A scenario-based framework for characterising subsurface uncertainties in shallow mine thermal flow systems

Deep beneath the cities of the United Kingdom lie vast, silent labyrinths of abandoned coal mines. For over a century, these tunnels were the arteries of industry, but today they are flooded with groundwater, forming a hidden, subterranean ocean. Scientists have long known that this water holds heat, warmed naturally by the Earth's interior, and that pumping it to the surface could provide a clean, low-carbon way to heat homes and buildings. This concept, known as mine-water geothermal energy, offers a promising path to decarbonise our heating systems. However, turning these flooded ruins into reliable power plants is fraught with uncertainty. The underground world is not a uniform sponge; it is a chaotic mix of open tunnels, collapsed rubble, and solid rock, all hidden from view. Without knowing exactly how these spaces connect, engineers cannot predict if water will flow smoothly between wells or if the heat will vanish too quickly to be useful.

A team of researchers from Heriot-Watt University and Delft University of Technology has tackled this problem by building a sophisticated digital twin of a flooded mine in Glasgow. Using real data from a deep monitoring station, they created a virtual model to test how different underground shapes and materials affect the flow of water and heat. Their work reveals a surprising truth about these systems: the total amount of heat stored in the ground matters far less than the specific pathways the water can actually travel. In their simulations, they found that if the tunnels between an injection well and a production well are not directly connected, the system fails quickly, regardless of how much heat is theoretically available nearby. They also discovered that the rock surrounding the mine plays a critical role, acting as a thermal battery that can recharge the system, but only if the mine's internal tunnels are too blocked to let the water move freely on its own.

The researchers began by constructing a detailed three-dimensional map of the subsurface, drawing on geological records and borehole data from the Glasgow UK Geoenergy Observatory. They did not rely on a single guess of what lay underground. Instead, they built an ensemble of many different possible realities, each representing a different arrangement of the mine's history. Some models featured wide, open tunnels with few obstacles, while others included partial walls of coal or large sections filled with rubble and waste rock, known as backfill. They simulated the movement of water and heat through these virtual mines over a period of twelve years, watching how the temperature changed as cold water was pumped in and warm water was drawn out.

The results highlighted a fundamental tension in designing these systems. The ability to pump water through the mine, known as deliverability, depends on having a wide, open path between the two wells. The more open the tunnels, the easier it is to move water, and the less energy is required to pump it. However, this high flow rate comes with a cost. When the path is wide and open, the cold water injected into the system rushes straight to the production well, creating a "thermal short circuit." This happens in just a few years, causing the water temperature to drop below the level needed for heating long before the mine's stored heat is fully used. Conversely, if the path is blocked by barriers or rubble, the water is forced to take a longer, more winding route. This slows down the flow, making it harder to pump, but it also delays the arrival of the cold water, allowing the system to produce useful heat for a much longer time.

Perhaps the most striking finding emerged when the researchers kept the total volume of empty space in the mine constant but changed how that space was connected. In one scenario, they created a mine where the tunnels were largely isolated from each other, forcing the water to bypass the main storage areas. Even though the total amount of heat stored in the ground was identical to a well-connected mine, the system with poor connectivity lost its useful heat twenty times faster. In these disconnected scenarios, the thermal production lifetime dropped from over three years to just a few weeks, while the ability to pump water remained almost unchanged. This demonstrates that the mere presence of heat in the ground is not enough; the heat must be in the specific part of the mine that the water can actually reach.

The study also examined the role of the rock surrounding the mine tunnels. In some models, the rock was treated as a tight, sealed barrier that water could not penetrate. In others, the rock was considered "leaky," allowing water to seep in and out of the mine tunnels. When the mine tunnels themselves were blocked or poorly connected, the leaky rock became a vital resource. The water would seep into the surrounding rock, drawing heat from the massive thermal mass of the Earth itself. This process acted as a natural recharge, extending the life of the system significantly. In the most blocked scenarios, this geological recharge doubled the time the system could produce heat. However, if the mine tunnels were already wide open, the surrounding rock had little effect, as the water preferred to stay in the easy path of the tunnels.

The researchers also looked at the impact of backfill, the rubble and waste rock often left behind in abandoned mines. They found that the material properties of this backfill, such as how porous or permeable it was, mattered very little unless it was placed directly in the path of the water flow. If the backfill was located off to the side, the water simply flowed around it, and its composition had no effect on the system's performance. It was only when the backfill blocked the main route between the wells that it became a significant obstacle, reducing the flow and altering the heat recovery. This suggests that the location of obstacles is far more important than their chemical or physical makeup.

These findings challenge the way engineers might traditionally assess the potential of a mine for geothermal energy. Rather than simply calculating the total volume of water and heat stored in the ground, the study suggests that the priority must be mapping the specific connections between the injection and production points. The most critical factor is the "swept volume"—the specific portion of the mine that the circulating water actually touches. If the water cannot reach the heat, that heat is effectively lost. The study indicates that a successful project requires a careful balance: enough connectivity to allow water to flow, but enough complexity or barriers to prevent the cold water from rushing straight through.

The work also underscores the importance of uncertainty in these projects. Because the underground architecture of abandoned mines is often poorly documented, relying on a single model is risky. The researchers' approach of testing many different scenarios provides a safer way to understand the range of possible outcomes. It shows that a mine that looks promising on paper might fail in practice if the hidden tunnels are not connected as expected. Conversely, a mine that seems blocked might still be viable if the surrounding rock can provide the necessary thermal recharge.

Ultimately, this research provides a clearer picture of how to turn old mines into new energy sources. It moves the focus from the total amount of heat available to the specific pathways that allow that heat to be captured. By understanding that the geometry of the tunnels and the nature of the surrounding rock are the primary controls on performance, engineers can design better systems. They can choose well locations that maximise the connected flow paths while using natural barriers to extend the system's life. The study confirms that while the potential for mine-water geothermal energy is significant, realising it requires a deep understanding of the hidden, complex world beneath our feet.