SUMMARY REPORT
NITROGEN SUPERSATURATION
in the
COLUMBIA and SNAKE RIVERS
JULY, 1971
ENVIRONMENTAL
PROTECTION
AGENCY
REGION X
SEATTLE, WASH.
-------
4162
SUMMARY REPORT
NITROGEN SUPERSATURATION IN THE
COLUMBIA AND SNAKE RIVERS
TECHNICAL REPORT NO. TS 09-70-208-016.1
ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF WATER PROGRAMS
SEATTLE, WASHINGTON
REGION X
July, 1971
-------
CONTENTS
I.
II.
II
IV
V."
INTRODUCTION
EFFECTS ON FISH
CAUSES .
CONCLUSIONS . . .
RECOMMENDATIONS
Page
1
1
2
6
9
-------
INTRODUCTION
Evidence strongly indicates that the future of salmon and
steelhead in the Columbia and Snake Rivers is seriously jeop-
ardized by high levels of nitrogen supersaturation in these
waters during spring fish migration periods. Under present
conditions, toxic concentrations of dissolved nitrogen per-
sist from the most upstream dams to the mouth of the Columbia
River during periods of spill.
The anadromous fish of the Columbia Basin, a major regional
and national resource, could be reduced to 10 percent of its
present size within three years. The urgency of this situation
requires that immediate action be taken by Congress and Federal
agencies to implement an aggressive strategy for an early and
effective reduction of dissolved nitrogen concentrations in
the Columbia and Snake Rivers. Also, since this reduction may
take several years to accomplish, a concurrent program of
interim measures to reduce the exposure of migrating salmonids
to the existing high nitrogen levels must also be established.
Nitrogen supersaturation has been shown to be a toxicant
which causes mortality to aquatic life and as such, violates
the Interstate Water Quality Standards criteria adopted pursuant
to the Water Quality Act of 1965. In accordance with the res-
ponsibilities transferred to the Environmental Protection Agency,
as set forth in Executive Order 11507 of February 4, 1970,
Mr. William D. Ruckelshaus, Administrator, directed the Northwest
Regional Office of the Environmental Protection Agency to pre-
pare a report outlining the scope of the problem in the Columbia
Basin with recommendations for its solution.
The objective of this report is to summarize the major
findings and recommendations of a more detailed technical
report (available on request from Environmental Protection
Agency, Region X). Courses of action are explored to implement
the short- and long-term solutions to the nitrogen supersatura-
tion problem.
EFFECTS ON FISH
Nitrogen supersaturation and its effect on fish (gas bubble
disease) has been known since before 1900 but has primarily
been associated with fish hatchery operation. Only in recent
-------
years have the mysterious disappearance of adult salmon and
steelhead between the dams and the losses of juvenile downstream
migrants been associated with high levels of dissolved nitrogen.
The relationship between migrant fish mortalities, nitrogen
supersaturation, and reservoir spillage has now been well docu-
mented by field observations and laboratory studies. Nitrogen
supersaturation and resultant gas bubble disease were implicated
in an estimated loss of 20,000 summer chinook salmon between
Bonneville Dam and McNary Dam on the Columbia River in 1968
(see Columbia River Basin Map, Figure 1). Similarly, high
spillway discharge at Lower Monumental and Little Goose Dams
was the primary factor in an estimated 70 percent loss of Snake
River downstream migrants during the spring of 1970. Dams
have converted the Columbia River from a flowing stream into
a series of slow moving reservoirs, which retard the out-
migration of juvenile salmonids to the ocean and thereby
subject them to increased stresses, including temperature,
diseases, predation, and high nitrogen levels.
CAUSES
Nitrogen supersaturation in the waters of the Columbia
and Snake Rivers is caused when river flow must be passed over
the spillways of the main stem hydroelectric dams. Air, which
is 79 percent nitrogen, is entrained in the flow as it passes
over the spillways and, as the water plunges into the deep
stilling basin below the dams, the increased pressure and
turbulence force the -gasses in the air to dissolve in the
water. Under these conditions the water may contain dissolved
nitrogen and other gasses much in excess of surface equilibrium
concentrations. Nitrogen concentrations up to 147 percent
of surface equilibrium values have been observed below spill-
ways under these conditions. Water passing through the power-
generating turbines at the dams, or through other non-air-
entraining facilities, does not increase in nitrogen content.
In a free-flowing river, with normal turbulence, the super-
saturated water would soon equilibrate with the atmosphere
and tend to produce a solution in which concentrations of
dissolved gasses are roughly 100 percent of their normal
saturation values. The construction of dams on the Columbia
and Snake Rivers, however, has converted these rivers from
free-flowing streams into a series of reservoir pools, reducing
mixing and exposure of the liquid to the atmosphere, and
retarding the dissolution of the gasses and return of the
supersaturated water to equilibrium.
-------
COLUMBIA
RIVER BASIN
f =. MAJOR RESERVOIR PROJECTS
*-—•' UNDER CONSTRUCTION
Anadromous Fish
Spawning Area
EXISTING MAJOR PROJECTS
Grand
Coulee
ky Reach SPOKANE
anopum
Little
---. Lower Coot* f—">
•^n \Monuirmaal ..2K* Lower ,A (
^_ .. aniie '
HANFORD WORKS"--
RESERVATION
N V.
Figure 1. Columbia River Basin Map
-------
In some cases, the series of dams can have an accumula-
tive effect, increasing nitrogen levels at each successive dam
downstream. In the more general case, the effect of most dams
is to reinforce and maintain a supersaturation level, thereby
offsetting the limited amount of equilibration which occurs
between dams.
Spilling occurs at a hydroelectric project when total
river flow exceeds the volume which can be passed through the
turbine generator units to produce power. Therefore, during
the periods of high runoff in the spring and early summer,
spilling often occurs for prolonged periods at most of the
dams in the system. Even during periods of lower flows,
spilling at many projects occurs because of the differences
in hydraulic capacities of turbines among the various main stem
hydroelectric projects. There is relatively little usable
storage capacity available in most of these "run-of-the-river"
projects on the Columbia and Snake Rivers. Consequently,
flows past one dam are roughly equal to the release at the
first upstream project plus any tributary inflow between the
two. This means, for example, that if Wells Project (Douglas
County Public Utility District) on the middle Columbia, were
passing flows through its turbines equal to their capacity
of 230,000 cubic feet per second (cfs), Rock Island (Chelan
County Public Utility District) downstream, with a present
turbine capacity of only 80,000 cfs, would have to pass at
least 150,000 cfs over its spillway.
Regional power demands on the network of generating fac-
ilities also may act to limit the amount of flow which can be
passed through turbines and thereby affect the amount of
spilling required. With the present hydraulic design of tur-
bone generator units in the Columbia system, it is not practical
to allow them to pass water without generating power. Passing
flow through these units with low loads produces cavitation
and potential damage to turbines and water passageways. Oper-
ating units in this manner has also been shown to produce high
mortality rates in downstream migrant fish which pass through
the units. It is the total energy load then—not the hydraulic
capacity-- that presently determines how much water can be
passed through the turbines. Unfortunately, the annual period
of lowest power demand coincides with the spring-summer period
of highest river flows and peak fish migration. Recognizing
this, the Corps of Engineers and Bonneville Power Administration
are currently conducting studies to devise methods of increasing
-------
power exports and operating turbine-generator units at very
low loads.
Storage in reservoirs located principally in headwater
areas of the Columbia and Snake Rivers is used to regulate
downstream flows. However, even with the addition to the
system of some 19 million acre-feet of active storage in Mica,
Libby, and Dworshak Reservoirs (now under construction), the
average-year floods in the Columbia Basin will still require
.spill at some projects. Significant spilling and high nitrogen
levels can be expected in the future, assuming the reservoirs
and power generation facilities are operated within existing
constraints. The use of the available reservoir storage
specifically for reduction of spilling and nitrogen levels
during the spring and early summer flood flow period is gen-
erally compatible with use of this storage for power generation,
flood control, navigation, and irrigation. Flexibility in the
operation of the reservoir system was demonstrated through
an unprecedented cooperative effort of the Corps of Engineers,
Bonneville Power Administration, state and federal fisheries
agencies, and public utility and private hydroelectric project
operators, in a program to reduce spilling during a five-day
period in April of this year to improve nitrogen conditions
during a massive release of hatchery-reared juvenile chinook
salmon and steelhead trout. Additionally, juvenile chinook
salmon and steelhead trout migrants were captured at Little
Goose Dam in an experimental program and transported downstream
by truck for iel ease below Bonneville Dam to demonstrate the
feasibility of this method to reduce exposure of the fish to
excess nitrogen and help minimize losses of the Snake River
run. Limitations on the flexibility of the reservoir and power
system include economic losses due to loss of some power
revenues, and costs from potential increased flood damage.
Legal complications include certain conditions imposed on the
operation of reservoir storage in Canada under the terms of
the Columbia River Treaty of 1964, contractual agreements for
operation of public utility and private hydroelectric projects,
and commitments in Bonneville Power Administration contracts
for power distribution and marketing.
The Corps of Engineers, National Marine Fisheries Service,
Bonneville Power Administration, and Oregon and Washington
fisheries agencies have worked diligently during the past year
toward the development of short-term and long-range solutions
to the problem of nitrogen supersaturation. Research and
development efforts have been concentrated in the areas of
operational and structural modifications to the regional
-------
reservoir and power generation system to reduce spills and
nitrogen concentrations; modifications in hatchery release
schedules; capture and transport of migrants around reaches
of critical nitrogen levels; cause-effect relationships between
spillage, dissolved nitrogen concentrations, and gas bubble
disease, and monitoring and analytical procedures for measuring
and predicting nitrogen supersaturation levels. The Corps of
Engineers has developed and successfully tested on prototype
scale, a scheme for passing an additional 60,000 cfs through
empty turbine bays in the Lower Snake River dams. This will
reduce considerably the amount of spills and nitrogen entrain-
ment at those projects. An additional spillway modification
(flip bucket) that could be added to an existing project
shows promise for reduction of nitrogen. The effect of this
device would be to prevent low to moderate spillway flows
from plunging deeply into the stilling basin.
However, even if the operational and structural measures
demonstrated to be feasible to date could be fully implemented,
the nitrogen levels in years of average or higher flows will
continue to be high enough to constitute a serious hazard to
juvenile and adult anadromous fish migrating in the Columbia
and Snake Rivers. To insure the safety of this major fishery
resource, an effective strategy for reduction of nitrogen levels
to safe concentrations must be developed and implemented, along
with measures to minimize exposure of migrant fish to the
existing and future high levels of nitrogen. The following
conclusions and recommendations are pertinent to this strategy
and are supported in the technical report.
CONCLUSIONS
1. Essentially all reaches of the Columbia and Snake Rivers
downstream from hydroelectric dams are significantly super-
saturated with dissolved atmospheric gasses during periods of
high spillway discharge.
2. The salmon and steelhead runs in the Columbia and Snake
Rivers are seriously jeopardized under present conditions by
supersaturation of dissolved gasses.
3. Nitrogen supersaturation conditions in the Columbia
and Snake Rivers are caused at main stem hydroelectric dams
when river flow must be passed over the spillways during the
spring flood period. Nitrogen in the air is entrained in the
-------
flow as it passes over the spillway and, as the water plunges
into the stilling basin below the dam, increased pressure
forces the nitrogen from a gaseous state into solution in the
water.
4. Water passing through the power-generating turbines
or otherwise not exposed to the aerating effect of the spill-
ways does not become supersaturated with nitrogen.
5. Nitrogen supersaturation levels above 105 percent
produce symptoms of gas bubble disease in fish, and levels
above 120 percent are lethal.
6. Nitrogen, because of its greater proportionate
volume in air, has been measured and related to toxic effects
on fish. However, it is known that the other atmospheric
gasses are also involved in the gas bubble disease phenomenon.
7. Hypothetical flow regulation studies show that, with
no operational or structural modifications for control of
nitrogen, spillway flows at existing dams will be great enough
to produce lethal levels of nitrogen supersaturation in years
of average and higher river flows, but not in low flow years.
An average year is one with flows which, statistically, would
be equalled or exceeded one year out of two. Although it is
not certain at what specific river flow levels nitrogen problems
begin to occur, it is apparent that the threshold level is
somewhere between low and average flows.
8. Under present plans to expand the Columbia Basin hydro-
electric system through the year 1980, the volume of spills
at the various projects will be reduced. However, without
additional control measures, the reduction in the'volume of
spills will not be great enough to reduce nitrogen supersatura-
tion to levels considered safe for fish during even a year of
average flows.
9. Installation of slotted gates and bypass of flow through
empty turbine bays in the Lower Snake River dams will further
reduce the spills in that river, but nitrogen levels hazardous
to fish will still occur in years of average or higher flows.
10. A basic requirement to a solution of the problem is
to devise methods whereby volume in excess of flows through
turbines can be passed by dams without entraining atmospheric
gasses.
-------
8
11. Lower Granite Dam, presently under construction, will
add to already critical nitrogen problems in the lower Snake
River unless the completed project includes provisions for
passing flood flows without entraining atmospheric gasses.
12. Unless provisions can be made to pass maximum flows
through turbines, installation of turbine generator units in
dams before they are needed to meet regional power demands may
eliminate flexibility in the use of the otherwise empty turbine
bays to bypass flow,- and thus may aggravate the nitrogen problem.
13. Increasing use of the hydroelectric system for peak
power generation could cause wide diurnal fluctuations in
the amount of spilling unless pumped storage, pondage, and
thermal power facilities are effectively used to reduce the
fluctuations.
14. The critical nature of the problem in 1971 caused an
unprecedented level of coordination and cooperation among the
Corps of Engineers, Bonneville Power Administration, and the
Northwest fishery agencies in developing partial immediate
solutions to the problem.
15. Flexibility within the federal dam-power complex
contributed much to the success of a flow reduction program
in April 1971 to make the best conditions for a mass hatchery
fish release. Although the program required increased
coordination, the cost to the Government was minimal. Con-
tinued annual use of the flexibility of the federal system
is a necessary partial solution to the nitrogen problem.
16. The Washington Public Utility Districts and Idaho
Power Company cooperated in the nitrogen reduction program
in 1971, but even greater interchange between the federal and
non-federal systems should be permitted by necessary changes in
regulations and licenses. It is necessary to include the
Federal Power Commission in future planning.
17. Trapping of downstream migrants from the Snake River
and transport to the estuary are a necessary partial solution
to the problem. Additional trapping at McNary Dam would
salvage fish from the upper Columbia and residual from the
Snake River.
-------
18. Delay in passage of adult upstream migrants below
the dams and in fishways is a significant factor in increasing
mortalities from direct or secondary effects of gas bubble
disease. Changes in the water management system within
allowable flexibility could relieve the adult fish delay
problem.
19. Nitrogen conditions are more severe at some dams
than at others because some necessitate more spilling than others
Also, spillway and stilling basin designs have an effect.
Dams newly completed have required almost complete spilling
of flood flows until the power-generating turbines are in-
stalled—for example, at John Day Dam, completed in 1968.
RECOMMENDATIONS
1. The states of Oregon, Washington, and Idaho, together
with the Environmental Protection Agency, should initiate
immediate action to establish a water quality standards criter-
ion for dissolved nitrogen in the Columbia and Snake Rivers.
a. The standards criterion should establish a maxi-
mum allowable concentration of dissolved nitrogen
in the Columbia and Snake Rivers at 110 percent
of saturation based on analytical procedures
presently being followed by the National Marine
Fisheries Service.
b. Immediate research and development efforts should
be initiated to develop a method to measure
total dissolved gas partial pressures to be
related to dissolved gas concentration data.
c. Expanded research and development efforts are
also needed to relate dissolved gas partial
pressure data to effects on fish. The results
of these studies may be used for future reviews
of the standards criterion for nitrogen.
d. The Environmental Protection Agency should work
with the National Marine Fisheries Service to
develop an effective regional program for dis-
solved nitrogen monitoring in the Columbia and
Snake Rivers. A formalized monitoring program
will be needed to measure progress towards
compliance with the standards criteria.
-------
10
2. A regional council should be established immediately
to make specific recommendations on administrative, legisla-
tive, and policy actions required to implement an effective
nitrogen control program. Membership on the council should
include high-level representatives of regional environmental,
fisheries, water resource development, and water management
agencies. The regional council for nitrogen control should
periodically report on its review of the nitrogen supersatura-
tion problem, assessment of progress towards control of nitrogen
supersaturation effects on fish, and recommendations for further
action to the council's constituent agencies, the Council on
Environmental Quality, and other affected agencies. Elements
of a regional nitrogen control program should include the
following:
a. Nitrogen supersaturation reduction program.
(1) Maximum utilization of the flexibility of
the regional system of reservoirs and power-
generating facilities to reduce spills
during periods of fish migration.
(2) Provisions at each dam of a method to pass
any flows which are in excess of those
required for power generation through or
over facilities which will not entrain
excessive atmospheric gasses in the water.
The Corps of Engineers should be fully
supported in their suggested program of
structural modifications for this purpose
at their projects and others.
(3) Revisions of generator installation schedules
and construction methods to maintain maxi-
mum bypass capability at dams during the
period of construction. Congressional
authorization and appropriation schedules
may need to be adjusted appropriately.
b. Expanded program to physically limit exposure
of migrating juvenile salmonids to nitrogen super-
saturated waters.
(1) Expand the installation and operation of
fish collecting screens in all turbine units
at Little Goose Dam by April 1972. The
collection screens could later be
-------
11
transferred to Lower Granite Dam at the
time of its completion. Lower Granite should
be designed to allow the installation and
operation of both skeleton bays and fish
collection screens.
(2) Install and operate fish collection screens
in all turbine units at McNary Dam by
April 1973.
(3) Continue to expand the program initiated in
1971 to make mass fish hatchery releases
during periods of reduced nitrogen. Trans-
port fish, where feasible, for release in
the Lower Columbia River below Bonneville
Dam.
3. Any new dam in the Columbia Basin must include facili-
ties for adequate control of nitrogen supersaturation. Specifi-
cally, adequate controls at Lower Granite Dam (now under
construction) must be provided before completion of this project.
The design goal for these control facilities should be to main-
tain dissolved nitrogen concentrations downstream from the
project below 110 percent of saturation in a ten-year flood.
Additionally, existing plans for further alteration
or expansion of the Columbia River water resource development
system, including hydroelectric power-peaking aspects of the
hydro-thermal program, should be reviewed by the river manage-
ment agencies to insure that the plans are consistent with the
above strategy for abatement of nitrogen supersaturation problems.
In order that control strategies can be developed and implemented
on a timely basis, this review should be completed and provided
to the regional nitrogen control council by January 1972.
------- |