Anyone running clamp-on sand/solids meters on 3" 1502 treating iron to keep slurry velocity above deposition and cut elbow erosion? We trialed a clamp-on ultrasonic plus a Coriolis downstream last week at 80–90 bpm and about 2.3 ppg; holding 28–32 ft/s and tuning FR kept Reynolds high and pressure drops predictable. Curious which sensors and placements you’ve found reliable during zipper ops without adding too much backpressure.
On 3" 1502 we’ve had best luck clamping the ultrasonic 8–10D upstream of the first elbow (straight spool between the missile and the goat head) and wrapping it so the couplant temp stays steady — giving the meter a quiet room — then set a simple alarm when measured solids deviates >0.2 ppg from blender setpoint to nudge FR before deposition… Small caveat: don’t span a hammer-union land or a chewed spool or it’ll drift with vibration; have you tried a short mag spool in the zipper to sanity-check the Coriolis when you’re at “80–90 bpm”?
Quick data point: our clamp-on ultrasonics on 3" 1502 were most reliable 12D upstream of the first elbow, transducers clocked about 30° with gel pad couplant + heat tape; otherwise the couplant creeps and you get ‘lean’ spikes whenever you jump 80→90 bpm at about 2.3 ppg, which drives me nuts… We pair that with the Coriolis as you did, but alarm on deviation between the two (>0.2 ppg) rather than absolute sand so FR tweaks don’t mask deposition. @marvin_v1970 have you tried a sacrificial wrap on the spool to keep the clamp pressure consistent once the iron hits 150–160°F?
We’ve had better stability using a dual-path clamp-on on the straight pup after the missile and a second head about 5D downstream, then alarm when the downstream solids signal lags >8% — it’s a cheap speed trap for sand. Biggest gotcha is thermal drift as the iron heats; wrap the heads and keep the couplant warm or you’ll chase ghosts. @k.wright, a quick ‘sand step’ at pad start ties the clamp-ons to the Coriolis and we set a low‑velocity alarm near 26 ft/s.
One tweak that’s saved our elbows: pair the clamp-on’s backscatter trend with your downstream Coriolis density and alarm on mismatch, not just absolute counts. During your 28–32 ft/s runs, do a 2–3 minute clean-fluid baseline and set a +3–4 dB threshold plus a 1–2 s lag window; if backscatter spikes without a density bump, we’ve seen sand start to settle ahead of the elbow even though velocity looks fine. , what drives me nuts is weld heat-affected zones — move the head onto a forged spool or a pup without HAZ and the noise floor drops; @marvin_v1970 have you tried that?