Hydro-Quebec studying generator cooling flow
To maximize load on its existing hydro equipment without compromising reliability, Hydro-Quebec has undertaken several research programs at its Institut de Recherche (IREQ). One of these programs involves studying the thermal behavior of air-cooled hydro generators to define the potential for power increase based on the thermal limits of the stator winding, says Sebastien Houde, a hydro-dynamic specialist at IREQ.
Hydro-Quebec's 340 turbine-generator units range in size from 10 MW to 400 MW.
To balance the complexity of the flow associated with large variations in length scales (associated with generator cooling flow) and the requirements of the simulation process, Hydro-Quebec developed a specific methodology. This methodology uses a coupling between two numerical simulations:
— Global simulations of the rotor and stator; and
— Detailed simulations of one slot of the stator.
The global simulations were based on a rotor section that encompasses one pole and a stator section representing a sufficient number of bars to yield circumferentially coherent geometries. The flow was solved with a commercial Navier-Stokes solver and a frozen rotor interface between the rotor and stator. The boundary condition for the simulations came from measured field data and an in-house thermal analysis model. The geometry of the stator cooling ducts was taken into account using a porous losses model. When compared with in-situ measurements, researchers determined that the global simulations showed a slightly offset flow balancing between the end windings and the stator core but a flow repartition between the core cooling ducts that was similar to the measured data, Houde says.
The detailed simulations were based on one stator slot but a complete representation of the cooling ducts. The inflow condition came from the global simulation. The detailed simulations exhibited a flow repartition between the end-windings and core that was close to the experimental values; however, this value was offset when considering the repartition within the duct. Further refinements are under way to minimize the experimental error that is currently deemed too large for a definitive validation of the CFD simulations, Houde says.
Houde says the study demonstrated the usefulness of computational fluid dynamics (CFD) for cooling flow analysis of a large hydro-generator. Hydro-Quebec plans to work to improve the reliability of this technology by conducting further validation. Furthermore, work is under way to add coupling with other physical simulations (thermal, electromagnetic, and structural). The utility plans to use this work to improve its power output by either relaxing some security factors or redesigning some components of the generators.
Study to determine effect of release timing on fish survival
Chelan County Public Utility District (PUD) is performing a study at its 1,287-MW Rocky Reach project to determine if the time of day during which study fish are released upstream affects migration rates, arrival time at the dam, and survival.
Studies have shown that predation may play a significant role in smolt survival at the dam, says Steve Hemstrom, senior fisheries biologist with Chelan County PUD. Before 2009, tagged sockeye smolts were released only during the day, from the upper end of the reservoir. These fish typically migrate downstream through Rocky Reach Reservoir, on the Columbia River in Washington, and arrive at the dam about 24 hours later, during daylight hours when predation rates are highest.
Hemstrom says the goal of the 2009 study was to determine if run-of-river sockeye smolts that migrate downstream past the dams at night, thereby avoiding piscivorous (pikeminnow) and avian predators that roam the reservoir and congregate near the dam, have a natural survival advantage.
For this study, Chelan County PUD personnel released more than 2,000 acoustic tagged sockeye smolts in groups during the day and night. Utility personnel made survival estimate comparisons for groups released during the day, and at night, over a 24-day period.
To determine dam passage timing, Chelan County PUD used split-beam hydroacoustics (echo-sounding equipment) to detect the arrival time of smolts, Hemstrom says. In addition, two crews working at the smolt bypass sampling facility conducted round-the-clock sampling to determine the proportions of smolts that passed during the night and day.
Results from the survival and dam passage studies are expected to be available in November 2009.
Chelan County PUD conducts survival tests as part its commitment to achieve juvenile salmon and steelhead survival standards set forth in its Habitat Conservation Plan for the Rocky Reach project. The plan calls for a three-year 93 percent average survival standard.