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.
A useful lost-circulation decision tree does not begin with a product name. It begins by confirming that volume is actually leaving the circulating system, protecting well control, locating the most probable loss interval and separating permeability losses from fractures, vugs, induced losses and ballooning. Only then can the team choose a proportionate response and verify that it worked.
The decision path at a glance
- Stabilize and account: confirm pit, trip-tank, transfer and flow measurements; check for surface leaks and unrecorded movements.
- Verify the event: compare expected displacement with measured returns and establish the rate, duration and cumulative deficit.
- Classify the loss: seepage, partial, severe or total—using the project definition and the trend, not one isolated reading.
- Rank mechanisms: permeable matrix, natural fracture/vug, induced or reopened fracture, ballooning, or a non-formation explanation.
- Locate the interval: combine timing, depth, lithology, connection/trip history, ECD/PWD and offset evidence.
- Select a response: remove controllable pressure first where appropriate, then choose a compatible and passable treatment for the ranked mechanism.
- Verify and escalate: test against pre-agreed acceptance criteria; stop repeating an ineffective response.
Gate 1 — Is it a real formation loss?
Reconcile active volume, reserve transfers, slug or pill movements, surface line-up, tank calibration and sensor quality. During a connection, distinguish the expected flow-back signature from a continuing deficit. A measured pit fall with no equivalent surface explanation is stronger evidence than a single flow-out fluctuation.
Record the baseline: timestamp, depth, hole section, activity, pump rate, standpipe pressure, surface density/rheology, ECD/PWD where available, flow-in/flow-out, active volume and cumulative loss.
Gate 2 — What changed immediately before the loss?
| Observation | Raises the rank of | Still check |
|---|---|---|
| Loss starts after a pump-rate or rheology increase | Induced/reopened fracture from higher ECD | Cuttings bed, restriction, sensor lag and weak-zone exposure |
| Loss begins at a new fractured or vuggy interval | Natural fracture/vug | Pressure history and whether the fracture was hydraulically opened |
| Volume returns after pumps-off | Ballooning / fracture breathing | Net cumulative balance; do not use flow-back alone to dismiss influx |
| Gradual seepage in a permeable formation | Matrix invasion or small openings | Filter-cake quality and differential pressure |
| Abrupt total returns loss | Large conductive feature or major fracture | Surface line-up, well-control status and exact onset depth |
Gate 3 — Can the pressure driver be reduced?
Compare static mud weight with ECD, annular friction, surge, pump rate, ROP, cuttings loading, gel-breaking pressure and any restriction signature. If the event is pressure-induced, repeatedly placing material without correcting the driver can reopen the path. Any change must remain inside the approved pressure and hole-cleaning envelope.
Gate 4 — Choose by mechanism, not habit
| Ranked mechanism | Design question | Evidence before escalation |
|---|---|---|
| Permeability / small openings | Can filtration and invasion be reduced with a compatible fine-to-medium bridge? | Loss trend, filter-cake response and stable returns |
| Natural fracture or vug | Does the blend contain a passable coarse bridging tail and enough supporting grades? | Placement confidence, reduced leak rate and sustained response |
| Induced/reopened fracture | Has ECD/surge been controlled before sealing or strengthening? | Stable pressure margin at the required operating condition |
| Uncertain mechanism | What low-risk diagnostic action will reduce uncertainty fastest? | New data that changes the ranked hypotheses |
| Persistent severe/total loss | Has the pre-approved threshold for specialized treatment or abandonment of the current tactic been reached? | Decision-authority review and documented contingency |
Compatibility and passability gate
Before any particulate treatment, check the smallest practical restriction through the complete path: mixing and transfer equipment, bit nozzles, motor, MWD/LWD/RSS, screens and completion hardware. Confirm carrier-fluid stability, temperature, contamination tolerance, reservoir-damage constraints and the planned solids-control handling. Sealing capability and tool passability are separate qualifications.
Define success before pumping
Success may mean restored full returns, a loss rate below an agreed threshold, a stable static level, the ability to circulate at the required rate, or a demonstrated operating-pressure margin. State the observation period and the test condition. “Losses improved” is not an acceptance criterion.
Escalation triggers
- No material reduction after the defined observation window.
- Repeated recurrence at the same pressure or activity.
- Uncertain well-control status or evidence of simultaneous influx.
- Tool restriction, inability to place the treatment, or unexpected pressure.
- Cumulative loss reaches the approved environmental, logistics or cost threshold.
- The working diagnosis no longer explains the data.
Common decision failures
- Calling every pit loss a formation loss before reconciling transfers.
- Selecting LCM only from loss rate, without an opening/mechanism hypothesis.
- Repeating the same pill while ECD remains above the weak-zone limit.
- Using pumps-off flow-back as proof that the well is safe.
- Reporting a peak seal pressure without duration or leak rate.
- Continuing treatment without a predefined stop or escalation rule.
Technical basis
Loss-rate transients and fracture-leakage models can help constrain fracture behavior, but interpretation remains non-unique and depends on geometry, rheology and boundary conditions. The decision tree therefore uses ranked hypotheses and verification rather than claiming certainty from one surface signal.
Common questions
Should treatment selection start from the loss rate?
No. Rate is important, but mechanism, opening uncertainty, pressure history, placement and passability determine whether a treatment is credible.
When should a team stop repeating an LCM pill?
When the predefined verification window shows no meaningful response, the loss recurs under the required condition, or new evidence invalidates the diagnosis.
Does restored circulation prove the zone is permanently sealed?
No. It proves the system met the stated test at that time. Durability requires monitoring under the expected operating condition.
