Hydrocyclones — desanders and desilters — are the only major piece of solids-control equipment with no moving parts. They run entirely on geometry and pressure. That makes them cheap and reliable, but it also means they fail silently: when they stop cutting, nothing breaks, nothing alarms, and the only symptom is a slow, expensive rise in dilution.
The variable that decides whether a cone is working is not flow rate — it is head: the pressure energy at the cone inlet that spins the slurry hard enough to throw solids to the wall and down to the apex while clean fluid spirals up the vortex finder and out the top.
How a cone separates
Fluid enters the cone tangentially and is forced into a tight spiral. The resulting centrifugal field — many times gravity near the wall — drives the denser solids outward and down to the narrow apex, where they leave as underflow. Lighter fluid reverses into an inner spiral and exits the overflow at the top. The strength of that spin is set almost entirely by the inlet head; lose the head and you lose the separation, even though fluid still flows.
The thirty-second check every mud engineer can run
Feed head converts directly from the manifold pressure your gauge already shows, so it costs nothing to check each tour:
where h is feed head (ft), p is feed pressure (psi) read at the manifold, and ρ is mud weight (ppg). The 19.2 is simply 1 ÷ 0.052, the pressure-gradient constant — so the formula is just the standard hydrostatic relation rearranged for head.
| Cone type | Typical size | Design feed head | Cut range |
|---|---|---|---|
| Desander | 6–12″ cones | ~75 ft (runs a touch lower) | Sand: down to ~40–74 µm |
| Desilter | 4″ (and smaller) cones | ~75 ft | Silt: down to ~15–40 µm |
Why head, not flow, is the master variable
It is tempting to judge a cone bank by how much fluid is moving through it. But a bank running high volume at low head is just passing dirty fluid through a funnel — it looks busy and removes almost nothing. The spin does the separating; the spin comes from head. Two banks moving identical volumes can have completely different cut points if their feed heads differ.
What steals your head
- A worn feed-pump impeller — by far the most common culprit. Impellers erode gradually, so the head loss creeps in over days and no one notices the shift the day it crosses the line.
- Too many cones open for the pump’s capacity — spreading the available head too thin across the manifold.
- Blocked or eroded apexes — changing the underflow geometry the cone depends on.
- Air or gas in the feed — cutting effective head and destabilising the air core (one reason the degasser belongs upstream of the cones).
Reading the cones by eye
Head tells you the cause; the discharge tells you the symptom. Walk the bank and look at every apex:
- A fine, umbrella-shaped spray — healthy. The cone is developing full spin and rejecting solids with minimal fluid.
- A rope-like, wet stream (“roping”) — the apex is overloaded or the head is too low. You are losing fluid, not solids.
- A bone-dry apex that occasionally plugs — the cut is too aggressive for the feed; the apex needs opening or the feed adjusting.
You can diagnose half the problems on a cone bank before you ever touch a gauge.
Key takeaways
The fix for low head is usually cheap — a fresh impeller, the right number of cones open, a cleared apex. The cost of ignoring it is not: every circulation at low head returns fines to the active system, and those fines drive dilution, wear and disposal volume for the rest of the section. One gauge reading and one division is the highest-return thirty seconds in solids control.