Hydrology Study, Bow and Elbow River Updated Hydraulic Model Project, Rev
March 2010
HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT, REV. A Submitted to: Alberta Environment REPORT Report Number: 09-1326-1040
HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT
Executive Summary
Golder Associates Ltd. (Golder) was commissioned by Alberta Environment (AENV) to conduct a hydrology study for the “Bow and Elbow River Updated Hydraulic Model” project. The City of Calgary (the City), in partnership with Alberta Environment (AENV), plans to create a HEC-RAS (Hydraulic Engineering Center River Analysis System) hydraulic model of the Bow and Elbow Rivers through the City. The model implementation is primarily for supporting emergency response planning and operations through flood inundation mapping. It will provide additional perspective on current flood hazard area management and will provide a basis for increased understanding of fish habitat, river morphology and erosion, water quality and storm water runoff impacts.
The scope of work for the hydrology study included:
1) Generation of naturalized daily flow series at the major storage facilities on the Bow River above Bearspaw Dam and on the Elbow River at Glenmore Reservoir;
2) Estimation of naturalized and regulated 1:2 to 1:1,000 year flood flows based on flood frequency analysis of naturalized and/or recorded peak flow series at relevant locations along the Bow River and its tributaries (including the Elbow River);
3) Development of synthetic inflow flood hydrographs for tributaries to the Bow River with hydropower developments and at the Water Survey of Canada (WSC) Bow River at Banff station;
4) Routing of the synthetic flood hydrographs through all major storage reservoirs along the Bow River above Bearspaw Dam and through the Glenmore Reservoir;
5) Commentary on the effects of climate change and the impact of stormwater runoff on flood estimates as well as on the seasonality of flood peaks; and
6) Comparison of the new flood flow estimates with those of AENV’s 1983 Calgary Floodplain Study (the 1983 study). Overview of the Bow and Elbow River Basins The Bow and Elbow Rivers are the primary water courses within the City. The river flows originate in the snowpack and glaciers of the Rocky Mountains and traverse the foothills before reaching the City on the edge of the Prairies. The upstream reach of the Bow River and its tributaries are controlled by several hydropower structures. Within the City, the Bow River is joined by several streams (Elbow River, Nose Creek, Fish Creek and Pine Creek). A short distance downstream of the Nose Creek confluence with the Bow River, water is diverted from the Bow to the Western Irrigation District (WID) canal.
According to historical records, major floods occurred on the Bow River in 1879, 1897, 1902, 1929, 1932, 1995 and 2005. Records also indicate that major floods occurred on the Elbow River in 1915, 1923, 1929, 1995 and 2005. Major floods are commonly associated with high rainfall or rain-on-snow events in the May to July period. The study area includes the Bow River and its tributaries within the city upstream of its confluence with the Highwood River. The drainage area of the Bow River at Calgary (WSC Station 05BH004) is 7,868 km2. The drainage area of the Elbow River below Glenmore Dam (WSC Station 05BJ001) is 1,236 km2.
March 2010 Report No. 09-1326-1040 ii
HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT
Hydrologic Analysis of Flood Peak Flows The hydrology of the Bow River basin was examined in detail by reviewing past work (e.g., the 1983 study), utilizing the latest methods of analysis and considering changes in flows due to flow regulation by hydropower infrastructure, water diversion and land use changes.
Naturalized daily flow series were developed at all major storage facilities on the upper Bow River, including inflows to Bearspaw Reservoir, Ghost Lake, Lake Minnewanka, Spray Lakes, Barrier Lake, as well as Lower and Upper Kananaskis Lakes and Glenmore Reservoir on the Elbow River. The flow naturalization used the same project depletion method and the Streamflow Synthesis and Reservoir Regulation (SSARR) channel routing procedures used by AENV. Naturalization of daily flows was facilitated by the Natural Flow Computation Program (NFCP) developed for the Prairie Provinces Water Board (Optimal Solutions, 2009).
Frequency analyses were completed based on flood flow series derived from naturalized daily flows for the Bow River and its tributaries above Bearspaw Dam and for the Elbow River above Glenmore Dam, based on the annual maximum daily flow series. Synthetic inflow flood hydrographs (shaped according to recorded flood events) at major storage facilities on the Bow River and its tributaries including the Elbow River at Glenmore Reservoir were developed for each flood flow obtained from frequency analyses of the naturalized flow series.
The design flood hydrographs created for all return periods were routed through the storage facilities according to the existing flood control and operation guidelines and selected starting water levels. The arrival times of peak flows from each reservoir and from the natural contributing areas were based on travel times provided by AENV and estimated flood travel times developed by TransAlta as part of its Flood Action Plan. The peak daily flows, after routing, were transformed into peak instantaneous discharges based on relationships established between recorded annual maximum daily and annual maximum instantaneous flows.
Historic flood peaks that pre-date systematic flow monitoring were available for the Bow River at Calgary for 1879, 1897 and 1902. The flow estimates for these floods were originally determined as part of the 1983 study, based on recorded high water marks, newspaper reports and information from local residents, and were adopted without change for the current study. A technique respecting the probability distribution of the systematic record was applied to in-fill missing annual peak flows in the historic period of the complete record. The historic floods were also considered in the estimation of flood flows on the Bow River at Bearspaw Reservoir.
The flood flows for the lower reaches of the Bow River (the reach between its confluence with the Elbow River and the Highwood River) were derived using two approaches. In the first approach, the peak flow of a specific frequency for a specific tributary was added to the peak flow of the same frequency in the reach of the Bow River just upstream of the tributary to estimate the peak flow of the same frequency downstream of the confluence. The second approach was based on the methodology used in the 1983 study, and involved developing maximum instantaneous discharges for the Bow River downstream of its confluence with the Elbow River.
The potential effect of climate change on flood flow estimates was analyzed based on recorded data and predictions from climate models. The result of this analysis showed that the recent patterns in the timing of these floods are similar to what were observed at the beginning of the century. Hence, there is no clear evidence that the patterns in magnitude or timing of flood peaks have changed significantly over the past 100 years.
The naturalized peak flow series in the Bow River below Bearspaw Dam and the Elbow River below Glenmore Dam were analyzed to identify whether one flood mechanism (snowmelt or rainfall) dominates flooding. Based
March 2010 Report No. 09-1326-1040 iii
HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT
on recorded data over the past 100 years, about 80% of the annual peak flows in the Bow River and the Elbow River occurred at the end of May and in June. This suggests that snowmelt or rain-on-snow events dominate generation of peak flows in the Bow River and Elbow River watersheds. However, it should be noted that the most significant historic floods are associated with high rainfall or rain-on-snow events.
An analysis of the effect of stormwater runoff on flood flow estimates was conducted using the 2005 flood, the largest flood recorded in the Bow River and Elbow River watersheds since 1932. The result of this analysis indicates that the stormwater runoff contribution to peak flows in the Bow River was relatively small, and that flood flow estimates therefore do not require adjustment to reflect stormwater runoff. Conclusions The following main conclusions are supported by the results of this hydrology study: The review of the available hydrometric data indicated that data were missing at several locations on the Bow River and its tributaries for a number of years. The missing data may have a minor affect on the level of uncertainty in flow statistics derived from the naturalized flow series; Results of the statistical tests conducted for various locations on the Bow River and its tributaries indicate that most of the recorded flood flow series are independent, random and homogeneous and have no trend. The flood flow series recorded for the Bow River above Bearspaw Dam displayed some trend and non- homogeneity, primarily due to a combination of long-term variability in the climate regime and the construction of major facilities on the Bow River and its tributaries. However, the entire recorded flood flow series was considered appropriate for frequency analysis; The inclusion of historic floods (data that pre-date systematic flow monitoring) in the flood frequency analysis increases the magnitudes of peak flows for floods with low probability of occurrence (i.e., floods with long return periods), compared to the estimates obtained based on the systematically-recorded flood flow series only; The flood flow moderating capacity of the major storage facilities on the Bow River is relatively small and there is minor flood regulation capacity by Glenmore Reservoir on the Elbow River; and The flood flows estimated in this study for the Bow River and its tributaries through the City are smaller than those reported in the 1983 study, mainly due to inclusion of additional flood flow data from 1981 to 2008 in the frequency analyses. Recommendations Table i summarizes the recommended regulated instantaneous flood peak discharges for various return periods for the Bow River and its tributaries through the City of Calgary for use in the HEC-RAS hydraulic model. These recommendations are made based on the results of this study, and incorporate the effects of regulation by the Glenmore Reservoir. Corresponding naturalized discharge estimates for the same return periods, which do not include the effects of regulation at the Glenmore Reservoir, are available in the main report.
March 2010 Report No. 09-1326-1040 iv
HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT
Table i: Recommended Instantaneous Flood Peak Discharges (m3/s) for the Bow River and its Tributaries Bow Elbow Elbow Bow Bow Bow Bow Nose Fish Pine Bow River Return River River River River River River River Highwood Creek at Creek at Creek at below Period above above below below below below below River at Bow Bow Bow Highwood (Years) Elbow Glenmore Glenmore Elbow Nose Fish Pine Bow River River River River River River Dam Dam River Creek Creek Creek 2 423 64.4 51.5 475 6 481 39 520 6.92 526 191 717 5 606 146 99.4 705 14 719 86 805 16.3 822 396 1220 10 774 238 193 967 23 990 134 1,120 24.7 1,150 583 1730 20 983 353 274 1,260 35 1,290 198 1,490 34.3 1,520 804 2,330 50 1,350 553 445 1,790 60 1,850 317 2,170 49.3 2,220 1,160 3,380 100 1,710 737 699 2,410 89 2,500 444 2,940 62.7 3,000 1,480 4,480 200 2,170 929 922 3,090 130 3,220 618 3,840 78.2 3,920 1,850 5,760 500 2,980 1,240 1,220 4,200 214 4,420 946 5,360 102 5,460 2,420 7,880 1,000 3,810 1,510 1,490 5,290 310 5,600 1,300 6,900 123 7,020 2,930 9,950
March 2010 Report No. 09-1326-1040 v
HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT
Table of Contents
1.0 INTRODUCTION ...... 1
1.1 Background ...... 1
1.2 Scope of Work ...... 2
1.3 Study Area ...... 2
2.0 HYDROLOGIC ANALYSIS OF FLOOD PEAKS ...... 5
2.1 Generation of Naturalized Daily Flow Series ...... 5
2.2 Frequency Analysis of Naturalized Peak Flow Series ...... 7
2.2.1 Statistical Tests on Peak Flow Series ...... 7
2.2.2 Frequency Analysis of Flood Series ...... 10
2.2.3 Historic Floods ...... 10
2.3 Inflow Flood Hydrographs ...... 15
2.4 Flood Hydrograph Routing and Regulated Peak Flow Frequency Estimates ...... 16
2.5 Peak Flow Estimates for Major Tributaries ...... 18
2.5.1 Nose Creek at Calgary ...... 18
2.5.2 Fish Creek at the Mouth ...... 19
2.5.3 Pine Creek at the Mouth ...... 20
2.5.4 Highwood River near the Mouth ...... 20
2.6 Peak Flow Estimates for the Lower Reach of Bow River ...... 21
3.0 DISCUSSION ...... 23
3.1 Commentary on Effects of Climate Change on Flood Estimates ...... 23
3.2 Commentary on the Seasonality of Flood Peaks ...... 26
3.3 Magnitude and Timing of Stormwater Runoff Inputs and Diversions ...... 27
4.0 COMPARISON OF FLOOD FLOW ESTIMATES ...... 30
5.0 CONCLUSIONS AND RECOMMENDATIONS ...... 33
5.1 Conclusions ...... 33
5.2 Recommendations ...... 33
6.0 REFERENCES ...... 36
7.0 LIMITATION OF LIABILITY ...... 36
March 2010 Report No. 09-1326-1040 vi
HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT
TABLES Table 1: Summary of Existing Hydrometric Data ...... 4 Table 2: Summary of Hydrometric Data Review at Major Storage Facilities on the Bow River and its Tributaries ...... 6 Table 3: Results of Statistical Tests of Annual Maximum Daily Flows ...... 8 Table 4: Estimated Daily Flood Flows (m3/s) for Various Return Periods, Derived from Naturalized Annual Maximum Daily Flows ...... 11 Table 5: Estimated Daily Flood Flows (m3/s) for the Bow River above Bearspaw Dam, Derived from Naturalized Annual Maximum Daily Flows and Historic Data ...... 14 Table 6: Comparison of Flood Inflows and Outflows for the Bow River at Bearspaw Reservoir, Derived from Naturalized Flows (m3/s) ...... 17 Table 7: Comparison of Inflow and Outflow Flood Flows for the Elbow River below Glenmore Dam, Derived from Naturalized Flows (m3/s) ...... 18 Table 8: Instantaneous Flood Peak Discharges (m3/s) for Pine Creek, Provided by AMEC (2007) ...... 20 Table 9: Instantaneous Flood Peak Discharges (m3/s) for Bow River Tributary Streams ...... 21 Table 10: Instantaneous Flood Flows along the Bow River, Assuming Coincident Events for all Rivers and Streams, Regulated Scenario ...... 21 Table 11: Instantaneous Flood Flows along the Bow River, Based on Method Used in the 1983 Study ...... 22 Table 12: Time of Occurrences of Annual Maximum Daily Flood Events in the Bow and Elbow Rivers ...... 24 Table 13: Flood Magnitudes Derived Using Seasonal Flood Series ...... 26 Table 14: Comparison of Instantaneous Flood Flows (m3/s) Derived for the Bow River Upstream of its Confluence with the Elbow River ...... 30 Table 15: Comparison of Instantaneous Flood Flows (m3/s) Derived for the Elbow River below Glenmore Dam(d) ...... 32 Table 16: Instantaneous Flood Flows along the Bow River at Various Locations, Based on Method Used in the 1983 Study ...... 32 Table 17: Recommended Instantaneous Flood Peak Discharges (m3/s) for the Bow River and its Tributaries, Regulated Scenario ...... 34 Table 18: Recommended Instantaneous Flood Peak Discharges (m3/s) for the Bow River and its Tributaries, Naturalized Scenario ...... 34
FIGURES Figure 1: Bow River Basin above Highwood River ...... 3 Figure 2: Flood Frequency Curve for the Bow River above Bearspaw Dam ...... 13 Figure 3: Flood Series for the Bow River at Calgary (WSC Station 05BH004) and the Elbow River at Bragg Creek (WSC Station 05BJ004) ...... 24 Figure 4: Timing of Recorded Floods for the Bow River at Calgary and the Elbow River at Bragg Creek...... 25 Figure 5: Sub-Catchments Contributing Stormwater Runoff to the Bow River and its Tributaries (Source: City of Calgary) ...... 27 Figure 6: Diversions from the Bow River by the Western Irrigation District (WID) ...... 29
APPENDICES APPENDIX A Naturalized Daily Flows and Flood Routing at Key Locations APPENDIX B Results of Flood Flow Analyses
March 2010 Report No. 09-1326-1040 vii
HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT
1.0 INTRODUCTION The City of Calgary (the City), in partnership with Alberta Environment (AENV), plans to create a new HEC-RAS (Hydraulic Engineering Center River Analysis System) hydraulic model of the Bow and Elbow rivers through the City. The model implementation is planned primarily to support emergency response planning and operations through flood inundation mapping. It will provide additional perspective on current flood hazard area management and will provide a base for increased understanding of fish habitat, river morphology and erosion, water quality and storm water runoff impacts.
In November 2009, the City and AENV engaged Golder Associates Ltd. (Golder) to undertake the “Bow and Elbow River Updated Hydraulic Model” project. The first component of the project, a Hydrology Study, was contracted through AENV and is the subject of this report. The hydrology study is required to determine the flood magnitudes for Bow River and its tributaries through the City for use in the HEC-RAS hydraulic model. 1.1 Background The Bow and Elbow Rivers are the primary watercourses within the City. The most recent hydraulic models of the Bow and Elbow Rivers were prepared by AENV as part of the 1983 Calgary Floodplain Study (the 1983 study), with minor updates in 1996 (AENV, 1983 and 1996). The Bow River hydraulic model reach extends from the Bearspaw Dam downstream to the confluence with the Highwood River. The Elbow River hydraulic model reach extends from the City limit at 101 Street SW (located upstream of Glenmore Reservoir) downstream to the Bow River confluence.
The Bow and Elbow Rivers originate in the snowpack and glaciers of the Rocky Mountains and traverse the foothills before reaching the City on the edge of the Prairies. The rivers flow through a mix of Alpine, Subalpine, Boreal Foothill and Aspen Parkland eco-regions. Land use ranges from urban in the City, to agricultural in parts of the foothills, and to forest in the remainder of the foothills and in the mountains.
The main weather systems in southern Alberta are cyclonic and travel from the Pacific Ocean over the cordillera of British Columbia before reaching Alberta. The cordillera is effective at removing moisture from the Pacific air mass leaving it relatively dry by the time it reaches the leeward side of the mountain ranges. In addition, Alberta receives cold, dry air masses from the Arctic and warm, moist air masses from the Gulf of Mexico. In southern Alberta, these air masses collide, usually between May and July, creating heavy rainfall events.
Major floods occurred on the Bow River in 1879, 1897, 1902, 1929, 1932, 1995 and 2005. Records also indicate that major floods occurred on the Elbow River in 1915, 1923, 1929, 1995 and 2005. Major floods are commonly associated with high rainfall or rain-on-snow events in the May to July period. Prior to construction of the Bearspaw Dam in 1953 and subsequent winter flow regulation changes, the Bow River was also subject to ice jam flood events. Although ice jam floods still occur, they are not as large as the open water floods and not the primary concern in flood risk assessment for the Bow and Elbow Rivers in the City.
March 2010 Report No. 09-1326-1040 1
HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT
1.2 Scope of Work The scope of work for the hydrology study includes the following: Generation of a naturalized daily flow series at the major storage facilities along the Bow River above Bearspaw Dam and for the Elbow River at Glenmore Reservoir; Estimation of naturalized and regulated 1:2 to 1:1,000 year flood flows, based on flood frequency analysis of naturalized and/or recorded peak flow series at relevant locations along the Bow River and its tributaries, including the Elbow River; Development of synthetic inflow flood hydrographs for rivers with hydropower developments that are tributaries to the Bow River and at the Water Survey of Canada (WSC) station Bow River at Banff; Routing of the synthetic flood hydrographs through all major storage facilities along the Bow River above Bearspaw Dam and for the Elbow River at Glenmore Reservoir; Commentary on the effects of climate change and the impact of stormwater runoff on flood estimates as well as on the seasonality of flood peaks; and Comparison of the new flood flow estimates with those of the 1983 study. 1.3 Study Area The study area includes the Bow River and its tributaries upstream of its confluence with the Highwood River.
The reported drainage area of the Bow River at Calgary (WSC Station 05BH004) is 7,868 km2. The upstream reach of the Bow River and its tributaries are controlled by several hydropower structures. Within the City, the Bow River is joined by several streams (Elbow River, Nose Creek, Fish Creek and Pine Creek). A short distance downstream of the Nose Creek confluence with the Bow River, water is diverted from the Bow River to the Western Irrigation District (WID) canal.
The Elbow River is a contributor to flood flows in the City. The Elbow River flows into Glenmore Reservoir. The reported drainage area of Elbow River below Glenmore Dam (WSC Station 05BJ001) is 1,236 km2.
The setting of the Bow River and its tributaries for this study is shown in Figure 1. Table 1 provides a summary of the basic hydrometric information available from WSC for the Bow River and its tributaries within the study area.
March 2010 Report No. 09-1326-1040 2
400000 450000 500000 550000 600000 LEGEND *# HYDROMETRIC STATION
Red Deer River STORAGE STRUCTURE ROAD MAJOR RIVERS
BOW RIVER BASIN ABOVE HIGHWOOD 5 ³ CITY 4 LAKE PRAIRIE FARM REGIONAL ADMINISTRATION BASIN BOUNDARY 5700000 5700000
Nose Creek 2B
*# 05BG002 05BG001 05BD003 *# AIRDRIE 05BD002 *# 05BE005 *#*# 05BD003 05BE006 *# # Lake Minnewanka * 05BH005 Bow River Ghost Lake *# 05BB001 *# *# 05BH005 CALGARY*# 05BH906 *# 05BC001 UV1A 1 05BH003 05BE004 *# UV Bearspaw Reservoir *# *# 05BH008 *# 05BH003 05BC008 05BF001 Elbow River *# 05BF025 05BH004 *# *#*# Barrier Lake UV8 Glenmore Reservoir *# 05BJ001 5650000 *# 5650000 Spray Lake *# 05BC006 05BJ010 *# 05BJ008 Fish Creek 05BJ005 *# 05BK002 05BC007 Bow River Basin *# 22x 05BK003 *#*# 05BC003 UV *# 05BK001 05BC002 Pine Creek UV2 *# 05BL024
05BF008 05BF003 *# 05BF009 *#*# Highwood River Lower Kananaskis Lake *# 05BF010 3 Upper Kananaskis Lake *# 05BF011 *# 05BF005 05BF002 05BF021
5600000 5600000 REFERENCE Hydrography and city data obtained from Natural Resources Canada. Hydrologic regions, basin and sub-basin data obtained from Alberta Environment. ALBERTA Projection: Alberta 10TM False Easting 500,000 at 115 ° W. Datum: NAD 83 20 0 20
SCALE 1:750,000 KILOMETRES
PROJECT Edmonton ALBERTA ENVIRONMENT BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL
TITLE BOW RIVER BASIN ABOVE HIGHWOOD RIVER Calgary
PROJECT No. 09-1326-1040 SCALE AS SHOWN REV. 0 DESIGN BD 10 Feb. 2010 GIS LY 03 Mar. 2010
5550000 5550000 CHECK GB 27 May 2010 Calgary, Alberta FIGURE: 1 400000 450000 500000 550000 600000 REVIEW BD 27 May 2010 I:\2009\09-1326\09-1326-1040\Mapping\MXD\RSA\Hydrology\BowRiverBasin_91.mxd HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT
Table 1: Summary of Existing Hydrometric Data
Length Type of Condition of WSC Drainage Period of of Recorded Recorded Station Station Name Area Record Record Hydrometric Hydrometric ID (km2) (Years) Data Data
05BB001 Bow River at Banff 2,210 1909-2008 100 Flow Natural
05BE004 Bow River near Seebe 5,170 1923-2008 86 Flow Regulated
05BE006 Bow River below Ghost Dam 6,550 1933-1989 57 Flow Regulated
05BH005 Bow River near Cochrane 7,412 1916-2008 93 Flow Regulated
05BE005 Ghost Lake near Cochrane 6,480 1929-2008 80 Level Regulated
05BH008 Bow River below Bearspaw Dam 7,770 1983-2008 26 Flow Regulated
05BH004 Bow River at Calgary 7,868 1911-2008 98 Flow Regulated
05BD003 Lake Minnewanka near Banff 647 1916-2008 101 Level Regulated
05BC006 Spray Reservoir at Three Sisters Dam n/a 1949-2008 60 Level Regulated
05BC002 Spray River near Spray Lakes 360 1915-1939 25 Flow Natural
05BC001 Spray River at Banff 751 1910-2008 99 Flow Regulated since 1951 05BF005 Upper Kananaskis Lake 151 1932-2008 77 Level Regulated
05BF009 Lower Kananaskis Lake 359 1932-2008 77 Level Regulated
05BF025 Kananaskis River below Barrier Dam 899 1975-2008 34 Flow Regulated
05BF001 Kananaskis River near Seebe 933 1911-1962 52 Flow Regulated since 1932 05BJ010 Elbow River at Sarcee Bridge 1,190 1979-2008 30 Flow Natural
05BJ005 Elbow River above Glenmore Dam 1,220 1933-1977 45 Flow Natural
05BJ008 Glenmore Reservoir at Calgary 1,224 1976-2008 33 Level Regulated
05BJ001 Elbow River below Glenmore Dam 1,236 1908-2008 101 Flow Regulated since 1932 05BH003 Nose Creek at Calgary 893 1911-1986 76 Flow Natural
05BK001 Fish Creek near Priddis 261 1908-2008 101 Flow Natural
05BL024 Highwood River near the Mouth 3,952 1970-2008 39 Flow Regulated n/a – Not Applicable
March 2010 Report No. 09-1326-1040 4
HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT
2.0 HYDROLOGIC ANALYSIS OF FLOOD PEAKS The hydrology of the entire Bow River basin was examined in detail by reviewing past work (e.g., the 1983 study), utilizing the latest available methods of analysis and taking into consideration changes in flows due to flow regulation by hydropower infrastructure, water diversion and land use changes.
The updated hydrologic analyses of the Bow River and its tributaries are required to estimate flood flows for input into the HEC-RAS hydraulic model of the Bow and Elbow River study reaches. Naturalized and regulated estimates are required for the 1:2, 1:5, 1:10, 1:20, 1:50, 1:100, 1:200, 1:500 and 1:1,000 year return periods, corresponding to the 50, 20, 10, 5, 2, 1, 0.5, 0.2 and 0.1 percent probabilities of annual exceedance. Flood frequency discharge estimates are required for the Elbow River above and below Glenmore Dam, as well as for the Bow River above and below Bearspaw Dam, above and below the Elbow River confluences and at confluences with major tributaries, including Nose Creek, Fish Creek, Pine Creek and Highwood River.
The hydrology of the Bow and Elbow Rivers upstream of Calgary is influenced by the operation of several storage reservoirs on the Bow River (Bearspaw Reservoir, Ghost Lake, Lake Minnewanka, Spray Lakes, Barrier Lake, Lower and Upper Kananaskis Lakes) and one on the Elbow River (Glenmore Reservoir). The tasks performed for estimating the flood discharges are outlined in Section 1.2, and a detailed methodology for each task is provided in the sections that follow. 2.1 Generation of Naturalized Daily Flow Series Naturalized daily flow series were developed at all major storage facilities on the upper Bow River, including inflows to Bearspaw Reservoir, Ghost Lake, Lake Minnewanka, Spray Lakes, Barrier Lake, as well as to Lower and Upper Kananaskis Lakes and Glenmore Reservoir on the Elbow River. The flow naturalization used the same project depletion method and Streamflow Synthesis and Reservoir Regulation (SSARR) channel routing procedures used by AENV. The approach is explained in more detail in Appendix A. Naturalization of daily flows was facilitated by the Natural Flow Computation Program (NFCP) developed for the Prairie Provinces Water Board (Optimal Solutions, 2009).
A detailed review was conducted of available hydrometric data at major storage facilities. It was originally assumed that the complete hydrometric data (reservoir levels and outflows) required for naturalization of flows at major storage facilities would be available from AENV to develop a long time series of about 100 years (from 1911 to 2008) of naturalized and/or recorded daily flows that could be used to generate flood flows. Once the study was initiated, it became apparent that data for a number of years were missing at several locations. The information available from AENV would only allow calculation of naturalized flows for the following periods: Lake Minnewanka (66 years), Spray Lakes (59 years), Barrier Lake (55 years) and Upper and Lower Kananaskis Lakes (24 and 34 years respectively) (Table 2). Golder consulted AENV and the City on the limitations of the readily available data. Golder was directed to proceed with the work using the available data to produce natural flow series and to conduct frequency analysis on these data sets.
One consequence of the limitation in the data is that there is some loss in the reliability of the study results. AENV may wish to consider addressing this data gap after this study is completed.
March 2010 Report No. 09-1326-1040 5
HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT
Table 2: Summary of Hydrometric Data Review at Major Storage Facilities on the Bow River and its Tributaries
Available Natural Flow from WSC Total Number of Years Regulated Flow and Storage Level Data Useable Stations at Site Prior to Construction of of Data Available for for Analysis Location Storage Facilities Naturalization of Flows
Useable Period of Station ID Station ID Period of Record Record 05BC006 (Levels) 1970-2008 Spray Lakes 05BC002 1915-1939 59 05BE007 (Outflows) 1975-2008
Upper Kananaskis 05BF005 (Levels) 1974-2008 - - 24 (Interlakes) TransAlta Data (Outflows) 1985-2008
Lower Kananaskis 05BF0009 (Levels) 1973-2008 - - 1932-1955, 34 (Pocaterra) 05BF003 (Outflows) 1975-2008 05BF024 (Levels) 1970-2008 Barrier Lake 05BF025 1912-1933 55 05BF025 (Outflows) 1975-2008
05BD003 (Levels) 1943-2008 Lake Minnewanka (Cascade) - - 05BD002 (Outflows) 1917-1941 92 05BD004 (Outflows) 1943-2008
05BE004 1923-2008 05BE005 (Levels) 1929-2008 Ghost Reservoir 79 05BG001 1911-1983 05BE006 (Outflows) 1933-1989
AENV Data (Levels) 1955-2008
Bearspaw Reservoir - - 05BH004 (Outflows) 1911-2008 98
06BH008 (Outflows) 1983-2008
- - 05BJ008 (Levels) 1976-2008
Glenmore Reservoir - - AENV Data (Levels) 1933-1988 101
- - 05BJ001 (Outflows) 1908-2008
March 2010 Report No. 09-1326-1040 6
HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT
2.2 Frequency Analysis of Naturalized Peak Flow Series The purpose of the frequency analysis of naturalized peak flow series is to estimate 1:2 to 1:1,000 year flood flows using the naturalized maximum daily flow series. Frequency analyses were completed for the Bow River above Bearspaw Dam and the Elbow River above Glenmore Dam based on the annual maximum daily flow series. Synthetic inflow hydrographs corresponding to floods of various return periods were then developed and routed through the storage reservoirs to simulate the effects of regulation. Peak daily flows were transformed into peak instantaneous discharge based on relationships established between recorded annual maximum daily and annual maximum instantaneous flows. For the Bow River below Bearspaw Dam, this relationship was established from recorded annual maximum daily and instantaneous flows at WSC Station 05BH004 (Bow River at Calgary). For Elbow River below Glenmore Dam, the relationship was established based on recorded maximum daily and instantaneous flows at WSC Station 05BJ001 and on historic large floods estimated for the Elbow River by T. Blench & Associates (1965).
The flood flows for the reach of the Bow River between Bearspaw Reservoir and its confluence with the Highwood River were determined using the two approaches described in Section 2.6. 2.2.1 Statistical Tests on Peak Flow Series The Consolidated Frequency Analyses (CFA) computer program from Environment Canada was used to determine if the flows in each peak flow series (Bow River and its tributaries) can be considered to be independent (not serially correlated), without trend, random and homogeneous. The results of statistical tests performed on the annual maximum daily flows at various locations on the Bow River and its tributaries are provided in Table 3.
The results of statistical tests indicate that: The majority of the recorded and/or naturalized flood peaks for the Bow River and its tributaries are independent, random, homogeneous and do not display any significant trends; Inflow floods to Lake Minnewanka display non-randomness at the 1% and 5% levels of significance; Local flood inflows to Spray River at Banff display non-homogeneity at the 5% level of significance but appear homogeneous at the 1% level of significance; Local flood inflows to Spray River at Banff also display non-randomness at the 1% and 5% levels of significance; Flood inflows to Barrier Lake display serial dependence, non-randomness and non-homogeneous at the 5% level of significance. At the 1% level of significance the flood inflows appear random and homogeneous, but still display serial dependence; Flood flows for Bow River at Banff display a trend at the 5% level of significance, but shows no trend at the 1% level of significance; Flood inflows to Bearspaw Reservoir display a trend and non-homogeneity at the 1% and 5% levels of significance; and Flood inflows to Glenmore Reservoir displays non-homogeneity at the 5% level of significance, but shows homogeneity at the 1% level of significance.
March 2010 Report No. 09-1326-1040 7
HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT
Table 3: Results of Statistical Tests of Annual Maximum Daily Flows Inflow to Local Inflow to Inflow Bearspaw Local Inflow Inflow to Inflow to Inflow to Bow Inflow to Inflow to Station Upper to Reservoir, into Lower Lake Spray Spray River at Ghost Glenmore Name Kananaskis Barrier Without Kananaskis Minnewanka Lakes River at Banff Lake Reservoir Lake Lake Historic Banff Data Serial Correlation Coefficient Test for Independence
S1 -0.229 0.035 0.165 0.149 0.190 0.275 -0.056 -0.005 0.158 0.100 t -1.312 0.191 1.579 1.458 1.459 2.587 -0.550 -0.047 1.556 0.996 t( =0.05) -1.696 1.697 1.662 1.661 1.672 1.664(a) -1.661 -1.665 1.661 1.661 t( =0.01) -2.453 2.457 2.369 2.367 2.394 2.373(a) -2.365 -2.376 2.366 2.365 Spearman Rank Order Correlation Coefficient Test for Trend
rs 0.025 0.154 0.140 0.066 0.246 0.192 0.245 0.139 0.284 0.072 t 0.139 0.870 1.339 0.640 1.930 1.787 2.505 1.234 2.904 0.713 t( =0.05) 2.037 2.040 1.987 1.985 2.002 1.989 1.984(a) 1.991 1.985(a) 1.984 t( =0.01) 2.738 2.744 2.632 2.629 2.663 2.636 2.627 2.641 2.628(a) 2.626 Mann-Whitney Split Sample Test for Homogeneity Size of Earlier Sample 20 20 46 49 30 45 50 20 23 25 z -0.315 -1.253 -0.508 -0.148 -2.004 -2.148 -1.479 -0.316 -3.558 -1.826 z( =0.05) -1.645 -1.645 -1.645 -1.645 -1.645(a) -1.645(a) -1.645 -1.645 -1.645(a) -1.645(a) z( =0.01) -2.326 -2.326 -2.326 -2.326 -2.326 -2.326 -2.326 -2.326 -2.326(a) -2.326 Test of General Randomness (Runs for Above or Below the Median) Median 18 23 51 66 12 71 208 346 383 53 N1(for Q>=Median) 17 17 46 49 30 43 50 40.0 49 51 N2(for Q March 2010 Report No. 09-1326-1040 8 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT Observed trends for inflows to Bearspaw Reservoir are mainly the result of non-homogeneities in the data series and do not appear to be due to some other large-scale climatic changes. Based on recorded data, 7 of the flood flows recorded prior to 1934 (including historic floods), are significantly higher than any of the flood flows recorded after 1934. Statistical tests were performed on flood inflows to Bearspaw Reservoir using recorded data from 1933 to 2008. The test results indicate that the data do not display significant serial dependence or trend and appear random and homogeneous. In addition, statistical tests of flood inflows to Ghost Lake (1930- 2008), which is located a few kilometres upstream of Bearspaw Reservoir and flood inflows to Glenmore Reservoir, which is the largest tributary to the Bow River, do not display significant serial dependence or trend and appear random and homogeneous. Non-homogeneities in data are caused when artificial changes affect the statistical properties of the observations through time. In reality, obtaining perfectly homogeneous data is almost impossible, as unavoidable changes such as flow regulation, forestry operations and river diversions in the watershed will often affect the data. Forestry operations, river diversions and reservoir regulation have all increased since the start of systematic streamflow monitoring in 1908 (AENV, 1983). There is evidence that extensive forest fires occurred during the late 1800s and early 1900s in the vicinity of the Bow River drainage basin. In his study of the landscape change in Banff National Park, Byrne (1968) expressed the opinion that climatic conditions during the latter half of the nineteenth century were drier and, hence, conducive to forest fire. Apparent depletion of forest cover might have resulted in increased peak runoff rates in the early period of flood record for the Bow River. Conversely, the recovery of the forest cover might have resulted in reduction of peak runoff rates in the later period of record. This may partially account for non-homogeneity of the flood record. However, forest cover may be depleted or replenished at any time by man or nature and should not be considered as a permanent change in the flood pattern. Thus the pre-1934 records and the post-1933 records should be combined regardless of the influence of forest cover. Consumptive water diversions on the Bow and Elbow Rivers upstream of the City are limited to minor licensed water supply systems for municipal, industrial and domestic uses. These diversion quantities are insignificant compared to peak flood flows in the Bow and Elbow Rivers. There are two non-consumptive diversion schemes on the Bow River upstream of the City: the Smith-Dorrien Creek diversion and the Ghost River diversion. The Smith-Dorrien Creek diversion diverts water from a drainage area of about 35 km2 into the Spray Lakes reservoir system through the Mud Lake Diversion canal. This area was naturally draining into the Kananaskis River. Based on recorded flows, the diversion through the Smith-Dorrien Creek is less than 0.5% of record peak flows at the City and will have insignificant effect on estimation of flood flows. The Ghost River diversion scheme diverts water from a drainage area of about 210 km2, located in the headwaters of the Ghost River, into Lake Minnewanka. This diversion represents less than 3% of Ghost River flood runoff flows and hence has negligible flood moderating influences on major floods recorded at the City. A system of 6 reservoirs and 11 hydropower plants was established in the Bow River valley upstream of the City between 1911 and 1955. Detailed analyses of the storage effects of each system are presented in Appendix A. Flood control potential from these reservoirs are dependent on available storage. A routing model of the Bow River from Banff to Bearspaw Reservoir was established to evaluate flood reduction potential from this reservoir. The regulated flows were naturalized by removing the effect of reservoir storage based on recorded reservoir levels and outflows. The result of flow naturalization showed that flow regulations have a minor effect on peak March 2010 Report No. 09-1326-1040 9 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT flood flows (a reduction of less than 2%). Hence, storage reservoirs on the Bow River system have only a minor influence in moderating floods and could not account for significant differences between the pre-1934 record and later record. The non-homogeneity and trends for inflows floods to Bearspaw Reservoir might have been caused as a result of a combination of variations in the climate system and construction of major storage facilities on the Bow River and its tributaries. Therefore, the entire flood record series for the Bow and Elbow Rivers are considered appropriate for frequency analysis. 2.2.2 Frequency Analysis of Flood Series Golder has developed an in-house application of the CFA software developed by Environment Canada using the same multiple parameter probability distribution functions (i.e., Three Parameter Log-Normal, Extreme Value and Log-Pearson Type III) as the CFA software. The application incorporates a customized interface, modern boot strapping, two methods for parameter estimation (method of moments and maximum likelihood), the ability to estimate confidence intervals for a flood estimates of standard return periods and the Anderson-Darling methods (Stephens, 1974) to identify the best-fit probability distribution. This in-house application was used to determine the best-fit probability distribution for each peak flow series and the confidence intervals for flow estimates of all specified flood frequencies. The annual maximum daily flow series generated for the Bow River and its tributaries including the Elbow River are provided in Appendix A. The flood flows for various return periods at various locations on Bow River and its tributaries are summarized in Table 4. Detailed results of the flood frequency analyses including the 95% confidence intervals are provided in Appendix B, Section B1. 2.2.3 Historic Floods Historic floods have been documented only for the Bow River above the Elbow River. Historic flood peaks that pre-date systematic flow monitoring have been established for the Bow River at Calgary for 1897 and 1902 based on recorded high water marks. Newspaper reports and local residents indicated that the 1879 flood observed on the Bow River was as severe as the flood of 1897. These floods were also considered in the estimation of the flood flows on the Bow River at Bearspaw Reservoir. The historic flood estimates for the Bow River upstream of the Elbow River are 2,265 m3/s in 1879 and 1897, and 1,557 m3/s in 1902. The largest recorded flood after systematic monitoring started in 1911 occurred in 1932, with a magnitude of about 1,520 m3/s. Precipitation and temperature data recorded at three climate stations (Calgary Airport, Banff and Kneehill) were analyzed to assess the validity of historic floods recorded on the Bow River in 1897 and 1902. Plots of precipitation and temperature data recorded in 1897 and 1902 are provided in Appendix B, Section B2. The 1897 flood occurred on June 18 and the 1902 flood occurred on July 4. March 2010 Report No. 09-1326-1040 10 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT Table 4: Estimated Daily Flood Flows (m3/s) for Various Return Periods, Derived from Naturalized Annual Maximum Daily Flows Inflow to Local Local Inflow Local Inflow Bearspaw Return Inflow to Inflow to Inflow to Inflow to Bow Inflow to Inflow to to Lower to Upper Reservoir, Period Lake Spray Spray Barrier River at Ghost Glenmore Kananaskis Kananaskis Without Minnewanka Lakes River at Lake Banff Lake Reservoir (Year) Lake Lake Historical Banff Data Drainage Area 151 209 647 520 230 899 2,210 6,550 7,770 1,236 (km2) 2 17.0 22.5 53.5 65.4 11.7 72.1 205 348 378 55.9 5 20.7 26.7 77.0 86.5 18.5 112 257 458 510 107 10 23.2 29.4 95 103 24.8 144 289 535 613 154 20 25.7 32.0 113 122 32.4 180 317 612 724 211 50 29.0 35.4 140 150 44.5 235 353 716 891 302 100 31.6 37.9 163 174 55.5 283 377 797 1,040 385 200 34.2 40.5 188 202 68.4 339 401 881 1,200 481 500 37.8 44.0 224 245 88.4 425 430 997 1,440 632 1,000 40.6 46.6 255 284 106 501 450 1,090 1,660 766 March 2010 Report No. 09-1326-1040 11 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT The recorded precipitation data indicates that: Recorded annual total precipitation in 1902 is the highest recorded precipitation (more than 870 mm at Calgary Airport Station and more than 770 mm at Banff) since 1885; Precipitation over a period of three months (May to July) in 1902 (a total precipitation of about 580 mm at Calgary, about 400 mm at Banff and about 360 mm at Kneehill) is significantly higher than the 1897 precipitation over the same period (May to July); Daily precipitation recorded in June 1897 indicates that a significant amount of precipitation (about 100 mm) was recorded at the Calgary Airport and Kneehill stations from June 14 to June 18. The total precipitation at the Banff station over these four days was about 63 mm; During the occurrence of the 1897 flood, significant precipitation was recorded at all three stations, which indicates that the majority of the upper watershed of the Bow River received significant precipitation over a period of four days (June 14 to June 18); During the occurrence of the 1902 flood, significant precipitation was recorded at the Calgary Airport station and less precipitation were recorded at Banff and Kneehill stations, which indicates that precipitation was not uniform across the upper watershed of the Bow River; and Both the 1897 and 1902 floods are mainly associated with occurrence of rainfall, since recorded air temperature values were greater than the freezing point for more than one (1) month in 1879 and for more than two (2) months in 1902 before occurrence of the major floods. Analyses of precipitation and temperature data indicate that the magnitude of historic floods observed at the City was dependent on the distribution and magnitude of precipitation that occurred in the upstream watershed. Though there is some uncertainty associated with the magnitudes of the historic floods, it is clear that both the 1897 and 1902 floods were caused by significant precipitation events (over the majority of the upstream watershed in 1897 and over the area downstream of Banff in 1902). Hence, the historic floods were considered together with the systematically recorded data to determine the peak flood flows for various return periods along the reach of the Bow River within the City. The inclusion of historic floods in the analysis provides more reliable estimates of probabilities of future large floods. The appropriate frequencies for plotting the historic data were first estimated using the method outlined in United States Geological Survey Bulletin 17B (USGS, 1982) and used in the 1983 study. Direct interpretation of these floods when plotted on the resulting frequency plot (Figure 2) indicates that the 1879 and 1897 floods approximately correspond to 1:200 year events, while the 1902 flood approximately corresponds to a 1:80 year event. A comparison of the frequency curve for systematically-recorded floods (i.e., after 1911) and the plotting positions of the historic floods suggests that the flood frequency curve for systematically-recorded floods tends to underestimate the low probability floods. March 2010 Report No. 09-1326-1040 12 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT 10000 Naturalized data (1908-2008) Naturalized and Historic data- Corrected Plotting Position Naturalized, Historic and Generated data Extreme Value Type II fit to Naturalized data Extreme Value Type II fit to Naturalized and Historic data Extreme Value Type II fit to Naturalized, Historic and Generated data 1000 Q (m³/s) 100 2 5 10 20 50 500 100 200 1.05 1.25 1000 1.003 Return Period (Years) Figure 2: Flood Frequency Curve for the Bow River above Bearspaw Dam A second approach was then used to incorporate the historic floods into the series of systematically-recorded floods. The procedure was as follows: 1) Using parameters of the flood frequency distribution fitted to the naturalized floods of systematically- recorded data (i.e., parameters of the Extreme Value Type II distribution), flood data for the 29 years (1880- 1896, 1898-1901 and 1903-1910) were generated randomly from the probability distribution. 2) A flood frequency analysis was then conducted using the combined series of historic data, naturalized floods derived from systematically-recorded flood data from 1911 onwards and the 29 years of flood data generated from the probability distribution of the systematically-recorded floods. 3) Parameters of the flood frequency distribution fitted in Step 2 were then estimated. The generated flood flows for the 29 years (1880-1896, 1898-1901 and 1903-1910) were then re-generated using the new set of parameters. This step was necessary to make sure that the generated flood flows are from the same population that includes both historic and naturalized floods derived from the systematically-recorded flood record. This ensures that all the flood data from 1879 to 2008 are identically distributed. March 2010 Report No. 09-1326-1040 13 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT The bias that may be introduced by using the systematic record to fill the missing historic flood period would be due to the underlying assumption that the generated series are from the same population as the systematically- recorded data. To minimize this bias, once the missing period was in-filled with the generated data (that did not include the historic floods) the in-filled floods were then re-generated using the combined series of historic data, randomly generated data for 29 years and systematically-recorded data in an iterative manner. The objective of this approach was to ensure that the generated floods belong to the same population as the entire series including the historic floods. This systematic approach removes bias introduced by the generating floods for the intervening years of the missing historic flood period. In addition, the generated data were scanned to make sure that none of the generated floods exceeded the magnitude of the known historic floods, as it is likely that any such event would have been documented. Table 5 and Figure 2 provide the results of the flood flows for Bow River above Bearspaw Dam using the historic and systematically-recorded data. Table 5 also provides a comparison of the approach adopted in this study (i.e., randomly generating 29 years of missing flood flows prior to 1911) to the method used in the 1983 study (Bulletin 17B). Detailed comparisons of the results of the 1983 study with those from the current study are provided in Section 4.0. The results indicate that the frequency distribution fit to the systematically-recorded, historic and generated data (i.e., Extreme Value Type II) and shown in Figure 2, provides the best fit to the flood flows (as shown in Appendix B) with plotting position corrected according to the method outlined in Bulletin 17B. Hence, it is recommended that this frequency distribution fit be used to estimate peak floods for various return periods. The estimated peak floods flows were also qualitatively checked against the empirical Creager C method. From the Creager diagram the peak unit discharge for the Probable Maximum Flood (PMF) for the Bow River above Bearspaw Dam is about 1.6 m3/s/km2, assuming a C value of 60. The estimated PMF value is about 8 times the 1:100 year flood and about 3 times the 1:1,000 year flood. This is consistent with the relationship observed for other regional streams such as the Oldman and Highwood Rivers (Creager et. al., 1945). Table 5: Estimated Daily Flood Flows (m3/s) for the Bow River above Bearspaw Dam, Derived from Naturalized Annual Maximum Daily Flows and Historic Data Based on Systematically- Based on Systematically- Return Period Based on Systematically- Recorded and Historic Data Recorded and Historic Data (Years) Recorded Data (Using Bulletin 17B method) (Using Method developed for this Study) 2 378 371 382 5 510 541 544 10 613 699 691 20 724 - 874 50 891 1,220 1,190 100 1,040 1,550 1,500 200 1,200 1,940 1,900 500 1,440 2,620 2,600 1,000 1,660 - 3,290 March 2010 Report No. 09-1326-1040 14 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT 2.3 Inflow Flood Hydrographs Synthetic inflow flood hydrographs at major storage facilities on the Bow River and its tributaries (including Glenmore Reservoir on the Elbow River) were developed for each flood flow obtained from flood frequency analysis of the naturalized flow series. The synthetic flood hydrographs were developed as follows: For a specific flood magnitude (for example, the 1:100 year flood), flood flows of equal or similar magnitude was identified from the naturalized daily flow series; The flood volumes, times to peak and the hydrograph time bases of the selected flood magnitudes were determined from the naturalized flow series identified for each specific flood event; and A dimensionless Gamma function, with mean, standard deviation and skewness parameters similar to those of the observed hydrograph, was developed for each flood event wherever possible. The flood volumes and flood hydrograph time base estimates were used to scale the ordinates of the dimensionless Gamma functions such that the peak flow rate, the time to peak flow and the total volume of flood closely matched the target values for a given return period. For example, the recorded 311 m3/s flood of May 23, 1932, was used as a surrogate to derive the 1:50 year inflow hydrograph for Glenmore Reservoir (design peak of 302 m3/s).The time to peak for the recorded surrogate flood was about 2 days. The recorded mean discharge and corresponding standard deviation, along with a hydrograph time base of about 7 days, a base flow of 52 m3/s and a skewness coefficient of 0.582, were used to determine the parameters for the Gamma function. It was not always possible to find a surrogate flood from the recorded data for each flood magnitude, especially for floods with greater than 1:50 year return periods. In addition, for some surrogate floods, it was not possible to fit the Gamma function if the skewness coefficient is either small or negative. For those cases, the shape of the Gamma function was assumed to be the same as the inflow design hydrographs derived for any other flood of a different return period. For example, in deriving design inflow hydrographs for Glenmore Reservoir, the only surrogate hydrograph that could be identified and fitted to a Gamma function was the flood hydrograph corresponding to the 1:50 year flood. Hence, the shapes of inflow design hydrographs for other return periods were assumed to be the same as the 1:50 year flood. Flood hydrographs for each frequency were developed for the following basins with hydropower developments that are tributaries to the Bow River: Lake Minnewanka (Cascade), Spray Lakes, Upper and Lower Kananaskis Lakes, Barrier Lake, Ghost Lake and Bearspaw Reservoir. Flood hydrographs were also developed for the Bow River at Banff (WSC Station 05BB001), which records natural flows from a drainage area of 2,210 km2. The sum of the areas drained by the hydro developments and that at Banff is about 7,750 km2. An aggregated natural contributing area of about 1,280 km2 (i.e., the area contributing runoff directly between Ghost Lake and Banff, excluding Spray Lakes, Lake Minnewanka and Upper and Lower Kananaskis Lakes) would normally require its own flood hydrograph. However, the flood hydrographs for this aggregated area were not considered because the volume of inflow runoff from the upstream watersheds to Bearspaw Reservoir is already close to the inflow hydrograph back-calculated from the outflow records at Ghost Lake, without additional runoff from this intervening catchment area. Inflow hydrographs were also developed for Elbow River at Glenmore Reservoir. All the inflow flood hydrographs are provided in Section B3, Appendix B. March 2010 Report No. 09-1326-1040 15 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT 2.4 Flood Hydrograph Routing and Regulated Peak Flow Frequency Estimates The design flood hydrographs created for all return periods were routed through the storage facilities according to the existing flood control and operation guidelines and selected starting water levels. A hydraulic routing model that includes all major storage facilities and outflow structure capacities (combined spillway and hydropower plant capacities) was set up to enable flood routing of the flood hydrographs for each specified frequency. The resulting peak flows from the routed outflow hydrographs provided regulated design flows for input into a hydraulic model for the Bow River below Bearspaw Dam. Similarly, a hydraulic routing model was set up to route the naturalized inflow flood hydrographs of each specified frequency through Glenmore Reservoir, such that the resulting peak flows from the routed outflow hydrographs provided regulated design flows for input into the hydraulic model for the Elbow River. The timings of the arrival of peak flows from each reservoir and the natural contributing areas were based on travel times provided by AENV and estimated flood travel times developed by TransAlta as part of its Flood Action Plan. This information, supplemented by analysis of observed flood times on the main stem of the Bow River, was used to determine best estimates of lag times for the flood hydrographs arriving at Ghost Lake. The combined flood hydrographs were then routed through Ghost Lake and Bearspaw Reservoir. The starting water levels for major storage facilities were conservatively assumed to be at TransAlta’s target level (the “key reservoir elevation” identified in TransAlta’s Flood Action Plan) in mid-June, which corresponds to the period when most of the recorded flood flows occur on the Bow River and its tributaries. The regulated peak flows resulting from routing of naturalized flood hydrographs were determined for each of the specified flood frequencies. The maximum instantaneous flow for each specified frequency for each river was estimated based on the established relationship between recorded annual maximum daily and instantaneous flows. For the Bow River below Bearspaw Dam, this relationship was established using the recorded annual maximum daily and instantaneous flows at WSC Station 05BH004 (Bow River at Calgary) (Figure B2, Appendix B). For the Elbow River above Glenmore Dam, the relationship was established using the recorded maximum daily and instantaneous flows at WSC Station 05BJ005 (Elbow River above Glenmore Dam) and on historic large floods estimated for Elbow River by T. Blench & Associates (1965) (Figure B3, Appendix B). Table 6 provides a comparison of naturalized inflow floods with peak flood flows routed through Bearspaw Reservoir. For the case of systematically recorded floods, inflow hydrographs at Bearspaw Reservoir were obtained by routing through all the major upstream facilities. Normally, the peak flood flows routed through Bearspaw Reservoir should be less than the inflow design floods derived from naturalized flows. However, the peak flood flows routed through Bearspaw Reservoir are greater than the inflow design floods between 1:2 year and 1:20 year return periods, as shown in Table 6. This situation results from the assumption in the hydraulic routing that the peak flood flows from all head-watersheds are coincident (e.g., if a 1:2 year flood occurs at Banff, a 1:2 year flood will also occur at Bearspaw Reservoir). In reality, a 1:2 year flood at Banff will not cause a flood of the same frequency at Bearspaw Reservoir. The consequence of this assumption is that the volume of water during the basin-wide coincident 1:2 year event is significantly larger than the typically-observed 1:2 year flood hydrograph at Bearspaw Reservoir. Furthermore, the time base of the basin-wide coincident 1:2 year event, especially the duration of the peak flow period, is relatively long. March 2010 Report No. 09-1326-1040 16 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT Table 6: Comparison of Flood Inflows and Outflows for the Bow River at Bearspaw Reservoir, Derived from Naturalized Flows (m3/s) Naturalized Flows of Naturalized Flows of Naturalized Flows of Systematically Naturalized Flows of Systematically- Systematically Recorded Data and Systematically Recorded Data, Routed Recorded Data and Historical Floods, Recorded Data through Storage Historic Floods Routed through Storage Return (Inflow) Facilities (Inflow) Facilities(a) Period (Outflow) (Outflow) (Years) Using Method Described in Section 2.2.3 Maximum Maximum Maximum Maximum Instantaneous Instantaneous Instantaneous Instantaneous Daily Daily Daily Daily Discharge Discharge Discharge Discharge Discharge Discharge Discharge Discharge 2 378 418 445 494 382 423 378 418 5 510 568 551 614 544 606 536 597 10 613 685 671 751 691 774 682 763 20 724 811 751 842 874 983 863 970 50 891 1,000 862 969 1,190 1,350 1,180 1,330 100 1,040 1,170 950 1,070 1,500 1,710 1,490 1,700 200 1,200 1,360 1,070 1,210 1,900 2,170 1,890 2,160 500 1,440 1,640 1,340 1,520 2,600 2,980 2,580 2,970 1,000 1,660 1,890 1,560 1,780 3,290 3,810 3,300 3,810 a) Routing of inflow hydrograph was completed through Ghost Lake and Bearspaw Reservoir only. For the case of systematically recorded and historic floods, the inflow hydrographs were obtained by routing through Ghost Lake and Bearspaw Reservoir only because historic information is not available for upstream tributaries so that their hydrographs could be adjusted. The routing of flood flows through Ghost Lake and Bearspaw Reservoir resulted in a reduction of less than 2% in flood magnitudes. The major storage facilities on the Bow River’s main stem were not designed to provide flood storage capacity. Similarly, tributary storage facilities associated with the upper and lower Kananaskis Lakes, and Barrier Lake hydropower developments have limited flood storage capacity. Spray Lakes and Lake Minnewanka have some flood storage capacity and significant flood attenuation capacity, however the drainage area contributing flows to these facilities is less than 15% of the total drainage area contributing runoff to the Bow River at Calgary. Hence, there is limited flood storage capacity that will result in attenuation of the large flood flows along the reach of the Bow River. Table 7 provides the results for the Elbow River below Glenmore Dam. Corrections to peak flood flows routed through Glenmore Reservoir were made to account for historic floods corresponding to those observed for the Bow River. The corrections were made using a similar procedure as the 1983 study since there was no historic flood recorded on the Elbow River. In the 1983 study, the frequency plot for the systematically recorded floods for the Elbow River was shifted to mimic the shape of the frequency plot for the Bow River which includes the historic floods. The resulting change in peak flood flows for the Elbow River are as follows: a 3% decrease for the 1:2 year flood, a 12% increase for the 1:5 year flood, a 24% increase for the 1:10 year flood, a 32% increase for the 1:20 year flood, a 42% increase for the 1:50 year flood, and a 47% increase for the 1:100 year flood. Similar percentage changes were applied for this study. An increase of 47% was assumed for all events greater than the 1:100 flood. March 2010 Report No. 09-1326-1040 17 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT Table 7: Comparison of Inflow and Outflow Flood Flows for the Elbow River below Glenmore Dam, Derived from Naturalized Flows (m3/s) Naturalized Flows of Naturalized Flows Routed Naturalized Flows Routed through Systematically-Recorded through Storage Facility and Storage Facility Return Data Corrected for Historic Floods (Outflow) Period (Inflow) (Outflow) (Years) Maximum Maximum Daily Instantaneous Maximum Daily Instantaneous Instantaneous Daily Discharge Discharge Discharge Discharge Discharge Discharge 2 55.9 66.8 45.1 53.3 43.6 51.5 5 107 132 73.7 89.1 82.2 99.4 10 154 193 126 156 156 193 20 211 267 165 206 219 274 50 302 389 245 313 349 445 100 385 501 366 476 538 699 200 481 633 477 626 701 922 500 632 841 624 830 917 1,220 1,000 766 1,030 757 1,020 1,110 1,490 2.5 Peak Flow Estimates for Major Tributaries The purpose of this task was to generate 1:2 to the 1:1,000 year flood flow estimates for each major tributary entering the Bow River between Bearspaw Reservoir and the Highwood River confluence (i.e., Nose Creek, Fish Creek, Pine Creek, and Highwood River). A frequency analysis of the annual maximum instantaneous flow series for each tributary was performed using the same approach described in Section 2.2.2. 2.5.1 Nose Creek at Calgary Flow records for Nose Creek at Calgary (WSC Station 05BH003) are available for the 1911-1919 and 1973-1986 periods. The annual series of recorded maximum instantaneous discharges, maximum daily discharges and corresponding dates are provided in Section B4, Appendix B. An attempt was made to establish a regional relationship between Nose Creek and other regional streams (Fish Creek, Jumpingpound Creek and Threepoint Creek) to extend the flood flow series. However, coincident floods in Nose Creek and other regional streams occurred only for 2 or 3 years over a period of more than 20 years. Hence, it was not possible to establish a meaningful regional relationship to extend the flood series for Nose Creek. Moreover, the Nose Creek watershed has significantly larger non-contributing areas compared to the other regional watersheds. A relationship between annual maximum daily discharges and maximum instantaneous discharge was established as shown in Section B4, Appendix B, to estimate annual maximum instantaneous discharge for those years where instantaneous flows were not available. The flood flow estimates for Nose Creek at Calgary for various return periods are provided in Table 9. In the 1983 study, the flood flow estimates for the Nose Creek were based on recorded data up to 1980. In the current study, the flood flow estimates for the Nose Creek are based on recorded data up to 1986. The 1:100 March 2010 Report No. 09-1326-1040 18 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT year flood estimate for Nose Creek in the current study (89.0 m3/s) is comparable to that of the 1983 study (81.0 m3/s), especially in consideration of the confidence limits of their statistical estimates. 2.5.2 Fish Creek at the Mouth Flow records for Fish Creek at the Mouth (WSC Station 05BK003) are available only for the 1989-1993 period. Hence, an estimation of flood flows for Fish Creek at the mouth requires a transfer of the data recorded at Fish Creek near Priddis (WSC Station 05BK001). The transfer of flood flows from the upstream station to the Mouth was accomplished using the following steps. 1) Establish a regional relationship between drainage area and flood peaks for various return periods. 2) Establish a relationship between annual maximum daily flows and annual instantaneous flows for Fish Creek near Priddis. 3) Transfer the annual maximum daily flood flows from Fish Creek near Priddis to Fish Creek at the Mouth based on the regional relationship established in Step 1. 4) Generate instantaneous flood flows for Fish Creek at the Mouth based on the relationship established in Step 2. Threepoint Creek near Millarville (WSC Station 05BL013) has a drainage area of 507 km2. A review of recorded flow data at this station indicated a high level of similarity with the flow regime at Fish Creek near Priddis (drainage area of 261 km2).), The annual maximum daily flows at both stations occurred on the same date for 23 years over a period of 25 years (1967-2007), and lagged by 1 day for the remaining 2 years. Similarities between the closely located Fish Creek and Threepoint Creek watersheds permitted the establishment of a regional relationship (provided in Section B4, Appendix B) that allowed for the transfer of the flood flow data from Fish Creek near Priddis to Fish Creek at the Mouth (drainage area of 442 km2). A relationship between annual maximum daily discharges and annual maximums instantaneous discharges was established using the Fish Creek near Priddis data, as shown in Section B4, Appendix B. This relationship was used to derive the annual maximum instantaneous discharges for Fish Creek at the Mouth from the corresponding maximum daily discharges. The resulting flood flow estimates for Fish Creek at the Mouth for various return periods are provided in Table 9. In the 1983 study, the flood flow estimate for the Fish Creek was based on the recorded data up to 1980. In the current study, the flood flow estimate for the Fish Creek is based on the recorded data up to 2008. The regional relationships established between the Threepoint Creek near Millarville and the Fish Creek near Priddis and used to transfer of the flood flow data from the Fish Creek near Priddis to the Fish Creek at the Mouth, are different between the 1983 and current studies. In addition, the relationship between annual maximum daily discharges and annual maximum instantaneous discharges established using the Fish Creek near Priddis data are different between the 1983 and current studies. Because of these differences, the 1:100 year flood flow estimate for Fish Creek in the current study (444 m3/s) is about 28% higher than that of the 1983 study (348 m3/s). March 2010 Report No. 09-1326-1040 19 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT 2.5.3 Pine Creek at the Mouth The only recorded flow data available for Pine Creek was collected by Golder (2008) between 2005 and 2007 as part of baseline monitoring completed for the City. As part of a Pine Creek drainage study, AMEC (2007) derived flood flows for Pine Creek based on regional analysis as well as on flows simulated using the Hydrologic Simulation Program Fortran (HSPF) model (1966- 2005). The flood flows derived from hydrologic simulation are about 23% to 32% of those derived from regional analysis, as shown in Table 8. The flood flows obtained from simulations with the HSPF model are comparable to the lower bound estimates derived from the regional analysis. The regional analysis was updated in this study to include data up to 2008. Figure B8 in Appendix B provides the relationship established between drainage area and daily maximum flood flows for various return periods. An instantaneous to daily discharge ratio of 1.34 was estimated using measured flood data in June 2005 on Pine Creek at Highway 2. The resulting flood flow estimates for Pine Creek at the Mouth for various return periods are provided in Table 8. Table 8: Instantaneous Flood Peak Discharges (m3/s) for Pine Creek, Provided by AMEC (2007) Percent of HSPF Flood Estimates Using Flood Estimates Using Return Period (Years) Simulated to Regional HSPF Simulated Flows Regional Analysis Derived Data (%) 2 1.6 7.0 23 5 5.1 17.0 30 10 7.9 25.6 31 20 11.1 35.2 32 50 15.8 49.6 32 100 19.7 62.1 32 2.5.4 Highwood River near the Mouth Flow records for Highwood River near the Mouth (WSC Station 05BL024) are available for the period from 1970 to 2008. The annual series of recorded annual maximum instantaneous discharges, annual maximum daily discharges and corresponding dates are provided in Section B4, Appendix B. A relationship between annual maximum daily discharges and annual maximum instantaneous discharge was established as shown in Section B4, Appendix B. The relationship was used for estimating to estimate annual maximum instantaneous discharges for those years where they were not recorded. The resulting flood flow estimates for Highwood River near the Mouth for various return periods are provided in Table 9. March 2010 Report No. 09-1326-1040 20 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT Table 9: Instantaneous Flood Peak Discharges (m3/s) for Bow River Tributary Streams Return Period Nose Creek Fish Creek Pine Creek Highwood River (Years) Drainage Area (km2) 893 442 212 3,950 2 6.15 39.0 6.92 191 5 14.2 85.7 16.3 396 10 23.0 134 24.7 583 20 35.3 198 34.3 804 50 60.2 317 49.3 1,160 100 88.8 444 62.7 1,480 200 130 618 78.2 1,850 500 214 946 102 2,420 1,000 310 1,300 123 2,930 2.6 Peak Flow Estimates for the Lower Reach of Bow River The flood flows for the lower reach of the Bow River (the reach between its confluences with Elbow River and Highwood River) were derived using two approaches. In the first approach, the peak flow of a specific frequency for a specific tributary was added to the peak flow of the same frequency in the reach of the Bow River just upstream of the tributary to estimate the peak flow of the same frequency downstream of the confluence. This approach results in conservative estimates of peak flows downstream of tributary confluences, but it is consistent with the direction provided in the Request for Proposals for this project to assume “coincident events for all rivers and streams”. The resulting flood flows at various locations along the Bow River are provided in Table 10. Table 10: Instantaneous Flood Flows along the Bow River, Assuming Coincident Events for all Rivers and Streams, Regulated Scenario Bow River below Bow River below Bow River below Bow River below Bow River below Return Period Confluence with Confluence with Confluence with Confluence with Confluence with (Years) Elbow River Nose Creek Fish Creek Pine Creek Highwood River 2 470 476 515 522 713 5 696 711 796 813 1,210 10 956 979 1,110 1,140 1,730 20 1,240 1,280 1,480 1,510 2,320 50 1,780 1,840 2,150 2,200 3,390 100 2,400 2,490 2,930 2,990 4,470 200 3,080 3,210 3,830 3,910 5,760 500 4,190 4,400 5,350 5,450 7,870 1,000 5,300 5,600 6,900 7,020 9,950 March 2010 Report No. 09-1326-1040 21 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT The second approach was based on that used in the 1983 study, and involved developing maximum instantaneous discharges for the Bow River downstream of its confluence with the Elbow River. The maximum instantaneous discharges in the Bow River below its confluence with the Elbow River were derived using one of the following criteria: 1) simultaneous occurrence of instantaneous peak flows on both the Bow River and the Elbow River; 2) instantaneous discharge on the Bow River plus daily flow on the Elbow River; 3) daily flow on the Bow River and Instantaneous discharge on the Elbow River; and 4) daily flow on the Bow River and daily flow on the Elbow River. The selection of the governing criterion was based on an examination of the concurrent data records for the Bow River and the Elbow River. Once the maximum instantaneous discharges are generated downstream of the Elbow River, a frequency analysis of the annual flood flow series was undertaken to determine flood magnitudes of each specified frequency. The annual series of maximum instantaneous discharges along with the corresponding dates for the Bow River below its confluence with the Elbow River are provided in Section B5, Appendix B. The flood flows for the remaining tributaries (Nose Creek, Fish Creek, Pine Creek and Highwood River) are assumed to be coincident with the corresponding flood flows in the Bow River for various return periods. The flood flow estimates for various locations along the Bow River within the City for various return periods are provided in Table 11. Table 11: Instantaneous Flood Flows along the Bow River, Based on Method Used in the 1983 Study Bow River below Bow River below Bow River below Bow River below Bow River below Return Period Confluence with Confluence with Confluence with Confluence with Confluence with (Years) Elbow River Nose Creek Fish Creek Pine Creek Highwood River 2 410 416 455 462 653 5 632 646 732 748 1,140 10 850 873 1,010 1,030 1,610 20 1,140 1,170 1,370 1,400 2,210 50 1,660 1,720 2,040 2,090 3,250 100 2,220 2,310 2,754 2,810 4,290 200 2,960 3,090 3,710 3,790 5,640 500 4,350 4,570 5,510 5,620 8,040 1,000 5,830 6,140 7,440 7,560 10,500 Confidence intervals for each flood frequency estimate (using the conventional frequency curves if they are not affected by historic data or the modified frequency curves otherwise) are presented as the lower and upper 95% bounds in Appendix B. March 2010 Report No. 09-1326-1040 22 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT 3.0 DISCUSSION 3.1 Commentary on Effects of Climate Change on Flood Estimates Most of the flows in the Bow and Elbow River watersheds originate in the mountains, where water is stored as snow and glacier ice. Recent studies on the effect of climate change (e.g., Martz et al., 2007; Valeo et al., 2007) indicate that climate change could result in increased temperature, more frequent drought and water shortages, increased precipitation in some areas and increased flooding. As a result of climate change and variability, many regions of Canada, including the prairies could experience warmer air temperatures and changes in streamflow magnitude and timing (e.g., higher winter streamflow and lower summer streamflow). Depending on the climate model used for prediction of future scenarios, precipitation is projected to increase in Alberta, with less precipitation falling as snow and more rainfall-on-snow precipitation events (Valeo et al., 2007). Hence, it is anticipated that such changes in precipitation patterns could increase the frequency and intensity of extreme events (flood, drought, hail and windstorms). For the Bow River watershed, it is predicted that if rain-on- snow events occur more frequently and the snowpack begins to melt earlier, then flood events could occur earlier in the spring. Using the predictions from the Canadian Regional Climate Model, Valeo et al. (2007) showed that May precipitation could increase by more than 35% under a double carbon dioxide (2xCO2) scenario. As a result, expected increases in precipitation during the month of May could nearly double spring flood peak flows. Based on recorded data over the past 100 years, the observed flood magnitudes in recent years do not appear to be increasing with time in either the Bow River or Elbow River as shown in Figure 3. The flood in 2005 in the Bow River and Elbow River basins is the only significantly large flood since 1932, however, the magnitude of this flood is well within the variability in flood flows since 1910. About 80% of the recorded flood peaks in the Bow River and the Elbow River occurred at the end of May and in the month of June as shown in Table 12 and Figure 4. The frequency of floods occurring outside this period (earlier or later) also does not appear to be changing with time. The recent patterns in the timing of these floods are similar to what were observed at the beginning the century. There is no clear evidence that the patterns in magnitude or timing of flood peaks have changed significantly over the past hundred years. March 2010 Report No. 09-1326-1040 23 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT 1200 1100 1000 Bow River Elbow River /s) 900 3 (m 800 Discharges 700 Flood 600 Daily 500 400 Maximum 300 Annual 200 100 0 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2010 Year Figure 3: Flood Series for the Bow River at Calgary (WSC Station 05BH004) and the Elbow River at Bragg Creek (WSC Station 05BJ004) Table 12: Time of Occurrences of Annual Maximum Daily Flood Events in the Bow and Elbow Rivers Bow River at Calgary, Occurrences of Annual Elbow River and Bragg Creek, Occurrences of Month Maximum Daily Flood Events Since 1911 Annual Maximum Daily Flood Events Since 1911 Number % Number % April - - 4 4 May 7 7 23 23 June 70 72 55 55 July 16 16 6 6 August 3 3 8 8 September 1 1 4 4 March 2010 Report No. 09-1326-1040 24 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT Figure 4: Timing of Recorded Floods for the Bow River at Calgary and the Elbow River at Bragg Creek. March 2010 Report No. 09-1326-1040 25 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT 3.2 Commentary on the Seasonality of Flood Peaks The naturalized peak flow series for the Bow River below Bearspaw Dam and the Elbow River below Glenmore Dam were analyzed to identify whether one flood mechanism (snowmelt or rainfall) dominates flooding for various return periods. Based on recorded data over the past 100 years, about 80% of the annual peak flows for the Bow River and the Elbow River occurred at the end of May and in June, as shown in Table 11, Section 3.1. This suggests that snowmelt or rain-on-snow events dominate generation of peak flows in the Bow River and Elbow River watersheds. However, it should be noted that the most significant historical floods are associated with high rainfall or rain-on-snow events. Frequency analyses were conducted on the seasonal flood series that were generated from the naturalized (1930-2008) and recorded flows (1911-1930) to compare the flood estimates with those derived based on annual flood series. The spring flood series period was defined to occur from April to mid-June, while the summer flood series period was defined to occur from mid-June to September. Table 13 provides a comparison of the flood magnitudes generated for various return periods for the Bow River below Bearspaw Dam and the Elbow River below Glenmore Dam. Table 13: Flood Magnitudes Derived Using Seasonal Flood Series Bow River below Bearspaw Dam Elbow River below Glenmore Dam Return Based Based Based Percentage Based Based Based Percentage Period on on on Compared to on on on Compared to (Years) Spring Summer Annual Annual Floods Spring Summer Annual Annual Floods Floods Floods Floods Floods Floods Floods Spring Summer Spring Summer 2 346 330 378 -9% -13% 48 37 56 -15% -33% 5 474 439 510 -7% -14% 95 67 101 -6% -33% 10 570 512 613 -7% -16% 139 95 144 -4% -34% 20 670 584 724 -8% -19% 192 129 200 -4% -35% 50 814 677 891 -9% -24% 279 190 300 -7% -37% 100 934 748 1,036 -10% -28% 358 250 403 -11% -38% 200 1,065 820 1,198 -11% -32% 451 328 539 -16% -39% 500 1,255 916 1,444 -13% -37% 598 467 788 -24% -41% 1,000 1,415 990 1,657 -15% -40% 729 606 1,045 -30% -42% For the Bow River, the flood peaks derived using spring flood series are less than those derived using annual flood series by about 7% to 15%. The flood peaks derived using summer flood series are significantly smaller than those derived using annual flood series, especially for floods with greater than 1:50 year return periods. For the Elbow River, the flood peaks derived using spring flood series are less than those derived using annual flood series by about 4% to 16% for floods with less than the 1:200 year return period. The peaks derived using summer flood series are significantly smaller (by more than 30%) than those derived using the annual flood series. The flood peak estimates based on the annual series are higher than those based on either spring or summer flow series, because the annual flood peaks can occur in spring or summer (Table 12). Therefore, flood peaks for various return periods should be estimated based on the annual flood flow series. March 2010 Report No. 09-1326-1040 26 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT 3.3 Magnitude and Timing of Stormwater Runoff Inputs and Diversions Within the City, the Bow River and its tributaries (Elbow River, Nose Creek, Fish Creek and Pine Creek) receive significant stormwater runoff from urban and developed areas. Stormwater runoff contributing to the flood flows of the Bow River above its confluence with Elbow and to the tributary streams such as Elbow River, Nose Creek, Fish Creek and Pine Creek has already been accounted for in the recorded flow values. The analysis of the effect of the stormwater runoff was limited to areas that are directly contributing runoff to the Bow River which were not accounted for in the recorded data (i.e., BR-3 shown in Figure 5 is a tributary sub-catchment contributing runoff in the reach of Bow River between its confluence with the Elbow River and Fish Creek) (Figure 5 was obtained from City of Calgary in 2008 as part of Bow River Loading Update Study). Figure 5: Sub-Catchments Contributing Stormwater Runoff to the Bow River and its Tributaries (Source: City of Calgary) March 2010 Report No. 09-1326-1040 27 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT The 2005 flood is the largest flood recorded in the Bow River and Elbow River watershed since 1932. The 2005 flood peak flow has an approximate 1:20 year return period for the Bow River below Bearspaw Dam and an approximate 1:50 year return period for Elbow River below Glenmore Dam. This flood was used as a reference to assess the effect of stormwater runoff (magnitude and timing) on the flood peaks estimated for the Bow River below its confluence with Elbow River. The 2005 peak flood flow on the Bow River occurred on June 18 at 18:15, while the peak flow on the Elbow River occurred on June 19 at 04:05. The 24-hour rainfall amounts recorded on those days were about 46 mm at the Calgary Airport and about 97 mm at Elbow Ranger Station. The rainfall amount recorded at the Calgary Airport has an approximate 1:5 year return period while amount recorded at Elbow Ranger Station has an approximate 1:100 year return period. As part of a Bow River loading study, a stormwater runoff simulation was completed for the entire City using the QHM (CWRS, 1997) continuous simulation model (Golder, 2007). The simulation period for this study includes the stormwater runoff during the June 2005 rainfall event. The daily average peak storm runoff discharge from tributary sub-catchments contributing to the Bow River between its confluence with the Elbow River and Fish Creek was estimated to be 20.5 m3/s on June 17. The peak discharge (average runoff over a 1 hour period) was about 50 m3/s, and occurred on June 17 at 23:00. The instantaneous stormwater runoff from the tributary sub-catchments was less than 7% of the peak discharge recorded on the Bow River on June 18. The lag-time between the flood runoff from the tributary sub-catchment and the peak flood on the main stem of the Bow River was about 19 hours. The peak stormwater runoff occurred during the rising limb of the Bow River flood hydrograph and therefore the actual contribution of the stormwater runoff during the peak flow in the Bow River was less than 7%. The contribution of the receding stormwater runoff to the Bow River peak flow was estimated to be 20.5 m3/s, about 3% of the recorded peak flood flow. The analysis of the 2005 flood shows that the contribution of the storm-runoff from the sub-catchment to the peak flood flows in the Bow River was small. This indicates that the flood peak flows estimated in Section 2.6 for the Bow River would change insignificantly through the City. Flows diverted from the Bow River to the Western Irrigation District (WID) have been recorded at WSC Station 05BM015 since 1998. Based on the recorded data, annual diversion from the Bow River start in April and continues to the end of September or to mid-October in some years. As shown in Figure 6, maximum annual diversions have historically occurred at the end of June or in July and have varied between about 15 and 24 m3/s (except in 1999, when the diversion was about 13 m3/s). During the June 2005 flood, the diversion through the WID canal was 2.5 m3/s. This is less than 0.3% of the peak flood flow observed in the Bow River. Hence, diversion flows are expected to be relatively small when compared to peak flows on the Bow River’s main stem. March 2010 Report No. 09-1326-1040 28 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT 30 25 20 /s) 3 15 Diversion (m Diversion 10 5 0 Figure 6: Diversions from the Bow River by the Western Irrigation District (WID) March 2010 Report No. 09-1326-1040 29 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT 4.0 COMPARISON OF FLOOD FLOW ESTIMATES Table 14 shows a comparison of flood flows derived for the Bow River above its confluence with Elbow River in this study with those in the 1983 study. The 1983 flood frequency analysis was based on recorded data up to 1980. The current study includes data up to 2008, reflecting an additional 28 years of data. Table 14: Comparison of Instantaneous Flood Flows (m3/s) Derived for the Bow River Upstream of its Confluence with the Elbow River Re-run of the Naturalized 1983 Study Naturalized Flow Data up Natural and Naturalized Re-run of Flood Data with Flow Data to 2008 using Regulated Flow Data the 1983 Estimates Method Used up to 2008 Method in the Flow Data up up to 2008 Study Data Reported in the Current Using Current Study to 2008 Routed Return Using in the 1983 Study for Bulletin for Using through Period Bulletin 17B Study(b) Incorporating 17B Incorporating Bulletin 17B Bearspaw (Years) Method(b) Historic Method(a) Historic Method(a) Reservoir(b) Floods(a) Floods(a) (1879- (1879-1980) (1879-1980) (1879-2008) (1879-2008) (1879-2008) (1879-2008) 1980) 2 351 392 397 441 423 358 418 5 617 605 589 603 606 552 597 10 886 810 777 783 774 738 763 20 1,190 - 1,020 - 983 970 50 1,630 1,520 1,460 1,387 1,350 1,370 1,330 100 1,970 1,970 1,930 1,760 1,710 1,770 1,700 200 - 2,540 2,540 2,220 2,170 2,270 2,160 500 - 3,520 3,680 3,010 2,980 3,140 2,970 1,000 - - 4,880 - 3,810 - 3,810 (a)Include both historic and generated data (1880-1897, 1898-1901 and 1903-1910) (b) Include historic data only. A comparison of the 1983 and current study results suggests the following: A re-run of the data used in the 1983 study using the Bulletin 17B method indicates that the 1:100 year flood is the same as what was reported. However, the 1:5, 1:10 and 1:50 year floods are less than what were reported, while the 1:2 year flood is slightly higher than what was reported in the 1983 study. By extending the data set used in the 1983 study to include data after 1980 and up to 2008, the naturalized 1:100 year flood estimated using the Bulletin 17B method is reduced by about 11%, from 1,970 m3/s to 1,760 m3/s. The addition of 28 more years of data is the main cause for the decrease in the flood estimates relative to those in the 1983 study. The recorded natural flood series (prior to 1932) and the recorded regulated flood series (post-1932) (i.e., without naturalization) were combined in one series for additional comparative analyses. Frequency analysis of this series suggests a 1:100 year flood estimate of 1,770 m3/s. This confirms that the flow naturalization for the Bow River has little effect, the difference between 1,760 m3/s and 1,770 m3/s is less than 1%. March 2010 Report No. 09-1326-1040 30 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT Using the method in this study for generating 28 years of missing flood data and accounting for the historic floods, results in a 1:100 year flood estimate of 1,930 m3/s using the data up to 1980 and 1,710 m3/s using the data up to 2008. The results using the method of the current study are within 2% to 3% of the 1:100 year flood estimate derived using the method of the 1983 study. This confirms that both approaches used to account for historic floods produce similar flood estimates. The same conclusion can be made by comparing the two approaches using data up to 2008. In this study, the design inflow hydrographs were routed through Ghost Lake and Bearspaw Reservoir. This resulted in a further reduction of the 1:100 year flood by about 2%, to 1,700 m3/s. In the 1983 study, the effect of flow regulation was assessed for both the Bow and Elbow Rivers. For the Bow River, the conclusion was that some of the regulated flows were higher than the naturalized flows. Hence, no corrections were made in the 1983 study to the flood estimates due to flow regulation. The comparative analyses noted above suggest differences between the flood estimates reported in the 1983 study and the current study are mainly due to inclusion of additional flood data from 1981 to 2008, as shown in Table 14. The results obtained using the method adopted in this study are consistent with those using the method used in the 1983 study (Bulletin 17B). For the Elbow River, the analyses indicated that there is a decrease in some of the flood estimates compared to the values reported in the 1983 study. The decrease reflects the additional 28 year of data available up to 2008, effects of flow regulation and use of a different relationship between maximum daily and instantaneous flows. In the 1983 study, the relationship between maximum daily and instantaneous flood flows was based on instantaneous discharges estimated by T. Blench & Associates for the Elbow River above Glenmore Dam. Since 1977, recorded data for both annual maximum daily and annual maximum instantaneous flows are available for the Elbow River below Glenmore Dam. A new relationship was established in the current study using the recorded data and 3 of the largest floods estimated by T. Blench & Associates. Figure B3, Appendix B, provides a comparison of the relationships developed in the 1983 study and the current study. The flood flows estimated using the relationship established in the current study results in an increase in the 1:2 year flood, as shown in Table 15. Floods of greater return period show decreases compared to the estimates of the 1983 study. The relationship established in the current study is believed to reflect the conditions when Elbow River flood flows are routed through Glenmore Reservoir. Table 16 shows a comparison of flood flows derived in this study with those in the 1983 study at various locations along the Bow River and its tributaries. The 1983 flood estimates are comparable considering the confidence limits of their statistical estimates. March 2010 Report No. 09-1326-1040 31 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT Table 15: Comparison of Instantaneous Flood Flows (m3/s) Derived for the Elbow River below Glenmore Dam(d) Re-run of Flood Re-run of Natural and Natural and the 1983 Estimates the 1983 Naturalized Naturalized Regulated Regulated Study Data Return Reported in Study Data Flow Data Flow Data Flow Data Flow Data Using Period the 1983 Using CFA up to 2008(a) up to 2008(b) up to up to Bulletin 17B (years) Study Method 2008(a)(c) 2008(b)(c) Method (1908-1980) (1908-1980) (1908-1980) (1908-2008) (1908-2008) (1908-2008) (1908-2008) 2 58.9 66.3 66.5 64.8 66.8 67.9 69.7 5 152 137 136 136 132 132 129 10 238 210 207 206 193 194 182 20 331 303 294 267 271 248 50 462 489 478 443 389 408 361 100 566 678 660 584 501 547 471 200 926 900 754 633 725 611 500 1,380 1,330 1,030 841 1,040 852 1,000 1,780 1,280 1,030 1,370 1,090 (a) Instantaneous discharge calculated based on the relationship established in 1983, using estimated data from T. Blench & Associates (1965) for the Elbow River above Glenmore Dam. (b) Instantaneous discharge calculated based on the relationship established in the current study, using the recorded data at WSC Station 05BJ001 (Elbow River below Glenmore Dam) (1975-2008) and Three Large Floods from T. Blench & Associates (1965). (c) The naturalized flow data include flows naturalized to remove the effects of Glenmore Reservoir. The regulated flow data in the Table indicates natural flow data for Elbow River below Glenmore Dam (1908-1932) and for Elbow River above Glenmore Dam (1933-1977), as well as for regulated flows recorded downstream of Glenmore Dam. (d) Corrections for historic flood were not applied in this table. Table 16: Instantaneous Flood Flows along the Bow River at Various Locations, Based on Method Used in the 1983 Study Bow River Nose Elbow Fish Creek Bow River Bow River Bow River Return above Creek at River below at Bow below Elbow below Nose below Fish Period Elbow Bow River Glenmore River River Creek Creek (Years) River Confluence Dam Confluence Confluence Confluence Confluence Confluence 1983 Study Results (AENV, 1983) 2 5.55 56.9 32.3 351 377 382 413 5 19.7 169 96.8 617 705 725 821 10 32.6 185 151 886 971 1,010 1,160 20 46.4 340 209 1,190 1,430 1,480 1,690 50 65.7 597 289 1,630 2,140 2,210 2,490 100 80.7 759 348 1,970 2,670 2,750 3,110 Current Study Results (2010) 2 6.15 51.5 39.0 418 410 416 455 5 14.2 99.4 85.7 597 632 646 732 10 23.0 193 134 763 850 873 1,010 20 35.3 274 198 970 1,140 1,170 1,370 50 60.2 445 317 1,330 1,660 1,720 2,040 100 88.8 699 444 1,700 2,220 2,310 2,750 March 2010 Report No. 09-1326-1040 32 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT 5.0 CONCLUSIONS AND RECOMMENDATIONS 5.1 Conclusions The hydrology study for the Bow River and its tributaries through the City is required to determine the flood flows to be used in a HEC-RAS hydraulic model for the Bow and Elbow River within the City. As part of the hydrology study, naturalized flow series were generated at the major storage facilities on the Bow River and its tributaries upstream of Bearspaw Reservoir and on the Elbow River at Glenmore Reservoir. This study included a detailed review of the available hydrometric data. The review indicated that data were missing at several locations for a number of years. The missing data may affect to a small extent the level of certainty on flow statistics derived from the naturalized series. Statistical tests were conducted on annual maximum daily flows at various locations on the Bow River and its tributaries to assess if the data were independent (not serially correlated), without trend, random, and homogeneous. Most of the recorded flood flow series for the Bow River and its tributaries were found to be independent, random and homogeneous, and have no trend. Flood flow series recorded for the Bow River above Bearspaw Dam displayed some trend and non-homogeneity, primarily due to a combination of long-term variability in the climate regime and construction of the major facilities on the Bow River and its tributaries. However, the entire recorded flood flow series were considered appropriate for frequency analysis. Historic floods that pre-date systematic flow monitoring were considered in the estimation of flood flows on the Bow River at Bearspaw Reservoir. The estimates for these floods were originally determined as part of the 1983 study and were adopted without change for the current study. The inclusion of historic flood flow data in the analysis increased the magnitude of peak discharges significantly for floods with low probability of occurrence (i.e., floods with long return periods), compared to the estimates obtained based on the systematically-recorded flood flow series only. In general, the results showed that the flood moderating capacity of the major storage facilities on the Bow River was relatively small. Minor flood regulation capacity is provided on the Elbow River by Glenmore Reservoir. The flood flows estimated in this study for the Bow River and its tributaries through the City are somewhat smaller than those reported in the 1983 study. For example, the estimated 1:100 year flood flows on the Bow River above and below the Elbow River confluence are 1,700 and 2,410 m3/s respectively, as compared to 1,970 and 2,670 m3/s reported in the 1983 study. The differences are mainly due to inclusion of additional flood flow data from 1981 to 2008 in the frequency analyses. 5.2 Recommendations Tables 17 and 18 summarize the recommended naturalized and regulated instantaneous flood peak discharges with various return periods for the Bow River and its tributaries through the City for use in the HEC-RAS hydraulic model. These recommendations are based on the results of this study. March 2010 Report No. 09-1326-1040 33 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT Table 17: Recommended Instantaneous Flood Peak Discharges (m3/s) for the Bow River and its Tributaries, Regulated Scenario Bow Elbow Elbow Bow Bow Bow Bow Nose Fish Pine Bow River Return River River River River River River River Highwood Creek at Creek at Creek at below Period above above below below below below below River at Bow Bow Bow Highwood (years) Elbow Glenmore Glenmore Elbow Nose Fish Pine Bow River River River River River River Dam Dam River Creek Creek Creek 2 418 64.4 51.5 470 6.15 476 39.0 515 6.92 522 191 713 5 597 146 99.4 696 14.2 711 85.7 796 16.3 813 396 1,210 10 763 238 193 956 23 979 134 1,110 24.7 1,140 583 1,730 20 970 353 274 1,240 35.3 1,280 198 1,480 34.3 1,510 804 2,320 50 1,330 553 445 1,780 60.2 1,840 317 2,150 63.5 2,200 1,160 3,390 100 1,700 737 699 2,400 88.8 2,490 444 2,930 62.7 2,990 1,480 4,470 200 2,160 929 922 3,080 130 3,210 618 3,830 78.2 3,910 1,850 5,760 500 2,970 1,240 1,220 4,190 214 4,400 946 5,350 102 5,450 2,420 7,870 1,000 3,810 1,510 1,490 5,300 310 5,600 1,300 6,900 123 7,020 2,930 9,950 Table 18: Recommended Instantaneous Flood Peak Discharges (m3/s) for the Bow River and its Tributaries, Naturalized Scenario Bow Elbow Elbow Bow Bow Bow Bow Nose Fish Pine Bow River Return River River River River River River River Highwood Creek at Creek at Creek at below Period above above below below below below below River at Bow Bow Bow Highwood (years) Elbow Glenmore Glenmore Elbow Nose Fish Pine Bow River River River River River River Dam Dam River Creek Creek Creek 2 423 64.4 64.4 488 6.15 494 39.0 533 6.92 540 191 731 5 606 146 146 753 14.2 767 85.7 853 16.3 869 396 1,270 10 774 238 238 1,013 23 1,040 134 1,170 24.7 1,190 583 1,730 20 983 353 353 1,340 35.3 1,370 198 1,570 34.3 1,610 804 2,410 50 1,350 553 553 1,900 60.2 1,960 317 2,280 63.5 2,330 1,160 3,390 100 1,710 737 737 2,450 88.8 2,530 444 2,980 62.7 3,040 1,480 4,520 200 2,170 929 929 3,100 130 3,230 618 3,850 78.2 3,930 1,850 5,780 500 2,980 1,240 1,240 4,220 214 4,430 946 5,380 102 5,480 2,420 7,900 1,000 3,810 1,510 1,510 5,310 310 5,600 1,300 6,900 123 7,020 2,930 9,950 March 2010 Report No. 09-1326-1040 34 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT 6.0 REFERENCES Alberta Environment. (1983). 1983 Calgary Floodplain Study. Creager, W.P., Justin, J.D., and Hinds, J. (1945). Engineering for Dams, Vol. 1, General Design. Wiley, New York. Center for Water Resources Studies (CWRS). (1997). QHM Model User Manual. DelTech University. Halifax, Canada. Golder Associates Ltd. (Golder). (2007). Watershed Runoff Modeling Analysis in Support of the Bow River Loading Modeling Study. Golder Associates Ltd. (Golder). (2008). Aquatics Baseline Information for Pine Creek, 2005-2007. Martz, Lawrence, Joel Bruneau and J. Terry Rolfe. (2007). Climet Change and Water: SSRB Final Technical Report. Optimal Solutions Ltd. (2009). Natural Flow Computation Program (NFCP) User's Manual, submitted to Prairie Provinces Water Board PWB in Nov. 2009. Stephens, M. A. (1974). EDF Statistics for Goodness of Fit and Some Comparisons, Journal of the American Statistical Association, Vol. 69, pp. 730-737. T. Blench and Associates Limited. (1965). Flood Protection – Elbow River at Calgary, August 1965. United States Geological Survey (USGS). (1982). Guidelines for Determining Flood Flow Frequency, Revised Bulletin 17B of the Hydrology Committee, U.S. Water Resources Council. Valeo, C., Xiang, Z., Bouchart, F.J.C., Yeung, P., and Ryan, M.C., (2007). Climate change impacts in the Elbow River watershed. Canadian Water Resources Journal. Vol. 32(4): 285-302. 7.0 LIMITATION OF LIABILITY This report has been prepared by Golder Associates Ltd. (Golder) for the benefit of the client to whom it is addressed. The information and data contained herein represent Golder's best professional judgment in light of the knowledge and information available to Golder at the time of preparation. Except as required by law, this report and the information and data contained herein are to be treated as confidential and may be used and relied upon only by the client, its officers and employees. Golder denies any liability whatsoever to other parties who may obtain access to this report for any injury, loss or damage suffered by such parties arising from their use of, or reliance upon, this report or any of its contents without the express written consent of Golder and the client. c:\users\kdecastro\desktop\wolf\09-1326-1040_final_hydraulic model, v2.docx March 2010 Report No. 09-1326-1040 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT APPENDIX A Naturalized Daily Flows and Flood Routing at Key Locations March 2010 Report No. 09-1326-1040 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT Table of Contents A1 AVAILABILITY OF INPUT DATA ...... 1 A2 PROJECT DEPLETION METHOD ...... 1 A3 GENERATION OF DAILY NATURALIZED FLOWS AT SPRAY LAKES ...... 5 A4 GENERATION OF DAILY NATURALIZED FLOWS AT LAKE MINNEWANKA ...... 9 A5 GENERATION OF DAILY NATURALIZED FLOWS AT THE UPPER AND LOWER KANANASKIS LAKES ...... 11 A6 GENERATION OF DAILY NATURALIZED FLOWS AT THE BARRIER LAKE ...... 14 A7 GENERATION OF DAILY NATURALIZED FLOWS AT GHOST RESERVOIR ...... 18 A8 GENERATION OF DAILY NATURALIZED FLOWS AT BEARSPAW RESERVOIR ...... 21 A9 GENERATION OF DAILY NATURALIZED FLOWS AT GLENMORE RESERVOIR ...... 23 A10 ROUTING OF DESIGN HYDROGROGRAPHS ...... 25 A11 FLOOD ROUTING AT GLENMORE RESERVOIR ...... 25 A12 FLOOD ROUTING THROUGH THE UPPER BOW SYSTEM ...... 26 TABLES Table A1: Summary of Daily Naturalized Peak Flow Estimates for Spray River at Banff (05BC001) ...... 8 Table A2: Summary of Daily Naturalized Peak Flow Estimates for Cascade River at Lake Minnewanka ...... 10 Table A3: Summary of Daily Naturalized Peak Flow Estimates for Kananaskis River at Interlakes and Pocaterra ...... 13 Table A4: Summary of Daily Naturalized Peak Flow Estimates for Barrier Lake ...... 17 Table A5: Summary of Daily Naturalized Peak Flow Estimates for Ghost Reservoir ...... 20 Table A6: Summary of Daily Naturalized Peak Flow Estimates for Bearspaw Reservoir ...... 22 Table A7: Summary of Daily Naturalized Peak Flow Estimates for Glenmore Reservoir ...... 24 Table A8: Summary of Flood Routing Results for the Glenmore Reservoir ...... 26 Table A9: Maximum Flows through Hydro Power Plants ...... 27 Table A10: Peak Flows of Routed Design Flood Hydrographs ...... 29 March 2010 Report No. 09-1326-1040 Ai HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT FIGURES Figure A1: Sample Schematic for Calculation of Naturalized Flows ...... 2 Figure A2: Spray Lakes and Hydro Power Diversion ...... 6 Figure A3: Interlakes Peak Flow vs Total (Interlakes and Pocaterra) Peak Flow ...... 12 Figure A4: Pocaterra Local Inflow vs Total (Interlakes and Pocaterra) Natural Flow ...... 12 Figure A5: Approximate Conversion of Mean Weekly Flows to Mean Daily Flows ...... 15 Figure A6: Annual Peak Flows at Barrier Lake pre and post 1943 ...... 16 Figure A7: NFCP Modelling Schematic ...... 19 Figure A8: Key and Target Elevations for a Typical Upper Bow Storage Reservoir ...... 28 Figure A9: Upper Bow River Basin Modelling Schematic ...... 30 March 2010 Report No. 09-1326-1040 Aii HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT A1 AVAILABILITY OF INPUT DATA The process of developing naturalized flow estimates requires the use of historic records of reservoir levels and outflows from the date each structure started its operation. It was initially assumed by Golder that the necessary time series of daily data were available and could be obtained from Alberta Environment (AENV). However, the assumption of complete data series being available was not correct. Both Alberta Environment’s Hydrol database as well as Water Survey of Canada’s data series that were of interest in this study had missing records that had to be filled to enable assessment of daily natural flows. The remaining sections within Appendix A provide a more detailed description of the methodology for calculating naturalized flows as well as how this methodology was implemented at each location of interest given the available data. It should be noted that the primary purpose of developing naturalized flows was to properly assess incoming annual peak flows into all major structures, and for as many years as possible. Naturalized annual peak flows constitute a principal input into the hydrologic frequency analyses, which is a key step in the development of design flood hydrographs with target return periods in the upper Bow River and its tributaries. Therefore, where seasonal flow data were available (i.e. data from April to October), they could still be used to help assess the annual maximum flows. Annual peak flows were assessed for as many years as possible, as further explained in subsequent sections that provide more information about each storage site where naturalized flow series were developed. A2 PROJECT DEPLETION METHOD Alberta Environment uses the Project Depletion Method to calculate naturalized flows on all major rivers in Alberta. The same methodology is employed by the Prairie Provinces Water Board (PPWB). The PPWB consists of representatives from Environment Canada (representing the Federal Government) and the representatives of the three Prairie Provinces. The short summary that explains the project depletion method in this section follows closely the documentation of the Natural Flow Computation Program (NFCP) program used in this study. Technical specifications for NFCP were approved by the PPWB in November of 2008. Natural flows are river flows that would have been observed at selected locations in a river basin assuming there had been no human intervention by operation of large storage reservoirs or withdrawals. The most common approach to estimate naturalized flows from regulated flows is the Project Depletion Method, which is essentially aimed at “undoing” the impacts of human intervention in a systematic way, reach by reach, in a downstream progression. March 2010 Report No. 09-1326-1040 A1 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT The calculation procedure is explained below for a small example shown in Figure A1 that contains all elements found in complex river basins. There are two river reaches with a reservoir R1 at their confluence. In this example naturalized flow is calculated at the reservoir site. There is one diversion into the reservoir (D1) and one return flow (RT1) into the reservoir, one diversion channel out of the reservoir (D2), and regulated outflow from the reservoir into natural channel reach C3. The general approach to calculate naturalized flows at any location is to estimate local runoff which originates between the given location and the closest upstream locations at which natural flows had already been evaluated. Denote the naturalized flow at reservoir as QR1 and the local runoff between natural flows Q1, Q2 and the reservoir as LR. The naturalized flow at the reservoir site can then be calculated as: QR1 = Q1 + Q2 + LR (1) I 1 D2 RT1 Q1 NE1 1 C1 C3 R1 C2 D1 LEGEND: Reservoir Q2 River Reach 2 Diversion Channel Return Flow Channel Consumptive Use Figure A1: Sample Schematic for Calculation of Naturalized Flows Consequently, the principal component of deriving naturalized flows is determination of the local runoff LR. Assuming Qr1 and Qr2 are the recorded flows at locations 1 and 2, LR for the reservoir in Figure 1 can be calculated using the following equation assuming average flow over time step t: LR = QC3 + QD2 – QRT1 – QD1 + ∆V/t – Qr1 – Qr2 (2) where: QC3 the recorded flow in channel C3 QD1 flow in diversion channel D1 QD2 flow in diversion channel D2 QRT1 flow in return flow channel RT1 ∆V/t reservoir storage change over time step t March 2010 Report No. 09-1326-1040 A2 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT Reservoir storage change is further evaluated using the starting and ending storage (Vs and Ve) for a time step, along with adjustments for net evaporation (evaporation minus precipitation) for a given time interval t. Note that the sign for net evaporation is reversed since the consideration is to remove the effect of net evaporation (i.e. put the evaporation loss back in the river since this loss would not have happened if the reservoir had not been built): ∆V Ve −Vs (E − P)[A(Ve) + A(Vs)] = + (3) t t 2t where: 3 Ve volume at the end of time step t (m ) 3 Vs volume at the start of time step t (m ) P total precipitation over time step t (m) E total evaporation from the reservoir surface over time step t (m) 2 A(Ve) surface area (m ) corresponding to the ending volume Ve 2 A(Vs) surface area (m ) corresponding to the starting volume Vs To summarize, local runoff LR can in general be assessed by conducting a water balance calculation for a sub- catchment which is delineated by the downstream point for which LR is evaluated and the upstream control points where recorded flow series are available. The general expression is: m n l ∆Vk LR = ∑Qi − ∑Qj + ∑ (4) i=1 j =1 k =1 t where: Qi average outflows (i=1,m) from a sub catchment within time step t Qj average inflows (i=1,m) into a sub catchment within time step t while the storage change term∆V/t is summed up over all storage reservoirs in the sub -catchment area under consideration. Inflows and outflows into a sub-catchment include all diversions and return flows into it, as well as diversions out of it. Normally, naturalized flows should be calculated at all on-stream reservoir locations, especially when reservoirs have sizeable live storage. Equation (1) suggests that naturalized flows be first determined at upstream locations (e.g. locations 1 and 2 in the example in Figure A1). The calculation then proceeds in the above manner for all requested locations in the river basin in a downstream progression. It should be noted that for short (e.g. daily) time steps, the length of river reaches along channels C1 and C2 may required the use of channel routing, such that the routed outflow March 2010 Report No. 09-1326-1040 A3 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT from these channels takes part in the mass balance calculation at the reservoir node, both for calculating local runoff LR, which would require the routing of recorded flows along these channels, and for calculating the naturalized flow, which would require routing of the naturalized flow estimates previously made at nodes 1 and 2. Alberta Environment uses the Wilson’s routing equation which was built into the Streamflow Synthesis and Reservoir Regulation (SSARR) model. A brief description of Wilson’s equation is provided below. As with the other river routing methods, the governing equation is related to channel storage change over a time step, which is a function of average inflow and outflow: It − 1 + It Ot − 1 + Ot ∆S − = (5) 2 2 t By subtracting both sides of the above equation with Ot-1 , multiplying by t/(Ot-Ot-1) and by letting ∆S/(Ot-Ot-1) = TS, the above equation becomes: It − 1 + It − Ot − 1 ⋅t 2 Ot = + Ot − 1 (6) t TS + 2 where the term TS represents the average travel time along a river reach for given flow conditions, evaluated either by reading from the TS vs Q table or by using a functional form of the travel time vs flow curve as: Kts = (7) TS n Ot − 1 + Ot 2 The routing coefficients Kts and n must previously be determined by finding the best fit curve for a given set of the available (Ts,Q) coordinates. Usually, Ts can be determined for any given flow rate by linear interpolation from a table of (Ts,Q) points (these tables were provided by Alberta Environment and used in this project). In the above definition of Ts, the base of the denominator shown below represents the average channel flow over a time step as the arithmetic average of the outflows at the beginning and the end of the time step: Ot − 1 + Ot 2 Various implementations of the SSARR method may rely on different estimates of the average channel flow during a given time step (which may also include inflows into the channel in some form). The method relies on the established empirical relationship between the travel time and flow for a given channel. Once this relationship is available, the calibration consists of deciding how many sequential phases a given river reach should be divided into, which is conducted using repeated simulation trials until the observed downstream hydrograph closely matches the simulated channel outflow. All work on calibration of the SSARR method in the South Saskatchewan River basin had already been done by Alberta Environment. The upper Bow River Basin schematic that was obtained from Alberta already has the channels broken into lengths that work as single phase channels (i.e. no further subdivision is required). March 2010 Report No. 09-1326-1040 A4 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT The use of the SSARR channel routing method is optional (i.e., naturalized flows can be calculated with or without channel routing), since calculating naturalized flows using sufficiently long (monthly, seasonal or annual) time steps does not require channel routing. In this study, the time step was daily, while the travel time between the most upstream point of interest (Banff) and the most downstream point at Bearspaw Reservoir is well above one day, indicating that channel routing is required. The use of the SSARR routing method in the calculation of naturalized flows does not change the methodology. Instead, it merely introduces a more realistic account of flow changes in large river systems where the total travel time is greater than the simulation time step. It should also be noted that the SSARR river routing is as accurate as the available input data. As any other differential equation, a coarse time step and a large variation between the flows in two subsequent time steps may jeopardize the accuracy of the results. It should be noted that the Wilson’s equation deals with the transformation of surface water movement aimed to account for the channel storage change from day to day. However, large channel storage changes may happen in the Bow River during the formation of ice cover in November, and its subsequent melting in April. These transformations of water into ice and back into liquid can also be understood movements of flow into and out of storage, but such channel storage changes are not modelled by the Wilson’s equation. A3 GENERATION OF DAILY NATURALIZED FLOWS AT SPRAY LAKES Figure A2 shows the Spray Lakes and diversion to the three downstream power plants that eventually discharge into the Bow River at Canmore. The Spray River drains in the northerly direction and joins the Bow River at Banff. Flow monitoring station 05BC001 is located just upstream of the confluence. The Spray Lakes is a naturally occurring lake, which was dyked off to divert water through three hydro power plants over a shorter distance and a large drop in elevation. The diversion route is indicated in orange color in Figure A2. The total catchment area of Spray River at Banff is approximately 751 km2, of which about 520 km2 drain into Spray Lakes. Although the remaining catchment area downstream of the Lake is roughly one third of the total, the upstream runoff into Spray Lakes is considerably higher due to higher water yield, such that roughly 85% of the total annual natural flow of Spray River at Banff (station 05BC001) originates upstream of the dam. March 2010 Report No. 09-1326-1040 A5 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT Figure A2: Spray Lakes and Hydro Power Diversion Proper calculation of naturalized flows at Spray Lakes requires the historic daily lake levels as well as all daily outflows from the lake. The old Spray River channel still collects the runoff downstream of the lake, but after construction of the dam in 1949 it also serves as a potential spillway conduit to evacuate large floods that exceed the capacity of the diversion route. The spillway into the Spray River has historically been used only once in 1974. There are several issues associated with assessing daily natural flows at the dam site: a) Daily diversion flows for power generation from Spray Lakes were only available after 1975, which makes it impossible to properly assess natural flows at the dam site between 1949 and 1975; b) Between 1918 and 1939 peak flows at the dam site were based on recorded flows at station 05BC002. This station is now submerged by the lake. Its original drainage area was 360 km2, which was increased by a factor of 1.3 to account for additional catchment area into the lake that was created by the dam (520 km2) less the adjustment for the considerable increase in the lake surface area which is handled by net evaporation directly in the water balance equation. Regression equation Y=0.77X+10.543 assessed on the basis of the available data for both series has a good fit (R2 = 0.86); March 2010 Report No. 09-1326-1040 A6 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT c) For periods from 1912 to 1917 and 1940 to 1948, peak flows were generated based on the relationship established between the natural Spray River flows at Banff 05BC001 (drainage area 750.6 km2) and the peak inflows into the Spray Lakes from 1918 to 1939 from records available at station 05BC002, which is expressed by the regression equation mentioned under b) above; and d) For the 1949 to 1975 period, both peak flows and continuous daily flows were estimated based on inferential relationship developed between the recorded flows of Spray River at Banff and the calculated natural inflows into the Spray Lakes after 1975. Different models were used for assessing annual peak flows and continuous time series, due to the difficulties of relating the reduced flows at Banff to the flows at the dam for the available period after 1975. These models are further discussed below. Infill of daily continuous naturalized flow estimates for Spray River at Banff is required to enable proper flood routing from Banff all the way to Ghost Reservoir, as well as to allow more accurate calculation of naturalized flows at Ghost reservoir. The model used for the development of continuous daily naturalized flows at Banff for the 1949 – 1975 period was calibrated to meet the two following objectives: a) Fit the regression between the recorded Spray River at Banff after 1975 (with reduced catchment area) and the sum of natural flows at the dam and flows at Banff for the same period (R2 = 0.923) ; and, b) Make sure that the resulting flow estimates also follow the same statistical distribution and have similar mean and standard deviation as the available estimates of natural flow at 05BC001 for the periods with available data. The above two objectives were met. However the resulting annual peak flows from this model were not distributed according to the same frequency that was found in the years of available data. It was felt that if the peak flows from this series were used as input into the frequency analyses, they may reduce the anticipated design peak flows at the dam, which would unnecessarily bias the estimates towards lower values. This concern was also driven by comparison of the flows at 05BC001 for the 1949 – 1975 period with the flows at the same site for 1975 – 2008 period. Recorded peak flows at 05BC001 between 1949 and 1975 are higher on average than the flows encountered in the post 1975 period, as shown in Table A1. Also, daily flows at Banff and at the dam site may differ in the range of 2 to 10 times with a high variance. Although the in-filling of daily flow series at 05BC001 was completed for the 1949 – 1975 period, it is suggested that future updates to the database of naturalized flows provided in this study include retrieval of the recorded outflows from the dam from Transalta, in order to allow proper evaluation of daily natural inflows into the Spray Lakes. March 2010 Report No. 09-1326-1040 A7 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT Table A1: Summary of Daily Naturalized Peak Flow Estimates for Spray River at Banff (05BC001) Peak Flow at Peak Flow at Peak Peak Flow Peak Flows Year the Dam Year the Dam Flows at Year at the Dam 3 3 3 at 05BC001 (m /s) (m /s) 05BC001 (m /s) 1912 65.72 1949 143.81 39.90 1976 62.31 9.17 1913 75.12 1950 84.97 19.90 1977 47.55 8.50 1914 76.90 1951 190.59 55.80 1978 72.11 11.20 1915 60.71 1952 68.49 14.30 1979 43.31 8.09 1916 108.42 1953 88.20 21.00 1980 59.35 11.40 1917 62.72 1954 83.79 19.50 1981 83.04 15.70 1918 96.86 1955 73.49 16.00 1982 70.50 11.00 1919 74.74 1956 69.96 14.80 1983 58.17 7.03 1920 76.90 1957 52.16 8.75 1984 65.80 9.01 1921 63.10 1958 49.84 7.96 1985 149.65 8.72 1922 57.71 1959 56.72 10.30 1986 102.21 17.40 1923 111.50 1960 56.13 10.10 1987 43.47 7.64 1924 52.85 1961 89.38 21.40 1988 63.62 10.30 1925 65.11 1962 48.33 7.45 1989 64.81 7.79 1926 43.68 1963 49.84 7.96 1990 75.18 17.50 1927 78.59 1964 85.26 20.00 1991 71.82 12.30 1928 73.81 1965 83.49 19.40 1992 52.43 10.10 1929 64.26 1966 76.43 17.00 1993 52.91 9.09 1930 57.24 1967 72.90 15.80 1994 44.79 6.40 1931 52.70 1968 48.66 7.56 1995 108.95 18.80 1932 120.75 1969 65.25 13.20 1996 84.02 14.90 1933 130.76 1970 70.25 14.90 1997 72.47 9.35 1934 73.20 1971 68.19 14.20 1998 58.82 9.99 1935 57.86 1972 155.58 43.90 1999 71.90 10.40 1936 62.72 1973 67.02 13.80 2000 47.24 6.95 1937 48.92 1974 114.68 30.00 2001 39.77 8.36 1938 70.35 1975 48.83 7.62 2002 93.78 19.40 1939 51.39 2003 53.10 12.10 1940 48.07 2004 57.61 10.30 1941 45.22 2005 70.31 23.50 1942 56.78 2006 73.10 11.50 1943 61.79 2007 95.24 17.70 1944 35.20 2008 58.08 11.60 1945 55.93 1946 70.96 1947 53.78 1948 101.48 25 percentile 55.93 56.43 10.20 53.10 8.72 Median 63.10 70.25 14.90 64.81 10.30 75 percentile 75.12 85.11 19.95 73.10 12.30 March 2010 Report No. 09-1326-1040 A8 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT Annual peak flow estimates at the Spray Lakes were based on the regression established for the post-1975 period between the recorded flows at Banff (station 05BC001) and the sum of these flows with the naturalized flows at the dam site, which represent the equivalent of the natural flows at Banff. This regression equation is Y = 2.942X + 26.415, however the regression fit is not as consistent (R2 = 0.59), and the regression line slope is subject to significant changes if selected outliers are removed. Table A1 provides a listing of all peak flows for the three distinct periods (prior to 1949, 1949 – 1975, and post 1975). Station 05BC001 shows the Spray River flows that were measured after the dam was constructed. As mentioned earlier, the peak flows at this station were considerably higher in the 1949 to 1975 period than in the post 1975 period. This justifies higher annual peak flows at the dam estimated for the same period. It should also be noted that Mud Lake diversion, which transfers some of the runoff that would naturally occur into the Lower Kananaskis Lake, was included in the estimates of naturalized flows at the dam since it is functions as a permanent modification to the watershed. The Mud Lake diversion would continue to operate during floods. A4 GENERATION OF DAILY NATURALIZED FLOWS AT LAKE MINNEWANKA Lake Minnewanka is also a naturally occurring lake that was raised to increase its storage. In addition to this, about 15% of the total runoff into Lake Minnewanka comes from the Ghost River catchment via the Ghost River diversion. This diversion was built in 1941 and a flow monitoring station has operated until 1994. Although the station has been discontinued, this diversion continues to operate, and it would also operate during floods. Consequently, in this study Ghost diversion was considered as integral part of runoff into Lake Minnewanka for calculation of naturalized flows into the lake. Naturalized flows into Lake Minnewanka were estimated based on its recorded outflows available from Water Survey of Canada stations 05BD002 until 1941 and from station 05BD004 from 1942. The Lake elevations (WSC station 05BD003) are available from 1917. There are many days with missing data prior to 1943, when the continuous daily records began. However, the records between 1917 and 1943 are usually available at least once or twice a week, which allowed the use of linear interpolation as the first approximation for infilling the missing data, particularly since this lake is large and the levels do not vary substantially from day to day. Until 1941, the dam was operated in the range between 1450 and 1455 m, with flow regulation having a relatively small impact on the downstream Bow River flows at Ghost ad Bearspaw Dams The mean annual flow of Cascade River at the mouth is 8.4 m3/s, compared to about 90 m3/s for the Bow River at Bearspaw dam. Effects of flow regulation are more pronounced after 1942, when the lake levels were raised to operate in the range between 1465 and 1475 m and Ghost diversion started to operate. Table A2 shows a comparison of daily peak flows at Lake Minnewanka for three periods (prior to 1943, 1943 – 1975, and after 1975). March 2010 Report No. 09-1326-1040 A9 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT Table A2: Summary of Daily Naturalized Peak Flow Estimates for Cascade River at Lake Minnewanka Peak Flow at the Peak Flow at the Peak Flow at the Year 3 Year 3 Year 3 Dam (m /s) Dam (m /s) Dam (m /s) 1917 89.03 1943 53.14 1976 40.63 1918 72.60 1944 37.92 1977 38.13 1919 155.04 1945 51.00 1978 51.20 1920 105.85 1946 47.83 1979 29.77 1921 53.56 1947 47.44 1980 58.75 1922 39.73 1948 130.12 1981 88.97 1923 123.85 1949 29.15 1982 48.10 1924 53.57 1950 87.09 1983 38.07 1925 36.61 1951 78.05 1984 36.93 1926 39.03 1952 71.54 1985 29.58 1927 50.03 1953 130.07 1986 75.52 1928 65.21 1954 72.08 1987 27.75 1929 69.07 1955 59.38 1988 48.42 1930 60.10 1956 63.75 1989 38.30 1931 34.45 1957 46.51 1990 94.39 1932 115.91 1958 44.45 1991 61.32 1933 84.89 1959 47.84 1992 40.88 1934 39.66 1960 40.94 1993 38.58 1935 50.11 1961 61.53 1994 32.53 1936 38.11 1962 51.49 1995 102.27 1937 30.57 1963 52.92 1996 60.33 1938 47.32 1964 70.60 1997 43.59 1939 46.81 1965 122.47 1998 49.55 1940 30.63 1966 62.50 1999 44.98 1941 18.06 1967 73.47 2000 24.43 1942 53.34 1968 37.96 2001 41.27 1969 82.98 2002 58.83 1970 88.84 2003 49.24 1971 77.43 2004 53.83 1972 77.42 2005 80.06 1973 87.94 2006 32.85 1974 105.01 2007 85.16 1975 36.73 2008 43.62 25 percentile 39.19 47.63 38.13 Median 51.72 62.50 44.98 75 percentile 71.72 72.77 58.83 March 2010 Report No. 09-1326-1040 A10 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT Table A2 shows that expected peak naturalized flows were higher in the 1943-1975 period than in the post 1975 period. There is presently no plausible explanation for this statistical discrepancy, but it can be noted that the diversions from the Ghost River were higher in the period prior to 1980s, and that the more recent modification of the diversion weir restricts the diverted flows to the maximum of 400 cfs (11.33 m3/s), which did not exist in the earlier period between 1943 and 1975. A5 GENERATION OF DAILY NATURALIZED FLOWS AT THE UPPER AND LOWER KANANASKIS LAKES Upper and Lower Kananaskis Lakes are also naturally occurring lakes that were raised to higher elevations to provide more balancing storage and head for hydro power generation. The upper Kananaskis Lake receives runoff from the catchment area of about 150 km2 and it drains into the lower Kananaskis Lake, which in addition to receiving outflow from the Upper Kananaskis Lake also receives inflow from a local catchment area of roughly the same size (151 km2), although some of the runoff originating from this area is diverted into the Spray Lakes via the Mud Lake diversion. The Lower Kananaskis Lake eventually drains into the Barrier Lake, located some 50 km downstream of it. The Upper Kananaskis Dam had its normal full supply level raised in 1942 by about 14 m. The Lower Kananaskis Dam was built in 1955 with an increase in full supply elevation by 11 m and allowed annual fluctuation of water levels by 14 m. Prior to 1942, both lakes have been operated as natural water bodies. The hydro power plants associated with these lakes are known as Interlakes and Pocaterra. The following procedure is needed to calculate naturalized inflows into both lakes: a) Use the Upper Kananaskis storage levels and outflow records to remove the effect the storage and calculate naturalized flows at the Upper Kananaskis Lake; and b) Use the Lower Kananaskis storage levels, outflows and the outflow from the Upper Kananaskis as a part of the total inflow to assess the net runoff from the local contributing catchment into the Lower Kananaskis Lake. The data available for full implementation of the above procedure are only available for the 1985 to 2007 for both reservoirs. Water levels at both lakes and the outflows from the Lower Kananaskis Lake are also available for 1975-1985 period, but the turbine flows out of the Upper Kananaskis Lake are missing. It was decided to naturalize flows from 1975 to 1985 for the combined Upper and Lower Kananaskis sub-catchments by using the storage change at both reservoirs and outflows from the Lower Kananaskis Lake. Daily peak flows for the years with missing data were developed based on the regression established between total naturalized flow for both catchments and each individual catchment for the 1985 – 2007 data. This approach provided 10 more years of data (1975-1984) that can provide input into statistical frequency analyses. The resulting regression fit and equations are shown in Figure A3 and A4. Table A3 shows the annual maximum daily naturalized flow estimates for the Upper and Lower Kananaskis Lakes (indicated as Interlakes and Pocaterra in Figures A3 and A4). March 2010 Report No. 09-1326-1040 A11 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT 40 y = 0.50397x + 5.25924 R² = 0.67288 35 30 25 20 15 Annual Max. Flow at Interlakes 10 20 25 30 35 40 45 50 55 60 Joint Annual Max. Flow (m3/s) Figure A3: Interlakes Peak Flow vs Total (Interlakes and Pocaterra) Peak Flow 35 y = 0.46612x + 1.24776 R² = 0.74173 30 25 20 Annual Max. at Pocaterra 15 10 20 25 30 35 40 45 50 55 60 Joint Annual Max. Flow Figure A4: Pocaterra Local Inflow vs Total (Interlakes and Pocaterra) Natural Flow March 2010 Report No. 09-1326-1040 A12 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT Table A3: Summary of Daily Naturalized Peak Flow Estimates for Kananaskis River at Interlakes and Pocaterra Joint Annual Max Flow Estimated Annual Max Estimated Annual Max Year below Pocaterra (includes Local Inflow into Flow above Interlakes total catchment) Pocaterra 1975 35.54 23.17 17.81 1976 26.68 18.70 13.68 1977 30.26 20.50 15.35 1978 37.56 24.19 18.76 1979 38.86 24.84 19.36 1980 34.37 22.58 17.27 1981 41.33 26.08 20.51 1982 37.92 24.37 18.92 1983 31.51 21.13 15.93 1984 33.33 22.05 16.79 1985 27.63 17.75 13.57 1986 52.57 37.78 27.02 1987 46.87 33.65 13.70 1988 37.11 25.22 19.29 1989 31.89 20.87 14.47 1990 39.03 26.08 19.03 1991 38.59 28.46 20.20 1992 33.95 25.72 17.79 1993 29.75 19.25 16.37 1994 26.90 20.99 12.68 1995 49.27 22.61 26.65 1996 45.04 27.17 22.05 1997 35.46 20.48 16.66 1998 34.78 25.05 19.15 1999 32.79 21.64 17.90 2000 25.44 16.48 13.62 2001 22.04 16.48 11.94 2002 54.70 29.43 29.83 2003 29.67 20.13 13.60 2004 29.65 19.19 16.67 2005 32.51 18.61 16.00 2006 36.29 25.56 19.26 2007 41.92 22.59 19.91 2008 14.64 – 14.64 The total runoff in 2008 was comparatively low compared to the other years. Consequently, no attempt was made to estimate its portion as inflow into the Upper Kananaskis Lake. A total of 34 years of data is not a reliable input for frequency analyses as far as the length of the series is concerned, but these are all the data that are available at this point. Flows from Interlakes to Pocaterra are readily available from Transalta in electronic format only for the 1985 – 2007 period, while data for years prior to 1985 are only available on microfiche. March 2010 Report No. 09-1326-1040 A13 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT A6 GENERATION OF DAILY NATURALIZED FLOWS AT THE BARRIER LAKE Proper estimation of daily naturalized flows at the Barrier dam requires the following information: a) Estimated daily naturalized flows at both the Upper and Lower Kananaskis Lake. This was only available for the 1985 – 2007 period. b) Continuous daily recorded outflows from the Lower Kananaskis Lake (Pocaterra). Some data are available prior to 1955 but mainly as seasonal records, and there is a large section of data completely missing from 1955 to 1975, when the continuous records begin. c) Calibrated travel time vs flow relationship for the 50 km Kananaskis River reach between Pocaterra and Barrier to allow proper channel routing. This information is not currently available. Calculation of actual flows conducted by Alberta Environment’s model does not use any channel routing on this reach. d) Historic daily elevations of Barrier Reservoir. These have been made available by Alberta Environment from 1948 to 1988, and they are also available from Water Survey of Canada as a continuous record from 1970 to 2008 (station 05BF024). These two datasets do not always agree well in the overlapping period from 1970 to 1988. e) Historic daily outflows from the Barrier Reservoir. Genuine turbine flows should be available from Transalta, however these are only available from 1985 to 2007. For periods prior to 1985, there are two Water Survey of Canada stations. Station 05BF025 is located immediately downstream of the dam, with a total catchment area of 899 km2 and a continuous flow record from 1975 to 2008 (which eliminates the need to use Transalta’s data), while station 05BF001 is located further downstream and includes a slightly larger catchment of 933 km2 due to a small tributary upstream of it. The data record at this station is available from 1911 to 1962, but there were many missing data gaps that required in-filling, including the complete blackout period from 1963 to 1974 inclusive. Alberta Environment maintains a weekly natural flow database, which contains mean weekly outflows from the Barrier dam from 1948 to 2001. The mean weekly recorded flow data were used as a basis for infilling the missing Barrier outflows, using the linear interpolation model the middle of one week to the middle of the subsequent week, as shown in Figure A5. The model shown in Figure A5 is considered acceptable for providing continuous estimates of daily flows from the Kananaskis River in the 1963 – 1974 period, and for in-filling occasional missing data prior to 1963, since the continuous stream records are required for estimation of naturalized inflow hydrographs at Ghost and Bearspaw reservoirs. Although daily flow estimates obtained in this way were used to remove the effect of the Barrier Reservoir storage, the resulting flows were not considered appropriate for estimation of peak daily flows at the Barrier Reservoir site, because this model obviously underestimates annual daily peak flows which are never higher than the annual weekly peak flows. Therefore, the 1963 – 1974 (inclusive) period was not used in the assessment of peak annual flows. March 2010 Report No. 09-1326-1040 A14 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT 13 12 11 /s) 3 10 9 Flow (m 8 Mean Weekly Flows 7 Daily Flow Estimates 6 56 61 66 71 76 81 Time (days) Figure A5: Approximate Conversion of Mean Weekly Flows to Mean Daily Flows The Barrier dam was built in 1948, before raising the Pocaterra levels in 1955. Prior to 1948, the additional raising of the Upper Kananaskis Lake level starting in August of 1942. Hence, there was a reason to suspect that the annual peak flows at the Barrier dam site exhibited statistical difference before and after 1943, and especially after 1955 due to additional inclusion of the balancing effect of the Lower Kananaskis Lake. To investigate this assumption, the annul peak flows at the Barrier Reservoir site were separated into two series (pre and post 1943) and plotted using the probability plot with the standard Weibul plotting position formula [r/(n+1), where r is the sequential number of data point in the sorted order, while n is the total number of data points]. The resulting plot is in Figure A6. March 2010 Report No. 09-1326-1040 A15 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT 300 1912-1942 1943-2008 250 200 150 100 Annual Peak Peak Daily Annual Flow (m3/s) 50 0 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Probability of Occurrence Figure A6: Annual Peak Flows at Barrier Lake pre and post 1943 It is obvious that the peak flows encountered after 1942 are smaller. For example, the median peak flow in the post 1943 period of 50 m3/s corresponds to the median peak flow of about 70 m3/s in the pre 1943 period. This model was further refined by using the moving average for both of the above plots, which were then fitted with higher order polynomials to define a functional relationship for each curve. The post 1943 naturalized peak flows obtained by removing the effect of the Barrier Lake were then adjusted to include increases that correspond to the difference between the two probability curves shown above. The actual differences that were added were depended on the magnitude of the post 1943 flood. For very small peak flows, the added adjustments were also small and in a few cases of very small peak flows they were left unchanged. Table A4 shows the resulting annual peak flow estimates for pre and post 1943 period, excluding the 1963-1974 period for reasons outlined above. March 2010 Report No. 09-1326-1040 A16 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT Table A4: Summary of Daily Naturalized Peak Flow Estimates for Barrier Lake Peak Flow at the Initial Estimates of Adjusted Estimates Year Dam - Natural Year Peak Flows at the of Peak Flows at the Conditions (m3/s) Dam (m3/s) Dam (m3/s) 1912 92.30 1943 49.30 71.35 1913 60.90 1944 26.60 32.69 1914 67.10 1945 62.90 85.39 1915 152.00 1946 66.50 93.31 1916 209.00 1947 50.10 70.64 1917 85.80 1948 147.61 228.67 1918 93.40 1949 104.21 182.25 1919 75.00 1950 95.46 159.24 1920 71.40 1951 97.78 166.92 1921 68.00 1952 57.17 77.43 1922 65.70 1953 106.17 188.23 1923 144.00 1954 90.93 143.57 1924 75.90 1955 61.68 82.82 1925 68.00 1956 44.65 71.28 1926 48.70 1957 32.09 52.26 1927 103.00 1958 44.66 71.15 1928 82.10 1959 46.08 71.50 1929 146.00 1960 47.84 71.66 1930 69.70 1961 65.38 89.75 1931 45.90 1962 103.09 177.09 1932 283.00 1975 44.63 71.40 1933 111.00 1976 35.82 60.79 1934 55.80 1977 31.41 46.81 1935 49.30 1978 38.03 64.35 1936 63.40 1979 31.82 50.47 1937 43.00 1980 49.98 71.22 1938 82.70 1981 75.96 118.14 1939 41.30 1982 37.15 62.90 1940 36.50 1983 30.18 43.87 1941 21.70 1984 24.91 28.74 1942 66.30 1985 31.63 48.69 1986 68.89 98.71 1987 28.05 38.13 1988 32.94 56.94 1989 91.61 149.82 1990 71.98 109.52 1991 45.91 71.95 1992 40.42 67.25 1993 38.56 65.25 1994 24.05 25.34 1995 105.76 186.63 1996 46.27 70.93 1997 32.29 53.86 1998 52.83 72.86 1999 26.88 35.03 March 2010 Report No. 09-1326-1040 A17 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT Peak Flow at the Initial Estimates of Adjusted Estimates Year Dam - Natural Year Peak Flows at the of Peak Flows at the Conditions (m3/s) Dam (m3/s) Dam (m3/s) 2000 17.92 17.92 2001 21.05 21.05 2002 69.31 102.72 2003 32.57 55.44 2004 29.50 41.42 2005 88.12 135.37 2006 48.38 71.32 2007 54.19 73.96 2008 55.96 75.79 25 percentile 58.35 32.36 54.26 Median 69.70 47.05 71.34 75 percentile 92.85 68.29 97.36 A7 GENERATION OF DAILY NATURALIZED FLOWS AT GHOST RESERVOIR The impoundment of Ghost Reservoir started in August of 1929. Prior to 1929, the only possible change to natural flows upstream of Ghost Reservoir would be due to the operation of Lake Minnewanka on Cascade River, which at that time was operating in the lower range between 1450 and 1455 m. Although the Cascade River flows have been adjusted for the effects of Lake Minnewanka storage, the missing link required for estimating natural flows at the Ghost Reservoir site prior to 1929 are the recorded flows in the vicinity of the Ghost Reservoir site. Station 05BE006 (Bow River below Ghost dam) started operation in 1933. The upstream flow monitoring station at Seebe (05BE004) started operation in 1923. The Ghost River tributary (station 05BG001) that also contributes significant flows to the Ghost Reservoir storage has missing data between November 1920 and December 1928. The recorded flow data series that represents recorded outflows (or inflows) into the Ghost Reservoir do not exist prior to 1930. Hence, it was only possible to assess natural flows into Ghost reservoir between 1930 and 2008, following required in-filling of the missing data. When the outflows were missing, the storage change and both inflows (Station 05BE004 at Seebe and the Ghost River tributary station 05BG001) were used to generate the outflow estimates. When the storage levels were missing and both outflows and inflows were available, storage levels were estimated by balancing inflows and outflows and by solving for the end of day storage as the only unknown. A regression was also developed between the Bow River below Ghost Reservoir (05BE006) and the Bow River at Calgary (05BH04), and it was used occasionally for in-filing the missing data when there was no other option. The following information was necessary to estimate daily naturalized flows at the Ghost Reservoir for the 1930 – 2008 period: a) recorded and naturalized flows of the Bow River downstream of confluence with the Spray River; b) recorded and naturalized flows of the Cascade River at Lake Minnewanka; March 2010 Report No. 09-1326-1040 A18 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT c) recorded and naturalized flows of the Kananaskis River at Barrier Reservoir; d) time of travel vs flow for all routing reaches of the Bow river between Banff and the Ghost Reservoir as well as on the Kananaskis river below the Barrier dam; e) end of day Ghost Reservoir elevations; f) recorded Ghost Reservoir outflows; and, g) precipitation and evaporation on Ghost reservoir. The NFCP model was run on a daily basis from 1930 to 2008 to obtain estimates of naturalized flows at the Ghost Reservoir. The model schematic is shown in Figure A7. The calculation procedure starts at the most upstream nodes, and evaluates naturalized flows at each node in a downstream progression. This procedure requires separate routing of recorded flows from the three upstream control points, which are the confluence of the Bow and Spray Rivers, Lake Minnewanka and the Barrier Reservoir downstream to the Ghost Reservoir. These routed flows were used to calculate the local inflow into the Ghost Reservoir by subtracting them from the Ghost Reservoir outflow adjusted for storage change. Once the local inflow is calculated in this manner, it is added to the sum of routed natural flows starting from the same three control points and ending at the Ghost Reservoir. This is a complex procedure that is significantly aided by the existing computer model designed for this purpose. Figure A7: NFCP Modelling Schematic Denote with R recorded flows, N naturalized flows, F( ) the channel routing function, LI local inflow, ∆V / ∆ t storage change and NE net evaporation on the Ghost Reservoir. The calculation of local inflow into the Ghost Reservoir process can then be mathematically expressed as: LI58 = ∆V / ∆ t + NE + R158 – F( R150 + R151 + R155) March 2010 Report No. 09-1326-1040 A19 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT It should be understood that the actual routing of flows happens channel by channel, and that two channels that meet at a node provide the sum of routed flows at their downstream ends which is further routed through the downstream channel. In that sense, the term F( R150 + R151 + R155) schematically represents the result of combined routing of all three control flows through all sequential channels in the schematic, resulting with the routed outflow from channel 157. The naturalized flow is then the sum off the three upstream naturalized flows routed to the Ghost Reservoir site, and adjusted by the evaluated amount of local inflow LI58, hence: N58 = LI58 + F( N150 + N151 + N155) The same considerations are valid regarding the routing of naturalized flows as for the routing of recorded flows (i.e. channel routing must follow the sequence of channels shown in the schematic). There are inaccuracies in this procedure that can be related to many missing values that had to be filled using various regression models, as well as ice formation which resulted in erroneous flow or elevation measurements. In some days in the winter the calculated naturalized flows have negative values, which imply that more work is required to modify the data and rectify these situations. However, the high flow events show reasonable hydrographs which provide insight into the duration and timing of flood events, in addition to the peak flows. Table A5 shows the resulting summary of peak flows obtained from the daily 1930 – 2008 series. It is noted that the first half of the natural flow series (1930 – 1969) has the three benchmark statistics (25 percentile, median and 75 percentile) of flows which are on average 50m3/s higher than the peak flows in the second half of the series from 1970 to 2008. Table A5: Summary of Daily Naturalized Peak Flow Estimates for Ghost Reservoir Peak Flow at the Dam - Peak Flow at the Dam - Year 3 Year 3 Natural Conditions (m /s) Natural Conditions (m /s) 1930 454.05 1970 342.50 1931 312.60 1971 382.37 1932 815.54 1972 444.77 1933 575.13 1973 327.61 1934 430.26 1974 518.37 1935 407.52 1975 245.84 1936 432.32 1976 268.31 1937 281.81 1977 227.32 1938 401.83 1978 297.82 1939 294.47 1979 205.01 1940 301.57 1980 325.70 1941 178.96 1981 474.64 1942 314.27 1982 297.75 1943 345.95 1983 240.63 1944 233.75 1984 270.49 1945 319.06 1985 228.42 1946 385.35 1986 389.92 1947 374.77 1987 221.69 1948 676.29 1988 308.06 1949 204.47 1989 312.45 1950 461.43 1990 527.99 1951 432.15 1991 424.81 1952 378.49 1992 278.70 March 2010 Report No. 09-1326-1040 A20 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT Peak Flow at the Dam - Peak Flow at the Dam - Year 3 Year 3 Natural Conditions (m /s) Natural Conditions (m /s) 1953 536.79 1993 242.75 1954 408.21 1994 238.68 1955 385.58 1995 594.43 1956 399.23 1996 381.74 1957 280.64 1997 333.75 1958 281.89 1998 318.41 1959 313.82 1999 346.16 1960 254.79 2000 213.81 1961 400.40 2001 222.20 1962 269.93 2002 400.07 1963 345.73 2003 261.27 1964 402.83 2004 280.70 1965 582.13 2005 562.16 1966 370.40 2006 306.93 1967 454.49 2007 508.55 1968 260.59 2008 381.59 1969 356.29 25 percentile 299.79 253.56 Median 376.63 312.45 75 percentile 430.73 386.14 A8 GENERATION OF DAILY NATURALIZED FLOWS AT BEARSPAW RESERVOIR Naturalized flows at Bearspaw Reservoir were also generated as part of the NFCP run, as shown in Figure 8 which also includes the Bearspaw Reservoir. Daily reservoir elevations were received from Alberta Environment, and they were also available from Transalta from 1985 to 2007. There are occasionally large differences between the reservoir levels available from those two sources, requiring judgement in the final data selection. Outflows from the Bearspaw Reservoir are available from Transalta only after 1985. There are two Water Survey of Canada stations on the Bow River below the Bearspaw dam. Station 06BH008 is close to the dam, but its record begins in 1983 and it is incomplete. The only other long term flow monitoring station in the relative vicinity of the Bearspaw dam is the Bow River at Calgary (05BH004). This station also needed some in-filling of missing data between 1930 and 2008, which was accomplished by developing a relationship with the Bow River below Ghost Dam (05BE006). The use of the Bow River at Calgary data is not a perfect solution, since there is a time lag of 6 to 8 hours between the Bearspaw dam and the flow monitoring station, but it was the only choice at this point. Bearspaw Reservoir outflows should be obtained from Transalta for the period since the dam started operation. Water Survey of Canada does not monitor Bearspaw reservoir levels. March 2010 Report No. 09-1326-1040 A21 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT Time series of the Bearspaw Reservoir levels provided by Alberta Environment begins in 1955, but this series does not contain the data related to the initial impoundment. Hence, for the period prior to 1955, the levels were kept constant with the evaporation and precipitation set to zero. Calculation of naturalized flows at Bearspaw Reservoir was conducted within the NFCP using the following steps: a) Calculate local inflow between the Ghost and Bearspaw reservoirs as: LI59 = ∆V / ∆ t + NE59 + R162 + D163 – F( R158 ) where R162 is the outflow from the Bearspaw dam into the Bow River, D163 is the diversion by the city of Calgary, and F (R158 ) is the routed recorded outflow from the Ghost Dam. Calculate the natural flows by adding the local inflow LI59 to the routed naturalized flow at Ghost Reservoir: N58 = LI59 + F( N58 ) The resulting naturalized flows show that the local runoff that originates between the Ghost and Bearspaw reservoirs does not significantly increase the natural flows at the Ghost dam. In fact, during large floods, the peak flows are often lower at Bearspaw than at the Ghost dam, since the channel attenuation overcomes the additional flow increases due to local runoff. This is likely caused by the lower channel slope and larger width compared to the river sections upstream of the Ghost dam. Table A6 provides the annual flood inflow series to Bearspaw Reservoir. Table A6: Summary of Daily Naturalized Peak Flow Estimates for Bearspaw Reservoir Year Peak Flow at the Dam Year Peak Flow at the Dam 1930 425.34 1970 413.84 1931 310.59 1971 419.41 1932 1223.32 1972 448.08 1933 619.14 1973 347.09 1934 453.63 1974 526.75 1935 404.68 1975 248.55 1936 419.55 1976 281.69 1937 273.19 1977 244.89 1938 404.24 1978 303.89 1939 300.79 1979 197.90 1940 279.99 1980 320.59 1941 184.87 1981 497.81 1942 335.81 1982 307.66 1943 358.71 1983 245.64 1944 238.36 1984 271.23 1945 348.21 1985 215.05 1946 402.71 1986 486.17 1947 414.36 1987 211.99 1948 687.90 1988 317.70 1949 207.84 1989 301.71 1950 488.20 1990 538.60 1951 431.86 1991 408.45 1952 447.19 1992 331.47 1953 598.72 1993 251.44 1954 446.54 1994 289.70 March 2010 Report No. 09-1326-1040 A22 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT Year Peak Flow at the Dam Year Peak Flow at the Dam 1955 397.85 1995 563.25 1956 397.82 1996 356.65 1957 299.24 1997 329.48 1958 311.97 1998 318.24 1959 326.94 1999 374.76 1960 312.13 2000 211.42 1961 428.04 2001 204.65 1962 287.29 2002 371.55 1963 349.54 2003 264.90 1964 403.69 2004 268.64 1965 521.10 2005 676.17 1966 372.10 2006 314.63 1967 424.11 2007 485.51 1968 281.10 2008 355.27 1969 439.16 25 percentile 311.63 266.77 Median 400.28 318.24 75 percentile 433.68 411.14 A9 GENERATION OF DAILY NATURALIZED FLOWS AT GLENMORE RESERVOIR Glenmore reservoir level data are available from Water Survey of Canada (station 05BJ008), with data records from 1976 to 2008. Additional water level data were obtained from 1933 to 1988 from Alberta Environment. There are two Water Survey of Canada stations downstream of the dam (05BJ005 and 05BJ001). Together they provide continuous data coverage for the 1912 – 2008 period. There is also one more flow monitoring station upstream of the dam (05BJ010), with data available after 1979. The reservoir level data begin abruptly with the reservoir being close to full in 1933, so the actual impoundment of the dam could not be included in the calculation of naturalized flows. Consequently, the reservoir was modelled with fixed elevation (i.e. zero storage change) and zero net evaporation for all years prior to 1933, implying that the recorded flows downstream of the dam was equal to natural flows. Natural flows after 1933 were assessed by using the outflows from the dam adjusted for Glenmore reservoir storage change. The results of the updated bathymetric survey were used in this study to incorporate the latest estimate of the storage capacity curves. Table A7 provides the resulting summary of flood series derived from the daily flow series. March 2010 Report No. 09-1326-1040 A23 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT Table A7: Summary of Daily Naturalized Peak Flow Estimates for Glenmore Reservoir Peak Flow at the Peak Flow at the Peak Flow at the Year 3 Year 3 Year 3 Dam Site (m /s) Dam (m /s) Dam (m /s) 1908 159.00 1933 55.33 1971 83.20 1909 94.00 1934 23.53 1972 42.03 1910 18.60 1935 29.19 1973 44.01 1911 89.50 1936 29.97 1974 62.27 1912 122.00 1937 53.64 1975 45.59 1913 38.80 1938 59.96 1976 38.06 1914 28.90 1939 91.41 1977 15.79 1915 239.00 1940 36.53 1978 45.61 1916 146.00 1941 34.33 1979 33.52 1917 147.00 1942 122.86 1980 52.80 1918 35.40 1943 30.94 1981 87.26 1919 72.50 1944 24.33 1982 36.10 1920 67.70 1945 73.85 1983 30.85 1921 37.40 1946 50.56 1984 21.05 1922 26.50 1947 69.10 1985 54.82 1923 331.00 1948 128.24 1986 53.24 1924 59.50 1949 36.99 1987 29.13 1925 66.50 1950 34.94 1988 32.88 1926 88.10 1951 135.72 1989 21.47 1927 83.30 1952 81.86 1990 121.82 1928 100.00 1953 129.13 1991 49.40 1929 382.00 1954 45.64 1992 108.06 1930 30.60 1955 45.92 1993 81.49 1931 22.90 1956 37.18 1994 60.44 1932 311.00 1957 30.64 1995 217.91 1958 55.17 1996 41.84 1959 47.74 1997 51.92 1960 29.95 1998 96.02 1961 50.35 1999 53.45 1962 30.17 2000 18.06 1963 121.93 2001 42.65 1964 63.98 2002 79.39 1965 102.56 2003 31.19 1966 36.99 2004 36.22 1967 192.37 2005 373.73 1968 50.49 2006 109.63 1969 127.65 2007 78.06 1970 94.74 2008 171.15 25 percentile 37.40 34.64 36.13 Median 83.30 50.53 50.66 75 percentile 146.00 71.48 80.97 Floods recorded prior to 1933 have greater median and 75 percentile values, than the other two periods that were compared (1933 – 1969) and (1970 – 2008) are statistically similar. March 2010 Report No. 09-1326-1040 A24 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT A10 ROUTING OF DESIGN HYDROGROGRAPHS Following development of inflow hydrographs for the various return periods, the next step was to route the hydrographs through the storage structures to account for the effects of storage and evaporation on the flood flows. Routing was conducted using the existing guidelines for operating these structures during floods. In some cases there is allowance for variability to provide flexibility to the operation such as “keep the outflow between 100 and 160 m3/s if inflow is above 100 m3/s”. Setting up a computer model to mimic the behaviour of an operator may therefore involve strengthening of the rules by the modeller, in order to ensure repeatability of simulation runs. The new version of the WRMM model from Alberta Environment was used in this study. This model includes both channel routing using the SSARR routing method, as well as reservoir routing. The routing equations are formulated as constraints to a linear program. The benefit of this model is that it has flexibility to represent various reservoir operating policies, including a mix of rule curves and multiple water use zones and priority factors that drive the simulation as cost parameters in the objective function. A11 FLOOD ROUTING AT GLENMORE RESERVOIR Glenmore reservoir was built in 1933 on the Elbow River to provide water supply for the City. Its original design did not include an allowance for flood storage. However, flood operational guidelines exist, and they provide instructions on the reservoir drawdown and release policies for the bottom outlet and spillway. Considerable judgement is required on the part of the operators in terms of the draw down of the reservoir during a flood event, ranging from 2 m to 4 m. Low level outlet releases are linked to the incoming flows and the achieved drawdown. The maximum low level outlet flows is 165 m3/s. This limit is achieved during the condition where it is necessary to evacuate the incoming flood and simultaneously lower the reservoir level to a desired target drawdown. Once the target drawdown is reached, the outflows will be equal to inflows, and if inflows exceed 165 m3/s the reservoir will begin to fill since the outflow remains set at 165 m3/s. Once the reservoir levels has reached the invert of the spillway, the flood will be evacuated through both the low level outlet and the spillway for the first 1.0 m of flow depth over the dam crest and with a gradual shut down of the bottom outlet. At a water level of 1.0 m above the dam crest, the low level outlet is closed and the spillway continues to discharge the flood as it has reached sufficient head to provide adequate outflow capacity. As part of the model setup, the results of the most recent bathymetric surveys from the City of Calgary were used in this study. Table A8 provides a summary of the initial assumptions, along with the key elevations and outflows obtained from applying the existing rules for routing design hydrographs through the Glenmore Reservoir. March 2010 Report No. 09-1326-1040 A25 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT Table A8: Summary of Flood Routing Results for the Glenmore Reservoir Reservoir 3 Initial Conditions Flows (m /s) Elevation (m) Return Period Spillway Starting (years) Starting Maximum Maximum Required Reservoir 3 Minimum Maximum Outflow (m /s) Inflow Outflow Elevation (m) 2 1,074.83 100 1,072.33 1,074.83 56 100 NO 5 1,074.83 100 1,072.33 1,074.83 107 100 NO 10 1,074.83 150 1,072.33 1,074.83 154 165 NO 20 1,074.33 165 1,072.33 1,074.33 211 165 NO 50 1,074.33 165 1,071.33 1,076.58 302 245 YES 100 1,074.33 165 1,071.33 1,076.83 385 366 YES 200 1,074.33 165 1,071.33 1,077.08 481 477 YES 500 1,073.66 165 1,071.33 1,077.36 632 624 YES 1,000 1,073.66 165 1,071.33 1,077.61 766 757 YES Table A8 shows that the low level outlet has sufficient capacity for floods that have magnitude which is less or equal to the 50 year return period, assuming the current operating rules are implemented in a timely manner. The analysis also shows that the use of the spillway will be required for floods with return flow period above 50 years. A12 FLOOD ROUTING THROUGH THE UPPER BOW SYSTEM Synthetic inflow hydrographs for the selected return flow periods were developed for all key storage structures in the Upper Bow River basin. Their routing through the storage structures and interconnecting river reaches was conducted based on the existing flood operating guidelines obtained from Translata Utilities. In their definition of the operating rules, Transalta distinguishes the Key Elevations for each reservoir as a guideline for triggering the use of the spillway. When the storage is below the Key Elevations, incoming flows are only released through the turbines, subject to the maximum capacity of each plant. Most hydro power plants in the system are designed to operate turbines during flows up to flows above the long term historic averages, and most are designed to also operate during floods together with spillway structures, thus allowing hydro power plants to act as low level outlets and add to the total outflow capacity, while simultaneously generating power. Table A9 provides a listing of maximum flow capacity for each hydro power plant in the Upper Bow River Basin. March 2010 Report No. 09-1326-1040 A26 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT Table A9: Maximum Flows through Hydro Power Plants Name of Hydro Power Plant Maximum Outflow (m3/s) THREE SISTERS 24.6 SPRAY 45.3 RUNDLE 62.3 INTERLAKES 19.8 POCATERA 31.2 BARRIER 36.8 CASCADE 39.6 KANANASKIS 120.7 HORSESHOE 104.8 GHOST 243.5 BEARSPAW 155.7 The Key Elevations at which the spillway operation begins have been established as part of earlier PMF studies, and they are adhered to every year to meet the existing safety requirements. These elevations typically involve a mandatory drawdown of reservoir levels in May of each year, with a gradual return to the full supply levels in June and early July. In addition to the Key Elevations, Transalta has also established target elevations for each storage reservoir as a function of time, as part of earlier operational studies. These target elevations were based on the overall integrated system modelling, with the existing downstream maintenance flow requirements in the Bow River. It was decided that the second week of June (Julian day 161) is the most likely timing for floods of high magnitude. Consequently, the target elevations provided a reasonable estimate of the starting reservoir levels for this week. These levels are somewhat below the Key elevations, but they are normally very close to them (usually lower but within 1 to 2 meters). A typical sketch of the key and target elevations in shown in Figure A8. March 2010 Report No. 09-1326-1040 A27 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT Figure A8: Key and Target Elevations for a Typical Upper Bow Storage Reservoir Transalta’s flood operating guidelines that were used in this study are summarized as follows: Start from the target elevation for the starting date of the flood, and attempt to bring the reservoir to the target elevation at the end of the current time step. This may be possible if there is sufficient inflow. If not, set the outflow to zero and attempt to bring the reservoir level as close as possible to the target elevation at the end of the time step. 1) If Inflow is above the hydro power plant outflow capacity (as may be anticipated during floods), set the outflow equal to the plant capacity and keep the surplus water in storage, thus ending at a reservoir level higher than the target elevation at the end of a time step. 2) Once the reservoir level reaches the Key Elevation, keep the plant operation at full capacity and open the spillways to evacuate the excess water and maintain the key elevations as close as possible. Some reservoirs have alternate outlet structures with a prescribed sequence of start up and shut down (examples are Spray Lake and Lower Kananaskis Lake). These rules were taken into consideration in the model setup. The highest retention capacity is available at Upper and Lower Kananaskis Lake, Lake Minnewanka and Spray Lake. Except for Spray Lake, these are also the locations that have the lowest peak inflows, such that their attenuation does not add much to the reduction of the incoming floods into the Ghost and Bearspaw reservoirs. The principal control points that determine runoff into Ghost Lake are the Bow River at Banff, Spray River at the mouth, and the Kananaskis River at Barrier Dam. The first two control points function as natural flow gauges, March 2010 Report No. 09-1326-1040 A28 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT while the Barrier Dam has negligible flood retention capacity. Therefore, all three control points can be assumed to essentially function without any flood retention capability, and subsequent routing of their runoff through Ghost Lake and Bearspaw reservoir also has little capacity to reduce the incoming flood peaks. The Upper Bow River Basin modelling schematic is illustrated in Figure A9. Included in the model are smaller structures, such as the Kananaskis hydro power plant, in order to include their outflow capacity tables. Channel routing is applied in all channels between Banff and Bearspaw, and on the channel between the Barrier reservoir and the confluence of the Bow and Kananaskis rivers, based on the information provided by Alberta Environment. The model was run using 6-hourly time steps, over an event base of 16 days, which translates to 64 consecutive time steps of 6 hour length. To achieve this, the WRMM was modified to increase its current limit to 52 time steps per year. Table A10 provides a summary of the routed peak flows for two simulations. The first simulation was based on the synthetic hydrographs which were derived for all key locations in the Upper Bow Basin. The other simulation was based on constructing flood hydrographs with added historic information available from high water marks. This simulation included only Ghost Lake and Bearspaw reservoir, since there was no historic information in the upper parts of the catchment that could be used to extend this analysis further upstream. It should be noted that in the base case simulation (based on the 79 years of daily natural flows estimated for the Upper Bow Basin), the design floods with the same return period were assumed to occur simultaneously at all key locations in the basin. This may not be a realistic assumption for more frequent flood events, and as a result the routed peak flows appear to be slightly higher than the ones that would have been obtained by conducting frequency analyses of the recorded flows downstream of Bearspaw reservoir. Table A10: Peak Flows of Routed Design Flood Hydrographs Naturalized Flood Flows of Return Period Naturalized Flows of Systematically Systematically Recorded Data and 3 (years) 3 Recorded Data (m /s) Historic Data (m /s) 2 378 445 5 536 551 10 682 671 20 863 751 50 1,178 862 100 1,492 950 200 1,893 1,068 500 2,584 1,338 1,000 3,301 1,564 March 2010 Report No. 09-1326-1040 A29 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT LAKE MINNEWANKA 207 713 CASCADE SPILL 212 607 KANANASKIS HORSESHOE 610 611 608 609 36 37 707 35 40 0 211 2 34 208 209 210 33 711 CALGARY CANMORE 708 709 710 32 63 GHOST BEARSPAW 6 31 BANFF EXISTING WITHDRAWAL 43 606 706 5 603 703 206 BARRIER 51 RUNDLE LEGEND: 203 62 42 RESERVOIR 4 SPILL 602 605 705 CHANNEL / RIVER REACH 702 HYDRO POWER PLANT 202 205 SPRAY RIVER GOAT POND LOWER KANANASKIS DIVERSION CHANNEL 61 41 SPILL INFLOW 3 701 WITHDRAWAL 601 801 604 704 THREE SISTERS 204 201 UPPER KANANASKIS LAKE SPRAY LAKE Figure A9: Upper Bow River Basin Modelling Schematic March 2010 Report No. 09-1326-1040 A30 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT APPENDIX B Results of Flood Flow Analyses March 2010 Report No. 09-1326-1040 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT Table of Contents B1 FLOOD SERIES FOR BOW RIVER AND ITS TRIBUTARIES ABOVE BEARSPAW DAM ...... 1 B2 PRECIPITATION AND AIR TEMPERATURE DATA ANALYSES ...... 8 B3 SYNTHETIC INFLOW FLOOD HYDROGRAPHS FOR BOW RIVER AND ITS TRIBUTARIES ...... 11 B4 FLOOD FLOW SERIES FOR MAJOR TRIBUTARIES OF BOW RIVER ...... 17 B5 FLOOD SERIES FOR BOW RIVER DOWNSTREAM OF BEARSPAW DAM ...... 30 B6 RESULTS OF FLOOD FREQUENCY ANALYSES FOR THE BOW AND ELBOW RIVER ...... 34 TABLES Table B1: Results of Flood Frequency Analysis for Bow River and its Tributaries Upstream of Bearspaw Reservoir – Based on Naturalized Annual Maximum Daily Flows (m3/s) ...... 1 Table B2: Recorded and Derived Annual Flood Flow Series ...... 18 Table B3: Relationship between Ratio of Flood Flows and Drainage Area of Fish Creek near Priddis and Threepoint Creek near Millarville ...... 28 Table B4: Regional Stations used to Estimate Flood Flows for Pine Creek ...... 29 Table B5: Annual Flood Series for the Bow River below its Confluence with the Elbow River ...... 31 FIGURES Figure B1: Flood Frequency Plots of Annual Maximum Daily Flows for the Bow River and its Tributaries upstream of Bearspaw Reservoir ...... 2 Figure B2: Relationship Between Recorded Annual Maximum Daily and Instantaneous Flows at WSC Station 05BH004 (Bow River at Calgary) ...... 7 Figure B3: Relationship between Recorded Annual Maximum Daily and Instantaneous Flows at WSC Station 05BJ001 (Elbow River below Glenmore Dam) ...... 7 Figure B4: Precipitation and Air Temperature recorded at Calgary and Banff Climate Stations...... 8 Figure B5: Synthetic Inflow Hydrographs ...... 12 Figure B6: Relationships between Annual Maximum Daily Discharge and Annual Maximum Instantaneous Discharge ...... 24 Figure B7: Flood Frequency Plots of Annual Maximum Daily Flows for the Bow River and its Tributaries Downstream of Bearspaw Dam ...... 26 Figure B8: Regional Flood Frequency Relationships Established to Estimate Pine Creek Flood Flows ...... 29 Figure B9: Flood Frequency Plot for the Bow River – Re-run of 1983 Study Data Using Bulletin 17B Method ...... 34 Figure B10: Flood Frequency Plot for the Bow River – Re-run of 1983 Data with Method Used in 2010 Study for Incorporating Historic Floods ...... 35 Figure B11: Flood Frequency Plot for the Bow River – Naturalized and Historic Data up to 2008 Using Bulletin 17B Method ...... 36 Figure B12: Flood Frequency Plot for the Bow River – Naturalized Data up to 2008 with Method Used in 2010 Study for Incorporating Historic Floods ...... 37 Figure B13: Flood Frequency Plot for the Bow River – Natural and Regulated Data up to 2008 with Using Bulletin 17B Method ...... 38 Figure B14: Flood Frequency Plot for the Elbow River – Re-run of 1983 Study Data Using Bulletin 17B Method ...... 39 Figure B15: Flood Frequency Plot for the Elbow River – Natural and Regulated Flood Data up to 2008 ...... 40 March 2010 Report No. 09-1326-1040 Bi HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT B1 FLOOD SERIES FOR BOW RIVER AND ITS TRIBUTARIES ABOVE BEARSPAW DAM Flood frequency analyses for the Bow River and its tributaries upstream of Bearspaw Reservor were conducted on naturalized annual maximum daily flows. Table B1 and Figure B1 (a) to (i) provide the results of fitting the flood series to various frequency distributions, including the 95 percent confidence limits for the best-fit distributions determined using the Anderson-Darling methods. Table B1: Results of Flood Frequency Analysis for Bow River and its Tributaries Upstream of Bearspaw Reservoir – Based on Naturalized Annual Maximum Daily Flows (m3/s) Inflow to Local Inflow into Inflow to Upper Local inflow to Inflow to Minnewanka Lake Spray Lower Kananaskis Kananaskis Lake Spray River at Banff Return Lakes Period (years) Frequency Distribution 3- 3-LN EV2 LP3 3-LN EV2 LP3 3-LN EV2 LP3 3-LN EV2 LP3 EV2 LP3 LN 2 17.0 17.1 17.4 22.5 22.5 22.4 54.0 53.5 53.8 66.2 65.4 68.2 11.7 12.0 12.8 5 20.7 20.8 20.8 26.7 26.8 26.6 78.1 77.0 77.6 88.0 86.5 88.7 18.5 18.2 19.0 10 23.2 23.3 22.8 29.4 29.7 29.4 95.2 95 94.9 104 103 102 24.8 23.7 23.3 20 25.7 25.8 24.6 32.0 32.4 32.1 113 113 113 120 122 114 32.4 30.4 27.6 50 29.0 29.1 26.9 35.4 36.0 35.8 136 140 137 142 150 130 44.5 41.8 33.5 100 31.6 31.6 28.5 37.9 38.7 38.6 155 163 158 160 174 142 55.5 52.9 38.2 200 34.2 34.1 30.1 40.5 41.3 41.6 174 188 179 179 202 154 68.4 66.7 43.0 500 37.8 37.6 32.2 44.0 44.9 45.6 201 224 209 205 245 170 88.4 90.3 49.7 1000 40.6 40.2 33.7 46.6 47.5 48.9 222 255 234 226 284 182 106 113 55.0 Inflow to Bearspaw Inflow to Barrier Inflow to Ghost Inflow to Glenmore Bow River at Banff Dam-Without Return Reservoir Reservoir Reservoir Historic Flood Period (years) Frequency Distribution 3-LN EV2 LP3 3-LN EV2 LP3 3-LN EV2 LP3 3-LN EV2 LP3 3-LN EV2 LP3 2 73.7 72.1 73.0 204 205 204 348 347 356 382 378 392 55.9 56.8 55.1 5 116 112 116 255 257 255 458 455 459 518 510 521 107 103 104 10 147 144 148 288 289 287 535 533 525 616 613 604 154 146 152 20 179 180 181 318 317 316 612 614 587 716 724 684 211 201 216 50 222 235 228 357 353 354 716 726 666 853 891 787 302 298 329 100 257 283 267 386 377 382 797 817 725 961 1036 864 385 397 445 200 294 339 308 415 401 410 881 915 783 1075 1198 942 481 524 595 500 346 425 366 452 430 447 997 1053 861 1234 1444 1045 632 753 863 1000 387 501 414 481 450 475 1089 1166 921 1361 1657 1125 766 987 1134 Notes: 3-LN: 3-parameters log-normal distribution EV2: Extreme Value Type II Distribution LP3: Log-Pearson Type II Highlighted column indicates distribution fit selected to estimate the peak floods March 2010 Report No. 09-1326-1040 B1 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT Figure B1: Flood Frequency Plots of Annual Maximum Daily Flows for the Bow River and its Tributaries upstream of Bearspaw Reservoir March 2010 Report No. 09-1326-1040 B2 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT (d)(d) Inflow Inflow to to Spary Spray Lakes 1000 Naturalized Annual Maximum Daily Data Extreme Value (Type II) Lower Bound (95%) Upper Bound (95%) Q (m³/s) Q 100 10 2 5 10 20 50 500 100 200 1.05 1.25 1000 1.003 Return Periods (Years) March 2010 Report No. 09-1326-1040 B3 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT March 2010 Report No. 09-1326-1040 B4 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT (h) SpraySpary RiveRiver at at Banff Banff 100 Recorded Annual Maximum Daily Data 3-Parameters Log-Normal (MLH) Lower Bound (95%) Upper Bound (95%) Q (m³/s) Q 10 1 2 5 10 20 50 500 200 100 1.05 1.25 1000 1.003 Return Periods (Years) March 2010 Report No. 09-1326-1040 B5 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT Figures B2 and B3 show a relationship between annual maximum daily and maximum instantaneous flood flows for the Bow River below Bearspaw Reservoir and the Elbow River below Glenmore Reservoir, respectively. March 2010 Report No. 09-1326-1040 B6 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT Figure B2: Relationship Between Recorded Annual Maximum Daily and Instantaneous Flows at WSC Station 05BH004 (Bow River at Calgary) 1600 Bow River At Calgary 1400 y = 0.9888x1.019 1200 R² = 0.9665 1000 800 600 Discharge (m³/s) Discharge 400 200 Annual Instantanuous Maximum 0 0 200 400 600 800 1000 1200 1400 Annual Daily Maximum Discharge (m³/s) Figure B3: Relationship between Recorded Annual Maximum Daily and Instantaneous Flows at WSC Station 05BJ001 (Elbow River below Glenmore Dam) ELBOW RIVER BELOW GLENMORE DAM 800 Based on Elbow River Below Glenmore Data (1975 to 2008) and 4-Highest Floods from 1983 Study 700 Based on T. Blench Data (1915-1964) y = 0.6589x1.1402 R² = 0.9471 600 Power (Based on Elbow River Below Glenmore Data (1975 to 2008) and 4-Highest Floods from 1983 Study) Power (Based on T. Blench Data (1915-1964)) 500 400 y = 0.9959x1.045 (m³/s) R² = 0.9619 300 200 100 Annual MaximumInstantaneous Discharge Discharge MaximumInstantaneous Annual 0 0 50 100 150 200 250 300 350 400 450 Annual Maximum Daily Discharge (m³/s) March 2010 Report No. 09-1326-1040 B7 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT B2 PRECIPITATION AND AIR TEMPERATURE DATA ANALYSES Figure B4 (a) to (d) provides precipitation and air temperature recorded at three stations in 1897 and 1902. Annual precipitation data recorded since 1884 at the Calgary Airport station shows that the precipitation patterns pre-1911 (i.e., period prior to beginning of the systematically record of flow data) is similar to precipitation patterns post- 1911. Figure B4: Precipitation and Air Temperature recorded at Calgary and Banff Climate Stations (a) Year 1897 1000 30 900 20 800 10 700 0 C) o 600 -10 500 -20 Peak Flood Occured on June 18 400 -30 Mean Air Temperature ( Cumulative PrecipitationCumulative (mm) 300 Banff _Precipitation -40 Calgary _Precipitation Keenhill _Precipitation 200 -50 Banff _Temperature Calgary _Temperature 100 -60 Keenhill _Temperature Banff_ Maximum Temperature 0 -70 01/01/1897 02/03/1897 01/04/1897 31/05/1897 30/07/1897 29/08/1897 28/10/1897 27/11/1897 31/01/1897 01/05/1897 30/06/1897 28/09/1897 27/12/1897 March 2010 Report No. 09-1326-1040 B8 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT (b) Year 1902 1000 30 900 20 800 10 700 0 C) o 600 -10 500 -20 400 -30 Mean Air Temperature ( Cumulative PrecipitationCumulative (mm) 300 Banff _Precipitation -40 Calgary _Precipitation 200 Keenhill _Precipitation -50 Banff _Temperature Peak Flood Occured on July 4 Calgary _Temperature 100 -60 Keenhill _Temperature Banff_ Maximum Temperature 0 -70 March 2010 Report No. 09-1326-1040 B9 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT (c) Annual Precipitation at Calgary Airport 1000 900 800 700 600 (mm) 500 Precipitation 400 Annual Annual 300 200 100 0 1881 1884 1887 1890 1893 1896 1899 1902 1905 1908 1911 1914 1917 1920 1923 1926 1929 1932 1935 1938 1941 1944 1947 1950 1953 1956 1959 1962 1965 1968 1971 1974 1977 1980 1983 1986 1989 1992 1995 1998 2001 2004 2007 March 2010 Report No. 09-1326-1040 B10 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT (d) Annual Precipitation at Banff 1000 900 800 700 600 500 400 Annual Precipitation Precipitation (mm)Annual 300 200 100 0 1887 1890 1893 1896 1899 1902 1905 1908 1911 1914 1917 1920 1923 1926 1929 1932 1935 1938 1941 1944 1947 1950 1953 1956 1959 1962 1965 1968 1971 1974 1977 1980 1983 1986 1989 1992 1995 B3 SYNTHETIC INFLOW FLOOD HYDROGRAPHS FOR BOW RIVER AND ITS TRIBUTARIES Figure B5 (a) to (k) provides all the synthetic inflow hydrographs developed for major facilities on the Bow River and its tributaries. March 2010 Report No. 09-1326-1040 B11 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT Figure B5: Synthetic Inflow Hydrographs 300 (a) Inflow to Spray Lakes 1000 yr 500 yr 250 200 yr 100 yr 200 50 yr 20 yr 150 10 yr 5 yr 100 Discharge (m³/s) Discharge 2 yr 50 0 0 1 2 3 4 5 6 Day March 2010 Report No. 09-1326-1040 B12 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT 45 (c) Local Inflow to Pocaterra 1000 yr 40 500 yr 200 yr 35 100 yr 30 50 yr 20 yr 25 10 yr 20 5 yr Discharge (m³/s) Discharge 15 2 yr 10 5 0 0 1 2 3 4 5 6 7 8 Time (Days) 300 (d) Inflow to Minnewanka Lake 1000 yr 500 yr 250 200 yr 100 yr 200 50 yr 20 yr 150 10 yr 5 yr Discharge (m³/s) Discharge 100 2 yr 50 0 0 2 4 6 8 10 12 Time (Days) March 2010 Report No. 09-1326-1040 B13 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT 450 (e) Inflow to Barrier Reservoir 1000 yr 400 500 yr 350 200 yr 100 yr 300 50 yr 250 20 yr 200 10 yr Discharge (m³/s) Discharge 5 yr 150 2 yr 100 50 0 0 2 4 6 8 10 12 14 16 Time (Days) 800 (f) Bow River at Banff 1000 yr 700 500 yr 200 yr 600 100 yr 500 50 yr 20 yr 400 10 yr 5 yr Discharge (m³/s) Discharge 300 2 yr 200 100 0 0 2 4 6 8 10 12 14 16 Time (Days) March 2010 Report No. 09-1326-1040 B14 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT 120 (g) Local Inflow to Spray River at Banff 1000 yr 500 yr 100 200 yr 100 yr 80 50 yr 20 yr 60 10 yr Discharge (m³/s) Discharge 5 yr 40 2 yr 20 0 0 5 10 15 20 25 Time (Days) 1200 (h) Bow River at Ghost Reservoir 1000 yr 500 yr 1000 200 yr 100 yr 800 50 yr 20 yr 10 yr 600 5 yr 2 yr Discharge (m³/s) Discharge 400 200 0 0 1 2 3 4 5 6 7 8 9 Time (Days) March 2010 Report No. 09-1326-1040 B15 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT 1800 (i) Bow River at Bearspaw Dam - Systematically Recorded Data Only 1000 yr 1600 500 yr 200 yr 1400 100 yr 1200 50 yr 20 yr 1000 10 yr 5 yr 800 2 yr Discharge (m³/s) Discharge 600 400 200 0 0 1 2 3 4 5 6 7 8 9 Time (Days) 4000 (j) Bow River at Bearspaw Dam- With Historic Data 1000 yr 3500 500 yr 200 yr 3000 100 yr 50 yr 2500 20 yr 10 yr 2000 5 yr 2 yr Discharge (m³/s) Discharge 1500 1000 500 0 0 1 2 3 4 5 6 7 8 9 Time (Days) March 2010 Report No. 09-1326-1040 B16 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT 900 (k) Elbow River inflow to Glenmore Reservoir 1000 yr 800 500 yr 700 200 yr 100 yr 600 50 yr 500 20 yr 400 10 yr Discharge (m³/s) Discharge 5 yr 300 2 yr 200 100 0 0 1 2 3 4 5 6 7 8 Time (Days) B4 FLOOD FLOW SERIES FOR MAJOR TRIBUTARIES OF BOW RIVER Table B2 (a) to (d) provides recorded and derived annual peak flow series for tributary streams to the Bow River. Figure B6 (a) to (d) provides the relationships established between annual maximum daily discharge and annual maximum instantaneous discharge for major tributaries of the Bow River. For those years with no date indicated in Table B2 (a) to (d), the instantaneous discharges were derived based on relationships established in Figure B6 (a) to (d). Figure B7 (a) to (d) provide the results of fitting the peak flow series to various frequency distributions, including the 95 percent confidence limits for the best fit distribution determined using the Anderson-Darling methods. Table B3 provided a regional relationship established between Fish Creek near Priddis (WSC Station 05BK001) and Threepoint Creek near Millarville (WSC station 05BL013). This relationship was used to transfer the flood flow data from Fish Creek near Priddis to Fish Creek at the mouth. March 2010 Report No. 09-1326-1040 B17 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT Table B2: Recorded and Derived Annual Flood Flow Series (a) Nose Creek At Calgary Name Nose Creek At Calgary (WSC Station 05BH003) 05BH003 Gauge Drainage Area 893 km² Effective Drainage Area 746 km² Lat 51°7'20" Long 114°2'45" Instantaneous Discharge Year Maximum Daily Discharge Date (Month-Day) (m³/s) Date (Month-Day) (m³/s) 1911 06-14 2.97 - 3.76 1912 03-27 2.66 - 3.36 1913 04-12 6.43 - 8.12 1914 06-19 1.36 - 1.72 1915 08-20 54.79 - 69.20 1916 05-31 32.80 - 41.43 1917 05-18 26.80 - 33.85 1918 04-04 4.25 - 5.36 1919 08-05 4.76 08-05 5.40 1973 03-22 8.10 - 10.23 1974 04-10 9.32 - 11.77 1975 04-13 3.31 04-13 5.21 1976 03-19 1.85 03-19 3.00 1977 03-14 0.19 04-05 0.49 1978 03-27 10.20 03-27 12.00 1979 03-08 2.28 03-07 3.23 1980 04-05 4.06 05-25 5.29 1981 05-07 2.05 07-14 5.23 1982 04-12 5.57 04-12 6.47 1983 04-25 1.66 04-24 3.79 1984 03-25 0.49 - 0.61 1985 09-13 5.68 09-12 18.00 1986 09-26 6.57 09-26 8.40 March 2010 Report No. 09-1326-1040 B18 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT (b) Fish Creek Near Priddis Name Fish Creek Near Priddis (WSC Station 05BK001) 05BK001 Gauge Drainage Area 259 km² Effective Drainage Area 259 km² Lat 50°53'10" Long 114°19'35" Year Maximum Daily Discharge Instantaneous Discharge Date (Month-Day) (m³/s) Date (Month-Day) (m³/s) 1911 08-08 45.30 - 67.01 1912 07-25 37.10 - 54.24 1913 06-29 8.78 - 11.80 1914 06-26 3.11 - 3.93 1915 06-26 199.00 06-26 200.00 1916 06-04 121.00 - 189.55 1956 05-13 5.66 - 7.41 1957 04-18 7.33 - 9.75 1958 04-16 13.60 - 18.75 1959 06-27 36.20 06-27 41.30 1960 03-24 4.79 - 6.21 1961 05-01 3.54 05-16 4.05 1962 04-04 8.64 - 11.60 1963 06-30 58.90 - 88.47 1964 05-03 17.30 - 24.19 1965 07-09 29.20 - 42.10 1966 07-03 14.80 - 20.51 1967 05-31 72.50 - 110.23 1968 06-08 7.25 - 9.64 1969 06-29 75.30 - 114.74 1970 06-14 34.50 - 50.23 1971 06-06 19.90 - 22.36 1972 04-06 7.16 - 9.51 1973 05-26 12.10 - 16.57 1974 04-29 20.70 - 29.25 1975 05-08 10.00 - 13.54 1976 08-10 3.26 - 4.14 1977 04-05 6.88 - 9.12 1978 05-31 21.10 - 29.85 March 2010 Report No. 09-1326-1040 B19 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT Name Fish Creek Near Priddis (WSC Station 05BK001) 05BK001 Gauge Drainage Area 259 km² Effective Drainage Area 259 km² Lat 50°53'10" Long 114°19'35" Year Maximum Daily Discharge Instantaneous Discharge Date (Month-Day) (m³/s) Date (Month-Day) (m³/s) 1979 04-27 6.45 - 8.51 1980 06-04 10.80 - 14.69 1981 06-14 17.00 - 23.75 1982 06-07 4.67 - 6.05 1983 04-25 16.00 - 22.27 1984 06-09 4.80 06-09 5.64 1985 09-13 6.52 09-13 8.26 1986 09-27 5.90 - 7.75 1987 04-01 4.97 - 6.46 1988 04-02 2.77 04-04 4.11 1989 04-23 4.44 04-23 5.01 1990 06-03 12.60 06-02 20.00 1991 06-22 6.65 06-22 9.70 1992 06-15 35.90 06-15 45.00 1993 06-16 16.20 - 22.57 1994 06-07 44.10 06-07 70.30 1995 06-07 59.30 06-07 84.20 1996 04-06 16.50 - 23.01 1997 05-26 20.70 05-26 26.60 1998 07-08 20.70 07-08 31.90 1999 08-13 18.40 08-13 31.30 2000 06-09 3.92 - 5.03 2001 04-28 9.85 - 13.33 2002 06-10 22.60 06-10 26.00 2003 04-26 15.10 04-25 25.60 2004 07-05 3.00 07-04 3.87 2005 06-18 166.00 06-18 482.00 2006 06-16 38.30 06-15 64.80 2007 06-18 10.10 06-18 12.60 2008 06-11 55.40 06-11 80.60 March 2010 Report No. 09-1326-1040 B20 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT (c) Threepoint Creek Near Millarville Name Threepoint Creek Near Millarville (WSC Station 05BL013) 05BL013 Gauge Drainage Area 507 km² Effective Drainage Area 507 km² Lat 50°46'16" Long 114°16'39" Year Maximum Daily Discharge Instantaneous Discharge (m³/s) Date (Month-Day) (m³/s) Date (Month-Day) 1909 05-24 23.10 - 36.89 1912 07-08 25.50 - 40.73 1913 05-14 15.50 - 24.75 1914 06-14 8.01 - 12.79 1915 06-26 60.00 06-26 159.00 1916 08-18 47.00 - 75.06 1966 06-05 24.60 06-10 32.60 1967 05-31 164.00 05-30 283.00 1968 06-08 20.20 06-08 26.00 1969 06-29 109.00 06-29 146.00 1970 06-14 56.10 06-14 75.90 1971 06-06 30.90 06-06 41.60 1972 04-06 18.00 04-06 27.30 1973 05-27 24.60 05-26 32.80 1974 04-30 34.30 04-30 53.80 1975 06-20 28.90 06-20 33.10 1976 08-10 13.00 08-09 17.40 1977 04-05 7.65 04-04 17.30 1978 05-31 27.30 05-31 37.10 1979 05-17 9.03 04-28 10.40 1980 06-04 31.70 06-04 38.50 1981 06-14 34.30 06-14 45.20 1982 06-30 14.30 06-28 18.10 1983 04-25 21.60 04-25 32.60 1984 06-09 7.12 06-09 7.86 1985 09-13 33.70 09-13 40.10 1986 05-22 11.50 09-27 13.00 1987 04-01 12.10 - 19.32 1988 08-21 3.19 08-21 3.45 March 2010 Report No. 09-1326-1040 B21 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT Name Threepoint Creek Near Millarville (WSC Station 05BL013) 05BL013 Gauge Drainage Area 507 km² Effective Drainage Area 507 km² Lat 50°46'16" Long 114°16'39" Year Maximum Daily Discharge Instantaneous Discharge (m³/s) Date (Month-Day) (m³/s) Date (Month-Day) 1989 04-15 8.25 04-14 14.00 1990 05-26 51.30 05-26 56.30 1991 06-22 18.20 - 29.07 1992 06-15 88.00 - 140.54 1993 07-13 49.40 - 78.90 1994 06-07 66.80 06-07 119.00 1995 06-07 104.00 06-07 129.00 1996 05-30 18.20 - 29.07 1997 05-26 43.00 - 68.67 1998 06-20 43.60 07-11 76.50 1999 07-05 15.50 08-13 22.60 2000 03-28 4.03 - 6.44 2001 06-06 19.10 06-06 22.80 2002 06-10 64.50 06-10 75.50 2003 04-25 12.50 04-25 32.80 2004 07-04 14.50 - 23.16 2005 06-07 229.00 06-17 389.00 2006 06-16 90.50 06-15 157.00 2007 06-18 27.10 06-18 34.10 2008 05-25 124.00 05-25 220.00 March 2010 Report No. 09-1326-1040 B22 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT (d) Highwood River at Mouth Highwood River at Mouth Name 05BL024 (WSC Station 05BL024) Gauge Drainage Area 3952 km² Effective Drainage Area 3940 km² Lat 50°46'59" N Long 113°49'15" W Year Maximum Daily Discharge Instantaneous Discharge Date (Month-Day) (m³/s) Date (Month-Day) (m³/s) 1970 06-14 331.00 06-14 430.00 1971 06-07 238.00 06-06 278.00 1972 06-01 205.00 06-01 219.00 1973 05-27 124.00 05-27 141.00 1974 06-17 215.00 06-17 232.00 1975 06-20 292.00 06-20 357.00 1976 08-06 178.00 08-06 213.00 1977 05-25 29.70 05-25 31.4.00 1978 06-06 154.00 06-06 163.00 1979 05-27 115.00 05-27 126.00 1980 06-04 161.00 06-04 178.00 1981 05-26 265.00 05-26 307.00 1982 06-15 84.00 06-15 88.40 1983 04-25 76.10 04-25 108.00 1984 06-16 51.50 05-31 54.90 1985 09-13 149.00 09-13 174.00 1986 05-29 153.00 05-29 162.00 1987 07-20 54.60 07-20 57.60 1988 06-09 75.30 06-08 85.80 1989 06-07 65.60 06-07 69.00 1990 05-26 383.00 05-26 447.00 1991 05-20 210.00 - 261.10 1992 06-15 325.00 06-15 375.00 1993 06-17 277.00 - 344.30 1994 06-07 178.00 06-07 269.00 1995 06-07 869.00 06-07 1,120.00 1996 06-04 146.00 06-04 164.00 1997 06-01 203.00 06-01 223.00 1998 06-20 389.00 06-20 462.00 March 2010 Report No. 09-1326-1040 B23 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT Highwood River at Mouth Name 05BL024 (WSC Station 05BL024) Gauge Drainage Area 3952 km² Effective Drainage Area 3940 km² Lat 50°46'59" N Long 113°49'15" W Year Maximum Daily Discharge Instantaneous Discharge Date (Month-Day) (m³/s) Date (Month-Day) (m³/s) 1999 07-05 81.30 07-05 89.80 2000 06-10 57.30 06-09 59.40 2001 06-06 136.00 06-06 153.00 2002 06-10 277.00 - 344.30 2003 04-26 124.00 04-26 164.00 2004 08-27 91.90 08-26 99.20 2005 06-08 1,040.00 06-18 1,340.00 2006 06-16 438.00 - 544.50 2007 06-07 107.00 06-07 113.00 2008 05-25 626.00 05-25 769.00 Figure B6: Relationships between Annual Maximum Daily Discharge and Annual Maximum Instantaneous Discharge (a) Nose Creek At Calgary 20 18 y = 1.263x R² = 1.000 16 14 12 (m³/s) 10 8 6 4 2 Annual Maximum Annual Maximum Instantanuous Discharge 0 0 2 4 6 8 10 12 Annual Maximum Daily Discharge (m³/s) March 2010 Report No. 09-1326-1040 B24 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT (b) Fish Creek near Priddis 600 500 400 y = 1.184x1.058 R² = 0.974 300 200 (m³/s) 100 0 0 50 100 150 200 250 Annual Maximum Annual Maximum Instantanuous Discharge Annual Maximum Daily Discharge (m³/s) (c) Threepoint Creek Near Millarville 450 400 350 300 250 200 y = 1.6212x (m³/s) R² = 0.956 150 100 50 0 Annual Maximum Annual Maximum Instantanuous Discharge 0 50 100 150 200 250 Annual Maximum Daily Discharge (m³/s) March 2010 Report No. 09-1326-1040 B25 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT (d) Highwood River At Mouth 1600 1400 1200 1000 800 y = 1.243x (m³/s) 600 R² = 0.993 400 200 0 Annual Maximum Annual Maximum Instantanuous Discharge 0 200 400 600 800 1000 1200 Annual Maximum Daily Discharge (m³/s) Figure B7: Flood Frequency Plots of Annual Maximum Daily Flows for the Bow River and its Tributaries Downstream of Bearspaw Dam March 2010 Report No. 09-1326-1040 B26 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT (b) Nose Creek at Calgary March 2010 Report No. 09-1326-1040 B27 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT Table B3: Relationship between Ratio of Flood Flows and Drainage Area of Fish Creek near Priddis and Threepoint Creek near Millarville Fish Creek near Threepoint Creek Return Period Priddis near Millarville Ln (Qf/Qt) Power Factor Drainage Area (km2) 259 507 2 14.7 32.5 0.795 1.184 5 34.2 67.3 0.677 1.008 10 54.6 99.3 0.599 0.891 20 80.9 137.4 0.529 0.788 50 126.7 198.4 0.448 0.667 100 171.3 253.7 0.393 0.585 200 225.9 317.9 0.342 0.509 500 316.3 418.1 0.279 0.416 1,000 400.6 506.8 0.235 0.350 Power Factor QThreepoint Creek/QFish Creek = (AThreepoint Creek/AFish Creek) Table B4 lists the regional stations used to estimate peak flows for the Pine Creek at the mouth. Figure B8 provides the results of the regional flood frequency analysis. March 2010 Report No. 09-1326-1040 B28 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT Table B4: Regional Stations used to Estimate Flood Flows for Pine Creek Drainage Period of Operational Station Station No. 2 Area (km ) Record Period Fish Creek near Priddis 05BK001 259 1911-2008 March – October Meadow Creek near the mouth 05AB026 130 1966-2008 March – October Pekisko Creek near Longview 05BL023 230 1967-2008 March – October Prairie Creek near Rocky Mountain House 05DB002 859 1922-2008 Continuous Prairie Blood Coulee near Lethbridge 05AD035 227 1970-2008 March – October Ray Creek near Innisfail 05CE010 42.5 1967-2004 March – October Rosebud River below Carstairs Creek 05CE006 753 1957-2008 March - October Threepoint Creek near Millarville 05BL013 507 1909-2008 March - October Trap Creek near Longview 05BL027 137 1979-2008 March – October Trout Creek near Granum 05AB005 441 1911-2008 March – October West Arrowwood Creek near Arrowwood 05BM014 775 1965-2008 February - October Figure B8: Regional Flood Frequency Relationships Established to Estimate Pine Creek Flood Flows March 2010 Report No. 09-1326-1040 B29 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT B5 FLOOD SERIES FOR BOW RIVER DOWNSTREAM OF BEARSPAW DAM The flood series for Bow River below the confluences with the Elbow River were derived by checking the following combinations of flows: 1) Simultaneous occurrence of instantaneous peak flows on the Bow River and the Elbow River; 2) Instantaneous peak flow on the Bow River plus corresponding daily flow on the Elbow River; 3) Instantaneous peak flow on the Elbow River and corresponding maximum daily flow on the Bow River; and 4) Simultaneous daily flow on the Bow River and the Elbow River. Each combination was investigated, on an annual basis, and the combination which gave the highest flow was selected to represent the maximum instantaneous discharge. The results of these analyses are provided below (Table B5) for the Bow River below its confluence with the Elbow River. March 2010 Report No. 09-1326-1040 B30 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT Table B5: Annual Flood Series for the Bow River below its Confluence with the Elbow River Maximum Daily Instantaneous Maximum Daily Instantaneous Daily Discharge Daily Discharge Floods on Bow Flood for Bow Floods For Elbow Floods For Elbow Floods For Bow Flood For Bow for Elbow River For Bow River at and Elbow River River below its Governing Date Date River Below River Below (a) River at Calgary River at Calgary Below Glenmore Calgary occurs on the Confluence with Combination Glenmore Glenmore (05BH004) (m³/s) (05BH004)4 (m³/s) (05BJ001) (m³/s) (05BH004) (m³/s) same Date Elbow River (05BJ001) (m³/s) (05BJ001)1 (m³/s) 25/06/1911 467.0 523.5 36.0 08/08/1911 89.5 109.3 428 No 559.55 2 10/07/1912 430.0 482.1 77.6 16/06/1912 122.0 151.8 357 No 559.67 2 12/06/1913 416.0 466.4 25.4 10/08/1913 38.8 45.1 235 No 491.77 2 18/06/1914 405.0 454.0 28.9 18/06/1914 28.9 33.0 405 Yes 487.04 1 26/06/1915 796.0 1130.0 239 26/06/1915 239.0 379.4 796 Yes 1,509.44 1 21/06/1916 767.0 810.0 111.0 29/06/1916 146.0 183.6 680 No 921.00 2 03/06/1917 436.0 488.8 147.0 03/06/1917 147.0 185.0 436 Yes 672.4 1 16/06/1918 532.0 596.4 24.1 10/06/1918 35.4 40.9 289 No 620.52 2 05/08/1919 411.0 459.0 46.4 06/08/1919 72.5 87.5 311 No 505.40 2 13/07/1920 510.0 515.0 67.7 13/07/1920 67.7 69.9 510 Yes 584.94 1 09/06/1921 337.0 377.8 27.9 25/05/1921 37.4 43.4 248 No 405.71 2 06/06/1922 365.0 377.0 22.6 17/05/1922 26.5 30.1 72.5 No 399.60 2 14/06/1923 714.0 841.0 119 01/06/1923 331.0 402.1 583 No 985.10 3 05/07/1924 377.0 385.0 28.3 04/08/1924 59.5 70.9 277 No 413.30 2 23/06/1925 385.0 394.0 30.0 12/06/1925 66.5 71.6 289 No 424.00 2 11/09/1926 259.0 277.0 88.1 11/09/1926 88.1 107.5 259 Yes 384.52 1 11/06/1927 547.0 564.0 78.4 10/06/1927 83.3 84.7 467 No 642.40 2 01/07/1928 558.0 575.0 60.6 19/06/1928 100.0 107.9 515 No 635.60 2 03/06/1929 1150.0 1320.0 382.0 03/06/1929 382.0 433.2 1,150 Yes 1,753.24 1 10/06/1930 399.0 428.0 23.8 31/05/1930 30.6 32.3 199 No 451.80 2 20/06/1931 306.0 328.0 16.4 08/04/1931 22.9 28.3 60 No 344.40 2 03/06/1932 1160.0 1520.0 311.0 03/06/1932 311.0 713.6 1,160 Yes 2,233.58 1 18/06/1933 612.0 686.1 26.0 05/06/1934 32.0 36.8 230 No 712.11 2 01/06/1934 447.0 513.0 10.5 16/06/1935 32.8 37.7 326 No 523.50 2 17/06/1935 374.0 394.0 31.1 08/06/1936 35.7 41.3 191 No 425.10 2 03/06/1936 408.0 428.0 27.2 14/06/1937 54.9 65.1 155 No 455.20 2 18/06/1937 257.0 297.0 24.1 24/05/1938 59.5 70.9 159 No 321.10 2 23/06/1938 413.0 459.0 21.0 17/06/1939 104.0 122.3 233 No 480.00 2 02/07/1939 309.0 320.0 30.3 30/04/1940 40.5 47.2 55.2 No 350.30 2 25/05/1940 271.0 323.0 18.1 04/06/1941 27.0 30.7 137 No 341.10 2 16/06/1941 178.0 197.0 11.3 12/05/1942 128.0 226.5 66 No 292.53 3 09/06/1942 303.0 365.0 53.8 25/04/1943 34.8 40.2 82.7 No 418.80 2 05/07/1943 323.0 365.0 22.1 14/06/1944 26.5 30.1 214 Yes 395.11 1 14/06/1944 214.0 251.0 26.5 26/05/1945 67.7 83.5 59.5 No 277.50 2 23/06/1945 306.0 377.0 40.5 29/09/1946 54.1 56.6 80.7 No 417.50 2 30/05/1946 354.0 402.0 38.5 11/05/1947 79.3 78.4 253 No 440.50 2 10/06/1947 374.0 425.0 41.9 24/05/1948 154.0 259.1 527 No 786.10 3 25/05/1948 558.0 595.0 89.8 22/05/1949 22.6 25.4 124 No 684.80 2 09/06/1949 200.0 228.0 5.75 16/06/1950 33.1 38.2 371 No 409.23 3 March 2010 Report No. 09-1326-1040 B31 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT Maximum Daily Instantaneous Maximum Daily Instantaneous Daily Discharge Daily Discharge Floods on Bow Flood for Bow Floods For Elbow Floods For Elbow Floods For Bow Flood For Bow for Elbow River For Bow River at and Elbow River River below its Governing Date Date River Below River Below River at Calgary River at Calgary Below Glenmore Calgary occurs on the Confluence with Combination(a) Glenmore Glenmore (05BH004) (m³/s) (05BH004)4 (m³/s) (05BJ001) (m³/s) (05BH004) (m³/s) same Date Elbow River (05BJ001) (m³/s) (05BJ001)1 (m³/s) 23/06/1950 419.0 490.0 32.0 31/08/1951 152.0 170.7 - No 522.00 2 24/06/1952 408.0 487.0 79.3 07/04/1952 80.4 90.9 - No 566.30 2 14/06/1953 513.0 578.0 103.0 04/06/1953 124.0 166.8 365 No 681.00 2 26/08/1954 388.0 428.0 47.6 22/05/1954 59.7 71.2 116 No 475.60 2 25/06/1955 365.0 416.0 24.2 20/05/1955 46.7 54.9 103 No 440.20 2 07/06/1956 306.0 377.0 20.8 05/07/1956 37.1 43.0 169 No 397.80 2 09/06/1957 270.0 320.0 38.2 10/06/1957 38.5 44.7 220 No 358.20 2 13/06/1958 250.0 281.0 18.2 14/07/1958 55.2 65.5 183 No 299.20 2 28/06/1959 306.0 326.0 48.7 27/06/1959 56.1 66.6 297 No 374.70 2 02/07/1960 289.0 306.0 21.4 18/05/1960 30.3 34.7 141 No 327.40 2 07/06/1961 391.0 396.0 28.3 29/05/1961 53.2 63.0 326 No 424.30 2 28/06/1962 249.0 264.0 6.68 18/06/1962 30.9 35.4 157 No 270.68 2 30/06/1963 317.0 450.0 99.1 30/06/1963 99.1 178.7 317 Yes 628.68 1 16/06/1964 351.0 377.0 43.6 09/06/1964 70.2 76.5 111 No 420.60 2 20/06/1965 456.0 473.0 56.9 19/06/1965 99.7 122.6 445 No 567.57 3 04/06/1966 345.0 354.0 29.7 03/07/1966 37.9 44.0 218 No 383.70 2 24/06/1967 382.0 428.0 26.1 01/06/1967 146.0 183.6 317 No 500.62 3 13/06/1968 256.0 265.0 21.7 11/06/1968 44.2 51.8 218 No 286.70 2 30/06/1969 399.0 481.0 118.0 29/06/1969 133.0 166.3 362 No 599.00 2 17/06/1970 331.0 334.0 83.8 14/06/1970 105.0 129.5 217 No 417.80 2 07/06/1971 362.0 388.0 77.9 07/06/1971 77.9 94.4 362 No 465.90 2 03/07/1972 402.0 425.0 17.2 11/06/1972 41.1 47.9 354 No 442.20 2 26/06/1973 334.0 354.0 29.2 28/05/1973 36.2 41.9 244 No 383.20 2 26/06/1974 476.0 484.0 38.5 17/06/1974 61.7 73.7 391 No 522.50 2 16/07/1975 224.0 242.0 7.19 20/06/1975 43.3 60.9 157 No 249.19 2 07/08/1976 249.0 253.0 32.3 06/08/1976 37.7 42.5 240 No 285.30 2 10/06/1977 208.0 233.2 10.6 23/08/1977 13.6 13.7 82.7 No 243.79 2 12/06/1978 238.0 292.0 24.0 06/06/1978 41.9 47.0 184 No 316.00 2 30/05/1979 141.0 160.0 13.6 27/05/1979 30.8 31.9 118 No 173.60 2 19/06/1980 302.0 370.0 23.4 03/06/1980 65.3 101.0 197 No 393.40 2 28/05/1981 425.0 427.0 60.3 27/05/1981 93.0 95.8 420 No 515.80 3 24/06/1982 276.0 305.0 20.9 16/06/1982 37.1 41.9 207 No 325.90 2 04/07/1983 202.0 225.0 14.5 28/05/1983 22.9 23.4 115 No 239.50 2 02/07/1984 217.0 234.0 8.35 10/06/1984 19.7 20.1 95.7 No 242.35 2 27/05/1985 189.0 198.0 3.53 13/09/1985 72.8 97.6 134 No 231.60 3 03/06/1986 420.0 427.0 29.2 29/05/1986 50.2 62.7 127 No 456.20 2 17/05/1987 197.0 207.0 6.3 20/07/1987 27.2 27.5 107 No 213.30 2 12/06/1988 227.0 242.0 11.8 08/06/1988 34.6 36.8 147 No 253.80 2 17/06/1989 238.0 244.0 16.1 07/06/1989 20.3 23.3 122 No 260.10 2 03/06/1990 525.0 530.0 72.9 26/05/1990 126.0 131.0 255 No 602.90 2 March 2010 Report No. 09-1326-1040 B32 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT Maximum Daily Instantaneous Maximum Daily Instantaneous Daily Discharge Daily Discharge Floods on Bow Flood for Bow Floods For Elbow Floods For Elbow Floods For Bow Flood For Bow for Elbow River For Bow River at and Elbow River River below its Governing Date Date River Below River Below River at Calgary River at Calgary Below Glenmore Calgary occurs on the Confluence with Combination(a) Glenmore Glenmore (05BH004) (m³/s) (05BH004)4 (m³/s) (05BJ001) (m³/s) (05BH004) (m³/s) same Date Elbow River (05BJ001) (m³/s) (05BJ001)1 (m³/s) 04/07/1991 357.0 399.0 29.8 19/05/1991 53.7 69.3 127 No 428.80 2 15/06/1992 275.0 289.0 111.0 15/06/1992 111.0 115.0 275 No 400.00 2 14/07/1993 219.0 245.5 49.5 02/06/1993 70.1 84.4 156 No 295.02 2 08/06/1994 251.0 285.0 64.1 07/06/1994 82.9 87.2 228 No 349.10 2 08/06/1995 452.0 499.0 123.0 07/06/1995 144.0 152.0 438 No 622.00 2 12/06/1996 280.0 282.0 11.5 04/06/1996 42.8 56.6 194 No 293.50 2 13/06/1997 275.0 278.0 36.9 01/06/1997 63.4 70.1 181 No 314.90 2 21/06/1998 293.0 331.0 77.4 23/05/1998 118.0 128.0 116 No 408.40 2 16/07/1999 342.0 391.0 48.6 15/07/1999 63.1 78.2 335 No 439.60 2 04/07/2000 173.0 222.0 7.41 08/06/2000 23.2 33.1 113 No 229.41 2 06/06/2001 167.0 187.0 42.6 05/06/2001 44.3 48.8 158 No 229.60 2 30/06/2002 339.0 341.0 30.1 11/06/2002 84.1 96.8 195 No 371.10 2 04/06/2003 220.0 238.0 15.9 26/05/2003 27.4 54.5 77.4 No 253.90 2 04/07/2004 207.0 238.0 22.4 27/08/2004 32.1 39.1 133 No 260.40 2 19/06/2005 602.0 791.0 251.0 19/06/2005 251.0 301.0 602 No 1,042.00 2 16/06/2006 244.0 255.0 108 16/06/2006 108.0 122.0 244 No 366.00 3 09/06/2007 331.0 360.0 38.4 06/06/2007 73.5 87.7 304 No 398.40 2 11/06/2008 287.0 322.0 103.0 25/05/2008 160.0 181.0 247 No 428.00 3 (a) Governing combination number is related to bulletins lists indicated in Section B4. March 2010 Report No. 09-1326-1040 B33 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT B6 RESULTS OF FLOOD FREQUENCY ANALYSES FOR THE BOW AND ELBOW RIVER Figures B9 to B17 provide frequency plots for the analyse completed to compare the flood flows derived in this study with those in the 1983 Calgary Floodplain study for the Bow River above its confluence with the Elbow River and Elbow River below Glenmore Reservoir. Figure B9: Flood Frequency Plot for the Bow River – Re-run of 1983 Study Data Using Bulletin 17B Method March 2010 Report No. 09-1326-1040 B34 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT Figure B10: Flood Frequency Plot for the Bow River – Re-run of 1983 Data with Method Used in 2010 Study for Incorporating Historic Floods 1000000 Naturalized, Historic and Generated Data 3 Parametes Log-Normal (Method of Likelihood) Extreme Value Type II Log Pearson Type III (Method of Likelihood) 100000 Q (m³/s) Q 10000 1000 2 10 20 50 100 1.05 1.25 200 500 1.003 1000 Return Period (Years) March 2010 Report No. 09-1326-1040 B35 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT Figure B11: Flood Frequency Plot for the Bow River – Naturalized and Historic Data up to 2008 Using Bulletin 17B Method March 2010 Report No. 09-1326-1040 B36 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT Figure B12: Flood Frequency Plot for the Bow River – Naturalized Data up to 2008 with Method Used in 2010 Study for Incorporating Historic Floods 10000 Naturalized, Historic and Generated Data 3 Parametes Log-Normal (Method of Likelihood) Extreme Value Type II Log Pearson Type III (Method of Likelihood) 1000 Q (m³/s) Q 100 2 10 20 50 100 1.05 1.25 200 500 1.003 1000 Return Period (Years) March 2010 Report No. 09-1326-1040 B37 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT Figure B13: Flood Frequency Plot for the Bow River – Natural and Regulated Data up to 2008 with Using Bulletin 17B Method March 2010 Report No. 09-1326-1040 B38 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT Figure B14: Flood Frequency Plot for the Elbow River – Re-run of 1983 Study Data Using Bulletin 17B Method March 2010 Report No. 09-1326-1040 B39 HYDROLOGY STUDY, BOW AND ELBOW RIVER UPDATED HYDRAULIC MODEL PROJECT Figure B15: Flood Frequency Plot for the Elbow River – Natural and Regulated Flood Data up to 2008 100000 Natural and Regulated Flow 3 Parametes Log-Normal (Method of Likelihood) Extreme Value Type II Log Pearson Type III (Method of Likelihood) 10000 Flood Flows (cfs) 1000 100 2 10 20 50 100 1.05 1.25 200 500 1.003 1000 Return Period (Years) March 2010 Report No. 09-1326-1040 B40 Golder Associates Ltd. 102, 2535 - 3rd Avenue S.E. Calgary, Alberta, T2A 7W5 Canada T: +1 (403) 299 5600