Use these ranges and decisions with the approved drilling program, local well-control procedure, mud engineer input and operator policy. Lost-circulation diagnosis is a ranked field hypothesis until pressure, volume and geology evidence converge.
The first depth written beside a mud loss is often the bit depth. That is a timestamp, not proof of location. The thief zone can be at the bit, behind the BHA, at the casing shoe or anywhere in exposed open hole. A defensible diagnosis reconstructs the event from verified volume, operating state, pressure and depth—then states the location as a confidence range, not a guess.
The key distinction: detection depth is not loss-zone depth
Surface instruments tell you when the circulating system changed. The driller’s depth at that moment marks the first confirmed detection under those conditions. Because the annulus connects the whole exposed interval, it does not automatically identify where fluid crossed into the formation. A pre-existing fracture above the bit can begin taking fluid when ECD rises, when cuttings loading changes, or when the pipe position changes.
Report two depths: (1) bit depth / hole depth at first confirmed loss and (2) the interpreted loss interval with confidence. Keeping them separate prevents a timestamp from becoming a false geological fact.
Step 1: prove the event with a clean timeline
Start before interpreting geology. Synchronize pit-volume totalizer, trip tank, pump strokes or flow-in, flow-out, standpipe pressure, hookload/pipe movement, ROP, depth and any downhole pressure-while-drilling channel. Correct for clock offsets between systems. Mark every transfer, dilution, slug, drain, sample, dump and surface spill.
- Use a stable interval immediately before the event as the baseline.
- Verify active-pit selection and sensor calibration.
- Inspect the surface circuit for leaks or an unintended valve line-up.
- Reconcile expected displacement during connections and trips.
- Record both the first deviation and the point at which the event is confirmed.
A sharp active-volume fall with a matching flow-out deficit during steady pumping is strong evidence. A pit change without supporting operating-state data is not enough.
Step 2: bracket the exposed interval
The broadest possible location is bounded by the last hydraulic barrier and current hole depth. Review the casing shoe, liner top, cement integrity information, leak-off or formation-integrity test context, known weak formations and every open-hole lithology. Then ask what changed since the last loss-free circulation.
| Observation | How it changes the location hypothesis | Limitation |
|---|---|---|
| Loss begins immediately on drilling a new interval | Raises confidence near the newly exposed formation. | An older zone may open because circulating pressure changed. |
| Loss begins after pump-rate increase | Raises suspicion of a pressure-dependent weak zone anywhere in open hole. | Does not locate that zone by itself. |
| Loss begins after a connection or trip | Review surge/swab, fill-up, ballooning and static/dynamic behavior. | Trip displacement errors can mimic losses. |
| Loss repeats at the same depth in offsets | Raises confidence in a formation-controlled interval. | Offset trajectory and pressure history may differ. |
| Downhole pressure anomaly near a marker | Can tighten the depth bracket when clocks and depth are aligned. | Tool position is not the same as the flow-entry point. |
Step 3: separate pressure-dependent from pressure-independent behavior
Plot loss rate against operating state. If the loss starts or sharply increases as pump rate—and therefore annular friction/ECD—increases, an induced or reopened pressure-sensitive path becomes more plausible. If substantial loss continues with pumps off, static hydrostatic pressure may still drive flow into the interval or a naturally conductive fracture/vug system may be present.
This is not a licence to run an improvised pressure test. Use only changes authorized by the drilling and well-control program. The value comes from analysing operating changes that already occurred or approved diagnostic steps.
Step 4: use loss-rate shape, not just the peak
The time response carries information about the receiving path. An abrupt step may coincide with intersecting a highly conductive fracture or crossing a pressure threshold. A loss that declines at constant conditions can reflect near-wellbore restriction, filter-cake growth, finite storage or changing fracture flow. An accelerating loss can indicate an enlarging conductive path or increasing pressure differential. None of these shapes is unique, so use them to rank hypotheses rather than declare a mechanism.
Semi-analytical fracture-flow research demonstrates why this matters: fluid rheology, fracture aperture/conductivity, pressure and time interact to control the observed leakage curve. A single peak rate throws away much of that diagnostic signal.
Step 5: align geological and drilling evidence
- Lithology and cuttings: fractured carbonates, vuggy intervals, depleted sands and faulted rock support different mechanisms, but surface cuttings arrive after lag.
- ROP and drilling response: a break, drilling break or torque change can support a newly entered zone, not prove it.
- Logs and images: caliper, image logs and fracture indicators can refine the interval when available.
- Offset wells: compare stratigraphic tops and pressure history, not measured depth alone.
- Pressure window: compare static mud weight, ECD/PWD and available integrity information with uncertainty.
- Treatment response: where an engineered treatment is authorized, the volume/pressure response can update the hypothesis, but successful sealing still may not reveal a unique depth.
Lag time: useful, but often misapplied
Cuttings and gas observed at surface represent an earlier downhole event after annular lag. Pit loss, however, is a whole-system volume imbalance observed as the system responds; do not “lag-correct” every pit movement as though it were a cutting sample. Use lag to align geological markers carried to surface, while using synchronized hydraulics to time the loss itself.
Likewise, distinguish bit depth from sensor depth. A PWD tool sits above the bit; its measurement depth and time must be shifted correctly before comparing it with formation tops or a surface event.
Step 6: assign a confidence level
A useful conclusion states what the data can actually support:
- Confirmed event: verified net loss with transfers and surface leaks excluded.
- High-confidence interval: loss begins directly after exposing a new interval and is supported by repeatable pressure/geological evidence.
- Probable interval: several lines of evidence agree, but an older open-hole zone remains possible.
- Unresolved: the event is real, but data cannot distinguish two or more exposed zones.
“First confirmed at 3,842 m MD while drilling at 3,846 m MD. Net loss increased from 0 to 28 bbl/h after the pump-rate step; no surface transfer or leak was found. The newly exposed fractured carbonate from 3,832–3,846 m is the probable thief interval, but a pressure-dependent zone below the 3,710 m shoe remains possible. Confidence: medium.”
A practical evidence matrix
| Hypothesis | Evidence that supports it | Evidence that weakens it |
|---|---|---|
| Newly drilled natural fracture/vug | Immediate loss after new footage; geological marker; offset history; loss persists at lower circulating pressure. | Loss began before the marker or only after a later pressure increase. |
| Induced/reopened fracture above bit | Strong dependence on pump rate/ECD; stops or reduces when friction falls; weak interval already exposed. | Large static loss independent of circulation. |
| Wellbore breathing/ballooning | Repeatable take while pumping and partial return after pumps stop; low net loss over full cycle. | Continuing net deficit with no return signature. |
| Surface-volume error | Transfer, valve line-up, sensor drift or unmeasured addition explains the balance. | Independent flow-out deficit and repeatable calibrated pit decline. |
What surface data cannot do alone
Standard rig sensors rarely provide a unique downhole coordinate for fluid exit. They can narrow timing, pressure state and plausible intervals, but exact location may require additional logs, downhole measurements or a planned diagnostic operation. The technically honest answer is sometimes a ranked interval, not a single depth.
Technical basis
The workflow combines mass-balance practice with research showing that fracture leakage depends on hydraulic aperture, conductivity, pressure, time and non-Newtonian fluid rheology. Large field datasets also show that reliable event classification draws on multiple drilling parameters and depth context—not one sensor threshold.
Common questions
Is the loss zone always at the bit when losses start?
No. The first-loss bit depth is the detection depth. Fluid can leave through any hydraulically connected interval in the exposed open hole, including an older weak zone or the casing-shoe region.
Which surface signals are most useful?
A synchronized combination of calibrated active volume, flow-in, flow-out, pump state, standpipe pressure, depth and pipe movement. ECD/PWD and geological markers strengthen the interpretation when correctly time- and depth-aligned.
Can pit volume identify the exact loss-zone depth?
Usually not by itself. Pit volume confirms and quantifies a net loss after transfers and displacement are reconciled; locating the zone requires pressure, depth, geology and operating-state evidence, and may remain an interval rather than an exact depth.