Libby VARQ Flood Control Impacts on Kootenay River Dikes

Author: B.G.C.Engineering Inc

The Columbia River Treaty Review Team (CRTRT), a branch of the Ministry of Energy, Mines and Natural Gas (MEM), has retained BGC Engineering Inc. (BGC) to evaluate the concerns of the local residents, who are amalgamated into a number of Diking Districts. The proposed scope of work is to investigate whether implementation of VARQ FC has had a significant negative impact on diking infrastructure adjacent to the Kootenay River between the Canada-US border and Kootenay Lake.

This study has investigated whether implementation of variable flow flood control (VARQ FC) has had a negative impact on diking infrastructure adjacent to the Kootenay River between the Canada-US border and Kootenay Lake. In this study reach, there are approximately 93 km of dike, which are maintained by five different diking authorities and the Lower Kootenay Band.
Prior to the construction of Libby Dam, major freshet floods caused extensive damage to the diking system (e.g. 1948). However, since Libby Dam operation commenced in 1973, average mean annual floods on the Kootenay River, as measured at the Canada-US Border, have decreased by more than a factor of two. Peak channel velocities also decreased by a factor of about 2 following dam construction.

Flow Ramping and Load Following
Conversely, bank erosion rates appear to have increased following dam construction, perhaps stabilizing in the late 1990s when flow ramping restrictions were instituted. Up until 1992, operation of Libby Dam was driven primarily by flood control and power needs. Flow ramping during the fall and winter months was a common practice, when the dam was operated to maximize hydroelectric power values. During load-following operations, it was not unusual to observe daily fluctuations in water level on the order of 0.5 to 1.0 m in the study reach. Similarly, daily fluctuations in average channel velocity on the order of 0.5 m/s also occurred.
In a 1999 study, NHC observed that a notch had developed along the Kootenay River banks, which they attributed to load following (fluctuations in dam releases that correspond to changes in power demand). NHC concluded that It is considered probable that the development of this notch is more pronounced now that the river level is controlled by Libby Dam in comparison to pre-Libby Dam, when the river level fluctuated over a wider range and the short duration releases from Libby Dam did not occur. The more limited range of water levels, greater fluctuations in flows during the winter season, and more frequent cycles of wetting and drying appears to induce a weakening of the banks resulting in toppling of soil wedges. The Corps (2006) have also concluded that past practices of load following at Libby Dam contributed to the erosion of the toe slope of much of the levee system in the Kootenai Valley, making the levees less stable.
Load following at Libby was common practice until the late 1990s, when it was realized that this practice may be having a significant impact on downstream fish habitat. Since the late 1990s, Libby Dam has operated with restricted flow ramping procedures. By prescribing maximum ramp rates from Libby Dam, there has been a considerable reduction in daily flow fluctuations in the study reach. These measures were implemented not only to protect resident fish and prey organisms in the Kootenay River, but also to help minimize dike/levee erosion along the river. Analysis of Kootenay River hydrographs shows that post-1999 hydrographs are more stable and the rapid water level fluctuations of the past have been eliminated. Therefore, it is not unrealistic to expect bank erosion rates to decrease in the future, as long as flow releases from Libby Dam continue to be managed for both fish habitat and bank erosion. This expectation is consistent with observations by the Corps who have noted that Kootenay River levees in the US are becoming stabilized by vegetation due to the curtailment of load following, especially daily fluctuations, since 2000 (Corps, 2006).
A lack of riparian vegetation was also likely a significant factor in the observed erosion. Many of the banks with protective dikes are either vegetated with grasses and shrubs only. In contrast where the bank is not protected by either riprap or a dike, the riparian vegetation is well established (with black cottonwoods in particular) providing a stabilizing influence against bank erosion.

Standard FC
Under Standard FC (1973-1992), Libby Dam would generally release high flows from January through April in order to increase storage capacity to capture the spring runoff in May, June and July. Because Lake Koocanusa drafted a large amount of water storage under Standard FC, Libby Dam historically released little water from May to July in order to refill. Eventually it was realized that this strategy was detrimental to several fish species. Between 1993 and 2002, Libby Dam was operated with increased flow releases during the freshet to accommodate different fish species, several populations of which have been listed for protection under the Endangered Species Act (ESA). In this transitional period Standard FC procedures continued to be used, but power operations were secondary to operations for fish.

VARQ FC
In December 2000, the USFWS and NOAA Fisheries each issued a Biological Opinion formalizing measures to protect endangered species including sturgeon, bull trout, salmon and steelhead. Recommended measures included implementation of VARQ FC at Libby Dam with the intent of ensuring reservoir refill in years when flood control allows it. As a consequence, VARQ should allow more assured provision of flows for endangered Kootenai River white sturgeon, threatened bull trout in the Kootenay and Flathead Rivers, and various listed stocks of salmon and steelhead in the Columbia.
With VARQ FC, the release during refill varies according to the reservoir level, water supply forecast and the estimated duration of flood control. VARQ FC is intended to provide a similar level of flood protection as Standard FC, but with improved flow augmentation for fish. Standard and VARQ Flood Control have the same storage space for flood control when the water supply forecast is greater than 120% of normal. In practice, there is only a difference between the two methods when the inflow forecast falls between 80% and 120% of normal (Corps, 2004). Within this range some of the water that would be stored during the refill period under Standard FC is instead passed through the dam under VARQ FC.
Using a calibrated hydraulic model, BGC has evaluated the difference in channel velocities between the Standard FC (1973-1992) and VARQ FC (2003-2012) periods. This assessment indicates that the velocity regimes between Standard FC and VARQ FC are not significantly different on a cumulative basis. On a seasonal basis, the peak monthly average velocity during freshet is about 30% higher when comparing VARQ FC (~0.65 m/s) and Standard FC (~0.5 m/s). This increase in velocity is a result of increased flow releases during the freshet to augment fish habitat.
However, peak flow velocities during VARQ FC remain well below the pre-Libby Dam period (~0.95 m/s). The pre-dam period is considered to be a better measure of typical shear stresses that induce meaningful channel changes (i.e. scour and bank erosion) along this section of the Kootenay River. It is therefore our opinion that the implementation of VARQ FC has not had a significant negative impact on diking infrastructure adjacent to the Kootenay River between the Canada-US border and Kootenay Lake.
In contrast, the past practice of load following did have a significant negative impact on diking infrastructure. Hence the curtailment of load following in the late 1990s has significantly reduced the erosion potential on the diking infrastructure.
It is also noted that peak freshet flows and channel velocities were well above average in 2011 and 2012. These above average flows may have resulted in some bank erosion along the study reach, increasing the perception that VARQ FC has had a negative impact on diking infrastructure. In the US, the Corps observed some levee damage along the Kootenai River in 2011 and 2012 (pers. comm., 2012). The Corps attributed this damage to a long duration snowmelt in 2011 and above average snowmelt in 2012 compounded by above average rainfall in June and July. In both years, these conditions resulted in saturated dikes that were more susceptible to erosion.

This study has investigated whether implementation of variable flow flood control (VARQ FC) has had a negative impact on diking infrastructure adjacent to the Kootenay River between the Canada-US border and Kootenay Lake. In this study reach, there are approximately 93 km of dike, which are maintained by five different diking authorities and the Lower Kootenay Band.
Prior to the construction of Libby Dam, major freshet floods caused extensive damage to the diking system (e.g. 1948). However, since Libby Dam operation commenced in 1973, average mean annual floods on the Kootenay River, as measured at the Canada-US Border, have decreased by more than a factor of two. Peak channel velocities also decreased by a factor of about 2 following dam construction.

Flow Ramping and Load Following
Conversely, bank erosion rates appear to have increased following dam construction, perhaps stabilizing in the late 1990s when flow ramping restrictions were instituted. Up until 1992, operation of Libby Dam was driven primarily by flood control and power needs. Flow ramping during the fall and winter months was a common practice, when the dam was operated to maximize hydroelectric power values. During load-following operations, it was not unusual to observe daily fluctuations in water level on the order of 0.5 to 1.0 m in the study reach. Similarly, daily fluctuations in average channel velocity on the order of 0.5 m/s also occurred.
In a 1999 study, NHC observed that a notch had developed along the Kootenay River banks, which they attributed to load following (fluctuations in dam releases that correspond to changes in power demand). NHC concluded that It is considered probable that the development of this notch is more pronounced now that the river level is controlled by Libby Dam in comparison to pre-Libby Dam, when the river level fluctuated over a wider range and the short duration releases from Libby Dam did not occur. The more limited range of water levels, greater fluctuations in flows during the winter season, and more frequent cycles of wetting and drying appears to induce a weakening of the banks resulting in toppling of soil wedges. The Corps (2006) have also concluded that past practices of load following at Libby Dam contributed to the erosion of the toe slope of much of the levee system in the Kootenai Valley, making the levees less stable.
Load following at Libby was common practice until the late 1990s, when it was realized that this practice may be having a significant impact on downstream fish habitat. Since the late 1990s, Libby Dam has operated with restricted flow ramping procedures. By prescribing maximum ramp rates from Libby Dam, there has been a considerable reduction in daily flow fluctuations in the study reach. These measures were implemented not only to protect resident fish and prey organisms in the Kootenay River, but also to help minimize dike/levee erosion along the river. Analysis of Kootenay River hydrographs shows that post-1999 hydrographs are more stable and the rapid water level fluctuations of the past have been eliminated. Therefore, it is not unrealistic to expect bank erosion rates to decrease in the future, as long as flow releases from Libby Dam continue to be managed for both fish habitat and bank erosion. This expectation is consistent with observations by the Corps who have noted that Kootenay River levees in the US are becoming stabilized by vegetation due to the curtailment of load following, especially daily fluctuations, since 2000 (Corps, 2006).
A lack of riparian vegetation was also likely a significant factor in the observed erosion. Many of the banks with protective dikes are either vegetated with grasses and shrubs only. In contrast where the bank is not protected by either riprap or a dike, the riparian vegetation is well established (with black cottonwoods in particular) providing a stabilizing influence against bank erosion.

Standard FC
Under Standard FC (1973-1992), Libby Dam would generally release high flows from January through April in order to increase storage capacity to capture the spring runoff in May, June and July. Because Lake Koocanusa drafted a large amount of water storage under Standard FC, Libby Dam historically released little water from May to July in order to refill. Eventually it was realized that this strategy was detrimental to several fish species. Between 1993 and 2002, Libby Dam was operated with increased flow releases during the freshet to accommodate different fish species, several populations of which have been listed for protection under the Endangered Species Act (ESA). In this transitional period Standard FC procedures continued to be used, but power operations were secondary to operations for fish.

VARQ FC
In December 2000, the USFWS and NOAA Fisheries each issued a Biological Opinion formalizing measures to protect endangered species including sturgeon, bull trout, salmon and steelhead. Recommended measures included implementation of VARQ FC at Libby Dam with the intent of ensuring reservoir refill in years when flood control allows it. As a consequence, VARQ should allow more assured provision of flows for endangered Kootenai River white sturgeon, threatened bull trout in the Kootenay and Flathead Rivers, and various listed stocks of salmon and steelhead in the Columbia.
With VARQ FC, the release during refill varies according to the reservoir level, water supply forecast and the estimated duration of flood control. VARQ FC is intended to provide a similar level of flood protection as Standard FC, but with improved flow augmentation for fish. Standard and VARQ Flood Control have the same storage space for flood control when the water supply forecast is greater than 120% of normal. In practice, there is only a difference between the two methods when the inflow forecast falls between 80% and 120% of normal (Corps, 2004). Within this range some of the water that would be stored during the refill period under Standard FC is instead passed through the dam under VARQ FC.
Using a calibrated hydraulic model, BGC has evaluated the difference in channel velocities between the Standard FC (1973-1992) and VARQ FC (2003-2012) periods. This assessment indicates that the velocity regimes between Standard FC and VARQ FC are not significantly different on a cumulative basis. On a seasonal basis, the peak monthly average velocity during freshet is about 30% higher when comparing VARQ FC (~0.65 m/s) and Standard FC (~0.5 m/s). This increase in velocity is a result of increased flow releases during the freshet to augment fish habitat.
However, peak flow velocities during VARQ FC remain well below the pre-Libby Dam period (~0.95 m/s). The pre-dam period is considered to be a better measure of typical shear stresses that induce meaningful channel changes (i.e. scour and bank erosion) along this section of the Kootenay River. It is therefore our opinion that the implementation of VARQ FC has not had a significant negative impact on diking infrastructure adjacent to the Kootenay River between the Canada-US border and Kootenay Lake.
In contrast, the past practice of load following did have a significant negative impact on diking infrastructure. Hence the curtailment of load following in the late 1990s has significantly reduced the erosion potential on the diking infrastructure.
It is also noted that peak freshet flows and channel velocities were well above average in 2011 and 2012. These above average flows may have resulted in some bank erosion along the study reach, increasing the perception that VARQ FC has had a negative impact on diking infrastructure. In the US, the Corps observed some levee damage along the Kootenai River in 2011 and 2012 (pers. comm., 2012). The Corps attributed this damage to a long duration snowmelt in 2011 and above average snowmelt in 2012 compounded by above average rainfall in June and July. In both years, these conditions resulted in saturated dikes that were more susceptible to erosion.