You’ve got sustained casing pressure (SCP) on an annulus. You’ve bled it off multiple times, maybe even run a negative test, but the pressure keeps returning. The first thing you check is the cement bond log, and it looks clean, showing good isolation. So, what’s actually happening?
The common assumption is that internal pressure causes radial cracking in the cement sheath. While that mechanism can occur, it’s rarely the primary culprit. The dominant path to lost isolation under cycling is debonding: the cement separating from the casing at the interface. This creates a microannulus, often just a fraction of a millimeter across, but entirely sufficient to conduct gas.
The Engineering Reality: Why Debonding Happens
The core reason for debonding is a fundamental material mismatch between steel and set cement. They have different thermal expansion coefficients and vastly different stiffnesses. When the casing heats up, it expands more than the surrounding cement, pushing outwards and compressing the sheath. When it cools, the casing shrinks away.
The cement, being far less elastic and having accumulated some permanent deformation during the compressive phase, does not fully follow the casing back. This leaves a small, permanent gap at the interface. Each subsequent cycle adds a little more permanent strain, and the microannulus appears gradually after a number of cycles, not necessarily after a single severe event. This is why failure often correlates with a well’s operating history rather than any specific incident.
What Drives the Interface Open?
Three primary loads interact to drive this interface separation:
- Temperature: This is the largest driver on most wells. Production heats the casing, shut-in cools it, and cold fluid injection cools it hard and fast. Steam injection and cryogenic injection are extreme examples, both notorious for sheath failure due to severe thermal cycling.
- Internal Pressure: Pressuring the casing expands it, compressing the sheath. Bleeding off pressure allows the casing to contract. A stimulation treatment is a single large pressure cycle; a well repeatedly fractured through the same casing experiences many.
- Pore Pressure Change in the Cement: Set cement is porous. Changes in the fluid pressure within its pore structure alter the effective stress and can cause shrinkage independent of mechanical loads.
Worked Example: Radial Movement at the Interface
Consider a 7-inch casing (OD 7.000 in) where the internal temperature falls 60 °C during a cold injection cycle. Using a steel thermal expansion coefficient (α) of 1.2 × 10−5 /°C:
- Radial contraction = (7.000 / 2) × 1.2 × 10−5 × 60 ≈ 0.00252 inches ≈ 64 micrometers (µm).
If the cement only partially recovers, say 60% of that movement, the residual gap is roughly 26 µm. For context, the diameter of a human hair is around 50-100 µm.
Gas flow through an annular slot scales with the cube of the aperture. Comparing a 26 µm microannulus against a well-bonded interface with a theoretical 5 µm gap:
- (26 / 5)³ ≈ 140
This means a gap smaller than a human hair, produced by a routine temperature change, can increase conductivity by more than two orders of magnitude. This explains why a bond log showing acceptable cement can coexist with a well that flows gas to the annulus. “The cement is good” and “the annulus is tight” are distinct statements.
Operational Approaches to Mitigate Debonding
While a complete design-away from this mismatch isn’t possible, several measures aim to improve sheath durability. However, each has limitations:
- Low-Modulus (Flexible) Cement: Designed to deform more elastically with the casing, reducing plastic strain accumulation. The trade-off is often lower compressive strength, which may not suit all load cases.
- Expanding Additives: These aim to place the interface in residual compression, so a small contraction doesn’t immediately open a gap. The expansion is finite and occurs once; large or repeated cycles can still exceed its capacity.
- Fiber Reinforcement: Primarily resists radial crack propagation. This addresses cracking, not the more common debonding mechanism for cycling failures.
- Latex and Elastomeric Additives: Improve the toughness and reduce the permeability of the cement matrix. However, matrix permeability is rarely the leak path; the interface is the critical point.
The most crucial, yet often overlooked, measure is good mud removal and centralization during the primary cement job. Debonding initiates at pre-existing defects: a mud film left on the casing, a channel, or a poorly displaced pocket. A sheath that started with a clean, fully bonded interface tolerates far more cycling because there’s nowhere for separation to begin.
Most of a cement sheath’s long-term durability is determined during the hours of the primary job. Decisions about mud conditioning, centralization, and displacement rate are made under schedule pressure and rarely revisited. By the time anyone cares about long-term annulus integrity, that primary job is often years in the past, performed by a different crew.
Decision Checklist for Managing Annulus Integrity
When dealing with wells prone to or experiencing annulus pressure, consider these operational factors:
- Limit Rate of Temperature Change: The magnitude of a thermal cycle drives strain, but the rate of change drives the thermal gradient through the sheath and the peak stress at the interface. Where practical, manage injection or production rates to avoid rapid temperature swings.
- Avoid Unnecessary Full Pressure Cycles: Repeatedly bleeding an annulus fully to zero and then re-pressuring it is a full cycle. Doing this monthly as a test regime means 120 cycles over a decade, accelerating degradation.
- Treat Cold Injection as a New Load Case: If cold injection is planned for a well not originally designed for such conditions, expect the cement sheath to behave differently. Re-evaluate potential annulus issues.
- Record Cycle Count: Sheath degradation is cumulative. A well’s thermal and pressure cycling history is a better predictor of future annulus problems than its age alone.
- Interpret Bond Logs Carefully: Always consider the pressure and temperature conditions under which a bond log was run. A microannulus can close under internal pressure and reopen when bled off, making a log run under pressure appear better than one run at ambient conditions.
Failure Modes and Lessons Learned
There is no remedial treatment that fully restores a debonded interface to its original condition. Squeeze cementing can seal paths it can reach, but it cannot re-bond the entire casing-cement interface. This is a critical point when a primary cement job is being value-engineered: the immediate cost saving is real, but the long-term cost appears a decade later as an annulus that cannot be made tight at any price.
A crucial diagnostic test for differentiating a microannulus from genuinely absent or channeled cement is to run a bond log at two internal pressures. If the bond response improves markedly when the casing is pressured (indicating the gap has closed), the cement is present but debonded. If the response doesn’t change, the cement is absent or severely channeled. These two conditions require entirely different remediation strategies, and a single-pressure log cannot tell them apart.
The industry often measures cement quality at the moment of setting and then treats that result as a permanent property. It is not. A sheath logged as excellent on day one but subjected to two hundred thermal and pressure cycles is a different object. Where a well has a long history of cycling and an annulus has started to build pressure, the reasonable prior is a debonded interface rather than a cement job that was defective from the start. This distinction fundamentally changes both the diagnosis and what remediation can realistically achieve.
Have a question about your well? Reach out via the contact page.