Sensor Fish gets redesign
Pacific Northwest National Laboratory scientists are now miniaturizing the Sensor Fish so it can be surgically implanted in the bodies of real fish. "That way there's no question that it's reacting to flows like a real fish because it is inside a real fish," said Tom Carlson, who manages the project. As smaller components become available, researchers will finalize a new design that will be about one-quarter by one-and-a-half inches in size and one-fourth the cost of the tube-shaped Sensor Fish.
PNNL scientists have used the Sensor Fish to study spillway and turbine passage routes for migrating salmon at several Columbia River dams. Scientists are trying to determine what hydraulic conditions contribute to mortality rates at these dams and what alternatives, structural or operational, might reduce injuries to fish.
In addition to the accelerometers and pressure sensors that were part of the original Sensor Fish, the Sensor Fish implant will include devices called rate gyros. The combination of linear and angular accelerometers will mimic the environmental sensing functions of the inner ear of a fish. "We're hypothesizing that the vestibular system of fish may be disrupted during passage," Carlson said. The rate gyros would measure the rate of change in pitch, roll and yaw of the fish as it moves through the water. "A number of salmon smolts we see below stilling basins and turbines are temporarily disoriented. They're dizzy. As they move to the surface, toward the light, to re-orient, they become vulnerable to birds and other predation," Carlson said.
By implanting sensors in live fish, releasing the fish through dam passage routes such as spill or turbines and recovering the fish afterwards, scientists can link fish injuries to the conditions experienced during passage. "This technology will provide a unique capability for assessing these environments and help us to more quickly identify deleterious conditions and find solutions that will provide safer fish passage," Carlson said.
One drawback, the Sensor Fish does not record its location as it passes through the stilling basin, making it difficult to determine the location and type of hydraulic conditions it encounters. To make this connection, scientists use three-dimensional computational fluid dynamics (CFD) modeling to create predictive simulations of the hydraulic environment and link them to information gathered by the Sensor Fish. After dumping data from the Sensor Fish into a computer, researchers set up the same physical parameters and run the same scenarios in CFD models. The models use moving particles to represent the Sensor Fish and produce the same sets of statistics as the Sensor Fish.
"CFD results are realistic, but not exactly what the Sensor Fish experienced. The Sensor Fish provides the linkage between simulations and data on actual flow conditions. It measures what happens to a body that's moving in real fluid under actual operating conditions," said Marshall Richmond, chief engineer, who manages the CFD component of the project.
Comparing the two allows scientists to better understand the hydraulic conditions that may be responsible for fish injury and to make recommendations for improvements. "We are trying to develop this process to predict as accurately as possible how changes to dam operations and design will affect fish survival."