EPA - 908/ 5-80-003
DRAFT ENVIRONMENTAL IMPACT STATEMENT
IMPACT OF CANADTAN POWER PLANT
DEVELOPMENT AND FLOW APPORTIONMENT
ON THE POPLAR RIVER BASIN
EPA
Prepared by
U.S. Environmental Protection Agency
Montana Office
Region VIII
Federal Building
Helena, Montana 59601
ApDroved b’
er 1 iams
Date: eg al Administrator

-------
Lf J7 3
SUMMARY
Draft (X) Final ( ) Environmental Statement
Type of Action : Administrative (X) Legislative ( )
2. Brief Description of Action : A coal-fired power plant is currently being
constructed by Saskatchewan Power Corporation on the East Fork of the Poplar
River about 4 miles north of the International Boundary. Since the power plant
and associated cooling water reservoir will result in a consumptive water use
and will modify the natural transboundary flow of the Poplar River, it is antici—
pateci that a flow apportionment agreement between the U. S. and Canada will be
established. This Environmental Impact Stateement has been prepared to address
the impacts of several flow apportionment alternatives in addition to other
potential impacts of the power plant on the U. S. part of the Poplar River Basin.
3. Summary of Environmental Impacts : The Poplar River flows will be reduced
under conditions of the recommended apportionment up to 50 percent of the natural
flow in the West Fork and at least 60 percent of the natural flow on the Middle
Fork. Summer flows in the East Fork will be about 35 to 65 percent less with
one power plant operating and up to 80 percent less with four power plants oper-
ating. These reduced flows will result in less water being available for
future irrigation expansion. These lower flows may also impact furbearers,
waterfowl, fish and other organisms now using these riverene habitats. Water
quality will become lower than at present with total dissolved solids levels
increasing.
The highest concentrations of airborne pollutants will occur southeast
of the proposed power plant. The highest annual mean concentrations of S02,
NOx and particulates were 4.8, 1.6 and 0.4 ug/rn 3 , respectively, for a 1200 MWe
plant. The predicted concentrations were less than the U. S. National Ambient
Air Quality Standards and the Montana Ambient Air Quality Standards for 1-hour,
3—hour, 24—hours, and the annual mean. The maximum increases in $02 concentrations
predicted from 24 to 62 percent of maximum allowable increases for Class II areas.
The particulate loading in the impact area was only about 2 percent of maximum
allowable limits.
Fumigation under very stable conditions could result in S02 concentrations
between 2000 and 4000 ug/m3 at a distance of 10 to 20 km downwind of a 1200 MWe
plant. If a concentration of 4000 ug/m 3 persisted for 30 minutes, the allowable
3 hour increase of 512 ug/rn 3 for Class II areas would be exceeded. Under more
typical conditions the $02 concentrations were estimated to be less than 1850 ug/m 3 .
The impact of fumigation events would be minimal because of the small area
affected by one event and the short time period (less than 45 minutes). Moreover,
the frequency of occurrence is estimated as 11.1 and 35-45 percent for the very
stable and typical atmospheric conditions, respectively.
The predicted S02 concentrations with two 300 MWe units and no S02 control
would exceed the Class I PSD regulations at the proposed Class I area of
the Fort Peck Indian Reservation but not at the Medicine Lake National Wildlife

-------
-2-
Refuge. The regulations could be met at the Fort Peck Indian Reservation
with 60 percent control for two 300 MWe units and 90 percent control for four
300 MWe units.
Comparison of the predicted concentrations of SO 2 , NO and particulates
with acute and chronic threshold limits for selected plant species indicated no
detectable impacts on the terrestrial vegetation. No impacts were predicted
due to accumulation of trace metals in soils. Thus, no impacts were predicted
on any rare and endangered species.
Expenditures of Canadian workers in Daniels County have resulted in
direct and secondary impacts of $216,000 to $324,000, representing 6.3 to
9.5 percent of total retail sales. Secondary impacts of $53,000 to $79,000
were projected for Roosevelt County.
Analysis of the secondary effects of flow apportionment resulting in
reduced development of irrigated land and poorer quality irrigation water
indicates that in both Daniels and Roosevelt counties the losses of personal
income (due to changes in farm income and secondary impacts) amount to
about 1 percent in the year 2000. In 1985, the projected personal income
changes are -0.7 percent and +0.3 percent for Daniels and Roosevelt counties,
respectively. The 1985 apportionment levels result in maximum reductions
of farm proprietors’ income of 1.5 percent and less than 3.1 percent for
Daniels and Roosevelt counties, respectively. These respective impacts
increase to 2.2 percent and 6 percent at the year 2000 level of development.
4. Alternatives Considered :
A. Air Quality - various levels of control for S02 (flue gas desulfuri-
zation) and particulates are evaluated in Chapter 3.
B. Various flow apportionment regimes have been evaluated.
An apportionment of waters on a 50% Canada - 50% U.S. basis may
be approved.
5. Con nents Requested From the Following :
see attachment
*Note - Only those who supply substantive comments on the draft EIS will
be sent copies of the final EIS.
6. Date Draft EIS Made Available to EPA and the Public :
AUG 15 1980

-------
Comments on the Draft Environmental Impact Statement will be requested from the
following:
International
International Joint Cornniision
International Poplar River Water Quality Board
Canada
Province of Saskatchewan
Saskatchewan Power Corporation
Congressional
U. S. Senator John Meicher
U. S. Senator Max Baucus
U. S. Representative Ron Marlene
Montana
Governor of Montana
Lt. Governor of Montana
Department of Agriculture
Cooperative Extension Service
Department of Community Affairs
Department of Health and Environmental Sciences
Department of Fish, Game and Parks
Department of Natural Resources and Conservation
Environmental Quality Council
Historical Society
Old West Regional Commission
Federal
U.S. Department of State
Advisory Council on Historic Preservation
Army Corps of Engineers
Department of Agriculture
Soil Conservation Service
Department of Comerce
Department of the Interior
Bureau of Indian Affairs
Bureau of Land Management
Bureau of Mines
Bureau of Outdoor Recreation
Bureau of Reclamation
Fish and Wildlife Service
Geological Survey
Heritage Conservation and Recreation Service

-------
—2—
Federal (Continued)
Missouri River Basin Coninission
National Park Service
Department of Housing and Urban Development
Department of Health Education and Welfare
Local Government
Sheridan County Comissioners
Daniels County Comission
Roosevelt County Commission
Town of Scobey
Town of Plentywood
Town of Wolf Point
Town of Poplar
Other Organizations and Individuals
Fort Peck Indian Reservation
Northern Plains Resource Council
Three Corners Boundary Association
Environmental Information Center
Montana Wildlife Federation
National Wildlife Federation
Friends of the Earth
Trout Unlimited
Ducks Unlimited
Daniels County WIFE - Zelpha Danielson - Scobey
Dale Chabot -Scobey
Lowell Burgett - Scobey
Jim Sims - Soil Science Lab, MSU, Bozeman
Delmer Safty - Whitetail
Dr. Merele D. Fitz - Scobey
Tittinger Brothers - Scobey
Environmental Studies Lab., U. of M. - Missoula
Glasgow Public Library
Plentywood Public Library
Scobey Public Library
Wolf Point Public Library
Hugh Baker — Whitetail
Dr. Robert R. Bell, DVM, Culbertsori
Charles Cassidy - Scobey
Bill Cromwell - Flaxville
Ken Lee - DVM, Scobey
Farver Brothers - Scobey
Lyle Hang - Scobey
Hellickson Brothers, Inc. - Scobey
Gordon Holte - Plentywood

-------
-3—
Other Organizations and Individuals (Continued)
Lee Hurnbert - Scobey
Art Lund - Scobey
Dennis Nathe - Redstone
Lee Rovig - Outlook
Robert Schneekloth - Redstone
Ted Skornogoski - Scobey
Ron Stoneberg — Circle
Bill Tande - Scobey
Boyd Tymofichuk - Scobey
Eddie Lund - Scobey
Daniels County Leader - Scobey
Plentywood Herald
Glasgow Courier
Wotanian - Poplar
Wolf Point Herald

-------
DRAFT ENVIRONMENTAL IMPACT STATEMENT
IMPACT OF CANADIAN POWER PLANT DEVELOPMENT
AND FLOW APPORTIONMENT ON THE
POPLAR RIVER BASIN
U.S. Environmental Protection Agency
Montana Office Region 8
Helena, Montana
Prepared with assistance of Tetra Tech staff
Tetra Tech, Incorporated
3746 Mt. Diablo Boulevard, Suite 300
Lafayette, California 94549
(415) 283-3771
EPA Contract No. 68-O1- 873
Tetra Tech Report No. TC-3254
EPA 908/80-003
JULY 1980

-------
TABLE OF CONTENTS
1. SUM 1ARY
1.1 WATER QUANTITY IMPACTS
1.2 WATER QUALITY IMPACTS
1.3 AIR QUALITY IMPACTS
1.4 SOCIOECONOMIC IMPACTS
2. PURPOSE AND NEED
3. ALTERNATIVES INCLUDING THE PROPOSED ACTION
3.1 ATMOSPHERIC EMISSIONS AND CONTROL
3.2 FLOII RELATED ALTERNATI’JES
3.2.1 Alternative Flow Apportionments
3.2.3 Demand Releases
3.3 WATER QUALITY
3.3.1 Mitigation of Water Quality Impacts
3.2.2 No Action Case
Pa ge
1
1
2
3
4
5
9
9
13
13
15
16
17
17
• . . . 22
22
• . . . 25
• . . . 28
• • . . 28
30
30
33
33
37
42
42
49
53
• . • . 53
53
53
53
ndian
54
57
57
59
60
60
62
62
64
64
4. AFFECTED ENVIRONMEN-T
4.1 LOCATION
4.2 GEOLOGY AND SOILS
4.3 LAND USE
4.3.1 General
4.3.2 Land Use Surveys
4.3.3 Agricultural Activities .
4.4 HYDROLOGY
4.4.1 Surface Water
4.4.2 Ground Water
4.5 WATER QUALITY
4.5.1 Surface Water
4.5.2 Ground Water Quality
4.6 WATER USE
4.6.1 Municipal Use
4.6.2 Industrial Use
4.6.3 Agricultural Use
4.6.3.1 Montana Water Use
4.6.3.2 Water Use on the Fort
Reservation
4.6.3.3 Canadian Water Use
4.7 VEGETATION AND WILDLIFE
4.8 AQUATIC BIOTA AND FISHERIES
4.9 METEOROLOGY AND AIR QUALITY
4.9.1 Meteorology
4.9.2 Existing Air Quality
4.10 SOCIAL AND ECONOMIC PROFILES . . .
4.10.1 Population Profile
4.10.2 Archaeological and Historical
4.10.3 Economic Profile
1
Peck I
Sites

-------
TABLE OF CONTENTS (continued)
5.2 WATER
5.2.1
5.3 WATER
5.3.1
5.3.2
Page
66
66
66
66
67
67
67
72
75
75
75
79
80
80
89
90
91
91
91
91
95
108
110
• • 117
• . . 117
121
122
• . 125
• . 126
126
• . • 127
• . 129
• . 134
134
ts
136
141
142
142
143
5. ENVIRONMENTAL CONSEQUENCES
5.1 AIR QUALITY IMPACTS
5.1.1 Air Quality Model
5.1.2 Power Plant Emissions
5.1.3 Model Input Parameters
5.1.4 Modeling Results
5.1.4.1 Sulfur Dioxide (SO2)
5.1.4.2 Oxides of Nitroaen (NOx)
5.1.4.3 Particulates
5.1.5 Impact Assessment
5.1.5.1 Sulfur Dioxide Impact
5.1.5.2 NO Impact
5.1.5.3 Particulate Impact
5.1.5.4 Fumigation Impact
5.1.5.5 Visibility Impacts...
5.1.5.6 Health Effects
OUALITY IMPACTS
Methodology
5.2.1.1 Flow Scenarios
5.2.1.2 Model Description
5.2.1.3 Water Uses
5.2.2 Predicted Flows
5.2.2.1 Flow Conditions at the 90
Level
5.2.2.2 Flow Conditions at the 10
Level
5.2.3 Direct Impacts
5.2.3.1 Municipal Water SuDply
5.2.3.2 Uses Dependent on Spring Runoff
5.2.3.3 Summer Flows
5.2.3.4 Winter Flows
QUALITY IMPACTS
Description of Quality Models
Boron
5.3.2.1 Boron Impacts on Crops
5.3.2.2 Other Boron Impacts
5.3.3 Salinity and Sodicity
5.3.3.1 Salinity and Sodicity Impacts on
Crops
5.3.3.2 Impact on Crops of Combined Effec
of Salinity, Sodicity, and Boron
5.3.3.3 Other Salinity Impacts
5.3.4 Sulfate (SO 4 )
5.3.5 Mitigative Measures to Reduce Impacts of
Saline Irrigation Waters
Percent
108
Percent
ii

-------
TABLE OF CONTENTS (continued)
5.3.5.1 Mitigative Practices for Salinity
Control in Soils
5.3.5.2 Mitigative Irrigation Practices
for Salinity Control in Return
Flows
5.3.5.3 Source Control of Salinity .
5.4 SOCIOECONOMIC IMPACTS OF POWER PLANT CONSTRUCTION
5.4.1 Introduction
5.4.2 Description of the Construction Work Force
5.4.3
5.4.4
Economic Impacts from the Plant
Economic Impacts fran the Construction
Workers
5.4.5 Secondary Impacts
5.5 SOCIOECONOMIC IMPACTS OF APPORTIONMENT
5.5.1 Introduction
5.5.2 Future Conditions
5.5.3 Impacts on Income
5.5.4 Other Impacts
5.6 BIOLOGICAL IMPACTS
5.6.1 Impacts of Atmospheric Emissions on
Terrestrial Biota
5.6.1.1 Gaseous Emissions
5.6.1.2 Particulate Emissions
5.6.2 Impacts of Atmospheric Emissions on Aquatic
Biota
5.6.2.1 Acidification and Nitrogen Loading
5.6.2.2 Trace Element Contamination
5.6.3 Impacts of Water Quality Changes on Fish
and tlildlife
5.6.3.1 Effects of Thermal Discharges
5.6.3.2 Effects of Dissolved Solids
Increases
5.6.3.3 Effects of Dissolved Oxygen
Changes
Other Constituents
of Flow Modifications on Fish and
5.6.3.4
Impacts
Wildlife
5.6.4.1 Wildlife and Furbearers . .
5.6.&2 Fish
6. REFERENCES . . . . . . 211
7. LIST OF PREPARERS
226
Page
143
146
149
151
151
151
153
153
155
158
158
158
160
165
167
167
• • 167
• . 175
177
177
180
181
184
188
189
• . 191
191
• . 191
192
5.6.4
ii-’

-------
LIST OF FIGURES
Figure
4.1-1 Location of the Poplar River Basin
4.1—2 Location of Poplar River Power Plant Site
4.2-1 Quaternary Geology of the U.S. Part of the
Poplar River Basin
4.3-1 Land Use in Area I Surveyed by Duggan
4.3-2 Land Use in Area 2 Surveyed by Duggan
4.4-1 Major Sub-Basins of the Poplar River System
4.4-2 Outflow Hydrograph for the Poplar River Near
Poplar, Montana, October, 1975, to September,
1976
4.4-3 Schematic of Typical Reach of the Poplar River
4.4-4 Ground Water Flow Regime in Canadian Part of
Poplar River Basin
4.4-5 General Ground Water Flow Regime in U.S. Part
of Poplar River Basin
4.5-1 Location of Water Quality Sampling Stations
4.5-2 Ranges of Selected Chemical Parameters in Water
Samples From the Fort Union Formation and Fox
Hills-Hells Creek Formation in U.S. Part of the
Poplar River Basin
4.5-3 Ranges of Selected Chemical Parameters in Water
Samples From Quaternary Alluvium, Glacial Outwash
and Flaxville Formation in the U.S. Part of the
the Poplar River Basin
4.6—1 Historical Water Use in the U.S. Part of the
Poplar River Basin
4.6-2 Historical Water Use on the Fort Peck Indian
Reservation
4.9-1 Normal Monthly Precipitation at Scobey, Montana
4.10-1 Location of Historic Sites in the Poplar River
Basin and Adjacent Areas
5.1-i Spatial Distribution of the Highest 1-Hour SO 2
Concentrations (pg/rn 3 ) Obtained From the CRSTER
Model for 1964, Assuming a 600 MW (1200 MW)
Poplar River Power Plant with Zero Percent
Emission Control
5.1-2 Spatial Distribution of the Highest 3-Hour 502
Concentrations (pg/m 3 ) Obtained From the CRSTER
Model for 1964, Assuming a 600 MW (1200 NW)
Poplar River Power Plant with Zero Percent
Emission Control
5.1-3 Spatial Distribution of the Highest 24-Hour SO 2
Concentrations ( ig/m 3 ) Obtained From the CRSTER
Model for 1964, Assuming a 600 MW (1200 MW)
Poplar River Power Plant with Zero Percent
Emission Control .
Pa g
23
24
26
31
32
36
39
40
41
43
44
50
51
55
56
6 1
65
68
69
70
iv

-------
LIST OF FIGURES (continued)
Fiqure Page
5.1-4 Spatial Distribution of the 1964 Annual S02
Concentrations (ug/m 3 ) Obtained From the CRSTER
Model , Assuming a 600 MW (1200 MW) Poplar River
Power Plant with Zero Percent Emission Control
5.1-5 Spatial Distribution of the Highest 1-Hour NO
Concentrations (ug/m 3 ) Obtained From the CRSTER
Model for 1964, Assuming a 600 MW (1200 MW)
Poplar River Power Plant
5.1-6 Spatial Distribution of the 1964 Annual NO
Concentrations (3Jq/m 3 ) Obtained From the CRSTER
Model , Assuming a 600 MW (1200 MW) Poplar River
Power Plant
5.1-7 Spatial Distribution of the Highest 2d-Hour
Particulate Concentrations ( ig/m 3 ) Obtained
From the CRSTER Model for 1964, Assuming a
600 MW (1200 MW) Poplar River Power Plant
with 99.5 Percent Emission Control
5.1-3 Spatial Distribution of the 1964 Annual Particulate
Concentrations (iig/m 3 ) Obtained From the CRSTER
Model, Assuming a 600 MW (1200 MW) Poplar River
Power Plant with 99.5 Percent Emission Control
5.2-1 Location of Stations with Flow Results
5.2-2 Canadian Water Uses on the East Fork
5.2-3 Canadian Water Uses on the Middle Fork
5.2-4 Canadian Water Uses on the West Fork
5.2-5 U.S. Water Uses on the East Fork of the Poplar
River
5.2-6 U.S. Water Uses on the Middle Fork of the Poplar
River Above the Confluence with East Fork .
5.2-7 U.S. Water Uses on West Fork of the Poplar River
(Includes Indian and non-Indian Uses)
5.2-8 U.S. Water Uses on Main Stem of Poplar River
Above Fork Peck Indian Reservation
5.2-9 U.S. Water Uses on Main Stem of Poplar River
Within Fork Peck Indian Reservation
5.2-10 Flows at East Fork of Poplar River at International
Border
5.2-11 Flows at East Fork of Poplar River at Scobey
5.2-12 Flows at Middle Fork of Poplar River at
International Border
5.2-13 Flows of Main Poplar River at Fort Peck Indian
Reservation
5.2-14 Flows of West Poplar at the International Border
5.2-15 Flows of Poplar River at Poplar
5.2-16 Flows at East Fork at International Border .
5.2-17 March Flows on the East Fork 1933-1974
5.2-13 April Flows on the East Fork 1933-1974
• . 71
73
74
• 76
• 77
92
96
• 97
• 98
101
102
103
194
105
109
111
• . 112
113
114
• . 115
116
118
119
V

-------
LIST OF FIGURES (continued)
Fiqure Page
5.3-1 Boron Concentrations in July on East Fork and Main
Stem of Poplar River for Scenarios 28, 29, 31,
and 32 128
5.3-2 Boron Concentrations in July on East Fork and Main
Stem of Poplar River for Scenarios 4A and 8A . . . 130
5.a-1 Estimated Construction Work Force Profile,
Saskatchewan Power Plant Unit 1, 1975 through
1980 152
5.6-1 Dose-Injury Curves for (a) S0 2 -Sensitive Plant
Species, (b) Plant Species of Intermediate SO 2
Sensitivity, and (c) S02-Resistant Plant Species . . . 170
5.6-2 River Heat Release Summary 185
5.6-3 Sediment Flux Versus Flow at the International
Boundary East Fork Poplar River 198
5.6-4 Relationship Between Flow and Year-Class Formation
of Game Fish in the East and Middle Forks of the
Poplar River 205
vi

-------
LIST OF TABLES
Table Page
3.1-1 Maximum S02 Concentrations in the U.S. Expressed as
a Percentage of the Maximum Allowable Increase
(PSD) in Class II Areas 10
3.1-2 A Comparison of Class I P 50 Values and Maximum
Predicted S02 Concentrations (pg/rn 3 ) 30 Miles
Southeast of Plant 12
3.2-1 Description of Apportionment Alternatives 14
3.3-1 Projected Boron Concentrations (mg/i) at Station 1 . . 18
3.3-2 Projected Sodium Absorption Ratios at Station 1 . . . 19
3.3-3 Projected TDS Concentrations (mg/i) at Station 1 . . . 20
4.3-1 Land Use Characteristics of the U.S. Poplar River
Drainage in 1967 (Daniels, Roosevelt and Sheridan
Counties) Expressed in Acres and as the Percent
of the Total Area 29
4.3-2 Livestock Inventory for Daniels County 34
4.3-3 Acres in Irrigated and Non-Irrigated Crops in
Daniels County, 1975 34
4.3-4 Crop Production, Roosevelt County, 1975 35
4.3-5 Livestock Inventory, Roosevelt County 35
4.4-1 Perennial Stream Length and Drainage Basin Areas
for the Poplar River Basin 38
4.4-2 Comparison of Expected Annual Flows, Mean Flows
and the 1975 Flows in the Poplar River 38
4.5-1 Water Quality Statistics for Stream Sampling
in Locations in Canada 45
4.5-2 U.S. EPA Water Quality Criteria Contraventions on
the Poplar River, 1975 48
4.6-1 Estimates of Existing Water Use for Gravity/Pump
Irrigation in the U.S. Poplar River Basin 58
4.10-1 Population in Daniels and Roosevelt Counties,
1970-1975, and Percentage Chanqe, 1970-1975 63
5.1-1 Projected Maxima Pollutant Concentrations (pg/rn 3 )
in United States Impact Area Compared with
Applicable Standards 78
5.1-2 Calculated Increases in Air Concentrations of Trace
Elements Near the Poplar River Plant 81
5.1-3 1975 Background Trace Element Concentrations
(pg/rn 3 ) Measured Near Glasgow, Montana 82
5.1-4 Estimates of Maximum Ground-Level Concentrations
(pg/rn 3 ) Durinci Morning Fumigation 84
5.1-5 Estimates of Maximum Ground-Level SO 2 Concentrations
(pg/rn 3 ) During Morning Fumigation Resulting From
Typical Meteorological Conditions at Scobey,
Montana 85
5.1-6 Estimates of l 1aximum Ground-Level NOx Concentrations
(pg/rn 3 ) During Morning Fumigation Resulting From
Typical Meteorological Conditions at Scobey,
Montana 86
vii

-------
LIST OF TABLES (continued)
Table Page
5.1-7 Estimates of Maximum Ground-Level Total Suspended
Particulate (TSP) Concentrations (iig/m 3 ) During
Morning Fumigation Resulting From Typical
Meteorological Conditions at Scobey, Montana 87
5.1-8 Average Meteorological Conditions (100-200 M Layer)
and Plume Heights During the Morning for the
Mid-Seasonal Months at Scobey, Montana 88
5.1-9 Expected Health Effects of Air Pollution on Selected
Population Groups
5.2-1 Reservoir Releases on the East Fork on the Poplar
River
5.2-2 Summary of Flow Scenarios 94
5.3—1 Irrigation Requirements and Dilution Factors for
Alfalfa and Small Grains 133
5.3—2 Salinity, SAR, and 504 Concentrations at Selected
Stations 135
5.3-3 Available Studies on Salinity and Sodicity Hazards . . 137
5.3-4 Average Chemical Data for Upper Basin Soils and
Soils Within Ft. Peck 140
5.3-5 Relative Tolerances of Various Crops and Forage
Species to Salinity Arranged According to
Decreasing Tolerances Within Groups; Modified
From Allison (1964) and From Hem (1970) 145
5.4-1 Total Personal Farm and Nonfarm Income, Daniels
and Roosevelt Counties 1972-1977 154
5.4-2 Total Retail Sales Daniels and Roosevelt Counties,
1973-1977 156
5.5-1 Projected Population and Employment, Daniels and
Roosevelt Counties -- 1980, 1985 and 2000 159
5.5-2 Change in Farm Income Resulting From Apportionment
Only 163
5.5-3 Change in Yield and Per-Acre Revenues for Wheat . . . 162
5 .5-4 Impacts of Apportionment on Personal Income in
Daniels and Roosevelt Counties 164
5.5-5 Maximum Possible Impacts on Farm Income 165
5.6-1 Reported Threshold Limits of Important Native Plant
and Cultivated Species Found in the Impact Area to
Gaseous (SO 2 , Ox and SO 2 ÷ NOx) Emission Exposures. . 168
5.6-2 Projected Deposition Rates, Soil Concentrations and
Plant Accumulation of 17 Trace Elements Resulting
From Particulate Emissions 178
5.6-3 Minimum and Maximum Concentration of Trace Elements
in Poplar River Basin Vegetation (‘jg/g) Samples
Collected During the Late Summer of 1977 179
5.6-4 Trace Element Concentrations (PPM) in Poplar River
Coal Ash Samples 182
5.6-5 Average Trace Element Concentrations ( ig/z) in the
Poplar River and Projected Increases due to
Atmospheric Emissions of the Poplar River Power
Plant 183
viii

-------
LIST OF TABLES (continued)
Table Page
5.6-6 Bed Material Size Distribution 195
5.6-7 Comparison of 1977 and 1978 Spring Flows (cfs) at
Selected Poplar River Stations 202
5.6-8 Predicted Average April Flows (cfs) in the East
Fork Poplar River 203
5.6-9 Predicted Impact of Flow Apportionment on Young-of-
the-Year Class Strength of Poplar River Game Fish . . 206
5.6-10 Recommended Instream Flows for the East Fork of
the Poplar River 209
lx

-------
1. SUMMARY
A coal-fired power plant is currently being constructed by Saskatche-
wan Power Corporation on the East Fork of the Poplar River about 4 miles
north of the International Boundary. Since the power plant and associated
cooling water reservoir will result in a consumptive water use and will
modify the natural transboundary flow of the Poplar River, it is antici-
pated that a flow apportionment agreement between the U.S. and Canada will
be established. This Environmental Impact Statement has been prepared to
address the impacts of several flow apportionment alternatives in addition
to other potential impacts of the power plant on the U.S. part of the
Poplar River Basin.
1.1 hATER QUANTITY IMPACTS
The Poplar River is a meandering prairie stream with a mean annual
flow near Poplar, Montana of 83,860 ac-ft (115.8 cfs). About 32 percent
of that flow comes from Canada. The flow has a wide seasonal and yearly
variability. For example, the annual flow on the East Fork at the border
has rancied between 0 and 32,000 ac-ft. The winter and summer are low flow
periods with about three-fourths of the annual flow coming from snowmelt
in March or April. Surface water uses in the Poplar Basin include munici-
pal uses, stock reservoirs, and irrigation.
The flows will be reduced under conditions of the recommended appor-
tionment up to 50 percent of the natural flow on the West Fork and at
least 60 percent of the natural flow on the Middle Fork. Flows on the
East Fork will be made up of a continuous release from the Cookson Reser-
voir of 1 to 3 cfs. The summer flows on the East Fork will be about 35 to
65 percent less with one power plant operating and up to 80 percent less
with four power plants operating. Alternative flow apportionments were
considered but the 50/50 division of flow between Canada and the U.S. with
the specifications above is the best from a water quantity viewpoint.
The impacts of the recommended apportionment, construction of the
Cookson Reservoir, and operation of one to four 300 MWe power plants are
summarized for municipal, stock and irrigation uses in this section.
Municipal uses at Scobey on the Lower East Fork can be met with 1975 de-
mands of 350 ac-ft and 1985 demands of 400 ac-ft. The full demands in-
cludinq outdoor uses in August and September in the year 2000 cannot be
met although the basic domestic uses of 30 ac-ft can be met. Spreader
irrigation demands in March cannot be met from the Poplar River in one
year out of ten. Stock reservoirs would be filled from the smaller
tributaries during the spring runoff. Sprinkler and flood irrigation
demands in June in 1975 of 4,841 acres and in 1985 of 16,288 acres can
be met 50 percent of the time at all the stations. Flows are not ade-
quate to meet demands ten percent of the time. In 2000, the full irri-
gation demands of 29,555 acres cannot be met, since only about 4,186
acres can be irrigated near Poplar instead of the projected acreage of
1

-------
20,618. The large increases in irrigated acreage from 1975 to 1985 and
2000 are mostly due to the projected increases on the Fort Peck Indian
Reservation after construction of two reservoirs. These reservoirs sub-
stantially increase the water available for irrigation on the Fort Peck
Indian Reservation even with the apportionment. Erregardless, the full
29,555 acres cannot be irrigated.
Changes in the flow of the Poplar River due to apportionment may im-
pact organisms (e.o., furbearers, waterfowl and fish) requiring specific
aquatic habitats in the basin. The highest potential for impact will
occur in the East Fork. The greatest adverse impact will probably be a
long-term change in channel morphology resulting in a loss of spawning
and overwintering habitat for game fish. Prior to the long-term impacts,
however, successful game fish reproduction will most likely occur in the
East Fork under mean flows with up to two units in operation. The contin-
uous low-level release during the winter will increase winter survival of
fish under the Ice cover. Predicted levels of macrophyte encroachment in
the upper reaches of the East Fork will result in decreased duck produc-
ti on.
1.2 WATER OUALITY IMPACTS
Water quality in the Poplar River was simulated using Karp III and
the Modified Montreal Engineering (MME) model of the Cockson Reservoir.
The quality of transboundary flow was calculated on the basis of upstream
water quality plus ground water seepage, irrigation return flow and run-
off as well as simulated reservoir water quality.
For the purposes of evaluating the impacts of the proposed action on
the quality of irrigation and potable water supplies, scenarios 28 through
32, which represent projected 1975, 1985, and 2000 conditions with recom-
mended flow apportionments but without ash lagoon decant were compared
with historical conditions as well as 1975 conditions with the Cookson
Reservoir in place. These impacts were compared with conditions under
scenarios 4.A and 8A which had all the ash lagoon decant added to Cookson
Reservoir. The impacts of alternative apportionment scenarios representing
increased Canadian water usage were also evaluated.
The impoundment of water in Cookson Reservoir has resulted in improved
water quality on the East Fork at the International Border in all seasons
except spring. Operation of a single 300 MWe unit will not lead to signif-
icant degradation of water quality for irrigation and other purposes.
Addition of more units would result in increased concentrations of lOS and
boron and higher sodium adsorption ratio (SAR) values which would cause
decreased crop yields. The severity of the impacts on crops will depend
partly on how the ash lagoons at the power generating site are operated.
Flow on the East Fork would be made up of a small cont nuous release
supplemented by larger demand releases from Cockson Reservoir. The amount
of the releases is based on the March to May flow of the Middle Fork at
2

-------
the International Boundary. The flow on the East Fork is the same
under the alternative apportionments so the water quality on the East
Fork would not improve if an alternative apportionment is implemented.
The Poplar River supports populations of game fish species -
walleye and northern pike. Operation of two or more units will result
in TDS concentrations in April exceeding tolerance limits for successful
game fish reproduction in the East Fork one out of every ten years.
With four units, the TDS concentrations in the East Fork would also
limit successful spawning in all years. No adverse effects of increased
TDS levels on fish are projected for other parts of the basin. Changes
in other constituents are also not expected to result in adverse effects
on game fish populations.
1.3 AIR QUALITY IMPACTS
The highest concentrations of airborne pollutants will occur south-
east of the proposed power plant based on the results from the CRSTER
model . The highest annual mean concentrations of S02, 1 Ox and particu—
lates were 4.8, 1.6, and 0.4 g/m 3 , respectively, for a 1200 MPe plant.
The predicted concentrations were less than the U.S. National P mbient
Air Quality Standards and the Montana Ambient Air Quality Standards for
1-hour, 3-hour, 24-hours, and the annual mean. The maximum increases
02 concentrations predicted from 24 to 62 percent of maximum allow-
able increases for Class II areas. The particulate loading in the im-
pact area was only about 2 percent of maximum allowable limits.
Fumigation under very stable conditions could result in SO 2 concen-
trations between 2000 and 4000 pg/m 3 at a distance of 10 to 20 km down-
wind of a 1200 MWe plant. If a concentration of 4000 ig/m persisted for
30 minutes, the allowable 3 hour increase of 512 pg/rn 3 for Class II areas
would be exceeded. Under more typical conditions the SO 2 concentrations
were estimated to be less than 1850 pg/rn 3 . The impact of fumigation
events would be minimal because of the small area affected by one event
and the short time period (less than 45 minutes). Moreover, the frequency
of occurrence is estimated as 11.1 and 35-45 percent for the very stable
and typical atmospheric conditions, respectively.
The predicted SO 2 concentrations with two 300 MWe units and no SO 2
control would exceed the Class I PSO regulations at the proposed Class J
area of the Fort Peck Indian Reservation but not at the Medicine Lake
National Wildlife Refuge. The regulations could be met at the Fork Peck
Indian Reservation with 60 percent control for two 300 MWe units and 90
percent control for four 300 1We units.
Conoarison of the predicted concentrations of SO 2 , NOx and particu-
lates with acute and chronic threshold limits for selected plant species
indicated no detectable impacts on the terrestrial vegetation. No im-
pacts were predicted due to accumulation of trace metals in soils. Thus,
no impacts were predicted on any rare and endangered species.
3

-------
1.4 SOCIOECONOMIC IMPACTS
Both Roosevelt and Daniels counties have low population densities
and stable populations with only small increases projected to the year
2000. The area’s economy is dominated by agriculture, with spring wheat
being the most important crop. There are currently no major industries
in the area; however, the development of a potash mine near Scobey is
projected for the year 2000.
Expenditures of Canadian workers in Daniels County have resulted
in direct and secondary impacts of 5216,000 to $324,000, representing
6.3 to 9.5 percent of total retail sales. Secondary impacts of 553,000
to 579,000 were projected for Roosevelt County. The increased expendi-
tures may have prevented Daniels County from experiencing the decline
in retail sales that occurred in Roosevelt County between 1974 and 1975.
Analysis of the secondary effects of flow apportionment resulting
in reduced development of irrigated land and poorer quality irrigation
water indicates that in both Daniels and Roosevelt counties the losses
of personal income (due to changes in farm income and secondary impacts)
amount to about 1 percent in the year 2000. In 1985, the projected per-
sonal income changes are -0.7 percent and +0.3 percent for Daniels and
Roosevelt counties, respectively. Such changes are far less than his-
torical variations resulting from annual weather variations. The 1985
apportionment levels result in maximum reductions of farm proprietors’
income of 1.5 percent and less than 3.1 percent for Daniels and Roosevelt
counties, respectively. These respective impacts increase to 2.2 percent
and 6 percent at the year 2000 level of development.
4

-------
2. PURPOSE AND NEED
In 1972, Saskatchewan Power Corporation (SPC) submitted an appli-
cation to the Saskatchewan Government for water rights on the Poplar
River, a basin which crosses the International Boundary in Eastern
Montana. Following the issuance of water rights in 1974, SPC formally
proposed the construction of a steam electric generating station on the
East Fork of the Poplar River, about four miles north of the Interna-
tional Boundary. The generating complex was also designed to include a
reservoir on the East Fork to be used as a once-through cooling water
source and a lignite coal mine near the site. On 29 April 1975, SPC
was authorized by the Canadian Government to construct the power plant
works, including darn and reservoir, subject to compliance with future
water apportionment agreements between the U.S. and Canada.
The Poplar River Basin falls within the geographical boundaries
(Eastern Montana and North Dakota) of the Souris Red River Basin and
under the jurisdiction of the International Joint Commission (IJC). The
IJC was established in 1948 by the governments of the U.S. anc Canada
to investigate and report on water requirements arising from projects
on river basins comon to both countries.
In responding to the planned Canadian development on the East Fork
of the Poplar River, the IJC requested, on 8 April 1975, its Interna-
tional Souris-Red Rivers Engineering Board to conduct a study of flow
apportionment in the Poplar River Basin. The Board then appointed the
Poplar River Task Force to carry out the apportionment studies.
The Task Forces’ recommendatio is for flow apportionment in the
Poplar River Basin were endorsed by the IJC and were published in a
1978 report entitled, “Water Appertionment in the Poplar River Basin.”
The recommended apportionment essentially involves an equal division
of transboundary flows between the U.S. and Canada. The following is
a direct transcript of the Board’s recommended apportionment which will
subsequently be referred to as Apportionment VI in this document.
The aggregate natural flow of all streams and tributaries
in the Poplar River Basin crossing the International Boundary
shall be divided equally between Canada and the United States
subject to the following conditions:
1. The total natural flow of the West Fork Poplar River
and all its tributaries crossing the International
Boundary shall be divided equally between Canada and
the United States but the flow at the International
Boundary in each tributary shall not be depleted by
more than 60 percent of its natural flow.
2. The total natural flow of all remaining streams and
tributaries in the Poplar River Basin crossing the
International Boundary shall be divided equally
5

-------
between Canada and the United States. Specific
conditions of this division are as follows:
a) Canada shall deliver to the United States a
minimum of 60 percent of the natural flow
of the Middle Fork Poplar River at the
International Boundary, as determined below
the confluence of Goose Creek and Middle Fork.
b) The delivery of water from Canada to the
United States on the East Poplar River shall
be determined on or about the first day of
June of each year as follows:
i) When the total natural flow of the
Middle Fork Poplar River, as determined
below the confluence of Goose Creek,
during the immediately preceding March
1st to May 31st period does not exceed
4690 cubic decameters (3800 acre-feet),
then a continuous minimum flow of 0.028
cubic meters per second (1.0 cubic feet
per second) shall be delivered to the
United States on the East Poplar River
at the International Boundary throughout
the succeeding 12 month period commencing
June 1st. In addition a volume of 370
cubic decameters (300 acre-feet) shall be
delivered to the United States upon de-
mand at any time during the 12 month
period commencing June 1st.
ii) When the total natural flow of the Middle
Fork Poplar River, as determined below
the confluence of Goose Creek, during the
imediately preceding March 1st to May
31st period is greater than 4690 cubic
decameters (3800 acre-feet), but does not
exceed 9250 cubic decameters (7500 acre-
feet) then a continuous minimum flow of
0.057 cubic meters per second (2.0 cubic
feet per second) shall be delivered to
the United States on the East Poplar
River at the International Boundary dur-
ing the succeeding period June 1st through
August 31st. A minimum delivery of 0.028
cubic meters per second (1.0 cubic feet
per second) shall then be maintained from
September 1st through to May 31st of the
following year. In addition, a volume of
617 cubic decameters (500 acre-feet) shall
6

-------
be delivered to the United States upon
demand at any time during the 12 month
period commencing June 1st.
iii) When the total natural flow of the Middle
Fork Poplar River, as determined below
the confluence of Goose Creek, during the
immediately preceding March 1st to May
31st period is greater than 9250 cubic
decameters (7500 acre-feet), but does not
exceed 14,800 cubic decameters (12,000
acre-feet), then a continuous minimum
flow of 0.085 cubic meters per second
(3.0 cubic feet per second) shall be
delivered to the United States on the
East Poplar River at the International
Boundary during the succeeding period June
1st through August 31st. A minimum
delivery of 0.057 cubic meters per sec-
ond (2.0 cubic feet per second) shall then
be maintained from September 1st through
to May 31st of the following year. In
addition, a volume of 617 cubic decameters
(500 acre-feet) shall be delivered to the
United States upon demand at any time
during the 12 month period commencing
June 1st.
iv) When the total natural flow of the Middle
Fork Poplar, as determined below the con-
fluence of Goose Creek, during the immedi-
ately preceding March 1st to May 31st
period exceeds 14,800 cubic decameters
(12,000 acre-feet) then a continuous mini-
mum flow of 0.085 cubic meters per second
(3.0 cubic feet per second) shall be
delivered to the United States on the East
Poplar River at the International Boundary
during the succeeding period June 1st
through August 31st. A minimum delivery
of 0.057 cubic meters per second (2.0
cubic feet per second) shall then be main-
tained from September 1st through to May
31st of the following year. In addition,
a volume of 1,230 cubic decameters (1,000
acre-feet) shall be delivered to the United
States upon demand at any time during the
12 month period commencing June 1st.
7

-------
c) The natural flow at the International Boundary
in each of the remaining individual tributaries
shall not be depleted by more than 60 percent
of its natural flow.
3. The natural flow and division periods for apportion-
ment purposes shall be determined, unless otherwise
specified, for periods of time commensurate with the
uses and requirements of both countries.
The IJC also authorized study of the water quality impacts of the
power plant to be done by the Poplar River Task Force. The results were
published by the IJC in 1979 in a report titled “International Poplar
River Water Oual I ty Study’ 1 . The IJC held publ i c hearings on this report
at Scobey, Montana and Coronach, Saskatchewan in September, 1979. The
IJC will review the Board report and hearings and will make recommenda-
tions to the Canadian desk at the U.S. State Department and the Canadian
Department of External Affairs. The U.S. EPA has a liaison person work-
ing with the State Department and the IJC. A water quality monitoring
program is being developed by the State of Montana and the U.S. State
Department.
Since the SPC power plant is being constructed outside of the U.S.
boundaries, there is no direct obligation for the preparation of an
Environmental Impact Statement (EIS). However, a federal action is
required for the U.S. Government to enter into a flow apportionment
agreement with Canada. Therefore, an EIS is required prior to such an
agreement which investigates all types of impacts not just flow or
water quality related impacts.
This environmental statement has been prepared to analyze the
integrated impacts of power plant operation and flow apportionment on
the Poplar River Basin and associated areas within the U.S. It is
designed to meet NEPA requirements associated with the final acceptance
of a flow apportionment agreement between the governments of the U.S.
and Canada.
8

-------
3. ALTERNATiVES INCLUDING THE PROPOSED ACTION
3.1 ATMOSPHERIC EMISSIONS AND CONTROL
Current plans for the SPC Power Plant include operation with 0 per-
cent 502 control (i.e., no flue gas desulfurization equipment). The
plant is also expected to operate at a particulate control rate of 99.5
percent by the use of electrostatic precipitators.
The impact analyses presented in Section 5.1 indicate that there
will be no violations of ambient state or Federal air quality standards if
the plant is operated with up to four 300 MWe units with no SO 2 controls.
The highest predicted S02 concentrations will occur during atmospheric
inversions which result in emission plume fumigations. Although speci-
fied levels of S02 concentrations may be exceeded during fumigation
events, it is improbable that these events will be of sufficient fre-
quency to violate Montana’s 1-hour SO 2 standard. Tne highest mean annual
increases above background SO2 concentrations are predicted by the CRSTER
model to occur along the 120° to 1700 azimuths (generally southeast) from
the plant site; predicted S02 concentrations, even at locations near the
International Boundary, are well below standards. Moreover, the pre-
dicted maximum concentrations and duration of exposures are also below
acute and chronic threshold limits for the most sensitive plant species.
Therefore, although the SPC power plant will result in elevated concen-
trations of atmospheric constituents (SO 2 , NOx and particulates), there
are rio projected impacts on terrestrial ecosystems or human health in the
area.
Although no air quality impacts are predicted with 0 percent SO 2
control, the effects of flue gas desulfurization (FGD) are also consid-
ered. Current FGD systems are capable of SO 2 control ranging from about
60 to over 90 percent reduction. Therefore, the alternatives of 60 and
90 percent SO 2 control were considered. Table 3.1-1 illustrates the
effocts of S02 stack controls and number of operating units on ambient
S02 concentrations. When the maximum allowable increases for Class II
areas are compared with predicted increases resulting from the Poplar
River Plant, the maximum values with four operating units and 0 percent
S02 control represent only about 37.5 percent and 24.0 percent of the
3-hour and annual allowable values, respectively.
A similar comparison may be made for particulate emissions. At a
99 percent level of ESP control , the maximum values are only 14.0 per-
cent and 2.1 percent of the maximum allowable (PSD) 24-hour and annual
increases, respectively. These percentages would be reduced to 7 and 1
percent with the use of 99.5 percent ESP’s (as planned). It should be
emphasized that the preceding discussion has considered the maximum con-
stituent concentrations occurring near the International Boundary. The
predicted maximum concentrations occurring in most of the U.S. part of
the Poplar River Basin are much less than those near the boundary. For
example, maximum 507 concentrations predicted for most of Daniels County
9

-------
Table 3.1—1
MAXIMUM CONCENTRATIONS IN THE U.S. EXPRESSED AS A PERCENTAGE
OF THE MAXIMUM ALLOWABLE INCREASE (PSD) IN CLASS II AREAS
Level of
Control
Number of Units
Two
3-hour
Four
18.7
37.5
O percent
60 percent
90 percent
annual
12.0
24.0
3-hour
7.5
15.0
annual
4.8
9.6
3-hour
1.9
3.8
annual 1.2
2.4
10

-------
are about an order of magnitude (ten times) less than those occurring
near the border. Therefore, the relationship of predicted ambient con-
centrations to maximum allowable increases is considerably more diver-
aent -in most of Daniels County than is indicated by the preceding com-
parisons at locations near the border.
The potential for morning fumigation events is quite high in the
Poplar River Basin and the predicted ground-level concentrations of SO 2
are about equal to the acute threshold limits for sensitive species.
Although no adverse impacts on terrestrial vegetation due to fumigation
are predicted due to the short duration and isolated spatial occurrences,
the projected SO 2 concentrations during fumigation would be reduced to
below acute threshold limits with the addition of FGD.
The Medicine Lake National Wildlife Refuge is a class I area located
approximately 65 miles downwind from the power plant site. This distance
is beyond the reliable predictive limits of the CRSTER model; however,
the potential impacts of atmospheric emissions must be considered due to
the very limited concentration increases allowed based on PSD regulations
(1977 Clean Air Act). Additionally, the Fort Peck Indian Reservation
represents an area which is now being considered for class I designation.
The northern boundary of this area is directly south of the power plant
site at a distance of approximately 30 miles.
With 99.5 percent control (1200 MW, four units), the particulate
concentrations at a downwind distance of 30 miles (the limit of model-
ing results) are about 0.8 and 0.08 ug/m 3 for the maximum 24-hour and
annual mean concentrations, respectively. These values are well be-
low the PSD values of 5 and 10 pg/ma for the corresponding exposure
times. With 99 percent control the predicted concentration at 30
miles downwind would still be only twice the 99.5 percent values, and
would therefore remain well below the class I PSD values. Moreover,
dispersion effects in the next 35 miles to the refuge would result In
considerable further reduction in the particulate concentrations.
The predicted maximum SO 2 concentrations for two power plant
units (zero percent control) at the modeled distance limits (30 miles)
equal or exceed the class I PSD regulations for 24-hour and 3-hour
averages (Table 3.1-2). t Ihile these levels would result in violations
within the proposed class I area, the values presented are for areas
located about 35 miles from the class I wildlife refuge. Extrapolation
of the modeled diffusion rate indicates that ambient increases in SO 2
concentrations at the refuge would be well below PSD regulation values.
As indicated in Table 3.1-2, the use of 60 percent or higher SO 2 control
on a two-unit plant would result in increased concentrations less than
class I PSD values at a distance of 30 miles. Therefore, with the FGD
systems, a 600 MW plant would have an extremely low probability of PSD
contravention at the refuge.
11

-------
Table 3.1-2
A COMPARISON OF CLASS I PSD VALUES AND MAXIMUM PREDICTED
SO 2 CONCENTRATIONS (pG/M 3 ) 30 MILES SOUTHEAST OF PLANT
Maximum S07 Concentrations
Class I
PSD Regulations Two Unit Four Unit
( ‘ ig/m 3 ) 0* (__60* 90* 0* 60* 90*
Annual 2 1 0.4 0.1 2 0.8 0.2
24-hour 5 10 4 1 20 8 2
3-hour 25 25 10 2 50 20 5
*Percent of SO 2 Control
The operation of a 1200 MW plant without EGO would result in pre-
dicted maximal 24-hour SO 2 concentration increases of about 20 ig/m 3 at
a distance of 30 miles (Table 3.1-2). Such increases above ambient
levels are about four times the class I PSD values, and, based on the
approximate predicted diffusion rate for 24-hour concentrations beyond
the 30-mile modeling distance, they would remain slightly above the PSD
values (7.1 ug/m 3 ) at the 65-mile distance to the refuge. Using the
same approach for annual or 3-hour values would result in predicted con-
centrations below the allowable limits. Operation of four 300 MW units
with at least 60 percent EGO systems would result in predicted concen-
trations below the maximum allowable increases at the Medicine Lake
Refuge.
Although the previous approximate calculations result in a potential
for a 24-hour S02 violation (with four units and zero percent control)
of the PSD regulations for the Medicine Lake Wildlife Refuge, the actual
probability for such an occurrence is extremely low due to the nature of
the modeling methodology. The predictive value of a Gaussian Plume model
such as CRSTER is quite low at distances beyond 25 to 30 miles from the
source. The resultant concentrations at the distance limits of the model
should be considered very conservative (i.e., they are probably much
higher than the actual concentrations under normal meteorological condi-
tions). This is especially true under stable conditions. The conservative
nature of the predicted SO 2 concentrations is further enhanced when the
limits are extended to 65 miles. The CRSTER modeling results indicate a
possibility of contravention of class I PSD regulations in the northern
portion of the Fort Peck Indian Reservation, which is a potential class I
area. As indicated in Table 3.1-2 the operation of four units with 60
percent control would result in the violation of the 24-hour standard at
a distance of 30 miles from the power plant. All the class I standards
could be met if the SO 2 control was increased to 90 percent.
12

-------
3.2 FLOW RELATED ALTERNATI’/ES
3.2.1 Alternative Flow Apportionments
Four alternative apportionments were considered by the Poplar River
Task Force. Three of the alternatives assume an equal division of the
total natural flow of the Poplar River. One alternative is a division
of flow such that Canada would be entitled to 70 percent of the flow and
the U.S. 30 percent. The alternatives are summarized in Table 3.2-1.
The no-action case, i.e., no apportionment, would allow Canada to
use all of the flow within the Canadian part of the basin for the worst
case. The scenarios (4 to 6) were run with 1975, 1985, and 2000 levels
of development in the U.S. The model scenarios did not include the
Cookson Reservoir or power plants. Another scenario (3) could be con-
sidered a no-action case in that it included the Cookson Reservoir but
no power plants and 1975 levels of development.
The June flcws at the Lower West and Middle forks at a 10 percent
frequency are the same for all the alternative apportionments (see
Appendix Table F-2). Flows occurring at a 50 percent frequency in 1975
are lowest for Apportionment V and highest for Apportionment IVa and VI.
In 2000 with four power plants the flows are 8.1 ac-ft higher under
Apportionment V. Flows occurring at a 90 percent frequency are 24.3 ac-
ft higher under Apportionment IVa in 1975 and 2000. Water demands cannot
be met with water available 9 out of 10 years at station 7 under any of
the apportionments in 1975 or 2000 and at station 11 in 2000. Water
demands could not be met 50 percent of the time in 2000 at station 7
under Apportionment V. Thus, although slightly more water is available
in June under Apportionment IVa, it is not enough to meet demands.
Comparing flows under apportionments IVa and VI for July, August,
and September at stations 7 and 11 shows that flows are zero 9 out of
10 years for these months. Water available 50 percent of the time is
zero at station 7 for both apportionments and zero in July and September
at station 11. Flows in August are 8.1 ac-ft higher at station 11 under
Apportionment IVa at the 50 percent frequency, but 57 ac-ft less in March
under Apportionment IVa. At the 90 percent frequency the flows are 8
ac-ft higher in July at station 7, but 16.2 ac-ft lower at station 11
under Apportionment IVa. This same pattern is also true for March flows.
The March flows are critical for fish and wildlife and for spreader
irrigation. The Middle Fork has better fish habitat than the West Fork,
but less spreader irrigation. Demands for spreader irrigation in March
can be met at both the Middle and West Forks under both apportionments
IVa and ‘11 through 2000. Summer irrigation demands are exceeded for the
same months and frequencies for both apportionments. Overall , 40 ac-ft
more water would be available 50 percent of the time under Apportionment
VI and IVa for these months. Eight ac-ft more water would be available
10 percent of the time under Apportionment VI.
13

-------
Table 3.2-1
DESCRIPTIONS OF APPORTIONMENT ALTERNATIVES
Percent of Flow to United States ——
Model
Apportionment West Fork Middle Fork East Fork Other Trib. Scenarios
II 50 percent division no restrictions --
lIla 40-60 60 30-50 40 1O_ 12*
IIIb 60 40-60 30-50 40 7_9*
IVa 40-60 60 Releases** 100 18-22
IVb 60 40-60 Releases 100 13-17
V 30 30 Releases 100 23-27
VI 50 60 Releases 60 28-32
*These scenarios did not include Cookson Reservoir or power plants, so will not
be discussed.
*AThe volumes of releases are at least 1 cfs. The specific releases are listed
in Table 5.2-2.

-------
Apportionment IVb would provide at least 60 percent of the natu-
ral flow of the West Fork, but only 40 percent of the natural flow of
the Middle Fork. Flow at the Lower Middle Fork (station 7) would be
slightly less than scenarios 28 to 32, and demands could not be met
from June to September for the same percentage of time as for scenar-
ios 28 to 32. Flows in the West Fork would be higher, but demands
still cannot be met for the same percentages of time from May through
September.
Thus, none of the flow apportionment alternatives allow full
water demands to be met. Apportionment VI appears best from a water
quantity viewpoint, although the increased March flows in the Lower
Middle Fork under ha would be beneficial to fish and wildlife.
3.2.2 No Action Case
The no-action case under existing 1975 conditions with the Cook—
son Reservoir and the worst case under 1975, 1985, and 2000 levels of
development are compared to historical flows (scenario 2). Flows are
compared for March which is important for stock, fish and wildlife,
and spreader irrigation and June through September which is the main
irrigation season.
Flows under existing conditions for the Middle and West Forks
(stations 4, 7, 9 and 11) are the same for scenarios 2 and 3. Flows
on the East Fork at the border are higher at low flows for scenario 3
due to seepage from the reservoir. The median flows in March are
about 85 percent less under scenario 3 than scenario 2. Summer median
flows for scenario 3 vary from 19 percent less in September to 20 per-
cent more in July. March high flows are about 46 percent less under
scenario 3. Summer high flows were 35 percent less in June to 39 per-
cent more in August. Low flows at Scobey (station 3) are zero in
March for both scenarios 2 and 3. Summer flows at station 3 were 2
to 240 percent higher under scenario 3 at low flow conditions because
of the continuous seepage from the reservoir. Median flows were 7 to
13 percent less and high flows were 38 to 30 percent less in June and
September due to the increased water uses of 1975. July and August
flows are 75 to 240 percent higher under low flow conditions and 38
percent higher under median conditions due to delayed irrigation re-
turn flows. The high flows were 7 percent less in July and 23 percent
higher in August.
The low flows under existing conditions are adequate to meet
municipal, stock, and irrigation demands at Scobey (station 3). Irri-
gation and stock demands can be met on the Middle Fork only in March,
April, and May. Stock and irrigation demands can be met on the West
Fork for all months except August and September. Stock and irrigation
demands on the main Poplar (station 8) cannot be met in March or June
through September. Demands on the Lower Main Poplar (station 12) can-
not be met in March, July or August. Under high flow conditions de-
mands can be met at all staticns except the Middle Fork (station 7)
in August and September.
15

-------
Flows for the worst case (scenarios 4 to 6) at the border stations
of the three forks are zero during the summer and for the low to median
flows in March (see Appendix Table F-5). March low flows are the same
as historical conditions in 1975 for all but station 7 and are only 2
percent less there. Under 2000 levels of development the flows are zero
at station 3 and 47 percent less at station 11. Median and high flows
are significantly less but water demands can be met at stations 3, 7,
and 11 through the year 2000. Demands at station 8 could be met only
with high flows. Demands at station 12 could be met in 1975 and 1985
with the high flows only, although water t.,iould be available from the
Poplar reservoirs. Low flows in the summer are about the same at all
stations except 3 which is zero for the worst no-action cases. Median
and hiah flows in 1975 are between 25 and 50 percent less. By 1985 and
2000 the increased water demands result in zero flows most of the time
for the low flow and median flow cases. The 1985 and 2000 demands can
be met under high flow conditions at all stations in June and July and
under 1985 demands in August and September. The months when demands
cannot be met are the same as for scenario 3, but fewer acres could be
irriaated because of the decreased amount of water.
3.2.3 Demand Releases
The actual impact of the flow apportionments will depend partly on
the timing and flow rate (cfs) of the one-time releases. The schedule
used in the model spread the release over the period from May to Septem-
ber as recommended by the Poplar River Task Force:
Month Percent
May 12
June 18
July 32
August 27
September 11
This division was based on need for irrigation rather than the timing of
a specific method of irrigation. However, the average number of irriga-
tion applications in the Poplar Basin was 2.4 with a maximum of 4. One
approach to increase the water available for irrigation would be to drop
the release in May and increase the releases in July through September
when more water is needed. Another consideration is the flow rate - a
higher flow rate for a period corresponding to the length of the normal
irrigation application would be beneficial.
16

-------
Another approach would be to use the September release amount to
miticiate the impact of the apportionment on fish and wildlife. The re-
lease could then be made in April of 33 ac-ft. If the release was made
in one day the flow rate would be 16.6 cfs and if in two days the flow
rate would be 8.3 cfs. These flow rates are not adequate for maintain-
ing channel morphology but would be a significant addition to the appor-
tionment flows on the East Fork and would increase the flow rate to the
recommended flow of 10 cfs.
3.3 WATER QUALITY
Evaluation of alternative transboundary flows, i.e., 70/30 apportion-
ment (V), indicates that the operation of one to four 300 MW units at
successive levels of development will have the same effect on water
quality at station 1 as the identical sequence of development at the
recommended apportionments (scenarios 28 to 32). The predicted boron,
1DS and SAR levels at station 1 provided by the water quality simulation
model for scenarios 23 to 27 are shown in Tables 3.3-1 to 3.3-3. These
values at the 90 and 50 percent probability levels are identical to those
predicted for scenarios 28 to 32 (i.e., equal apportionment). Similar
levels of boron, TDS and SAR are predicted at stations 3 and 8 under
scenarios 28 through 32 and 23 through 27. Boron concentrations result-
ing from scenario 32 will result in levels exceeding acceptable limlts
for semi-tolerant crops 10 percent of the time. The levels of these con-
stituents at stations 3 and 8 are presented in Appendix Tables H-i through
H-6.
Modeling scenarios 4, 5 and 6 represent the worst case action in
terms of apportionment alternatives. Under these conditions Canada would
make unlimited use of available water resources. These scenarios do not
consider the impacts of either water storage (reservoir in place) or
power plant utilization, and, therefore, represent unlikely alternatives.
Based on water quality sinulations, scenarios 4, 5 and 6 would result in
both the unavailability of sufficient supplies and significant deterio-
ration in water quality at station 1 during the period of March through
September. Although SO 4 and SAR levels would be expected to remain within
acceptable limits, predicted boron and TDS levels at station 1 would be
high durinci the period July through September (50% probability). Water
quality would not be affected by scenarios 4, 5 and 6 at stations 3 or 8.
3.3.1 Mitigation of Water Quality Impacts
The increased salinities of East Fork waters under the apportionment
alternatives are due to mine and power plant related inflows, Fife Lake
overflows, and evaporation in Cookson Reservoir. Modifications in the
ash settling lagoons to elininate decanting to the reservoir and minimize
ground ‘ iater seepage would decrease TDS concentrations in the East Fork
(station 1) by about 10 percent. Boron concentrations could be reduced
significantly from a maximum predicted 90 percent probability level of
9.3 to about 2.1 mg/9 .. The proposed mitigation measures by SPC include
17

-------
Table 3.3—1
PROJECTED BORON CONCENTRPTIONS (MG/2 ) AT STATION 1
Percent Month
Probabi 1 i ty
Scenario Level March April j June July August September
j90 - - - - - - -
150 - - 2.1 2.0 3.2 4.4 3.5
190 - - - - - - -
150 - - 2.2 2.0 3.4 4.9 3.8
190 - - - - - - -
6 150 - - 2.2 2.2 4.3 5.5 4.9
9O 1.9 1.8 1.9 1.9 1.9 2.0 2.0
23
50 0.9 0.8 0.8 0.8 0.8 0.9 0.9
j9O 2.4 2.2 2.2 2.2 2.3 2.4 2.5
24 50 1.2 1.0 1.0 1.0 1.0 1.1 1.2
90 3.7 2.9 2.9 3.0 3.1 3.3 3.6
25 50 1.6 1.3 1.3 1.3 1.4 1.5 1.6
j90 3.7 2.9 2.9 3.0 3.1 3.3 3.6
26 5O 1.6 1.3 1.3 1.3 1.4 1.5 1.6
590 8.0 4.8 3.8 4.1 4.7 5.5 6.5
27 150 2.2 1.5 1.3 1.3 1.6 1.8
18

-------
Table 3.3-2
PROJECTED SODIUM ABSORPTION RATIOS AT STATION 1
Percent i
Probability onth
Scenario Level March April May June July August September
100 - - - - - - -
4
5O - - 3.9 3.7 5.2 6.1 5.5
(90 - - - - - - -
5
‘50 - - 3.9 3.7 5.3 6.4 5.7
190 - - - - - - -
0
4.0 3.8 6.0 6.6 6.4
9O 17.1 14.8 13.6 12.8 12.2 12.0 11.9
23
‘50 8.5 5.7 5.1 4.9 4.8 4.8 4.8
24 190 19.7 15.7 14.7 13.8 13.2 13.1 13.2
50 10.6 7.0 6.0 5.8 5.7 5.7 5.8
(90 22.2 17.6 16.4 15.7 15.1 15.1 15.4
25
‘50 12.6 8.4 7.3 7.1 7.0 7.1 7.1
190 22.2 17.6 16 . 15.7 15.1 15.1 15.4
26
50 12.6 8.4 7.3 7.1 7.0 7.1 7.1
27 j90 31.2 22.9 18.9 18.6 17.7 16.6 16.4
50 12.8 8.5 6.0 6.0 6.1 6.3 6.4
19

-------
Table 3.3-3
PROJECTED TDS CONCENTRATIONS (MG/SQ) AT STATION 1
Percent Month
Probability
Scenario Level March April May June July August September
(90 - - - - - -
4
‘50 - 1192 1103 1635 2279 1833
190 - - - - - -
5
50 - 1207 1127 1745 2536 1960
(90 - - - - - -
6
50 - 1243 1207 2243 2853 2530
9O 984 930 935 946 958 992 1028
23
5O 583 530 507 522 547 573 596
24 1249 1107 1110 1131 115d 1208 1268
50 747 653 53 1 653 678 711 744
90 1967 1459 1462 1506 1559 1668 1793
25
50 1003 859 816 842 885 936 991
(90 1967 1459 1462 1506 1559 1668 1793
26
‘50 1003 859 816 842 885 936 991
190 4796 2551 2381 2546 2867 3358 3978
27
‘50 1347 999 847 884 965 1068 1146
20

-------
recirculat-jon of the ash lagoon decant to the power plant instead of
discharge to the Cookson Reservoir and compaction of a 300 rr n clay layer
as a base for the ash lagoons. These measures would limit total seepage
to the East Fork and reservoir to less than 2 1/sec (IJC, 1979). The
concentrations in the East Fork could be reduced even further if compact-
ion was extended to a 600 mm layer which would limit seepane to approxi-
mately 0.7 Z/sec (IJC, 1979). The resulting boron concentrations in the
East Fork were estimated to be a maximum of 3.3 mg/2. if no ground water
mounding occurred and a maximum of 5 mg/P. if mounding at the ash lagoon
sites occurred.
Other mitigation methods to reduce impacts of saline irrication
waters would most likely involve methods to control salts in irrigation
return water and salt buildup in soils. These methods are fully de-
scribed in Section 5.3.4. A leaching fraction of 0.2 minimizes the
impacts of the poor quality irrigation water on crop yields. Increasing
the leaching fraction to 0.3 results in only a small improvement in
yield which would probably be offset by the loss of acreage which could
be irrigated with the available water.
21

-------
4. AFFECTED ENVIR0NI 1ENT
4.1 LOCATION
The Poplar River basin is located in Southern Saskatchewan
(Canada) and Northeastern Montana between 49° 301 and 48° north lati-
tude and 104° 45’ and 106° 45’ west longitude (Figure 4.1-1). The
Poplar River originates in Saskatchewan as three main tributaries:
West Fork, Middle Fork and East Fork. A major portion of the Middle
and East Forks drainage basins lies in Canada. Both of these streams
join to form the main stem of the Poplar River near the town of
Scobey, Montana, located approximately 12 miles south of the inter-
national boundary. The West Fork sub-basin is located mainly in
Montana and joins the Poplar River at a point approximately 37 miles
below the U.S.-Canadian boundary.
The Poplar River may be divided into five major segments which
will form the basis for subsequent discussions concerning apportion-
ment alternatives and associated flow regimes. In addition to the
three aforementioned forks, the main stem of the Poplar River may be
divided into a 25—mile segment between the confluence of the East and
Middle Forks and the confluence of the West Fork and a 40-mile segment
between the West Fork confluence and the Missouri River.
Three small transboundary tributaries originate in Canada and
join the East, Middle and West Forks near the international boundary:
Coal Creek, Cow Creek and the East Tributary. Other tributaries such
as Butte Creek, Cottonwood Creek and Police Creek are located entirely
in the U.S. part of the basin.
The Poplar River Power Plant is currently being constructed at a
site located approximately 4.3 miles north of the International Bound-
ary (Figure 4.1-2). The site is located adjacent to a 1600-acre im-
poundment (Cookson Reservoir) which will be used as a water source
for the plant’s once-through cooling water system. The total storage
volume of Cookson Reservoir is approximately 30,000 acre-feet. Al-
though only one 300-MW coal-fired plant is currently under construc-
tion, an additional unit of equal size is planned for the site. The
addition of two other 300-MW units is also considered feasible if the
r ecessary increased cooling water is available at the site.
In addition to an adjacent surface coal mine, the following an-
cillary facilities are planned: coal handling plant, water treatment
facilities and wet-ash disposal system (lagoon). The lagoon location
is indicated in Figure 4.1-2.
22

-------
Canada
USA
Oplieini - - Glentana
\ .

4 0 4 8 12
6 0 6 12 18
Kiloinieters
Saskatchewan
-- — - Sheridan County -- - - Montana
Westhy
Redstone
-- -“
Plentywood
‘p
Northern Boundar
Fork Peck l.R. -
Medicine) Medicine
Lake Natlondi
Wildlife Refuge
- Roosevelt County
Froul
LOCATION OF THE POPLAR RIVER BASIN
N
Miles
La
Brockton —- - - - 2
Figure 4.1-1

-------
Figure 4.1-2
LOCATION OF POPLAR RIVER POWER PLANT SITE
24
Scale
0123456
km

-------
4.2 GEOLOGY AND SOILS
The Poplar River Basin is located in the Northern Great Plains.
The area is characterized by flat to gently dipping sedimentary strata
overlain by glacial deposits and alluvium in the river valleys. The
study area is at the northern edge of the Williston Basin, a large
syncline in Montana and North Dakota. The formations exposed in the
area are the Bearpaw Shale,Fox Hills sandstone, and Hell Creek Formation
of Cretaceous aae; the Fort Union arid Flaxville Formations of Tertiary
Age, and Ouaternary gravels, alluviun, and glacial deposits. The gene-
ral stratigraphic section for the region and local variations within
the Poplar Basin are discussed for the formations in Appendix A-i.
The Quaternary period can be divided into the Pleistocene series
dominated by glacial activity and the Holocene series dominated by
fluvial activity. The oldest formation which predates the glacial
period is the Wiota Gravels. The location of glacial and recent
deposits in the Poplar River Basin are shown in Figure 4.2-1. Gen-
erally, the till deposits consist of unstratified clay, silt, sand
and gravel. The average thickness is 10 feet. Other types of glacial
deposits include eskers, kames, glacial outwash sediments, and lake
deposits.
Recent deposits include alluvium in the river valleys, colluvium
in slope-wash and alluvial fans, landslide material, and sand dunes.
The alluvium is composed of locally-derived sand, silt and gravel and
may be up to O feet thick. Samples of the river banks analyzed by
the Montana Department of Natural Resources and Conservation (1978) at
nine locations along the entire Poplar River’ system were as follows:
Flood—plain 85-1OO Silty Loam
5-15 Gravel
Upper Bank 85-95 Silty Loam
O-5 Sand
2.5-15 Gravel
Lower Bank 85-95 Silt/Silty Loam
5-15 Sand arid Gravel
Channel 40-85 Silt
Channel 5-2O Sand
10-10OO Gravel
Along the Mainstem of the Poplar River below the confluence with the
West Fork and the middle of the East Fork, the lower bank and channel
are 100 percent gravel . Sand dunes occur southeast of Four Buttes and
northeast of the town of Poplar. The dunes consist of gray sand with
a relief of 10 feet. Vegetation stabilizes most of the dunes.
25

-------
Tfg
Tf J
: Qewd 5” ’ •• — -
Tfg V \
\\ Qal T g QaI
_. - —- _: := Qsq “ -
3
- T g ) /
—:‘ r -G - ’ — -‘ .‘ If
r Tfg
L
_
\
j/ j N ) ‘ ‘ ‘ Qewd
• -‘ 7 I / \ J \, Qj
1’
Qewd c / / “
( ) Q g\\ ;//f ’ .:’\ Qd, ’\

Key afrer Howard, 1960 0 MILES 6
Qd Dune Sand Qewd Glacial Till (Early W,sconsin age)
Oic Ice contact deoosits (Kames, Eskers) Tfg F!axvil le Gravel
Qsg Stratified Drift Deposits (Wisconsin age) Glacial Channels
Ocg Crane Creek Gr ve1 Glacial Bars
Qrnc Glacial Till ( ‘ankato Drift - Wisconsin age) Moraine Topograohy
Qal Alluvium - flooa plain deposits
Figure 4.2-1 QUATERNARY GEOLOGY OF THE U.S. PART OF THE
POPLAR RIVER BASI’J
26

-------
The Canadian portion of the Poplar River Basin has a similar geo-
logic history to the U.S. part of the basin and thus has similar rock
types. The general stratigraphic sequence found in the area from
youngest to oldest strata is:
o Alluvium
o Glacial Till
* Empress Group
o Ravenscrag Formation
o Frenchman Formation
o Bearpaw Shale
• Older sedimentary rocks
Ouaternary deposits include the alluvium, glacial till, and the
Empress Formation. Alluvial deposits occur in the floodplain of the
East Fork of the Poplar River and Girard Creek. The glacial till is
less than 70 feet thick.
The soils in the Poplar River Basin are derived from the glacial
till and out iash deposits and the locally exposed bedrock formations.
The soils are predominantly a loam type. The soils in the Canadian
part of the basin belong to the Brown and Chernozemic Brown groups. At
least 12 different soil series are present in the U.S. part of the
basin. Soils developed on the glacial till are predominantly Williams
and Zahill Soil Associations. Floodplain soils are mostly Haverlon,
Trembles, and Lohler soil series. Soils developed on till over bedrock
in the southern part of the basin belong to the Phillips-Scohey-Thoeny
Soil Associations. These soils are loam to clay loam with a lime zone
at shallow depths. Detailed descriptions of the U.S. and Canadian soil
types are provided in Appendix A-i.
Permeability ranged from low (0.06 inches/hour) to moderate (0.6
to 2 inches/hour) at the sampled sites in the U.S. part of the basin.
Low permeability soils occur in parts of the floodplain and can restrict
agricultural use of these areas. In the lower Poplar River Basin exces-
sively high permeability soils (2 to 20 inches/hour) are found on hilly
parts of the glacial till and sand dunes. The Banks soil series is
sandy ‘ iith high permeability and low water holding capacity (approxi-
mately 3.8 inches in the first four feet) (Klages, 1976).
Most of the irrigated lands in the Poplar River Basin have soils
belonging to three soil series (Haverlon, Trembles, and Lohler). Ap-
proximately 75 percent of the irrigated soils in the area belong to
the Haverlon and Trembles series (Smetana, 1979). Both of these soils
are loam types. The Lohler soil is a fine loam iith some montrnoril-
lonite. The average ‘. iater holding capacity in the deep loam soils
27

-------
was estimated at 10 inches (USDA, 1964). Thus, the water requirement
for spreader irrigation systems was estimated as 10 inches by the Poplar
River Task Force. The crop requirement for gravity or pumping irriga-
tion was estimated by the Task Force as five inches since the applica-
tions should take place when the soil moisture is 50 percent depleted.
The total water requirement due to losses by seepage and evaporation was
therefore estimated as 7.7 inches (Montana DNRC, 1978).
The chemical properties of the soils are an important factor in
determining the suitability of the soils for agriculture. The CEC
(cation exchange capacity) of soils on the Fort Peck Indian Reservation
ranged between 15 and 44 meq/iOO g. The pH of soils through the U.S.
part of the Poplar Basin ranged between 7.5 and 9.5 (see Table A.1-2).
Analyses of major cations including boron are included in Appendix A-i.
Mineral resources of the Poplar River Basin are associated with
the fiat-lying sedimentary strata of Cretaceous and Tertiary Age and
the glacial deposits of Quaternary Age. Detailed descriptions of the
fuel and non-fuel resources are included in Appendix A-i. Fuel re-
sources include oil and coal. Oil has been produced from five fields
in the Poplar River Basin. Several lignite coal fields exist in the
U.S. part of the Poplar River Basin, although no coal is presently
being mined. The fields have a low probability for development due
to overburden and low Btu value (Montana Energy Advisory Council, 1976).
There are 12 coal fields in the Canadian part of the basin. The Hart
coal seam will be mined for the SPC power plant at a rate of 50 million
tons per 300 MWe unit over the life of the plant.
Non-fuel resources include potash, quartzite, bentonite, sand,
gravel, clay and marl. Potash is mined in Saskatche ian by the Pitts-
burgh Plate Glass Industries. The Farmer’s Potash Company has plans
for a mine and processing plant near Scobey, Montana.
4.3 LAND USE
4.3.1 General
The Poplar River Basin (Daniels and Roosevelt Counties) and ad-
jacent Sheridan County are rural, sparsely-populated areas in which the
majority of land is devoted to agricultural use (Table 4.3-1). In
Daniels and Roosevelt counties over 98 percent of the total acreage is
classified as cropland, pasture or rangeland. Almost all of the crop-
land is used for dryland farming.
On the Fort Peck Indian Reservation the majority (68 ) of the
reservation lands was used for open grazing in 1972. Only 30 percent
of this land belongs to the Indians. The other major land USC iS dry
land farming (29%). Sixty percent of the dry land farming is on land
used by the Indians.
28

-------
Table 4.3-1
LAND USE CHARACTERISTICS OF TilE U.S. POPLAR RIVER DRAINAGE IN 1967
(DANIELS, ROOS [ VELT AND SHERiDAN COUNNES) EXPRESSED IN ACRES
AND AS TILE PERCENI OF TILE FU1AL AREA
County
Total Area
Cropland
Pasture
Range
Forest
Other
Daniels
911,938
543,166
4,550
352,896
0
11,326
59.6%
0.5%
38.7%
0.0%
1.2%
Roosevelt
1,501,671
731,915
20,090
722,902
8,745
18,025
48.71%
1.34%
48.14%
0.58%
1.2%
Sheridan
1,047,532
638,310
32,500
368,081
0
8,641
58.7%
3.0%
33.8%
0.0%
0.8%
TOTAL
3,461,147
1,913,391
57,110
1,443,879
8,745
37,992
55.3%
1.6%
41.1%
.30%
1.1%
Source: USDA, SCS, 1973 and 1976

-------
The majority of the land is privately owned in Daniels, Roosevelt,
and Sheridan counties. Detailed tables for land ownership are given in
Appendix A-2. In Daniels County the amount of federally-owned lands is
relatively small (0.09%) and is comprised of about 850 acres of National
Resource Lands and U.S. Fish and Uildlife Service refuge. The relative
amount of federally owned lands in Roosevelt and Sheridan counties is
greater and comprises 3.6 and 2.5 percent of the total areas, respec-
tively. In Roosevelt County most of the federal land is owned by the
Bureau of Indian Affairs (non—trust lands), while in Sheridan County
almost all of the federal land is associated with the Medicine Lake
National Wildlife Refuge.
The total land area of Fort Peck Indian Reservation is 2,093,124
acres. However, over 50 percent of the reservation land is owned by
non-Indians. All deeded land, state and county, within the reservation
is included in this ownership category, as well as all federally-owned
land not assigned specifically for use by Indians. The remaining land
is divided into three ownership categories, Tribal Indian Trust land
o’,ined by the tribe (11.3%), and individual allotted (30.2 ), and fed-
eral government land assigned for use by Indians (4.1%).
4.3.2 Land Use Surveys
In an effort to describe land use classifications in detail in
the direct down-wind region from the plant, aerial infra-red surveys
were conducted in 1978 (Duggan, 1979). The area surveyed extends
about 30 miles into the U.S. The area was selected to correspond to
the most prevalent wind direction (i.e., north-westerly). Therefore,
these areas would have the highest potential for impact resulting from
atmospheric emissions from the Poplar Plant. The land use classifica-
tions for two of the surveyed areas are included as Figures 4.3-1 and
4.3-2.
The proportion of cultivated land in the U.S. (Areas 1 and 2)
was quite high and ranged from 70.9 to 73.9 percent. These values
are higher than the countywide values of about 60 percent cropland
(Table 4.3-1). There was also a corresponding decrease in the amount
of rangeland when compared with the overall county figures.
Wetlands comprised from 3.5 to 4.3 percent of the U.S. areas
surveyed. In Area 1 the wetlands mapped were adjacent to the East
Fork Poplar River and Cow Creek. In Area 2 the wetlands were directly
associated with Eagle Creek and Whitetail Creek, both of which are
located in the Big Muddy Creek drainage basin.
4.3.3 Agricultural Activities
About 271,000 acres in Daniels County were harvested in 1975.
Less than one percent of this acreage was irrigated. The number of
30

-------
w
nt t .iod i C I f . on
I URII) IN LIII I IVATLI) 3 RTFl( [ I Alit) 4 WIT) AT))) 5 WAILIT
I - U S 0 0 73 9 At / 4 3 0
A.., I ) n )o 2 (3. 11 1 II 9 3 3 (3 il
Hgure 4.3-1 LAND USE IN AR [ A 1 SUkVEY [ D DY DUGGAN (1978)

-------
0
2
M 1
o
2 3
,J
Percent Land IJ’ e llasstlir.itton
I IIRBAFI 2 C l ii TIVAT(D 3 Rlktd,LLAFI 1) WLF(AUD S WAILR
Area2-US 02 709 253 I OI
C,-)
I ”)
Scale
Figure 4.3-2 LAND USE TN AREA 2 SURVEYED BY DUGGAN (1978)

-------
acres harvested has increased by seven percent since 1960. However,
during those 15 years there was a great deal of fluctuation in acres
harvested per year. For example, between 1974 and 1975, the total
number of acres harvested decreased by nine percent or 26,800 acres.
Cattle-, hog-, and sheep-raising also occur to some extent in the
county (Table 4.3-2).
The total number of farms ‘ iith sales of S2,500 and over in the
county increased bet,ieen 1969 and 1974 by 19 from 435 to 454. The
number of irrigated farms as a percent of total farms dropped from
3.9 percent in 1969 to 3.1 percent in 1974, even though the number of
irriaated acres increased from 1,269 to 2,016. The majority of the
farms (88%) are either family or individually operated.
Daniels County, in spite of its small area (1,443 square miles),
ranked third in spring wheat production and fourth in durum wheat
production in the state in 1975. The crops produced in the study area
are grown largely on nonirrigated soil. The crops are usually planted
in strips, leaving unused areas between to preserve water for the acres
planted Barley and oats contribute to the feed grains production
within Daniels County. Others crops raised in the county include flax
seed, winter wheat, and hay. Most irrigated land in the county is
used for producing wild and alfalfa hay (Table 4.3-3).
Roosevelt County is ranked number one statewide in the number of
acres of spring wheat harvested and number three in overall wheat pro-
duction. The crops are grown almost entirely on nonirrigated soil.
Only a small portion of the spring wheat and oats are irrigated. Hay
is tne only crop that is irrigated to any extent and then only about
30 percent of the crop, or 12,700 of the harvested acres. Other crops
planted in the county include barley, oats, rye, corn, flax, safflower,
and mustard (Table 4.3-4). The number of farms with sales of S2,500
and over increased from 636 in 1969 to 659 in 1974. At the same time,
the number of irrigated farms as a percent of total farms dropped from
6.1 percent in 1969 to 4.7 percent in 1974, with the total number of
irrigated acres declining from 6,815 acres to 4,402 acres.
Approximately 675,000 acres within the county were used for graz-
ing in 1977. All of the livestock inventory showed a sharp decline
between 1973 and 1976. The number of sheep and lambs decreased by al-
most 50 percent. Table 4.3-5 shows changes in livestock between 1974
and 1976.
4.4 HYDROLOGY
4.4.1 Surface Water
The Poplar River Basin consists of four sub-basins draining a
total of 3,002 square miles (Figure 4.4-1). The West, Middle, and

-------
Table 4.3-2
LIVESTOCK INVENTORY FOR DANIELS COUNTY
YEARS
Livestock 1974 1975 1976
All cattle and calves 2 ,6OO 24,800 24,600
Milk cows and heifers that have
calved 750 100 100
Beef cows and heifers that have
calved 14,700 15,400 14,000
Stock sheep and lambs 5,000 4,500 5,000
Hogs and pigsk 5,400 5,900 4,800
Chickens* 2,700 2,000 1,300
*Inventory for the years 1973-1975.
Source: Montana Department of Agriculture, Montana Agricultural
Statistics, December, 1976.
Table 1 3_3
ACRES IN IRRIGATED AND NON-IRRIGATED CROPS
DANIELS COUNTY, 1975
Irrigated Non—Irrigated Percent
Crop Acreage Acreage Total Acreage of Total
Spring wheat * 170,000 170,000 63
Barley * 51,400 51,400 19
Durum wheat 0 16,600 16,600 6
Flaxseed 0 5,900 5,900 2
Oats * 2,900 2,900 1
Winter wheat 0 1,500 1,500 1
All hay 2,500 20,200 22,700 8
Totals 2,500 268,500 271,000 100
*Spring wheat, barley, and oats are planted and irrigated as a cover
crop in rotation with alfalfa, but total acreage of irrigated small
grains is less than 1 percent of the total grain production (Montana
Department of Agriculture, 1931-1977).

-------
Table 4.3_Li
CROP PRODUCTION, ROOSEVELT COUNTY, 1975
Crop
Bushels
All wheat
Winter wheat
Spring wheat, except durum
Barley
Oats
All hay
*Tofls
8,211 ,900
1,158,700
5,954,200
1,954,200
466,100
79,000*
Source: USDA and Montana Department of Agriculture,
Montana Agriculture, County Statistics,
‘I. XVI, 197 and 1975.
Table 4.3-5
LIVESTOCK INVENTORY, ROOSEVELT COUNTY
Years
Livestock
1974 1975 1976
Percent Change
All cattle and calves
Milk cows and heifers
that have calved
Beef cows and heifers
that have calved
Stock sheep and lambs
Hogs and pigs*
Chi ckens*
*Inventory is for the year’s 1973-1975
Sour-ce: USDA and Montana Department of Agriculture, Montana
Agricultural Statistics, V. XVI, County Statistics 1974
and 1975.
50,400 44,000 35,000
-30
200
100
100
-50
27,900
26,200
21,700
-22
4,500
2,600
1,900
-48
7,700
5,800
5,500
-29
10,600
9,200
5,800
-44
35

-------
ion-contrbuting areas
— — Sub-basin boundaries
Figure 4.4-1
MAJOR SUB-BASINS OF THE POPLAR RIVER SYSTEM
36
POPLAR RiVER 8ASIP
& __ T T.1--- ;
- -._= - -.c —

-------
East Poplar ri vers all have their headwaters in southern Saskatchewan.
The East and Middle forks join at approximately 48°17’ North latitude
to fork the main stem of the Poplar River. These are joined by the
West fork at about 48°30’ North latitude. The main stem is 455 river
miles in lengtr from the International border to the confluence iith
the Missouri River and drops 450 feet over that distance (L. Brown,
personal communications). Estimates of perennial channel length are
shown in Table Ll.4_1 with drainaae areas for the four major sub-basins.
The lengths of perennial channel for the West, Middle, and East Poplar
rivers were calculated based on a sinuosity/projected length ratio for
the lower Poplar. Therefore, they may tend to be high because of in-
creased meandering of the river as the volume/gradient ratio increases.
The mean flow of the Poplar River at the basin outlet is 115.8 cfs.
The mean flows are compared to 1975 flows for the river outlet and
border stations in Table 4.4-2. The outflow hydrograph for the Poplar
River at Poplar, Montana is shown in Figure 4.4-2. The flow in the
water year shown is a one in 8.33 year event, somewhat above the ex-
pected value perhaps due to the strong peak in late June and early July
caused by relatively heavy precipitation during that period. The shape
of the hydrociraph is typical however, showing the strong spring peak
flow and low winter flow.
The Poplar River Basin has mostly low gradient, meanderina
alluvial streams. The alluvium through which the streams flow is two
to six feet thick and is underlain by a deep glacial till. Therefore,
the banks are constructed of silts and sands, while the streams have
cut into the glacial till making the beds a combination of unconsoli-
dated gravels, cobbles and loose sand. A schematic of a typical reach
of the Poplar River is shown in Figure 4.4-3.
A detailed discussion of surface water hydrology, including annual
flow frequencies and sediment transport characteristics, is included in
Appendix A-3.
4.4.2 Ground Water
There are three major aquifers in the Canadian part of the basin -
the glacial drift including the Empress Formation, the Ravenscrag For-
mation, and the Frenchman Formation. Ground water is used for domestic
and stock purposes. A description of the aquifers and well yields in
the U.S. and Canadian parts of the basin is included in Appendix A-4..
The general ground water flow regime in the Canadian part of the
Poplar River Basin is shown in Figure 4.4-4. The upper part of the
basin flows into Fife Lake. In dry or normal years there is no over --
flow from Fife Lake into Girard Creek. The only loss is from evapo-
ration. In wet years ther-e nay be overflows as occurred in 1952-53
and 1975-76. The ground water in the lower Girard Creek sub-basin
flous southeastt.iard into the East Fork of the Poplar River. There is
another loss to the basin in the East Fork sub-basin ‘ ihere the ground
37

-------
Table 4.4-1
PERENNIAL STREAM LENGTH AND DRAINAGE BASIN AREAS
FOR THE POPLAR RIVER BASIN
Basin Segment
Length of
Perennial Stream
(River-miles)
0
Major
rder
of
Drainage
Area*
Drained
(mi 2 )
Drainage
Density
(mi l
West Poplar
390
2
1009
.39
Middle Poplar
2â4
2
(582)
.34
East Poplar
212
2
(485)
.32
Poplar
341
3
926
.37
TOTAL
1187
-
(3002)
.36
*Numbers in parenthesis indicate drainage areas without Fife Lake
non-contributing area.
Table 4.4-2
COMPARISON OF EXPECTED ANNUAL FLOWS, MEAN FLOWS AND
THE 1975 FLOWS IN THE POPLAR RIVER
River
Segment
Expected
Annual Flow
(ac-ft)
Return
Period
(yrs)
Mean
Flow*
(cfs)
1975 Flow
(ac-ft)
West Fork @
I.B.
3,445
3.0
4.76
9,200
Middle Fork @
I.B.
11,790
2.7
16.28
34,040
East Fork @
I .B.
11,850
2.9
16.37
34,040
Poplar River @
Poplar, Montana
83,860
3.0
115.83
323,000
*Period of Record: 1933-1974.
38

-------
I I I I
I I
I I I I I I I I
0 N D J F M A M J J A S
1975
1976
Figure 4.4-2 OUTFLO’J H’(DROGRAPH FOR THE POPLAR RI’JER
NEAR POPL, R, MONTANA, OCTOBER, 1975,
TO SEPTEMBER, 1976
100000
10000 -
1000 —
100 —
10 —
0
39

-------
/
b
ELEVATION
b-b
Figure 4.4-3
SCHEMATIC OF TYPICAL REACH OF THE POPLAR RIVER
/
/
/
/
/
/
/
a
a
I
I
a’
a’
ELEVATION
a—a
\
BANKFULL
STR EAM
WIDTH
Xb’
\
N
N
\
\
IFFLE
b’
POINT
BAR
7
/
/
/
/
40

-------
0 5
5 0 5
MILES
K I LOM E Tfl ES
Groundwater Flow
Direction
Figure 4.4-4
GROUNDWATER FLOW REGIME IN CANADIAN PART OF POPLAR RIVER BASIN
OPEN PIT COAL MINE
(AREA LIMIT FOR
FIRST 15 YEARS)
LEGEND
I)rainage Basin Boundary
SuI Basin Bound .iiy
5
106000.
Saskac liewan
Montatia
after Saskmont
Engineerirlg, 1978

-------
water flows northward into a series of lakes including Bonneau, Rivard
and Montague. The lower East Fork sub-basin flows south toward the
border. The East Fork area just north of the International border is
a ground water discharge area.
The major aquifers in the U.S. part of the basin are the Fox
Hills-Hell Creek Formations, Fort Union Formation, Flaxville Formation,
Wiota gravels, glacial deposits, and alluvium. The wells supply water
for stock and domestic purposes but few wells have high enough yields
for large-scale irrigation. The flow regime in the U.S. part of the
basin is similar to the topography. Feltis (1978) conducted a detailed
study in the East Fork sub-basin from the Canadian border south to the
northern boundary of the Fort Peck Indian Reservation. A generalized
ground water contour map for the shallow unconfined aquifers is shown
in Figure 4.4-5. A detailed ground water level contour map also show-
ma the chemical composition is included in Appendix A-4. The direc-
tion of flow in the recharge areas at higher elevations is into the
upper aquifers--the glacial outwash and Flaxville Formation. The
ground water then flows into the Fort Union Formation and partly into
the streams and deeper formations. Ground water may flow down the
river valley for several miles before discharging into a section of
the Poplar River (Feltus, 1978). The detailed flow regime in the
southern part of the Poplar River was not available. The expected
flow directions are from the basin divides toward the river and then
down to the Missouri River.
Primary recharge areas in the U.S. part of the basin are the
plateaus and terraces along the sub-basin divides. Most of the re-
charge is from precipitation within the basin with additional inflow
across the International border. Irrigation along tributaries and in
the floodplain provides some recharge to the shallow ground water
aquifers. At the time of spring runoff (usually March or April) the
Poplar River may recharge the Quaternary alluvium. During other per-
iods the stored water would return to the river.
4.5 WATER QUALITY
4.5.1 Surface Water
Surface water quality data for the baseline year 1975 as requested
by the EPA are compared to the appropriate water quality standards and
criteria. The sampling stations are located in Figure i.5-1. A statis-
tical summary of the data is included in Appendix A-S.
Comparing sites on the Canadian East Poplar River (Cl and C6) ,
water quality appears to deteriorate substantially movina downstream.
In general, concentrations of the following water quality constituents
increase (based upon comDarisons in Table 4.5-1):

-------
- _- -
43
Ground water flow
direction
Non-contributing areas
for surface water
— Sub-basin boundaries
I
I
$
I
. 5 .-
‘I,
5-
>
5-
5.—
0
4-
0
S .-
C)
E
a)
C
C)
43
-C
0
S .-
S.-
a)
C .)
U
C)
4
4
I
I
,f
/
3
\ ____
I
Key

-------
Figure 4.5—1
LOCATION OF WATER QUALITY SAMPLING STATIONS
In
I
I

-------
Table 4.5-1
WATER QUALITY STATISTICS FOR STREAM SAMPLING LOCATIONS IN CANADA
9 2 .120 .
.00 lIE
I0(sl0
Sampling Dates
5 . 1/ 0 4 7 , 2929)
Sb.tiitbc,
toM b 102.2 Cood.o-
022.2 *11.1 12vl2 turbid - Sulfite Cob M ig Potts. lobe? lot.)
E u /b 5 ,9/b j.t.11/ ity Color 409 sq/a nJ. 0191u. C.luI. n..l,a Oudlie slum i2.,dn.s. 0 P0 4 1,1.1 P tltt.I. to tOll
721 C.20 2 C a L l 2 1a .229 9PttA P0/ b 90 P4/b .9/9 mg/I soJ 9 sq/I Sq/ I .9/I . 9/I sq/b mg/I ag/I ‘o/°
9u llnq p, .iIlon Cl ,
24 19 P.pIsr 42.,r,
2. 9 1 02 Co,onaco
Sn nlfl—, 414
I/li I/ I 2/’O
2/29, 0/20. 0/24 ,
9/IS, 9/00 9/22
Si/I 9/I l b /Il
I/ I nit 0/29
SI) 291/I 20/22
Il /lu i l/Il
a
la b
nno
92..
oO 00 20 20 9 20 20 20 20 20 20 20 20 20 20 20 20 20 10
110 04 0 292 9(0 0 0 ito 400 290 9 20 22 0 9 90 20 0 6 20 99 0 0 0 00 00 0 0 I I 0
992 990 202 20 0 204 292 0 42 90 0 22 0 92 2 920 29 2 392 0222 0220 202 401 22 2
003 909 2990 229 190 999 209 24 0 292 20/ 990 299 200 991 02,9 290 91/3 2 091 9.11 0
204 109 29 0 22 9 290 62 0 2 29 2 92 29 2 22 2 99 0 2 42 99 9 0149 0994 290 / 04 Ii I
9..plloj (.1 ,191.10 C O
l..i Poplin Il..r .9
l.Inu.Il..n,i 9.1,nd. ’
2/24 2/72, 2/20.
7/ ,0 2/20 2/24
4/IS 4/20 9/22
‘2 0/10 C/lI
1/2 2/11 21/9
0119 9/ I, 20/I
il./00 Il/l b 22/22
s
# 1 ,
P .. .
9..
5
72 9 22 22 22 20 22 22 22 72 22 22 22 22 22 22 22 22 22 22
2 40 240 209 230 490 00 290 ES 0 0020 240 09 0 200 22 0 99 204 00 020 00 02 202
35 4 022 2290 209 22 0 2020 0/2 079 066 92 0 96 2 249 22 0 020 0920 0942 2 51 511 1’ 0
0 92 040 122 2020 29 0 900 iro too 209 22 0 900 620 9?) 04 0 444 290 29’ I 20 I 2/ 94 0
29 9 240 929 2) 9 2/ 2 202 Li 9 00 4 22 22 0 20 0 220 299 24 0 09.47 02/2 92.92 it) It 9

-------
e Total alkalinity . Sodium
o Conductance 0 Potassium
• Color o Orthophospnate
• Total dissolved solids a Total phosphorus
• Sulfate • Total nitrate
• Chloride 0 Ammonia
o Silica a COD
• Macnesium
These parameters probably exhibit an increase because of the influx of
Girard Creek water, which includes Fife Lake discharge. Since Fife
Lake discharges only intermittently, the waters are high in dissolved
solids due to evaporative concentration. Only turbidity, calcium, and
total hardness do not increase or decrease slightly.
Mercury values measured in walleye tissue from the Cookson Reser-
voir were up to 1.5 ppm which exceeds the Canadian standard of 0.5 ppm.
Mercury levels measured in walleye tissue from the East Fork on the
U.S. side were up to 0.9 ppm which is close to the new FDA standard of
1 ppm (U.S. EPA, personal communication, 1980). Limited water column
measurements of mercury indicate that mercury is present in all the
forks and that the accumulation rate in the fish is high. Mercuric
acetone is widely used as a fungicide for treatment of wheat seeds
throughout the basin. More work is currently being done by the USGS
and EPA on this problem.
The East Fork below the International boundary had total dissolved
solids concentrations between 1050 and 1750 mg/Q in 1975. Sodium con-
centrations are fairly high with a mean up to 486 mg/i. Boron concen-
trations ranged between 2.5 and 3.7 mg/i in the summer of 1975 at the
border and 1.5 to 3.2 mg/Q at station G, close to the confluence with
the Middle Fork. Dissolved oxygen varied from 4.4 to 12 mg/i suggesting
that levels may be low for fish during the winter. As elsewhere in the
Poplar River system, the water is very hard, calcium-magnesium hardness
typically being in the 300-400 mg/i range (as CaCO 3 ).
Sulfate levels in the water are high (mean of 306 mg/i) as are both
nitrogen (mean of 1.46 mg/i) and phosphorus (mean of 0.09, maximum of
0.4 mg/i).
Data for the West Fork are limited to three stations in the summer
of 1975. Water quality appears relatively uniform, especially with
respect to dissolved solids concentration which is low (mean of 700-800
mci/i TDS) relative to the East Fork. Dissolved oxygen appears to de-
crease slightly from upstream to down, but is high in all samples.
Boron levels rance from about one-half to slightly over 1 ppm.
46

-------
Data for the Middle Fork are also limited. Dissolved solids in
the Middle Fork are high (maxima slightly above 1000 mg/P.. TDS). Sus-
pended solids, are similarly hicih (up to 76 mg/i SS) and may, at
times, be stressful to aquatic biota. Boron levels are uniformly high
(observed values of 0.36 to 2.0 ma/i) in the Middle Fork, and the water
is likely to be toxic to sensitive crops grown under irrigation. The
water is also hard with a range of 200-300 ppm of calcium-magnesium
hardness. Dissolved oxygen concentrations are quite variable with a
range of 4.6 to 8.2 mg/i at the border station.
Water quality data in the mainstem of the Poplar River show mean
TDS concentrations above 1000 mg/Q. Sodium concentrations are high
with means of 300 mg/i. Calcium and magnesium levels are similar to
those in the other forks, and the water is classified as very hard.
Boron levels are high, observations as high as 3.4 ppm having been
made at station PR-4. Apparently, however, boron levels decrease down-
stream as the water approaches the confluence with the Missouri River.
Phosphorus and nitrogen levels (based upon only three observations at
USGS station 06181000) appear significantly lower in this part of the
system than elsewhere. The dissolved oxygen concentrations are about
5 mg/i.
Federal Drinking Water Quality Standards . Available data for
1975 show that the only Federal primary standard for drinking water
quality contravened was lead on the main stem and the East Fork. How-
ever, since the concentration values were reported as being “less than
100 ug/Z,” there may, in fact, have been no contraventions at all.
The secondary standards for iron, manganese, pH, and TDS were contra-
vened in at least one observation on all four branches of the river.
Sulfate concentrations exceeded the secondary drinking water standard
on the Middle and East Forks and on the main stem. There were insuffi-
cient data to evaluate barium, silver, chlorinated hydrocarbons, tur-
bidity (because of units) and coliforms (all primary contaminants) and
methylene blue active substances, hydrogen sulfide, and odor (secondary
contaminants).
Federal Water Quality Criteria . Table 4.5-2 shows criteria and the
locations where one or more contraventions occurred in 1975. The criteria
for boron and iron are contravened in all four branches of the river. In
all but the main stem, mercury values have apparently exceeded 0.05 iig/i,
and in the East Fork, manganese has exceeded 0.1 mg/i. With respect to
barium, fecal coliforms, color, cyanide, and nickel, data were inadequate
to evaluate for criteria contraventions. In all other cases, existing
data for 1975 did not show levels to be in excess of criteria.
State Water Quality Standards . Montana State water quality standards
for waters of B-D 2 classification (including the Poplar River Basin) are,
with the exception of toxic substances and fecal coliforms, expressed in
terms of anthropogenic increases above, or changes in, natural conditions.
Standards cover pH, dissolved oxygen, sediment, turbidity, and color.
47

-------
Table d.52
U.S. EPA WATER QUALITY CRITERIA CONTR. VE TIONS O I
THE POPLAR RIVER, 1975
Branch of River 3
Parameter Standard ’ F West Fork Middle Fork East Fork Mainstem
Alkalinity as CaCO 3 20
Ammonia .O2
Arsenic .10
Barium 1.0 X
Beryllium ii ug/Q
Boron .75 o a
Chromium .10
Fecal coliforms 6 X
75 Pt-Co
Color units
Cyanide .005 X
Iron 1.0 9 0
Manganese .10 a
Mercury .05 ig/Z’ a
NickeF x
Dissolved Oxygen 5 9 a
pH 5-9 units a a a
Hydrogen Sulfide .002
Notes ‘In mg/ except as noted.
ifl this column means inadequate data to evaluate.
‘“a” in these columns means one or more kno ’ in contraventions.
Un onized ammonia, a function of pH ann temperature
5 Most stringent, for protection of aquatic life in soft, fresh .sater .
°Depends upon method of assaying.
7 For protection of fresft.iater aquatic life and iildlife.
Defined in terms of the LC5O for aquatic life.

-------
Toxic substances are considered in the same way, i.e., with respect to
anthropogenic increases, but also with regard to Federal drinking water
standards. These were discussed earlier. Fecal coliforms are consid-
ered in terms of proportions of samples having more than some number of
organisms per unit volume during a set time period. The available data
are inadequate to evaluate coliforms contraventions.
4.5.2 Ground Water fluality
Ground water quality data were available for 65 wells in Canada
and 20 in the U.S. Detailed water quality data tables are included in
Appendix A-S. The following section describes the general chemical
water types found in the basin. The water in all formations except
the Flaxville Formation is alkaline and high in total dissolved solids
(TDS). The data show a sodium bicarbonate type water in the alluvium
and Fox Hills-Hell Creek Formation although the latter may contain
some carbonate. The Fort Union Formation may have a sodium or magne-
sium bicarbonate water. One sample from the Flaxville Formation was
lower in dissolved solids with more calcium than magnesium or sodium.
The available data for the Canadian wells show a similar pattern to
the U.S. wells with the glacial drift water a calcium and magnesium
bicarbonate type. The Ravenscrag Formation samples were a sodium
bicarbonate-sulfate type water with very low calcium and magnesium.
One of the major uses of ground water in the basin is for domestic
water supplies. The available data were compared to the U.S. EPA pri-
mary and secondary drinking water standards. Figures 4.5-2 and 4.5-3
show the range of water quality data for selected parameters and the
standards. A detailed description of the standards is included in
Appendix A-S.
Ground water is used for stock watering in approximately 72 per-
cent of the wells and springs in Daniels and Roosevelt counties
(Kiarich, 1978). Water quality guidelines for livestock from the EPA,
Montana, and the California State Water Quality Control Board were
compared with the available data. The Fox Hills-Hell Creek Formation
samples exceeded the guidelines of 1 mg/2 Fluoride and 500 mg/2. HCO 3 .
Water samples from the Fort Union Formation and Quaternary alluvium
also contained HCO3 above 500 rng/Q. One sample from a well completed
in the Fort Union Formation had sulfate levels above the threshold
value of 500 mg/Z. One sample from a well completed in the Fort Union
Formation had sulfate levels above the threshold value of 500 mg/9. but
belo i the limiting value of 1000 mg/2 . Guidelines were met for the
following selected cations and heavy metals:
49

-------
FORT L4ON
Fo . I1 L!s I I Cl Il Fn,
1 10 1000 10000
TOO — —
S —
flLC* I’.ITV
“I , .
0*00 1 . 1 5 0 —
C I M 1 1
100 .N
W I _____,
S
C ’
C. —
———-—C————
N. _____ _____
0
• I -
S
I , _•____
—“j r
* 1
Pb
.4 , 0
Cd
‘4” —--—- ———--
C.
. 4 ( 0
5.
‘ 4, .
LI
C-
4” S
— R .GE
MEAN
US F’ IARY DO .Cl 1 .Q
WATER ST .\OL005
— FORT UT.I( lTI FORVAT O ’
FOb HILLS ‘(t CR 500 FC’°S1AT ,C
Data from Feltis, 1978 and
U.S. EPA. 1977
Figure ‘1.5-2 RANGES OF SELECTED CHEMICAL PARAMETERS IN WATER
SAMPLES FROM THE FORT UNION FORMATION AND
FOX HILLS-HELLS CREEK FORMATION IN U.S. PART
OF THE POPLAR RIVER BASIN
50

-------
CUiTER MRY . LLUVlU’.
Glacial Oorwasn inc FIan,llt Fc,r,at.on
0 IC O
• —3—-
—4,, = U I
•—-o-
£LkALI&Itv c.— I
IC. MV
:————--
C—.
Cl —
“‘4,’
- -
-V. - - -,
‘.4
-0
F. • >— —
I ” , ” U

4, ’ ___Q____.
. 4 ,
U
—-—3.—
“4
Pb -
i’ ll
Cd
A,
—4
9.
-3
S. I
— ‘ —————4
EEl
— RA ’ aGE
• US P I . 4RY DRI I ’G
5d 9TER SI. AI’)S
— FORT U%ION FO allaN
FOX “ILLS HELL CPEEE FCPMATION
__________ - _j
Data from Feltis, 1978 and
U.S. EPA, 1977
FiQure 4.5-3 RANGES OF SELECTED CHEMICAL PARAMETERS IN
WATER SAMPLES FROM QUATERNARY ALLUVIUM,
GLACIAL OUTWASH AND FLAXVILLE FORMATION
IN THE U.S. PART OF THE POPLAR RIVER BASIN
51

-------
• Calcium o Copper
• Magnesium • Fluoride
o Sodium • Lead
• Arsenic a Mercury
o Aluminum • Nitrate + Nitrite
• Boron • Nitrite
o Cadmium o Selenium
o Chromium o Vanadium
• Cobalt o Chloride
The TDS concentrations in the ground water would classify the water as
good (less than 2,500 mg/Z) by Montana’s salinity classification.
Water from some of the wells completed in the Quaternary and Fort Union
Formations would be classified as excellent (less than 1,000 mg/2 )
under the EPA’s system (EPA, 1975).
Small numbers of wells are used for irrigation. The ground water
in the basin would be assigned to Class II of McKee and Wolf (1963) on
the basis of the high conductance, TDS, sulfate, and sodium concentra—
tions. Class II ground water has a medium to very high salinity hazard
and is suitable for irrigating tolerant to semi-tolerant crops includ-
ing barley, hay, alfalfa, and wheat. A recent study by Klages (1976)
investigated the irrigation potential and hazards of irrigation in the
Poplar River Basin. This study recommended a limit of 1,500 imhos/cm
for conductance (at 25°C) and SAR values (sodium adsorption ratio)
less than 7.5 to 8. Water from the glacial outwash deposits, the
Flaxville Formation and a few wells in the Quaternary alluvium and Fort
Union Formations could be used for irrigation based on these criteria.
Boron levels in the glacial outwash and Flaxville Formation are less
than 0.67 mg/2 so the water could be used for sensitive crops. Boron
concentrations in the Quaternary alluvium and Fort Union Formation are
mostly between 1.4 and 2.4 mg/9 . which would be suitable for semi-
tolerant or tolerant crops. Other elements which exceeded guidelines
in at least one well were manganese, fluoride, molybdenum, iron, and
selenium.
In the Canadian part of the basin most wells are used for domestic
water supplies and stock-watering. The range of chemical parameters
and the Canadian and Saskatchewan standards for drinking ‘.iater, stock
use and irrigation are discussed in detail in Append x A-5
52

-------
4.6 WATER USE
d.6.1 Municioal Use
In the U.S. part of the basin the only municipal water use is by
the City of Scobey, Montana. The water is pumped from shallow wells
adjacent to the Poplar River at a current capacity of about 750 gpm
(3.31 acre-feet/day). The wells are considered to be recharged from
the Poplar River. During 1975, municipal water use by Scobey was 210
gal/person/day, or a total of 0.95 acre-feet-day (Montana DNRC, 1978).
This represents an annual use of about 347 acre-feet.
In Canada the only municipal water use is a shallow well located
adjacent to Coronach Reservoir serving the Village of Coronach, Sas-
katchewan. In recent years the annual water use has ranged from 20 to
dO (average 36) acre-feet. It is assumed that almost all of the use
represents a surface water depletion (Poplar River Task Force, 1976).
4.6.2 Industrial Use
There is currently no industrial use of water in the U.S. or
Canadian parts of the Poplar River Basin. The Farmer’s Potash Company
has applied for a permit in Montana but this has been delayed pending
outcome of the apportionment.
4.6.3 Agricultural Use
Consumptive water use for agricultural purposes is associated
with stock watering and irrigation. Water consunption for stock
watering activities results from two separate actions: evaporation
from stock ‘.‘iater holding reservoirs and actual consumption by livestock.
In the U.S. part of the basin, information on water use projects
for agricultural purposes has been estimated based on water resource
surveys, field inspections and aerial photographs. In the Canadian
basin data were also available from project owner’s records of pumping
rates. This information was then used in conjunction with assumptions
on crop requirements, soil characteristics and water availability to
estimate total water use. The assumptions used for both U.S. and
Canadian parts of the basin are listed in the Poplar River Task Force
Report (1976).
4.6.3.1 Montana Water Use
For the U.S. part of the basin outside of the Fort Peck Indian
Reservation, the 1975 water use for livestock was 432 acre-feet for
stock watering and 452 acre-feet due to reservoir evaporation (Montana
DNRC, 1978).
53

-------
The historical use of water for irrigation by gravity/pump diver-
sion is variable depending upon river flow (Figure 4.6-1). Since 1960
there has been a large increase in irrigation usage which has generally
fluctuated between 2,500 and 5,000 acre-feet per year in the U.S. part
of the basin.
In 1975, a total of 5,900 acre-feet was used for irrigation by
gravity and pumping and 3,076 acre-feet by spreader irrigation in the
U.S. basin exclusive of the Indian Reservation. This relatively high
water usage reflects the correspondingly high availability of water
during 1975, which allowed for a total of four applications. During
the last 45 years there have been only three years, however, in which
water availability in the basin allowed for four applications. During
the period of record, the average number of applications was 2.4
(Dooley, 1975). A breakdown by sub-basin of irrigation water usage in
the non-Indian Reservation U.S. part of the basin is shown in Appendix
A-5.
4.6.3.2 Water Use on the Fort Peck Indian Reservation
On the Fort Peck Indian Reservation all existing (1975) water uses
are associated with agricultural activities. A detailed breakdown by
sub-basin is shown in Appendix A-5. In a survey conducted by the Fort
Peck Sioux and Assiniboine Tribes and Morrison-Maierle, Inc. (1978), a
total of 283 stock ponds were identified on the reservation. The total
water uses for livestock consumption and evaporation were 287 acre-
feet and 600 acre-feet, respectively.
A total of 12 existing irrigation projects were also identified.
Of these, four projects used gravity or pump diversion systems to irri-
gate alfalfa, native hay, truck gardens or barley. The total acreage
used for gravity/pump systems in 1975 was 306. Estimates were also
available for the actual water use in irrigation projects during 1975.
Since 1975 was a high runoff year, four irrigation applications were
conducted. Based on a total application of about 31 inches, the esti-
mated use was 785 acre-feet. All of the 1975 use associated with
gravity/pump irrigations was from two sub-basins of the Mainstem of
the Poplar River (boundary to West Fork and Hay Creek to Missouri River).
An additional area of 299 acres on the Indian Reservation was irrigated
by eight spreader dike systems during 1975. Based on a single 10-inch
aoplication, the total water use was estimated at 299 acre-feet. Crops
irrigated by spreader dikes were alfalfa and native grasses.
During the period from 1955 to 1975, consumptive water uses on the
Indian Reservation for stock watering and spreader irrigation have re-
mained relatively constant (Figure 4.6-2). The annual water use for
gravity/pump irrigation has fluctuated from about 300 to over 1,500
acre-feet. During this period the total area irrigated by gravity/pump
54

-------
5.0
4.0
3.0
2.0
1.0
0
1955
Ficiure 4.6-1
1960 1965 1970 1975
YEAR
HISTORICAL WATER USE IN THE U.S. PART
POPLAR RIVER BASIN 1955 THROUGH 1974.
from Poplar River Task Force, 1976)
OF THE
(Data
C
I-
x
U-
U
LU
C l )
LU
I.-.
55

-------
2000
1500 —
a,
C.,
LU 1000
‘I )
LU
I . .-
500 —
I I I I I I I I I I I I
I I I I I I I I I
1955 1960 1965 1970 1975
YEAR
Figure 4.6-2
HISTORICAL WATER USES ON THE FORT PECK
INDIAN RESERVATION 1955 THROUGH 1965.
(Data from Morrison-Maierle, Inc. 1978)
Stock Consumption . & a .
Stock Pond Evaporation •
Gravity/pump •
Spreader .
0_
A
56

-------
systems varied from 236 to 786 acres. Much of the fluctuation in water
use was due, however, to water availability and subsequent variations
in the total number of irrigations per season.
Since 1975 was a high runoff year, the actual irrigation uses are
probably not representative of the average existing water use at the
1975 level of development. Existing water use for gravity/pump irriga-
tion is more appropriately defined by using 1975 levels of development
(i.e., irrigated acreage) and average runoff values. Table 4.6-1 pre-
sents estimates of existing water use in the U.S. part of the basin
assuming 2.4 irrigations per year and 0.641 feet per application
(Poplar River Task Force, 1976). The use of these assumptions modifies
only the gravity/pump irrigation uses, since stock uses and spreader
dikes are less dependent upon the amount of runoff.
4.6.3.3 Canadian Water Use
In the Canadian part of the basin water use for agricultural pur-
poses is associated with stock watering (consumption and pond evapora-
tion), irrigation, and evaporation from several large reservoirs.
Agricultural uses for stock watering and irrigation under exist-
ing conditions are 182 and 319 acre-feet, respectively. The existing
uses for each sub-basin in Saskatchewan are given in Appendix A-5.
With an additional 620 acre-feet use due to evaporation at three reser-
voirs (Clarke Bridge, Coronach and West Poplar), the total agricultural
water use is 1,523 acre-feet.
4.7 VEGETATION AND WILDLIFE
The Poplar River Basin is dominated by cropland (wheat, barley
and alfalfa) and grassland areas. The grasslands are comprised of a
variety of grasses including: crested wheatgrass, yarrow, blue gramma
and bluegrasses. Although coulees and breaks may contain areas of
shrubs (e.g., rose, silver sage and chokecherry), there are very few
stands of deciduous trees. The existing aspen poplars and green ash
are confined primarily to the banks of the Lower Poplar River near its
confluence with the Missouri River.
The Poplar River Basin provides important habitat for upland game-
birds, big game and waterfowl. Major upland gamebird species include
the ring-necked pheasant, Hungarian partridge and sharp-tailed grouse.
White-tailed deer is the main big game species; however, lower densi-
ties of mule deer and pronghorn antelope are also present. The occur-
rence of rare and endangered species is discussed in Appendix A-6.
57

-------
Table 4.6-1
ESTIMATES OF EXISTING WATER USE FOR
GRAVITY/PUMP IRRIGATION IN THE U.S. POPLAR RIVER BASIN
Estimated
Acres Water Use*
Sub-basin Irrigated ( acre-feet )
International Boundary to
Fort Peck I.R .
East Fork 65 100
Middle Fork 1,269 1,950
West Fork 389 598
Poplar River (main stem) 976 1,500
Maternach Coulee 137 211
Sub-Total 2,836 4,359
Fort Peck I.R. to
Missouri River
Poplar River to West Fork 250 384
Poplar River-West Fork
to Missouri River 56 71
Sub-Total 306 470
TOTAL 3,142 4,829
*Based on 2.4 applications per year and 0.641 ft of water per
application.
Acreage is partly irriqated by small tributaries to the East Fork.
The only diversions are near the confluence with Middle Fork.
Source: Poplar River Task Force (1976)
-J

-------
The aquatic habitat provided by the Poplar River is important for
significant breeding populations of waterfowl such as mallards and
American wigeons. The potential use of the river by waterfowl is an
important consideration since flow modifications due to apportionment
could result in changes in available nesting and rearing habitat.
Studies of waterfowl utilization of the Poplar River have been con-
ducted by DeSimone (1979) and are discussed fully in Appendix A-6.
Also included in the Appendix are detailed discussions of the vege-
tation types and terrestrial wildlife studies conducted in the study
area.
4.8 AOUATIC BIOTA AND FISHERIES
The Poplar River supports relatively abundant game fish popula-
tions. In this regard it is distinct from most northern plains streams
which are not generally considered as game fish habitat. The primary
game species in the Poplar River Basin is walleye; however, northern
pike are also common in most of the area. Smailmouth bass have been
stocked and are apparently reproducing successfully in the main river
near its confluence with the West Fork. Goldeye are also utilized by
anglers; however, they are primarily restricted to the lower river.
Although the Poplar River supplies good fishing quality throughout
much of the drainage basin, the actual angler utilization is low.
Montana Department of Fish and Game has estimated that the entire
Poplar River received a fishing pressure of only 2660 angler-days per
year (State of Montana, DNRC, 1978). This represents an average usaae
of only 7.3 anglers per day of fishing throughout the year. The low
fishing pressure is also reflected in the local Daniels County fishing
license sales, which have averaged only 349 per year (1966-1976).
Due to the low population density, lack of projected population
increases (see Section 4.10) and the proximity of popular fishing areas
(e.g., Fort Peck Reservoir), it is anticipated that the Poplar River
will continue to be utilized by anglers at a low level of fishing in-
tensity. It is also expected that the usage will be primarily by local
anglers.
Studies conducted by Montana Department of Fish and Game (1978)
reveal that gamefish spawn throughout much of the U.S. Poplar River.
Both walleye and northern pike require distinct spawning habitats in
the Poplar River Basin. Walleye spawn over gravel bottoms in shallow
water riffle areas while pike commonly use streamside vegetation as a
spawninG substrate. Both species spawn during high flow conditions
following the spring ice breakup. Therefore, modifications of the
natural flow regimes due to the apportionment may potentially impact
gamefish spawning success and subsequent recruitment.
59

-------
Studies of periphyton, macrobenthos and fish of the Poplar River
have been conducted by Bahis (1977), Montana Department of Fish and
Game (1976 and 1978), Saskmont Engineering (1978) and the IJC, 1979
(DeSimone, 1980). The results of these studies provide for a general
discussion of the aquatic biology of the river which is included as
Appendix A-7.
4.9 METEOROLOGY AND AIR QUALITY
4.9.1 Meteorology
The climate of the Poplar River Basin is controlled largely by
high pressure areas which move into the area from the north during the
winter and from the west during the summer. The major tracks of low
pressure centers pass either to the north or south of the area
throughout the year. The area is in the rain shadow of the Rocky
Mountains, resulting in a semiarid climate with a large annual range
in temperature. Cold winters and warm summers are characteristics of
the regional climate. The average monthly temperature at Scobey is
above freezing between April and October. Detailed temperature data
are presented in Appendix A-8.
The precipitation at Scobey ranged from a low of 7 inches to a
high of 21.9 inches for the period 1940 to 1978. The average precipi-
tation at Scobey from 1941 to 1970 is 13.6 inches. Monthly precipita-
tion for Scobey is shown in Figure 4.9-1. While the normal annual
precipitation for the area is only 11 to 14 inches, 76 percent of it
falls from April through September, with May and June accounting for
34 percent of the annual total. Winter precipitation nearly always
falls as snow; but as a rule, although snow seldom accumulates to any
great depth, it usually is formed into drifts in the open, unprotected
areas. Accumulated winter snow remains on the ground until about March.
Wind plays an important role in the dispersion and dilution of
pollutants emitted into the atmosphere. Pollutant concentrations are
inversely proportional to wind speeds, i.e., the stronger the wind the
lower the pollutant concentrations. The Poplar River Basin is charac-
terized by relatively high winds and few periods of calm. Mean monthly
wind speeds at Scobey typically range from 8 to 13 knots with calm
winds occurring from 0.5 to 6 percent of the time on an annual basis.
The higher wind speeds occur in the autumn and spring months from the
northwest. The prevailing winds at Scobey are from the northwest or
southeast. Wind rose plots are included in Appendix A-8 for Scobey
and Glasgow, Montana.
60

-------
Ficiure 4.9-1 Normal Monthly Precipitation at Scobey, Montana
61
5
4
3
2
0
U)
-c
U
C
a
C
0
0
0
U
0)
3-
Months
J A
N

-------
4.9.2 Existing Air Quality
The existing air quality in the study and impact area is very
good. Recent measurements taken by the Montana Air Quality Bureau,
during 1979, in northeastern Montana have shown very low concentrations
of S02, NO 2 and particulates (Geihaus, 1977, 1979). Since there are
no major sources of pollutant emissions in northeastern Montana, these
results might have been anticipated. A comparison of these values with
the National and Montana ambient air quality standards shows that S02,
N02 and suspended particulates concentrations are well below standards.
A further discussion of existing air quality and air quality standards
in included in Appendix A-8.
4.10 SOCIAL AND ECONOMIC PROFILES
4.10.1 Population Profile
The population and average annual percentage change in population
between 1970 and 1975 are shown for Daniels and Roosevelt counties in
Table 4.10-1. Population has changed very little in Daniels County,
remaining stable at about 3,100 throughout the five-year period, 1970-
1975. Slight variations in population shown in Table 4.10-1 were due
to changes in estimating procedure rather than actual changes in popu-
lation. The size of the population also remained stable in Roosevelt
County. There were approximately 10,300 people living in the county
between 1970 and 1975 (see Table 4.10-1). The occasional fluctuations
are due either to estimating techniques or to actual small population
changes. Detailed population profiles by age, sex, and race are in-
cluded in Appendix A-9.
The total population on the Fort Peck Indian Reservation was 9,898
in 1970. Indians made up 34 percent of the population, or 3,406. The
number of Indians almost doubled in a three-year period to 6,202 in
March, 1973.
Small population increases are projected for Daniels and Roosevelt
counties. In Daniels County, between 1980 and the year 2000 an increase
of 300 persons is projected, increasing the population from 3,100 to
3,400. The increase will be largely due to the construction and opera-
tion of a proposed potash plant which should begin in roughly 1990. In
Roosevelt County, the population is expected to increase from 10,700 to
11,500 between 1980 and 2000. The increase in employment due to oil
and gas industries and exploration will increase the county population
and counteract the declining agricultural opportunities (Department of
Community Affairs, 1978).
62

-------
Table 4.10-1
POPULATION IN DANIELS AND ROOSEVELT COUNTIES, 1970-1975,
AND PERCENTAGE CHANGE, 1970-1975
Average Annual
Percent Change
County 1970 1971 1972 1973 1974 1975 1970 — 1975
Daniels 3,083’ 3,0002 3,1OO 3,100’ 3,200’ 3,1OO 0.1%
p p
Roosevent 10,365’ 10,4002 10,6O0 10,300’ 10,500’ 10,3OO -0.1%
Sources:
‘Bureau of the Census, Current Population Reports, Federal-State Cooperative Program
for Population Estimates, Series P-26, No. 109, May, 1975.
2 Bureau of the Census, Current Population Reports, Population Estimates and Projections,
Series P-25, No. 517, May, 1974.
3 Bureau of the Census, Current Population Reports, Federal-State Cooperative Program
for Population Estimates, Series P-26, No. 53, February, 1974.
1 ’Ibid, Series P-26, No. 76-26, July, 1977.
p = preliminary

-------
4.10.2 Archaeological and Historical Sites
There have been no archaeological sites identified in Daniels
County to date. However, the county is rated as having high future
archaeological potential. Seven archaeological sites of various types
have been identified within Roosevelt County. The county is also
rated as having a high future archaeological potential.
There are no nationally registered historical sites within
Daniels County. However, within Scobey there is a private historical
site, Pioneer Town. It is a representative pioneer homesteader town
of the early 1900’s that has been created by combining many authentic
early structures and objects from throughout the county.
The Fort Peck Agency and Fort Union Trading Post are national
registered historical landmarks within Roosevelt County. Fort Peck
was originally a fur-trading post, then an Indian Agency. Fort Union
Trading Post was an important early upper-Missouri fur trade depot.
There are also six private historical sites within the county,
four of which were trading posts. Fort Jackson and Fort Poplar are
located near Poplar. Fort Kipp and Fort Stewart are located west of
Culbertson. Disaster Bend and Snowden Bridge are the other private
historical sites (Figure 4.10-1).
4.10.3 Economic Profile
The size of the labor forces in Daniels and Roosevelt counties
declined between 1960 and 1970; however, both counties have experienced
increasing labor forces since 1972. The 1977 unemployment rates in
Daniels and Roosevelt counties were 2.9 percent and 5.1 percent, re-
spectively. Detailed information on employment rates, income and
business activities are included in Appendix A-9.
64

-------
Historic Places
0 National Park Service Areas
Location of Disaster Bend, Fort Jackson and Fort Poplar
Figure 4.10—1
LOCATION OF HISTORIC SITES IN THE POPLAR
RIVER BASIN AND ADJACENT AREAS
-- — Saskatchewan
asS - Sheridan County - - ci1oi tana
P,cneer TCwfl
Plentywood “.
Northern Boundary
I ForkPeck IR
Tt
Hills Medicine Lake
• National
Wildlife Refuge
i - Roosevelt County
Miles
o 4 8 12
o 6 12 18
Kilometers
Fort Kipp
Fort Stewart
cSrt ..1 CS
Trading st
65

-------
5. ENVIRONMENTAL CONSEQUENCES
5.1 AIR QUALITY IMPACTS
5.1.1 Air Quality Model
To estimate the projected air pollution concentrations in the
“impact area”, the Environmental Protection Agency air quality Single-
Source (CRSTER) Model was selected. This model was employed because
it was determined that the Poplar River Power Plant would be the only
significant contributor to the pollutant burden in the impact area
through the year 2000. This model is also suitable to the flat or
slightly rolling terrain which is found in the impact area.
The Single Source (CRSTER) Model is a steady-state Gaussian plume
dispersion model designed for point-source applications. It calcu-
lates pollutant concentrations for each hour of a year at 180 receptor
sites on a radial grid. The hourly concentrations are averaged to
obtain concentration estimates for time increments of specified length,
such as 3-hour, 8-hour, 24-hour and annual. The model contains the
concentration equations, the Pasquill-Gifford dispersion coefficients
and the Pasquill stability classes. Plume rise is calculated according
to Briggs. No depletion of the pollutant is considered. A complete
description of the model is presented in Appendix C.
5.1.2 Power Plant Emissions
The emission rates for the power plant were obtained from the U.S.
Environmental Protection Agency (EPA) Regional Office in Denver, Colo-
rado. The following emission rates were employed in the model.
Emission Rate for 600 MI.J Plant
Pollutant ( pounds/hour) ( g/seconci )
Sulfur Dioxide (SO 2 ) 10,732 1352.2
Particulate (TSP) 450 56.7
Oxides of Nitrogen (NO ) 3,600 453.6
The emission rates are for a 600 MW plant with a single stack.
The °2 emission rate was calculated on the basis of 1.94 pound S0 2 /10 6
Btu heat input and about 8 percent sulfur retention. The particulate
emissions employed 99.5 percent control and 0.08 pounds/1O° Btu. The
NO emissions utilized a 0.60 pound/10 6 Btu rate. The EPA new source
performance standards for 502, TSP, and NO < are 1.2, 0.10, and 0.70
pounds per i06 Btu, respectively.
66

-------
5.1.3 Model Input Parameters
The emission rates employed in the model are shown in the previous
section (5.1.2). The stack data that were employed for plume rise
height computation were obtained from the EPA Regional Office in Denver
and are as follows:
Stack Data for 600 MW Plant
• Number - 1
• Height, meters - 122.0
o Diameter, meters - 7.4
• Exit velocity, meters/sec - 24.4
o Exit temperature, °K - 424
The meteorological data employed were hourly surface observations
from Glasgow, Montana and twice daily mixing heights for the year 1964.
The year 1960 was also employed to determine if there were any signifi-
cant differences in peak concentrations of pollutants for the two years
since temperature and wind conditions were quite different during the
months of January and February. The mean temperatures for these months
were 10 to 15 degrees colder, and the frequency of calms was about seven
percent higher in 1960 than in 1964. The meteorological data were ob-
tained on magnetic tape from the National Climatic Center.in Asheville,
North Carolina. Recently, a continuous meteorological monitoring station
has been set up at Scobey. The air quality model will be rerun with the
new data for the EPA by the Montana Air Quality Bureau.
5.1.4 Modeling Results
5.1.4.1 Sulfur Dioxide (SO 2 )
For SO 2 , the model was employed to obtain maximum concentrations
and second highest concentrations in the impact area for averaging
times of 1 hour, 3 hours, and 24 hours. The annual concentrations
for the year were also obtained. Values were computed at 2 kilometer
(km) intervals along 10° radials out to a distance of 50 kilometers
from the plant site. Calculations were not made for greater distances,
because the basic assumptions incorporated into the model cause the
results to be suspect at greater distances. However, concentrations
at distances greater than 50 kilometers should become less as the
distance increases.
Spatial distributions of the highest maximum short term concen-
trations (1 hour, 3 hours, and 24 hours) and the 1964 annual concen-
trations are presented in Figures 5.1-1 through 5.1-4. The isopleths
67

-------
OCKGLEN ) AcH , ‘
/ - - - :
CANADA “ I
• _-1 UUIT ,k AT - - Saskatchewan ___ - , - -___
9 (1 6%) ç1
(1201 N I t.JLII I f’-, \ WHITETAIL OUTLOOk (
‘ ‘ - / ( /4 1 t( V ‘ . 4 RAYMOND
6 C. .. .. - / - ‘i,.
120) . I 80
L .‘ / i t t
RICKLANO 4 \I “ I REDSTONE
I, •FOUR ‘
60 1 OUTTE ______________
(20 ). 0 PEE (58 PlENTY WOOD
op 3 %‘• . .
I eo
U u (160)
ln 4I
\\
(120) \ 80
____ ___________________________________
DANIELS COUNTY (120) (170) DANIELS COUNTY SII(RIDA CO
ROOSEVELT COUN Y
—4a° I I MEDICINE LAKE
_______________ I? ISII )
I 1 La.,
/ lllPlLISL I III
(
0 I MIlPItINC (ARt NAIIOIIAL 1 1 01I 1 1 RUILL
FORT PECK JINDIAN RESERVATION
/ IIWS, )Ia4DL , ( I ROOS(VELT COUNTY
> o i ) 7 ‘Rob

Figure 5.1-1 SPATIAL DISTRIBUTION OF THE HIGHEST 1-HOUR SO 2 CONCENTRATIONS (pG/M 3 ) OBTAINED FROM THE
CRSTER MODEL FOR 1964, ASSUFIING A 600 MW (1200 MW) POPLAR RIVER POWER PLANT WITH ZERO
PERCENT EMISSION CONTROL

-------
R ESERVATIO
0 OUTLOOK
IIlO,AN AISIAVATIO, , SOUPOAP,
COOUU Oo .LMpY
SOILLTII
CONCCNTHA1IONS N i /n
N1LRM O ISOPLETH
Q , in ’i lAO
__ L __ - ———-- •A i --
,ll t. .
Figure 5.1-2
SPATIAL DISTRIBUTION OF THE HIGHEST 3-HOUR SO? CONCENTRATIONS (pG/M 3 ) OBTAINED FROM THE
CRSTER MODEL FOR 1964, ASSUMING A 600 MW (12O MW) POPLAR RIVER POWER PLANT WITH ZERO
PERCENT EMISSION CONTROL
•
FORT
PE C K

-------
0
rr j T9 T 1
flOUft S S 0 5 W IS
I II ? ((IIIAI
I II ?
—5—
I ,op L III
CONCINTSIATIONSIN pq1r.
RESERVATIO
• OUTLOOK
IlnI III??
I III — — £ 3L______.. — •
$I,rI__ AI
Figure 5.1-3
SPATIAL DISTRIBUTION OF THE HIGHEST 24-HOUR SO 2 CONCENTRATIONS (pG/M 3 ) OBTAINED FROM THE
CRSTER MODEL FOR 1964, ASSUMING A 600 MW (1200 MW) POPLAR RIVER POWER PLANT WITH ZERO
PERCENT EM! SSION CONTROL
FO RI
PE C K

-------
14
(28) 14
LANT (28)
R ESERVAflO
b31 (0140
50PI ( I II
CONCL0iT11ATtONSINP l&’
INTEIMLO ISOPL1T
Q, I0vHKI


I:Igure 5.1-4
SPATIAL DISTRIBUTION OF THE 1964 ANNUAL SO 2 CONCENTRATIONS (jiG/M 3 ) OBTAINED FROM TIlE
CRSTER MODEL, ASSUMING A 600 MW (1200 MW) POPLAR RIVER POWER PLANT WITH ZERO
PERCENT EMISSION CONTROL
FORT
PE C K

-------
are labeled in pg/rn 3 units of concentration. They represent result-
ing concentrations with zero percent emission controls. The basic
labels are for a 600 MW plant, while those in parentheses are for a
1200 MW plant. Operation of one 300 MW plant would result in pre-
dicted concentrations equal to one half of the 600 MW concentrations.
The concentration patterns indicate that the highest values occur
in general toward the southeast with some secondary peaks toward the
south and southwest. The greatest 1-hour concentrations in the U.S.
are about 214 .jg/rn 3 for a 600 MW plant and 428 pg/rn 3 for a 1200 MW
plant. The highest 3-hour concentrations are 96 pg/rn 3 and 192 pg/rn 3
while the 24-hour concentrations are 21.7 ug/m 3 and 43.4 pg/rn 3 for
the 600 MW and 1200 MW plants, respectively. The highest annual con-
centrations are about 2.1 pg/rn 3 and 4.2 .ig/m 3 . Detailed tables of
model output for 600 MW and 1200 MW plants with zero, 60, and 90 per-
cent S02 control are included in Appendix C.
Both the tabular and graphical presentations of model outputs
indicate that the highest predicted SO 2 concentrations will occur in
southerly and southeasterly directions from the plant site. Regions
of maximum predicted concentrations are generally confined to the
extreme northerly part of the impact area within 15 miles of the Inter-
national Boundary. For example, for a 600 MW plant with zero percent
S02 control, the maximum annual concentrations (1.8-2.0 pg/m 3 ) would
occur on azimuth 120° in the northeast corner of Daniels County
(Figure 5.1-4). The predicted elevation in annual SO 2 concentrations
for most of Daniels County would be less than 0.6 pg/rn 3 (0.00023 ppm).
Although Roosevelt County is beyond the predictive range of the CRSTER
model, the annual SO 2 concentrations there (for 600 MW, no control)
would be expected to be less than 0.2 pg/rn 3 (0.000076 ppm).
5.1.4.2 Oxides of Nitrogen (N0 )
The spatial distributions of the highest 1-hour concentrations
and the annual mean concentrations of NO for the year 1964 are shown
in Figures 5.1-5 and 5.1-6, respectively. The general patterns for
the NOx distribution are similar to the SO 2 distributions. This is
because the meteorological input to the model is the same for both
pollutants. In the U.S. the maximum 1-hour concentrations of NOx are
about 74 pg/rn 3 and 148 pg/rn 3 for the 600 MW and 1200 MW plants, re-
spectively. The highest 1-hour maxima and the second highest 1-hour
maxima for 120, 170, and 260 degree azimuths are included in Appendix
C for the 600 and 1200 MW Poplar River plants.
72

-------
0 OUTLOOk
RESERVATION
St If CUll 4 (I
I 45?
I4.D4AN RIS(N 1tA1 1044 OtJ4.DASt
COIINIV WUNDAHV
Pil l.il W*V
ISOPLETI I
CONLI NTUAIIONS IN 4
INStrHM O ISOPLLTII
Q t, l rpu
- - -____ - - 454 (4?
S4ITLIS
Figure 5.1-5
SPATIAL DISTRIBUTION OF THE HIGHEST 1-HOUR NOx CONCENTRATIONS (jiG/M 3 ) OBTAINED FROM THE
CRSTER MODEL FOR 1964, ASSUMING A 600 MW (1200 MW) POPLAR RIVER POWER PLANT
FORT
PE C K

-------
0 OUTLOOK
ERVATION
MUSS S 0 5
k uMtI(MS 5 0 5 0 5
5—
1)401*50)5)5*0110)4 EIUPIO*IIV
COWIITY IOUNOASY
ISOPLE TI)
CONCI NtI)ATIONSIN jo/,.,
INTI I IMED ISOPLSII4
I l lS) lOSt
— . ‘ I 4* ) — — _________ — — •0 J —
5ILTL, *l
• 000LEY
Figure 5.1-6
SPATIAL DISTRIBUTION OF THE 1964 ANNUAL NOx CONCENTRATIONS (pG/M 3 ) OBTAINED FROM TIlE
CRSTER MODEL, ASSUMING A 600 MW (1200 MW) POPLAR RIVER POWER PLANT
• POUR
/
FORT
PE C K

-------
5.1.4.3 Particulates
The spatial distributions of the highest 24-hour concentrations
and the annual mean concentration of particulates for the year 1964
are shown in Figures 5.1-7 and 5.1-8. These concentrations are for a
99.5 percent controlled particulate emission. For a 99 percent emis-
sion control all the isopleth concentration values should be multiplied
by a factor of 2. From these maps the following maximum 24-hour and
annual particulate concentrations ( ig/m 3 ) were estimated for Montana:
600 MW Plant 1200 MW Plant
99 Percent 99.5 Percent 99 Percent 99.5 Percent
Control Control Control Control
24 hour 2.0 1.0 4.0 0.18
Annual 0.18 0.09 0.36 0.18
5.1.4.4 Comparison of Model Outputs for the Years 1964 and 1960
A comparison of SO 2 concentrations was made between the years
1964 and 1960. As indicated previously, the winter months for these
two years differed in both wind and temperature conditions. A compar-
ison of the 1-hour and 3-hour maximum concentrations showed very little
difference between the two years. However, the maximum 24-hour con-
centrations in 1960 were about 31 percent higher than in 1964, and the
1960 annual concentrations were about 10 percent higher than those
obtained in 1964.
5.1.5 Impact Assessment
Table 5.1-1 presents the estimated maximal pollutant concentrations
occurring in the U.S. impact area as a result of the operation of a
600 MW or a 1200 MW Poplar River power station. For comparative pur-
poses, the U.S. and Canadian National Ambient Air Quality Standards
and the Montana and Saskatchewan standards are also shown in Table
5.1-1. The 502 concentrations assume a zero percent emission control
while the particulate concentrations are based on a 99 percent emis-
sion control. All the concentrations are based on the maximal values
obtained from the CRSTER Model for the years 1960 and 1964.
5.1.5.1 Sulfur Dioxide Impact
The SO 2 measurements made by the Montana Air Quality Bureau showed
the 24-hour averaged SO 2 concentrations to be less than 26 iig/m 3 , which
was the minimum detectable level of the instrumentation. 1axima 24-
75

-------
R ESERVATIO
OUtLOOK
—
INDIAN AISAIIVAIIOSI IOUNOANI
COL,NI I IOUIIIJAIIY
V . 1 , 11 WA V
ISflPI liii
CON( ,ENTU*TIONS IN
INTEHMEO ISOPL(TII
Q I,,sI SA IL
__L L_ .. i _.____ AA
5 ,LSL.. I l
Figure 5.1-7
SPATIAL DISTRIBUTION OF THE HIGHEST 24-HOUR PARTICULATE CONCENTRATIONS (jjG/M 3 )
OBTAINED FROM THE CRSTER MODEL FOR 1964, ASSUMING A 600 MW (1200 MW) POPLAR
RIVER POWER PLANT WITH 99.5 PERCENT EMISSION CONTROL
FORT
PE C K

-------
O
(0
—
/
,
0 OUTLOOK
1 10 , 1* A,
usopi r r
C(JNCINTHATIONS IN p
INTLHMLD I OPtiTiI
I lls’,”,
_ IIA __ “‘°‘
A41 1 AM
ERVATIO
Figure 5.1-8 SPATIAL DISTRIBUTION üí THE 1964 ANNUAL PARTICULATE CONCENTRATIONS (pG/M 3 ) OBTAINED FROM
THE CRSTER MODEL, ASSUMING A 600 MW (1200 MW) POPLAR RIVER POWER PLANT WITH 99.5 PERCENT
EMISSION CONTROL
.1 1T(llft .U
I4At
FORT
PE C K

-------
Table 5.1-1
PROJECTED IIAXIMA POLLUTANT CONCENTRATIONS ( 1g/m 3 ) IN UNITED STATES IMPACT AREA
COMPARED WITH APPLICABLE STANDARDS
*
Pollutants
600 MW
Plant
.
Maxima
1200 MW
Plant
.
Maxima
U.S.
**
NAAQS
Montana
AAQS
Max. Allowable
Increase (PSD)
for Class II Area
Canada Max.
.
Desirable
Saskatchewan
AAQS
SO?
1 Hour
3 Hour
24 Hour
Annual
214
96
28
2.4
428
192
56
4.8
-
l300t
365
80
655
-
265
55
-
512
91
20
450
-
150
30
450
-
150
30
NO
1 Hour
Annual
74
0.79
148
1.6
-
100
-
-
-
-
-
60
400
100
Particulates
2.6
0.2
5.2
0.4
l5O
75
200
75
37
19
-
60
120
70
24 Hour
Annual
Concentrations Assume Zero Percent Control; Particulate Concentrations Assume 99 Percent Control.
MOS = Ambient Air Quality Standards.
1 Secondary Standard.

-------
hour SO 2 concentrations, added to the background concentrations by a
1200 MW power plant, are about 56 pg/rn 3 . The total concentration of
82 pg/rn 3 is well below the Montana AAQS of 265 1.Jg/m 3 , and the NAAQS of
365 pg/rn 3 . The highest 1-hour concentrations are estimated to be near
428 pg/rn 3 for a 1200 MW plant. This concentration is below the 655
pg/rn 3 Montana AAQS but approaches the Canadian standard of 450 pg/rn 3 .
Both the highest 3-hour concentrations and the annual concentrations
are well below all air quality standards.
Although the SO 2 concentrations resulting from a 1200 MW plant
are below U.S. and Montana standards this will result in some deteri-
oration of air quality in the U.S. impact area. To prevent significant
deterioration of air quality the U.S. Clean Air Act prescribes a maxi-
mum allowable increase in concentrations of SO 2 for Class II areas.
These values are shown in Table 5.1—8. The maximum increases in SO 2
concentrations predicted for the impact area range from 24 to 62 per-
cent of the maximum allowable increases.
It should be emphasized that the preceding comparisons are based
on the maximum predicted SO 2 concentrations. The predicted concentra-
tions for most of the study area are considerably less than the maximum
values. For example, the predicted annual SO 2 concentrations for most
of Daniels County and all of Roosevelt County are about an order of
magnitude less than the maximum values which occur near the Inter-
national Boundary. Such low concentrations (0.2 pg/rn 3 for 600 MW, no
control) are only about 1 percent of the Maximum Allowable Increase
(PSD, Class II area). Moreover, it represents only about 0.4 percent
of the Montana Ambient Air Quality Standards. Operation of a single
300 MW unit with no SO 2 control would result in predicted concentrations
of one half of the preceding 600 MW values.
A similar comparison can be made for short-term SO2 concentrations.
The overall maximum 1-hour concentration of 96 pg/m 3 /600 MW, (no control)
would occur at or near the International Boundary. Most of Daniels
County would experience highest 3-hour SO 2 exposures of 20 to 40 pg/rn 3 ,
while maximum concentrations in Roosevelt and Sheridan Counties would
be less than 20 pg/rn 3 . These values are quite low when compared with
the National Ambient Air Quality Standard of 1300 pg/rn 3 . Furthermore,
the maximum concentrations in Roosevelt and Sheridan Counties ( 3O
pg/rn 3 ) represent only 6 percent of the maximum allowable increase for
Class II areas.
5.1.5.2 NO Impact
The predicted increase in NOx annual concentrations for a 1200 MW
plant amount to only 1.6 pg/rn 3 . This is well below the 100 pq/m 3 NAAOS.
The U.S. and the State of Montana do not have a 1-hour NOx standard.
However, the Canadian 1-hour standard is 400 pg/rn 3 . The predicted
highest NO 1-hour concentration in the impact area of 1L18 pg/rn 3 is
well below the Canadian standard.
79

-------
5.1.5.3 Particulate Impact
The present particulate loading in the impact area has an annual
geometric mean of about 20 to 25 ug/m 3 . A 1200 MW plant with 99 per-
cent control will only add about 0.4 iJg/rn 3 to the present background
concentrations. This is only about 2 percent of the maximum allow-
able particulate increase permitted by the prevention of significant
deterioration (PSD) regulations. The Montana and U.S. annual partic-
ulate standards of 75 jg/m 3 would therefore not be exceeded in the
impact area.
The present 24—hour maximum particulate concentrations measured
in the impact area by the Montana Air Quality Bureau range from 100
to 109 ug/m 3 . The predicted maximum increase of 5.2 ig/m 3 for a 1200
MW plant with 99 percent control would result in maximum 24-hour con-
centrations of 105 to 115 g/m 3 . These values are below the Montana
standard of 200 ug/m 3 and the U.S. secondary standard of 150 ig/m 3 .
The predicted increase is about 14 percent of the maximum allowable
increase under PSD regulations for Class II areas.
The trace metal analyses of collector ash from Poplar River Coal
were used in conjunction with model predictions to estimate potential
elevations in atmospheric trace metal concentrations. The maximum
annual increase in particulate concentrations was assumed to be about
0.1 pg/m 3 (600 MW plant with 99.5 percent ESP’s). Such concentrations
could be expected to occur very near the International Boundary; how-
ever, the annual particulate increases over most of the impact area
would be approximately one order of magnitude less ( 0.01 iig/m 3 )
(Figure 5.1—7).
The resultant predicted trace element concentrations are presented
in Table 5.1-2. These values were then compared with background trace
element concentrations measured by lesich and Taylor (1975) at Glasgow,
Montana (Table 5.1—3). The comparisons indicate that operation of a
600 MW power plant would result in very minimal increases in background
concentrations. The predicted annual increases range from 0.02 to 2.2
percent of the background levels (Table 5.1—2). These differences
would not be detectable with air quality monitoring equipment. Even
if the hi9hest 24-hour particulate concentrations were considered
(1.0 Jg/mJ), the resultant increases over background would be only
about 2 to 20 percent. Moreover, such increases would occur only at
locations near the International Boundary.
5.1.5.4 Fumigation Impact
The phenomenon known as “fumigation” is discussed in Appendix A-C
Inversions. The CRSTER Model does not incorporate fumigation in esti-
mating pollutant concentrations. However, Turner (1970) has suggested
a method for estimating peak concentrations resulting from plume
fumigation. Following this technique and employing the emission and
80

-------
Table 5.1-2
CALCULATED INCREASES IN AIR CONCENTRATIONS OF TRACE ELEMENTS
NEAR THE POPLAR RIVER PLANT
Maximum Increase 1 Percent Increase
in Annual Concentration ( g/m 3 ) Above Background
Antimony 9.8 x 1O 0.5
Arsenic 7.4 x 10 0.3
Beryllium 4.6 x io8 0.3
Cadmium 2.4 x io_8 0.05
Chromium 6.8 x io6 0.1
Copper 3.7 x 1O 0.2
—7
Germanium 1.8 x 10 1.4
Lead 1.6 x 10 0.3
Manganese 6.0 x 1O 2.2
Nickel 1.1 x i0 6 0.06
Selenium 2.9 x 1O 0.2
Silver 1.7 x io6 0.06
Vanadium 3.2 x io_6 0.4
Zinc 2.5 x io 6 0.02
1 Assuming atmospheric particulate concentration of 0.1 pg/rn 3 .
81

-------
Table 5.1—3
1975 BACKGROUND TRACE ELEMENT CONCENTRATIONS (1.ig/m 3 )
MEASURED NEAR GLASGOW, MONTANA
Source: Mesich and Taylor, 1976.
Composite
Samples
2nd Quarter
3rd
Quarter 4th Quarter
Element
Aluminum
1st
Quarter
0.5
1.0
0.6
1.0
Antimony
<6
x 1O
2
x 10
3
x 1O
3
x 1O
Arsenic
2
x 1O
2
x 1O
3
x 1O
2
x 1O
Beryllium
<2
x 1O
<2
x 1O
<2
x 1O
<6
x io6
Cadmium
2
x 10
3
x 1O
1
x 1O
3
x 1O
Chromium
2
x 1O
9
x 1O
7
x 1O
2
x 1O
Copper
3
x 1O
2
x 1o 2
3
x io 2
1
x io2
Germanium
<5
x io 6
1
x 10
3
x 1O
<6
x io 6
Lead
3
x 1O
9
x io
1
x io2
3
x io
Manganese
<4
x 1O
<3
x 1O
<3
x 1O
1
x 1O
Nickel
2
x 1O
2
x 1O
5
x 1O
<7
x 1O
Selenium
2
x 1O
1
x 1O
2
x 1O
2
x 1O
Silver
1
x 1O
5
x 1O
3
x 1O
2
x 1O
Vanadium
4
x 1O
7
x 1O
2
x 1O
3
x 1O
Zinc
4
x iü
1
x io 2
2
x io2
1
x io2
82

-------
stack parameters utilized in the CRSTER Model, concentrations were esti-
mated for Pasquill stability categories E and F. The results are pre-
sented in Table 5.1-4.
For very stable conditions (CAT F) SO 2 concentrations of 2000 to
4000 pg/rn 3 may occur 10-20 kilometers downwind of a 1200 MW plant. Con-
centrations of NO may range from 850 to 1350 pg/rn 3 , while particulate
concentrations of 110 to 170 pg/rn 3 may occur. These concentrations are
those that might occur when wind speeds are light and a very strong sur-
face inversion is present.
Furniaation concentrations under more typical meteorological condi-
tions for Scobey, Montana are presented for the midseasonal months in
Tables 5.1-5 throuqh 5.1—7. A summary of the meteorological conditions
for these months and the height of the plume above the surface are pre-
sented in Table 5.1-8. The meteorological conditions are based on
upper-air temperature soundings and wind rose tables for Scobey, Montana.
These data were presented and discussed by Gelhaus, et al. (1979). The
plume heights were calculated using the Briggs’ (1969, 1970, 1972) plume
rise equations. Both the plume heights and the inversion intensities
( T,’AZ) showed little seasonal variation. The plume heights were close
to 200 meters, while the inversions were of moderate to strong intensity.
The concentrations shown in Tables 5.1-5 through 5.1-7 are only
one-half to one-third the values resulting from the more severe meteoro-
logical conditions (Table 5.1-4). The highest concentrations occur in
the winter with a minimum in the autumn. Also, concentrations decrease
and the plume width increases with downwind distance. At 10 kilometers
(krn) downwind from the source, the concentrations at a distance of 500
meters from the plume centerline are only 25 percent of those found at
the centerline. At 20 km, the concentrations drop to 16 percent of
centerline concentrations at a distance of 1000 meters fron the center-
line.
Wintertime SO 2 fumigation concentrations, along the plume center-
line, range from 1800 pg/rn 3 at 10 km to 482 pg/rn 3 at 40 km for a 1200 MW
plant. In the autumn, when the S02 concentrations may be the lowest,
the values range from about 1100 pg/m 3 at 10 km to 298 pg/rn 3 at 40 km.
The N0 wintertime concentrations along the plume centerline may
range from 610 pg/rn 3 at 10 km to 162 pg/rn 3 at 40 km for a 1200 MW plant,
(Table 5.1-6), while particulate concentrations may range from 76 pg/rn 3
to 20 pg/rn 3 at 10 and 40 km, respectively (Table 5.1-7).
The frequency of plume fumigation will most l1kely be the greatest
in the spring and summer and the lowest in the autumn and winter. Sta-
ti sti cs on the frequency of stabi 11 ty categories dun ng the hours from
0000 to 0600 MST, presented by Gelhaus, et al. (1979) for Scobey, Montana,
show that the F stability category (moderately stable) occurs about 25
to 30 percent of the time in autumn and winter and 35 to 45 percent in
83

-------
Table 5.1-4
ESTIMATES OF MAXIMUM GROUND-LEVEL CONCENTRATIONS (jiG/M 3 ) DURING MORNING FUMIGATION
Downwind
Stab Cat [ (Wind Speed 5 iii/s)
Stab Cat F (Wind
Speed 3 m/s)
Distance (Kin)
600 MW
1200 MW
600 MW
1200 MW
SO
2
NO
x
TSP
SO
2
NO
x
TSP
SO
2
NO
x
TSP
SO
2
NO
x
TSP
10
804
270
34
1608
540
68
2016
676
85
4032
1352
170
15
634
213
27
1268
426
54
1571
527
66
3142
1054
132
20
508
171
21
1016
342
42
1301
436
55
2602
872
110

-------
Table 5.1-5
ESTIMATES OF MAXIMUM GROUND-LEVEL S02 CONCENTRATIONS (pG/M 3 ) DURING MORNING FUMIGATION
RESULTING FROM TYPICAL METEOROLOGICAL CONDITIONS AT SCOBEY, MONTANA
DOWNWIND DIST. (km)
600 MW
JAN APR JUL OCT
1200 MW
JAN APR JUL OCT
Along Plume Centerline (CL)
10
20
30
40
912 863 822 568
484 457 434 301
321 303 288 195
241 227 216 149
1824 1726 1644 1136
968 914 868 602
642 606 576 390
482 454 432 298
500 m From CL
10
20
30
40
232 218 207 143
308 291 276 191
258 244 232 160
211 199 189 131
464 436 414 286
616 582 552 382
516 488 464 320
422 398 378 262
1000 m From CL
10
20
30
40
3.8 3.5 3.3 2.3
79.9 75.1 71.3 49.5
134 126 120 83.3
143 134 128 88.4
7.6 7.0 6.6 4.6
160 150 143 99.0
268 252 240 167
286 268 256 177
Note: Estimates are based on meteorological conditions shown in Table 5.1-8 with zero SO 2 control.

-------
Table 5.1-6
ESTIMJ\TES OF MAXIMU 1 GROUND-LEVEL NOx CONCENTRATIONS (pG/M 3 DURING MOF NING FUMIGATION
RESULTING FROM TYPICAL METEOROLOGICAL CONDITIONS AT SCOBEY, MONTANA
DOWNWIND 01ST. (km)
600 MW
JAN APR JUL OCT
1200 MW
JAN APR JUL OCT
Along Plume Centerline (CL)
10
20
30
40
306 290 276 191
162 153 146 101
108 102 96.6 65.4
80.8 76.2 72.5 50.0
612 580 552 382
324 306 292 202
216 204 193 131
162 152 145 100
500 m From CL
10
20
30
40
77.8 73.1 69.4 48.0
103 97.6 92.6 64.1
86.6 81.9 77.8 53.7
70.8 66.8 63.4 43.9
156 146 139 96.0
206 195 185 128.2
173 164 156 107
142 134 127 87.8
1000 m From CL
10
20
30
40
1.3 1.2 1.1 0.8
26.8 25.2 23.9 16.6
45.0 42.3 40.3 27.9
48.0 45.0 42.9 29.7
2.6 2.4 2.2 1.6
53.6 50.4 47.8 33.2
90.0 84.5 80.6 55.8
96.0 90.0 85.8 59.4
Note: Estimates are based on meterological conditions shown in Table 5.1-8.

-------
Table 5.1-7
ESTIMATES OF MAXIMUM GROUND-LEVEL TOTAL SUSPENDED PARTICULATE (TSP) CONCENTRATIONS
(jiG/M 3 ) DURING MORNING FUMIGATION RESULTING FROM TYPICAL METEOROLOGICAL CONDITIONS
AT SCOBEY, MONTANA
DOWNWIND DIST. (km)
600 MW
JAN APR JUL OCT
1200 MW
JAN APR JUL OCT
Along Plume Centerline (CL)
10
20
30
40
38.2 36.2 34.5 23.8
20.3 19.2 18.2 12.6
13.5 12.7 12.1 8.2
10.1 9.5 9.1 6.3
76.4 72.4 69.0 47.6
40.6 38.4 36.4 25.2
27.0 25.4 24.2 16.4
20.2 19.0 18.2 12.6
500 m From CL
10
20
30
40
9.7 9.1 8.7 5.2
12.9 12.2 11.6 8.0
10.8 10.2 9.7 6.7
8.9 8.3 7.9 5.5
19.4 18.2 17.4 10.4
25.8 24.4 23.2 16.0
21.6 20.4 19.4 13.4
17.8 16.6 15.8 11.0
1000 m From CL
10
20
30
40
0.2 0.2 0.1 0.1
3.4 3.2 3.0 2.1
5.6 5.3 5.0 3.5
6.0 5.6 5.4 3.7
0.4 0.4 0.2 0.2
6.8 6.4 6.0 4.2
11.2 10.6 10.0 7.0
12.0 11.2 10.8 7.4
Note: Estimates are based on meterological conditions shown in Table 5.1-8 with 99.5 percent control.

-------
Table 5.1-8
AVERAGE METEOROLOGICAL CONDITIONS (100-200 M LAYER) AND
PLUME HEIGHTS DURING THE MORNING FOR THE MID-SEASONAL
MONTHS AT SCOBEY, MONTANA
PARAMETERS
MONTH
JAN APR JUL OCT
T (°C)
AT/AZ (°C/m)
Pasquill Stability Category
Wind Speed (mis)
Plume Height (m)
—17 2 12 2
0.013 0.017 0.018 0.016
F F F F
6.3 6.8 7.2 10.3
216 210 208 211
88

-------
the sprino and summer. Thus, the probability of local plume fumigations
with concentrations of the magnitude shown in Tables 5.1-5 through 5.1-7
is relatively high.
Fumigation events will result in high concentrations of SO2 within
the impact area but they will be of short duration and confined to rela-
tively small areas. While there are currently no national air quality
standards (primary or secondary) for periods less than or equal to 1-
hour, the potential exists for exceeding Montana’s 1-hour standard,
which states that SO 2 concentrations shall not exceed 655 ug/m 3 for more
than 1-hour in any four consecutive days. As indicated in Table 5.1-5,
concentrations in excess of the state’s standard may occur in that area
between 10 and 20 km from the emission source for the 600 MW facility
and in that area less than 30 km from a 1200 MW plant during fumigation
events. However, due to variable wind conditions and the fact that pre-
dicted SO 2 concentrations at a distance of 500 m from the plume center-
line are less than the state standards at all distances greater than or
equal to 10 km from the emission source, it is improbable that excess
502 concentrations will occur for more than 1-hour during the specified
time interval. -
Since there are no short-term (<1 hour) state or national standards
for NOx or particulates, there is no reasonable basis for comparison
with predicted values. Potential effects of these constituents and SO2
will be discussed, however, in the vegetation impacts section (5.6.1).
5.1.5.5 Visibility Impacts
Mathematical models describing the visual effects of power plant
emission plumes are in the early stages of development. Recently, for
example, a model has been developed to estimate the spatial and temporal
distribution as well as the magnitude of visibility impairment resulting
from power plant emissions of SO2 and NOx at various control levels
(Latimer, 1980). Although the results of a generic application of this
model for power plants in the West have been published, the generating
capacity, meteorological and operating assumptions for the model plant
were such that a comparison with the SPC power plant was not appropriate.
Therefore, in the absence of suitable results, a quantitative estimate
of visibility impairment in Montana resulting from emissions of the SPC
power plant will not be provided.
In qeneral, visibility impairment resulting from a coal-fired power
plant may be classified into two types - short-range (<100 km) and long-
range (>100 km). The short-range type, known as plume blight, occurs
when a plume is perceptible because of its coloration. The long-range
type, where the plume itself is not visible, results in regional haziness,
and the visual range is reduced by plume aerosol. The short-range type
generally results from NOx emissions. Nitrogen dioxide (N02) formed in
the atmosphere from nitric oxide (NO) emissions is visible as a yellow
89

-------
or brown haze, particularly on mornings with stable, light-wind meteoro-
logical conditions. The long-range type results from SO 2 emissions,
which are slowly converted in the atmosphere into sulfate aerosol , which
reduces visual range by contributing to regional haze.
5.1.5.6 Health Effects
Epidemiological investigations have shown relationships between
increased mortality and disease in populations exposed to elevated levels
of atmospheric pollutants (National Academy of Sciences, 1975). The com-
pounds primarily associated with health effects are sulfur oxide-particu-
late complexes. These are formed by the combination of S02 gas and its
conversion products (sulfuric acid and sulfate aerosols) with suspended
particulate matter.
The values presented in Table 5.1-9 provide an indication of the
concentrations of atmospheric pollutants associated with health effects.
Based on the projected levels of S02, NO and TSP associated with the
operation of the proposed Canadian power plants, it is concluded that
there will be no adverse health effects in populations residing within
the Poplar River Basin.
Table 5.1-9
EXPECTED HEALTH EFFECTS OF AIR POLLUTION ON SELECTED POPULATION GROUPS
Value
( gJm ) Causing Effect
Excess Mortality
Worsening of
Visibility
and 1-lospital
Patients with
Respiratory
and/or Human
Pollutant
Admissions
Pulmonary Disease
Symptoms
Annoyance Elfc ts
SO °
500
(daily average)
500—250’
(daily average)
100
(annual anthmetic
mean)
80
(annual geometric mean)
500
(daily average)
250
(daily average)
100
(annual arithmetic
mean)
80
(annual geometric mean)
° British Standaid Practice (Ministry ot Technology. 19ô6) Values for sulfur dioxides and suspended partieulates apply onl’
ri conjunLtlon with each other They may have to be adjusted when translated into terms of results obtained by other
procedrii es
These values represent the differences ot opinion within the committee of experts
Based on high-volume samplers
Source: Shy, 1978
go

-------
5.2 WATER QUANTITY IMPACTS
5.2.1 Methodology
5.2.1.1 Flow Scenarios
Flow impacts were assessed using predicted monthly flows for 1933-
1974 from the Karp II model of the Poplar River and the Modified Mon-
treal Engineering (MME) model of Cookson Reservoir. Results are given
at 12 stations (Figure 5.2-1). The Karp II Model was run for natural,
historical, and existing conditions, six apportionment plans, unlimited
Canadian uses, and two cases with ash lagoon decant. This section will
discuss the impacts of the apportionment recommended by the Poplar River
Task Force. The other apportionments will be discussed as alternatives.
The natural flows were estimated using available flow data and
interstation correlations to complete the flow record (Poplar River Task
Force, 1976b). The historical water uses were then added to the ob-
served flows to give the natural flows under predevelopment conditions
with no apportionment. Historical flows were also estimated from avail-
able flow data and interstation correlations (Poplar River Task Force,
1976b). The existing flows were developed for 1975 levels of develop-
ment with the Cookson Reservoir but with no power plants in operation.
The unlimited Canadian water uses scenarios allow all Canadian water
to be used to represent a ‘worst case.’
The recommended apportionment splits the total natural flow of the
Poplar River and tributaries at the International Border evenly between
Canada and the U.S. In addition to this principal requirement the flow
of the West Fork (stations 9 and 10) must be at least 50 percent of the
natural flow and the flow of the Middle Fork (station 4) must be at
least 60 percent of the natural flow. Flow in Cow and Coal Creeks
(station 2 and 5) must also be at least 60 percent of the natural flow.
Flow in the East Fork depends on the flow at station 4 because different
size releases are made as shown in Table 5.2-1. If flow releases are
needed to meet the apportionment, the releases are made from stations 2
and 5 first and station 4 second.
5.2.1.2 Model Description
Flow scenarios with the recommended apportionment were run for five
levels of development. The various scenarios used in analysis of im-
pacts are summarized in Table 5.2-2. The scenarios 28 through 32 simula-
tions included power plant evaporation but did not include decant from
the ash lagoons as an inflow to Cookson Reservoir. Scenarios 4A, and 8A
were simulated using the MME model to provide a “worst-case” where all
the ash lagoon decant reaches Cookson Reservoir. Since the modeling
work was done SPC has made plans to line the lagoons and recirculate the
water instead of discharging the decant to Cookson Reservoir. The effect
91

-------
Figure 5.2—1
LOCATION OF STATIONS WITH FLOW RESULTS
.i.
.11 .
:i
I
$
I
$

-------
Table 5.2-1
RESERVOIR RELEASES ON THE EAST FORK OF THE POPLAR RIVER
Flow at Station 4
Ac re-Feet
Continuous Release
Acre-Feet Months
Demand
Release
Acre-Feet
Months++
All year
September-May
June—August
September-May
June—August
September-May
June — Au gus t
300 May-September
500 May-September
500 May-September
Sum of March through May flows at Middle Fork at border.
++Schedule for releases is based on irrigation need as follows:
Month May June
Percent 12 18
Amount of releases from scenario
Environmental Sciences.
0-3,800
3,801—7 ,500
7,501—12,000
>12,000
60
60
120
120
180
120
180
1 ,000 May-September
July August September
32 27 11
descriptions of Montana Health and
93

-------
Table 5.2-2
SUMMARY OF FLOW SCENARIOS
Scenario No. Flow Type Level of Development No. of Power Plants
1 Natural Predevelopment 0
2 Historical Historical 1933—1974 0
3 Existing 1975 & Cookson Res. 0
28 Recomm. App 1975 1
29 Reconvn. App. 1985 2
30 Recomm. App. 1985 3
31 Recomm. App. 2000 3
32 Recomm. App. 2000 4
4A With Ash Lagoons 1975 1
8 With Ash Lagoons 1985 2
* Recommended Apportionment of Poplar River Task Force (1979).
94

-------
of the ash la oons is discussed in more detail in Section 5.3 Water
Quality Impacts . A brief summary of the Karp II and MME models will be
given here with a detailed discussion included in Appendix D. The
Karp II model uses the estimated monthly natural flows and the esti-
mated monthly flow at the East Fork border station from the reservoir
model as input for the years 1933 to 1974. The net water uses discussed
in the following section are subtracted from these flows corresponding
to the appropriate scenario. Table E-2 in the Appendix gives the
detailed uses and amounts of the surface and subsurface return flows.
The output of the model is monthly flows for 1933 to 1974 for each of
the 13 stations (see Figure 5.2-1). A cumulative frequency distribu-
tion was computed and the flows exceeded 10, 50, and 90 percent of the
time were listed in a summary section along with the mean flow.
The MME model was used to simulate the Cookson Reservoir as
affected by natural processes and operation of the coal mine and
Canadian power plants. Inputs include tributary inflows, precipita-
tion, natural and forced evaporation, inflows from ash lagoons, mine
dewatering, releases, and groundwater seepage. The Cookson Reservoir
was assumed to be full at the start of the simulation. The maximum
reservoir volume was 32,550 ac-ft. The flow on the East Fork station
at the border is determined from reservoir releases specified in the
apportionment scenarios (see Table 5.2-1) and seepage from the ash
lagoons and groundwater.
5.2.1.3 Water Uses
Canadian Future Uses (1985 )
Except in the indicated cases, water use estimates of future uses
for the model studies were taken from the International Souris - Red
River Engineering Board, Poplar River Task Force, Appendix C (1976).
These estimates are portrayed in Figures 5.2-2 through 5.2-4. (See
Appendix Table E-4 for supporting values.)
Municipal future uses were developed for the Village of Coronach
based on inflated estimates of population levels due to anticipated in-
creases due to development of coal deposits in the area. The expected
increase cited by the Task Force was 114 ac-ft bringing the 1985 total
usage to 150 ac-ft/year. If per capita consumption remains constant,
this represents a population increase to 1250 persons in contrast to
the 1975 population estimate of 300. At the year 2000 level, assuming
constant per capita consumption, 5300 persons would be the projected
population. The town of Rockglen derives its present water supply
from groundwater and it was assumed that future increases could be met
by that resource.
95

-------
8.000
7 ,000
6,000
U
5,000
LU
LU
C
4,000
-j
POPLAR IVI WATER USU
z
z
LCGENO
3,000 I RlGAT$ON ]
ESE VOIR
IVAPO ATIQN
2.000

oo iu•ric
1.000 ______
WILOLtFE
PC ER
0 PtANTS
1975 2000
LEVEL OF DEVELOP .1ENT
Figure 5. -2 CANADIAN WATER USES Oi’ ih . LA
1985
96

-------
POPLAR !IVER WA1E USES
LEGEND
I IGATION
REURVOI
EVAPORATION
E UNICZPAL
DOrIESTIC
WILOLIPE lift
Figure 5.2-3
LEVEL OF DEVELOPMENT
CANADIAN WATER USES ON THE MEDDLE FORK
2
U
2
LU
LU
0
LU
-J
2
z
8 ,000
7,000
6,000
5,000
4,000
3,000
2,000
1.000
0
1975 1985 2000
97

-------
8,000 —
7,000 —
6.000 —
U
5,000—
Lu
Lu
C
4,000 —
-J
POPLAR RIVER WATER USIA
z
z
LEGEND
3,000 — — IRRIGATION J
RESERVOIR
EVAPORATION
2,000 —
OOIESTIC
POWER ffl1ffl1 [
— PI.ANTS
1,000 — ______
1975 1985 2000
LEVEL OF DEVELOPMENT
Figure 5.2-4 CANADIAN WATER USES ON THE WEST FORK
98

-------
Future industrial uses are also predicted primarily on the develop-
ment of coal resources in the area. Water requirements in 1985 would
include 3,600 ac-ft for natural evaporation due to installation of a
reservoir for coal-fired power plants. Installation of two power plants
would require an additional 2180 ac-ft for forced evaporation and plant
consumptive uses. These figures are taken from Volume One of the Poplar
River Final Environmental Assessment for Saskatchewan Power Corporation
(Saskmont Engineering, 1978). Forced evaporation was calculated on the
basis of a 5.2°C temperature excess above normal lake water temperature
for two 300 MW power plants. Using this excess and other meteorological
data daily evaporation was calculated as follows:
E = b(T - Td) f (U)
where E is the daily evaporation (Btu Ft 2 Day 1 )
b = .255 - .0085 T + .00024 T 2
and - T = (T 5 + Td) / 2
T = water surface temperature
Td = dew point temperature
f(U) is a windspeed function equal to 70 + 0.7 U 2
where U is the wind speed in miles per hour.
Losses due to an increase in thermal loading were computed as
— b(K - 15.7 ) HP
dE— (0.26+b) R
where dE is the increase in evaporation (Btu Ft 2 Day )
HP is the thermal loading and
K is the coefficient of surface heat exchange equal to
15.7 + (0.26 + b) f(U)
and b and f(U) are as previously defined. The volume of water evapo-
rated was then evaluated by making use of the latent heat of evaporation
of water (about 1032 Btu/lb).
Wildlife uses were estimated at 300 ac-ft/year and divided equally
between the East and Middle Forks. These uses were estimated in 1985
by Ducks Unlimited to be 1,220 ac-ft. The 300 ac-ft fiqure was derived
as a result of considering that only 25 percent of the projects would
be implemented.
99

-------
Domestic uses were estimated based on linear extrapolations of
historical domestic water usage increases. The Task Force predicted a
1985 use of 730 ac-ft. There is a discrepancy between this and Table
E-4 in that the total domestic 1985 use is only 708 ac-ft. The differ-
ence is the usage in other small tributaries crossing the International
Boundary. The uses were proportioned among the three forks based on
the same percent as the 1975 percent of the total.
The Poplar River Task Force report indicated irrigation water usage
by private developers to be 380 ac-ft by 1985. Altogether, based on
linear extrapolation of historical growth rates, the increase was ex-
pected to be 120 ac-ft over the 1975 level of 300 ac-ft. The 120 ac-ft
was proportioned among the three forks in the same way that domestic
uses were. In addition, an irrigation project amounting to 100 ac-ft
was planned for the West Poplar, bringing the 1985 total irrigation
usage to 520 ac-ft.
Canadian Future Uses (2000 )
All irrigation development presumed to occur was expected to have
taken place by 1985. No further development was estimated for the year
2000.
The construction of a reservoir for power plant development
accounts for the additional 4960 ac-ft of reservoir evaporation loss
from the Middle Fork. This was computed similarly to the reservoir
evaporation losses under the 1985 level of usage. The 1181 ac-ft of
forced evaporation and plant consumptive use was also predicted as per
the 1985 levels. From a water balance, 1827 ac-ft of water diverted
from the West Fork was deemed necessary to allow for power plant devel-
opment on the Middle Fork.
Municipal water uses for the Village of Coronach were estimated
as 500 ac-ft over the 1985 level of consumption for a total of 650 ac-
ft. To meet this demand an additional source of water is required and
it was assumed that 333 ac-ft would be diverted from the Middle Fork to
the East Fork for this purpose.
Domestic (stock-watering) and wildlife demands were not expected
to increase past the 1985 levels.
United States Future Uses
The estimates of future uses in the U.S. portion of the basin are
shown in Figures 5.2-5 through 5.2-9 (see Appendix Table E-5).
The municipal demands that were used in modeling studies were
assumed to be exclusively by the Town of Scobey, Montana. Projections
for the future population estimates based on historic trends were
made by the Montana Department of National Resources and Conservation.
100

-------
8.000
POPLAR RIVER WATER USES
LEGS ND
IRRIGATION
MUNICIPAL [ J
OO IIE ST1C
Figure 5.2-5
U.S. WATER USES ON THE EAST FORK
OF THE POPLAR RIVER
7 000
6 000
C.,
5.000
4,000
-J
z
2
3.000
2.000
1.000
0
1975 19C5 2000
LEVEL OF DEVELOPMENT
101

-------
8.000 —
7.000 —
6.000 —
U
4
5,000 —
4,000 —
POPtAR RIVER WATER I. E5
2
2
4 LEGEND
3,000 — IRRIGATION
_ //// /
OOHESIIC
2.000
1,000 —
0—— —
1975 1905 2000
LEVEL OF DEVELOPMENT
Figure 5.2-6 U.S. WATER USES ON THE MIDDLE FORK
OF THE POPLAR RIVER ABOVE THE
CONFLUENCE WITH EAST FORK
102

-------
8.000
1975 1985 2000
LEVEL OF DEVELOPMENT
POPLAR RIVER WA lER USES
LEGEP D
IRRIGATION
DOINSTIC —
Figure 5.2—7
U.S. WATER USES ON WEST FORK
OF THE POPLAR RIVER (INCLUDES
INDIAN AND NON-INDIAN USES)
U
4
I-
z
LU
LU
a
LU
-J
4
2
2
4
7.000
6,000
5,000
4,000
3,000
2.000
1,000
0
103

-------
8,000 —
7.000 —
6,000—
w
U
5,000 —
4,000—
_______________ P P1 .AR P.WER WATER !S
UGEND
3,000 — — IRRIOAION
2.000— OO?IEETIC
1,000 —
0— —
1975 1985 2000
LEVEL OF DEVELOPMENT
Note: West Fork uses are excluded.
Figure 5.2-8 U.S. WATER USES ON MAIN STEM OF POPLAR RI’/ER
ABOVE FORT PECK INDIAN RESERVATION AND
BELOW CONFLUENCE OF MIDDLE AND EAST FORKS
104

-------
70,000
1975 1985 2000
Figure 5.2-9
LEVEL OF DEVELOPMENT
U.S. WATER USES ON MAIN STEM OF
POPLAR RIVER WITHIN FORT PECK INDIAN
RESERVAT ION
60,000
U
2
LU
LU
C
LU
-I
2
2
50.000
40,000
30,000
2,000
1,000
POPI.AA RIVER WATER USES
LEGEND
IRRIGATION
OO? ESTIC
0
105

-------
In the past 15 years the population of Scobey has decreased steadily.
The 1985 projection is 1445 persons. Water use per capita has increased
but was assumed to level off at 250 gal/day. Using this per capita
usage and the 1985 population estimate the value of 400 ac-ft in the
table was derived. The development of a potash facility was expected
to reverse the population trend by the year 2000 and bring it to 2160
persons. Using the same per capita usage the 600 ac-ft figure was
obtained. The model diverted the water for Scobey from station 3 but
the wells are located just below the confluence with the Middle Fork.
Stock and domestic water requirements for the non-Indian lands were
estimated by the Montana Department of Natural Resources and Conserva-
tion for 1985 and 2000. They were based on linear extrapolation of the
1931-1975 trends. The numbers include stock uses and stock pond evapo-
ration.
The Montana DNRC also made projections of the number of irrigated
acres for each of the West, East and Middle Forks (non-Indian portions)
based on linear extrapolations of 1960 through 1975 data for spreader
dike irrigation systems. (The acreage estimates for all types of irri-
gation systems appear in Appendix Table E-3.) These systems were
assumed to require a single 10 inch (.833 ft) application per year.
Multiplying the projected acreages by this figure the requirements for
these sub-basins were obtained. The spreader acreages for the Indian
lands were assumed to remain the same as the 1975 levels through 1985
and 2000. Using the same 0.833 irrigation depth requirement, the total
requirements for Indian lands were calculated. For the West Fork, the
Indian and non-Indian land requirements were summed. This sum appears
in Table E-5. For the Lower Poplar the indicated value is the product
of estimated Indian spreader acreage and the irrigation depth alone.
Acreage projections for flood irrigation and sprinkler irrigation
were also estimated with regression techniques by the Montana DNRC.
Data from the period of 1961 through 1975 were used. Morrison-Maierle,
Inc. (1978) estimated that the flood irrigation acreages would remain
constant for the Indian lands through 2000. Estimates of seasonal
diversion requirements by the Montana Health and Environmental Sciences
Department, Water Quality Bureau, were 39 inches/acre for flood irri-
gation and 29 inches/acre for sprinkler irrigation.
Sprinkler irrigation acreages on Indian lands were expected to
increase drastically due to the availability of water from the con-
struction of two large reservoirs. The seasonal requirement for
sprinkler irrigation operations was based on the requirements for
alfalfa grown on class II lands. This requirement is roughly 18.5
inches/acre. Assuming a 65 percent field and conveyance efficiency
about 29 inches/acre need to be diverted. Multiplying this by the
total acreage in sprinkler irrigation for each sub-basin and converting
inches to feet gives the volumes shown in Table E—5. Although the
Morrison-Maierle report (1978) indicates the need for a much larger
consumptive use per acre and hence a larger diversion volume, similar
depth requirements were used for the entire basin (see next section).
Water applications were made between May and September according to
percentages given in Appendix Table E-1.
106

-------
Domestic and stock water demands for Indian lands also include
both livestock uses and evaporation from stock ponds. The substantial
increase in the 1985 and 2000 figures over the 802 ac-ft of 1975 is due
to the expected installation of the two reservoirs indicated above and
resultant evaporation from them. Their combined storage is 152,400
ac-ft and their combined surface area is about 7700 acres. The annual
evaporation from these large reservoirs is 4924 ac-ft. Including the
802 ac-ft from existing stock consumption and evaporation, the total
domestic and livestock requirement tallies to 5726 ac-ft.’
Discussion of Irrigation Requirements
In the majority of cases, estimates for future water requirements
from different sources who had performed analyses of the Poplar River
system were supportive. Occasionally, the sources were not in agree-
ment. This was the case in the determination of irrigation water re-
quirements on the Fort Peck Reservation. Estimates of the water require-
ment for alfalfa were derived using a modified Blaney-Criddle approach
and were calculated at 33 inches. Subtracting annual precipitation and
dividing by a 63 percent combined field and conveyance efficiency gave
a seasonal estimate of 34.9 inches or 2.9 ac—ft (Morrison-Maierle, 1978).
Tetra Tech calculates the water requirement for alfalfa as follows.
Alfalfa can be grown from the first frost-free day of spring to the last
frost-free day in fall. The Climatic Atlas of the U.S. gives these dates
as May 30 to roughly September 20. From Figure 4.5-2 the total potential
evaporation for June through September is 21.5 inches. Multiplying by
the upper limit of the Blaney-Criddle evapotranspiration coefficient
(Schwab, etal., 1966) the consumptive use for alfalfa would be 18.3 in-
ches. Subtracting the June through September normal precipitation depth
of 6.5 inches gives a net requirement of 11.8 inches. Dividing by the
field and conveyance efficiency of 63 percent gives a gross diversion
recuirement of 18.7 inches per acre. This is roughly half of the
Morrison-Maierle estimate.
Agricultural irrigation design, however, is predicated upon risk.
While evapotranspiration will not vary widely from year to year, pre-
cipitation can vary greatly. Cumulative precipitation probabilities for
Havre, MT were used (Dept of Commerce, 1968). The April through Sep-
tember normal precipitation for Havre and the Poplar River Basin (taken
from Glasgow, Montana) is similar (8.5 inches). The cumulative proba-
bility curve shows that 90 percent of the time a rainfall depth of 5.4
inches (April through September) will be equalled or exceeded. This
means that there is a 10 percent change that less than 5.4 inches will
fall. Adjusting this value for only June through September by the
‘Montana Health and Environmental Sciences Department, 1979.
107

-------
ratio 6.5/8.5 we can estimate that there is a 10 percent chance of
receiving less than 4.1 inches during the alfalfa growing season. In-
corporating this into the design, the consumptive use of 18.3 inches
would be decreased by 4.1 inches of precipitation. The difference
divided by the efficiency of 63 percent gives 22.5 inches. Demand for
this quantity of water should only be exceeded on the average once
every ten years. This appears to be a realistic design quantity for
alfalfa. Most other crops grown in the area would require less water
with the possible exception of sugar beets.
Based on these estimates it appears that values for irrigation
volumes predicted by the Soil Conservation Service approach are more
reasonable than those made by other sources. The estimate made by
Tetra Tech is lower than those made by other sources due to the fact
that April and May are not considered as part of the irrigation season.
Even if ambient temperatures during these months are conducive to plant
growth it is doubtful that large quantities of irrigation would be re-
quired since snowmelt conditions would tend to keep the soils at mois-
ture contents suitable for plants at least for portions of these months.
5.2.2 Predicted Flows
Pertinent results discussed here are the flows under scenarios 1,
2, 3, 28, 29, 30, 31, and 32 (see Table 5.2.2). Scenarios 28-32 use
the apportionment whereby the U.S. receives 50 percent of the natural
flow of the West Fork, 60 percent in the Middle Fork and reservoir re-
leases as scheduled in Apportionment VI in the East Fork (see Table
5.2-1).
Monthly flows in the Poplar River were calculated for the 10, 50,
and 90 percent frequency for the scenarios listed above with develop-
ment (i.e., irrigation, municipal use, domestic use) at the commensurate
levels. The percentages associated with each are the chance that flows
will be less than the flows represented in the figures. For instance,
in the first plot (Figure 5.2-10) there is a 90 percent chance that
the flow at the International Boundary on the East Fork in April will be
less than 15,941 ac-ft under scenario 28. The flows at each station at
the 10 and 90 percent frequency are tabulated in the Appendix (Tables
F—i and F-2).
5.2.2.1 Flow Conditions at the 90 Percent Level
Monthly flows at the International Boundary on the East Fork are
shown in Figure 5.2-10. The change from scenario 1 to scenario 3 in-
volves the installation of Cookson Reservoir. Essentially only the
spring peak flows have been affected under scenario 3. Spring flood
peaks are primarily the result of snowmelt. The runoff in 1975 was
high so the flows are only decreased by 13 percent. The addition of
power plants to the system has the effect of drawing the storage in the
108

-------
20,000
16000
12,000
8,000
4,000
C
0
w
U
0
0)
2
0
I —
U)
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC
Figure 5.2-10
FLOWS AT EAST FORK OF POPLAR RIVER AT INTERNATIONAL BORDER

-------
reservoir down and thereby decreasing the spillage during peak periods.
The model predicts that the spring runoff flows will occur in March or
April as occurs under natural conditions. The spring runoff flows oc-
curred in May and June at station 1 (USGS, 1976, 1977) in 76—77 after
construction of the reservoir.
The summer flows are less under the apportionment conditions than
natural conditions. The percent reduction from scenario 1 to scenario
28 during 1ay is 44 percent. The percent reduction from scenario 1 to
scenario 29 is 53 percent. Scenario 30 and 31 bring about a 65 percent
reduction and scenario 32 causes an 80 percent reduction from scenario
1. In June all scenarios bring about a 63 percent reduction in the
natural flows. In July the natural flows are reduced by 35 percent for
all scenarios and in August the demand release flows under all scenarios
are the same as the natural flows. Winter flows are higher under the
apportionment scenarios than natural or historical flows. Basically,
the same patterns that exist under the different scenarios at the Inter-
national Boundary on the East Fork also are manifested in the flows on
the East Fork near Scobey (Figure 5.2-11).
The flows on the Middle Fork at the International Boundary (Figure
5.2-12) are 60 percent of the natural flows for all scenarios under
Apportionment VI except scenario 32. Until a storage reservoir is con-
structed on the Middle Fork above the border as predicted by year 2000,
the actual flows will be close to those under natural conditions since
only 188 ac-ft is used annually in 1975 and 396 ac-ft in 1985.
In the Main Poplar at the boundary of the Fort Peck Reservation
(station 8), flow reductions increase with increasing levels of develop-
ment. Figure 5.2-13 shows that in 90 percent of the years the flow in
August and September went to zero under the year 2000 level of develop-
ment. During the other summer months, the flows were reduced from 20
to 50 percent of the natural flows.
The flows in the West Poplar (station 9 — Figure 5.2-14) under
Apportionment VI are specified as 50 percent of the natural flew. Thus,
all scenarios (28-32) have the same flow at the International Boundary.
At Poplar, Montana on the Main Poplar (station 12 - Figure 5.2-15),
with the exception of scenario 28 all future use scenarios have flows
that are zero in the months October through February due to construction
of two reservoirs on the reservation. Scenarios 31 and 32 also have
zero flows in March 90 percent of the time. During the spring, flow
reductions are greater as the level of development increases.
5.2.2.2 Flow Conditions at the 10 Percent Level
On the East Fork at the International Boundary the effect of in-
stalling Cookson Reservoir is seen by comparing scenarios 1 and 3.
Flows generally are damped out in the spring through storage. Ground-
water seepage increases flows in October through March. Flows for sce-
narios 28-32 are representative of the demand releases from the reser-
voir (Figure 5.2-16). The changes in flow will impact the downstream
110

-------
28,000
C
0
U-
a ’
‘I
I— a
I— .
I— ’ 0 )
0
4
I-
12,000
4,000
A A
-
.—-—- —.
JAN FEB MAR
APR MAY
JUN
JUL AUG
SEP OCT NOV
FLOWS AT EAST FORK OF POPLAR RIVER AT SCOBEY
DEC
Figure 5.2—11

-------
20 OO0
16,000
12,000
8 000
4,000
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC
Figure 5.2-12
FLOWS AT MIDDLE FORK OF POPLAR RIVER AT INTERNATIONAL BORDER

-------
60000
50,000
40000
‘C
C
a
U-
30,000
0
z
0
I-
I-
20000
10,000
- Sd
0—————•O Sc28
•Sc29
___ — •Sc30
p OSc3l
•—————. Sc32
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC
Figure 5.2-13
FLOWS OF MAIN POPLAR RIVER AT FORT PECK INDIAN RESERVATION

-------
8000
6000
C
0
U
C 4000
0
2.000
- S d
o—————-o Sc28
SL2O
___ - •Sc30
p Sc31
.—————. Sc32
JAN FEB MAR APR MAY
JUL AUG
Figure 5.2-14
FLOWS OF WEST POPLAR AT THE INTERNATIONAL BORDER

-------
120 000
A A
__—___
0 —o
C
0
w
U
0
z
0
I-
U,
100 000
80 000
60 000
40000
20 000
JAN FEB MAR APR MAY
FLOWS OF POPLAR RIVER AT POPLAR
Figure 5.2-15

-------
- S d
•Sc3
p pSc28
Sc3O
p pSc3l
- 0 .. fl
0
0
U
0
z
0
4
(I )
SAOOQ
/
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC
FLOWS AT EAST FORK AT INTERNATIONAL BORDER
Figure 5.2-16

-------
fish and wildlife populations during the spring. The flows in March
are expected to be 52 to 63 percent less, 19 to 88 percent less in
April, and 30 to 51 percent less in May with the recommended apportion-
ment under 1975 levels of development and one 300 MWe unit operating.
With four 300 MWe units operating and year 2000 levels of development
the flow decreases in March, April , and May are 63 to 98, 69 to 88, and
42 to 80 percent, respectively. The drop in spring flows are shown in
Figures 5.2-17 and 18.
In the East Fork near Scobey the flows are capable of meeting water
requirements under all scenarios during the months of May and June.
Under scenarios 31 and 32 flows are zero in March, April, July, August,
and September (see Figure F-i in Appendix). In March, however, histori-
cal flows are also zero 10 percent of the time. During all other months
the flows are adequate to meet the demands with the scheduled releases.
On the Middle Poplar at the International Boundary the zero flows
for future uses coincide with the zero flows under the natural conditions.
April peak runoff is reduced on the order of 26 to 29 percent under fu-
ture uses (see Figure F-2 in Appendix). Flows on the Main Poplar at
the Fort Peck Reservation boundary are zero in March and June through
September (see Figure F-3 in Appendix).
In the West Fork at the International Boundary, flows under sce-
narios 28 through 32 will be zero for all months. Natural flows are
zero 10 percent of the time for all months except May, October and No-
vember (see Figure F-4 in Appendix).
The flow of the Main Poplar at Poplar will be zero for all months
under scenarios 29 through 32 depending on the operating schedule of
the proposed reservoirs (see Figure F-S in Appendix). Under scenario
28, flows will be adequate to meet demands in all months except March,
July, August and September.
5.2.3 Direct Impacts
5.2.3.1 Municipal Water Supply
The only municipal uses of Poplar River water are for Scobey. The
town operates several wells close to the river which receive a large
percentage of their recharge from the river. The model used a conser-
vative approach in that all the water is withdrawn from the river at
station 3 with no return flow. However, the wells are actually located
below the confluence with the Middle Fork so recharge would come partly
from the Middle Fork. The municipal demand varies by month as shown
below depending on need for outdoor water:
117

-------
I I I I I I I 1 1 1. 1,1 1 1 I i i I I I .. i . . I n_I
MARCH
STATION I - EAST
S March Sd
.. MaichSc3
_______________ March Sc28
1940
1950
Figure 5.2-17
1960
1070
MARCH FLOWS ON THE EAST FORK 1933-1974
100,000 -
10000-
1 , 0 0i) -
100 —
10 -
a
U
4

-------
100.000— I ii ii I I i i. •iii I Ii
19 0 1060
1970
APRIL
STATION 1 - EAST
t April Sd
o_____.- ApnISc3
April Sc28
APRIL FLOWS ON THE EAST FORK 1933-1974
10,000—
1,000 —
100—
10 -
1940
Figure 5.2—18

-------
Monthly Demand, ac-ft/month
January July Total
1975 14 56 350
1985 16 64 400
2000 24 96 600
Ten percent of the time or one in ten years the flows in March in
the East Fork at Scobey are zero under existing (post-reservoir), his-
torical , and the recommended apportionment. The water uses in March
(1975) were estimated as 789 ac-ft for spreader irrigation, 105 ac-ft
for stock and 17.5 ac-ft for municipal water supply. The available
water before diversions is estimated as 56.7 ac-ft (from model results
at stations 1 and 2). Thus, the municipal demand could be met with an
addit 4 onal 39.2 ac-ft left over which could be used for instream uses,
stock or irrigation. The municipal demand for March is expected to in-
crease to 20 ac-ft in 1985 and 30 ac-ft in 2000. Thus, the water avail-
able for irrigation would decrease to 36.7 and 26.7 ac-ft which would
supply about 44 and 32 acres, respectively. The model overestimated
the impact since an additional 24.3 ac-ft and 8.1 ac-ft of water would
be available from the Middle Fork in 1985 and 2000, respectively. Flow
in April is adequate under 1975 and 1985 levels of development but not
for year 2000 uses. The municipal demand would be 36 ac-ft. This demand
could be met out of the estimated available water of 56.7 ac-ft leaving
20 ac-ft for other demands including stock and spreader irrigation.
Flows in August and September were adequate to meet demands for
1975 and 1985 but not for year 2000. The water uses for year 2000 are
estimated as follows:
Uses August September
Flood Irrigation 146.4 62.7 acre-feet
Municipal 90 60
Available Water 56.7 56.7
The municipal demands are higher in these months reflecting outdoor
uses for a total of 250 gallons per capita/day. If the basic water
supply needs of 30 ac-ft were met 26.7 ac-ft would be available for
irrigation. The municipal demands can be met in the other months. The
inability to meet the high summer demands could be resolved by drilling
another well away from the river or encouraging the public to conserve
water. Some water could come from the Middle Fork although flows are
zero 50 percent of the time for historical and existing conditions. In
some years flows have also been zero on the East Fork at Scobey in late
summer.
120

-------
5.2.3.2 Uses Dependent on Spring Runoff
Spring runoff peak flows are expected to decrease under the appor-
tionment plan at all the border stations but will decrease the most on
the East Fork (see Figures F-6, F-7, F-8 and F-9 in Appendix). The
peak flows supply water for filling stock reservoirs and irrigating
land by the spreader method. The peak flows also scour the river chan-
nel and clean out sediment deposited under low flow conditions. This
scouring function is important in maintaining fish and wildlife habitat
as will be discussed in more detail in Section 5.6.4.
Model output gives the volume of discharge in ac-ft rather than a
peak discharge in cfs. Flow data for two years since the Cookson Res-
ervoir was constructed provide an indication of what daily peak flows
might be on the East Fork. The flows in March, 1976, reached a maximum
of 827 cfs. The peak flows in the 76-77 water year came in May with
peak flows of only 58 cfs. The peak flows in the 77-78 water year in
May were 270 cfs. The variability of peak flows occurs also under
natural conditions (see Figures 5.2-13 and 5.2-14). Water demands on
the East Fork between stations 1 and 3 include 789 ac-ft for spreader
irrigation and 105 ac-ft for stock. One year out of ten the full
demands cannot be met in 1975, 1985, or 2000. If the maximum amount
was used for irrigation in 1975 about 47 acres could be irrigated
instead of the total of 1,023 acres. The lack of adequate supplies for
irrigation and filling stock reservoirs in March is less serious if
adequate flows occurred in April or May. In four of the nine years
with zero flows in March the main runoff came in April. In the other
five years of low spring runoff most of the runoff occurred in March.
The water available in April for these uses and instream flows is
adequate in 1975 and 1985 but not year 2000. The available water would
be 20 ac-ft which would not meet the total projected demand of 816.2
ac-ft. These low flows occurred in five out of 43 years when all but
one year had a low spring runoff.
March flows on the main Poplar (station 8) are zero 10 percent of
the time for historical and existing conditions and the recommended
apportionment. This is not surprising since flows at station 3 further
upstream at Scobey are also zero 10 percent of the time. Available
water from the Middle Fork varies between 8 and 24.7 ac-ft for the rec-
ommended apportionment. The demands for water at station 8 include 334
to 465 ac—ft for stock/domestic uses and 287.6 to 816.8 ac-ft for
spreader irrigation. These demands appear to be high since the esti-
mated natural flow was only 186 ac-ft 10 percent of the time. However,
stock reservoirs are filled mostly from smaller tributaries to the main
forks of the Poplar River whenever the spring runoff occurs.
Prior to construction of two reservoirs on the Fort Peck Indian
Reservation March flows at Poplar (station 12) are zero 10 percent of
the time under historical, existing and the recommended apportionment
conditions. Thus, the spreader irrigation demand of 228.8 ac-ft would
121

-------
not be met 10 percent of the time. April flows would be zero 10 per-
cent of the time in 1985 and 2000 but no diversions are required al-
though there would be impacts on fish and wildlife. The water needed
for stock and sprinkler irrigation is estimated as 33,384 ac-ft in 1985
and 61,042 ac-ft in 2000. Available water flowing into the reservoirs
in dry years (one out of ten years) from October through March would be
about 2,000 ac-ft in 1985 and 1,652 ac-ft in 2000. The stock and domes-
tic demand of 802 ac-ft could be met and 1,199 and 850 ac-ft would be
available for irrigation in 1985 and 2000, respectively. This amount
could supply 450 acres in 1985 and 319 acres in 2000.
Stock and spreader irrigation demands can be met at the Lower
Middle Fork (station 7) and Lower West Fork (station 11). Flows at the
Upper West Fork (station 9) are zero in March and April under natural,
historical , and the recommended apportionment conditions one year out
of ten but there are no diversions for water uses.
5.2.3.3 Summer Flows
Summer flows are diverted for flood and sprinkler irrigation. The
model distributed the demand according to the monthly schedule shown in
Tables E-1 and E-2 in the Appendix. The estimated actual and projected
acreages under irrigation were compared to the number of acres which
could be irrigated with the available water based on four applications
per year with an average of 6.1 inches per month. The information will
be summarized below with the complete table included in the Appendix
(Table F-i). This is a worst case since the average number of appli-
cations per year has been 2.4 under historical conditions with little
or no irrigation in August and September.
Irrigation demands on the Upper East Fork can be met from a quantity
standpoint in 1975 and 1985 at least 90 percent of the time. The demand
for irrigation of 1,533 acres in the year 2000 exceeds the available
supply in August and September. Flows have been zero 10 percent of the
time under historical and existing conditions. The number of acres that
could be irrigated in these months is about 188 depending on the method
and extent of municipal uses.
Flows in the upper Middle Fork (station 4) are zero in August and
September, an estimated one out of ten years, for the recommended
apportionment as well as for natural and existing conditions but there
are no diversions.
Flows in the Lower Middle Fork (station 7) are zero 10 percent of
the time in June for the historical, existing, and apportionment condi-
tions. The rest of the summer can be even drier with flows zero 50 per-
cent of the time in July and 90 percent of the time in September under
existing and apportionment conditions. Irrigation demands (net) are as
follows:
122

-------
1975 2000
June 381.1 acre—feet 674.2 acre—feet
July 522.4 1051.7
August 412.8 833.5
September 137.6 280.6
Only a part of this demand could be met at the ten percent frequency
even under natural conditions when only 194.4 (June), 32.4 (July), and
8 ac-ft (August and September) would be available. Fewer irrigation
applications under 1975 conditions would be made since only 408 acres
could be irrigated in June, 344 in July, 36 in August and none in
September. The full irrigation demand could be met in June more than
50 percent of the time but less than 50 percent of the time in July
and less than 10 percent of the time in August and September.
During the summer months of June to September flows on the Main
Poplar (station 8) are zero one out of ten years under the recomended
apportionment and historical and existing conditions. For the year 2000
level of development the flows for August and September are zero in 40
out of 42 years. Water demand in June for irrigation in 1975 was 472.9
ac-ft. This is estimated to increase to 810.9 ac-ft by the year 2000
which is more than the 640 ac-ft under natural conditions. Available
water would provide about half of the 1975 demand and could irrigate
about 708 acres depending on the type of crop and irrigation method
used. For the year 2000 the available water could irrigate about 440
acres. Irrigation demand in July was 718 ac-ft in 1975 and is projected
to increase to 1248 ac-ft in the year 2000. Under the recommended
apportionment the number of acres which could be irrigated averages 492
in 1975 and 24 in 2000. The available water is much less than the net
irrigation demand of 560.7 ac-ft in 1975 and 982.3 ac-ft in 2000. The
response of farmers in dry years such as these would most likely be not
to irrigate the third and fourth time. The case in September is similar
with a net irrigation demand of 199.8 ac-ft in 1975 and 322.5 ac-ft in
2000. The number of acres which could be irrigated is 172 in 1975 and
148 in 1985.
Flows on the Upper West Fork (station 9) are zero ten percent of
the time under historical, existing, and the recommended apportionment
conditions in March, April, August, and September. No diversions are
made from the river at station 9. As expected, flows in the Lower West
Fork (station 11) are zero in August and September under historical
existing and the recommended apportionment. By the year 2000 May and
June flows are estimated to be zero 10 percent of the time. Water uses
for irrigation include 356.2 ac-ft in May and 582.8 ac—ft in June in
2000. These demands cannot be met 10 percent of the time since there
is no available water from upstream. Full irrigation demands in July
123

-------
cannot be met 50 percent of the time in 1985 and 2000, with water avail-
able in 1985 to irrigate only about 24 acres. August and September
flows are zero 10 percent of the time under historical , existing and
the apportionment scenarios and 50 percent of the time for the recom-
mended apportionment scenario. For September in the year 2000 the flow
is zero 100 percent of the time if maximum available water was with-
drawn. The irrigation demands are as follows:
August September
1975 204.2 acre-feet 70.2 acre-feet
2000 725.3 246.1
Water would be available to meet these demands less than 50 percent of
the time. The water available less than 10 percent of the time could
irrigate about 48 acres.
Meeting the 1975 irrigation demands during the summer on the Fort
Peck Indian Reservation is possible 50 percent of the time. The full
demand can be met 10 percent of the time in June and September. In
July only about 296 acres can be irrigated instead of 618 acres. In
August no water is available for irrigation.
Expansion of irrigation is projected based on construction of two
reservoirs on the Main Poplar. The design capacity of these reservoirs
is 152,400 ac-ft (Morris-Maierle, 1978). After construction of the
reservoirs the projected demand in 1985 for 10,618 acres can be met at
least 50 percent of the time. The irrigation demand in 2000 for 20,618
acres cannot be met 50 percent of the time if no carryover storage from
wet years is included. The full irrigation demand can be met if carry-
over storage from one year is included or irrigation is delayed until
May.
The releases below the two new reservoirs are needed for flood
irrigation. Flows in May and June are adequate to meet demands in 1975
but not in 1985 and 2000 at least 10 percent of the time. Water used
for flood irrigation is estimated to be 39 ac-ft in May and 64.3 ac-ft
in June. Flows in July and August are zero under historical , existing,
and the recommended apportionment 10 percent of the time so the flood
irrigation demands of 99.7 ac-ft in July and 78.7 ac-ft in August would
not be met. The September water demand of 26.2 ac-ft for flood irriga-
tion could be met in 1975 but not in 1985 or 2000 at least 10 percent
of the time. Water could be made available if small releases were made
in the sumer during the low-flow years.
1 2â

-------
5.2.3.4 Winter Flows
Winter flows are needed to maintain the fish and wildlife habitat.
The upper reaches of the Poplar River can freeze and so have no flow.
The no flow conditions occur under natural and the recommended apportion-
ment scenarios at stations 2, 4, 5, 9 and 10 in one out of ten years.
Flows are zero under natural conditions only at stations 1, 3, 6 and 7.
Flows in the Lower Main Poplar (station 12) would be zero after con-
struction of the reservoirs in 1985 if no releases are made. The Upper
West Fork winter flows (station 9) (December to February) are zero for
all years under natural and all other conditions. In November under
the recommended apportionment the flows are zero 90 percent of the time.
The lack of flow in the winter partly explains the poor fish habitat.
The winter conditions on the Upper Middle Fork (station 4) are similar
with zero flows 50 percent of the time.
125

-------
5.3 WATER QUALITY IMPACTS
5.3.1 Description of Quality Models
llater quality in the Poplar River was simulated using Karp III and
the Modified Montreal Engineering (MME) model of the Cookson Reservoir.
Appendix D includes a detailed description of both of these models. The
quality at border, Poplar Reservoir, and downstream stations was modeled
differently. The basic method was to develop regression equations for
specific conductivity or total dissolved solids (TDS) based on the aver-
age monthly discharge. Regression equations were then developed to
predict sodium, sulfate, boron, and hardness concentrations from TDS.
At downstream stations a mass-balance approach is used. The water qual-
ity is calculated from the quality of upstream water plus any ground-
water seepage, irrigation return flow, or runoff. Average yearly
groundwater quality and seepage flows were used for each station. The
quality of surface irrigation return flow was calculated at 10 percent
higher than the quality of the diverted water due to evaporation and
salt pickup.
A hypothetical reservoir on the Middle Fork of the Poplar River
in Canada was included in the year 2000 scenarios to meet Canadian water
demands. The quality in this reservoir was based on mass-balance of all
inflows and outflows and complete mixing in the reservoir. The quality
of the outflow was calculated as the average of the reservoir quality
at the beginning and end of the month. After construction of the Poplar
Reservoir water is released from this reservoir if needed to meet the
flow apportionment rather than Cookson Reservoir.
The Cookson Reservoir was simulated by the MME model (described in
Saskmont Engineering, 1978). The model input includes natural and
forced evaporation, tributary inflows, direct precipitation, ash lagoon
discharges, mine dewatering discharges, power plant releases, and
groundwater seepage. Overflows from Fife Lake occurred in March through
October of 1952-54. The model includes power plant operations such as
addition of sulfuric acid for reverse osmosis treatment of boiler feed-
water, demineralization, and chlorination of condenser cooling water.
The effects on the ash lagoons of these processes are modeled for each
lagoon as if it were a separate reservoir. The outflow from the ash
lagoons is partly surface flow to the reservoir and about 30 ac-ft/month
of subsurface flow into the East Fork between the dam and the border.
Additional groundwater seepage also flows into this section of the river
and is included in the computation of water quality at the border. The
outflow quality from the reservoir is determined by complete mixing of
the reservoir and mass balance of all reservoir inflows and outflows.
The appropriate releases are made with the quality equal to the result-
ing quality of the reservoir at the end of the month. The quality at
the East Fork border station is used as input for the river quality
model, Karp III. The MME model was used for scenarios 4A and 8A. Sce-
narios 28 throuqh 32 include forced evaporation but not the ash lagoon
inflow to Cookson Reservoir.
126

-------
Using model results a discussion is presented below regarding
effects of power plant operation and apportionment on criteria and stan-
dards associated with irrigation stock watering, and human consumption
uses. Effects on aquatic life and secondary impacts on agricultural
land use are discussed in Sections 5.6.3 and 5.5, respectively.
Figure 4.6-1 shows points where projected water quality simulation
results are available. In terms of irrigation and stock watering uses,
as well as impacts of power plant operation, the stations of interest
are 1, 3, 8, and 12. For potable water supply use, stations 3 and 7 are
important. This is the case because the Town of Scobey draws its water
from the East Fork (below the confluence with the Middle Fork) and the
Town of Poplar uses groundwater. The remainder of the system, while
having altered flows, will have only nominally altered quality.
For purposes of the primary impacts discussion, modeling scenarios
2 and 3 and 28 through 32 are considered. It is important here to note
the following: scenario 2 represents historical, pre-Cookson Reservoir
conditions. Scenario 3 represents 1975 conditions with Cookson Reser-
voir in-place. Scenario 28 represents 1975 conditions with a single
operating power plant. Scenarios 29 and 30 represent projected 1985
conditions with two and three power plants, respectively. Scenarios 31
and 32 represent year 2000 conditions with three power plants and four
power plants, respectively. Scenarios 4A and 8A with one and two plants,
respectively, are also discussed.
5.3.2 Boron
Plants can be sensitive to boron but trace amounts are required.
Thus, concentrations in water used for irrigation are important. The
months during which most irrigation occurs in the Poplar River Basin are
April through September, inclusive. The impacts of the increased boron
concentrations are discussed both for scenarios 28 through 32 which do
not include the ash decant and scenarios 4A and 8A which do. The con-
centrations at stations 1, 3, 8, and 12 are discussed for both types of
scenarios. Summary tables of model output are included in Appendix G.
The impacts of these concentrations on crops (alfalfa, barley, oats, and
wheat) and livestock are discussed next.
Boron concentrations decrease in the downstream direction as shown
in Figure 5.3-1. Scenarios 28, 29, and 31 represent the more likely
cases. All boron concentrations for stations 1, 3, 8, and 12 are below
3 mg/2. for scenarios 28 and 29. Boron concentrations for scenario 31
at stations 1 and 3 are less than 4 mg/2 during the irrigation season
but have a 10 percent probability of exceeding 4.1 mg/i. during the
winter. For the four power plant case (scenario 32) boron concentrations
at station 1 during the irrigation season may range between 0.4 and 0.8
mg/9. at the 10 percent probability level and between 3.8 and 8.0 mg/Z
at the 90 percent probability level. The boron concentrations for
scenario 32 at the other stations were less than 4 mg/9. during the irri-
gation season, although winter concentrations at station 3 were up to
6.3 mg/2. at the 90 percent probability level.
127

-------
8
KEY:
r 10T090 PERCENT
1. PROBABILITY LEVEL (ppI)
— Sc 28,5OppI
6 sc 29,9OppI
Sc. 31, 9OppI
I—
Sc 32, 9OppI
STATION WHERE MODEL
Al OUTPUT AVAILABLE
o
2
o —j- .-
o
z -
O 2 — . —-.
00 2’5 50 7’5
Al h3 8 A 12ft
APPROXIMATE DISTANCE DOWNSTREAM OF INTERNATIONAL BORDER, miles
Figure 5.3-1 BORON CONCENTRATIONS IN JULY ON EAST FORK AND MAIN STEM OF
POPLAR RIVER FOR SCENARIOS 28, 29, 31, AND 32

-------
Boron concentrations are higher when the ash decant is included
(scenarios 4A and 8A) (Figure 5.3-2). The resultant concentrations
during the irrigation season are discussed for stations 1, 3, 8, and 12.
The concentrations at station 1 with one power plant are between 3.7
mg/ at the 10 percent probability level and 7.5 mg/2. at the 90 percent
probability level during the irrigation season. With two power plants
the concentrations increase at station 1 for the same probability levels
above to 3.6 mg/L and 13.9 mg/2.. At station 3 the concentrations for
the one power plant case range between 1.1 mg/2 . at the 10 percent proba-
bility level and 6.5 mg/.Q at the 90 percent probability level and between
1.0 and 11.4 ma/9 . for the two power plant case. Concentrations at sta-
tion 8 exceed 4 mg/2 .. only for the two power plant case at the 90 percent
probability level. Concentrations at station 12 are below 4 mg/L at all
probability levels.
5.3.2.1 Boron Impacts on Crops
The effects of boron toxicity in crop-soil systems is poorly defined
partly because the soil-water-plant system is so complex and variable.
Experimental data are sparse and are collected under a variety of soil
and climatic conditions. Soil chemical data were available at a limited
number of locations, mostly adjacent to the Poplar River and only a few
water quality parameters were modeled. For these reasons the crop yield
changes presented here should be considered as best estimates for compara-
tive purposes only. The crops studied were alfalfa, wheat, barley, and
oats, although at present the latter two crops are not usually irrigated.
These crops were included since it is important to determine if the
diversity of crops grown could be limited by future water quality.
Compared to salinity and sodicity few studies have been performed
relating to boron toxicity. Only three studies were found that related
boron (B) in the soil solution or irrigation water to the yield of alfalfa,
wheat, barley or oats. Eaton (1944) grew alfalfa, barley, and oats using
irrigation water with boron concentrations ranging from a trace to 25 mg-
B/2,. Fox (1968) showed that calcium treatments had no effect on alfalfa
grown with boron concentrations in the nutrient solution ranging from
0.25 to 32 mg-B/2 . Yields, on the average, decreased with the increas-
ing boron concentrations. Chauhan and Powar (1978) had decreasing wheat
yields when boron concentrations in the irrigation water increased from
0.5 to 8 mg-B/2..
Other work done by Hatcher, etal. (1959) showed that plants respond
to soluble boron in the soil solution and not to that adsorbed or held
in mineral complexes. Whei irrigation waters containing boron are applied
to soils some of the boron is adsorbed. Therefore it is difficult to
determine the amount of boron that the plant is actually responding to.
In the experiment of Chauhan and Powar (1978) the soil solution concen-
tration of boron was measured. At low concentrations of added boron the
soil solution concentration exceeded the concentration in the irrigation
water. However at high concentrations of added boron the solution con-
centration was only half the concentration in the irrigation water.
Eaton (1944) and Fox (1968) only measured the boron concentration in the
129

-------
14
KEY
T 10 TO 90 PERCENT
I PROBABILITY LEVEL (ppl)
— SC.4A,5Oppl
SC. 8A, 9OppI
— — sc. 8A,5OppI
S —
S
S
S
S.
APPROXIMATE DISTANCE DOWNSTREAM OF INTERNATIONAL BORDER, miles
Figure 5.3-2 BORON CONCENTRATIONS IN JULY ON EAST FORK AND MAIN STEM
OF POPLAR RIVER FOR SCENARIOS 4A AND 8A
S.
5 ’
5’
S.
S.
5’
5’
5’
S.
12
l0•
. 5
5 ’
5 ’
5’
5’
S.
5’
S.
S.
S.
E
0.
z
0
I—
F-
z
U
0
0
0
z
0
0
aD
6
. 5. -.
5 .-
\
4
\
2
Al A3 8â I2

-------
irrigation water therefore their data are subject to some interpretation
since some boron may be adsorbed by the soil and is not available directly
to the plant. Boron adsorption is discussed in more detail in Appendix G.
The approach used to estimate the effect of boron on crop yields was
to develop regression equations between the boron concentration in the
soil solution and the percent yield using data from the literature. Al-
falfa appears to be the most boron tolerant of the four crops based on
the projection of the regression line to zero yield (see Figures G-1
through G—4 in Appendix G). The zero yield projection for alfalfa is
41.3 mg-B/9 (soil solution) followed by oats (13.1), wheat (12.3) and
barley (11.4). The regression equations for percent yield of these crops
with mg-B/9 . (soil solution) as the independent variable are:
Percent Alfalfa Yield* = 88.9 - 2.15 B (r = -0.56)
Percent Oats Yield = 75.1 - 5.7 B (r = -0.78)
Percent Wheat Yield = 105.7 - 8.71 B (r = -0.95)
Percent Barley Yield = 86.6 - 7.61 B (r = -0.91)
*Note: In reality zero boron concentrations would not result in maximum
yield since trace amounts are required.
The data for the Poplar River Basin were obtained in the following
manner. Average properties were estimated from available data (Horpestad,
1978) for the soils in the East Fork sub-basin above the Fort Peck
Indian Reservation and within the reservation. Boron concentrations in
the irrigation water were averaged over the period April through September
for stations 1 and 3 for the East Fork sub-basin and stations 8 and 12
for the Fort Peck Reservation at the 90, 50, and 10 percent cumulative
probability level for scenarios 4A, 8A, 28, and 29. Boron concentration
in the soil solution was estimated by an equilibrium approach using
Langmuir adsorption constants (for details see Appendix G). The dilu-
ting effect of growing season precipitation was then accounted for at the
90, 50, and 10 percent probability levels.
The dilution factor was determined using precipitation from the
three closest weather stations to Scobey for which April to September
probabilities have been tabulated (Thomas and Whiting, 1977) Havre,
Montana, Billings, Montana, and Williston, North Dakota. Distances from
Scobey to these stations were determined and inverse distance weights
for these stations were used to calculate precipitation values for the
90 percent, 50 percent and 10 percent probability levels at Scobey.
These values are 14.2, 10.5 and 7.0 inches, respectively. The seasonal
water requirement for alfalfa (from June to September, assuming that in
the months of April and May the soil will retain sufficient moisture
from snowmelt, requiring no irrigation) is 18.5 inches. This number is
computed by using potential evaporation for June to September and multi-
plying by the upper limit of the Blaney-Criddle consumptive use factor
(0.85) (Schwab, etal., 1966). Subtracting the 90, 50, and 10 percent
probability rainfall magnitudes from the total consumptive use require-
ment give the 90, 50, and 10 percent irrigation requirements for this
crop of 4.3, 8.5, and 11.5 inches, respectively.
131

-------
Similarly irrigation requirements for oats, wheat and barley (ex-
pected irrigation season from May through July) can be determined. The
dilution factor can be calculated for these crops at each rainfall
probability level (Table 5.3-1). Multiplication of the boron concen-
tration in the soil (Be) by the dilution factor gives a reasonable
seasonal average concentration to which the plants respond. In the
analyses which follow it is assumed that the distribution of rainfall
and distribution of water quality parameters are independent. This
assumption becomes more favorable as impoundments are constructed in
the basin.
The expected changes in yield from those yields which would occur
in the upper basin (East Fork) and within the Ft. Peck Reservation if
these crops were presently being irrigated were computed for the four
crops. The changes were derived by taking the average boron (water
soluble) content of the soils from existing data and entering the yield
functions to estimate a present percent yield. The Be for scenarios
4A, 8A, 28, and 29 were then computed and the rainfall dilution factors
applied. These concentrations were used in the yield functions and the
differences between these yields and present yields computed. The
changes in yield were also estimated for the combined effects of boron,
salinity, and sodicity. These results are discussed following the
section on salinity and sodicity effects.
The present yields for 1975 for alfalfa, wheat, barley and oats
due to soil boron are estimated to be 87 percent, 97 percent, 79 per-
cent and 70 percent of the optimum for the East Fork sub-basin and 84
percent, 85 percent, 69 percent and 62 percent within the Fort Peck
Reservation, respectively. These present yields are based on the aver-
age chemical composition of the soil saturation extract and no moisture
stress. The projected yield reductions due to boron are less than the
projected reductions due to salinity. The greatest yield reductions
for scenarios 4A and 8A due to boron were as follows:
Crop Percent Yield Reduction From Present Yield
Alfalfa 6
Wheat 27
Barley 24
Oats 18
Reductions were greater for the two power plant case (scenario 8A) in
the East Fork sub—basin but almost equal in the Fort Peck Reservation.
Yield increases for alfalfa, barley, wheat and oats were predicted for
scenarios 28 and 29 but boron inputs from the ash lagoons were not
included. The expected yields are similar for the East Fork sub-basin
and the Fork Peck Reservation, although present yields in the Fort
Peck Reservation appear lower. Expected yields were the same for the
one and two power plant cases.
132

-------
Table 5.3-1
IRRIGATION REQUIREMENTS AND DILUTION FACTORS FOR ALFALFA
AND SMALL GRAINS
Crop
Rainfall Probability Level
90%
50%
10%
DF*
IR+
DF
JR
DF
IR
Alfalfa
Small grains
.23
.43
4.3
6.7
.46
.58
8.5
9.0
.62
.72
11.5
11.2
*
DF = Dilution Factor, dimensionless
+IR = Irrigation Requirements, inches
133

-------
5.3.2.2 Other Boron Impacts
With respect to stock watering boron probably is of little conse-
quence. Boron fed to the dairy cow as boric acid at the rate of 16-20
g/d for 40 days had no ill effects ( Water Quality Criteria , 1972).
Even at 4 mg/Z of boron, this represents about 220 gallons per day, or
more than seven times what milk cows consume and about 21 times what
beef cattle consume (“Water-use in Montana”, Montana DNRC Inventory
Series Report 13, 1975. A suggested limit for livestock was 5 mg/2.
although this was based on the maximum concentration found in a survey
of lakes and rivers and not on specific effects of boron ( Water Quality
Criteria , 1972). The limit was not included in the 1976 update Quality
Criteria for Water . Boron concentrations at stations 1 and 3 exceed
5 mg/9. for part of the year at the 50 and 90 percent probability levels
for scenarios 4A and 8A. Boron concentrations at station 8 exceed 5 mg/2.
in February with one power plant and in four additional months with two
power plants at the 90 percent probability level. Concentrations at
station 12 do not exceed 5 mg/9 for either scenarios 4A or 8A. Boron
concentrations exceed 5 nig/2.. for some months at station 1 only if four
power plants were operating (scenario 32) without the ash decant input.
There is no drinking water standard for boron. The highest boron
concentrations without the ash input to the Cookson Reservoir at the
90 percent probability level are 4.1 mg/9 . with up to three power plants
and 8 mg/9. with four power plants and year 2000 level of development.
With all the ash decant entering the Cookson Reservoir the concentrations
may reach 20 mg/2 . at station 1 at the 90 percent probability level. Con-
centrations of 5 to 20 g of boric acid (one-time ingestion) may cause
death in adults (Stecher, P.G., 1968). The concentrations in the East
Fork are considerably less than this and the concentration in the water
supply would be even smaller due to adsorption as the water moved through
the soil to the well and due to dilution by the Middle Fork and ground
water.
5.3.3 Salinity and Sodicity
The salinity and sodicity of the river are higher in August and
September than June as shown in Table 5.3-2. The highest concentrations
at all stations occur in the winter (see Appendix G). SAR (sodium ad-
sorption ratio) is not a true conservative parameter so actual downstream
values are higher than given by the model. These parameters and sulfate
can be used to classify irrigation water (Klarich, 1978) as shown below:
Salinity as
Water SO 4 TDS
Class SAR ( mg/2 .) ( mg/2 . )
I <1.0 - 4.2 <192 - 480 <700
II 1.0 - 11.6 192 - 960 350 - 2100
III >9.0 - 11.6 >576 - 960 >2100 - 3000
134

-------
Table 5.3-2
SALINITY, SAR, AND SO 4 CONCENTRATIONS AT SELECTED STATIONS
Stations
Scenarios*
Salinity (TDS), mg/i.
SAR
June Sept
SO 4 , mg/2 .
June Sept
June
Sept
1
3
4A
28
844
1099
946
907
1285
1028
4.5 4.6
5.3 5.6
4.8 4.9
229 244
354 439
260 284
3
3
4A
28
827
1095
884
950
1277
988
5.0 5.7
5.4 5.8
5.1 5.8
226 268
336 415
240 279
8
3
4A
28
918
974
935
1087
1189
1087
6.1 7.1
6.0 6.6
6.1 7.1
248 324
275 367
264 321
12
3
4A
28
1147
1135
1147
1367
1364
1367
8.7 9.5
8.7 8.8
8.7 9.4
242 293
242 306
242 243
Values shown give concentration which is exceeded only 10 percent
of the time. Monthly tables at the 10, 50, and 90 percent
probability levels are included in Appendix G.
*Scenarios are defined as follows: 3 = existing conditions with
Cookson Reservoir but no power plants, 4A = 1975 development
with one power plant and ash lagoon input, 28 = 1975 development
with one power plant and no ash lagoon input.
135

-------
Based on this classification system, the river water at stations 3, 8,
and 12 would be considered class II for scenarios 28 through 32, 4A and
8A and for historical (scenario 2) and 1975 conditions (scenario 3).
The water at station 1 would be class II for scenarios 2, 3, 28 through
31, 4A and 8A; class I for scenario 28 at the 10 and 50 percent proba-
bility level and class II at the 90 percent probability level; and
class III for scenario 32.
5.3.3.1 Salinity and Sodicity Impacts on Crops
The effects of salinity on crops are considered to be primarily
osmotic. Plants respond to the total ion content of the soil solution
as well as to specific ions. In this investigation the electrical con-
ductivity of the saturation extract (ECse) has been used to determine
the osmotic stress on plants due to salt buildup in the soil profile.
The symptoms of salinity poisoning to crops are retarded growth as
evidenced by fewer plants with smaller and fewer leaves (Rhoades, 1972).
Sodium, however, has not only a toxic effect but also damages soil
structure reducing permeability. High concentrations of a single ion,
such as sodium, also may upset the competitive uptake of other bene-
ficial nutrients (Kamphorst and Bolt, 1976).
The sodium absorption ratio (SAR) is the water quality index with
which the sodicity hazard has been estimated. It is a measure of the
availability of the monovalent sodium ion to the availability of the
beneficial calcium and magnesium divalent species. The SAR values
used are for the soil saturation extract as calculated in Appendix A.
The exchangeable sodium percentage (ESP) has also been used as an
index of sodicity (Chang and Dregne, 1955; Bains, etal., 1970;
Agarwala, etal., 1964). The SAR is used in this investigation because
it is consistent with Gapon’s equation (Sposito and Mattigod, 1977) and
the Poisson-Boltzman equation (Bower, 1961) for double layer exchange.
Numerous studies have been done to determine the effects of salin-
ity and sodicity (Table 5.3.3) including one study by Werkhoven (1964)
in which he grew alfalfa and wheat on three loam soils from southeastern
Saskatchewan. Several studies also investigated the effect of leaching
fraction on the salinity and sodicity hazard. Additional information
on the salt and sodium tolerance of these agricultural crops can be
found in Biggar and Fireman (1960), Agarwala, etal. (1964), Wahhab
(1961), Bernstein, etal. (1974), Hanks, etal. (1977), and Shourbagy
and Wallace (1965).
Methodology for Estimating Crop Yields
From most of the literature surveyed on salinity and sodicity it
is unclear whether the decrease in crop yield is a result of an increase
in SAR or conductivity in the soil solution since in most cases they
136

-------
Table 5.3—3
AVAILABLE STUDIES ON SALINITY AND SODICITY HAZARDS
Reference
Bernstein and Pearson (1956)
Chang and Dregne (1955)
Chang (1961)
Werkhoven (1964)
Bower, Ogata and Tucker (1968)
Bower, Ogata and Tucker (1969)
Bernstein and Francois (1973)
Ingvalson, Rhoades and
Page (1976)
Ayers, Brown and Wadleigh
(1952)
Mehrotra and Gangwar (1964)
Asana and Kale (1965)
Torres and Bingham (1973)
Elgabaly (1955)
Bains, etal. (1970)
Hassan, etal. (1970)
Patel and Dastane (1971)
Number of
Leaching
Soil Types Fractions Used
Pachappa loam,
Chino clay
Gila clay loam
Gila loam
6 SE Saskatchewan
soil types
Pachappa sandy
loam
Pachappa sandy
loam
Pachappa sandy
loam
Pachappa fine
sandy loam
Pachappa fine
sandy loam
6 Indian soils
Indian soil
sand culture
resin sand culture
sandy loam
Hobbs silt loam
several types
+Crop types are A = alfalfa, W = wheat, B = barley, and 0 = oats.
+
Crops
A,W,B,O
A
A
A,W
A
A
A
A
W,B
W,B,0
W
W
B
B
B
B
1
1
1
1
4
>1
>1
1
1
1
1
1
1
1
1
4
137

-------
vary coincidentally. The two variables are related best to percent yield
using the following function:
f(SAR, EC)se = in (SAR x EC)se
where SAR is computed with Na+, Ca 2 +, Mg2+ expressed in meq/2., EC has
units of pmho/cm, and the subscript se stands for saturation extract
measurement. The yield data from each experiment were reduced to rela-
tive yield by dividing each yield by the maximum yield for each soil
type. These percent yield data were then plotted against the corres-
ponding f (SAR, EC)se value for each crop. Each crop responded similarly
as shown in the plots (Appendix G). At an f (SAR, EC)se value of approx-
imately 1.5 yield reduction began to become evident and decreased approx-
imately linearly. Below this threshold value, yields were consistently
around 100 percent. At some very low values of the function, barley,
wheat and especially oats seemed to experience ion “starvation”.
Relative yield functions were obtained from linear regression on
the “declining yield” portion of the plots. The functions are as
follows:
Percent Alfalfa Yield = 100% for x < 1.7
and
= 129.81 - 17.41 (x), x > 1.7
(r = -0.71)
Percent Wheat Yield = 100% for x < 0.4
and
= 105.12 — 11.75 x, x > 0.4
(r = -0.62)
Percent Barley Yield = 100% for x < 1.5
and
= 126.15 - 17.22 x, x > 1.5
(r = —0.75)
Percent Oat Yield = 100% for x < 1.8
and
= 153.55 - 29.38 x, x > 1.8
(r = -0.95)
where x = in (SAR) EC)se in all the above equations.
138

-------
Based on these empirical relationships it appears that oats have the
highest threshold tolerance for SAR and EC of the saturation extract
but are also most sensitive to increments of the two (having the high-
est slope). Wheat apparently has the lowest threshold tolerance and
is least sensitive to incremental increases in SAR 5 e and ECse. Alfalfa
and barley appear to be almost equally tolerant.
The average soil properties and dilution factors were estimated
as described in section 5.3.2.1 Boron Impacts. The soil chemical data
(Table 5.3-4) show that the soils within the Fort Peck Reservation are
more nutrient enriched. The quality of irrigation waters applied to
these soils was tabulated at the 90, 50, and 10 percent probability
level. The steady-state values of ECse and SAR 5 e resulting from appli-
cation of irrigation water to these soils for three leaching fractions
(0.1, 0.2, and 0.3) were then computed as a seasonal average for April
through September. These are the values of EC and SAR to which the
plants would respond. The derivation of the ECse values from the TDS
and SAR of the irrigation water using the method of Kamphorst and Bolt
(1976) is described in Appendix G.
Yield Changes
The present yields relative to optimum of alfalfa, wheat, barley,
and oats due to salinity and sodicity of the soil are estimated to be
91, 79, 89, and 88 percent for the East Fork sub-basin and 90, 78, 86,
and 86 percent for the Fort Peck Reservation, respectively. These
yields are based on the average chemical composition of the soil satu-
ration extract and no moisture stress. The maximum yield reductions
due to salinity for scenarios 4A and 8A were as follows:
Crop Percent Yield Reduction From Present Yield
Alfalfa 48
Wheat 35
Barley 51
Oats 87
Additional irrigation water applied to the soil removes or leaches the
salts built up by evaporation and thus decreases the salinity of the
soil. For example, a leaching fraction of 0.1 indicates that 10 per-
cent more water is applied than required by the crop. The yield re-
ductions are significantly less with a leaching fraction of 0.2 instead
of 0.1. The improvement in yield with a leaching fraction of 0.3 is
most pronounced for the poorer irrigation water quality at the 90 per-
cent probability level, as would be expected. The improvement with a
leaching fraction of 0.3 is small for the other probability levels.
139

-------
Table 5.3 4
AVERAGE CHEMICAL DATA FOR UPPER BASIN SOILS AND SOILS
WITHIN FT. PECK
B Ca Mg K Na CEC pH EC e SAR e
Location mg/L nieq/100 g - mno/cm (meq/ )
Upoer Basin
(U.S. portion) 11.7 37.7 16.4 - 141.2 ll 79 1.9 4.8
Ft. Peck
Reservation 1.6 60.0 63.0 170. 255. 29 8.1 1.8 5.5
+Estinlated value
The yield reductions were also estimated for the scenarios without
the ash decant - 28 (1975 with one power plant) and 29 (1985 with two
power plants). The maximum reductions which occur for scenario 29 for
leaching fractions of 0.1 and 0.3 are as follows:
Crop Percent Yield Reduction From Present Yields
Alfalfa 39 ( 7)*
Iheat 30 ( 8)
Barley 41 (14)
Oats 75 (25)
*
Values in parenthesis are for a leaching fraction of 0.3.
Under high rainfall conditions and the median water quality conditions
with a leaching fraction of 0.1 yield increases in alfalfa due to lower
salinity and SAR values may occur in both the East Fork sub-basin and
the Fort Peck Reservation. When the leaching fraction is increased to
0.2 yield increases for alfalfa also may occur under the median rain-
fall conditions. Yield increases for barley, wheat, and oats may occur
with a leaching fraction of 0.2 under high rainfall and median water
quality conditions and under the median rainfall with water quality
conditions at the 10 percent probability level. The change from one to
two power plants results in yield decreases generally of zero to 2 per-
cent with a maximum of 5 percent.
140

-------
5.3.3.2 Impact on Crops of Combined Effects of Salinity, Sodicity,
and Boron
The effects of boron and salinity/sodicity were considered to be
additive. The yield reductions were computed for scenarios 4A, 8A,
28 and 29. Summary tables are included in Appendix G. Since the SAR
values from the model are low, the yield reductions for the Fort Peck
Indian Reservation could be greater.
Scenarios 4A and 8A will be discussed first. For alfalfa grown
in the East Fork sub-basin yield reductions of up to 50 percent may
occur with a leaching fraction of 0.1 in low rainfall years. Yield
increases can occur when the leaching fraction is increased to 0.2
under certain rainfall and water quality conditions. In general, yield
reductions within the Fort Peck Reservation for alfalfa would be less.
The increased reduction due to the addition of two power plants is 2 to
8 percent in the East Fork sub-basin and zero to 1 percent in the Fort
Peck Reservation. Yield reductions of up to 60 percent are possible
for wheat for a leaching fraction of 0.1 and generally seem greater
than for alfalfa in the East Fork. Within the Fort Peck Reservation
yield reductions for wheat are estimated to be less than those for
alfalfa. Barley yield reductions are estimated to be up to 75 percent
in the East Fork. Yield reductions within the Fort Peck Reservation
are about 5 percent less than for alfalfa.
Oats were shown to be the most sensitive crop from the salinity
standpoint and with low rainfall and leaching fraction of 0.1 no crop
would be produced if irrigated. Severe yield reductions of up to 60
percent would also be realized in the Fort Peck Reservation.
The yield decreases for scenarios 28 and 29 without the ash decant
are less than for scenarios 4A and 8A. The maximum yield decreases in
low rainfall years with a leaching fraction of 0.1 are as follows:
Crop Percent Yield Reduction From Present Yield
Alfalfa 38
Wheat 28
Oats 74
Barley 42
These yield reductions are 20 to 30 percent less than for scenarios 4A
and 8A. Yield increases may occur at high leaching fractions in years
with at least the median rainfall for all crops. Oats could be irri-
gated under all conditions although the yields were estimated to be
low. The change from one to two power plants results in yield decreases
between zero to 9 percent.
141

-------
5.3.3.3 Other Salinity Impacts
TDS concentrations exceed the recommended limit for drinking water
of 500 mg/i (EPA, 1977) at stations 3 and 7 for all scenarios including
historical conditions during all but a few winter months at the 10 per-
cent probability level. Waters containing in excess of about 1300 mg/i
TDS may be considered unacceptable by consumers. TDS concentrations
above 1300 mg/i were projected at station 3 for scenarios 29 through 32
and 8A at the 90 percent probability level. Diluting flow from the
Middle Fork is available only in the spring and winter except in high
rainfall years.
TDS concentrations above 3000 mg/i can cause effects in poultry.
Concentrations exceed 3000 mg/i for scenario 32 in the winter at the
90 percent probability level at stations 1 and 3 but not for any other
scenarios. All concentrations are below 5000 mg/i so would be suitable
for livestock.
5.3.4 Sulfate (SO 4 )
Secondary Drinking Water Standards (EPA, 1977) for SO 4 have been
established at 250 mg/i. This is due to the fact that SO 4 , when present
in potable water in high concentrations or when in moderate concentra-
tions and consumed by individuals unaccustomed to it, may have a laxa-
tive effect.
The water quality modeling results (Appendix G) indicate that the
SO 4 standard for drinking is exceeded on a historical basis about 10
percent of the time during low-flow months (August-March) in the Lower
East Fork near Scobey (Appendix Table G.2-3). Under the apportionment
agreement and with up to two power plant units in operation (1985), the
standard would be exceeded 10 percent of the time every month. However,
the maximum 90 percentile concentration under scenario 29 is 383 mg/i
(during January) and 386 mg/i for scenario 2. The standard would be ex-
ceeded 50 percent of the time for scenarios 4A and 8A with a maximum 90
percent concentration of 665 mg/i for station 3 and 280 mg/i for station 7.
Operation of three or four units at 1985 or 2000 level of develop-
ment would result in significant SO 4 elevations. With four units (i.e.,
scenario 32) the drinking water standard would be exceeded at least 50
percent of the time during low-flow months. Maximum concentrations
occurring at a frequency of 10 percent during low-flow months would ex-
ceed 800 mg/i.
Sulfate is also an important water quality constituent in relation
to agricultural uses. Sulfate concentrations are an indication of poten-
tial salinity and may also be directly toxic to plants at concentrations
greater than 500 mg/i. Sulfate ion concentrations at all stations gener-
ally increase both with increasing numbers of power plants on the
reservoir and with increasing irrigation development. Concentrations
tend to remain relatively constant as the irrigation season progresses
in the East Fork whereas they tend to increase during the summer at the
142

-------
lower stations on the main stem of the Poplar. The only time when
levels are unacceptable in the system are at the International Boundary
in March under scenarios 30 through 32 and 8A at the 90 percent proba-
bility level. Therefore, sulfate ion concentration does not appear to
be a crucial factor to consider with regard to suitability of waters
for irrigation.
5.3.5 Mitigative Measures to Reduce Impacts of Saline Irrigation Waters
Due to the high potential for increased salinity of Poplar River
waters following power plant developnent and associated apportionment
agreements between the U.S. and Canada, it is appropriate to discuss
potential mitigative measures. High salinity water used for irrigation
may have two potential impacts:
• Direct salinity damage to irrigated crops.
• Downstream effects on a variety of water uses
due to highly saline irrigation return flows.
In arid regions, when rainfall is less than 20 inches/year, little
leaching occurs and the alkalinity of the soils remains high. Normally
a layer with higher levels of caustic calcium carbonate than the sur-
rounding material can be found; the lower the rainfall , the closer this
layer is to the surface. Poplar River Basin soils have a pH ranging
from about 7.5 to 8.5 making them midly alkaline. Soils with pH in
this range, having high soluble salt concentrations, can have detrimen-
tal effects on plants due to plasmolysis. That is, concentration grad-
ients tend to move water out of plant tissues into the soil until plant
cells collapse. Irrigation with saline waters aggravates this situation.
Irrigation appears to be the practice which will dominate agricul-
tural water consumptive use in the Poplar River Basin in the future.
Irrigatton can be extremely beneficial to the economy of the basin.
However, if improperly managed, it can have detrimental effects. Over-
irrigating can cause mineral salts to be leached from the soils and
introduced into shallow ground water aquifers which recharge the river,
impairing water quality downstream. Downstream impacts include not
only problems of saline waters for municipalities and stock watering
operations, but also for other irrigators who, in turn, must irrigate
their crops with more saline waters. Irrigating with a very low leach-
ing fraction can cause salts to build up in the soil to toxic levels.
Potential mitigative measures for each of the aforementioned impacts
of saline water irrigation are discussed in the following sections.
5.3.5.1 Mitigative Practices for Salinity Control in Soils
There are three methods by which salinity in soils can be con-
trolled. These are leaching with irrigation water, conversion of alkali
carbonates into sulfates and control of evapotranspiration (Brady, 1974).
143

-------
Leaching of salts from the soil profile should be done with irri-
gation waters having small amounts of exchangeable sodium. This works
well on pervious soils whose salts tend to be neutral and high in cal-
cium and magnesium. Removal of neutral salts may increase alkalinity
because of increased sodium saturation and consequent increases in
hydroxyl ion concentrations. This can be avoided by treating with gyp-
sum or sulfur which will convert sodium carbonates or bicarbonates to
sodium sulfate.
Conversion of alkali carbonates to sodium sulfate takes place
according to the reaction
Na 2 CO 3 + CaSO 4 CaCO 3 + £4a 2 SO 4
where sodium sulfate is a leachable salt. This method leaves the cal-
cium carbonate in the soil. Several tons of gypsum per acre may be
required and the soil should be kept moist to hasten the reaction.
Application of sulfur to soils in the presence of water creates
sulfuric acid which not only converts the carbonates but reduces
alkalinity as well. The reaction is
NaCO +HSO ‘CO +HO+NaSO
2 3 2 4 2 2 2 4
Using this method the sodium sulfate can be leached and the carbonate
radical is eliminated.
Control of evapotranspiration to control salts is still untenable
for large acreages at this time. However, this method can be utilized
in gardens by covering areas between rows with dark polyethylene films.
This not only prevents evapotranspiration from these areas, which
causes salts in solution to migrate upward in the soil profile, but it
prevents losses due to weed growth and consumption.
Although steps are taken to reduce salinity in soils, the fact re-
mains that some crops simply are more tolerant to elevated levels of
salinity than others. Therefore, good management of saline soils re-
quires selection of the proper crop. Table 5.3-5 shows crops that
could conceivably be grown in the Poplar basin and their relative tol-
erance to salinity.
Tolerance to salinity is governed by many factors and is generally
difficult to predict. Stage of growth and rooting habits are among
these factors. Seedlings are in general more susceptible than estab-
lished plants. Deep-rooted crops are more tolerant than shallow-rooted
species. Plants ajso respond differently to similar concentrations of
different salts. Selection of plant species must be done primarily
based on experience but with proper salinity management a wider variety
can be grown with satisfactory yields.
144

-------
Table 5.3-5
RELATIVE TOLERANCES OF VARIOUS CROPS AND FORAGE SPECIES TO SALINITY ARRANGED
ACCORDING TO DECREASING TOLERANCES WITHIN GROUPS; MODIFIED FROM ALLISON (1964) AND FROM HEM (l97O).*
Tolerant to —Semi-Tolerant and Moderately Tolerant— to _S nsitiv
Field, Truck, and Fruit Croj s
Barley
Sugar Beets
Garden Beets
Ka 1 e
Aspa ragus
Spinach
Rye
Wheat
Oats
Corn
Flax
Sunfl ower
Tomato
Broccol i
Cabbage
Cauliflower
Lettuce
Sweet Corn
Potato
Bell Pepper
Carrot
Onion
Peas
Squash
Cucumber
Field Beans
Radish
Green Beans
Apple
Boysenberries
Blackberries
Raspberries
Strawberries
Forage Species
Sal tyrass
Bermudagra ss
Tall wheatgrass
Rhodesgrass
Canada wildrye
Western wheatgrass
Tall fescue
Barley (hay)
Birdsfoot trefoil
Sweetcl over
Perennial ryegrass
Mountain brome
Harding grass
Beardless wildrye
Strawberry clover
Dall isgrass
Sudangrass
Hubam clover
Alfalfa
Rye (hay)
Wheat (hay)
Oats (hay)
Orchardgrass
Blue grama
Meadow fescue
Reed canary
Big trefoil
Smooth brome
Tall meadow
oatgrass
Mu kvetch
Sourcl over
White dutch
clover
Meadow foxtail
Alsike clover
Red clover
Ladino clover
Burnet
*/\fter Kiarich, 1978

-------
5.3.5.2 Mitigative Irrigation Practices for Salinity Control in Return
Flows
Surface return flows or tail waters are very rarely treated and
sub-surface return flows are virtually impossible to treat. It is
therefore imperative that practices be adopted to mitigate the water
impacts of increased irrigation water utilization in the field. These
measures can be broken down into two basic groups, either improved
water conveyence or improved on-farm management.
Improving conveyance
Seepage out of irrigation diversion ditches is a major source of
salinity resulting from irrigation. The hydraulic head causes leaching
of mineral salts out of soils directly beneath the ditches and can move
salts to shallow ground water aquifers, from which they may enter the
river. Additionally, for irrigation water already carrying salts,
evaporation from these shallow conveyences may result in an increase in
concentration of salts which may be harmful to plants. Both of these
conveyence losses decrease the efficiency of the irrigation system and
cause the gross diversion requirements to increase. Conveyence effi-
ciency can be increased by lining channels with a concrete slip or a
plastic membrane. This prevents losses to ground water and also leach-
ing beneath ditches. The next level practice would be conduits of buried
PVC pipe which would preclude both seepage and evaporative losses.
Friction in pipe runs may require that pumps be installed to provide the
necessary flows. Naturally, these practices can be adopted for both
irrigation canals and for laterals. Proper management of turn-out
volumes by metering can also reduce over-irrigating and subsequent
impairment of return flow quality.
On-Farm Management Practices
On-farm management practices for reducing salinity in return waters
can be broken down into three categories:
o Lining of on-farm conveyences
• Altering irrigation practices or switching to a
more effective system
• Education of operators
the first of which has been discussed in the previous section.
Salinity pickup by irrigation waters cannot be wholly eliminated
because of the need to provide a flow of water through the root zone
which ameliorates the concentrating effects due to evapotranspiration.
This quantity of water over and above the crop needs required to leach
146

-------
salts away from plant roots is known as the leaching fraction. Due to
non-homogeneity of the soils water is normally applied such that the
species with the highest water holding capacity will receive the re-
quired leaching fraction resulting in over-irrigation on the remainder
of the field. Obviously, one method of reducing salt pickup is to in-
crease the efficiency of water applications. The various methods will
be briefly discussed in increasing order of efficiency, with suggestions
for efficiency improvement within each discussion.
Spreader Dike or Level Border Irrigation
Border irrigation is so named because water is spread over the
field by a system of earth dikes or borders which keep the water within
specified bounds. Border systems are well suited for alfalfa and small
grains and consequently have been popular in the Poplar River Basin.
With level border irrigation water is turned onto level plots to form
basins from which infiltration takes place. There is no surface return
flow from these systems.
The average U.S. efficiency for level border irrigation is about
59 percent (Evans, etal., 1978). Efficiences of 80 to 90 percent can
be realized with proper management. Good management requires level
fields so that water will not collect disproportionately in certain
areas, insuring uniform infiltration. The amount of water turned into
the system should be precisely regulated so that over-topping of dikes
and excessive deep seepage does not occur.
Flood or Furrow Irrigation
Flood irrigation involves the application of water into small fur-
rows which traverse the field slope. To apply water efficiently, the
inlet discharge and duration of flow must be a function of soil proper-
ties such as water-holding capacity and infiltration rates, field slope
and furrow geometry. As water is inlet to the furrow it advances down
slope but initial resistance to wetting by the dry soil causes water
not to infiltrate i iimediate1y. Behind the flood “wave” a wetting front
advances down the furrow. Ideally, the flow rate and duration can be
adjusted so that the quantity of water that infiltrates all along the
furrow is uniform. Differences in soil characteristics will cause over-
irrigation in instances where the soil with the least water holdina capa-
bility is properly irrigated. In these systems tail-water or surface
return flow is the result of over-irrigation in addition to augmented
seepage losses. According to the soil type, salt pickup from excessive
infiltration or sediment erosion from surface return can be the major
consideration.
There are several ways by which flood irrigation can be made more
efficient. The first involves irrigation scheduling. Normally, irri-
gation is used when the soil water content drops to 50 percent. The
147

-------
determination of moisture content should always be made at the lower
end of the field such that the minimum intake opportunity time can be
determined. Flow in the furrow should be adjusted so that the time
required for the flow to advance to the end of the field is about 25
percent of the minimal intake opportunity time (Walker, 1978).
A second method is called cut-back furrow or flood irrigation.
The inlet must be automated such that a large “wetting” flow is admit-
ted at the head of the furrow and then the flow is decreased or cut
back to a smaller flow to finish the irrigation. By using this method
tailwater flows are reduced substantially. Both efficiency and exces-
sive percolation losses leading to salt build-up of return flows
are decreased.
Finally, an alternative whereby tailwater flows could be elimina-
ted completely is through collecting and repumping of surface runoff
waters. This represents an increase in cost over the cut—back alterna-
tive.
For flood, as with spreader-like irrigation, land grading will aid
the application uniformity and therefore increase water use efficiency
and salinity control capability. According to Evans (1978) application
efficiencies in Colorado for this type of irrigation is about 64 percent
with the possibility of 85 to 90 percent efficiencies under well managed
systems.
Sprinkler Irrigation
Irrigation by sprinklers is desirable because of the high unifor-
mity of application and excessive losses due to percolation and surface
runoff can be effectively minimized. Application efficiencies are
normally in the 80 to 90 percent range. This method has the additional
advantage that the leaching fraction may be reduced without serious
crop effects. Evans, etal., (1978) reported that studies at the U.S.
Salinity Laboratory have shown that the leaching fraction can be reduced
by one-fourth of present values without a reduction in alfalfa yields.
Because percolation losses are minimized, nutrient losses are reduced
also. In fact, some water soluble fertilizers can be applied directly
through the sprinkler system.
As a rule of thumb, with systems yielding increasing application
efficiency the water quality of the water used for irrigating should
be better. Particulates in water flowing through sprinkler systems can
cause corrosion and wear on piping and valves, and can clog sprinkler
nozzles, whereas this causes no problems in surface systems. Water of
poor quality can leave deposits on plants and cause burning which may
lower the aesthetic quality of the fruit or cause physiological effects.
Deposition of fine particles on the soil surface can also cause crust-
ing which may reduce infiltration rates and promote runoff from both
natural rainfall and sprinkler applications.
148

-------
In addition to better water quality requirements, sprinkler systems
involve a higher capital outlay and higher operating costs, if pumping
is required. Labor costs usually are substantially less and can be
made even more so because of the compatibility of sprinkler irrigation
and automation. Soil moisture sensors (tensiometers) have been used
with microprocessor systems to integrate soil moisture readings from
all over the field and automatically trigger an irrigation when soil
water tension falls below a specified level. This type of irrigation
naturally requires advanced scheduling techniques and technically qual-
ified operators.
Trickle Irrigation
Because of the types of crops grown in the area, trickle irrigation
does not seem to be a viable alternative for improving irrigation effi-
ciency and reducing salinity. It is normally used on widely spaced
crops (orchards, for example) and it is more cost effective for those
types of crops. At this time, perforated tapes are available such that
trickle systems have begun to be used for row crops. These are generally
high value crops such as tomatoes or other similar produce. The high
salt content of the Poplar River would make trickle irrigation difficult.
Operator Education
The best irrigation system or management practices for a given sys-
tem are only as good as the expertise of the operator allows. Normally,
systems which are capable of increased efficiences and therefore better
salinity control require increased levels of operator technical capa-
bility. Experience has shown that most operators know when to apply
water but not how much water to apply and many operators tend to over-
irrigate. Many states have county programs which educate operators in
scientific irrigation scheduling and water application. Such a program,
if it does not exist, is recommended if projected levels of irrigation
in the Poplar River Basin are realized.
5.3.5.3 Source Control of Salinity
Increased TDS concentrations can also be potentially controlled at
the source. There are two main alternatives:
• Control of Fife Lake overflows. The relatively high
salinity of Fife Lake overflows may exert considerable
influence on reservoir TDS concentrations. A poten-
tial control measure would involve diking to control
releases at Fife Lake during periods of maxirnun poten-
tial for downstream impact.
149

-------
• Prevention of discharge or treatment of ash lagoon
decant, ash lagoon seepage and mine dewatering
effi uent.
As discussed in Section 5.3.1 the routing of the ash decant to the
Cookson Reservoir increases the TDS and boron concentrations in the
upper East Fork significantly. Saskatchewan Power Corporation has pro-
posed to recycle the ash decant to the plant and minimize seepage from
the ash lagoons (IJC Hearings, 1979). The proposed plan would be to
compact 300 mm of glacial till to limit the seepage to less than 2 2./sec
of which an estimated 0.2 2,/sec would enter the reservoir. The remainder
(approximately 1.4 2./sec) enters the Empress Gravel aquifer and flows
toward the East Fork. Dilution of approximately 10:1 would occur from
underfiow from Cookson Reservoir and recharge from the glacial till based
on ground water modeling by SPC. The estimated boron concentrations with
two power plants in the upper East Fork would be 2.3 to 3.3 mg/ 2 . if no
ground water mounding occurred and 2.6 to 5 mg/2. if mounding occurred
(IJC Hearings, 1979). The latter concentrations are about 50 percent of
the values for scenario 8A. Thus, the impacts on crops of scenarios 4A
and 8A represent a “worst case”. With effective control of the ash
lagoon seepage the concentrations could be reduced nearly to the values
for scenarios 28 and 29.
150

-------
5.4 SOCIOECONOMIC IMPACTS OF POWER PLANT CONSTRUCTION
5.4.1 Introduction
This section describes the socioeconomic impacts that may occur in
the Poplar River Basin as a result of the construction of a series of
coal-fired plants by Saskatchewan Power north of the U.S. —Canadian
border. As in previous sections, the impact area for this portion of
the study will be in Daniels and Roosevelt counties. The potential
economic impacts that can occur during the construction phase are ex-
penditures on materials for the plant itself, and expenditures made
by the construction workers in the local area for goods and services
during their stay. Both of these sources of impacts are examined, and
an estimate is made of the amount of money spent annually in Daniels
and Roosevelt counties due to the construction project.
Construction of the first of Saskatchewan Power’s coal-fired
plants near Coronach, Canada began in August, 1975 and is expected to
be completed in the spring of 1980 (Mathew, 1979). The site is roughly
five miles southeast of the small town of Coronach, which is about
seven miles north of the U.S. -Canadian border. A second unit is ex-
pected to be constructed between the fall of 1979 and the winter of
1982.
5.4.2 Description of the Construction Work Force
The work force at the Poplar River Plant consists of workers from
various union contractors. The size of the construction crew has
ranged from about 50 at the beginning of the project to a peak of ap-
proximately 600 workers in October, 1978. Figure 5.4—1 shows the
fluctuation in the work force over the course of the project. In gen-
eral , the crew averaged about 450 to 500 workers during the major
portion of the construction, decreasing in the winter months when no
outside work could be done, and gaining strength during the rest of
the year. Between May and August, 1978, however, a strike at the plant
reduced the work force to almost zero (Cairnes, personal communication).
Most of the construction force lives in the construction camp
built on the site of the first unit. The camp consists of housing, a
dining hall, and a recreational center providing workers with a gym,
movies, television, pooi table, and other types of entertainment.
Workers who brought their families to the area cannot live in the camp
and live instead in Coronach and other small Canadian towns in the
plant’s vicinity (Mathew, personal communication).
No workers live across the border in the U.S. because of the re-
strictions on the hours that the border is open. Traffic can only move
across the border on Highway 13 between 9 A.M. and 6 P.M. Therefore,
there have been no direct impacts of plant construction on U.S. employ-
ment rates.
151

-------
650
600
, 550
500
450
o 400
I-
0
350
c c
I-
300
0
o 250
LI
0
200
150
100
50
0
SPRING SPRING
1975 1976 1977 1978 1979 1980 1980
YEAR BY QUARTERS
ESTIMATED HISTORICAL CONSTRUCTION WORK FORCE
PREDICTED FUTURE CONSTRUCTION WORK FORC
Source SRI International estimate using data from Ken Cairnes, telephone conversat ion, May 1979
Figure 5.4-1
ESTIMATED CONSTRUCTION WORK FORCE PROFILE, SASKATCHEWAN
POWER PLANT UNIT 1, 1975 THROUGH 1980
‘ I
I
I

-------
5.4.3 Economic Impacts from the Plant
No spending on plant materials and equipment has taken place in
Daniels or Roosevelt counties. Saskatchewan Power representative, Wayne
Nordquiest, estimated that approximately $50 million of the $190 million
total construction costs of Unit 1 were spent on parts and machinery in
the U.S.; however, none of this money went to companies in Daniels and
Roosevelt counties. Instead, the $50 million was spent on specialized
pieces of equipment that were purchased outside these counties.
5.4.4 Economic Impacts from the Construction Workers
Interviews with plant representatives and others indicate that the
construction workers have been purchasing many goods and services in
the U.S. (Bowler, Kenny, and Cairnes, personal coniliunications). In
spite of custom charges, many goods are less expensive in Scobey and
Plentywood. In addition, Coronach is a small town (population of 379
in 1971), so the workers find a greater selection of goods and less
wait for services in the larger nearby towns of Scobey and Plentywood
(populations of 2,041 and 3,126, respectively, in 1970).
No precise estimate of the amount of money spent by construction
workers in Daniels and Roosevelt counties is available. A rough esti-
mate, however, can be made, and the impact that these expenditures are
having on the counties can be estimated by examining county personal
income and retail sales data. The supervisor of one construction crew
estimated that his workers spend approximately $100 per week when they
spend the weekend in a nearby U.S. town. Of this, about $70 is spent
on a hotel room and entertainment, and the remaining $30 is spent on
food, clothing, and miscellaneous goods and services (Sooley, personal
communication). Sooley further estimated that approximately 15 percent
of the construction crews went to the U.S. on any given weekend. As a
basis of estimating impact, it is assumed that a maximum of 10 percent
of the workers go to Scobey, and the remaining 5 percent go to towns
such as Plentywood and Glasgow in nearby counties.
Using these assumptions and estimates, it can be calculated that
the range of expenditures the construction workers would make in one
year in Daniels County would be between $200,000 and $300,000. This
figure assumes a construction work force of 450 to 500. It is estimated
that Roosevelt County receives no expenditures from the construction
workers because the largest town in the county, Wolf Point, is about
75 miles from the construction camp.
As indicated in Table 5.4-1, personal income fluctuates sharply
from year to year in both counties, because of large yearly variations
in agricultural harvests and crop prices. The percentage change in
nonfarm income also fluctuates greatly, though not as much as farm
income.
153

-------
Table 5.4-1
TOTAL PERSONAL FARM AND NONFARM INCOME
DANIELS AND ROOSEVELT COUNTIES 1972-1977
Source: Montana Department of Community Affairs, Division of Research and Information
Systems, “County Profiles,” unpublished. Adjusted to constant dollars.
Total Labor and Proprietors’ Income
(thousands of 1975 dollars)
Year Total
______— Roosevelt
Nonfarm Total Farm
Percentage
Annual Change in
Nonfarm Income
Daniels Roosevelt
1972
1973
1974
1975
1976
1977
Daniels
Farm
$12,740
21 ,8O2
11,658
12,671
10,051
6,209
$18,737
28,407
18,360
19,836
17,446
13,591
$5,997
6,605
6,702
7,165
7,395
7,382
$41,138
59,559
38,122
42,940
38,707
31 ,248
Non farm
$25,846
27,233
26,359
27,774
28,146
30,237
$15,292
32,326
11,763
15,166
10,561
1,011
10.1%
1.5
6.9
3.2
-0.2
5.4%
-3.2
5.4
1.3
7.4

-------
Table 5.4-2 shows a similar picture for retail sales. Sales vary
considerably from one year to the next. The large decline in sales
between 1973 and 1974 in Daniels County is probably related to the
large decline in income. Farm income decreased dramatically in
Roosevelt County between 1976 and 1977 due to a substantial decline in
prices received by farmers for spring wheat. Neither table shows a
clear increase in income or sales that might be attributable to con-
struction worker expenditures.
5.4.5 Secondary Impacts
Construction workers spend a portion of their wages for goods and
services in Daniels County. The recipients of those payments treat
them as additional income and likewise spend some. Therefore, the
initial expenditure is actually respent a number of times throughout
the economy, generating secondary income, which is a multiple of the
initial injection.
Multipliers are calculated by the Bureau of Economic Analysis
(BEA) for 53 industrial sectors in each of the 173 BEA economic areas
of the U.S. Daniels and Roosevelt counties are in BEA area number 93.
Industrial sector number 54 (Trade) was chosen for these counties,
which has a multiplier of 1.85 for Daniels and Roosevelt counties,
according to the U.S. Bureau of the Census (January, 1977).
Simply applying the multiplier to the change in expenditures would
not accurately represent the impact, however, because of the inter-
dependence between the nearby counties. Spillover effects between
Daniels County, the area of direct impact, and other counties occur,
which can be calculated using a procedure developed by J.A. Chalmers,
et al., (1977). In this study, they noted a hierarchical relationship
between different kinds of counties. For example, certain counties,
designated level-3 counties, appear to serve as central market areas
for groups of other counties (level-2). Level-2 counties, in turn,
support groups of still lower-level counties and receive spillover
effects from them. Following Chalmers’ ranking scheme, we assigned
each county in the BEA area a rank of 1, 2, or 3.
As in Chalmers, all level-i counties in BEA area 93 are allocated
12 percent of the indirect impacts resulting from the direct expendi-
ture of $200,000 to $300,000. The 12 percent is then distributed to
individual level 1 counties within the BEA region by the ratio of the
population of the level—i county to the total population of all level-i
counties in the BEA region.
Using this methodology, secondary impacts can be calculated.
Daniels County receives approximately Sl6,000 to S24,000 annually in
secondary impacts. Roosevelt County, although it receives no direct
155

-------
Table 5.4-2
TOTAL RETAIL SALES
DANIELS AND ROOSEVELT COUNTIES
1973- 1977
Total Retail Sales
(thousands of 1975
dollars) Percentage Annual Change
Year Daniels Roosevelt Daniels Roosevelt
1973 $7,379 $23,864
1974 3,748 23,852 -49.2% -0.1%
1975 3,407 21,681 - 9.1 -9.1
1976 3,425 20,853 0.5 —3.8
1977 3,552 20,291 3.7 -2.7
Source: Sales and Marketing Management Magazine, Survey of
Buying Power (1974-1978) and adjusted to constant
dollars.
156

-------
impacts, receives secondary impacts of about $53,000 to $79,000 annual-
ly--a larger share than Daniels County because of wholesale and retail
purchases generated by activity in Daniels County. This procedure may
tend to overestimate the secondary impacts in Roosevelt County since Wolf
Point is so far from Scobey.
In summary, direct and secondary impacts combined total about
$216,000 to S324,000, or 6.3 percent to 9.5 percent of retail sales in
Daniels County in 1975. The total impact of $53,000 to $79,000 repre-
sents 0.2 percent to 0.4 percent of retail sales in Roosevelt County
represents a noticeable addition to sales, and may be preventing the
county from experiencing the decline in retail sales that occurred in
Roosevelt County. Both counties had declines in sales of about 9 per-
cent between 1974 and 1975. Daniels County however, recovered in the
following two years, whereas Roosevelt County continued to decline by
about 3 percent to 4 percent annually.
Expenditure changes can also be related to personal income. Using
the fact that income is roughly 0.3 percent of gross output, or sales,
it can be calculated that the total increase in sales resulting from
the construction workers amounts to about 0.3 percent to 0.5 percent of
personal income in Daniels County in 1975, and 0.9 percent to 1.4 percent
of nonfarm personal income for the same period. In Roosevelt County,
the increase in sales represents an increase of less than 0.1 percent
for both total and nonfarm personal income in 1975.
Although the revenue from the construction workers represents a
benefit that local communities such as Scobey will miss when the con-
struction is ended, this impact is far less significant to Daniels
County than the impact of changes in prices received by farmers. The
effect of wheat prices greatly overshadows any possible impact that
construction workers could have in this area.
157

-------
5.5 SOCIOECONOMIC IMPACTS OF APPORTIONMENT
5.5.1 Introduction
The apportionment of water from the Poplar River will affect the
amount of land irrigated in Daniels and Roosevelt counties and the qual-
ity of the irrigation water. This section estimates economic conditions,
from 1975 to 2000, and the impacts of Apportionment VI on farm income
and total income in Daniels and Roosevelt counties.
5.5.2 Future Conditions
The Poplar River Basin is a sparsely populated, rural area that is
unlikely to change significantly before 2000*. The only significant
development expected to take place over the next 25 years is the con-
struction of a potash plant, and the likely population influx associated
with the plant is included in the projections. Significant deposits of
lignite occur in the region of the Fort Peck Indian Reservation. Unpub-
lished data indicate that, under reasonable projections of future coal
demand, it is unlikely that these deposits would be developed and mined
before 2025. Compared with other coal resources, the lirinite is econom-
ically and cieoloqically unattractive (Yabroff and Dickson, 1979). However,
federal policies promoting coal development to reduce dependence on
imported petroleum could overcome the market constraints. It is possible
that some development could occur in the vicinity of McCone County by
1990 although there is substantial local resistance to large energy
developments in McCone County (Parfit, 1980). Two coal-related projects
investioatinq sites near Circle, Montana are the Basin Electric Power
Cooperative and the Circle West Project. The planned development by
Basin Electric would be two 500 MWe coal-fired power plants with the
first unit to begin operation in 1988. Basin expects to select a site
in early 1980 (personal communication - EPA, 1980). Sites being consid-
ered for the Circle West Project are in McCone County.
Table 5.5-1 shows population and employment projections for 1980,
1985, and 2000. The ratio of employment to population for the period
1971 to 1977 was used in estimating future employment, based on the pop-
ulation projections. Because no discernible trend was identified, future
employment was estimated as the average of the employment-to-population
ratio for 1971 to 1977. The average ratios for Daniels and Roosevelt
counties, respectively, were 0.47 and 0.42.
*The projections presented here are based on population projections
made by the Montana Department of Community Affairs.
158

-------
Table 5.5-1
PROJECTED POPULATION AND EMPLOYMENT
DANIELS AND ROOSEVELT COUNTIES -- 1980, 1985 and 2000
Population’ Employment 2
Year Daniels Roosevelt Daniels Roosevelt
1975 3,100 10,300 1,480 4,430
1980 3,100 10,700 1,500 4,500
1985 2,900 10,900 1,400 4,600
2000 3,400 11,500 1,600 4,800
Sources:
1. Montana Department of Community Affairs, Research
and Information Systems Division, “Montana
Population Projections, 1980-2000) (July, 1978)
2. SRI estimate
Future income levels in Daniels and Roosevelt counties were esti-
mated as the sum of wage and salary income, proprietors’ income,
property income, and net transfer payments. Wage and salary income
was projected using projections of future wage and salary employment
and 1975 wage levels. Income of nonfarm proprietors was based on the
number of proprietors times the average proprietors’ income for 1975.
Projecting the income of farm proprietors was complicated by
having to account for additional income because of the increases in
irrigated lands. Based on data of the USDA Cooperative Extension
Service, conversion of existing dryland acreage to irrigated acreage
was estimated to increase net incomes by $40 to $50/acre (Luft, 1979;
Luft and Griffith, 1978). To be conservative and to account for pos-
sible yield and technological changes over the next 20 years, $50!
irrigated acre was used to project the additional agricultural income
expected in the two counties.
The change in irrigated acreage for 1985 and 2000 is taken as the
difference between irrigated acreage in the Poplar River Basin in 1975
and the desired irrigated acreages for 1985 and 2000. The projections
for 1985 and 2000 are based on the completion of two reservoirs on the
Poplar within the Fort Peck Indian Reservation. The combined storage
capacity of the reservoirs is app’oxiniately 152,400 ac-ft. It was
assumed that irrigated land would be converted from existing unirrigated
cropland. Because most cropland in Daniels County is within the Poplar
River Basin, no other factors are likely to affect agricultural income.
In Roosevelt County it was assumed that conditions outside the Poplar
River Basin would remain constant. This assumption will tend to over-
estimate impacts because the projected conditions in Roosevelt County
do not allow for additional irrigation or conversion of pasture and
range to cropland.
159

-------
Other sources of income, rent, interest, and net transfer payments
were estimated from their historical relationship to the total of wage,
salary, and proprietors’ income. The results are summarized below in
millions of 1975 dollars.
Farm Proprietors’ Total Personal
Income Income
Daniels County: 1975 $11.9 $25.5
1985 12.0 24.7
2000 12.1 26.8
Roosevelt County: 1975 13.6 55.6
1985 14.1 56.2
2000 14.6 59.2
In constant dollars, the projected gains are quite modest in
Daniels County. Income of farm proprietors is projected to increase
by only 1.7 percent between 1975 and 2000 and total personal income by
5.1 percent. In Roosevelt County, the gains are somewhat larger, but
hardly robust: 7.4 percent for income of farm proprietors and 6.5 per-
cent for total personal income.
These projections reflect the historical experience of agricultural
income in the region. The changes in income result from changing com-
modity prices--especially wheat. Real growth over the long term has
been slight. The outlook for the future is for continued fluctuations
but for no dramatic increases.
5.5.3 Impacts on Income
The impacts of the apportionment scheme on Montana agriculture in
larue part depend on the amount of land that is developed for irrigation.
Projections of irrigated acreage that are desired by Montana interests
are oresented above, and will be used as the basis for estimating im-
paces. The amount of irrigated land that is likely under the apportion-
ment will depend on the farmers’ perceptions of the risks and returns
invol ved.
Installation of a sprinkler irrigation system requires an invest-
ment of between $160/acre for a hand-move system and $340/acre for a
circular self-propelled system (Luft, 1979). Side-roll wheel-move sys-
tems can be installed for $250 to $270/acre (Luft, 1979). Analysis of
the side-roll wheel-move systems indicates that additional income of $46!
acre is possible with irrigated alfalfa. Conversion of dryland wheat to
irrigated wheat would result in additional income of $42.60/acre (Luft,
No Date Given).
160

-------
Side-roll irrigation systems are used as the basis for calculating
impacts because they provide a convenient middle ground between the ex-
tremes of hand-move and center-pivot systems. Buying an irrigation
system requires that the farm operator assume the risk of no return on
his investment in a dry year. He must therefore be convinced that over
the long run his investment will be profitable. That is, his returns
in years with sufficient water must exceed carrying costs in dry years.
The additional fixed costs of a side-roll irrigation system are approx-
imately $35/acre. If, as discussed above, the additional income from
irrigated wheat is $42/acre, a farmer could invest in an irrigation
system based on the mean flows from the Poplar River. In years when
flows are less than the mean, fixed costs could be covered as long as
yields do not drop more than 16 percent. However, a definite risk of
bankruptcy is encountered if several dry years occur consecutively. The
risks with irrigated alfalfa are slightly less, allowing on the average
a margin of Sli/acre over fixed costs which is equivalent to 24 percent
of average yields.
Based on the foregoing, it is assumed that farmers in the Poplar
River Basin will install irrigation systems on acreage irrigable with
the mean flows expected under the apportionment. This implies making
use of the acreage irrigable with the mean flows for June in 1985--13,573
acres. Moreover, it assumes that farmers can recover their investment
before the fourth power plant is completed, causing mean flows for June
to drop so that only 10,936 acres can be irrigated in 2000.
The preceding line of reasoning presents a plausible scenario for
estimating impacts. An investment of approximately $3.0 million in irri-
gation systems would be required for the more than 10,000 acres in the
Poplar River Basin. Given the region’s long history of dryland farming,
this constitutes a large, risky investment. The cautious farmer will
wait to ascertain how the apportionment affects flows before he decides
to make such an investment. However, in estimating impacts, the assump-
tions above provide a conservative approach.
The apportionment will cause a change in the irrigable acreage in
Daniels and Roosevelt counties (Table F-i) and a change in yields on
irrigated lands because of lowered water quality. The income loss caused
by the change in acreage is estimated as the difference between net in-
come for irrigated crops and net income for dryland farming of the same
crop. For the impact calculations this difference is assumed to be $50!
acre. This is higher than the increases in net income estimated by Luft
for alfalfa and wheat; however, it does allow for the possibility of
higher wheat and alfalfa prices in the future.
The increase in net income for irrigated crops is assumed to be
earned only when irrigation is possible throughout the growing season.
Therefore, the change in irrigable acreage is calculated using irrigable
acres for mean flows in August. The irrigable acres shown in Table F-i
161

-------
are based on water requirements and losses. In practice, additional
water may be applied to leach salts from the soils. It is assumed that
20 percent more water than is needed for plant growth will be used for
leaching. Table 5.5—2 shows the changes in irrigable acres and income
due to apportionment.
Income changes due to water quality are shown in Table 5.5-3 under
median water quality and rainfall conditions. Yield decreases up to 6
percent would be predicted if the leaching fraction decreased to 0.1 in
Roosevelt resulting in an estimated decrease in income of $6,798 for
1975 and -$116,794 for 1985.
Table 5.5-3
CHAr GE IN YIELD AND PER-ACRE REVENUES FOR WHEAT
Irrigated
Total Yield Per-Acre Acres Income
Location Scenario* Change Change Affected Change
Daniels 1975 (1 plant) -11% —$24 1,391 -33,400
County
1985 (2 plants) -20 — 44 1,520 -66,900
Roosevelt 1975 (1 plant) +8% +$17 515 +20,600
County
1985 (2 plants) +8 + 17 8,848 +150,400
*1975 is model scenario 4A. 1985 is model scenario 8A.
The change in income resulting from water quality changes depends
upon the sensitivity of a crop to boron and other salts and the mix of
crops planted. Of the four major crops in the area, alfalfa is the most
tolerant and oats the least tolerant. Wheat and barley are intermediate
in tolerance. The yield losses expected for wheat are less than those
for barley under similar conditions; however, because wheat has a higher
price, the income losses would be greater (on a per-acre basis) for
wheat. Water quality-caused reductions in yield are worst on the East
Fork. In Roosevelt County the water quality changes are underestimated
by the model so the yield could be less than predicted.
The change in farm income will result in secondary income changes
in other sectors of Daniels and Roosevelt counties, and in nearby trading
centers that serve residents and businesses in the two counties. BEA
Area 93 has a multiplier of 1.629 for the field crop sector in which
wheat and alfalfa fall. This is used to estimate the total changes in
personal income. Then, taking into account the hierarchical relationship
among counties in the BEA area, the total impacts are apportioned among
counties. (See the explanation in section on construction period impacts
and Chalmers, etal., 1977.) Table 5.5-4 summarizes the impacts on total
personal income.
162

-------
Table 5.5—2
CHANGE IN FARN INCOME RESULTING FROM APPORTIONMENT ONLY
Estimated and
Projected Irrigated Mean Acreage Difference Change as % of
Acres Affected by Irrigable in Irrigated Income Farm Proprietor’s
Project in August Acres Change Income
Daniels County
1975 2,826 1,390 —1,436 —$71,800 —0.6%
1985 3,575 1,519 —2,056 —102,800 —0.9
2000 5,342 1,521 —3,821 —191,050 —1.6
Roosevelt County
1975 515 515 0 0 0
1985 8,848 8,848 0 0 0
2000 17,182 3,488 —13,694 —684,700 —4.8

-------
Table 5.5-4
IMPACTS OF APPORTIONMENT AND WATER QUALITY ON PERSONAL INCOME
IN DANIELS AND ROOSEVELT COUNTIES
(Dollar Figures are in Thousands of 1975 Dollars)
County Change in Impact as
and Change in Induced Income Total Change Percent of
Year Farm Income in County in Personal Income Personal Income
Daniels
1975 $-105.2 $ -7.9 S-113.1 -0.4%
1985 -169.7 -12.8 -182.5 -0.7
2000 -258.0 -19.5 -277.4 -1.0
Roosevel t
1975 $ 20.6 $ 1.6 $ 22.2 *
1985 150.4 11.4 161.8 0.3%
2000 59.3 4.5 63.8 0.1
*Less than 0.05 percent.
In both Daniels and Roosevelt counties the secondary income changes
plus the change in farm income amount to 1 percent or less of personal
income in 2000. The losses in Daniels County are partially offset by
gains in Roosevelt County but the net impact is negative. The changes
estimated are far less than historical variations in personal income
that arise from weather variations.
Approximately 50 percent of the secondary impacts may flow to Minot,
North Dakota (Ward County), which is the major trading center for BEA
Area 93. This would have a minimal impact amounting to less than 0.1 per-
cent of personal income in Ward County in the year 2000.
The foregoing has presented an estimate of impacts that might result
from apportionment of the water flows in the Poplar River; however the
analysis is influenced by many underlying assumptions. Some of the more
important figures that have been included are the selling price of wheat
and alfalfa, the cost of water to farmers, the value of cropland, the
cost of sprinkler systems, and farm operating costs. In fact, the number
of possible combinations that could be considered is unwieldy. There-
fore, to indicate the upper range of possible impacts, the worst-case
approach was used. That is, it was assumed that the number of irrigated
acres affected by the apportionment was equal to the total projected
irrigated acreage affected by the power plants in the Poplar River Basin.
This is equivalent to cutting off all irrigation waters, leaving on’y
dryland farming in Daniels and Roosevelt counties--a possibility in ex-
tremely dry years or if water quality were reduced to harmful levels.
The impacts on farm income resulting from the conversion of all irrigable
lands to dry land farming are shown in Table 5.5-5.
164

-------
Table 5.5-5
MAXIMUM POSSIBLE IMPACTS ON FARM INCOME
Percent
County Change in Change in Farm Change of Farm
and Irrigated Lands Proprietors’ Income Proprietors’
Year ( acres) ( thousands of 1975$) Income in Region
Daniels
1975 -2,826 $ -140 -1.2%
1985 —3,575 -180 -1.5
2000 —5,342 -270 -2.2
Roosevelt
1975 -515 -26 -0.2
1985 —8,848 -440 -3.1
2000 -17,182 -860 -6.0
Total s
1975 —3,341 —170 -0.7
1985 -12,423 -620 -2.4
2000 -22,524 -1,130 -4.3
The largest absolute and percentage impacts would occur in Roosevelt
County where farm income could be reduced by 6 percent in 2000. Within
boundaries of the Fort Peck Indian Reservation, the impacts would be much
higher on a percentage basis. However, all the Indians do not own and
operate farmland there. Gross value of products (crops and livestock)
cirown on Indian lands by Indian operators is estimated to be less than
$10 million. In the case of no irrigated lands, impacts on total personal
income would be less than 2 percent of personal income in 2000. Spill-
over of secondary impacts into other regions would be negligible on a
percentage basis.
The preceding discussion of impacts has presented the worst case
situation. The actual changes in income after the apportionment may be
less than indicated above under the mean flow conditions. After con-
struction of the reservoirs more land can be irrigated in 1985 and 2000
than in 1975. The average number of irrigation applications per year
has been 2.4 so that even under present flow conditions water in August
is not adequate to irrigate the full number of acres. Thus, assuming
the loss of the full premium is very conservative.
5.5.4 Other Impacts
The apportionment could affect the values of irrigated land if water
is no longer available. Conceivably, between 1935 and 2000 land values
165

-------
could drop as the amount of irrigable land declines. Based on current
crop budgets for the region, the premium for irrigated land is between
$50 and $100/acre (Griffith, et al., 1978; Shaefer, et al., 1978A and
1978B).
A related impact is the loss of the undepreciated portion of any
investment in irrigation equipment. For new systems, the loss could be
more than S200/acre.
The impacts of the apportionment on farm income and total personal
income are unlikely to result in noticeable changes in population. If
irrigation water is unavailable, farmers will resort to what they know
best--dryland farming. The trend to fewer and larger farms would not
be affected by the changes estimated here. An increase in demand for
seasonal farm workers is likely to occur if all the growth in irrigated
lands takes place as projected. The converse would occur if the amount
of irrigated land, and, therefore, crop yields are diminished. It is
difficult to estimate the number of seasonal employees affected because
no data exist regarding the current seasonal farm employment available.
However, in Daniels and Roosevelt counties together fewer than 300 full
and part-time workers were employed in agriculture in 1977. The down-
ward trend in agricultural employment is expected to continue; therefore,
in 2000, less than 5 percent of total county employment will be in agri-
culture. As a result, the likelihood of significant impacts on employment
is small.
166

-------
5.6 BIOLOGICAL IMPACTS
5.6.1 Impacts of Atmospheric Emissions on Terrestrial Biota
5.6.1.1 Gaseous Emissions
The effects of primary gaseous pollutants on terrestrial vegeta-
tion and animals can be categorized in terms of exposure duration and
dosage. Three broad categories can be distinguished: acute, chronic
and long-term. Acute effects are those which result from short-term
exposures to relatively high concentrations of pollutants. Observable
effects which result from exposures to comparatively low levels of
pollutants for periods of months to years are referred to as chronic,
while the effects of exposure to pollutants for decades or longer are
classified as long—term.
Acute effects result from direct injury to the biota and are typ-
ified by leaf damage and lesions in an animal’s respiratory tract or
other serious debilities. While chronic effects may be expressed
similarly to those which have been described as acute, more often they
are expressed as subtle changes in the ecosystem. These subtle
changes, moreover, may represent a change in the species composition
of a well defined natural vegetation assemblage due to the loss of a
pollutant sensitive species, or they may represent a decrease in num-
bers of wildlife in a particular area due to concomitant habitat dep-
redation. Long-term effects on the other hand are the result of
secondary or tertiary interactions and are always manifested as subtle
changes in the ecosystem or an organism’s susceptibility to respiratory
diseases.
Sulfur dioxide (SO 2 ) is the most abundant gaseous pollutant pro-
duced by coal-fired electrical generating stations, and is therefore
of primary concern when considering the impacts on terrestrial eco-
systems. Nitrogen oxide emissions (NOx) represent another important
pollution source, but because expected concentrations produced in the
vicinity of the plant are generally anticipated to remain below acute
and chronic injury threshold levels, they are of secondary importance.
In addition to considering the impacts associated with the emission of
these common pollutants and their singular effects, the additive or
possible synergistic effects must also be addressed.
Reported threshold limits of selected species to acute and in
some cases chronic levels of SO2, NO and combined gaseous emissions
are provided in Table 5.6-1; included in this list are important
species of grasses, forbs and cultivated crops found in the Poplar
River impact area. These empirically determined threshold values rep-
resent upper levels of exposure at which no visible injury to the
plants was observed. The values presented in Table 5.6-1 are greater
than reported acute threshold levels listed for the most sensitive
167

-------
Table 5.6-1
REPORTED THRESHOLD LIMITS OF IMPORTANT NATIVE PLANT AND CULTIVATED
SPECIES FOUND IN THE IMPACT AREA TO GASEOUS (SO 2 , NOx AND SO 2 + NOx
EMISSION EXPOSURES
THRESHOLD LIMITS
(uo I JURY OBSERVED)
Dose Period of
Gaseous Pollutant Exposure
conti nuous
502 0.2 ppm - during growing
season
NO 1.0 ppm — 4 hrs
SO 2 + NO, (0.3 ppm + 0.1 ppm) - 4
Arteusia frigida
(finged sagewort)
502
NO
502 + NO,
1.0 ppm — 4 hrs
4.0 ppm - 4 hrs
(0.6 ppm + 0.1 ppm) -
Tingey. ! .t .L.. 1978
Tingey. etal.. 1978
4 hrs Tingey. etal.. 1978
Koeleria cr.xscata
(praire june grass)
Sc2pa comata
(needle and threadgrass)
502
SO 2
NO
SO 2 + NO
SO 2
SO 2
NO
SO 2 + NO
1.0 ppm - 4 hrs
1.2 ppm - 3 hrs (biweekly)
1.0 ppm - 4 hrs
(0.3 ppm + 0.1 ppm) - 4 hrs
1.5 ppm — 4 hrs
1.2 ppm - 3 hrs (biweekly)
1.0 ppm - 4 hrS
(1.2 ppm + (1.1 ppm) - 4 hrs
Tingey. et al. 1978
Wilhour, et at • 1979
Tingey. et al. • 1978
Tingey. !! !L. ’ 1978
Tingey. !!. L-. 1978
Wilhour , !S. J_. 1979
Tingey, !! AL .. 1978
Tingey, et al.. 1978
Medlcac7o sativa
(Alfalfa)
Stevens and
Hazelton, 1976
Stevens and
Hazelton, 1976
Stevens and
Hazelton. 1976
Wilhour, et aL, 1979
Drei Si nger
and McGovern, 1970
Drei singer
and McGovern, 1970
Drei singer
and McGovern, 1970
Drei singer
and McGovern. 1970
Wilhour, et al. • 1979
Tingey, £! i. . 1978
Wilhour, et al. • 1978
Tingey. ! _L . 1978
Tingey. et al. • 1978
Snec i es
Reference
soutelova oracilis
(blue grama)
SO 2
so 2
NO
502 + N0
1.2
(0.3
0.5
ppm —
0.5
ppm +
ppm — 4 hrs
3 hrs (biweekly)
ppm — 4 hrs
0.1 ppm) 4 hrs
Tingey,
WilPiour,
Tingey,
Tingey,
et al.,
et al. •
et al.,
etal.,
1978
1979
1978
1978
Agropyron
(Western
smithii
wheatgrass)
502
502
1.2
1.5
ppm -
ppm - 4 hrs
3 hrs (biweekly)
Tingey,
Wilhour,
et al..
et al. •
1978
1979
Dodd, !! .i. . 1978
Tingey, et al. , 1978
hrs Tingey. !! i•. 1978
502 1.15 ppm — 1 hr
502 0.50 ppm - 3 hrs
SO 2 0.25 ppm - 24 hrs or more
502 1.2 ppm - 3 hrs (biweekly)
502 0.70 ppm - 1 hr
SO 2 0.46 ppm - 2 hrs
SO 2 0.27 ppm - 4 hrs
SO 2 0.14 ppm - 8 hrs
SO 2 1.2 ppm . 3 hrs (biweekly)
Hordoum vuigare
(Barley)
Tritcum aest.’.,zum
(hyslop wneat)
SO 2
So 2
Nox
SO? + NO
0.6 ppm - 4 hrs
1.2 ppm - 3 hrs (biweekly)
2 0 ppm - 4 hrs
(0 6 ppm + 0 1 ppm) - 4 hrs
168

-------
plant species which are presented graphically in Figure 5.6-1. The
concentrations and corresponding exposure shown in Figure 5.6-1, how-
ever, represent threshold levels for plants growing under the most
sensitive conditions and stage of maturity. While such guidelines
for chronic exposures are not available, it has been estimated that
an ambient S02 concentration of 130 pg/rn 3 (0.05 ppm) on a seasonal or
annual average basis represents the threshold level (Mukammal , 1976).
Studies were conducted at the Corvallis Environmental Research
Laboratory during the spring and summer months of 1977 to investigate
the effects of both chronic and multiple S02 exposures on selected small
grains, native range grasses and alfalfa growing in the Poplar River
Basin (Wilhour, etal., 1979). The results indicated that the yield
of duram wheat and barley may be substantially reduced by weekly 72-hr
exposures to SO 2 concentrations of approximately 0.15 ppm (400 pg/rn 3 )
and that spring wheat, while more resistant to chronic exposures, may
also suffer decreased yield due to SO2 exposures at similar levels.
Multiple exposure of these species as well as alfalfa to frequencies
varying from once per week to once in five weeks with SO 2 concentrations
up to 1.2 ppm (320 pg/rn 3 ) for 3-hr periods, however, had no effect on
the yield of the plants. Finally, biweekly exposures for 3-hr at con-
centrations up to 1.2 ppm (3200 pg/rn 3 ) had no effect on the growth of
either the roots or tops of alfalfa or the following five species of
native grasses: crested wheatgrass (Agropyron desertorum) , western
wheatgrass (Agropyron smithii) , Russian wild ryegrass (Elymus junceus)
blue gramma grass (Bouteloua gracilis) and needle and thread grass
(Stipa comata).
The combined effects of gaseous emissions, especially the inter-
action of SO 2 and NOx, on vegetation are not well understood, and
some controversy exists concerning possible synergistic or additive
effects. Studies conducted in growth chambers (Bennett, et al. , 1975)
indicated that SO 2 and NO applied in combination may enhancithe
deleterious effect of these pollutants; relatively high doses of these
pollutants, however, were required to cause visible injury. For exam-
ple, one hour exposures to a 0.5 ppm S02 + NQx (960 pg/rn 3 NOx and 1301
pg/rn 3 SO 2 ) mixture or to 0.75 ppm (1950 pg/mi) were required to cause
visible :Foliar injury in the most sensitive species. Other studies
(Tingey, etal., 1978 and Hill, etal., 1974) have been unable to sub-
stantiate that mixtures of SO 2 + NO interact to cause greater foliar
injury than equal concentrations of single gaseous pollutants. The
acute tolerance levels of several plant species found within the im-
pact area to mixtures of SO 2 + NO are indicated in Table 5.6-1.
Based on the conservative projections provided by the CRSTER air
quality model, gaseous emissions resulting from the operation of the
proposed generating station will not have a detectable impact on the
terrestrial ecosystem. This analysis is based on the specific consid-
erations outlined below, but generally is the result of the absence
of industrial and urban development within the Poplar Basin and the
concomitant excellent baseline air quality conditions. As previously
169

-------
Figure 5.6-1
‘I
a
0
1
Y0R0AMA
0 I 2 3 4 5 6 7 8 -
DURATION OF EXPOSUREli,
(b)
I0
a G
I- )
a 7
0
0
4
3
2
(c)
8340
15720
13100
A
0480 a.
‘a
a
0
‘a
1860
5240
2620
0
26200
23580
20960 a
18340
15720 ‘ .
0
13100 ‘
10480
7860
5240
2620
a
DURATION OF EXPOSURE. Ii ,
(a)
INJURY OR DAMAGE
POSSIBLE
I I I I I I I
0 I 2 3 4 5 6 7 8
DURATION OF EXPOSURE. Dr
Source: U.S. EPA,
1973
DOSE-INJURY CURVES FOR (a) S0 2 -SENSITIVE PLANT SPECIES,
(b) PLANT SPECIES OF INTERMEDIATE SO2 SENSITIVITY, AND
(c) SO 2 -RESISTANT PLANT SPECIES.
170

-------
noted, however, the CRSTER model does not incorporate fumigation events
in estimating pollutant concentrations. Estimated maximum ground-level
concentrations of gaseous emissions during morning plume fumigation
for the worst case atmospheric conditions are of the same magnitude as
concentrations associated with acute injury levels for the most sensi-
tive plant species. The possible impact of estimated peak concentrations
resulting from plume fumigation are discussed in a subsequent section.
The area of primary concern for the evaluation of emission impacts
is that region within a 50 km (31.1 miles) radius of the plant site.
Assuming the operation of four 300 MW units, the yearly maximum 24-hour
concentration of SO 2 predicted by the air quality model to occur at a
distance of 50 km is 22.8 1g/m 3 or 0.008 ppm; this value is below the
minimum detectable limits of baseline measurements in this area. The
projected maximum 1-hour SO 2 concentration at this distance at this
same electrical generating capacity is 169 119/rn 3 or 0.06 ppm, and this
projected concentration is an order of magnitude less than reported
concentrations associated with acute injury threshold levels for sen-
sitive vegetation. The projected ambient concentrations of NO for
this region are similarly below threshold injury levels for sensitive
native plants and crops. These projected concentrations for SO 2 and
NOx are of particular importance since the air quality model is espe-
cially conservative at this distance from the point source. The region
described here as the area of primary concern is shown in Figures
5.1-1 through 5.1-4; it lies within Daniels and Valley Counties and
represents a total area of approximately 700 km 2 (270 square miles).
That portion of the impact area continuously exposed to the high-
est concentrations of gaseous emissions and therefore most likely to
display the acute and chronic effects of exposure to SO 2 and NOx is
within 6-15 km (3.7-9.3 miles) of the stacks. The yearly maximum 1-,
3- and 24-hour ambient SO 2 concentrations projected for any area within
this range (for a 1200 MW 2laflt with 0 percent SO 2 control) are 464
ig/m 3 (0.18 ppm), 274 jg/mi (0.10 ppm) and 50 pg/m 3 (0.02 ppm). The
projected maximum annual mean SO2 concentration at this distance is
3.064 ig/m 3 (<0.01 ppm). The projected SO 2 concentrations as well as
those of NO are considerably less than those associated with acute
and chronic threshold injury levels for sensitive vegetation. More-
over, the projected levels of these pollutants in combination are be-
low reported concentrations associated with combined or synergistic
effects. As previously noted, experiments using the most sensitive
native grasses found that 4-hour exposures to combined concentrations
of 0.3 ppm SO 2 and 0.1 ppm NO 2 were insufficient to induce foliar in-
jury (Tingey, etal., 1978).
The preceding comoarisons which indicate the absence of injurious
levels resulting from the Poplar River plant were all based on the
maximum projected emissions at the site, i.e., operation of four 300 MW
units with 0 percent SO 2 control. The maximum concentrations occurring
near the plant site were also considered. The actual concentrations
of gaseous constituents, and the corresponding potential for foliar
injury, would be considerably less due to the following factors:
171

-------
1) Operation of fewer than four units would result in
correspondingly lower emission rates.
2) The use of stack emission controls would result in
large reductions in ambient SO 2 concentrations.
3) The ambient concentrations of all constituents
decrease rapidly with distance from the site.
Therefore, most of the Poplar basin would be exposed
to maximum projected concentrations greater than
two orders of magnitude below the tolerance levels
for local agricultural plant species.
Areas not subjected to acute or chronic SO 2 exposure can neverthe-
less be affected by long-term, low level ambient SO 2 concentrations.
The most serious known effect of long-term exposures to elevated sulfur
levels is the acidification of soils (Nyborg, 1978). Additionally,
recent studies (Rice, et al. , 1979) suggest that a long-term buildup
of sulfur in plants could have a detrimental effect on crop yield and
quality. However, no studies have been conducted to determine a
relationship between excessive sulfur accumulation and plant damage.
The acidification of soils is associated with the deposition of
H 2 SO 4 in precipitation (acid rain) as well as the formation of H 2 SO 4
in the soil following direct dry deposition of SO 2 . The effects on
soils of increased acidity are: an increased solubility of aluminum
(Al) and manganese (Mn); deficiency in soil concentrations of calcium
(Ca) and magnesium (Mg); and increased hydrogen ion (H+) concentrations.
These changes in soil chemistry can adversely affect plant growth and
yield. The primary manner in which soil acidity affects plant growth
is through aluminum toxicity which begins to occur at a pH around 5.0 -
5.5. Calcium and magnesium deficiencies are primarily a problem associ-
ated with forest soils which are naturally acidic. Increased H+_ion
concentrations are generally toxic to plant roots only when soils are
extremely acidic (pH = 4.0 to 3.0). Nitrogen fixation by symbiotic
bacteria associated with alfalfa is inhibited at pH levels less than
6.0.
While the accumulation of sulfur can have deleterious effects, sul-
fur is also an essential nutrient. The amount of sulfur needed for
medium to high crop yields ranges from about 10 to 40 kg/ha yr-i (Noggle
and Jones, 1979). At least a portion of a plant’s sulfur requirements
can be met by direct uptake of SO 2 from the atmosphere if present at
low concentrations (Faller, 1971; Bromfield, 1972; Cowling, et al.,
1973; Noggle and Jones, 1979). Beneficial effects r f ambient low level
SO 2 concentrations on plant growth when soil concentrations are inade-
quate have also been demonstrated (Faller, 1970 and 1971). In addition,
sulfur in the form of calcium sulfate (gypsum) is often applied to soils
to change part of the caustic alkali carbonates into sulfates.
172

-------
An assessment of the impacts of sulfur deposition on soils must
consider not only SO 2 deposition rates but also buffering capacity of
the soils and sulfur loss through cropping and leaching. Sulfur emitted
to the atmosphere by a power plant can be introduced into the terres-
trial ecosystem by wet and dry deposition. There are two methods of
estimating the rate deposition of sulfur. The first is based on the
assumption that the total amount of sulfur in stack emissions will be
equally distributed within a specified area. A second method is based
on a calculation of deposition rate as the product of an annual mean
SO 2 concentration and an estimated deposition velocity (Fowler, 1973;
Garland, 1978).
The first method was used by the Montana Department of Health and
Environmental Sciences (MDHES) to predict sulfur deposition within the
Poplar River Basin (Geihaus and Roach, 1979). The estimate was based
on the following assumptions: the operation of a 300 MW plant with zero
percent SO 2 control, daily SO 2 emissions of 5.84 . iO kg (6.4 tons),
60 percent deposition of sulfur emissions within a 40 km radius of the
source, and equal deposition throu9hout the affected area. The predicted
deposition rate was 25.5 kg/ha yr—I of SO 2 or approximately 12.8 kg/ha
yr-i of elemental S. An annual deposition rate of 12.8 kg S/ha for the
300 MW plant is equivalent to deposition rates of 25.6 and 51.2 kg S/ha
yr-’ for the 600 and 1200 MW plants, respectively.
The second method focuses on the contribution of dry deposition to
total annual inputs of sulfur. Measurements of the deposition velocity
ranne from 0.3 cm/sec to 1.25 cm/sec over soil, crop and other vegetation
surfaces (Garland, 1978). These deposition velocities are substantially
greater over vegetative surfaces during periods of maximum growth as well
as during daylight periods. In order to estimate a dry deposition rate
of SO 2 within the Poplar River Basin, conservative values for both the
mean concentrations as well as deposition velocities were selected.
Accordinoly, the annual mean concentration of SO 2 was taken to be 4.2 ig/
m 3 (the maximum annual average SO 2 concentration predicted by the CRSTER
model for a 1200 MW power plant with zero percent S02 control). The
selected deposition velocity (0.8 cm/sec) is the estimated value for day-
light periods during the growing season. Based on these parameters
the estimated annual dry deposition rate is 10.6 kg/ha S02 or 5.3 kg/ha
elemental S.
Wet deposition represents another mechanism for transport of atmos-
phere sulfur to the soil. However, given the low levels of precipitation
in the Poplar River Basin as well as the low ambient concentrations of
SO 2 , expected sulfur inputs associated with precipitation would be less
than dry deposition rates and will most likely make only a fractional
contribution to total sulfur deposition.
The underlying assumptions of the two estimates of annual sulfur
deposition described above have a substantial effect on the predicted
values (51.2 kg S/ha yr—i for a 1200 MW plant in the first estimate ver-
sus 5.3 - 10.6 kg S/ha yr in the latter). The first estimate assumed
173

-------
60 percent deposition within 40 km of the source. However, the rate of
dry deposition (the most important source in semi-arid regions) in an
area surrounding the emission source is related to the product of the
inversion height and the ratio of the mean wind speed to the deposition
velocity. Concentration decay distances of several hundred kilometers
have been reported (Scriven and Fisher, 1975). The result of this sim-
plified accounting of total deposition is a very conservative estimation
of annual sulfur deposition on a per unit area basis.
Sulfur can also be removed by cropping following plant uptake, by
the leaching of soluble sulfates, and by surface drainage. For example,
Likens, ai.. (1967) found that the loss of sulfur (9.8 k S/ha) in the
drainacie for a catchment area in Flew Hampshire was approximately equal
to sulfur inputs associated with precipitation (10 kg/ha). Plants them-
selves provide a sink for sulfur emissions. A portion of the sulfur
that is in a form available to plants during the growing season (the
soluble inorganic sulfate, adsorbed sulfate) is assimilated by the vege-
tation. As much as 50 percent of this organic sulfur can be removed
from the system by cropping, while the remainder returns to the soil as
organic material. This organic material again becomes available to the
sulfate pool through mineralization, but this is a slow process and not
rapid enouch to promote maximum plant growth.
The nature of the soils also affects the degree of environmental
impact. The soils in the Poplar River Basin are moderately calcareous;
(average — 4 percent calcium equivalent on a weight basis, Smetana (per-
sonal communication)). While soil acidification can be a problem in
areas exposed to SO 2 emissions, the impact is minimized by calcareous
soils (Flyborg, 1978). Based on the buffering capacity of soils in the
impact area assuming a 4 percent CaCO 3 concentration, a soil depth of
50 cm, and the deposition of 50 kg/ha of elemental sulfur and the subse-
quent conversion of this total amount of sulfur to H 2 S0 4 no significant
change in soil pH would occur.
Plume fumigation may occur during certain stable atmospheric condi-
tions and result in the exposure of vegetation to substantially higher
than usual concentrations of gaseous emissions. The estimated maximum
exposure levels of S02 and N0 emissions are presented in Table 5.1-4.
The estimated concentrations of NOx emissions are below reported thres-
hold levels for exposures of short duration, but S02 levels expected to
occur 10-20 km downwind of a 1200 MW plant during fumigation are within
the lower limits of the threshold range associated with acute injury to
sensitive vegetation.
The concentrations of SO 2 expected to occur during fumigation under
the most stable atmospheric conditions and maximum power generating
capacity range between 1.0 and 1.5 ppm (Table 5.1-4), while concentra-
tions associated with average meteorological conditions at Scobey are
less than 0.7 porn (Table 5.1-5). These concentrations have been shown
174

-------
to cause foliar damage to some of the most sensitive plants when sub-
jected to short-term exposure experiments; however, the sensitivity of
these plants were determined in experiments in which the duration of
exposure was one or more hours (U.S. EPA, 1973). The predicted fumi-
gations are generally of short duration, persisting for periods up to
30-45 minutes (Portelli, 1975). Comparison of the estimated maximum S02
concentrations expected during fumigation with reported threshold limits
of important native and cultivated plants found in the impact area
(Table 5.6-1) indicate that the risk of S02 damage is minimal.
During fumigation events vegetation will be simultaneously exposed
to elevated concentrations of NOx, 03 and S0 . As an indication of the
increased susceptability of vegetation to injury from combinations of
gaseous emissions, a review of the literature (Linzon, 1978) found that
a low concentration of 0.10 ppm SO 2 in combination with either 03 or
NOx for periods of 4 hrs can injure a wide variety of plants. The
problem with projecting impacts due to the synergistic effects of emis-
sion gases based on values from the literature, however, is that the
determination of threshold limits has been made at relatively long
periods of exposure when compared to the duration of fumigation events.
Examination of Table 5.6-1, for example, indicates a threshold level
for hyslop wheat (Tritcurn aestivium) for 4 hr exposures at combined
concentrations of 0.6 ppm and 0.1 ppm for SO 2 and NOx, respectively.
While combined concentrations of S02 and NOx approximating these levels
may occur as maximum ground-level concentrations during fumigation, the
periods of exposure will be much less than 4 hours. Evidence to equate
predicted gaseous concentrations during fumigations to established
acute threshold levels was not found. -
Factors other than temporal instability will also act to minimize
the possibility for either losses in plant productivity or damage to
vegetation in the impact area. The ground-level area fumigated during
an inversion breakup is contained within a relatively narrow band be-
neath the originally stable plume. At a distance of 10 km from the
emission source, for example, the concentration of gaseous emissions is
decreased by a factor of four at a distance of 500 m from the plume
centerline and is less than 1 percent of the maximum concentration at
1000 m from the centerline. Moreover, the location of the plume fumi-
gation is determined by wind direction. The chances of detecting fumi-
gations, are very small unless the same area is fumigated repeatedly,
and only a very restricted area is exposed to these higher SO 2 concen-
trations during each event. Finally, the actual region within the impact
area which will be potentially exposed to fumigation consists of a 10-km
wide band between 10 and 20 km from the power plant and between the 1100
and 250° azimuths.
5.6.1.2 Particulate Emissions
Based on the results of the air quality simulation model, it was
concluded that no impacts on the terrestrial ecosystem would be ob-
served as the result of direct contact with particulate emissions in
175

-------
the atmosphere. Projected increased ambient particulate concentrations
are less than 0.4 .ig/m 3 (1200 MW plant with 99 percent control) within
the impact area and, as previously indicated, meet both U.S. and
Montana State air quality standards.
Further analyses, however, were conducted in order to assess the
possible long-term impacts associated with the deposition of trace
elements and their subsequent accumulation within the food chain.
Toxic levels of trace elements may accumulate in the tissue of plants
indirectly by the uptake of soluble elements from the soil. It has
further been shown that bioaccumulation of toxic concentrations of
trace elements can occur at successive levels in the food chain follow-
ing initial uptake and translocation by vegetation.
The methodology used in the impact assessment of trace element
deposition was similar to those utilized by Dvorak, etal., 1977.
Accordingly, the following conservative assumptions were made in order
to establish worse case projections:
1) The deposition of all emitted particulates occurs
within a 80.5 km (50 miles) radius of the
generation site.
2) The particulates are evenly distributed within
the impact area.
3) Emission rates of trace elements are the product
of the concentration of the dust collector ash
and the estimated particulate emission rate
(4899 kg/day).
4) All deposited trace elements reach the soil and
are retained in the top 3 cm.
5) Trace elements move into the root zone and remain
totally available for uptake by vegetation.
These assumptions must be considered extremely conservative due
to the physical characteristics of emitted particulates as well as the
behavior of trace elements in the soil. Particulate emissions escap-
ing 99.5 percent electrostatic precipators are very fine and exhibit
a gas-like behavior. Vaughan, et al., (1975), for example, predicted
that only 6 percent of the totaTeñiTssions would be deposited within
a 50-km radius of their model plant. Additionally, while it is assumed
that trace elements remain totally available to plant uptake and are
readily transported to edible portions of the plants, available evi-
dence indicates that the absorption and translocation of these elements
are, in fact, limited by a number of factors.
176

-------
The projected percent increases in soil and aerial plant parts
occurring over a 30-year period as well as the important intermediate
calculations made in this analysis are shown in Table 5.6-2. Based
on these results it is anticipated that the deposition of trace ele-
ments will not have an adverse impact on the terrestrial ecosystem.
The projected increased concentrations in the soil during the 30-year
period are less than 0.20 percent for each of the selected elements.
Moreover, the predicted increased concentration of trace elements in
aerial portions of plants are below reported toxic levels summarized
by Dvorak, etal., (1977).
The projected increases in several trace element concentrations
may also be compared with baseline concentrations in selected agricul-
tural plant species reported by Braun, 1978. The ranges of concentra-
tions measured in samples from 14 sites near the International Boundary
are presented in Table 5.6-3. The tabulated concentrations of lead,
cadmium, arsenic and selenium are for samples collected during the
late summer of 1977. Samples were also collected in early summer, at
which time most of the constituent concentrations were slightly less
than or about equal to the late summer values. The notable exception
is selenium, which displayed early-summer concentrations of 80-84 1g/g
in grasses and forbs at two of the sites.
A comparison of background trace element levels with predicted
concentrations following 30 years of power plant operation indicates
that there would be negligible increases in vegetation levels of the
four measured elements. For example, the predicted increase in lead
is only 10 percent of the lowest background lead concentration in
grasses and forbs (0.20 pg/g), and represents only 2.4 percent of the
corresponding maximum value. The percent increases above background
are even less for lead levels in stems and heads of spring wheat.
5.6.2 Impacts of Atmospheric Emissions on Aquatic Biota
Atmospheric emissions from the Poplar River Power Plant may poten-
tially effect the following surface water quality parameters: acidity,
nutrient availability and trace element concentrations. The potential
for each of these effects is discussed in subsequent sections.
5.6.2.1 Acidification and Nitrogen Loading
Emission of SO 2 and N0 may contribute towards increasing the
acidity of local precipitation by the formation of strong acids. Sub-
sequent reductions in the pH of surface waters may result if the natu-
ral soil or water buffering capacity is low. Regional acid fication,
presumably enhanced by combustion product emissions, has been observed
in several areas of the world, including the Adirondack Mountain Lakes
177

-------
Table 5.6-2
PROJECTED DEPOSITION RATES, SOIL CONCENTRATIONS AND PLANT ACCUMULATION
OF 17 TRACE ELEMENTS RESULTING FROM PARTICULATE EMISSIONS
Increased Soil Increase Over
Ti ace Element Concentration’ Avei aye Soil Tolal [ iidogeuious Plant Soil Increased Coi,c€,,t, at ic.i,s in
Eii,i 551 on Rate Deposit ion (,iy/y) Cor ,cc ,,tr,, t ion Concenti at ions Co,,ceii tra tiori Pioi i a 1 Plant Pa ts
9/day g/iii /3Oy ar 3 cm 0eil. ___ J y/y)___ P ci cent Ratio’ Jp jL weiyh1)_
Lead 784 4.0 x l0 .9 x lO 10 09 2 0 02
tie, culy 0 441 2 37 a 10 5.4 x io 6 26
Pintiltiony 48 2 58 x 10 6 0 x 1O - - - -
Cad i iium 1.18 6 35 a 1O 1 4 a 10 0 06 0 02 222 0 003
Silver 0 83 4 47 a 1O 1.0 x 10 - - -
Seleiiium 1421 765 x i0 6 17 x 10 05 0.03 4 00007
Pirsenic 3625 195 a 10_s 442 a 10” 60 0007 42 0002
ue,riian ium 8 8 4.74 a io 6 1 1 a I0 - - -
Zinc 122 5 6 59 a 10_s 1 5 a 10 50 0 003 40 0.06
Copper 181 3 9.76 a 10 22 x 10 20 0.01 1000 2.20
l ucI d 53 89 2 90 a 10 6 6 x l0” 40 .002 331 0 22
Cobalt 29 39 1 58 a 10 3.6 a 10 ” 8 .005 87 0 03
ManganeSe 2939 4 1 58 a 10 3.6 a io2 850 .004 3000 108 0
Ch,oiiiium 333 13 1.79 a 10 ’ 4 1 a 10 100 .004 250 1.03
Vai,ad,um 156 77 8 44 a 10 1 9 x 10 100 002 1 0.002
Boron 1469 7 7 91 a 10 I 8 a 10 10 0 0 18
Be iyliiurn 2 25 1.21 x 2 7 a l0 6.0 0005 16 0004
‘As uiiies bulk density of soil is 1 47 9/cm 3 .
?TIIOSC concentration ratios were derived by Vaughn, eta). (1975) and express the potential uptake capacity of various plant species for
indicated trace elenients
‘Ac unii: the concenttat lon of trace elements deposited in the top 3 cm of soil moves into the root zone and is totally available for
uptake by the vegetation

-------
Table 5.6-3
MINIMUM AND MAXIMUM CONCENTRATION OF TRACE ELEMENTS IN
POPLAR RIVER BASIN VEGETATION (pGIG) SAMPLES
COLLECTED DURING THE LATE SUMMER OF 1977
Pb Cd As Se
Grasses and Forbs 0.20- 8.3 0.05-0.32 <0.05-0.17 <0.05-0.52
Spring Wheat
- Stems 0.6 - 2.0 0.01-0.35 <0.05-0.20 0.08-0.38
- Heads 14.1 -67.6 0.01-0.16 <0.05-0.06 0.08-0.85
Alfalfa 0.2 - 1.0 0.05-0. 14 <0.05-0.08 0.07-0.35
179

-------
of Northern New York State. A regional analysis indicates, however,
that there is an extremely low potential for surface water acidifica-
tion in the Uorthern Great Plains (Dvorck, .i. , 1977). This re --
suits primarily from the alkaline nature of soils and surface waters,
and the clay component of soils in the Poplar River Basin.
Similarly, atmospheric NOx emissions are expected to result in no
appreciable increase in N loading of surface waters which would result
in increased eutrophication. The Upper East Fork currently displays
indications of a eutrophic status based on inorganic nitrogen and phos-
phorus concentrations and the abundant macrophyte growths. Based on
observations by Klarich (1978) the Poplar River system is N-limited;
therefore, increased inorganic nitrogen concentrations could poten-
tially result in a concomitant increase in algal or niacrophyte growth.
However, during the period of maximum potential plant growth (i.e.,
summer) atmosphere NO is generally not available for introduction
into surface waters due to the lack of precipitation. Even during
periods of heavy precipitation or runoff the transport of atmospheric
inorganic nitrogen to surface waters would be minimal due to the low
concentrations and utilization by terrestrial flora.
5.6.2.2 Trace Element Contamination
A number of trace elements occur in flyash following coal combus-
tion and are generally emitted in conjunction with the particulate
matter. Mercury is emitted as a vapor, however.
Trace elements emitted as stack discharges may eventually be in-
troduced into streams, resulting in increased concentrations above
background levels. The mechanisms and magnitude of transport into
waterbodies are quite complex, however, and depend on the following
factors:
1) terrestrial deposition rate
2) water flow characteristics (e.g., surface runoff
or groundwater)
3) soil chemistry (especially CEC and pH)
4) aquatic chemistry (especially suspended solids,
alkalinity, pH).
In order to assess the potential for increased concentrations of
trace elements in the Poplar River due to atmospheric emissions, the
potential riverine concentrations were calculated using a simplified
direct-transport approach. In estimating potential aquatic impacts,
the following assumptions were used:
180

-------
1) Sixty percent of the total particulate emissions
would be deposited with the Poplar River
drainage basin.
2) The total daily deposition of particulate matter
would be dissolved into the average expected
daily flow of the Poplar River.
Due to the very conservative nature of the aforementioned assump-
tions, the resultant estimates are actually the maximum potential con-
centrations. The particulate emission rate was assumed to be 4898.9
kg/day, which corresponds to two 300 MW units with 99.5 percent elec-
trostatic precipitators. Trace element concentrations in particulate
emissions were assumed to equal the concentrations measured in dust
collector ash of Poplar River coal samples (see Table 5.6—4).
The results of these analyses indicate that the addition of trace
elements derived from Poplar River Power Plant stack emissions would
not result in appreciable elevations of background concentrations
(Table 5.6-5). In most cases, calculated concentrations are consider-
ably less than 1 ug/2. and represent only a small percentage of natural
constituent concentrations. Moreover, when compared with Water
Quality Criteria for the protection of aquatic life (EPA, 1976), the
calculated increases do not result in any concentration (background
and addition) exceeding the corresponding criteria value (Table 5.6-5).
Trace elements can also accumulate in the snow and enter rivers as a
slug during the spring but the concentrations would still not be sig-
ni ficant.
It should again be emphasized that the preceding calculations
represent maximum potential values. The actual increases and potential
effects on aquatic biota would probably be considerably less due to:
1) Much lower terrestrial deposition rates. Vaughn,
eta]., 1975, indicates that only about 6 percent
of particulates would generally be deposited
within a 50—mile radius.
2) Retention of trace elements in soil will prevent
groundwater or surface water contamination. The
alkaline nature of Poplar River Basin soils and
the moderate cation exchange capacity would tend
to prevent groundwater transport.
3) Reduction of concentrations and toxic effects in
surface waters. The solubity and resultant
toxicity of heavy metals is considerably reduced
in waters of high alkalinity and pH.
5.6.3 Impacts of Water Quality Changes on Fish and Wildlife
Operation of the Poplar River Power Plant in conjunction with flow
apportionment will affect downstream water quality in the U.S. part of
the basin. Changes in a number of constituents have been predicted by
181

-------
Table 5.6-4
TRACE ELEMENT CONCENTRATIONS (PPM) IN POPLAR RIVER COAL
ASH SAMPLES
Upper Dust
Ash Collector Ash
Lead 110 160
Mercury 0.07 0.09
Antimony 5.6 9.8
Cadmium 0.24 0.24
Silver 0.17 0.17
Selenium 0.41 2.9
Arsenic 0.74 7.4
Germanium 0.91 1.8
Zinc 25 25
Copper 21 37
Nickel 7.5 11
Cobalt 5.1 6
Manganese 300 600
Chromium 170 68
Vanadium 48 32
Boron 100 300
Beryllium <0.10 0.46
Source: Accu—labs Research, Inc. (1978).
182

-------
Table 5.6-5
AVERAGE TRACE ELEMENT CONCENTRATIONS ( G/9.) IN THE
POPLAR RIVER AND PROJECTED INCREASES DUE
TO ATNIOSPHERIC EMISSIONS OF THE
POPLAR RIVER POWER PLANT
Input from Quality Criteria
Median Maximum Atmospheric for Aquatic Life Criteria
C.,ncen:ration’ Er issions 2 ( EPA. t975) Esti rate 3
Lead 2 5 1.6 0.01 96-h LC5O 2000
ercury 0 2.8 0.0009 0.05
Antimony - — 0.10
Cad iium <1 1 0.0025 12.0
Silver - - 0.0018 0.01 96-h LC5O
Selenium <0.2 1.6 0.030 0.01 96-h LCSO 20
Arsenic 2 27 0.076 50’
Germamium - - 0.019
Zinc 10 30 0.259 0.01 96-h LC5O 50
Copper 1 9 0.383 0.01 96-h LC5O ‘ 50
Nickel 3 6 0.114 0.01 96—h LC5O “ 200
Cobalt — — 0.062
Manganese 60 200 2.69 1,5006
Chrcnium 0 10 0.704 100
Vanadium 1 5 0.331
Beryllium 0 10 0.048 1,100
Boron 1,000 4,100 3.11 18 x
‘Average for Poplar River from Klarich, 1978.
2 Based on daily dilution in average Poplar River Flow.
‘Based on reported toxicities in water quality similar to that of Poplar River.
‘io criterion for protection of aquatic life
o criterion - domestic water supply criterion should protect aquatic life.
6 11o criterion - threshold for sensitive species.
‘ o criterion - lethal dose for minnows
183

-------
the water quality modeling studies outlined in Section 5.3. Several of
these constituents have a high potential for direct impacts on aquatic
biota. The effects of changes in thermal regimes, dissolved oxygen and
dissolved solids are discussed in the following sections. Other con-
stituents which are considered to be of very low potential for impact
on aquatic life are also discussed in less detail.
5.6.3.1 Effects of Thermal Discharges
The once-through cooling water discharge for the Poplar River
plant is located near the reservoir release point at the dam. There-
fore, there is a potential for heated waters to be released into the
East Fork during reservoir spill or demand releases. The following
section provides a discussion of the downstream temperature elevations
and the potential for effects on aquatic life.
The magnitude of temperature elevations (i T’s) downstream from
Cookson Reservoir was examined by Spraggs (1977). The methodology used
was a two-step modeling approach in which the thermal regime of Cookson
Reservoir was simulated with both one- and two-unit operation (300 and
600 MW). The output of the layered one-dimensional reservoir model was
then used as input into a fully mixed one-dimensional stream model to
predict downstream Al’s from the dam to the International Boundary.
The reservoir model indicated that for one-unit operation (at a
plant AT = 18°F) the reservoir releases would be at a maximum AT of 9°F.
For two units, the releases would be at the full power plant AT of 18°F.
When these release AT’s are used in the stream model, a complete return
to natural ambient temperature was predicted to occur within the Cana-
dian part of the basin during the low-flow months of August through
September. During June and July, an 18°F reservoir release (i.e., two
300 MW units) would result in temperature elevations at the Interna-
tional Boundary of only 2.3°F (Figure 5.6.2). The greatest downstream
effect on ambient temperature was predicted in May when a 9.2°F AT
would occur at the boundary during an 18°F AT reservoir release.
The modeling results are substantiated by using the surface heat
exchange equation of Edinger, etal. (1975):
TS = TN + TD_K(t ’
where
IS = predicted temperature
TN = natural temperature
TD = temperature of reservoir release
184

-------
Source: Spraggs, 1977
Temperature Increase (°C) at International Boundary
Figure 5.6-2
RIVER HEAT RELEASE SUMMARY
185
L)
0
a)
U,
a)
I-
a)
a)
S.-
S.-
a)
E
a)
I—
10
8
6
4
2
0
1 2 3 4 5 6

-------
K = surface heat exchange coefficient
A = area of river reach
Q = flow
This equation indicates that almost all of the heat will be dis-
sipated prior to the International Boundary at flows of less than 100
cfs. The maximum predicted AT at the boundary is about 7.2°F, which
corresponds to a flow of 250 cfs. Higher flows would result in greater
heat dissipation due to increased water surface area resulting from
overflow beyond the main channel.
It should be emphasized that the May flow condition (130 cfs) re-
sulting in the predicted 9°F increase at the boundary would generally
be a rare occurrence under the proposed apportionment alternatives (see
flows for Scenario 29 in Section 5.2). Under most conditions, the East
Fork flow would be less than 100 cfs; therefore, the temperature dissi-
pation would be considerably greater than predicted by Spraggs (1977).
The modeling results indicate that the only potential for signifi-
icant downstream temperature elevations in the U.S. part of the basin
would occur during high runoff conditions in the spring. The potential
effects would be considerably moderated, however, since the reservoir
would not usually have a natural spill during April—May, demand releases
would generally be less than 100 cfs and cooler ground water would also
enter the river downstream of the reservoir.
Although it is considered unlikely that significant temperature
elevations would occur in the U.S. part of the basin, it is an impor-
tant consideration due to the occurrence of gamefish spawning activi-
ties during high flow conditions. Eggs and larvae may be especially
susceptible to temperature changes. Therefore, recruitment to walleye
and northern pike fisheries is dependent on a given thermal regime
during incubation and juvenile development.
Egg and larvae data collected by Montana Department of Fish and
Game (1978) indicate that, during the spring of 1977, walleye and
northern pike were probably spawning during the period of mid-April to
early May. During the spawning period median daily river temperatures
ranged from about 48° to 59°F. Daily temperature variations were quite
high and ranged from 8 to 12°F. Larval walleye occurred in the river
until mid-May, at which time natural river temperature ranged from 55°
to 70°F.
Studies by Koenst and Smith (1976) on the temperature requirements
of young walleye indicate that optimal temperatures for egg incubation
range from 48.2° to 59.0°F. This temperature range closely corresponds
to the observed occurrence of ripe adults and larval walleye during
1977. The young walleye were also quite resistant to temperature
186

-------
changes since they tolerated up to a 18°F AT over a period of 72 hours
without adverse effects on survival. The high resistance to rapid
temperature fluctuation by young walleye was also demonstrated by
Ailbaugh and Manz (1964). In the laboratory studies there was also
good survival to hatch of walleye eggs incubated at temperatures up to
64°F, indicating that relatively small increases in natural river
temperature during the spawning season would not adversely affect sur-
vival.
The optimal temperature for growth of juvenile walleye was about
72°F (Koenst and Smith, 1976). The upper lethal temperature ranged
from 81 to 89°F as acclimation temperature increased from 46° to 79°F.
These data indicate that small elevations in ambient river temperature
(<6°F) during the juvenile rearing period would also not be detrimental
to walleye production. Conversely, it could be potentially beneficial
since ambient river temperature during much of May are actually sub-
optimal for walleye growth. This is an important consideration since
Johnson (1961) found that cold weather during the incubation period may
actually be an important limiting factor in determining walleye year-
class strength.
Based on available data for young northern pike (Hokanson, etal.,
1973), it appears that their temperature tolerances would also not be
exceeded by small temperature elevations (<6°F) in downstream areas.
Pike larvae are apparently more sensitive than walleye to rapid
temperature changes. However, in field studies Franklin and Smith
(1963) found good survival when temperature changes did not exceed
1.2°F/hour.
Furthermore, the upper lethal thresholds (96—hour) for both
northern pike and walleye are about 90°F. Since the maximum natural
river temperatures are generally about 80°F, limited heating (<5°F) of
the river during summer would not exceed the upper limits for gamefish
survival.
In summary, mathematical predictions indicate that the downstream
temperature elevations due to power plant operation on the Poplar River
are a function of:
• number of units
• flow in East Fork
For the two-unit operation (600 MW) the maximum predicted temperature
elevation at the boundary during expected spring and summer flows would
be less than 2°F. Moreover, during low flows all excess heat would be
dissipated prior to the International Boundary. Even under situations
of high flow (>130 cfs) the maximum excess AT at the border would be
187

-------
only about 9°F, and this would undergo rapid decay in the East Fork.
Therefore, even under the most extreme situations (i.e., two-units and
high flow) the natural stream temperatures would probably be reached
within the East Fork. The available information on the thermal re-
quirements of the two major game fish present in the Poplar River indi-
cates that there would be no adverse impacts due to thermal releases at
the reservoir.
5.6.3.2 Effects of Dissolved Solids Increases
Increased total dissolved solids (TDS) levels in freshwaters may
be potentially toxic to aquatic life. Tolerance of fishes to increased
TDS is highly variable according to species; however, it is generally
assumed that TDS concentrations above 5,000 to 10,000 mgI2. would gen-
erally represent adverse conditions for survival of most freshwater
organisms.
It should be emphasized that most freshwater species, especially
those inhabiting the Northern Great Plains waters, can tolerate a
relatively wide range of TDS concentrations. This is necessary due to
the high seasonal variability encountered in many surface water bodies.
For example, in the Poplar River, natural TDS concentrations may range
from <150 mg/9 .. during high runoff, to over 1000 mg/Z during low flow
conditions.
In general, the early life stages of fish (i.e., eggs and larvae)
are more sensitive to increases in TDS than adults of the same species.
Limited data are available on the salinity tolerances of eggs of the
two major species of Poplar River garnefish: walleye and northern pike.
PeterLa (1972) reports that eggs of both species displayed good hatch-
ing success in waters with a conductivity of 1300 pmhos. Based on the
conductivity - TDS relationship of Poplar River waters at 25°C, this
would represent a TDS concentration of about 1100 mg/9. . As the conduc-
tivity was increased to 4000 imhos (TDS 3500 mg/2 ). There was no
hatch of walleye eggs and very poor hatch of northern pike. The
studies of Petarka (1972) also indicated that fathead minnows (a for-
age fish in the Poplar River) experience no adverse effects on growth
and reproduction at TDS concentrations up to 7000 ng/2 .
The water quality modeling results indicate that the highest
potential for increases in TDS concentrations resulting from flow
apportionment will occur in the Upper East Fork (see Section 5.3.3).
With more than two units (1985 and 2000 level of development) the
April TDS concentrations in the entire East Fork will be at levels
where spawning of walleye and northern pike may be impaired in one
year out of every ten (i.e., 90 percent concentration is exceeded).
During 90 percent of the spawning seasons the egg tolerance limits
should not be exceeded in the East Fork. In all other parts of the
U.S. basin, TDS concentrations during April should be below the toler-
ance levels for successful hatch of gamefish eggs. This results from
dilution by lower salinity waters from the West and Middle Forks.
188

-------
The only condition which results in TDS concentrations exceeding
the limits for maintenance of fish population during the non-spawning
season is with four units at the year 2000 level of development. In
such cases relatively high TDS concentrations (4000-5000 mg/9 ) would
exist in the East Fork during low flow months in the fall and winter.
Although such concentrations would probably be toxic to gamefish, the
more important limiting factor under Scenario 32 would be the absence
of flow in the lower East Fork (see Sections 5.6.4).
5.6.3.3 Effects of Dissolved Oxygen Changes
Adequate dissolved oxygen is essential for the maintenance of fish
populations. This is especially true for gamefish species such as
those existing in the Poplar River, which are generally less tolerant
of low oxygen levels than “rough” or forage species. There are two
main conditions in which dissolved oxygen may be limiting factors for
the production of stream fish populations:
1) During the spawning period eggs and larvae may be
susceptible to low oxygen levels, especially in
species which deposit eggs in the bottom substrate.
2) In areas where surface waters freeze during the
winter, low dissolved oxygen conditions may exist,
especially in shallow waters at low flow rates.
Based on recent Quality Criteria for Water (EPA, 1976), a minimum
dissolved oxygen concentration of 5 mgR is recommended for the mainte-
nance of “good” fish populations. Based on available water quality
data, it appears that dissolved oxygen levels in the Poplar River are
well above minimum values for gamefish protection during the period of
April through November. Even in the upper reaches of the East and West
Forks, the dissolved oxygen always exceeded 86 percent saturation.
Moreover, these high levels were measured after closure of the dam on
the East Fork; therefore, during ice-free conditions no adverse impacts
of apportionment schedules on dissolved oxygen concentrations are ex-
pected to occur.
Dissolved oxygen levels also appear to be sufficient for success-
ful spawning and early development of walleye and northern pike. Oseid
and Smith (1971) indicate that the lowest dissolved oxygen concentration
for optimal hatching of walleye eggs is 5-6 mg/L Moreover, a relatively
high survival to hatch was observed at dissolved oxygen concentrations
of only 2 mg/2 .. Hatching time was extended, however, and the larvae
were smaller at hatching when incubated at 2 mg/Z.
Siefert, etal. (1973) found similar results for northern pike eggs
and larvae. Dissolved oxygen concentrations of about 5 mg/2. were ade-
quate for hatching and survival to the feeding stage. Pike eggs may be
189

-------
more sensitive to continuous lower oxygen concentrations since survival
was considerably reduced at a level of 3.4 mg/9 . However, pike larvae
are apparently able to tolerate short exposures to low dissolved oxygen
concentrations. Peterka and Kent (1976) found no reductions in survival
of embryos, yolk-sac larvae and larvae maintained for eight hours at
dissolved oxygen concentrations of 0.6, 2.0 and 4.0 mg/9.., respectively.
Since April dissolved oxygen concentrations in the East Fork after
dam closure (since October, 1975) are generally above 5 mg/2 ., no impacts
on fish spawning due to changes in dissolved oxygen levels are antici-
pated. This conclusion is based on the observed high spawning-period
oxygen concentrations under spring flow conditions on the East Fork
which are more severe (in regard to flow and oxygen availability) than
would be expected under the apportionment agreements. This results
primarily from the additional demand releases which would add a minimum
of 1 cfs to the natural flow resulting from groundwater accretion and
surface runoff.
During the period of normal ice cover on the Poplar River, low
dissolved oxygen may be an important limiting factor in fish survival.
In February-March, 1977, Stewart (1978) observed large numbers of dead
walleye, suckers and carp on the East Fork. In late February, dissolved
oxygen concentrations as low as 0.1 mg/9 .. were recorded under the ice.
Moreover, in some locations there was no water below the ice or the ice
depth exceeded 4.5 feet. This lack of available under-ice habitat and
low oxygen levels was apparently the cause of the fish kills. Similarly
low oxygen levels were observed on the West Fork and, in conjunction
with the low summer flows, may explain the lack of gamefish in that
part of the basin.
Based on the aforementioned considerations and the historical
occurrence of very low flows during January and February, it appears
that during years of average or low flows the apportionment of flows
and resultant continuous demand releases (1 or 2 cfs) will have a po-
tential beneficial effect on fish winter habitat. As indicated in
Section 5.2, the flows on the Upper East Fork at both 10 percent and
50 percent frequency will be greater than the corresponding historical
flows.
Oxygen concentrations in the East Fork during the previous winter
(1976-77) (Stewart, 1978) appear to be satisfactory for fish survival
(e.g., >4 mgI2 .). The average winter flows in 1976-77 (1.7-2.3 cfs)
were similar to the median winter flows under development Scenarios 28
through 32.
There are also no anticipated adverse impacts on winter fish
habitat in the Middle Fork due to the apportionment. During January
and February there is historically little flow at the International
Boundary of the Middle Fork or its major tributary, Coal Creek. There-
fore, during the winter most of the flow in the Lower Middle Fork has
occurred in the U.S. part of the basin below the confluence of Coal
Creek. This is true under historical conditions and under Scenarios
190

-------
28 through 30 (i.e., up to three units at 1985 water use). Scenarios
31 and 32 (3 and 4 units at 2000 water use) include a reservoir on the
Middle Fork and median (50 percent) or low (10 percent) winter flows
would actually be larger than historical flows in the Lower Middle Fork.
In summary, the apportionment schedule which results in the main-
tenance of stable winter flows in the East Fork will not adversely
impact fish survival in winter and should have a beneficial effect
in preventing the occurrence of fish kills during low—flow years.
5.6.3.4 Other Constituents
Maximum concentrations of boron up to 20 mg/2 could occur in the
Poplar River under the apportionment agreement. No impacts on local
fish and wildlife are predicted, however, due to the relatively low
toxicity of this constituent to animal life. Toxic concentrations
to fish are over 1000 times the osmosis predicted concentrations.
Turbidity and suspended solids measurements on the Poplar River indi-
cate that this constituent is not limiting to fish production. Even
during extreme high flow conditions the suspended solids are within
short-term tolerance ranges. Any effect of power plant development
on the East Fork would probably be evidenced as reduced suspended
solids loading due to settling in Cookson Reservoir. Therefore, no
adverse impacts of development would be predicted.
5.6.4 Impacts of Flow Modifications on Fish and Wildlife
5.6.4.1 Wildlife and Furbearers
Several species of mammals and birds are directly dependent on the
aquatic habitat provided by the Poplar River. The two main groups
which have a potential for impact from flow modifications are furbearers
and waterfowl. Both of these groups require specific aquatic habitat
characteristics which may be modified by variations in the amount of
water available in the river basin.
During the period from 1977 to 1978 the number of observations of
beaver and raccoon in the Poplar River drainage increased slightly. At
the same time, the observations of muskrat decreased, while the numbers
of mink remained unchanged (DeSimone, 1979). It is generally believed
that beaver, muskrat and min’k populations have declined during the
last 30 years primarily as a result of intensified agriculture and land
use practices in the area. It is not anticipated that the proposed
action will have an adverse impact on furbearers, since they are not
directly dependent upon the magnitude of peak flows.
191

-------
Inability to maintain existing channel morphology in the upper
nine miles of the East Fork due to reduced flows may result in encroach-
ment of emergent macrophytes and a concomitant loss in waterfowl pro-
duction. Recent studies (DeSimone, 1978) have shown that both waterfowl
breeding pair density and production was greater in this portion of
the East Fork than on similar areas studied on the Middle Fork, Main
River, West Fork, Coal Creek and Butte Creek. While it was concluded
that existing submergent and emergent macrophyte growth provided brood
cover and was responsible, in part, for the greater brood production
in the upper portion of the East Fork (DeSimone, 1979), substantial in-
creases in emergent growth would result in the loss of brood habitat
by restricting access to the preferred rearing grounds for surface-
feeding species of ducks. It has been shown that a greater than 50
percent areal coverage of emergents on a water body renders these areas
unsuitable for duck production (Evans and Black, 1956; Stoudt, 1971;
and Whitman, 1976).
In an assessment of the effects of altered flow regimes of the
Poplar River, Bahis (1979) stated that flow reductions with two power
plants and the recommended apportionment will result in an expansion of
the macrophyte community within the system. It was concluded that mac-
rophyte growth within and along the main channel will be most dramatic
in the East Fork and may result in 50 percent coverage during the life
of the project. The results of an aerial survey of emergent vegetation
in several segments of the East Fork Poplar River during July 1979
(DeSimone, 1980) indicated that at least in the area adjacent to the
International Boundary one section is approaching 50 percent coverage,
while emergent coverage in three other sections ranged from 5.8 to 21.6
percent. However, additional releases specified under the recommended
apportionment were not made. Based on the relationship between areal
coverage of emergents and waterfowl production presented above additional
encroachment in the uppermost reaches of the East Fork will result in a
reduction in waterfowl production. Progressive encroachment in the
upper nine miles of the East Fork Poplar River resulting in greater than
50 percent areal coverage would result in the loss of approximately
70-80 breeding pairs of ducks and the production of between 300 and 400
ducks annually (DeSimone, 1979). However, simultaneous increases in
macrophyte growth in the lower East Fork and Main Poplar River, where
abundances of emergents are substantially less than the upper East Fork,
could lead to an increase in waterfowl production in these areas and the
supplanting of the breeding areas to the north.
5.6.4.2 Fish
The fish populations of the Poplar River are susceptible to habitat
alterations resulting from changes in river flow rate. The two major
areas of potential impact are:
• Changes in spawning habitat
• Reduction in overwintering habitat
192

-------
The impacts of flow modifications resulting from apportionment on the
winter survival of fish during ice cover conditions are related pri-
marily to the low dissolved oxygen concentrations occurring under low
flow conditions in January and February. These impacts are discussed
in the water quality section (5.6.3.3).
The main emphasis of this discussion will involve consideration of
the flow impacts on spawning habitat for the two major game fish
species in the Poplar River: walleye and northern pike. Although the
magnitude of flows during non-spawning seasons may also be potentially
limiting to game fish production, no adverse impacts of apportionment
are expected due to the continued observed occurrence of gamefish
(Stewart, 1978) during low flow years such as 1976-77. In such cases
measured flows throughout the basin were less than those predicted
under the most likely apportionment alternative (VI). Therefore,
subsequent impact discussions will be concerned with three main aspects
of fish spawning habitat:
1) Maintenance of normal channel morphology
2) Inundation of shoreline vegetation
3) Prevention of siltation
Both northern pike and walleye normally spawn during April in the
Poplar River. Although their spawning seasons coincide, they require
quite different spawning habitats. Pike deposit adhesive eggs among
vegetation in shallow water (<15 inches deep). The eggs hatch in about
2 weeks. Walleye spawn over gravel bottoms in pool tails or riffles
(depth usually less than 1.5 feet). Their adhesive eggs settle into
the gravel and also hatch in about two weeks.
Walleye are an important game fish in the Poplar River and their
continued existence in a prairie stream such as the Poplar River is
dependent upon the maintenance of the existing channel morphology
consisting of alternating riffles and pools with clean gravel substrate
during the spring. Alteration of the morphology of the Poplar River
to that of a more typical prairie stream (e.g., uniform gradient, soft
bottom) would result in a loss of walleye production. Therefore, it
is important to consider the flow regimes required to maintain the
current degraded channel.
The hydraulics of the Poplar River are complex. The bed character-
istics and meandering channel form make velocities at any transect
extremely non-uniform. The abruptness with which flood peaks occur
make velocity computation by steady-state hydraulics equations inexact
at high flows. However, these are the important flows from the stand-
point of maintaining fish habitats in certain portions of the river.
193

-------
Therefore the actual numerical estimates made in this section to pro-
vide a flow and sediment characterization of the East Fork may be
considerably in error. However, the numbers do reveal some general
relationships such that conclusions can be drawn concerning impacts
of flow on fish.
In order to determine if the Poplar River can maintain the degraded
channel appropriate for spawning, two questions must be answered.
They are:
• What flows produce the tractive forces required to
suspend sediments, and
• What duration of the critical flow is necessary to
scour the sediments.
Flow uniformity in the channel must first be assumed. This will
necessarily underestimate the scouring potential over the riffle areas
and overestimate in pool areas. If pools occur in bends of the river,
radial acceleration may keep the scouring rates competitive with those
in riffle areas.
Brown (1978) conducted a study to determine minimum flow rates for
maintenance of fish habitats. In the investigations he collected the
particle size data shown in Table 5.6-6. The data are broken down
into riffle and pool areas. In general, velocities are lower in pools
and finer sediments tend to collect in those areas. To maintain the
spawning habitat of game fish on the East Fork, it is necessary to
scour the size fractions from approximately the medium sands to the
colloidal clays. The velocities required to pick up and transport
these particles range from about .52 to 1.04 feet/second. (Longwell,
et al., 1969).
The approach to finding the critical flow required on the East
Fork included the use of the Manning formula. This equation,
v = n R” S’
says that the average stream velocity (V) is proportional to the stream
hydraulic radius (R) and the energy slope (S) and inversely proportional
to channel roughness (n). Since the critical velocity (1.04 fps) is
known, the equation can be rearranged to calculate R. The roughness
coefficient (n) was taken from Schwab, etal., (1966) and has a range
of 0.45 to 0.6 from streams like the Poplar. S was found by dividing
the difference in elevation of the USGS stream gauge at the International
Border and the gauge near Scobey on the East Fork by the total estimated
distance in river miles between the two locations. The value of S used
was 0.00011 ft/ft. Using these values yielded a hydraulic radius R
and, assuming a rectangular channel, a new flow depth was calculated as
194

-------
Table 5.6-6
BED MATERIAL SIZE DISTRIBUTION
Sample Pool(P) D 036 D c 084 09
Loca tion Ri ffle(R) (ni ) (mni (nin (mmj (n im (nm) (n m
East Fork
2.5 miles below P 0.058 0.095 0.11 0.15 0.18 0.25 0.4
Inter. Border R 0.6 4.0 12.5 21 31 39 45
Cromwell Slab P 0.34 1.6 7.6 12.5 21 46 57
R 2.7 10.25 20 36.5 52 73 84
USGS gage P 0.27 1.0 4.8 9.1 16.5 31 49
near Scobey R 0.5 1.8 8.5 14 22 39 49
Middle Fork
McCarty Crossing p i.i 8 19 33 45 63 80
(J1 near mt. Border R 2.1 11 21 33.5 39 49 58
Hagfeldt Slab P 0.21 0.37 6.6 21 39 56 62
R 0.92 8.3 18 26 45 54 58
Ofstedal Slab P 0.18 0.255 0.30 0.345 0.44 0.79 1.20
R 4.7 10.5 19.5 28 37 68 75
West Fork
South of Peerless P 0.35 4.3 10.0 15.5 20.5 39.5 54
P 4.2 10.5 25.5 48 68 79 86
Susag Farm P 1.4 10.0 30 40 51 66 68
R 0.32 1.8 7.6 14 32 56 68
Main Stem P 0.062 0.099 0.135 0.165 0.21 0.33 0.45
R 0.31 1.4 8.0 15 22 41 48
Crowley Slab P 0.54 0.086 0.128 0.196 0.31 0.38 0.45
R 0.30 1.6 10.0 17 24 47 52
Source: Brown, 1978

-------
depth = w R
where w is the width of the stream (from the Brown (1978) data). Know-
ing the new depth and channel width, the cross-sectional area can
be calculated. The critical flow required to flush sediments from
the fish habitats was computed by continuity (Q AV) to be about 715
cfs. The flow required to overtop the main channel and begin inunda-
tion of the shoreline vegetation areas is somewhere between 220 and
330 cfs according to similar calculations. This range of values is
supported by the report of Saskmont Engineering (1976) who indicated
that a 250 cfs flow would escape the main channel. Therefore at flows
which would provide the required scouring velocities the pike spawning
habitats would be inundated.
Calculating the duration of the required 700 cfs scouring flow is
a more difficult problem. The first assumption that was made in this
solution was that the reservoir is a 100 percent effective sediment
trap for waters spilling from the upper East Poplar across the
International Boundary. Secondly, it is assumed that sediment loadings
to the river (excepting the within channel bank erosion) come only
from cultivated agricultural land.
Using the universal soil loss equation, the annual sediment
loads from cropland to the East Fork can be obtained. The equation is:
A = RKLSCP
where A is the annual sediment loss in tons/acre
R is the rainfall erosivity factor
K is the soil erodibility factor
LS is the slope-length factor
C is the crop factor, and
P is the practice factor.
The parameters were estimated from tabulations and figures in Stewart
(1976). The parameter values are given below.
R = 50
K = .30 (for a sandy clay to silt loam soil)
LS = .20 (slope length 1000 ft, .05 percent slope)
C = .13 (spring wheat)
P = .52 (fall seeded grain, low slope)
Using these parameters the gross erosion is about 0.20 tons/acre.
Zison, etal. (1977) has presented the relationship developed by McElroy
to determine the sediment delivery ratio. Using the drainage density
given for the East Fork and the proper soil type, this ratio is deter-
mined to be 0.12, yielding a net delivery to the stream of 0.024 tons/
acre. The contributing area to the portion of the East Fork from the
196

-------
International Boundary to the USGS gauge near Scobey is approximately
188 mi 2 . Of that, about 74 percent is under cultivation, or about
89,000 acres. This indicates an annual sediment loading in the U.S.
portion of the East Poplar of 2136 tons. This, of course, assumed
that loadings from uncultivated areas are negligible and that no loads
are added in between the reservoir and the U.S. boundary.
To estimate the scour at the critical flow rate, sediment concen-
tration and flow data were taken from USGS records for the East Fork
at the International Boundary. The suspended sediment data were
converted to mass flux data and plotted versus flow rate. This plot
is shown in Figure 5.6-3. A log-log least squares regression gave
the best fit line shown in the figure. The equation for sediment
flux in grams/seconds given flow in cubic feet/second is:
Flux = 1.16 (Flow) 0 97
Substituting 715 cfs into the above, the sediment flux is calculated
to be 680 g/sec. This represents the amount of suspended sediment
that would move past a point in the river in one second at that flow.
If the assumptions are made that the only sediment in suspension comes
from overland flow loading and that sediment must be transported
through the entire length of the reach, then the time required to
flush the East Fork would be 33 days. Realistically, the sediment
probably does not need to be moved entirely through the system in
order to cleanse critical areas. The spawning habitat in riffles is
probably maintained due to the movement of sediments off riffles into
more sluggish areas of the stream. These areas are usually where
point bars are building up on the inside of bends. Therefore, the
33 days is a nominal upper limit value.
According to Brown (1979) the average distance between areas
of severe bank cutting and hence the distance between corresponding
sluggish areas on the inside of the turns is about 2,000 feet. If
sediments are only required to move this distance, the duration of
the 700 cfs flow would necessarily be only about 6 hrs. Very likely,
the time required for scouring is between these limits. For instance,
the clays may move entirely through the system while coarse sands
may be picked up and deposited 2000 feet downstream. Fine sands might
travel several miles and come out of suspension.
If the outlet pipe of the reservoir could produce the required
715 cfs, the duration of such a flow (assuming half the total reservoir
storage could be utilized) would be only 11.5 days. If the 300 ac-ft
under the apportionment schedule were released at this rate the
flow duration would be only five hours. It seems infeasible that the
habitat can be maintained with the reservoir in place. However, his-
torical records indicate that the mean monthly flow in the East Fork
197

-------
1000
Flux 1.16 (Flow) 0 97
r 2 85
log-log
0 Prereservolr (1974 - Nov 1915)
Postreservoir (Dec 1915 - 1976)
0.1
0.1 1 10 100 1000
Flow (cfs)
Figure 5.6-3
SEDIMENT FLUX VERSUS FLOW AT THE INTERNATIONAL BOUNDARY
EAST FORK POPLAR RIVER
*
U
QJ
a.
U-
4-.
a)
E
a)
€1)
0
a)
-a
C
a)
U .
U)
100
10
1
0
0
Do
0
0
0
0
0
0

-------
is never sustained above 715 cfs. It is very likely, then, that
burgeoning agricultural development over the past century has caused
a slow aggradation of sediments in the East Fork. The following dis-
cussion illustrates this.
The sediment loadings to the East Fork appear to be quite small.
The U.S. East Fork from the International Border to the USGS gauge
at Scobey is approximately 100 feet wide during high flow on the aver-
age. Sediment with voids weighs about 165 pounds per cubic foot.
Using this information, the mean annual depth increase of sediment
on the stream bed can be estimated. This depth is 0.00069 ft/year.
At this rate, 120 years would be required to accumulate a one-inch
depth uniformly over the East Fork (assuming bank cutting is a negli-
gible source). If siltation occurs as a result of reduced flows, it
will likely be a very slow process.
Analysis of historical records indicates that on the upper East
Fork the peak discharge with a return period of 2 years is 948 cfs.
Although it would occur normally for only a short period (e.g., <2 days),
peak flows of this magnitude have obviously been sufficient to main-
tain adequate spawning substrate for walleye.
Since the water quantity modeling results are based on total
monthly flows, it is difficult to predict the exact effects of appor-
tionment on the magnitude or duration of peak flows which may occur
for short periods only. However, it is reasonable to assume that
peak flows on the East Fork are considerably reduced when compared
with pre-impoundment conditions. Since the demand releases would be
insufficient to affect significant scouring, it must be concluded that
the long-term effects of power plant operation and flow apportionment
will be a gradual change in channel morphology, especially in the East
Fork. These changes would eventually result in an adverse impact on
the availability of walleye spawning habitat in the form of riffles
with clean gravel bottoms.
Several factors suggest, however, that the adverse impacts on
walleye spawning habitat may not be manifested for some time after
initiation of power plant operation and flow apportionment:
• Low sedimentation rate. Reservoir construction has
actually reduced sediment load in the East Fork due
to sedimentation in the reservoir and reduction in
bank erosion due to lower peak flows.
• Continued spawning on East Fork after closure of
Morrison Dam. Flows on the East Fork during the last
two spawning seasons (1977 and 1978) have been
approximately equal to, or less than, median flows
predicted to occur with one power plant unit in
operation. Although poor spawning occurred in 1977,
a good year class of gamefish was apparently pro-
duced in the East Fork in 1978.
199

-------
The spawning success of pike and walleye (Stewart, 1978; Stewart,
1979) during the last two years will be examined in further detail in
order to predict effects of stable flows during April on gamefish
spawning success. These data are important since they were collected
after closure of Morrison Dam and two quite different flow regimes
were represented: an extremely low flow period in April of 1977,
and a higher flow in April of 1978 which was, however, still below
average for the historical period.
During both years walleye and northern pike in spawning condition
were found in all three Forks and in the main Poplar River. In the
low flow year of 1977, there was an apparent failure of the walleye
year class on the East Fork. In April of 1978, there were many more
larvae of both walleye and pike collected, indicating increased
spawning success when compared with 1977. The East Fork station
referred to as Cromwell had the highest mean density of walleye eggs
per sample when compared to all other 1978 sampling stations. More-
over, in 1978 numbers of young-of-the-year walleye per mile on the
Ea t Fork increased to about 13 times the densities measured in 1977.
Over the entire river basin 1978 densities of young-of-the-year
walleye were approximately double the 1977 densities. There was also
apparently a successful 1978 year-class formation of northern pike
on the East Fork as evidenced by the occurrence of ripe spawners and
young-of-the year fish. It must be emphasized, however, that there
are no actual baseline evaluations of normal fish spawning in the
Poplar River. Therefore, although 1978 was a good year for year
class formation, it may have actually been less than the potential
production under optimal flow conditions.
All of these data indicate that the April, 1977, flows on the East
Fork were not sufficient for successful gamefish spawning, while the
flow regimes in 1978 were adequate for the formation of a good year
class. The conclusion regarding the relative success of the 1978 East
Fork spawning of walleye is further substantiated because there probably
was a reduced spawning population due to winterkill. The large mortal-
ity of walleye observed in that area during February and March was due
to low dissolved oxygen concentrations below the ice cover (Stewart,
1978). Thus, the spawning success might have been even greater with a
normal size spawning population.
Pike typically spawn among submerged vegetation during high flow
or flood conditions. Several studies indicate that pike prefer sub-
merged terrestrial vegetation such as native grasses, mowed hay or
wheat stubble for spawning substrate (McCarraher and Thomas, 1972;
Forney, 1968). Others have found, however, high utilization of sub-
merged plants such as Myriophyllum (Frost and Kipling, 1967) or
emergents such as sedges or rushes (Franklin and Smith, 1963). Appar-
ently, flooding beyond the primary channel is not necessary for success-
ful pike spawning in the Poplar River. This is evidenced by the
200

-------
occurrence of a peak flow of only 80 cfs during April of 1978. In
order to overtop the main channel in the East Fork, a flow in excess
of about 200 cfs would be required.
Examination of the flow data presented in Table 5.6-7 reveals
that during 1977 the East Fork experienced very low flow conditions.
The East Fork April flows at the border and Scobey stations were
considerably less than (‘ 50 percent) natural or historical 10 percent
frequency flows (Table 5.6-8). Moreover, peak April flows reached
only 11 cfs. This condition probably resulted from a combination of
an extreme low runoff year and the blockage of flow at Morrison Dam.
The 1977 Middle Fork flows at the border were relatively higher
since there is no impoundment on that Fork. The April flows were about
20 percent higher than the 10 percent frequency flow on an historical
basis.
The April, 1978, East Fork flows (average and peak) at the
International Boundary were approximately equal to the 1977 measure-
ments. However, flow accretion in the U.S. East Fork sub-basin resulted
in higher peak and average flows in the lower East Fork near Scobey
than occcurred in 1977.
These comparisons indicate that average spawning-period flows of
2.4 to 6.6 cfs are not adequate for successful game fish spawning in
the Poplar River. However, as the average flow in the same river
reach increased to 2.7 to 24.5 cfs, a good year class was formed.
During the successful spawning period, peak flows ranged from 3.3 to
80 cfs, while the poor spawning period was characterized by peak flows
of 2.9 to 11 cfs.
Comparison of 1977-78 East Fork flows with predicted flows under
apportionment reveals that the measured 1978 average April flows were
only slightly higher at Stations 1 and 3 than the predicted flows
under scenario 29 (i.e., 2 units at 1985 water use)(Table 5.6-8).
Therefore, with up to two power plant units, the walleye and pike
spawning on the East Fork should be approximately equal to the 1978
level of success with a return period of 2 years. However, in one out
of ten years with one or more units in operation, the April East Fork
flow would average only 0 to 4.3 cfs, resulting in severe adverse
effects on game fish reproduction. Based on the observed spawning-
flow relationships in the East Fork, it is expected that the flow
apportionment has a lower potential for adverse impact on fish spawning
in the main river and middle forks. However, the apportionment will
also affect flows in the main river and long-term effects on stream
morphology may occur.
201

-------
Table 5.6-7
COMPARISON OF 1977 AND 1978 SPRING FLOWS (cfs) AT SELECTED POPLAR RIVER STATIONS
STAT ION
East Fork
at Border
East Fork
at Scobey
Middle Fork
at Border
Main River
at Poplar
Peak Average Peak Average Peak Average Peak Average
r- .
‘ —
a
—
MARCH
APRIL
MAY
5.8
2.9
58
2.6
2.4
13
22
11
164
8.3
6.6
17.9
15
15
82
9.5
11.1
12.4
168
210
42
112
122
31.3
•
MARCH
APRIL
MAY
6.2
3.3
3.9
3.6
2.7
3.0
220
80
10
69.7
24.5
6.1
829
709
56
225
75.4
25.6
4610
4630
205
1163
640
150
.-4
APRIL
MAY
270.0
139.0
143.0
43.9
-
-
-
-
1620
155.0
325.3
59.7
-
-
-
-

-------
Table 5.6-8
PREDICTED AVERAGE APRIL FLOWS (cfs)
IN THE EAST FORK POPLAR RIVER
PERCENTILE
Scenario 90 50 10
Station 1
East Fork
at Border
1
2
28
29
32
308
301
270
249
63.3
28.6
28.6
11.4
2
2.2
7.7
7.1
1.0
1.0
1.0
Station 3
East Fork
near Scobey
1
2
28
29
32
492
--
437
398
221
48.0
48.4
23.0
20.4
17.7
13.7
--
4.3
2.2
0.0
203

-------
Recent fish data collected by Stewart (1980) provide further evi-
dence for the influence of streamfiow on spawning success of walleye and
northern pike in the Poplar River. In 1979, there was above average
runoff in combination with spring releases from Cookson Reservoir. The
1979 average April flow at the East Fork Border station was 143.0 cfs,
compared to 1977 and 1978 flows of 2.7 and 2.4 cfs, respectively. The
higher 1979 flows resulted in the formation of strong year classes as
evidenced by high densities of 0+ walleye and northern pike in the East
and Middle Forks.
The relationship between spring flow and year class recruitment for
the period 1977-1979 is diagrammed in Figure 5.6—4. Abundances of
young-of-the-year walleye and northern pike at downstream stations show
a positive correlation with mean April flow near the border in both the
East and Middle Forks. Coefficients of determination (r 2 ) for the wall-
eye and northern pike were 0.84 and 0.73, respectively. Although these
data were from two different river sub-sections, and some data are esti-
mates due to low collection numbers, they do provide an indication of the
strong relationship between flow rate and year class recruitment for both
species. Flow appears to be a stronger predictor for young walleye sur-
vival than for northern pike as evidenced by the corresponding r 2 values.
In addition, the slope of the walleye data (82.29) is considerably higher
than that for northern pike (35.23). Therefore, as mean April trans-
boundary flow increases from 10 to 100 cfs the predicted density of young-
of the year walleye increases by 190/mi, while predicted northern pike
density increases only about 80/mi.
Comparison of 1977-1978 East Fork flows with predicted flows under
apportionment reveals that the measured 1978 average April flows were
only slightly higher at Stations 1 and 3 than the predicted flows under
scenario 29 (i.e., 2 units at 1985 water use) (Table 5.6-8). Therefore,
with up to two power plant units, the walleye and pike spawning on the
East Fork should be approximately equal to the 1978 level of success
with a return period of two years. However, in one out of ten years
with one or more units in operation, the April East Fork flow would
average only 0.0 to 4.3 cfs, resulting in severe adverse effects on
game fish reproduction. Based on the observed spawning - flow relation-
ships in the East Fork, it is expected that the flow apportionment has
a lower potential for adverse impact on fish spawning in the main river
and middle forks. However, the apportionment will also affect flows in
the main river and long-term effects on strea n morphology may occur.
The flow-spawning success relationship presented in Figure 5.6-4
may be used to provide a quantitative estimate of impacts on fish popu-
lations due to the power plant operation and apportionment. Based on
natural and historical April transboundary flows (scenarios 1 and 2)
the mean young-of-the-year class strength for walleye and northern pike
is about 235/ni and 77/mi, respectively (Table 5.6-9).
204

-------
500 +
400
300
200
E
0
U i
C-,
z
a
z
cZ
U i
>-
Ii
0
0
z
0
Figure 5.6—4
00
I0
0
- 0
100
MEAN APRIL TRANSBOUNDARY FLOW (cfs)
RELATIONSHIP BETWEEN FLOW AND YEAR-CLASS FORMATION
OF GAME FISH IN THE EAST AND MIDDLE FORKS OF THE
POPLAR RIVER
KEY:
WALLEVE
o NORTHERN PIKE
— BEST FIT LINE FOR WALLEYE
—— BEST FIT LINE FOR PIKE
0
.0
I 000
205

-------
Table 5.6-9
PREDICTED IMPACT OF FLOW APPORTIONMENT ON YOUNG-OF-THE-YEAR
CLASS STRENGTH OF POPLAR RIVER GAME FISH
Predicted
M 0+ Density* Percent
Power Ap l (No./Mile) Reduction
Scenario Plants Flow (cfs) Walleye N. Pike Walleye N. Pike
1-2, Natural- 0 28.6 235 77 0 0
Hi storical
28 Apportionment, 1 11.4 160 45 32 42
1975 use
29 Apportionment, 2 2 <20 <10 >90 >87
1985 use
*
0+ means young-of-the-year class.
206

-------
Operation of one power plant (scenario 28) would result in a 32
percent reduction in the young-of-the—year walleye density during a
median flow year. With two power plants and the 1985 level of develop-
ment (scenario 29) the predicted impact would be a greater than 90 per-
cent reduction in young—of-the-year walleye densities. The corresponding
reductions in northern pike densities for scenarios 28 and 29 are 42
percent and greater than 87 percent, respectively.
Several important points should be emphasized regarding the flow
related impacts on fish:
• The predicted impacts are for the lower portions of
the East Fork. The impacts on the already marginal
fish habitat in the upper East Fork will probably
be greater.
• The impacts are predicted for median flow conditions
under the indicated apportionments. During low flow
years (e.g., tenth percentile) the impacts under
apportionment would be considerably greater, probably
resulting in complete failure of the year class in
the East Fork.
• The predicted impacts are based on the maintenance
of currently existing channel morphology and sub-
strate type. Flow related impacts on such physical
conditions would compound the overall biological
impacts.
• Impacts on young-of-the-year gamefish are not neces-
sarily manifested directly as changes in the size of
adult fish populations. Density dependent mechanisms
may act to ameliorate the initial poor year class
formation, resulting in less effect on subsequent
older age groups.
Mitigation of flow related impacts on Poplar River fish populations
would require maintenance of specified instream flows by controlled
reservoir releases. Two types of flow regimes would be required:
• Maintenance of flow required to provide spawning
habitat for gamefish during April-May.
• Maintenance of peak flows to preserve channel
morphology.
207

-------
Based on the fish data collected by Stewart (1980), it appears that
a transboundary flow of at least 10 cfs during April and May would insure
the successful spawning and rearing of both walleye and northern pike in
the mid and lower East Fork. Although some recruitment would occur at
average flows between 2 and 10 cfs, the long-term effects of such flow
regimes would most likely be a decline in the East Fork gamefish popula-
tions. It should be emphasized, however, that the production of young
aamefish would increase at flows of 10 to 200 cfs. Therefore, a spawn-
ing period flow of 10 cfs should be considered as a minimum for maintain-
ing gamefish spawning and rearing at a level that would sustain gamefish
recruitment in parts of the East Fork.
Based on the flow modeling results, the operation of one power plant
with 1975 water use (scenario 28) would result in a median April trans-
boundary flow of 11.4 cfs. Under such flow conditions there would be
successful gamefish recruitment in the East Fork in average or high flow
years, if channel morphology were maintained. However, during low flow
years (e.g. , 10 percentile, Table 5.6-8) there would be insufficient flow
under the apportionment schedule for successful gamefish spawning. It is
durinq such low-flow years that flow augmentation by increased April-May
reservoir releases would provide mitigation against adverse impacts on
fish populations.
The previous analyses in this section indicate that a peak flow of
about 715 cfs would be required to provide sufficient scouring of fine
material to maintain the present riffle-pool configurations. The required
duration of the peak flow is unknown; however, it would most likely be in
the range of 2.5 days to 33 days. Since flows of this magnitude have not
been historically maintained for a monthly period, the required duration
of peak flows is probably nearer the lower limit of the indicated range.
In order to provide optimal mitigation, the peak flow release should be
provided shortly before the spawning period in April-May.
The Biological Resources Committee (BRC) of the International Poplar
River Water Quality Board (IPRWQB, IJC, 1979) has provided estimates of
water requirements in the East Fork necessary to meet various biological
objectives. The objectives for maintenance of gamefish spawning and
rearing habitat are 15 cfs in April and 10 cfs in May at the International
Boundary. The BRC’s April recommendation is higher than the estimate of
10 cfs contained in this document. The committee’s recorimended require-
ment for maintenance of channel integrity is an 18-day release, consisting
of at least 123 cfs for 16 days and a peak of 700 cfs for two days within
the period. The BRC’s peak flow recommendation is the same as the flow
requirements presented in this document; however, no duration of flow was
contained herein. Based on our evaluation of the overall range required
(2.5 to 33 days), the BRC estimate would be of comparable macnitude.
Montana Department of Fish and Game (1979) has recently published
recommendations for instream flows to protect walleye and northern pike
in the Poplar River. The April and May recommended flows are the same as
the BRC’s estimates, i.e., 15.0 and 10.0 cfs, respectively. Table 5.6-10
208

-------
Table 5.6-10
RECOMMENDED INSTREAM FLOWS FOR THE EAST FORK OF THE POPLAR RIVER
Month
January
February
March
April
May
June
July
August
September
October
November
December
Upper Reach*
2.0
2.0
5.0
15.0
10.0
5.0
3.0
3.0
3.0
3.0
2.0
2.0
3.0
3.0
5.0
15.0
10.0
5.0
4.0
4.0
4.0
4.0
3.0
3.0
*
Boundary to Highway 13 bridge
tHighway 13 bridge to mouth
East Fork Flow (cfs )
Lower Reacht
209

-------
presents the recommended flows for the remainder of the year for both upper
and lower reaches of the East Fork.
Flow recommendations (Montana Department of Fish and Game, 1979) were
based on the previously discussed spawning flow requirements and channel
maintenance flows. In addition, flows for the remainder of the year were
based on the observations that game fish seem to be maintained at flows
less than mean flows. The months of January and February are exceptions
in that the recommended flows are higher than mean flows due to the poten-
tial for under-ice mortalities at low flows.
The overall effects of apportionment on Poplar River game fish will
be primarily associated with changes in spawning period flows and channel
forming flows. The effects of modified instream flows during the remainder
of the year are much more difficult to quantify. Although predicted flows
under apportionment are less than recommended instream flows under low
flow conditions, the effects on fish are unknown. A beneficial effect of
apportionment may exist, however, during January—February of low flow
years. Under such conditions, natural (predevelopment) flows may reach
very low levels (see Figure 5.2-16). Under the recommended apportionment,
flow would be maintained at 1.0 cfs, only slightly less than existing
conditions (1975 use and reservoir) which would tend to mitigate against
under-ice mortality of ganefish. Therefore, the 1.0 cfs release would
tend to mitigate against winter gamefish mortalities although it is lower
than the Montana Department of Fish and Game recommendation of 2.0 cfs
for the upper East Fork. Actual flows would be higher due to ground water.
The evaluation of flow related impacts on fish has been oriented
towards walleye and northern pike, the two major game species in the Poplar
River. Modifications in flow may effect other species as well; however,
impact generalizations are difficult to develop due to the variety of fish
species present and the lack of specific life history infornation on non-
game species. Some species may be adversely affected by reduced flow,
while others may benefit from conditions under flow apportionment.
Modified flow regimes can also be expected to influence the corununity
composition of benthic macroirivertebrates in the Poplar River. Previous
studies have indicated that many of the Epheneroptera and Trichoptera of
the Poplar River are dependent upon flowing water habitat. The long—term
changes in stream morphometry predicted under apportionment will result
in a reduction of stream dependent taxa and an increase in macroinverte-
brate fauna adapted to ponds or slow moving streams. The overall effects
are difficult to assess due to the complex interrelationships of potentially
lowered densities due to the loss of riffle habitat and the potential in-
creases due to stimulation of macrophyte production.
210

-------
6. REFERENCES
Accu-Labs Research, Inc. 1978. 11485 West 48th Avenue, Wheat Ridge,
Colorado. Letter sent to Terry Thoem, EPA, concerning results of Poplar
River ash sample analysis.
Agarwala, S.C., N.K. Mehrotra and B.K. Sinha. 1964. Influence of ex-
changeable sodium of the growth and mineral composition of plants.
Journal of the Indian Soc. of Soil Sci. 12:7-23.
Allbaugh, C.A. , and J.V. Manz. 1964. Preliminary study of the effects
of temperature fluctuations on developing walleye eggs and fry. Prog.
Fish Cult. 16:175-180.
American Hospital Association Guide to the Health Care Field, 1977
Edition.
American Medical Association. 1976. Physician distribution and medical
licensure in the United States.
American Public Health Association (APHA). 1976. Standard Methods.
Asana, R.D. and V.R. Kale. 1965. A study of salt tolerance of four
varieties of wheat. Indian Journal of Plant Physiology. VIII:5-22.
Ayers, A.D., J.W. Broun and C.H. Wadleigh. 1952. Salt tolerance of
barley and wheat in soil plots receiving several salinization regimes.
Agronomy Journal 44:307-310.
Bahls, Loren L. 1977. Spring and summer phycoperiphyton of the Poplar
River system, Montana: A pre-energy development, baseline floristic
survey of a transboundary river. Prepared for Water Quality Bureau,
Montana Department of Health and Environmental Sciences.
Bahls, L.L. 1979. International Joint Commission - International
Poplar fliver Water Quality Board. Biol. Res. Corn. Final Report. Aquatic
Plant Section. 133 p.
Bains, et al. 1970. Relative salt tolerance of barley varieties.
Indian Journal of Agronomy. XV:356-358.
Bellrose, F.C. 1976. Ducks, Geese, and Swans of North t merica. Wild-
life Management Institute.
Bennett, S.H., A.C. Hill, A. Soleirnani and W.H. Edwards. 1975. Acute
effects of combination of sulphur dioxide and nitrogen dioxide on
plants. Environmental Pollution 9:127—132.
211

-------
Bernstein, L. and L.E. Francois. 1973. Leaching requirement studies:
sensitivity of alfalfa to salinity of irrigation and drainage waters.
Soil Science Soc. Am. Proc. 37:931-943.
Bernstein, L., L.E. Francois, and C.A. Clark. 1974. Interactive
effects of salinity and fertility on yields of grains and vegetables.
Agronomy Journal 66:412-420.
Bernstein, L. and G.A. Pearson. 1956. Influence of exchangeable
sodium on the yield and chemical composition of plants: I. green beans,
garden beets, clover and alfalfa. Soil Science Soc. Am. Pro. 82:247-258.
Biggar, J.W. and M. Fireman. 1960. Boron adsorption and release by
soils. Soil Science Soc. Am. Proc. 24:115-119.
Binqham, F.T., A.L. Page, N.T. Coleman, K. Flach. 1971. Boron adsorp-
tion characteristics of selected amorphous soils from Mexico and Hawaii.
Soil Science Soc. Am. Proc. 35:356-550.
Blood, D.A., G.G. Anweiler and J. Polson. 1976. Wildlife Impact Study -
Poplar River Thermal Power Project--Final Report. Prepared for the
Saskatchewan Power Corporation, August, 1976.
Bloom, and Botz. 1975. Water Quality Inventory and Management Plan.
Lower Missouri River Basin, Montana. pgs 20-21 (Biological conditions).
Bower, C.A. 1961. Cation-exchange equilibria in soils affected by
sodium salts. Soil Science Soc. Am. Proc. 32-35.
Bower, C.A. and L.V. Wilcox. 1965. Precipitation and solution of cal-
cium carbonate in irrigation operations. Soil Science Soc. Am. Proc.
29:93-94.
Bower, C.A., G. Ogata and J.M. Tucker. 1968. Sodium hazard of irriga-
tion waters as influenced by leaching fraction and by precipitation or
solution of calcium carbonate. Soil Science Soc. Am. Proc. 106:29-34.
Brady, N.C. 1974. The Nature and Properties of Soils. Mac Millian
Publishing Company, Inc., New York.
Braun, 0. 1978. Vegetation chemistry, In: Geihaus, J.W., Background
Air Quality Studies of the Poplar River Area of Northeastern Montana,
1977. Montana Department of Health and Environmental Sciences, Air
Quality Bureau.
Breeder, C.M. and D.E. Rosen. 1966. Modes of Reproduction in Fishes.
Natural History Press, Garden City, New York.
Briqgs, G.A. 1972. Discussion on chinney plumes in neutral and stable
surroundings. Atmospheric Environment 6:507-610.
212

-------
Brigys, G.A. 1969. Plume Rise . U.S. Atomic Energy Comission critical
review series, TID—25075, National Technical Information Service, Spring-
field, Virginia 22151.
Briggs, G.A. 1971. Some recent analyses of plume rise observations,
pp. 1029-1032 in Proceedings of the Second International Clean Air Con-
gress, edited by H.M. Englund and W.T. Berry, Academic Press, New York.
Broomfield, A.R. 1972. Absorption of atmospheric sulphur by mustard
growth in a glass house. J. Agric. Sci. 78:343-344.
Brown, L. September 11, 1978. Unpublished Report, Montana Department
of Natural Resources and Conservation, Hele’ia, Montana.
Brown, L. 1979. Personal Comunication, Hydrologist, Montana Department
of Natural Resources and Conservation, Helena, Montana. February 8,
1979.
Bureau of Indian Affairs. August, 1974. Profile of the Montana Native
American.
Cairnes, Saskatchewan Power Company, telephone conversation, May, 1979.
Chalmers, J.E., etal. 1979. An empirical model of spatial interaction
in sparsely populated regions, prepared for presentation at the Twenty-
Fourth North American Meeting, Regional Science Association in November.
Chana, C.W. 1961. Effects of saline irrigation water and exchangeable
sodium on soil properties and growth of alfalfa. Soil Science Soc. Am.
Proc. 91:29-37.
Chang, C.W. and H.E. Dregne. 1955. Effect of exchangeable sodium on
soil properties and on growth and cation content of alfalfa and cotton.
Soil Science Soc. Am. Proc. 19:29-35.
Chauhan, R.P.S. and S.L. Powar. 1978. Tolerance of wheat and pea to
boron in irrigation water. Plant and Soil 50:145-149.
Coiwell, W.E. and R.W. Cummings. 1944. Chemical and Biological Studies
on Aqueous Solution of Boric Acid and of Calcium, Sodium and- Potassium
Metaborates. Soil Sciences Soc. Am. Proc. 57:37-49.
Cowling, D.W., L.H.P. Jones and D.R. Lockyer. 1973. Increased yield
through correcting of sulfur deficiency in ryegrass exposed to sulfur
dioxide. Nature 243:479-480.
DeSimone, R. 1978. Poplar River wildlife study, Quarterly Progress
Reports (February, June, September). Montana Department of Fish and
Game, Ecological Services Division, sponsored by U.S. Environmental
Protection Agency.
213

-------
DeSimone, R. 1979. Poplar River waterfowl study: Final Report. Montana
Department of Fish and Game, sponsored by U.S. Environmental Protection
Agency.
DeSirnone, R. 1980. Poplar River waterfowl study progress report.
Montana Department of Fish and Game, sponsored by U.S. Environmental
Protection Agency.
Dodd, J.L., W.K. Laurenroth, R.K. Heitschmidt and J.W. Leetham. 1978.
First-year Effects of Controlled Sulfur Dioxide Fumigation on a Mixed-
grass Prairie Ecosystem. In: The Bioenvironmental Impact of a Coal-fired
Power Plant, E.M Preston and R.A. Lewis, eds. Third Interim Report, EPA-
600/3-78-021, U.S. Environmental Protection Agency, Corvallis, Oregon.
pp. 345-375.
Dreisinger, B.R. and P.C. McGovern. 1970. Monitoring atmospheric sulfur
dioxide and correlating its effect on crops and forests in the Sudbury
area. In: Proceedings of the conference on the Impact of Air Pollution
on vegetation. Toronto, Ontario, Canada.
Duggan, J.S. February, 1979. Poplar River study, Poplar River, Montana,
August, 1978. Lockheed Electronics Company, Inc. , for Office of Research
and Development, U.S. Environmental Protection Agency, Las Vegas, Nevada.
Dvorak, A.J., etal. 1977. The environmental effects of using coal for
generating electricity. NUREG-0252. Prepared for U.S. Nuclear Regulatory
Commission by Argonne National Laboratory, Argonne, Illinois.
Eaton, F.M. 1974. Deficiency, toxicity, and accumulation of boron in
plants. Journal of Agricultural Research 69:237-277.
Eaton, F.M. and L.V. Wilcox. 1939. Behavior of Boron in Soils. USDA
Technical Bulletin 696. 58 pp.
Edinger, J.E. and E.M. Buchak. 1975. A hydrodynamic, two-dimensional
reservoir model: The computational basis. J.E. Edinger Associated, Inc.
Wayne, Pennsylvania.
Elgabaly, M.I1. 1955. Specific effects of adsorbed ions on plant growth:
I. effect of different combinations of calcium, magnesium and sodium on
barley seedlings. Soil Science Soc. Am. Pro. 80:235-248.
Evans, C.D. and K.E. Black. 1956. Duck production studies on the prairie
potholes of South Dakota. U.S. Fish and Wildlife Service Spec. Sci. Rept.,
Wildi. flo. 32. 59 pp.
Evans, R.G., W.R. t’Ialker, G.V. Skogerboe and S.W. Smith. 1978. Evalua-
tion of irrigation methods for salinity control in Grand Valley. EPA-600/
2-78-161. U.S. Environmental Protection Agency, Ada, Oklahoma.
214

-------
Faller, N. 1970. Effects of atmospheric 502 on plants. Sulphur Inst.
J. 6:5-7.
Faller, N. 1971. Plant nutrient sulphur: SO 2 versus SO 4 . Sulphur
Inst. J. 7:5-6.
Feltis, R.D. 1978. Water resources of shallow aquifers in the Upper
Poplar River Basin, Northeastern Montana. U.S. Geological Survey,
Montana Bureau of Mines and Geology with U.S. Environmental Protection
Agency.
Forney, J.L. 1968. Production of young Northern Pike (Esox lucius) in
a regulated marsh. New York Fish Game J. 15(2):143-154.
Fowler, D. 1978. Dry deposition of SO 2 on agricultural crops. Atmos-
pheric Environ. 12:369-373.
Fox, R.H. 1968. The effect of calcium and pH on boron uptake from
high concentrations of boron by cotton and alfalfa. Soil Science Soc.
Am. Proc. 106:435-439.
Franklin, D.R. and L.L. Smith, Jr. 1963. Early life history of the
northern pike, Esox lucius L., with special reference to the factors
influencing the numerical strength of year classes. Trans. Amer. Fish.
Soc. 92(2):91-110.
Frost, W.E. and C. Kipling. 1967. A study of reproduction, early life,
weight-length relationships and growth rate of pike, Esox lucius L. in
Windermerl, J. Animal Ecology 36(3):651-693.
Garland, J.A. 1978. Dry and wet removal of sulphur from the atmosphere.
Atmospheric Environ. 12:349-362.
Gelhaus, J.W. and M.D. Roach (Eds). 1979. Issue paper by the Scientific
and Engineering Advisory Panel on Poplar River Air Quality. Air Quality
Bureau, Montana Dept. Health and Environ. Sci. 31 p.
Geihaus, J.W. and M. Machler. 1978. Background air quality studies of
Poplar River area of northeastern Montana, 1977. Part II, meteorology,
Montana Department of Health and Environmental Sciences, Air Quality
Bureau.
Gelhaus, J.W., M.G. Machler, S.Z. Schwartz and M.D. Roach. 1979. Back-
ground air quality studies of Poplar River area of northeastern Montana,
1978. Montana Department of Health and Environmental Sciences, Air
Quality Bureau.
Griffin, R.A. and R.G. Burau. 1974. Kinetic and equilibrium studies of
boron desorption from soil. Soil Science Soc. Am. Proc. 38:892-897.
215

-------
Griffith, D.W., etal. 1978. Enterprise costs for fallow, winter wheat,
spring wheat, barley, recrop safflower in Roosevelt County. Bulletin
1179, Cooperative Extension Service, Montana State University, Bozeman,
Montana.
Hadas, A. and J. Hagin. 1972. Boron adsorption by soils as influenced
by potassium. Soil Sciences Soc. Am. Proc. 113:189-193.
Hanks, R.J., T.E. Sullivan and V.E. Hunsaker. 1977. Corn and alfalfa
production as influenced by irrigation and salinity. Soil Science Soc.
Am. Proc. 41:606-609.
Hargreaves, G.H. 1973. Estimation of potential and crop evapotranspira-
tion. American Society of Agricultural Engineers. Paper No. 73-4537.
Hassan, N.A.K., etal. 1970. Influence of soil salinity on production
of dry matter, uptake and distribution of nutrients in barley and corn:
I. Barley. Agronomy Journal 62:43-45.
Hatcher, J.T. and L.A. Bower. 1958. Equilibria and dynamics of boron
adsorption by soils. Soil Science Soc. Am. Proc. 85:319-323.
Hill, A.C., S. Hill, C. Lamb and T.W. Barrett. 1974. Sensitivity of
native desert vegetation to 502 and S02 + NO 2 . Journal of Air Pollution
Control Association 24.
Hokanson, K.E.F., J.H. McCormick and B.R. Jones. 1973. Temperature
requirements of northern pike, Esox lucius, eggs and larvae. Trans.
Amer. Fish. Soc. 102: (In press).
Holzworth, G.C. 1972. Mixing heights, wind speeds, and potential for
urban air pollution throughout the contiguous United States, Environmen-
tal Protection Agency, 118 pp.
Horpestad, A. 1978. Analyses of irrigated soils from the Poplar River
Basin. Water Quality Bureau, Environmental Sciences Division, Montana
Department of Health and Environr ental Sciences. 26 pp.
Hosler, C.R. 1961. Low-level inversion frequency in the contiguous
United States, monthly weather review. pp. 319-339.
Howard, A.D. 1960. Cenozoic History of northeastern Montana and north-
western North Dakota with emphasis on the Pleistocene, U.S. Geological
Survey Prof. Paper 326, 107 pp.
Ingvalson, R.D. , J.D. Rhoades and A.L. Page. 1976. Correlation of
alfalfa yield with various index of salinity. Soil Science Soc. Am.
Proc. 122:145-153.
International Joint Commission. 1978. Water apportionment in the
Poplar River Basin. An IJC report to the Governments of Canada and the
United States.
216

-------
International Joint Commission. 1979. International Poplar River Water
Quality Study. Final Report and Appendixes.
International Joint Commission. 1979. Transcripts of Public Hearings
on the Study of the International Poplar River Water Quality Board held
at Scobey, Montana on the 10th day of September, 1979 and at Coronach,
Saskatchewan on the 11th day ofSeptember, 1979. 274 pp. and 290 pp.
Johnson, F.H. 1961. Walleye egg survival during incubation on several
types of bottom in Lake Winnibigoshish, Minnesota, and connecting waters.
Trans. Amer. Fish. Soc. 90:312-322.
Kamphorst, A. and G.H. Bolt. 1976. Saline and Sodic Soils in Bolt,
G.H. and M.G.M. Bruggenwert, ed., Soil Chemistry A. Basic Elements. New
York: Elsevier Scientific Publishing Company 171-191.
Karp, R.W. 1978. Poplar River model interim report. Water Quality
Bureau, Environmental Sciences Division, Department of Health and Environ-
mental Sciences. 54 pp.
Klages, M. 1976. Water quality requirements for irrigation in the Poplar
River Valley. Montana Health and Environmental Sciences, Helena. 8 pp.
Kiarich, Duane, A. 1978. Surface water site inventory of the Poplar
River drainage and preliminary considerations of potential water quality
impacts from the coal-fired, power generation development near Coronach,
Saskatchewan. U.S. Environmental Protection Agency, Grant No. R008115.
Koenst, W.M. and L.L. Smith, Jr. 1976. Thermal requirements of the
early life history stages of walleye, Stizostedion vitreum vitreum, and
sauger, Stizostedion canadense. J. Fish. Res. Board, Canada. Vol. 33.
Latimer, Douglas A. 1980. Power plant impacts on visibility in the
west: siting and emissions control implications. J. Air Poll. Cont.
Assoc. 30:142-146.
Likens, G.E., F.H. Bormann, N.M. Johnson and R.S. Pierce. 1967. Ecol.
48:772.
Lizon, S.N. 1978. Effects of air-borne sulfur pollutants on plants.
In: J.0. Nriagu, Ed. Sulfur in the environment. Part II. ecological
impacts. Wiley-Interscience. New York, pp. 109-162.
Longwell, C.R., R.F. Flint and J.E. Sanders. 1969. Physical Geology.
John Wiley and Sons, Inc., New York.
Lucas, Merle Source. August, 1974. Profile of the Montana native
American. Urban management consultants. -
Luft, L.D. 1979a. Feasibility--Side roll irrigation system, dryland
wheat to irrigated wheat, unpublished paper. Cooperative Extension
Service, Montana State University, Bozeman, Montana (North Dakota).
217

-------
Luft, L.D. and D.W. Griffith. July, 1978. Costs of owning and operating
eight hay harvesting, handling and feeding systems. Bulletin 1149,
Cooperative Extension Service, Montana State University, Bozeman, Montana.
Mathew, E. May, 1979. Senior resident engineer, Saskatchewan Power
Company, telephone conversation.
McCarraher, D.B. and R.E. Thomas. 1972. Ecological significance of
vegetation to Northern Pike (Esox lucius) spawning. Trans. Am. Fish.
Soc. 101(3):560-563.
Mckee, J.E. and H.W. Wolf, Eds. 1963. Water quality criteria. Publica-
tion 3-A, California State llater Resources Board, Sacramento, California.
Mehrotra, C.L. and B.R. Gangwar. 1964. Studies on salt and alkali tol-
erance of some important agricultural crops in Uttar Pradesh. J. Indian
Soc. Soil Sci. 12:75-84.
Mesich, E.G. and H.L. Taylor. 1976. Levels of trace elements in the
ambient air at selected locations in the Northern Great Plains. NTIS
PB-265863. 35 pp.
Montana Department of Agriculture. 1976. Montana agricultural statis-
tics.
Montana Department of Community Affairs. 1974. Human resources situa-
tion report for action for Eastern Montana (AEM).
Montana Department of Community Affairs. July, 1978. Montana population
projections 1980-2000. summary. Research and Information Systems Divi-
sion.
Montana Department of Community Affairs. 1978. Division of Research
and Information Systems. Profiles for Daniels and Roosevelt counties.
Montana Department of Community Affairs. 1979. Division of Research
and Information Systems, County profiles. Unpublished update, May.
Montana Department of Fish and Game. 1976. A summary of fish popula-
tion investigations in the Poplar River drainage. July 1, 1975 through
March 31, 1976.
Montana Department of Fish and Game. 1978. Lower Missouri River Basin
investigation. Planning inventory game. Ecological Services Division,
job progress report. July 1, 1977 through June 30, 1978.
Montana Department of Fish and Game. 1979. Instream flow evaluation
for selected streams in the upper Missouri River Basin. Department
Report. Helena, Montana. 250 p.
Montana Department of Health and Environmental Studies, Water quality
standards. MAC 16-2.14(10)-514480.
218

-------
Montana Department of Health and Environmental Sciences. 1978. Mining
and mineral fuel development in the Montana statewide 208 study area
implications to water quality.
Montana Department of Labor and Industry. 1977. Montana Employment
and Labor Force, Employment Security Division, Vol. 7, No. 7, July, and
Vol. 8, No. 3, March.
Montana Department of Natural Resources and Conservation. 1975. Water
use in Montana. Inventory series report No. 13, April, 1975.
Montana Department of Natural Resources and Conservation. 1978. Final
report on water uses in the Poplar River Basin.
Montana Deoartrnerit of Natural Resources and Conservation. 1978. Deter-
mination of minimum instream flows, sediment discharge, and stream chan-
nel formation related to the balance of natural maintenance and preser-
vation of fishery/wildlife habitat and manipulation of the Poplar River
by man.
Montana Energy Advisory Council. 1976. Coal development information
packet (updated in 1977).
Morrison-I’laierle, Inc. and Fort Peck Sioux and Assiniboine Tribes. 1978.
Poplar River existing and future uses, Fort Peck Indian Reservation,
Montana, for Environmental Protection Agency, Denver, Colorado.
Mukammal, E.I. 1976. Review of present knowledge of plant injury by
air pollution. WMO Tech. Mote No. 147. World Meteorological Organiza-
tion, Geneva, Switzerland.
National Academy of Sciences. 1975. Air Quality and Stationary Source
Emission Control. Report by the Comission on Natural Resources,
National Academy of Sciences, National Academy of Engineering, National
Resource Council, U.S. Senate Committee on Public Works, Serial 94-4,
U.S. Government Printing Office, Washington, D.C.
National Association of Counties, International City Managenent Associa-
tion. 1977. The county year book, profiles of individual counties.
National Ecological Research Laboratory. 1974. The bioenvironnental
impact of a coal-fired power plant. Colstrip, Montana. Corvallis,
Oregon.
Nelson, H., H.F. Travland, D.W. Anderson, P. Ag., R.E. McAllister, P.
Ag. and A.H. Klesse, P. [ rig. 1977. The reclamation of mine spoil at
Coronach, Saskatchewan.
Noggle, J.C. and H.C. Jones. 1979. Accumulation of atnospheric sulfur
by plants and sulfur-supplying capacity of soils. NTIS PB-298 875. 37 p.
219

-------
Northern Great Plains Resources Program. 1974. Surface resources work
group - regional profile discussion draft.
Nyborg, M. 1978. Sulfur pollution and soil. In: J.O. flriagu, Ed.
Sulfur in the environment. Part II: ecological impacts. Wiley-Inter-
science, New York, pp. 359-390.
Olson, R.V. and K.C. Berger. 1946. Boron fixation as influenced by pH,
organic matter content and other factors. Soil Science Soc. Am. Proc.
11:216-220.
Oseid, D.M. and L.L. Smith. 1971. Survival and hatching of walleye
eggs at various dissolved oxygen levels. Progressive Fish-Culturist,
April.
Parfet, M. 1980. A gathering storm over syn fuels on the big sky range.
Smithsonian Magazine - March issue. p. 70-78.
Parizek, R.R. 1964. Geology of the Willow Bunch Lake area (72-H),
Saskatchewan. Saskatchewan Research Council, Geology Division, Report
No. 4.
Patel, A.S. and N.G. Dastane. 1971. Effect of different soil textures,
levels of leaching and salinity on radish and barley. Indian J. Agron.,
16:327-33.
Peterka, John. 1972. Effects of saline waters upon survival of fish
eggs and larvae and upon the ecology of the fathead minnow in North
Dakota. North Dakota Water Resources Research Institute, Fargo, North
Dakota.
Peterka, J.J. and J.S. Kent. 1976. Dissolved oxygen, temperature, sur-
vival of young at fish spawning sites. U.S. Environmental Protection
Aciency Report. Environmental Research Laboratory, Duluth. Report No.
600/3-76-113.
Poplar River Task Force. 1976b. Joint studies for flow apportionment
Poplar River Basin in Saskatchewan and Montana, Appendix C: Probable
future use.
Portelli, R.V. 1975. Meteorological aspects of airborne pollutant
dispersion from the proposed Poplar River Thermal Generating Station,
Saskatchewan. Atmospheric Environment Service, report ARQT-3-75.
Rhoades, J.D. 1972. Quality of water for irrigation. Soil Science
113:277-284.
Rhoades, J.D., R.D. Ingvalson and J.T. Hatcher. 1970. Laboratory
determination of leachable soil boron. Soil Science Soc. Am. Proc.
34:871—87 5.
220

-------
Rice, P.M., L.H. Pye, R. Boldi, J. O’Louqhlin, P.C. Tourangeau and C.C.
Gordon. 1979. The effects of “low level S02” exposure on sulfur accurnu-
lation and various plant life responses of some major grassland species
on the ZAPS sites. In: Bio-environmental impact of a coal-fired power
plant. Fourth interim report, Coistrip, Montana. 1978. EPA-600/3-79-044.
pp. 494-591.
Sales and Marketing Management magazine, Survey of buying power (1974-
1978).
Saskatchewan Power Corporation. 1977. Poplar River, thermal power project
environmental assessment and mine reclamation. Summary report.
Saskatchewan Power Corporation. 1976. Coal and environmental programs:
Poplar River coal test burn at B.D.P.S. Unit #1, December 19, 1975 and
May 5, 1976.
Saskatchewan Power Corporation. 1978. Poplar River, thermal power project
environmental assessment and mine reclamation--supplementary report, water
quality and fisheries.
Saskmont Engineering Company, Ltd. 1978. Poplar River Reservoir, final
environmental assessment. Vol. I, January, 1978, for Saskatchewan Power
Corporation.
Schaefer, J., etal. 1978A. Costs of dryland crop production in Sheridan
and Daniels counties. Revised to accompany Bulletin 1133, Cooperative
Extension Service, Montana State University, Bozeman, Montana.
Schaefer, J. , et al. 19788. Costs of irrigated crop production in Blame
County. Circular 1146 (revised), Cooperative Extension Service, Montana
(July, 1978A).
Schwab, G.0., R.K. Frevert, T.W. Edminister and K.K. Barnes. 1966. Soil
and water conservation engineering. John Wiley and Sons, Inc., New York.
p. 543.
Scriven, R.A. and B.E.A. Fisher. 1975. The long range transport of air-
borne material and its removal by deposition and washout. Atmospheric
Environ. 9:49-68.
Shourbagy, l 1.E. and A. Wallace. 1965. Sodium accumulation and sodium
responses of five varieties of barley. Agronomy Journal 57:449-450.
Shy, C.M. 1978. Health consequences of human exposure. In: J.0.
Nriagu, Ed., Sulfur in the environment. Part II: ecological impacts.
Wiley-Interscience, New York. pp. 75-108.
Siefert, R.E., W.A. Spoor, and R.F. Synett. 1973. Effects of reduced
oxygen concentrations on northern pike (Esox lucius) embryos and larvae.
Journal Fish. Res. Board, Canada. 30:849-852.
221

-------
Singh, S.S. 1964. Boron adsorption equilibrium in soils. Soil Science
Soc. Am. Proc. 98:383-387.
Smetana, D.R. 1979. Soil Scientist, Soil Conservation Service, Wolf
Point, Montana. Personal Communication.
Sposito, G. and S.V. Mattigod. 1977. On the chemical foundation of the
sodium adsorption ratio. Soil Science Soc. Am. Proc. 41:323-329.
Spraggs, L.D. 1977. SRSWQ-Souris River study water quality model.
Canadian i Iater Resources Engineering Corporation Report to Environment.
Canada.
Stecher, P.G., Ed. 1968. The Merck Index. Rahway, New Jersey. The
Merck and Company, Inc. p. 160.
Stevens, T.H. and T.W. Hazelton. 1976. Sulfur dioxide pollution and
crop damage in the Four Corners region: A simulation and analysis.
New Mexico State University Agricultural Experimental Station Bulletin
No. 647.
Stewart, B.A. 1975. Control of water pollution from cropland Volume
I, a manual for guideline development. U.S. Environmental Protection
Agency - U.S. Department of Agriculture EPA-600/2-75-026a.
Stewart, B.W. 1978. Poplar River fisheries study progress report.
Sponsored by: U.S. Environmental Protection Agency. File EPA 18-6-2-8.
Stewart, B.W. 1979. Poplar River fisheries study progress report.
Montana Department of Fish and Game, sponsored by: U.S. Environmental
Protection Agency. 19 p.
Stewart, B.W. 1980. Poplar River fisheries study: Final report.
Montana Department of Fish and Game, sponsored by: U.S. Environmental
Protection Agency.
Stoneberg, Ronald P. August, 1977. Fish and game planning, lower
Missouri River - game. July 1, 1976-June 30, 1977, Montana.
Stoudt, J.H. 1971. Ecological factors affecting waterfowl production
in the Saskatchewan parkiands. Res. publication 99 of the Bureau of
Spt. Fish and Wildlife., Washington, D.C. 58 pp.
Sukava, L.F. 1977. Poplar River, a recreation development plan for
the East Poplar River Reservoir and Plant. Prepared by the Tourism and
Recreation Planning Branch, Department of Tourism and Renewable Re-
sources for the Saskatchewan Power Corporation.
Tanner, C.E. 1977. Landsat land cover study of Poplar River Basin,
Montana. Prepared for Environmental Monitoring and Support Laboratory,
U.S. Environmental Protection Agency, Contract 68-03-2153.
222

-------
Thomas, R.E. and D.M. Whiting. 1977. Annual and seasonal precipitation
probabilities. EPA—600/2-77-182. U.S. Environmental Protection Agency,
Ada, Oklahoma, August.
Tingey, D.T., L. Bard and R.W. Field. 1978. The relative sensitivity
of selected plant species to several pollutants singly and in combina-
tion. From: The bio-environmental impact of a coal-fired power plant,
third interim report: U.S. Environmental Protection Agency, Ecological
Research Series.
Torres, C. and F.T. Bingham. 1973. Salt tolerances of Mexican wheat:
I. effect of N03 and MaCi on mineral nutrition, growth, and grain
production of four wheats. Soil Science 37:711-715.
Turner, D. Bruce. 1970. Workbook of atmospheric dispersion estimates,
U.S. Department of Health, Education and Welfare, Cincinnati, Ohio.
U.S.D.A. and Montana Department of Agriculture, Montana Agriculture
County Statistics, Vol. XVI, 1974 and 1975.
U.S. Department of Agriculture. 1974. Irrigation Guide for Montana.
Soil Conservation Service, Bozeman, Montana.
U.S. Department of Commerce, Bureau of the Census, U.S. Census of Housing.
1960. Vol. 1, PT. 28; 1970, First and fourth count summary tape and
Montana housing.
U.S. Department of Commerce, Bureau of the Census. 1975. County business
patterns, Montana, 1975. (CBP75-28)
U.S. Department of Commerce, Bureau of the Census. 1976. Population
estimates: Estimates of the population of Montana counties and metropol-
itan areas; July 1, 1974 and 1975, (June, 1976).
U.S. Department of Commerce, Bureau of the Census. 1976. Population
estimates: Estimates of the population of North Dakota counties and
metropolitan areas; July 1, 1974 and 1975, (July, 1976).
U.S. Department of Commerce, Bureau of the Census. 1977. Industry-
specific gross output multipliers for BEA economic areas (January, 1977).
U.S. Department of Commerce, Bureau of the Census, current population
report, Federal-state cooperative program for population estimates,
Series P-26, No. 109, May.
U.S. Department of Commerce, Bureau of the Census, current population
reports, population estimates and projections, Series P-25, No. 517,
May.
U.S. Department of Commerce, Bureau of the Census, current population
reports, Federal-state cooperative program for population estimates,
Series P-26, No. 53, February and r Io. 76-26, July.
223

-------
U.S. Department of Commerce, Bureau of Economic Analysis, Regional
Economic Information System (magnetic tape).
U.S. Department of Commerce. 1968. Climatic Atlas of the United States,
Environmental Science Service Administration — Environmental Data
Services, June.
U.S. Environmental Protection Agency. 1973a. Effects of sulfur oxides
in the atmosphere on vegetation. Revised chapter 5 for air quality
criteria for sulfur oxides, EPA-R3-73-030; PB-226 314.
U.S. Environmental Protection Agency. 1975. Water quality criteria.
Committee on water quality criteria.
U.S. Environmental Protection Agency. 1976. Quality criteria for
water. U.S. Government Printing Office, Washington, D.C.
U.S. Environmental Protection Agency. 1977a. National interim primary
drinking water regulations. U.S. Government Printing Office, Washington,
D.C.
U.S. Environmental Protection Agency. 1977b. National secondary drink-
ing water regulations. Federal Register, Vol. 42, No. 62.
U.S. Geological Survey. 1968. Mineral and water resources of Montana.
U.S. Senate Document No. 98. U.S. Government Printing Office, Washington,
D.C.
Vaughan, B.E., et al. 1975. Review of potential impact on health and
environmental quality from metals entering the environment as a result
of coal utilization. Battelle energy progress report, Pacific Northwest
Laboratories - Battelle Memorial Institute, Richiand, Washington.
Wahhab, A. 1961. Salt tolerance of various varieties of agricultural
crops at the germination stage. In: Salinity Problems in Arid Zones:
Proceedings of the Teheran Symposium. p. 185-192.
tJalker, tI.R., G.V. Skogerboe and R.G. Evans. 1978. Best management
practices for salinity control in Grand Valley. EPA-600/2-78-162, July.
U.S. Environmental Protection Agency, Ada, Oklahoma.
Werkhoven, C.H.E. 1964. Boron in some saline and nonsaline soils in
southeastern Saskatchewan. Soil Science Soc. Am. Proc. 28:542-545.
Whetstone, R.R., W.O. Robinson and H.G. Byers. 1942. Boron distribution
in soils and related data. USDA Technical Bulletin 797.
Whitaker, S.H. and E.A. Christiansen. 1972. The Empress Group in
southern Saskatchewan, Canadian Journal of Earth Sciences, Vol. 9, No. 4
April, p. 353-360.
224

-------
Whitman, W.R. 1976. Impoundments for waterfowl. Canada Wildlife
Services Occas. Paper Mo. 22, 22 pp.
Wilhour, R.G., Grady E. Neely, David E. Weber and Louis Grothaus. 1977.
Response of selected small grains, native range grasses and alfalfa to
sulphur dioxide. Corvallis Environmental Research Laboratory, Corvallis,
Oregon.
Wilhour, R.G. and D.E. Weber. 1975. Response of small grains and
forages to sulfur dioxide: A literature review. Corvallis Environmental
Research Laboratory. Corvallis, Oregon.
Wilhour, R.G., G.E. Meely, D.E. Weber and L.C. Grothaus. 1979. Response
of selected small grains, range grasses and alfalfa to sulfur dioxide.
In: Bio-environmental impact of a coal-fired power plant. Fourth interim
report, Colstrip, Montana, 1978. EPA-600/3-79-O44. pp. 592-609.
Yabroff, I.W. and E.M. Dickson. 1979. Coal resources model: An impact
assessment tool, draft final report, SRI International, Menlo Park,
California (February, 1979).
Zison, S.W., K.F. Haven and W.B. Mills. 1977. Water quality assessment:
A screening method for nondesignated 208 areas. EPA-600/9-77-O23. U.S.
Environmental Protection Agency. August, 1977.
225

-------
7. LIST OF PREPARERS
Thomas C. Ginn, Ph.D . Dr. Ginn served as project manager and was respon-
sible for part of the biological impacts analyses. Dr. Ginn is an aquatic
ecologist specializing in the effects of environmental pollutants on the
distribution and physiology of freshwater and marine organisms.
Karen V. Summers, M.S . Ms. Summers served as co-project manager and con-
ducted most of the surface water quantity evaluations, water use analyses
and part of the water quality impacts. She is a geohydrologist specializ-
ing in modeling studies of ground water and rivers and design of field
monitoring programs.
Stanley W. Zison, Ph.D . Dr. Zison conducted the analyses of baseline
water quality. Dr. Zison is an environmental chemist and statistician
specializing in waste treatment management and mathematical modeling of
the fate of pollutants.
William B. Moreland, M.A . Mr. Moreland was responsible for the air qual-
ity modeling and prediction of air quality impacts. He is a meteorolog-
ical specialist in diffusion analyses, cliriatological analyses and effects
of atmospheric turbulence.
Thomas Grieb, M.A . Mr. Grieb conducted terrestrial biological effects
analyses for both air quality and water quantity considerations. Mr.
Grieb specializes in the impacts of energy-related operations on biologi-
cal systems.
J. David Dean, M.S . Mr. Dean conducted parts of the water quality impacts
analyses, especially as they related to agricultural use of water and
determined the impacts of boron and salinity on crops. He is an agricul-
tural engineer with considerable experience in the use of stochastic!
deterministic models for analyzing water demand by irrigated crops.
John W. Ryan, MBA . Mr. Ryan was responsible for partial development of
the socioeconomic methodology and impact assessment. He is a specialist
in the modeling of complex socioeconomic systems in the public and private
sectors.
Daniel T. Dick, Ph.D . Dr. Dick also contributed to the socioeconomic
methodology. He is an environmental economist with considerable experi-
ence in regional impact analysis of energy development.
Mary E. Gray, B.A . Ms. Gray collected and described information concern-
ing the baseline socioeconomic conditions. She has experience in resource
planning, impact assessment and energy development analysis.
226

-------
Susan Mara, M.A . Ms. Mara is a hydrologist and served as a consultant
on water-related matters in the socioeconomic analyses.
Leslie A. Young, M.S . Ms. Young analyzed the direct socioeconomic
impacts. She specializes in economic analyses and the development of
interview and survey techniques.
227

-------