Appendix 7C CALPUFF and CALMET Methods and Assumptions · 2016-06-17 · Appendix 7C: CALPUFF and CALMET Methods and Assumptions Maxim Power Corp. Page 7C-3 January 2009 In the first

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




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




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.




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




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)




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


CALMET Land-use Categories Specified for the Modelled Domain

7C-4

Project Site

Map Features

Barren Land

Coniferous Forest

Deciduous Forest

Mixed Forest




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


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


MM5 Grid Point




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




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)




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


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


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


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


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


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


Wind Vector




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




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




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.




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




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


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




Table 7C-7 CALMET Model Options Group 2: Grid Control Parameters


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


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




Table 7C-8 CALMET Model Options Group 3: Output Options (cont’d)


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


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




Table 7C-10 CALMET Model Option Group 5: Wind Field Options andParameters


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




Table 7C-10 CALMET Model Option Group 5: Wind Field Options andParameters (cont’d)


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




Table 7C-10 CALMET Model Option Group 5: Wind Field Options andParameters (cont’d)


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


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




Table 7C-11 CALMET Model Option Group 6: Mixing Height, Temperature andPrecipitation Parameters (cont’d)


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)




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.




Table 7C-12: Summary of Major Features of CALPUFF




Table 7C-12: Summary of Major Features of CALPUFF (Continued…)




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).




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



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.




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




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.




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




Table 7C-14 CALPUFF Dispersion Model User Options (Continued…)Input Group Parameter USEPA


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






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

- - - - - -




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)





DefaultApplied Input Group


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)





DefaultApplied Input Group


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




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.




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.




Figure 7C-12 Hourly Modelled and Observed NO2 and SO2 Concentrations




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.

Documents

Appendix 7C CALPUFF and CALMET Methods and Assumptions · 2016-06-17 · Appendix 7C: CALPUFF and CALMET Methods and Assumptions Maxim Power Corp. Page 7C-3 January 2009 In the first