Grand Coulee Dam and the Columbia Basin Project usa final Report: November 2000

 
3.3 Flood Control 
 
3.3.1 Predicted Benefits 
 
Flood control provided by GCD is best viewed over two different time periods: the time of project 
authorisation through to 1973, and after 1973, when the Columbia River Treaty projects came on line.
48
 
Despite the fact that the Rivers and Harbors Act of 1935, which authorised Grand Coulee Dam, begins “ 
. . . for the purpose of controlling floods” (US Congress, 1935), in the earliest project planning 
documents, it appears that flood control at GCD was only a minor consideration until after a disastrous 

Grand Coulee Dam and Columbia Basin Project 
 
         41 
 
This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. The views, conclusions, and 
recommendations contained in the working paper are not to be taken to represent the views of the Commission 
 
1948 flood. A recent Reclamation study concluded that the project’s anticipated flood control benefits 
were small and incidental to other operations (USBR (Economics Group), 1999: 20).  
 
Flood control was treated only briefly in the Butler Report, and Major Butler, in his testimony before the 
US House Committee on Irrigation and Reclamation, stated that “navigation and flood control are not 
important on this section of the river” (USACE, 1933: 867–71; US Congress, 1932: 17). One of the 
major oversights of the Butler Report relates to the way flood control was handled. The report states that 
the “valley of the Columbia within the United States above the Snake is not subject to inundation” 
(USACE, 1932: 867). However, the report did not consider the effect of upstream flood control on 
reducing flood damage at points below the confluence with the Snake River. This point was considered 
by Major Kuentz who produced a companion Corps of Engineers report on the development of the 
Columbia River below the mouth of the Snake River. While Major Kuentz recognised the significance of 
flooding in the lower Snake, he concluded that flood control was not important enough to be a major 
factor in the Corps’ plan for water resources development. 
 
In examining predicted flood control benefits, we looked to the Butler and Reclamation reports, but 
neither report treated this subject in depth. Butler’s discussion of potential flood control benefits 
concentrates on the risk of minor flooding at upstream locations on the tributaries, not on the mainstream 
Columbia (USACE, 1933: 867–71). The 1932 Reclamation Report makes no mention of flood control
. 
Little formal data exists on the projected flood control benefits of GCD for three reasons. First, both the 
Butler and Reclamation reports minimise the significance of flood control. Second, the project was 
authorised during a time when benefit-cost analyses were not required for federal water projects. Third, 
prior to the 1948 flood, it was generally believed that adequate flood control could be achieved solely 
through the construction of levees (USBR (Economics Group), 1999: 20). 
 
A local newspaper article from 1937 estimates that annual flood control benefits were $5 million (about 
$60 million in $1998) and dam backers said that the dam would end downstream flooding (Pitzer, 1994: 
258). An article in an issue of Atlantic Monthly in 1936 stated that GCD would control floodwaters 
“down the Columbia to 450 miles (724km) to the sea”. We have found no information on projected flood 
control benefits that is more quantitative or more carefully detailed. Documents we have reviewed, as 
well as our consultations with flood control personnel at the Corps of Engineers Portland Division 
Office, suggest that little, if any, detailed information exists on the pre-project estimates of monetary 
flood control benefits provided by GCD. The argument that local project proponents did not view flood 
control as a major purpose of GCD during the late 1920s and early 1930s is supported by Reclamation 
correspondence from then Commissioner Floyd Dominy, dated 1961, which indicates that, in fact, the 
dam was “not designed specifically for flood control operation” (Dominy as quoted by Pitzer, 1994: 
260). The estimated cost of flood control works relative to the project as a whole was modest. According 
to historian, Paul Pitzer, “when the government allocated the total cost of the dam to its various 
components, the amount indicated for flood control and navigation combined amounted to only $1 
million” (about $12 million in $1998) (Pitzer, 1994: 258). 
 
3.3.2  Flood Control Operations Before 1973 
 
During its early years of operation, GCD was not managed with flood control as a priority. Even though 
the storage capacity of the project (over 5 million acre feet, or 6 170 x 10
6
 m
3
) seems large, it is only a 
small percentage of the capacity needed to control the flow of the Columbia (Brooks 1999). During the 
first few decades of operation, the level of the Lake Roosevelt was generally kept high to maximise 
power production and the potential to provide irrigation water. The limited ability of GCD to control 
floods became obvious when it was unable to prevent severe flooding in 1948. That year, the peak flow 
at the dam was 638 000 cfs (18 100 m
3
/sec) (USBR, 1949). 
 
In the wake of the 1948 flood, the city of Vanport, then the second largest city in Oregon, was 
completely destroyed. The disaster caused over $100 million ($730 million in $1998) in property 
damage and killed 51 people (USBR, 1949). The enormous amount of damage from the 1948 flood 
demonstrated that the projects with flood control purposes in place at that time were not capable of 

Grand Coulee Dam and Columbia Basin Project 
 
         42 
 
This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. The views, conclusions, and 
recommendations contained in the working paper are not to be taken to represent the views of the Commission 
 
handling large flows. Afterwards, efforts at basin-wide flood control efforts (along with basin-wide 
hydropower management) received heightened attention and Congressional support (Pitzer, 1994: 334; 
Brooks 1999).  
 
3.3.3  Flood Control Operations After 1973 
 
From 1948 to 1973, significant improvements were made in flood control operations at GCD and other 
multi-purpose dams were brought on line to provide additional system-wide storage capability. The key 
event that secured flood control protection for the basin was the construction of the Columbia River 
Treaty projects. These projects, which became operational between 1967 and 1973, basically doubled the 
previously existing storage capacity of the system (USDOE et al. (Appendix E), 1995: 2-2). As detailed 
in Section 6, the construction of these projects followed ratification of the Columbia River Treaty 
between the US and Canada. These dams became fully operational in 1973. The Columbia River Treaty 
projects were able to store water and control releases in ways that greatly enhanced the generation of 
hydropower and the provision of flood control benefits basin-wide (For a detailed discussion of flood 
control operations in the Columbia River Basin, please see the Annex titled “System Operations — 
Hydropower, Flood Control, and Anadromous Fish Management Activities”)
. 
 
Most of the system storage capacity on the Columbia River actually resides in Canada. Out of 37 MAF 
(45 600 x 10
6
 m
3
) of available system storage, 43% (16 MAF) is provided by US facilities. GCD, with 
the largest reservoir on the US side, is the system’s main storage dam, accounting for about one-third of 
the total US storage capacity and 13% of total US and Canadian storage. The dam is the gateway for 74 
000 square miles (19 x 10
6
 ha) of river drainage and is operated for flood control in conjunction with 13 
other dams on the main-stem Columbia (USACE, 1991). Lake Roosevelt is the last major catchment for 
flood control on the river. Since the early 1970s, the combined operation of dams in the US and Canada 
has served to significantly reduce the potential of downstream flooding.
49
  
 
Although Reclamation had responsibility for constructing and operating GCD, under the Flood Control 
Act of 1944, the Corps has responsibility for specific operations concerning flood control (Brooks 1999). 
Spring rains and melting snow generate large volumes of runoff. In anticipation of these runoff events, 
the level of Lake Roosevelt is lowered to accommodate high river flow, thereby reducing the risk or 
minimising the impact of downstream flooding. The Corps regulates the level of Lake Roosevelt based 
on daily, weekly, and monthly forecasts of upcoming river runoff using a storage reservoir diagram. The 
diagram defines how much storage must be available in any project in the system at any given time 
under a specific set of flow conditions (See Annex titled “System Operations—Hydropower, Flood 
Control, and Anadromous Fish Management Activities” for GCD’s storage reservoir diagram). 
 
The Corps estimates that monetary benefits of basin-wide flood control since 1973 have been 
substantial. For example, Corps calculations indicate that the value of flood control operations basin-
wide for particularly wet years has been on the order of hundreds of millions of dollars (USACE, 1999). 
In 1974, the flood control capacity of the system was put to the test when a flood of roughly the same 
magnitude as 1948 occurred. This event was regulated sufficiently by upstream storage to below major 
damage levels (USDOE et al. (Appendix E), 1995: 2-1). Corps estimates for flood control benefits 
provided by the entire Columbia River system
50
 since 1973 range from $500 000 (in 1973) to $378 
million (in 1997) (USACE, 1999) (See Annex titled “System Operations—Hydropower, Flood Control, 
and Anadromous Fish Management Activities” for details on flood control benefits estimate). According 
to Reclamation records (USBR, 1998c), project costs allocated to both flood control and navigation
51
 
account for less than 3% of the total project cost. 
 
 
Interestingly, there are differing opinions concerning the validity of models used to derive flood control 
benefits approximations for the system. According to staff we interviewed at the Corps, the benefit 
estimates are likely to be low, since a formal empirical study of the flood prone area has not been 
undertaken since 1969 (Brooks 1999). According to other analysts, estimates may be inflated because of 
overestimation of potential flood risks and estimates of flood damages used in the benefits estimates 
methodology (National Research Council et al, 1995). More generally, environmental groups, such as 

Grand Coulee Dam and Columbia Basin Project 
 
         43 
 
This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. The views, conclusions, and 
recommendations contained in the working paper are not to be taken to represent the views of the Commission 
 
the Sierra Club argue that the Corps’ benefits estimates are inflated because calculations of damages 
wrought in areas where Corps projects promoted development in floodprone areas that are later 
inundated are not included in estimations of a project’s benefits and costs (Schildgen, 1999).  
 
The exact flood control benefits directly attributable to GCD are difficult to gauge because the storage 
capacities of all the major reservoirs that provide flood control are linked and operated as a system.
52
 
However, a 1999 Corps study of the annual flood control benefits derived from GCD storage puts the 
amount to be $20 200 000 in $1998 (USBR (Economics Group), 1999: 20). 
 
3.3.4  Unexpected Flood Control Benefits, Costs, and Impacts 
 
The main flood control operations at GCD are: (i) drawing down reservoirs in the winter to provide 
enough space to store runoff; and (ii) filling reservoir space during the spring in a strategic manner to 
minimise downstream damages (USDOE et al. (Appendix E), 1995: 2-3). These activities are closely 
tied to the operation of the project for hydroelectric and irrigation purposes. Therefore, it is not possible 
to link unanticipated impacts of reservoir operations to flood control operations. The major unintended 
impacts of general project operations are discussed in Sections 3.4, 3.5, and 3.7 of the report and include 
adverse effects on ecosystems (particularly anadromous fish), recreational benefits, and social effects.  
 
A phenomenon often associated with increased flood protection is increased settlement in floodplains. In 
the Columbia River system, the main damage potential is in the lower Columbia, which includes 
portions of the cities of Portland, Oregon and Vancouver, Washington, as well as extensive rural areas 
protected by diking districts (USDOE et al. (Appendix E), 1995: 2-4). The literature review we 
conducted did not yield data or studies that would allow definitive statements to be made as to whether 
increased flood protection in the lower Columbia River has led to a population increase in the floodplain.  
 
While critics of federal flood control projects have not specifically attacked projects in the Columbia 
River Basin, they have criticised the Corps and other agencies that construct physical flood control 
structures (eg, dams and levees). They argue that because current federal programmes focus on structural 
alternatives to flood control, these programmes encourage development in floodplain areas and facilitate 
resettlement in floodprone areas after a disaster occurs (Schildgen, 1999; Mount, 1995). Some analysts 
of flood plain management posit that people commonly view flood control works as being able to 
prevent floods, and this mis-conception
53
 has led to increased urbanisation behind levees (Schildgen, 
1999; Haeuber & Michener, 1998). In fact, levees and other structural flood control works sometimes 
fail. 
 
Additionally, there is a growing recognition that extensive structural flood control works affect natural 
ecosystems adversely by disrupting natural patterns of ebb and flow (Sparks, 1995). There is also some 
evidence that extensive structural works may even increase the stage of large floods, leading to higher 
damages associated with these rare events (Mount, 1995). 
 
3.4  Recreation and Tourism 
 
3.4.1 Predicted Benefits 
 
GCD and CBP have generated a number of recreational benefits, such as the creation of wetland and 
riparian corridors, although project planners of the 1920s and 1930s did not anticipate them. Neither the 
Reclamation nor the Butler reports mentioned recreation. The 1935 Rivers and Harbors Act, which 
authorised the Grand Coulee Dam, did not list recreation as a primary project purpose. However, a 
statement in the Act that lists “providing other beneficial uses” as a purpose of the project has, over time, 
been interpreted to include recreational activities (NPS, 1998: 169). The Columbia Basin Joint 
Investigations anticipated the recreational potential of Lake Roosevelt, and the report on “Problem 26” 
of the Investigations, completed in 1945, deals with this issue. The Joint Investigation committee 

Grand Coulee Dam and Columbia Basin Project 
 
         44 
 
This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. The views, conclusions, and 
recommendations contained in the working paper are not to be taken to represent the views of the Commission 
 
studying recreation in Lake Roosevelt saw multiple potential recreational uses of the area, including 
boating, fishing, picnicking, bathing, camping, hunting, and backpacking (USBR, 1945c: 4-6).  
 
The Columbia Basin Joint Investigations report committee investigating Problem 26 noted that pleasure 
boaters were using the reservoir even though it was only partially filled (USBR, 1945c: 4). A fish 
hatchery built to replace salmon losses was expected to create fishing opportunities, and the reservoir 
was predicted to create increased upland game and migratory bird populations for hunting (USBR, 
1945c: 4-5). The committee identified several locations where summer home communities could be 
built, and it discussed many educational opportunities related to history, archaeology, geology, and plant 
and animal life (USBR, 1945c: 7-10). The report noted that GCD received an average of 325 000 visitors 
annually before World War II, of which 40% were from outside Washington state. Noting the constantly 
changing conditions that could alter its projections, the committee predicted levels of future use of the 
reservoir. It foresaw a rise in visits to GCD with the end of World War II, including 200 to 400 pleasure 
boats with hundreds more smaller boats using Lake Roosevelt after peacetime, and a peak Sunday 
attendance for picnicking, bathing, and water sports of around 4 000 to 5 000 visitors (USBR, 1945c: 26-
7). The committee expected these activities to increase as population increased in the surrounding areas. 
While no specific projections were made of potential visits to the project area, the report states “It . . . 
seems reasonable to conclude that, on the basis of present trends and recreational habits of people in the 
district, new recreational areas to provide from a capacity of at least 40,000 to 80,000 additional people 
will be necessary during the next two decades” (USBR, 1945c: 26). 
 
3.4.2  Actual Recreational Facilities and Use 
 
Recreational facilities created by GCD and CBP can be broadly divided into two categories: (i) facilities 
associated with the Lake Roosevelt National Recreational Area (LRNRA), which includes GCD; and (ii) 
facilities associated with CBP’s irrigation command area. These two types of facilities are administered 
separately. The National Park Service (NPS) and the Colville and Spokane tribes manage recreational 
activities associated with the National Recreation Area; Reclamation operates recreational activities 
related to the dam site; and a variety of state and federal agencies manage recreational activities within 
the irrigation command area.
54
 In 1996, total visits to all these facilities were estimated to exceed three 
million (Olsen, 1996). (See Annex titled “Use Statistics for GCD and CBP-related Recreational 
Facilities” for a list of some major recreational facilities.) Generally, recreational activity is highest 
during the summer. For example, visitation between June and September accounts for about three-
quarters of total annual visits to the LRNRA (NPS, 1998: 68). 
 
Lake Roosevelt National Recreation Area 
 
LRNRA includes Lake Roosevelt, GCD, and surrounding areas. Lake Roosevelt is the largest 
recreational facility in LRNRA, extending 151 miles (243km) from the dam site to the Canadian border. 
It houses over 30 campgrounds, 10 swimming beaches, 18 boat ramps, and 28 boat docks. Six major 
facilities operated by the Colville Indian tribe offer houseboat and fishing boat rentals, fuel, and food. A 
25-mile (40km) stretch of the shoreline along the Spokane River is also included in LRNRA. 
 
Since 1980, the number of recreational visitors to LRNRA has increased from 800 000 visits in 1980 to 
1.4 million in 1997 (NPS, 1998: 67). A 1996 visitor use survey found that most respondents (62%) were 
between 15 and 44 years of age. Almost three-quarters of survey respondents (74%) were from the state 
of Washington, and about 13% were from Canada. Only about 7% of the respondents were US residents 
visiting from other states, and only 1% of respondents were from outside the US (NPS, 1998: 72). The 
most common purposes of visits to LRNRA are camping, swimming, motor boating, and fishing (NPS, 
1998: 72). 
 
LRNRA also offers a guided tour of the dam, the Third Powerplant, and the pump-generating plant. The 
area houses a visitors centre, where tourists can watch a video that explains the history of the dam and 
view various educational displays. In 1994, almost half a million people visited the dam (Olsen, 1996). 

Grand Coulee Dam and Columbia Basin Project 
 
         45 
 
This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. The views, conclusions, and 
recommendations contained in the working paper are not to be taken to represent the views of the Commission 
 
One of the most popular attractions at the dam is a 36-minute laser light show featuring animated 
graphics on the downstream surface of the dam. The show is put on every evening during the summer 
and routinely attracts large crowds at the dam site and at surrounding areas from which the spillway is 
visible. According to Reclamation staff we interviewed at the dam, the majority of attendees at the light 
show were from Washington State, with the remaining show visitors coming from other US states 
(Sprankle 1999d). 
 
3.4.2.1  Irrigation Command Area 
 
In addition to LRNRA, there are numerous recreational opportunities in the CBP area, and many of the 
recreation sites were created as a result of the application of irrigation water. Some of these recreation 
sites have been extensively developed. In total, there are 143 785 acres (58 188ha) of land and water 
available for recreational use that rely on CBP water (USBR, 1992b). There are six Washington State 
parks and 32 other major state and federally managed recreation facilities in the project area (USBR, 
1989a: III-299-300). 
 
 
In 1992 there were more than 1.34 million public visits to CBP-related recreation facilities south of 
Grand Coulee Dam (USBR, 1992b). There are several major fishing areas, such as Banks Lake, where 
about 120 000 sport fish were caught in 1992, and Potholes Reservoir, where 531 000 sport fish were 
caught in 1992 (USBR, 1992b). There are numerous smaller sport fishing locations. Hunting is also a 
popular recreational activity. In 1992, nearly 60 000 ducks were taken in the project area; pheasant and 
geese are also heavily hunted (USBR, 1992b). 
 
3.4.2.2 Benefits 
Estimate 
 
Few comprehensive studies exist that focus specifically on recreational benefits in the project area. 
However, a 1996 study by Olsen
55
 sponsored by the CBP irrigation districts estimates that the total direct 
net value (ie, economic value derived from primary activities minus the cost of provision), and 
secondary economic benefits of recreational purposes served by the project area. In Olsen’s analysis, the 
total direct net value of recreational benefits was between $23 and $77 million in $1998. This estimate 
does not include all forms of flat-water recreational activity such as motor-boating. In addition, Olsen 
estimated secondary economic benefits (eg, lodging, food, gas) to be about $24 million annually in 
$1998. Recreation-related business is an economic mainstay for many communities located on or near 
the major reservoirs, including Kettle Falls, Davenport, Coulee Dam, Grand Coulee, and Electric City 
(Lebret 1999; Halsey et al. 1999; NPS, 1998: 62).