The evolution of a gas plume injected into a curved axisymmetric porous channel

This paper investigates gas injection into curved, water-saturated porous channels modeling underground storage, deriving asymptotic evolution equations that reveal how axisymmetry and channel slope drive distinct buoyancy-influenced dynamics and five temporal spreading regimes, ultimately offering insights for enhancing the safety and efficiency of subsurface hydrogen and CO$_2$ storage.

The Gist

Imagine you are trying to store a giant, invisible balloon of gas (like hydrogen or carbon dioxide) deep underground. You pump it into a sponge-like rock layer that is already soaked with salty water. The goal is to keep that gas trapped safely for years, perhaps to use as fuel later or to keep it from polluting the atmosphere.

But the underground isn't a flat, boring floor. It's often shaped like a giant, upside-down bowl or a dome (geologists call these "anticlines"). The shape of this underground dome is crucial: if the gas gets too high, it might spill over the edge and escape into a neighboring area, never to be recovered.

This paper is a mathematical study of how that gas bubble moves and changes shape as it gets pumped into these curved, water-filled rock tunnels. The researchers used advanced math to predict exactly what happens, breaking it down into simple rules.

Here is the story of the gas bubble, explained through everyday analogies:

1. The Setup: A Curved Sponge

Think of the underground rock as a long, curved garden hose that is completely filled with water.

• The Gas: You are injecting a stream of gas (like blowing air into the hose). Because gas is lighter than water, it wants to float to the top.
• The Shape: The hose isn't straight; it curves upward like a hill.
• The Problem: If you blow too fast, the gas rushes to the top and shoots forward like a bullet. If you blow too slow, the gas just floats up and spreads out slowly. The researchers wanted to know: How does the curve of the hill change the game?

2. The Two Types of Hills

The team tested two specific shapes for their "underground hose":

• The Parabolic Hill: Imagine a hill that gets steeper and steeper the further you go. It never flattens out; it keeps curving up.
• The Gaussian Hill: Imagine a hill that starts steep but then gradually flattens out into a gentle plateau as you go further away.

3. The Five Acts of the Gas Bubble (The Parabolic Hill)

When pumping gas into the steep, ever-curving hill, the gas doesn't just move in a straight line. It goes through five distinct "acts" or stages, like a play:

• Act I: The Thin Film (The Speed Run)
  At first, the gas is pumped in so fast that it doesn't have time to float. It shoots along the very top of the water, forming a thin, fast-moving film of gas, like a skater gliding on the very top layer of a pond. It rushes forward, ignoring gravity for a moment.

• Act II: The Bulge (The Heavy Lifting)
  As the gas film gets longer, the "slope" of the hill starts to matter. Gravity kicks in. The gas realizes, "Hey, I'm light! I want to float!" It starts to thicken up, like a balloon inflating. The front of the gas slows down and starts to bulge. It's no longer a speed skater; it's a heavy swimmer trying to stay at the surface.

• Act III: The Drain (The Water Evacuation)
  This is the most interesting part. As the gas floats and thickens at the top, it pushes the water down. The water underneath the gas starts to drain away toward the bottom of the channel, like water draining out of a bathtub as you fill it with air. The gas layer gets thicker, and the water layer underneath gets thinner.

• Act IV: The Catch-Up
  Eventually, the water underneath is almost gone. The gas film is now very thin again, but this time it's a thin film of water at the bottom, with the gas taking up most of the space. The "front" of the gas (the upper edge) starts moving again, but now it's dragging the last bit of water along with it.

• Act V: The Flat Lake (The Final Calm)
  Finally, the water is completely drained from under the gas. The gas bubble has stretched out so much that it becomes a flat, horizontal sheet floating on top of the remaining water. It stops rushing forward and just sits there, perfectly level, like a calm lake. This is the safest and most efficient state for storage.

4. Why Does the Shape Matter?

The researchers found that the shape of the hill changes the rules:

• On the steep hill (Parabolic): The gas has to fight the slope. The slope eventually forces the gas to stop rushing and start floating, leading to that "Flat Lake" state.
• On the gentle hill (Gaussian): The gas behaves differently. If the hill flattens out, the gas can spread out sideways very quickly, almost like a ripple in a pond, rather than just moving forward.

5. Why Should We Care? (The Real-World Impact)

This isn't just about math; it's about energy safety and efficiency.

• Hydrogen Storage: We want to store hydrogen (a clean fuel) underground. If the gas rushes too fast and spills over the edge of the dome, we lose our fuel.
• Carbon Capture: We want to bury CO2 to stop climate change. If it leaks, it's useless.

The paper gives us a "rulebook" for engineers. It tells them:

• How fast to pump: If you pump too fast, the gas shoots ahead. If you pump at the right speed, the gas will naturally spread out and flatten itself against the roof of the cave, maximizing the space used.
• Where to stop: The math predicts exactly where the gas will stop moving forward (the "spill point"). If you know this, you can ensure your storage site is big enough so the gas never escapes.

The Big Takeaway

Think of the gas bubble as a lazy, floating balloon in a curved tunnel.

• If you push it hard, it zooms ahead.
• But if you let gravity do its work, the balloon naturally spreads out, flattens against the ceiling, and fills up the whole space efficiently.

The researchers figured out the exact timing and speed for this to happen, ensuring that when we store energy underground, we don't lose it to leaks, and we use every inch of the available space. It's like knowing exactly how to pour water into a curved glass so it fills perfectly without spilling over the rim.

Technical Summary

1. Problem Statement

The study addresses the fluid dynamics of injecting gas (e.g., hydrogen or CO₂) into water-saturated, curved, axisymmetric porous channels. These channels serve as idealized models for dome-shaped anticlines, which are promising geological formations for large-scale underground energy storage.

• Key Challenge: Unlike flat or gently inclined aquifers, curved anticlines introduce complex interactions between injection-driven viscous forces, buoyancy, and geometric curvature.
• Objective: To derive a mathematical framework describing how a gas plume evolves, spreads, and interacts with the channel boundaries in these curved geometries. The goal is to understand how geometry influences storage efficiency, plume containment (preventing migration beyond the "spill point"), and interface stability.

2. Methodology

A. Mathematical Formulation

The authors model the system using Darcy's law for two immiscible, incompressible fluids (gas and water) in a porous medium.

• Geometry: The channel has a fixed width $h$ and a curved centerline defined by a shape function $F(s)$. Two specific geometries are analyzed:
  1. Parabolic Channel: Weakly curved initially but develops large slopes over long distances ($z \sim -r^2$).
  2. Gaussian Channel: Weakly curved everywhere, asymptotically approaching a flat profile ($z \sim e^{-r^2}$).
• Coordinate System: A local curvilinear coordinate system $(s, n, \theta)$ is aligned with the channel centerline to handle the curvature.
• Scaling: The problem is non-dimensionalized using the channel width $h$, radial scale $R$, and injection rate $q$. Key dimensionless parameters include:
  • $\epsilon = h/R$ (slenderness ratio, $\epsilon \ll 1$).
  • $M = \mu_g / \mu_w$ (viscosity ratio, $M \ll 1$ for gas).
  • $\lambda$ (Darcy-Rayleigh number, representing the ratio of buoyancy to viscous forces).

B. Derivation of Evolution Equations

Exploiting the slenderness of the channel ($\epsilon \ll 1$), the authors derive a composite asymptotic approximation:

1. Pressure Assumption: Pressure is hydrostatic in the direction normal to the centerline.
2. Evolution Equation: A single partial differential equation (PDE) is derived for the interface position $H(s,t)$ (normalized distance from the lower boundary).
   • The composite model (Eq. 30a) retains exact buoyancy effects and is valid for large channel slopes.
   • A small-slope limit (Eq. 33a) is derived for weakly curved channels, simplifying the governing equations while retaining the essential physics of buoyancy and injection.

C. Asymptotic Analysis

For the limit of high gas mobility ($M \ll 1$), the authors perform asymptotic analysis to identify distinct temporal regimes.

• Parabolic Channel: The solution is partitioned into five temporal regimes (I–V), each characterized by different spatial regions and dominant physical balances (injection vs. buoyancy).
• Gaussian Channel: The analysis identifies an inner-outer structure. A distinguished limit is found where $\beta \equiv M^2\lambda = O(1)$, representing a balance between buoyancy and injection.

D. Numerical Validation

The analytical reduced-order models are validated against full numerical simulations of the governing PDEs using a method of lines with a second-order finite-volume scheme.

3. Key Contributions

1. Composite Asymptotic Model: Development of a unified evolution equation that accommodates both small and large channel slopes in axisymmetric geometries, bridging the gap between flat-channel models and complex curved anticlines.
2. Identification of Five Spreading Regimes (Parabolic): Detailed characterization of the temporal evolution in parabolic channels, revealing a complex sequence of events:
   • Regime I: Injection-dominated thin gas film along the upper boundary.
   • Regime II: Buoyancy thickens the film; upper contact line decelerates and plateaus.
   • Regime III: Buoyancy drives liquid drainage from beneath the gas; lower contact line accelerates.
   • Regime IV: Formation of a thin liquid film; upper contact line resumes motion.
   • Regime V: Complete drainage of the liquid layer; the interface flattens and advances as a single front.
3. New Distinguished Limit for Gaussian Channels: Discovery that the transition between injection-dominated and buoyancy-dominated spreading in axisymmetric channels occurs when $M^2\lambda \sim 1$ (parameter $\beta$), rather than $M\lambda \sim 1$ as previously suggested for planar channels. This highlights the role of geometric attenuation of velocity in axisymmetric systems.
4. Mechanism of Buoyancy Arrest: Demonstration that in curved channels, buoyancy can arrest the upper contact line (preventing it from reaching the spill point) for extended periods, allowing the plume to thicken and maximize storage volume.

4. Key Results

• Parabolic Channel Dynamics:
  • Initially, gas spreads rapidly along the upper boundary (Regime I).
  • As the slope increases, buoyancy creates a hydrostatic pressure gradient that thickens the gas layer and slows the upper contact line ($S_u$), causing it to plateau at $S_u \sim (q\mu_w R^2 / k_0 \Delta\rho g h)^{1/2}$.
  • The lower contact line ($S_l$) accelerates as the liquid drains, eventually catching up to the upper front.
  • In the final regime, the interface becomes horizontal, and both contact lines advance at the same rate ($t^{1/2}$), maximizing the use of the channel volume.
• Gaussian Channel Dynamics:
  • The flow exhibits an inner-outer structure. The inner region (near the injection point) is pinned or evolves slowly due to curvature.
  • The outer region evolves self-similarly. The spreading rate depends on $\beta$:
    • $\beta \ll 1$: Injection-dominated (similar to flat channels).
    • $\beta \gg 1$: Buoyancy-dominated (thinner film, faster spreading).
  • The critical parameter $\beta = M^2\lambda$ determines the regime boundary, a finding specific to the axisymmetric geometry.
• Comparison with Flat Channels: In flat planar channels, buoyancy often remains subdominant to injection for long periods. In curved axisymmetric channels, the attenuation of gas velocity due to radial expansion allows buoyancy to influence the dynamics much earlier and more significantly.

5. Significance and Implications

• Subsurface Storage Safety: The findings suggest that curved anticlines can naturally arrest gas plumes via buoyancy before they reach the spill point (the location where gas escapes to a neighboring aquifer). This "pinning" effect enhances storage safety.
• Storage Efficiency: The mechanism of interface flattening and liquid drainage (Regimes III–V) maximizes the utilization of the available channel volume. A horizontal interface is more efficient for storage than a tilted or fingered one.
• Hydrogen vs. CO2:
  • Hydrogen: Due to its very low viscosity ($M \approx 10^{-2}$), hydrogen injection often falls into regimes where buoyancy effects are delayed or require very large length scales to manifest fully.
  • CO2: Higher viscosity and typical injection rates suggest a broader parameter range where buoyancy significantly influences dynamics, potentially leading to thinner gas films and different storage efficiencies.
• Practical Strategy: The study implies that injection rates and site selection (curvature and permeability) can be optimized to leverage buoyancy for plume control, potentially allowing for higher injection rates without risking spillage.

In summary, this paper provides a rigorous theoretical framework for understanding gas storage in curved geological formations, revealing that geometry plays a central role in determining whether injection or buoyancy dominates the plume evolution, with direct implications for the design and safety of underground hydrogen and carbon storage facilities.