SC DrillTechDesander vs desilter: cone size, feed head, and the spray that tells the truth
Hydrocyclones are the quiet workhorses of the train — no moving parts, no screens, just pressure and geometry doing the second and third stages of removal. Yet more cyclone capacity is wasted through two mistakes than any other equipment: wrong feed head, and running them on the wrong mud. This week: desander versus desilter, the 75-foot rule, and the one glance that tells you a cone has failed.
Same physics, different cut
A hydrocyclone is an inverted cone. Slurry is injected tangentially at the wide top, spins down the wall at high centrifugal force, throws the heavy solids outward and down to the apex, and returns clean fluid up the centre through the vortex finder. Desanders and desilters are the same device — the only real difference is cone diameter, and cone diameter sets the cut:
- Desander — larger cones (typically 6–12 in, commonly 10 in). Cut point ≈ 40–75 µm (sand-size). Runs first, high flow — a 10-in cone handles ~500 gpm at 75 ft of head.
- Desilter — smaller cones (typically 4 in, sometimes 5 in). Cut point ≈ 15–44 µm (silt-size). Runs after the desander — a 4-in cone handles ~60–80 gpm at 75 ft of head, so you need many in a manifold.
Smaller cone, higher centrifugal force per unit volume, finer cut. That is the whole relationship. The correct order never changes: shale shaker → desander → desilter → centrifuge, with each stage's overflow feeding the next stage's suction tank.
The 75-foot rule (feed head is everything)
A hydrocyclone does not run on flow — it runs on head. The design target for most units is about 75 feet of head, roughly 32 psi, measured at the manifold, with desilters generally run a little higher than desanders (practical range ~30–50 psi). This single number makes or breaks the cut:
- Too little head → weak centrifugal force, the cut coarsens, and the underflow ropes. Removal collapses.
- Too much head → marginal separation gain at the cost of rapid apex erosion and wasted pump energy.
Feed head is set by the dedicated centrifugal pump — each cyclone bank needs its own. If you take one lesson from this issue: check the manifold pressure before you blame the cones.
Spray vs rope — the one-second diagnostic
The apex discharge tells you instantly whether a cone is working. A healthy cone throws a fine, umbrella-shaped spray — solids leaving with just enough liquid, cone at maximum efficiency. A rope — a thick, wet cord of underflow — means the cone is overloaded: too many solids for the apex, so solids are bypassed straight to the overflow and removal effectively stops. The fixes, in order: open or replace the apex, reduce the solids load, or add cones. Aim to keep underflow around 3% of throughput — enough to avoid apex plugging, not so much that you dump whole mud.
The caveat that saves your barite
Here is the mistake that quietly costs the most money: do not run desanders or desilters on weighted mud. Their cut point (15–75 µm) sits right on top of the barite size range (median ≈ 10–75 µm), so on a weighted system the cones throw your expensive weight material out the apex. On weighted mud the correct tool is the mud cleaner — cyclones mounted over a fine screen, so the barite passes back through the screen while the coarse drilled solids are discarded. Cyclones full-stream belong on unweighted water-based mud, run continuously to hold a low mud gradient.
A field example
An unweighted WBM desilter bank is discarding almost no solids and the mud gradient is climbing. The cones are roping. The instinct is "worn cones" — but the manifold gauge reads 20 psi, well under the 75-ft/32-psi target. The dedicated pump is undersized for the added cones fitted last hitch. Restoring feed head to ~32 psi turns the ropes back into sprays, the cut sharpens to its 15–44 µm design range, and the gradient falls — no new cones required. The problem was never the cones; it was the head feeding them.
The takeaway
Cyclones reward respect for two numbers and one picture: cone size sets the cut (desander coarse, desilter fine), feed head at ~75 ft/32 psi sets whether that cut actually happens, and the apex spray tells you at a glance if it is working. And never point them at weighted mud — that is what the mud cleaner is for. Get those right and the cyclones do the medium-to-fine removal cheaply, so the centrifuge only has to chase the true fines.
Educational field guidance based on standard hydrocyclone practice. Cut points and heads are typical design ranges — verify against your specific cone design, pump and fluid before acting on a well. For a fuller Q&A see the SC DrillTech solids-control questions.
This brief is the field summary. For the full reference, see:
Every SC DrillTech tool, in one bundle
- Field Secrets — the manual
- SC DrillTech Doctor — diagnostics
- FluidPro Suite — mud & fluids
- DrillTech ROC — cost & cuttings
- Training Series — decks & exams
- Every calculator in this brief
Prefer the live platform? Meet Rig IQ
Run every calculation in this series on your own rig data — dashboards, PSD, audits and reports, all in one place.
Explore Rig IQ →Subscribe to SC DrillTech Weekly
One field-tested technical brief in your inbox every week — no spam, one-click unsubscribe.
Subscribe free →Know an engineer who’d want this? Forward this to a colleague →