Integrity Log Review

Document-review skill for the evidence layer beneath pnge:well-integrity-barriers. It is designed for reports and test sheets that an engineer already has locally, not for public data sources.

Important: Do not collapse multiple report types into one conclusion too early. A failed pressure test, weak CBL response, and temperature anomaly each support different hypotheses and have different uncertainty.

---

Preferred Inputs

• MIT pressure test sheet with acceptance criterion
• Annulus pressure history or bleedoff / rebuild plots
• Cement evaluation logs: CBL/VDL, ultrasonic, bond summary
• Noise or temperature log report around suspected leak interval
• Corrosion / caliper inspection notes and workover history
• Current well schematic with depth references

---

Workflow

1. Normalize all depths and pressures to a common reference.
2. Separate evidence by report type.
3. Ask what each report can rule out, not just what it suggests.
4. Build a leak-path evidence matrix.
5. Recommend the next diagnostic only where the evidence is still ambiguous.

---

Module 1 - Pressure Test Review

def pressure_change_rate(p_start, p_end, hours):
    if hours <= 0:
        return None
    return (p_end - p_start) / hours

def percent_pressure_loss(p_start, p_end):
    if p_start == 0:
        return None
    return 100.0 * (p_start - p_end) / p_start

For each pressure test, document:

• test volume and fluid type
• hold period
• acceptance threshold
• temperature correction applied or not
• whether the reported pass/fail criterion was objective or narrative only

---

Module 2 - Evidence Matrix

def evidence_vote(*flags):
    """
    Count positive indicators across report types.
    """
    return sum(1 for f in flags if bool(f))

Use a matrix like this:

Evidence source	Supports tubing leak	Supports packer leak	Supports behind-pipe leak
MIT result	yes/no	yes/no	weak
Noise log	yes/no	yes/no	yes/no
Temperature anomaly	yes/no	yes/no	yes/no
Annulus rebuild	yes/no	yes/no	yes/no
CBL / ultrasonic	weak	weak	yes/no

One positive indicator is usually not enough for a confident call.

---

Module 3 - Report-Type Guidance

Report type	Best use
MIT sheet	Confirm containment over the tested volume and duration
CBL/VDL	Screen for cement quality and zonal isolation concerns
Ultrasonic imaging	Better pipe and cement interface detail
Noise log	Locate active fluid movement
Temperature log	Identify active flow or injection anomalies
Corrosion inspection	Support failure-mechanism diagnosis

---

Module 4 - Common Review Findings

Finding	Typical interpretation
Pressure test failed but no depth-localizing evidence	Containment issue exists, source still unresolved
Annulus rebuilds quickly after bleedoff	Active communication likely
CBL suggests poor bond but no active pressure behavior	Potential barrier weakness, not necessarily active leak
Noise and temperature coincide at same depth	Strong leak-path candidate
Corrosion or pitting aligns with pressure behavior	Failure mechanism more credible

---

Output Format

When using this skill, structure the answer as:

1. Documents reviewed and reference depth basis
2. What each document actually proves
3. Leak-path evidence matrix
4. Most likely interpretation with uncertainty
5. Best next log, test, or workover discriminator

---

Integration Points

• Use pnge:well-integrity-barriers for the barrier and pressure framework.
• Use api-well-standards for casing and cement design context.
• Use pnge:materials-fracture-mechanics when crack growth or fatigue is implicated.