Ever see a working scale model of a 600-foot-long dam and more than a half-mile of river through which water flows just as it would in real life as the dams infrastructure is altered?


Avista Corp. has had an engineering company build such a model of its Cabinet Gorge Dam in North Idaho, the dams two diversion tunnels, and the bed of the Clark Fork River to test how the rivers flows would be altered by a proposed $40 million project.


It is a 1/50th scale model of something on the order of 4,000 feet of river, with the projects facilities in the middle of that, says Chick Sweeney, a senior consultant in the Redmond, Wash., office of ENSR Corp., a Westford, Mass., environmental consulting and engineering firm that built the model.


The model is a behemoth20 feet wide, 90 feet long, and 6 feet highand Sweeney says, We have about 20,000 gallons of water in the thing and in the recirculating system that feeds water to the model.


The $300,000 model is located at ENSRs Redmond office, and Avista and ENSR are using it to test the feasibility, and to tweak the design, of Avistas proposed project, which the Spokane company hopes would reduce dissolved gases in the Clark Fork downstream from the dam.


During spring runoff, huge volumes of water sometimes plunge over the 208-foot-high dams spillway, driving atmospheric gases deep into the spillway basin, where high concentrations of the gases dissolve into the water, Avista spokesman Hugh Imhof says. Nitrogen in such supersaturated water can cause bubbles to form in the bloodstreams of fish, sickening and killing them, much as the affliction known as the bends sickens and kills divers who resurface too quickly from great depths.


As part of its relicensing agreement for its two Clark Fork dams, Cabinet Gorge and Noxon Rapids farther upstream, Avista has agreed to mitigate the dissolved gases, Imhof says.


We have to come up with something to mitigate that entrained gas, he says. Theyve noticed that its even in Lake Pend Oreille, and thats a ways downstream. We havent noticed any problems with the fish, but its possible there are problems, and if there are, we would mitigate it. One of the species of fish in the Clark Fork is the bull trout, which is protected under the Endangered Species Act.


While the dam was being constructed in the winter of 1951-52, the Clark Fork was diverted into a pair of 1,000-foot-long tunnels dug through a mountain on one side of the dam site. Those tunnels were plugged when the dam was completed, but Avista hopes that by reopening one or both of the tunnels it can divert part of the Clark Forks flows during spring runoff and reduce the amount of water that plunges over the dams spillway, easing and perhaps solving the dissolved-gas problem.


Its estimated that it would cost $40 million to open one of the tunnels or $60 million if Avista must open both tunnels to reduce gas concentrations sufficiently, Imhof says.


The plan now is to build just one tunnel and try it for a few years and see what the biological impacts are, and then build the second one if necessary, Imhof says. Construction would start in 2007 and be completed in 2010.


Avista would ask regulators for permission to recoup the cost of the project from ratepayers, but that wouldnt increase utility rates much because it would amortize its costs over many years, Imhof says. He says the Federal Energy Regulatory Commission, the Idaho Department of Environmental Quality, and the U.S. Fish and Wildlife Service all have signed off on the basic plan for the project, but the project still would have to go through a permitting process.


Avista built the model because we dont want to go out and build a $40 million remedy and find out later that it doesnt work, Imhof says. He adds, Basically, were going to do this unless the model says we shouldnt.


Sweeney says that in initial tests with the model, the projects design looks promising, although testing will continue through the end of the year.


The model is built so precisely that a one-inch-high wave in the model triggered by a design change would equate to a 50-inch wave in real life, Sweeney says.


To mirror the exact contours of the river bed in the model, ENSR used data gathered by Avista with an acoustic depth finder with many, many beams to map the Clark Forks bed, Sweeney says. The device was on a moving boat, with the transmitter immersed, he says. Basically, it maps everything thats under the water.


Also, ENSRs team has glued miniature boulders onto the models river bed to emulate the structure in the bottom and sides of the river. Says Sweeney, We had to add roughness to the modelwe call the boulders roughnessto get the same slope as in the field for the model flow rate.


The river bed is a concrete shell applied to a metal lath suspended from cardboard templates, he says. When it came to putting the concrete surface on the models river bed, there was interpretation and art involved, Sweeney says. We have a cadre of technicians who make their living doing that by hand troweling the concrete to form the shell.


The model is far from the first one ENSR has built. Its currently doing modeling tests on a removable spillway weir thats under study to improve fish passage at Lower Monumental Dam, on the Snake River, and its also modeling the spillway and part of the powerhouse for similar improvements at McNary Dam, on the Columbia River. It also has done models of Lower Granite and Ice Harbor dams, on the Snake; Bonneville and Rocky Reach dams on the Columbia; and other impoundments.


For Avistas project, the testing will attempt to simulate how the design would work, how it can be tweaked, and what would happen when one or both of the diversion tunnels were put into use, Sweeney says.


The model allows them to see that in 3-D, with the water flowing, he says.


Flows of water through the tunnels can be altered by doing such things as changing the location of control gates and rearranging outlet structures, Sweeney says.


Imhof says the outlets from the tunnels might have to be modified and widened. Theyre making them rectangular rather than round, he says. The engineers think thats the way to reduce the turbulence. They may end up doing other things like putting an angled end to it.


In one test using the model, the engineers put dye into the flowing water and record where the dye goes in the current, Imhof says.


The big question is whether using the tunnel creates more problems than it resolves, he says.


Turbulence in the tailrace from the tunnel could end up creating erosion in the river, Imhof says. Of course, it alters the river environment, and the project will send material downstream, he says. If Avistas road next to the river were washed out due to the flow changes, the companys personnel would lose ready access to the dam, he says.


Then theres the question of whats inside the plugged tunnels now, he says. Its probably coffer-dam material, but its been there 50 years, so its established an ecosystem.


A management committee of representatives of the groups and agencies that signed off on the relicensing of the Clark Fork dams, along with a subgroup that includes state and federal agencies, approved of studying the tunnel option. It had a good probability of meeting both the projects needs and state standards for water quality, says Nate Hall, an Avista natural resource manager on the Clark Fork project. Hall says that riffles can reduce gas-saturation levels in rivers, but there are few riffles in the seven-mile stretch of the Clark Fork between Cabinet Gorge and Lake Pend Oreille, and the dissolved gas from the dams spillway doesnt dissipate before the river empties into the lake.