The Marsh funnel is the most-used and most-misunderstood instrument on the rig. It is quick, rugged and needs no power, so it gets read every tour and quoted in every morning meeting. The trouble is what people do with the number. A funnel reading is a useful consistency check and a terrible rheology measurement — and confusing the two leads to real engineering mistakes.
What the funnel measures
The test is simple: fill the funnel through its screen to the bottom of the screen, then time how many seconds it takes for exactly one quart (946 mL) to drain out into the cup. That time, in seconds, is the funnel viscosity. The calibration reference is water: clean water at 70°F (21°C) should run out in 26 ± 0.5 seconds.
A reading of “46 seconds” means this mud takes 46 seconds to drain the same volume water drains in 26 — it is thicker, full stop.
Why it is not real viscosity
Here is the part that trips people up. The funnel gives a single number that blends together everything affecting flow — plastic viscosity, yield point, gel strength, even a little density — into one figure measured at one undefined, ever-changing shear rate (the flow slows as the funnel empties). True rheology needs at least two numbers, and they come from a different instrument.
| What you want to know | Marsh funnel | Rotational viscometer (Fann VG) |
|---|---|---|
| Quick “is it thicker than last tour?” | Yes — ideal | Overkill |
| Plastic viscosity (PV) | No | Yes (600–300 rpm) |
| Yield point (YP) | No | Yes |
| Gel strengths | No | Yes (10-sec / 10-min) |
| Inputs for hydraulics design | No | Yes |
Two muds can give the identical funnel reading and behave completely differently downhole: one high in PV (solids-loaded, hard to pump), the other high in YP (good carrying capacity, very different hydraulics). The funnel cannot tell them apart. Designing hole-cleaning or ECD off a funnel number is guessing dressed up as a measurement.
The solids-control connection
For a solids engineer, the funnel earns its keep as an early-warning gauge. When funnel viscosity drifts up tour over tour with no deliberate treatment, the usual cause is rising low-gravity solids — drilled fines the equipment is failing to remove, thickening the fluid. A creeping funnel reading is often the first hint, before the daily LGS or MBT confirms it, that the separation train has slipped. Treat the funnel as the smoke detector and the retort/MBT as the investigation.
Key takeaways
The Marsh funnel is a brilliant consistency check and a useless rheometer — respect both halves of that sentence. Read it every tour to catch change, but never quote it as viscosity, never design hydraulics from it, and never let a steady funnel reading convince you the rheology is fine. When the number moves, go to the VG meter and the retort to find out why — and on a solids-control job, a quietly rising funnel reading usually means fines are winning.