At 40 bpm slickwater through 4.5-in casing, is your observed Reynolds number enough to keep 100‑mesh off the low side, or do you still see a bed forming at toe turns from slip velocity? I’m comparing approaches for protecting elbows — any luck using dP variance or acoustic signatures as an early sand-bed indicator?
At 40 bpm in 4.5-in we kept 100‑mesh off the low side by forcing micro‑pulses: drop 2–3 bpm for about 10 s every 5 min, then ramp back. On the diagnostic side, “dP variance or acoustic signatures” — a bed shows up as a sluggish dP fall on the dip and a sharp elbow ping on ramp‑up; a cheap clamp‑on accelerometer at the first turn read cleaner than line mics. @OP have you tried a quick pulse test like that before your elbow?
Quick example: we’ll drop in a 25–30 bbl viscous sweep (20–25 lb/Mgal HPG) right before long toe turns; it cuts slip enough that 100‑mesh stops streaking low and the elbow doesn’t build a bed. Do you have fiber on this well, @OP? Our DAS cue is a brief “gravelly buzz” near about 300 Hz about 10–15 s before the dP wiggle, which lets us time the sweep.
I’ve had better luck catching a bed early by strapping a clamp‑on accelerometer to the toe‑side elbow and watching HF RMS/kurtosis; the “kurtosis kicks first” 10–30 s before any dP wobble, and a small 0.05–0.1 gpt FR bump right before the bend bumps wall shear without touching rate. If you’ve got DAS, you’ll see a 1–3 kHz energy rise too (example hits in OnePetro: https://onepetro.org/search?q=acoustic+sand), and it’s like hearing the popcorn before it burns — do you have a spot to mount a sensor on that elbow?
Even with high Re at 40 bpm in 4.5‑in, geometry at ‘toe turns’ still seeds a bed. I’ve had best results putting two 200–500 Hz quartz pressure taps across the elbow and watching cross‑spectral phase; a steady 20–40° lag in the 8–15 Hz band shows up well ahead of any mean dP shift and gives time to correct. Building on @mario_ross94’s acoustic angle, use a pump tach to strip harmonics first or you’ll chase false positives.