⚡ electrical engineering

Pressure transient analysis for detecting leakages along active and abandoned wells

This study demonstrates the feasibility of using Pressure Transient Analysis (PTA) on real-time downhole pressure data to detect and estimate leakage rates from induced fractures and casing failures in both active and abandoned wells within a representative offshore Norwegian sandstone reservoir.

Original authors: Joshua Mugisha, Anton Shchipanov, Hans Joakim Skadsem, Ingebret Fjelde

Published 2026-07-08

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Joshua Mugisha, Anton Shchipanov, Hans Joakim Skadsem, Ingebret Fjelde

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you are managing a giant, underground water tank (a reservoir) deep beneath the ocean floor. You are pumping water into it to store energy or capture carbon. The big worry? What if that water finds a way to escape through a crack in the tank's lid or leaks out the side of the pipe you used to fill it? If it escapes, it could contaminate fresh water sources or leak into the atmosphere.

This paper is like a detective story about how to catch these leaks early using a tool that is already installed in many of these wells: a permanent pressure gauge sitting deep underground.

Here is the breakdown of the study in simple terms:

The Detective's Toolkit: Pressure Transient Analysis (PTA)

Think of the pressure gauge as a very sensitive microphone listening to the "heartbeat" of the well. The researchers used a method called Pressure Transient Analysis (PTA).

The Analogy: Imagine you are blowing air into a balloon. If the balloon is perfect, the pressure rises smoothly. But if there is a tiny hole, the pressure behaves differently—it might drop suddenly or rise more slowly because air is escaping.
The Method: The researchers looked at how the pressure changed over time, specifically using a mathematical trick called a "derivative" (which highlights changes in the pressure curve). They also created a "safety zone" (called an Uncertainty Envelope) around the normal pressure curve. If the pressure curve steps outside this safety zone, it's a red flag that something is wrong.

The Two Main Scenarios They Investigated

The study looked at two specific ways leaks can happen, using computer simulations (since they couldn't go dig up real wells to test this).

1. The "Crack in the Lid" (Induced Fracturing)

The Problem: When you pump water in too fast or at too high a pressure, you might accidentally crack the rock layer (the "caprock") that sits on top of your reservoir. This creates a new path for the water to shoot up into shallow, fresh water zones.
The Test: They simulated a "Step-Rate Test," which is like turning the faucet up a little bit, then a little more, then a lot more, while watching the pressure gauge.
The Discovery: When the rock cracked and the water started leaking upward, the pressure gauge showed a distinct "dip" or drop in its reading. The computer model showed that if about 2% of the water started leaking out, the gauge could spot it. It's like hearing a specific "hiss" sound that tells you the balloon has a hole.

2. The "Leaky Pipe" (Behind-Casing Leakage)

The Problem: Wells are lined with steel pipes (casing) and sealed with cement. Over time, the cement can crack or pull away from the pipe or the rock. This creates a hidden tunnel behind the pipe where water can sneak up from the deep reservoir to the shallow layers.
The Scenario A (Active Well): The well is currently pumping. The researchers simulated water leaking through the bad cement. They found that if the cement gets "leaky" enough (about 10% as permeable as the rock itself), the pressure curve changes shape, showing a downward slope that signals a leak.
The Scenario B (Abandoned Well): What if the well is plugged and abandoned, but a neighbor well is pumping nearby? The pressure from the neighbor's pump can push water into the abandoned well through the bad cement. The researchers showed that even though the abandoned well is "shut off," its pressure gauge can still "feel" the leak and show a signature that proves water is moving where it shouldn't.

The Results: What Can We Actually Detect?

The study is a "feasibility check," meaning they asked: "Is this theoretically possible?"

Yes, it works: The computer simulations proved that these pressure gauges can detect leaks in both active and abandoned wells.
The Threshold: The method is good at catching leaks that are roughly 2% or more of the total fluid being pumped. If the leak is microscopic (less than 2%), the "noise" in the data might hide it, making it hard to see.
The Advantage: Unlike other methods that require expensive fiber-optic cables or seismic trucks (which are like sending a team to the surface to listen for vibrations), this method uses the gauges that are already there. It's a "free" upgrade to the software that reads the data.

The Catch (Limitations)

The authors are very honest about what they didn't do:

It's a Simulation: They didn't test this on a real, leaking well in the ocean yet. They only used computer models.
Simple Conditions: They assumed the fluid was just water and the temperature was constant. Real-world scenarios involving hot geothermal fluids or carbon dioxide (which changes state) are more complex and weren't tested here.
The "Noise" Problem: Real pressure gauges aren't perfect; they have static and jitter. The researchers used a "safety envelope" to try to ignore the jitter, but in the real world, distinguishing a tiny leak from a noisy gauge is harder than in a computer.

The Bottom Line

This paper proposes a new way to listen to our underground wells. By using the pressure gauges we already have and looking for specific "dips" or "slopes" in the pressure data, we might be able to catch leaks in active wells and even in old, abandoned wells before they become environmental disasters. It's like giving the well a stethoscope to listen for the sound of a leak, rather than waiting for the water to spill out.

Technical Summary: Pressure Transient Analysis for Detecting Leakages Along Active and Abandoned Wells

Problem Statement
Subsurface well integrity is critical for environmental safety and operational efficiency, particularly in large-scale geological injection projects such as water injection, geothermal energy, and Carbon Capture and Storage (CCS). A primary risk involves leakage from the target formation into shallow zones or the atmosphere through compromised wellbore integrity. Potential leakage pathways include induced fractures propagating through caprocks, failures at the casing-cement or cement-formation interfaces, and crossflow in legacy (Plug-and-Abandonment, P&A) wells. While monitoring technologies like fiber-optics and micro-seismic surveys exist, they often face limitations regarding cost, deployment complexity, or sensitivity to micro-leaks. Furthermore, existing Pressure Transient Analysis (PTA) methods have primarily focused on individual leakage scenarios rather than a comprehensive approach capable of diagnosing multiple pathways concurrently. This study addresses the need for a feasible, cost-effective monitoring solution using widely available Permanent Downhole Gauges (PDGs) to detect and characterize leakage in both active and abandoned wells.

Methodology
The study employs a mechanistic well-reservoir modeling approach based on a representative offshore Norwegian sandstone reservoir. The fluid flow simulations assume single-phase, isothermal, low-compressible fluid (water) flow. The research is divided into two primary investigative parts:

Induced Fracturing and Caprock Leakage:
- Modeling: A pseudo-coupling workflow using the reservoir simulator Rubis was utilized to simulate pressure-induced fracture opening and propagation. A pressure-dependent fracture permeability function, $k_f(p)$ , was applied to fracture grid blocks to model the transition from matrix flow to fracture flow once a specific fracturing pressure ( $p_f$ ) is reached.
- Scenario: A vertical injector was simulated in a three-layered system (target formation, caprock, shallow formation). The model tracked fracture growth from the target formation, through the caprock, and into the shallow layer.
- Analysis: Synthetic Step-Rate Tests (SRTs) were generated. The data was interpreted using a PTA-SRT approach enhanced with an Uncertainty Envelope (UE) concept. The Bourdet derivative of normalized pressure transients was calculated for each injection step. The UE was established based on a reference derivative (pre-fracture) to distinguish genuine leakage signatures from measurement noise.
Behind-Casing Leakage (Casing-Cement-Rock Integrity Failure):
- Modeling: A composite reservoir model was developed where the cement sheath behind the casing was assigned a distinct effective permeability ( $k_c$ ). This allowed for the simulation of fluid migration from the target formation to the shallow formation through a compromised cement zone without induced fracturing.
- Scenarios: Two distinct cases were simulated:
  - Active Wells: A constant-rate injection test followed by a shut-in period to analyze flowing and build-up pressure responses.
  - Abandoned (P&A) Wells: A nearby active injector was simulated to induce pressure interference in a plugged well. The abandoned well was monitored via a simulated wireless downhole gauge during a long shut-in period.
- Analysis: PTA was applied to the shut-in and flowing data. Diagnostic log-log plots of pressure and Bourdet derivatives were compared against reference limits representing "no-leakage" (single formation flow) and "full penetration" (two-formation flow).

Key Contributions and Results
The study demonstrates the feasibility of using PDG data combined with PTA methods to detect leakage in specific scenarios:

Induced Fracture Detection: The PTA-SRT method with an Uncertainty Envelope successfully detected the onset of induced fracturing. The signature of fracture opening was identified as a downward shift in the Bourdet derivative. The method detected fracture propagation into the shallow formation (out-of-zone leakage) when the leakage rate reached approximately 2% of the total injected volume. Below this threshold, leakage signatures were difficult to distinguish from noise in the simulated cases.
Behind-Casing Leakage in Active Wells: The study showed that increasing effective cement permeability ( $k_c$ ) resulted in a declining trend in the pressure derivative during the flowing period. Leakage was detectable when the relative cement permeability ( $k_c/k_m$ ) approached 0.1, corresponding to a leakage volume of approximately 2%.
Behind-Casing Leakage in Abandoned Wells: The PTA framework adapted for abandoned wells successfully identified leakage signatures during shut-in periods induced by a nearby injector. The method detected leakage when the cement permeability was equal to or greater than the target formation permeability, with leakage volumes exceeding 1% of the injected volume.
Noise Handling: The application of the Uncertainty Envelope concept proved effective in differentiating true leakage signals from the inherent noise of PDG measurements, provided the leakage signature exceeded the noise amplitude (e.g., >2 bar deviation in derivative vs. <1 bar noise).

Significance and Claims
The authors position this work as a feasibility study establishing a conceptual basis for applying PTA/PDG methods to leakage detection. The significance of the study lies in:

Operational Efficiency: The proposed methods utilize flowing transient interpretation (during injection) and standard shut-in data, potentially avoiding the costs and production losses associated with dedicated well tests.
Utilization of Existing Infrastructure: The approach leverages PDGs, which are already widely deployed in active wells and increasingly available in abandoned wells via wireless technology, offering a scalable and potentially lower-cost alternative to fiber-optics or micro-seismic surveys.
Complementary Monitoring: The study suggests that PTA-based leakage detection can serve as a pragmatic adjunct to Surface Casing Pressure (SCP) monitoring, offering depth-specific detection capabilities that surface measurements lack.

Limitations and Future Work
The paper explicitly states that the results are valid only for the specific assumptions made: vertical wells, homogeneous infinite-acting reservoirs, and single-phase isothermal flow. The authors acknowledge that the proxy modeling of fractures (without fully coupled geomechanics) and the assumption of static cement permeability are simplifications. They emphasize that the non-uniqueness of PTA interpretations requires integration with geological data and operational history. Crucially, the authors conclude that further testing and validation using field data, as well as cross-validation with other monitoring technologies, are necessary next steps to mature the methodology for practical industrial application.