Executive SummaryThis report summarizes the results of studies sponsored by the U.S. Department of Energy and conducted by Pacific Northwest National Laboratory (PNNL) to evaluate the biological performance of an advanced design turbine installed at Unit 8 of Wanapum Dam on the Columbia River in 2005. In all studies, paired comparisons were made between Unit 8 and a conventional Kaplan turbine, Unit 9. PNNL studies included an evaluation of blade-strike using deterministic and probabilistic models, integrated analysis of the response of the Sensor Fish to severe hydraulic events within the turbine system, and a novel dye technique to measure injury to juvenile salmonids in the field.The first study involved applying deterministic and stochastic blade-strike models to the advanced and existing turbine designs. Modeled probabilities were compared to results of Sensor Fish releases and injury/mortality of balloon-tagged fish under the same operational parameters. The new advanced design turbine had slightly higher modeled injury rates than the existing turbine design; however, there was no statistical evidence that suggested significant differences in blade-strike probabilities. Overall, injury rates predicted by the deterministic model were higher than experimental rates of injury while those predicted by the stochastic model were in close agreement. Fish orientation at the time of entry into the plane of the leading edges of the turbine runner blades is an important factor contributing to uncertainty in modeled results.A second study focused on expanded analysis of Sensor Fish data through the use of computational fluid dynamics (CFD) simulations and other analytical techniques. The Sensor Fish data was collected in conjunction with balloon tag tests involving juvenile Chinook salmon. Sensor Fish pressure and acceleration measurements were analyzed to identify characteristic signatures in each passage time history. These signatures were used, together with CFD and streamline plots, to identify the frequency and location of significant acceleration events such as collisions and shear. The Sensor Fish data were divided into four regions for each release condition (i.e., turbine, discharge, release pipe depth, and intake bay). Events were classified as either collision or shear and by severity level. The majority (80%) of severe acceleration events were due to collision. When all four regions were pooled, the two turbine units had similar probability of severe collisions (i.e., 20%). Unit 8 (advanced turbine) had fewer severe shear events (1.1%) than unit 9 (3.4%), but Unit 8 had more slight shear events (29.3%) than Unit 9 (19.7%). Although mean pressure nadirs recorded by Sensor Fish were consistently higher for Unit 8, values were highly variable. Thus, there was no overall evidence of correlation with Sensor Fish collision events and measured pressure nadir. The overall predicted shear injury rates were 3.1% for Unit 8 and 4.4% for Unit 9. This compared to observed injury rates of 1.1% and 0.9% for Units 8 and 9...