Anyone using clamp-on ultrasonic meters to ground-truth night flows in DMAs? I’ve been running a portable transit-time unit on a 300 mm ductile iron main at 02:00 to validate EPANET 2.2 calibration and tune PRV setpoints, but at sub-0.1 m/s the signal gets jumpy and Doppler is a non-starter; what models and mounting tricks are giving you stable, repeatable readings so we can target leaks and cut NRW without overspending?