maximum acceptable divergence in the divergenceminimization procedure
Proposed HR Milner Expansion Project Environmental Impact Assessment Report
Appendix 7C: CALPUFF and CALMET Methods and Assumptions
Maxim Power Corp. Page 7C-26 January 2009
Table 7C-10 CALMET Model Option Group 5: Wind Field Options andParameters (cont’d)
Parameter Default Project Comment
NITER 50 50 Maximum number of iterations in the divergenceminimization procedure
NSMTH 2,
(mxnz-1)*4
2,7,7,14,14,28,28,28
Number of passes in the smoothing procedure
NINTR2 99*8 99,99,99,99,
99,99,99,0
Maximum number of stations used in each layer for theinterpolation of data to a grid point
CRITFN 1.0 1.0 Critical Froude number
ALPHA 0.1 0.1 Empirical factor controlling the influence of kinematiceffects
FEXTR2(NZ) 0*8 0*8 Multiplicative scaling factor for extrapolation of surfaceobservations to upper layers
Barrier Information:
NBAR 0 0 Number of barriers to interpolation of the wind fields
KBAR NZ 8 Level (1 to NZ) up to which barriers apply
XBBAR - 0 X coordinate of beginning of each barrier
YBBAR - 0 Y coordinate of beginning of each barrier
XEBAR - 0 X coordinate of ending of each barrier
YEBAR - 0 Y coordinate of ending of each barrier
Diagnostic Module Data Input Options:
IDIOPT1 0 0 Surface temperature (0 = compute internally from hourlysurface observation)
ISURFT - 3 Surface meteorological station to use for the surfacetemperature
IDIOPT2 0 0 Domain-averaged temperature lapse (0 = computeinternally from hourly surface observation)
IUPT - 0 Upper air station to use for the domain-scale lapse rate
ZUPT 200 200 Depth through which the domain-scale lapse rate iscomputed (m)
IDIOPT3 0 0 Domain-averaged wind components
IUPWND -1 -1 Upper air station to use for the domain-scale winds
ZUPWND 1, 1000 1, 2500 Bottom and top of layer through which domain-scale windsare computed (m)
IDIOPT4 0 0 Observed surface wind components for wind field module
IDIOPT5 0 0 Observed upper air wind components for wind field module
Lake Breeze Information:
LLBREZE F F Lake breeze module
NBOX - 0 Number of lake breeze regions
XG1 - 0 X Grid line 1 defining the region of interest
Proposed HR Milner Expansion Project Environmental Impact Assessment Report
Appendix 7C: CALPUFF and CALMET Methods and Assumptions
Maxim Power Corp. Page 7C-27 January 2009
Table 7C-10 CALMET Model Option Group 5: Wind Field Options andParameters (cont’d)
Parameter Default Project Comment
XG2 - 0 X Grid line 2 defining the region of interest
YG1 - 0 Y Grid line 1 defining the region of interest
YG2 - 0 Y Grid line 2 defining the region of interest
XBCST - 0 X Point defining the coastline in kilometres (Straight line)
YBCST - 0 Y Point defining the coastline in kilometres (Straight line)
XECST - 0 X Point defining the coastline in kilometres (Straight line)
YECST - 0 Y Point defining the coastline in kilometres (Straight line)
NLB - 0 Number of stations in the region
METBXID - 0 Station ID’s in the region
Table 7C-11 CALMET Model Option Group 6: Mixing Height, Temperature andPrecipitation Parameters
Parameter Default Project Comment
Empirical Mixing Height Constants:
CONSTB 1.41 1.41 Neutral, mechanical equation
CONSTE 0.15 0.15 Convective mixing height equation
CONSTN 2400 2400 Stable mixing height equation
CONSTW 0.16 0.16 Over water mixing height equation
FCORIO 1.0E-4 1.0E-04 Absolute value of Coriolis (l/s)
Spatial Averaging of Mixing Heights:
IAVEZI 1 1 Conduct spatial averaging
MNMDAV 1 2 Maximum search radius in averaging (grid cells)
HAFANG 30 30 Half-angle of upwind looking cone for averaging
ILEVZI 1 1 Layer of winds used in upwind averaging
Convective Mixing Heights Options:
IMIXH 1 1 Method to compute the convective mixing height (Maul-Carson)
THRESHL 0.05 0.05 Threshold buoyancy flux required to sustain convectivemixing height growth overland (W/m
3)
THRESHW 0.05 0.05 Threshold buoyancy flux required to sustain convectivemixing height growth overwater (W/m
3)
ITWPROG 0 0 Option for overwater lapse rates used in convective mixingheight growth (1=use prognostic lapse rates)
ILUOC3D 16 16 Land use category ocean in 3D.DAT datasets
Proposed HR Milner Expansion Project Environmental Impact Assessment Report
Appendix 7C: CALPUFF and CALMET Methods and Assumptions
Maxim Power Corp. Page 7C-28 January 2009
Table 7C-11 CALMET Model Option Group 6: Mixing Height, Temperature andPrecipitation Parameters (cont’d)
Parameter Default Project Comment
Other Mixing Height Variables:
DPTMIN 0.001 0.001 Minimum potential temperature lapse rate in the stablelayer above the current convective mixing height (K/m)
DZZI 200 200 Depth of layer above current convective mixing heightthrough which lapse rate is computed (m)
ZIMIN 50 50 Minimum overland mixing height (m)
ZIMAX 3000 3000 Maximum overland mixing height (m)
ZIMINW 50 50 Minimum overwater mixing height (m)
ZIMAXW 3000 3000 Maximum overwater mixing height (m)
Overwater Surface Fluxes Method and Parameters:
ICOARE 10 10 COARE with no wave parameterization
DSHELF 0 0 Coastal/Shallow water length scale (km)
IWARM 0 0 COARE warm layer computation
ICOOL 0 0 COARE cool skin layer computation
Relative Humidity Parameters:
IRHPROG 0 1 3D relative humidity from observations or from prognosticdata
Temperature Parameters:
ITPROG 0 1 3D temperature from observations or from prognostic data
IRAD 1 1 Interpolation type
TRADKM 500 500 Radius of influence for temperature interpolation (km)
NUMTS 5 5 Maximum number of stations to include in temperatureinterpolation
IAVET 1 1 Conduct spatial averaging of temperatures (1 = yes)
TGDEFB -0.0098 -0.0098 Default temperature gradient below the mixing height overwater (K/m)
TGDEFA -0.0045 -0.0045 Default temperature gradient above the mixing height overwater (K/m)
JWAT1 - 999 Beginning land use categories for temperatureinterpolation over water
JWAT2 - 999 Ending land use categories for temperature interpolationover water
Precipitation Interpolation Parameters:
NFLAGP 2 2 Method of interpolation
SIGMAP 100 180 Radius of Influence (km)
CUTP 0.01 0.01 Minimum Precipitation rate cut-off (mm/h)
Proposed HR Milner Expansion Project Environmental Impact Assessment Report
Appendix 7C: CALPUFF and CALMET Methods and Assumptions
Maxim Power Corp. Page 7C-29 January 2009
7C.2 CALPUFF MODELLING
7C.2.1 Model Description
The following description of the CALPUFF model’s major model algorithms and options are all excerpts
from the CALPUFF model’s user manual (Scire et al. 2000b).
The CALPUFF model is a non-steady-state Gaussian puff dispersion model which incorporates simple
chemical transformation mechanisms, wet and dry deposition, complex terrain algorithms and building
downwash. The CALPUFF model is suitable for estimating ground-level air quality concentrations on
both local and regional scales, from tens of meters to hundreds of kilometres. It can accommodate
arbitrarily varying point sources and gridded area source emissions. Most of the algorithms contain
options to treat the physical processes at different levels of detail depending on the model application.
The major features and options of the CALPUFF model are summarized in Table 7C-12. Some of the
technical algorithms are briefly described below.
Proposed HR Milner Expansion Project Environmental Impact Assessment Report
Appendix 7C: CALPUFF and CALMET Methods and Assumptions
Maxim Power Corp. Page 7C-30 January 2009
Table 7C-12: Summary of Major Features of CALPUFF
Proposed HR Milner Expansion Project Environmental Impact Assessment Report
Appendix 7C: CALPUFF and CALMET Methods and Assumptions
Maxim Power Corp. Page 7C-31 January 2009
Table 7C-12: Summary of Major Features of CALPUFF (Continued…)
Proposed HR Milner Expansion Project Environmental Impact Assessment Report
Appendix 7C: CALPUFF and CALMET Methods and Assumptions
Maxim Power Corp. Page 7C-32 January 2009
Chemical Transformation: CALPUFF includes options for parameterizing chemical transformation
effects using the five species scheme (SO2, SO, NOx, HNO3, and NO) employed in the MESOPUFF II
model, the six species RIVAD/ARM3 scheme, or a set of user-specified, diurnally-varying transformation
rates. The RIVAD/ARM3 reactions separately model NO and NO2 rather than NOx. Calculations of
chemical transformations require, among other information, a knowledge of background concentrations of
ozone and ammonia.
Subgrid Scale Complex Terrain: The complex terrain module in CALPUFF is based on the approach
used in the Complex Terrain Dispersion Model (CTDMPLUS) (Perry et al., 1989). Plume impingement on
subgrid scale hills is evaluated using a dividing streamline (Hd) to determine which pollutant material is
deflected around the sides of a hill (below Hd) and which material is advected over the hill (above Hd).
Individual puffs are split in up to three sections for these calculations.
Puff Sampling Functions: A set of accurate and computationally efficient puff sampling routines are
included in CALPUFF which solve many of the computational difficulties with applying a puff model to
near-field releases. For near-field applications during rapidly varying meteorological conditions, an
elongated puff (slug) sampling function can be used. An integrated puff approach is used during less
demanding conditions. Both techniques reproduce continuous plume results exactly under the
appropriate steady state conditions.
Wind Shear Effects: CALPUFF contains an optional puff splitting algorithm that allows vertical wind
shear effects across individual puffs to be simulated. Differential rates of dispersion and transport occur
on the puffs generated from the original puff, which under some conditions can substantially increase the
effective rate of horizontal growth of the plume.
Building Downwash: The Huber-Snyder and Schulman-Scire downwash models are both incorporated
into CALPUFF. An option is provided to use either model for all stacks, or make the choice on a stack-by-
stack and wind sector-by-wind sector basis. Both algorithms have been implemented in such a way as to
allow the use of wind direction specific building dimensions.
Overwater and Coastal Interaction Effects: Because the CALMET meteorological model contains both
overwater and overland boundary layer algorithms, the effects of water bodies on plume transport,
dispersion, and deposition can be simulated with CALPUFF. The puff formulation of CALPUFF is
designed to handle spatial changes in meteorological and dispersion conditions, including the abrupt
changes that occur at the coastline of a major body of water.
Dispersion Coefficients: Several options are provided in CALPUFF for the computation of dispersion
coefficients, including the use of turbulence measurements (σv and σw), the use of similarity theory to
estimate σv and σw from modelled surface heat and momentum fluxes, or the use of Pasquill-Gifford (PG)
or McElroy-Pooler (MP) dispersion coefficients, or dispersion equations based on the Complex Terrain
Dispersion Model (CTDM). Options are provided to apply an averaging time correction or surface
roughness length adjustment to the PG coefficients.
Dry Deposition: A full resistance model is provided in CALPUFF for the computation of dry deposition
rates of gases and particulate matter as a function of geophysical parameters, meteorological conditions,
and pollutant species. Options are provided to allow user-specified, diurnally varying deposition velocities
to be used for one or more pollutants instead of the resistance model (e.g., for sensitivity testing) or to by-
pass the dry deposition model completely.
Wet Deposition: An empirical scavenging coefficient approach is used in CALPUFF to compute the
depletion and wet deposition fluxes due to precipitation scavenging. The scavenging coefficients are
specified as a function of the pollutant and precipitation type (i.e., frozen vs. liquid precipitation).
Proposed HR Milner Expansion Project Environmental Impact Assessment Report
Appendix 7C: CALPUFF and CALMET Methods and Assumptions
Maxim Power Corp. Page 7C-33 January 2009
7C.2.2 Model Initialization
7C.2.2.1 Computational Domain
Dispersion modeling was conducted using CALPUFF over a computational domain equal to the CALMET
meteorological grid defined in Section 2.0 of this Appendix. The CALPUFF computational domain is the
area in which the transport and dispersion of puffs are considered for the modelling.
7C.2.2.2 Meteorological Data
Meteorological data such as mixing heights, stability and winds determine the transport and dispersion of
pollutants within the CALPUFF model. To capture puff behaviour under a variety of meteorological
conditions, one year of modelling was considered for this application. Hourly three-dimensional
meteorological fields for the year 2007 were prepared using the CALMET model, as described in Section
2.0 of this Appendix.
7C.2.2.3 Emissions and Source Characteristics
CALPUFF was used to model the dispersion of emissions from the source combinations specified for
each of the four distinct cases presented in the Air Quality Technical Data Report (TDR). Rates of
emission for each species of concern as well as source characteristics used in the modelling are
discussed in the main body of the Air Quality TDR.
7C.2.2.4 Receptor Grids
Multiple receptor networks centered on the Project site were established for the purposes of dispersion
modelling. The grids and their corresponding receptor spacing are:
50 km by 50 km, with 1000 m spacing
20 km by 20km, with 500 m spacing
10 km by 10 km, with 250 m spacing
4 km by 4 km, with 50 m spacing
20 m spacing along the Project boundary and in areas of maximum predicted effect
A number of schools, hospitals and residences were selected as sensitive receptors within the study area
such that maximum predicted ground-level concentrations of air contaminants of interest could be
determined for these locations. Table 7C-13 shows the location of sensitive receptors included in
dispersion modelling within the air quality study area.
The areas of applicability of the Alberta ambient air quality objectives are not defined; however, they are
usually interpreted as applying to areas where there is public access (e.g. beyond the plant boundary). In
the case of large industrial facilities without clearly defined fencelines (e.g. mines and haul roads) the
definition of this boundary between the areas where the AAAQO is deemed to apply and not apply is
usually interpreted as a disturbed area boundary. In this assessment, the disturbed area boundary for
industrial facilities, mines and haul roads were defined by the emission source locations of all of the
individual area sources used to represent the sources, plus a 200 m buffer zone immediately surrounding
the individual sources. Areas with a reasonable expectation of public access, such as Highway 40, were
included as receptors.
Proposed HR Milner Expansion Project Environmental Impact Assessment Report
Appendix 7C: CALPUFF and CALMET Methods and Assumptions
Maxim Power Corp. Page 7C-34 January 2009
Table 7C-13: Sensitive Receptors Included in Dispersion Modelling
ReceptorUTM NAD83
mE mN Zone
Grand Cache Hospital 360684 5973530 11
Grand Cache Community High school 360244 5973173 11
Day Care 360183 5973022 11
Summitview Middle school 360404 5973011 11
Sheldon Coates Elementary School 359375 5972518 11
FN Muskeg See Pee Cooperative 392048 5974515 11
FN Susa Creek Cooperative 372311 5977528 11
FN Kamisak Enterprise 366583 5975791 11
FN Victor Lake Cooperative 362246 5971392 11
FN Joachim Enterprise 357550 5977271 11
FN Wanyandie Flats (West Cooperative) 366715 5989227 11
FN Wanyandie Flats (East Cooperative) 376724 5993049 11
Deadhorse Meadows Campsite (Kakwa) 307889 6000136 11
Trench Creek Cabin (Kakwa) 315347 5984356 11
Lower Kakwa Falls 320534 5996806 11
Sulphur Cabin (Willmore) 366012 5948587 11
Sheep Creek Patrol Cabin (Willmore) 325539 5970311 11
Onsite receptor 362224 5986378 11
Sheep Creek Lodge 348938 5982596 11
Grande Cache Lake Day Use Area 365619 5975347 11
Goat Cliffs 362131 5983713 11
Mount Hamel 355519 5982068 11
Muskeg River Corridor 369584 5982275 11
Red River/Prairie Creek Woodland Caribou Herd 340577 5988748 11
Caw Ridge 342530 5993953 11
Mountain Goat Corridor 368437 5978131 11
Wildlife Habitat 376430 5984256 11
Marv Moore Campground 362096 5973898 11
Eco Receptor 1 359748 5997932 11
Eco Receptor 2 366305 6013249 11
Eco Receptor 3 344887 6005344 11
Eco Receptor 4 369351 5995000 11
Eco Receptor 5 366170 6000809 11
Eco Receptor 6 385212 6001385 11
Grande Cache Fish, Game and Gun Club 358705 5980622 11
Grande Cache Airport 377069 5975806 11
Eco Receptor 7 352853 5993190 11
Grande Cache Institution - Minimum Security 358997 5970489 11
Cabin SRC 364901 5985440 11
Proposed HR Milner Expansion Project Environmental Impact Assessment Report
Appendix 7C: CALPUFF and CALMET Methods and Assumptions
Maxim Power Corp. Page 7C-35 January 2009
7C.2.2.5 Terrain Effects
The CALPUFF model was used to estimate concentrations, for each species considered, at each
receptor locations. Since, some of these receptors were located in terrain at elevations greater than puff
release points, terrain effects were considered. To account for the possible distortion of the plume
trajectory over elevated terrain, the Partial Plume Path Adjustment Method (PPPAM) was used to modify
the height of the plume.
The PPPAM employs a plume path coefficient (PPC) to adjust the height of the plume above the ground.
Default PPC values of 0.5, 0.5, 0.5, 0.5, 0.35, and 0.35 for Pasquill-Gifford (PG) stability classes A, B, C,
D, E, and F, respectively were used as recommended by the CALPUFF authors.
7C.2.2.6 Dispersion Coefficients
A fundamental parameter controlling plume dispersion in a Gaussian model such as CALPUFF are the
dispersion coefficients. These values, which must be specified for both the horizontal as well as the
vertical directions in the model, can be estimated using several different methods in CALPUFF. For this
application, dispersion coefficients were internally computed from turbulence estimates based on
micrometeorological data from CALMET (MDISP=2). This method was chosen over the more simplistic
default method (MDISP=3) to allow for a better characterization of dispersion in the model.
7C.2.3 Model Options
Table A-13 provides a detailed summary of all CALPUFF model user options selected for one of the
numerous CALPUFF simulations done for this assessment. Model default values, as recommended by
the United States Environmental Protection Agency (U.S. EPA 1998a), are presented for comparative
purposes. In most cases, these default values were used. Model options for CALPUFF Input Group 2 are
in accordance with the recommended values specified by the BC MOE in their current Guidelines for Air
Quality Dispersion Modelling in British Columbia (BC MOE 2006).
In addition, it should be noted that the parameterization shown in Table A-13 represents a specific
source-species-receptor combination. Therefore, application-specific model parameters such as the
number of sources and species modelled had different values for different model runs. For the simulation
specified in Table A-13, the sources are specified by the ‘Application Case’; the number of receptors is a
subset of the total nested gridded receptors, as described in Section 3.2.4.
Proposed HR Milner Expansion Project Environmental Impact Assessment Report
Appendix 7C: CALPUFF and CALMET Methods and Assumptions
Maxim Power Corp. Page 7C-36 January 2009
Table 7C-14 CALPUFF Dispersion Model User OptionsInput Group Parameter USEPA
DefaultApplied Description
Group 1: GeneralRun ControlParameters
METRUN 0 0 Run all period in met file
IBYR - 2007 Used only if METRUN=0
IBMO - 1 Used only if METRUN=0
IBDY - 1 Used only if METRUN=0
IBHR - 0 Used only if METRUN=0
XBTZ - 8 Time Zone, Pacific Standard Time
IRLG - 8760 Length of run in hours
NSPEC 5 6 Number of chemical species modelled
NSE 3 6 Number of chemical species emitted
ITEST 2 2 Continue with model execution after setup
MRESTART 0 0 Do not write a restart file
NRESPD 0 24 File updated every 24 periods
METFM 1 1 CALMET binary type of meteorological file
AVET 60 60 Averaging time is 60 minutes
PGTIME 60 60 PG Averaging time is 60 minutes
Group 2:TechnicalOptions
MGAUSS 1 1 Gaussian distribution used in the near field
MCTADJ 3 3 Partial Plume Path Adjustment Method of terrainadjustment
MCTSG 0 0 Subgrid-scale complex terrain not modelled
MSLUG 0 0 Near field puffs not elongated
MTRANS 1 1 Transitional plume rise applied
MTIP 1 1 Stack tip downwash applied
MBDW 1 2 PRIME method
MSHEAR 0 0 Vertical wind shear not modelled
MSPLIT 0 0 No puff splitting allowed
MCHEM 1 0 Chemical transformation not modelled
MAQCHEM 0 0 Aqueous phase transformation not modelled
MWET 1 0 Wet removal modelled
MDRY 1 0 Dry removal modelled
MDISP 3 2 Dispersion coefficients calculated from CALMETmicrometeorological variables
MTURBVW 3 3 Use direct turbulence measurements to estimatedispersion (Not Used)
MDISP2 3 3 Use PG coefficients when turbulencemeasurements not available
MROUGH 0 0 Sigma Y and Z are not adjusted for roughness
MPARTL 1 1 Model partial plume penetration of elevatedinversion
MTINV 0 0 Strength of temperature inversion is computed fromdefault gradients
MPDF 0 0 Use PDF to compute near-field dispersion underconvective conditions
MSGTIBL 0 0 Sub-grid TIBL module is not used
MBCON 0 0 Boundary conditions are not modelled
MFOG 0 0 Not configured for fog model output
MREG 1 0 Do not test options against defaults
Proposed HR Milner Expansion Project Environmental Impact Assessment Report
Appendix 7C: CALPUFF and CALMET Methods and Assumptions
Maxim Power Corp. Page 7C-37 January 2009
Table 7C-14 CALPUFF Dispersion Model User Options (Continued…)Input Group Parameter USEPA
DefaultApplied Description
Group 3: SpeciesList
CSPEC - SO2, SO4, NO,NO2, HNO3, NO3,NOx, CO, TSP,PM10, PM2.5
List of chemical species
- SO2 Modelled, Emitted
- SO4 Modelled
- NO Modelled, Emitted
- NO2 Modelled, Emitted
- HNO3 Modelled
- NO3 Modelled
- NOx Modelled, Emitted
- CO Modelled, Emitted
- TSP Modelled, Emitted
- PM10 Modelled, Emitted
- PM2.5 Modelled, Emitted
Group 4: GridControlParameters
PMAP UTM UTM Universal Transverse Mercator for Projection ofall X, Y
FEAST 0 0 False Easting (Not Used)
FNORTH 0 0 False Northing (Not Used)
IUTMZN - 11 UTM Zone
UTMHEM N N Northern Hemisphere
RLAT0 - 0N Latitude of Projection Origin (Not Used)
RLON0 - 0E Longitude of Projection Origin (Not Used)
XLAT1 - 0N Latitude of 1st
Parallel (Not Used)
XLAT2 - 0N Latitude of 2nd
Parallel (Not Used)
DATUM WGS-84 WGS-84 WGS-84 Reference Ellipsoid and Geoid, Globalcoverage (WGS84)
NX - 100 Number of X grid cells
NY - 100 Number of Y grid cells
NZ - 8 Number of vertical grid cells
DGRIDKM - 0.5 Grid spacing in X and Y directions (km)
ZFACE - 0, 20, 40, 80,160, 320, 600,1400, 2600
Vertical cell face heights of the NZ vertical layers
XORIGKM - 337 Reference Easting of SW corner of SW grid cellin UTM (km)
YORIGKM - 5961 Reference Northing of SW corner of SW grid cellin UTM (km)
IBCOMP - 1 X index of lower left grid cell for computation
JBCOMP - 1 Y index of lower left grid cell for computation
IECOMP - 100 X index of upper right grid cell for computation
JECOMP - 100 Y index of upper right grid cell for computation
LSAMP T F Sampling grid is not used
IBSAMP - 1 X index of lower left grid cell for sampling
JBSAMP - 1 Y index of lower left grid cell for sampling
IESAMP - 100 X index of upper right grid cell for sampling
JESAMP - 100 Y index of upper right grid cell for sampling
MESHDN 1 1 Nesting factor of sampling grid
Proposed HR Milner Expansion Project Environmental Impact Assessment Report
Appendix 7C: CALPUFF and CALMET Methods and Assumptions
Maxim Power Corp. Page 7C-38 January 2009
Table 7C-14 CALPUFF Dispersion Model User Options (Continued…)Input Group Parameter USEPA
DefaultApplied Description
Group 5: OutputOptions
ICON 1 1 Create binary concentration output file
IDRY 1 0 Create binary dry flux output file
IWET 1 0 Create binary wet flux output file
IVIS 1 0 Output file containing relative humidity is notcreated
LCOMPRS T T Apply data compression
IMFLX 0 0 Diagnostic mass flux option not applied
IMBAL 0 0 Do not report hourly mass balance for eachspecies
ICPRT 0 0 Do not print concentrations to list file
IDPRT 0 0 Do not print dry fluxes to list file
IWPRT 0 0 Do not print wet fluxes to list file
ICFRQ 1 24 Concentration print interval in hours
IDFRQ 1 24 Dry flux print interval in hours
IWFRQ 1 24 Wet flux print interval in hours
IPRTU 1 3 Output units are g/m3 for concentration andg/m2/s for fluxes
IMESG 2 2 Track progress of run on screen
- SO2 Concentrations are saved to the hard disk.Concentrations are not printed hourly.- SO4
- NO
- NO2
- HNO3
- NO3
- NOx
- CO
- TSP
- PM10
- PM2.5
LDEBUG F F Do not print debug data
IPFDEB 1 1 Debug options - First puff to track
NPFDEB 1 1 Debug options - Number of puffs to track
NN1 1 1 Debug options - Met period to start output
NN2 10 10 Debug options - Met period to end output
Group 6: SubgridScale ComplexTerrain Inputs
NHILL 0 0 Number of terrain features
NCTREC 0 0 Number of complex terrain receptors
MHILL - 2 Hill data created by OPTHILL (Not Used)
XHILL2M 1 1 Horizontal conversion factor to meters
ZHILL2M 1 1 Vertical conversion factor to meters
XCTDMKM - 0 CTDM X origin relative to CALPUFF grid
YCTDMKM - 0 CTDM Y origin relative to CALPUFF grid
Group 7: ChemicalParameters forDry Depositionof Gases
Diffusivity
Alpha Star Reactivity MesophyllResistance
Henry’s Law Coefficient
- - - - - -
Proposed HR Milner Expansion Project Environmental Impact Assessment Report
Appendix 7C: CALPUFF and CALMET Methods and Assumptions
Maxim Power Corp. Page 7C-39 January 2009
Table 7C-14 CALPUFF Dispersion Model User Options (Continued…)Input Group Parameter USEPA Default Applied
Group 8: SizeParameters forDry Depositionof Particles
Geometric Mass Mean Geometric Standard Deviation
- - -
Group 9:Miscellaneous DryDepositionParameters
RCUTR 30 30
RGR 10 10
REACTR 8 8
NINT 9 9
IVEG 1 1
Group 10: WetDepositionParameters
Liquid Precip Coef. Frozen Precip Coef.
- - -
Group 11:ChemistryParameters
MOZ 1 1 Monthly ozone values are used in chemistry
BCKO3 12*30 12*30 Monthly ozone values are used in chemistry
BCKNH3 12*1.0 12*1.0 Constant background concentration in ppb
RNITE1 0.2 0.2 Night time SO2 loss rate (% per hour)
RNITE2 2 2 Night time NOx loss rate (% per hour)
RNITE3 2 2 Night time HNO3 formation rate (% per hour)
BCKH2O2 12*1 12*1 Background H2O2 (Not Used)
BCKPMF 12*1 12*1 Background fine particulate matter (Not Used)
OFRAC 12*0.20 12*0.20 Organic fraction of fine particulate matter (NotUsed)
VCNX 12*50 12*50 VOC/NOx ratio for chemistry (Not Used)
Group 12:MiscellaneousDispersion andComputationalParameters
SYTDEP 550 550 Horizontal size of puff in meters beyond whichHeffer dispersion is applied
MHFTSZ 0 0 Do not use Heffer formulas for sigma Z
JSUP 5 5 Stability class used to determine plume growthrates for puff above the boundary layer
CONK1 0.01 0.01 Vertical dispersion constant for stable conditions
CONK2 0.1 0.1 Vertical dispersion constant for neutral/unstableconditions
TBD 0.5 0.5 Transition factor between Huber-Snyder andSchulman-Scire downwash schemes
IURB1 10 10 Lower range of land use categories for which urbandispersion is assumed
IURB2 19 19 Upper range of land use categories for which urbandispersion is assumed
ILANDUIN 20 VariesSpatially
Land use category for modelling domain
ZOIN 0.25 VariesSpatially
Roughness length in meters for domain
XLAIIN 3 VariesSpatially
Leaf area index for domain
ELEVIN 0 VariesSpatially
Elevation above sea level in meters
XLATIN -999 -999 Latitude of met location in degrees
XLONIN -999 -999 Longitude of met location in degrees
ANEMHT 10 10 Anemometer height in meters
ISIGMAV 1 1 Read sigma-v from profile file (Not Used)
Proposed HR Milner Expansion Project Environmental Impact Assessment Report
Appendix 7C: CALPUFF and CALMET Methods and Assumptions
Maxim Power Corp. Page 7C-40 January 2009
Table 7C-14 CALPUFF Dispersion Model User Options (Continued…)Input Group Parameter USEPA
DefaultApplied Input Group
Group 12:MiscellaneousDispersion andComputationalParameters
IMIXCTDM 0 0 Predicted mixing heights are used
XMXLEN 1 1 Maximum slug length
XSAMLEN 1 10 Maximum travel distance of a puff in grid unitsduring one sampling step
MXNEW 99 60 Maximum number of puffs released from onesource during one sampling step
MXSAM 99 60 Maximum number of sampling steps during onetime step for a puff
NCOUNT 2 2 Number of iterations used when computing thetransport wind for a sampling step that includestransitional plume rise
SYMIN 1 1 Minimum sigma Y in metres for a new puff
SZMIN 1 1 Minimum sigma Z in metres for a new puff
SVMIN 0.5,0.5,0.50.5,0.5,0.5
0.5,0.5,0.50.5,0.5,0.5
Default minimum turbulence velocities for eachstability class (Sigma-V)
SWMIN 0.2, 0.120.08, 0.06
0.03, 0.016
0.2, 0.12 0.08,0.06
0.03, 0.016
Default minimum turbulence velocities for eachstability class (Sigma-W)
WSCALM 0.5 0.5 Minimum wind speed allowed for non-calmconditions in m/s
XMAXZI 3000 3000 Maximum mixing height in meters
XMINZI 50 50 Minimum mixing height in meters
CDIV 0, 0 0, 0 Divergence criteria for dw/dz in meters
PLX0 0.07, 0.07,0.10, 0.15,0.35, 0.55
0.07, 0.07,0.10, 0.15,0.35, 0.55
Wind speed profile power-law exponents forstabilities 1 to 6
PTG0 0.02, 0.035 0.02, 0.035 Potential temperature gradient for stable classes
PPC 0.5, 0.5, 0.5,0.5, 0.35,0.35
0.5, 0.5, 0.5,0.5, 0.35, 0.35
Plume path coefficients for partial plume pathadjustment terrain method.
SL2PF 10 10 Slug to puff transition factor (Not used)
NSPLIT 3 3 Number of puffs that result everytime a puff is split(Not used)
IRESPLIT 0,0,0,0,0,0,0
0,0,0,0,0,0,0
0,0,0,1,0,0,0
0,0,0
0,0,0,0,0,0,0
0,0,0,0,0,0,0
0,0,0,1,0,0,0
0,0,0
Times of day when puff can be split after being splitpreviously (Not used)
ZISPLIT 100 100 Puff split only occurs if previous hours mixingheight exceeds this value (Not used)
ROLDMAX 0.25 0.25 Maximum allowable ratio previous hour mixingheight to maximum mixing height experience bypuff (Not used)
NSPLITH 5 5 Number of puffs that result from each split (notused)
SYSPLITH 1 1 Minimum sigma-y off puff before it may be split(Not used)
SHSPLITH 2 2 Minimum puff elongation rate due to wind shear,before it may be split (Not used)
CNSPLITH 1e-7
1e-7
Minimum concentration (g/m3) of each species inpuff before it may be split (Not used)
Proposed HR Milner Expansion Project Environmental Impact Assessment Report
Appendix 7C: CALPUFF and CALMET Methods and Assumptions
Maxim Power Corp. Page 7C-41 January 2009
Table 7C-14 CALPUFF Dispersion Model User Options (Continued…)Input Group Parameter USEPA
DefaultApplied Input Group
Group 12:MiscellaneousDispersion andComputationalParameters
EPSSLUG 1e-4
1e-4
Fraction convergence criterion for numerical slugsampling integration
EPSAREA 1e-6
1e-6
Fraction convergence criterion for numerical areasources integration
DSRISE 1 1 Trajectory step-length (m) used for numerical riseintegration
HTMINBC 500 500 Minimum height to mix boundary condition puffs(m)
RSAMPBC 10 15 Search radius (BC length segments) about areceptor for sampling nearest BC puff.
NDEPBC 1 0 Near surface depletion adjustment when samplingBC puffs
Group 13: PointSourceParameters
NPT1 - 70 Number of point sources modelled (ApplicationCase)
IPTU 1 1 Units used for emissions (g/s)
NSPT1 0 0 Number of source-species combinations withvariable emissions scaling factors
NPT2 - 0 Number of point sources with variable emissions
Group 14: AreaSourceParameters
NAR1 - 7 Number of polygon area sources modelled
IARU 1 1 Units used for emissions (g/m2/s)
NSAR1 0 0 Number of source-species combinations withvariable emissions scaling factors
NAR2 - 0 Number of area sources with variable emissions
Group 15: LineSourceParameters
NLN2 - 0 Number of buoyant line sources with variablelocation and emission parameters
NLINES - 0 Number of buoyant line sources
ILNU 1 1 Units for line source emission rates is g/s
NSLN1 0 0 Number of source-species combinations withvariable emission scaling factors
MXNSEG 7 7 Maximum number of segments used to model eachline
NLRISE 6 6 Number of distances at which transitional risecomputed
XL - 0 Average building length
HBL - 0 Average building height
WBL - 0 Average building width
WML - 0 Average line sources width
DXL - 0 Average separation between buildings
FPRIMEL - 0 Average buoyancy parameter
Group 16: VolumeSourceParameters
NVL1 - 0 Number of volume sources applied
IVLU 1 1 Units used for volume sources (g/s)
NSVL1 0 0 Number of source-species combinations withvariable emission scaling factors
NSVL2 - 0 Number of volume sources with variable locationand emission parameters
Group 17: Non-Gridded ReceptorInformation
NREC - 3558 Number of non-gridded discrete receptors thatcompose the series of nested grids, propertyboundary and sensitive receptors
Proposed HR Milner Expansion Project Environmental Impact Assessment Report
Appendix 7C: CALPUFF and CALMET Methods and Assumptions
Maxim Power Corp. Page 7C-42 January 2009
7C.2.4 NOx to NO2 Conversion
While the CALPUFF model can predict ambient NO and NO2 concentrations, the calculation has been
shown to overestimate ambient NO2 concentrations. For this assessment, the ozone limiting method
(OLM) was applied to account for this overestimation. The OLM assumes that the conversion of NO to
NO2 in the atmosphere is limited by the ambient ozone (O3) concentration in the atmosphere. The
approach assumes that 10% (on a volume basis) of the NO is converted to NO2 prior to discharge into the
atmosphere. For the remaining NO, the following is adopted:
If 0.9 (NO) is greater than the ambient O3 concentration then NO2 = 0.1 (NO) + 0.9 (O3). For thiscase, the conversion is not complete.
If 0.9 (NO) is less than the ambient O3 concentration then NO2 = 0.1 (NO) + 0.9 (NO) = NO. This isequivalent to the total conversion approach, since there is sufficient ozone to effect the completeconversion.
In the application of the OLM, the above relationships assume the concentrations are expressed on a ppb
basis.
Alberta Environment (2003a) recommends ambient ozone concentrations for 1-h, 24-h and annual
averaging periods (i.e., 50, 40 and 35 ppb for rural areas, and 50, 35 and 20 ppb for urban areas).
Alternately, hourly ambient ozone data can be used to calculate the NO to NO2 conversion on an hourly
basis. For consistency, the hourly ozone data should coincide with the meteorological data used in the
modelling. For the application of the OLM approach in this assessment, hourly ozone data from
Beaverlodge for 2007 were used to estimate hourly NO2 concentrations. The Beaverlodge ozone data are
discussed in Appendix 2B.
7C.2.5 Prediction Confidence
The evaluation of potential changes in air quality depends primarily upon air dispersion models that are
used to predict the change in expected ambient air concentrations. Air quality models, such as CALPUFF,
are as accurate as the inputs and assumptions employed in the model and the inputs.
Emission rates used in the modelling were estimated based on a combination of maximum permitted
emission limits, emission factors, and engineering estimates provided by MAXIM. In reality, actual
emissions vary from hour to hour and day to day. Due to the nature of this approach, there is a high
degree of confidence that estimated emissions over-estimate actual emissions.
Air quality dispersion models such as CALPUFF also employ assumptions to simplify the random
behaviour of the atmosphere into short periods of average behaviour. These assumptions limit the
capability of the model to replicate every individual meteorological event. To compensate for these
simplifications, one full year of meteorological data is applied to evaluate a wide range of possible
conditions. Additionally, regulatory models, such as CALPUFF, are designed to have a bias towards over
estimation of contaminant concentrations (i.e. to be conservative under most conditions).
To investigate relative model performance, ambient air quality data from the Milner monitoring station was
compared to model output for the monitor location for the Base Case (all existing and approved facilities).
Observed ground-level concentrations of NO2 and SO2 from the most recent available complete year of
monitoring (2007) were used as a basis of comparison against model output values. Although this
approach allowed for the maximum number of existing sources to contribute to the observed values,
some care must be taken in the interpretation. The Base Case modelling includes existing and
operational facilities operating at their approved rates, when in fact they may be operating at lower levels
of activity. Background sources (long-range transport) and biogenic (local) emissions were not
considered in the modelling.
Proposed HR Milner Expansion Project Environmental Impact Assessment Report
Appendix 7C: CALPUFF and CALMET Methods and Assumptions
Maxim Power Corp. Page 7C-43 January 2009
Table 7C-13 shows model-estimated and observed 1-hour ground-level nitrogen dioxide (NO2) and
sulphur dioxide (SO2) concentrations at the Milner station. For NO2 the estimated 99th, 98
th, and 90
th
percentile values are higher than the corresponding observed value for both average and peak emission
scenarios. The modelling of the highest hourly NO2 concentrations appears conservative as depicted at
the monitoring locations. A comparison of the remainder of the distribution (below the 90th
percentile)
shows a pattern of increasing conservatism – modelled concentrations are much higher than measured.
This may be because modelled NOx sources (haul trucks passing near the monitor) may not have been
as active in 2007 as depicted in the model. This entire NO2 distribution (predicted vs observed) is
presented in Figure 7C-12)
For SO2 the estimated 99th, 98
th, and 90
thpercentile values for the average emission scenario are lower
than the corresponding observed value. For the peak emission scenario the 99th
percentile value is
higher, while the 98th
and 90th
percentile values are lower than the corresponding observed value. The
modelling of the highest hourly SO2 concentrations based on the peak emission scenario appears
conservative as depicted at the monitoring locations. For the average emission scenario the predictions
of the highest hourly SO2 concentrations are within a factor of two of that measured – an indicator of good
model performance. A comparison of the remainder of the distribution (below the 90th
percentile) shows
that modelled concentrations are much lower than measured. This may be due to a known failure in the
SO2 monitor in 2007. An intermittent SO2 monitor failure resulted in occasional positive baseline drift, and
eventual failure of the instrument. This led to the invalidation of several months of data. This condition
may have biased values below the 90th
percentile upwards, resulting in higher than expected observed
values in that part of the distribution. The entire SO2 distribution (predicted vs observed) is presented in
Figure 7C-12).
Table 7C-15 Hourly Modelled and Observed NO2 and SO2 Concentrations
StatisticNitrogen Dioxide (µg/m
3) Sulphur Dioxide (µg/m
3)
Modelled1
Observed Modelled2
Observed
Count 8,759 8,185 8,760 8,190
Maximum 159.7 126.1 107.0 / 193.6 133.6
99th
Percentile 106.7 43.3 28.7 / 51.5 70.7
98th
Percentile 99.7 35.8 17.3 / 31.2 63.5
90th
Percentile 75.3 22.6 2.1 / 2.6 39.3
Note:1
The modelled concentrations are representative of both the average and peak emission conditions.2
Modelled concentrations representative of both the average and peak emission conditions are presented(average / peak).
The comparison of predicted vs observed hourly NO2 and SO2 concentrations presented is not intended
as a substitute for a rigorous model validation exercise. The summary statistics shown in the preceding
tables, and Figures 7C-12 and strongly suggest that the CALPUFF model has provided a reasonable
depiction of reality for the existing sources in proximity to the monitoring stations in this case.
Proposed HR Milner Expansion Project Environmental Impact Assessment Report
Appendix 7C: CALPUFF and CALMET Methods and Assumptions
Maxim Power Corp. Page 7C-44 January 2009
Figure 7C-12 Hourly Modelled and Observed NO2 and SO2 Concentrations
Proposed HR Milner Expansion Project Environmental Impact Assessment Report
Appendix 7C: CALPUFF and CALMET Methods and Assumptions
Maxim Power Corp. Page 7C-45 January 2009
7C.3 References
Allwine, K.J., and C.D. Whiteman. (1985). MELSAR: A mesoscale air quality model for complex terrain:
Volume 1 – Overview, technical description and user’s guide. Pacific Northwest Laboratory,
Richland, Washington.
Carson, D.J. (1973). The development of a dry, inversion-capped, convectively unstable boundary layer.
Quart. J. Roy. Meteor. Soc., 99: 450-467.
Douglas, S., and R. Kessler. (1988). User’s guide to the diagnostic wind model. California Air Resources
Board, Sacramento, CA.
Garratt, J.R. (1977). Review of drag coefficients over oceans and continents. Mon. Wea. Rev., 105: 915-
929.
Hanna, S.R., L.L. Schulman, R.J. Paine, J.E. Pleim, and M. Baer. (1985). Development and evaluation of
the Offshore and Coastal Dispersion Model. JAPCA, 35: 1039-1047.
Holtslag, A.A.M., and A.P. van Ulden. (1983): A simple scheme for daytime estimates of the surface
fluxes from routine weather data, J. Clim. and Appl. Meteor., 22: 517-529.
Liu, M. K. and M. A. Yocke. (1980). Siting of wind turbine generators in complex terrain. Journal of
Energy, 4: 10:16.
Maul, P.R. (1980). Atmospheric transport of sulfur compound pollutants. Central Electricity Generating
Bureau MID/SSD/80/0026/R. Nottingham, England.
O’Brien, J.J. (1970). A note on the vertical structure of eddy exchange coefficient in the planetary
boundary layer. J. Atmos. Sci., 27: 1213-1215.
Pasquill F. (1961). The estimation of the dispersion of wind-borne material. Meteorological Magazine,
90: 33-48.
Perry, S.G., D.J. Burns, L.H. Adams, R.J. Paine, M.G. Dennis, M.T. Mills, D.G. Strimaitis, R.J. Yamartino,
E.M. Insley. (1989). User’s Guide to the Complex Terrain Dispersion Model Plus Algorithms for
Unstable Situations (CTDMPLUS) Volume 1: Model Description and User Instructions.
EPA/600/8-89/041, U.S. Environmental Protection Agency, Research Triangle Park, NC.
Scire, J.S., F.R. Robe, M.E. Fernau, and R.J. Yamartino. (2000a). A User’s Guide for the CALMET
Meteorological Model (Version 5). Earth Tech, Inc., Concord, MA.
Scire, J.S., D.G. Strimaitis, and R.J. Yamartino. (2000b). A User’s Guide for the CALPUFF Dispersion
Model (Version 5). Earth Tech, Inc., Concord, MA.
Turner, D.B. 1963. A Diffusion Model for an Urban Area. Journal of Applied Meteorology. 3:83-91.
U.S. EPA. (1998a). United States Environmental Protection Agency. Interagency Workgroup on Air
Quality Modelling (IWAQM) Phase 1 Summary Report and Recommendations for Modelling Long
Range Transport Impacts. EPA-454/R-98-019.
Top Related