Most solids-control losses are invisible because the system never stops running. Nothing trips, nothing leaks onto the deck — the dilution rate just sits quietly higher than it should, and the money leaves the budget one barrel at a time. This is the anatomy of one such case (illustrative of the method), and what it took to catch and reverse it.
The situation
On an offshore jack-up, dilution on a 12¼″ section had crept to 165 bbl/day. The mud properties were in spec, the shakers were running, the cones were humming — by every casual measure the system was “fine.” No one had flagged a problem because there was no event to flag. The only thing wrong was a number nobody was calculating.
Two quiet failures, stacked
When removal efficiency was finally run from the daily reports, it came back at 58% — firmly in the “money on the table” band. Two independent failures had compounded:
- A worn desilter feed-pump impeller had dropped cone feed head below the ~75 ft target (the manifold had drifted to 41 psi, about 66 ft at the section mud weight), coarsening the cut and returning fines to the active system.
- An under-set shaker — a screen one step too coarse for the formation — was passing solids it should have captured at the first stage.
Each failure alone looked minor; either would have shaved a few points off η. Together they forced extra dilution to hold the low-gravity-solids concentration in spec — and that dilution is what showed up as cost.
The recovery
The interventions were ordinary maintenance, not capital projects: re-impeller the feed pump to restore head (back to ~49 psi / ~78 ft), and re-screen the shaker to the correct cut for the section. Within a day, dilution fell from 165 to 110 bbl/day and η climbed from 58% to 72%.
Over the remainder of the section, the reduced dilution recovered roughly $48,000 of fluid that would otherwise have been built, circulated once and discarded — before counting the disposal cost of the waste that fluid would have become. The payback on a few hours of maintenance was a matter of days.
Why the leak stayed open so long
Both faults were the slow kind. An impeller does not fail on a Tuesday; it erodes a little each day until, weeks later, the head has quietly fallen below the line. A screen that is one cut too coarse never announces itself — it simply lets a fraction of the solids through, every circulation, forever. Neither shows up in the parameters a driller watches. Only η, tracked against footage, makes them visible.
The real lesson
The dollar figure is not the lesson — it changes with rig, fluid and section. The lesson is that the leak was completely invisible until someone measured it. There was no alarm to wait for, because solids-control losses do not announce themselves. The only defence is a number, tracked daily, that turns “the system looks fine” into “η dropped four points yesterday — go look at the cones.”
Key takeaways
- Measure η every tour and plot it against footage.
- Walk the cones and shakers by eye — spray pattern, beach, screen condition.
- Treat a slow rise in dilution as a fault to diagnose, not a cost to absorb.
The rigs that do this routinely recover sums like this one without ever calling it a “save” — because they never let the leak open in the first place.