A major water treatment upgrade is underway in Nashville, Tennessee. Metro Water Services (MWS) owns and operates two 90-mgd conventional plants: KR Harrington and Omohundro, first constructed in the 1970s and 1920s, respectively. As part of the far-reaching improvements at both plants, MWS will expand Omohundro’s capacity to 150 mgd and implement several new core facilities: raw water pumping, pretreatment, and filtration. These core processes will be kept largely independent of the existing facilities, allowing the upgrades to take place with minimal impact to operations and providing a blank slate for establishing a highly robust process train fit to meet the City’s needs now and into the future. This presentation will focus on how extensive CFD modeling informed design decisions and optimized the flash mix, flocculation, and sedimentation designs. Initial vetting of pretreatment alternatives with MWS set design guardrails. MWS selected conventional flocculation and sedimentation with plate settlers for the reliability, familiarity, and simplicity of operation. Pumped diffusion flash mix was selected because of its effectiveness in quickly dispersing coagulant, flexibility across a wide range of plant flows, and its use of simple, off-the-shelf components. Vertical shaft mixers were selected for maintenance considerations. Having established a framework with basic decisions, more nuanced and interrelated questions remained: 1. A parallel-flow flocculator geometry with flow-splitting inlet weirs is well-proven and largely eliminates concerns over short-circuiting that could lead directly to higher settled water turbidity, but a serpentine flow path allows a smaller footprint, is hydraulically efficient, and is simpler for chemical application. What is the real difference in particle residence time distribution and how can stator baffles, baffle wall dimensions, and the layout of the openings in the ported walls be adjusted to mitigate the difference? 2. As flow transitions from the flocculation to the sedimentation basins, decreasing the total port area increases the port velocity, thus improving flow distribution along the width of the basin. However, this increased port velocity is in direct competition with protecting the flocs from zones of high shear. What is the sweet spot for a practical ported wall design? 3. Hydrofoil impellers and paddlewheel mixers are both capable of mixing the flocculator volume, but there are differences in mixing energy uniformity, hydraulics, and localized shear. How do these different styles of mixers compare in terms of total particle collision potential through the flocculation basin? The project team used computational fluid dynamics modeling to answer these questions, further industry knowledge of pretreatment basin performance, and design a system that will result in the highest settled water quality at the Omohundro Water Treatment Plant.