A desilter feed pump supplies smaller cones that are more sensitive to feed instability, plugging, viscosity and manifold imbalance. The pump must hold usable feed energy without forcing barite loss or flooding the mud-cleaner screen in weighted service.
Pressure-to-head conversion
Desilter pressure must be interpreted with mud density. The same psi is not the same pump head when mud weight changes.
Head (ft) = Pressure (psi) × 2.31 / SGExample: 32 psi at SG 1.30 gives 32 × 2.31 / 1.30 = 56.9 ft of head at the manifold.Worked example: many small cones
If twelve 4 in cones are online and each is expected around 50 gpm, the pump must deliver roughly 600 gpm at the required manifold pressure plus system losses. Isolating cones changes both total flow and manifold behavior, so cone count must be part of the pump check.
| Variable | Example | Engineering meaning |
|---|---|---|
| Active cones | 12 | Defines total process flow |
| Per-cone flow | About 50 gpm | Use actual OEM/process basis where available |
| Total flow | About 600 gpm | Pump must meet this at manifold head |
| Manifold behavior | Balanced discharge | Uneven cones indicate distribution or plugging issues |
Weighted mud caution
On weighted mud, desilter operation may conflict with barite conservation unless a mud cleaner screen is part of the strategy. The pump may be hydraulically correct while the process decision is economically wrong. That is why desilter feed duty must be tied to the mud program, not only cone pressure.
Field acceptance checks
| Check | Acceptable behavior | Failure signal |
|---|---|---|
| Manifold pressure | Stable for the online cone count | Pressure changes when cones plug or isolate |
| Cone underflow | Consistent discharge across cones | Some cones dry, some roping |
| Mud condition | Viscosity and solids loading within operating expectation | High PV makes cone performance look like pump failure |
| Screen return | If mud cleaner is used, screen can handle underflow | Underflow floods screen and returns solids |
Troubleshooting logic
Extra calculation: active cone capacity
Desilter pump duty changes every time cones are isolated or returned to service. The pump should be checked against the active cone count, not the original manifold nameplate.
Total process flow = Active cones × Expected flow per coneExample: 16 active cones × 50 gpm/cone = 800 gpm. If only 10 cones are open, the same pump and valve position may force a different manifold behavior.Fine-cone sensitivity
Small cones react strongly to viscosity, feed solids and plugging. A desilter problem is often blamed on the pump because pressure is visible, but the actual loss may be inside the cone bank: partially blocked inlets, worn apexes, mismatched cones, high PV, or weighted mud being treated without a barite strategy.
| Observed field sign | Likely meaning | Action |
|---|---|---|
| Several cones with no underflow | Plugging or poor distribution | Inspect feed entries and apex condition |
| All cones weak | Feed energy or suction issue | Check pump speed, suction and impeller wear |
| Good underflow but poor mud result | Wrong process strategy | Review weighted mud, screen recovery and solids trend |
Engineering depth: why desilter feed is less forgiving
Desilter cones are smaller and more sensitive to small changes in pressure, solids loading and manifold balance. A desilter bank can look online while part of the bank is starved, overfed or bypassed through uneven distribution.
| Failure mode | Field evidence | What to verify |
|---|---|---|
| Uneven cone loading | Some cones spray, others weak or dry | Header balance, plugged feeds and valve positions |
| Excessive solids loading | Roping and high underflow density | Upstream screens, desander performance and feed pit solids |
| Weighted mud misuse | Useful weighting material loss | Whether mud cleaner mode is required instead of direct discard |
| Low energy | Fine solids remain high | Pump curve, SG correction and active cone count |
Bank demand = active cones × flow per cone at target pressureNever calculate bank demand from installed cone count alone. Isolated, plugged or worn cones change the real duty.For quality control, compare feed, overflow and underflow character together. A clean-looking overflow is not proof if the feed sample was not representative or if part of the flow bypassed the cones.
