Using CFD,Virtual Fish calculate the impact of the complicated virtual flow environment on passing fish. It models fish as ellipsoid-type objects and assumes that fish have no free will – that is, they cannot react to the water forces that carry them through the plant. "That's probably a good assumption for the most part because the fluid forces inside the plant are so strong," Sotiropoulos says. "The fish may have little or no time to react."
Sotiropoulos formulated a set of equations that describe how the ellipsoid object is transported and rotated by the flow from the upstream reservoir, through the turbines and in the downstream tailrace river reach. Virtual Fish allows the user to calculate the various water-induced forces that tend to shear, squeeze, stretch, bend, rotate or twist the fish at every point along its path through the power plant. With biological input, Virtual Fish users can then determine whether these forces are harmful to fish. For example, the user could find out how many times the fish was spun by the flow, revealing whether it is likely that the fish became disoriented, and thus more vulnerable to predators downstream.
"The Virtual Fish model is a significant advancement, but it is still a very approximate thing," Sotiropoulos says. "It represents the turbulent and rapidly changing flow environment in the power plant with its statistical time average. Yet passing fish get injured by instantaneous water forces, whose magnitude could often be much higher than their statistical mean value. Also, the model doesn't account for the effects fish have on the flow. And the fish (in the model) is not flexible. It is a stiff body right now....
"But the model for Voith Siemens Hydro is a good one for the industry," he adds. "You have to strike a balance between model sophistication and getting fast answers. But if you want to further enhance the understanding of specific flow mechanisms responsible for injury and mortality, you have to understand the instantaneous flow structures at the fish's scale. To do this, we need to develop unsteady CFD models of the turbulent flow environment and account for the effect of the fish on the local flow field."
Thus, Sotiropoulos began work in the summer of 2001 on a four-year project with Oak Ridge National Laboratory (ORNL) researchers to create the next generation of Virtual Fish. The enhanced model will feature a flexible fish, which can interact with and be distorted by instantaneous flow. It will allow engineers and biologists to study and understand the interaction of a much more realistic fish-like object with a flow environment much closer to that encountered by live fish passing through a power plant.
"To do this, we have to modify the unsteady flow equations with terms that account for the effect of the fish on the flow and solve them at the same time with the equations describing the fish motion," Sotiropoulos explains. "So this model will be much more sophisticated. We will be using the massively parallel supercomputing facilities at Oak Ridge to do this work."
This research is funded through ORNL by the U.S. Department of Energy's Hydropower Program, which is focusing on the issue of what happens to fish as they pass through hydropower turbines. "The hydraulic stresses inside turbines are difficult to study because of their high velocities and chaotic structures," says Michael Sale, head of ORNL's Environmental Sciences Division. "Computer simulation is an important tool for understanding phenomena that we cannot measure directly.... (The enhanced version of) Virtual Fish will be an important method for predicting how real fish might respond to simulated velocity fields."
Meanwhile, Sotiropoulos is also collaborating with ORNL and TVA researchers on an experimental project. Researchers are equipping several TVA dams to measure flow forces. They will compare data from this experiment with results from CFD modeling done by Sotiropoulos.
The other CFD model important to hydropower industry officials is Virtual Bubbles, which assesses auto-venting turbine (AVT) technology. With it, air is aspirated into the water as it flows through the turbine whenever the water's dissolved oxygen level is below the minimum of 5 milligrams per liter.
photo by Gary Meek
|
|
Fotis Sotiropoulos, a Georgia Tech associate professor of civil engineering, models the flow of water through hydropower plants by numerically solving a set of non-linear mathematical equations – the Navier-Stokes equations – using state-of-the-art computational fluid dynamics methods.
(300-dpi JPEG version - 514k)
|
|
This level is recommended by the U.S. Environmental Protection Agency for supporting early-life stages of warm-water fisheries. AVTs rely on turbulent mixing and mass transfer to release as much dissolved oxygen as possible from the air bubbles into the water.
Hydropower operations reduce dissolved oxygen downstream, particularly in the summer. Daytime heating creates a warm layer of water at the top of a reservoir, while water at the bottom is colder. Minimal mixing of oxygen occurs in this situation, so the cold water below becomes depleted of dissolved oxygen. Power plants draw from the lower layers of the reservoir, so water that is released downstream is also depleted in dissolved oxygen. Such conditions create an unhealthy habitat for fish and other organisms and can create a smelly river not suitable for recreation.
TVA has installed several Voith Siemens Hydro AVTs to increase dissolved oxygen in its operations in the Southeast. The company is using Virtual Bubbles to evaluate and optimize the performance of this new turbine technology. The model uses CFD to calculate the flow environment downstream of the turbine and tracks the paths of individual air bubbles as they are carried by this simulated flow. The bubbles are allowed to exchange oxygen with the flow, collide and form bigger bubbles, and then break up into smaller bubbles.
Virtual Bubbles can help engineers determine: (1) whether AVTs are achieving maximum transfer of oxygen from the bubbles to the water; and (2) whether the new turbines are negatively affecting the energy efficiency of the power plant.
"Industry needed a tool to analyze possible scenarios and designs and to quantify the impact of air bubbles on efficiency and oxygen transfer," Sotiropoulos explains. "Again, this model is based on a balance between having enough sophistication to be useful, but not being so complicated that it requires so many resources that it is beyond industry's time scales."
For more information, contact Fotis Sotiropoulos, School of Civil and Environmental Engineering, Georgia Tech, Atlanta, GA 30332-0355. (Telephone: 404-894-4432)
(E-mail: fotis.sotiropoulos@ce.gatech.edu)
