Unlocking Rock Secrets: Seeing Inside Sandstone with X-rays and Neutrons

Hey there! Ever wondered what’s *really* going on deep inside a rock, especially something as porous and complex as sandstone, when you squeeze it or push stuff through it? It’s not just a solid, boring block! Understanding the nitty-gritty of how these rocks behave under pressure and flow is super important for things like stashing away CO2 underground, tapping into geothermal energy, getting oil and gas out, or cleaning up groundwater. Predicting how reservoirs will react? Yeah, you need to know the rock’s secrets.

The catch? Most of the time, we’ve only been able to measure what happens at the *edges* of a rock sample. Like trying to understand a whole party by just listening at the door! We knew things like structural quirks, unevenness (heterogeneity), and localized damage (like tiny cracks or shear bands) could totally mess with how easily fluids flow through the rock. And predicting how these properties change when the rock deforms? That’s been a real head-scratcher because lots of tiny things are happening all at once – grains cracking, sliding, breaking free, or cracks closing up.

Especially in sandstones with lots of empty space (high porosity), the stress you put on them changes *everything*. Squeeze them gently, they might crack in a brittle way. Squeeze them hard, they might flow more like something ductile. And as you squeeze, they first get smaller (volumetric compression, less empty space) then bigger (volumetric dilation, more empty space) before they fail. All sorts of tiny damage mechanisms are playing out.

Past studies, often relying just on those edge measurements, sometimes found conflicting results. Does permeability (how easily fluid flows) always go down as you squeeze? Not always! Some saw it decrease, others saw it increase, depending on the squeezing stages. It was clear we needed a way to see *inside* the rock as it was happening.

Peeking Inside: Our High-Tech Eyes

That’s where our awesome tools come in! We decided to get a much better look at the action using imaging techniques, specifically X-ray and neutron tomography. Think of tomography like getting a super detailed 3D scan, slice by slice. X-rays are great for seeing the rock structure itself – the grains and the pores (the empty bits). Neutrons, on the other hand, are fantastic for seeing water, especially when you use a clever trick with heavy water (D2O) and normal water (H2O). They look different to neutrons, giving us contrast to track fluid movement!

Our goal was to really understand the hydromechanical behavior of a specific porous sandstone, Idaho Gray. We wanted to link what we saw happening inside (the full 3D picture) with the overall measurements at the edges. We focused on two big questions:

• How does the initial pattern of empty spaces (porosity distribution) affect where and how the rock deforms when squeezed?
• How do that initial structure and the ongoing deformation influence where the fluid goes?

To do this, we took our sandstone samples to a special facility (NeXT-Grenoble) and performed coupled tests: squeezing them (triaxial compression) and pushing water through them, all while taking incredibly fast and detailed neutron tomography scans. We’re talking 4D here – 3D space plus time! We also used high-resolution X-ray scans beforehand to map out the initial structure in detail.

We used some pretty cool analysis techniques too. Digital Volume Correlation (DVC) is like tracking tiny patterns in the 3D images to see how the rock is moving and deforming internally (calculating strain). For the fluid flow, we tracked how the H2O front advanced through the D2O-saturated sample over time.

Getting Our Hands on the Rock (and the Data)

We used cylindrical samples of Idaho Gray sandstone, about an inch in diameter and two inches long. This sandstone is mostly quartz and feldspar with some clay, and its grains are cemented together by silica. The X-ray scans we did first were super high-resolution (15 µm voxels – tiny cubes!), letting us clearly see the grains and pores. We found the samples were quite porous, around 35%, with grains averaging about 300 µm.

The neutron scans during the experiments were done at different speeds and resolutions. High-resolution scans (43 µm voxels) were taken between squeezing steps to see the accumulated deformation. High-speed scans (170 µm voxels, taking only 1 minute!) were done *during* the fluid flow tests to capture the water moving in near real-time. The D2O/H2O contrast was key here.

We squeezed the samples under a relatively low confining pressure (1 MPa), putting them in the brittle failure regime. One sample (IG16) actually had a brief moment at a higher pressure (5 MPa) due to a system hiccup, but we adjusted and confirmed the pore fluids weren’t affected. We squeezed them slowly, measuring the force and overall volume changes.

During the flow tests, we injected H2O at a constant rate into the D2O-saturated samples, measuring pressures at both ends to calculate the overall bulk permeability. The neutron scans let us see the fluid front advancing.

Peeking Inside: What X-rays Showed Us First

The initial X-ray scans were eye-opening. These samples weren’t just uniformly porous. Nope, they had clear zones of higher and lower porosity. In one sample (IG24), the transition from high to low porosity was inclined across the sample, likely reflecting the original sedimentary layering (bedding). The other sample (IG16) had a more vertical high-porosity zone. These variations in structure are super important because they set the stage for how the rock will behave.

We also looked at other properties related to the pore space from the X-ray data: the average size of the pores (equivalent pore diameter) and how twisty the paths through the pores are (tortuosity). As you’d expect, higher porosity generally meant bigger pores and less tortuous paths. This initial mapping gave us a baseline to compare against what happened during the experiments.

Interestingly, when we looked at the orientation of the individual grains in the X-ray images, we found a preferred alignment, especially in the regions where shear bands eventually formed. This suggests the original grain orientation might have played a role in guiding where the rock decided to break.

Squeezing the Rock: How it Deformed

The overall mechanical behavior of both samples was typical for brittle rocks: they got stiffer as we squeezed, then started to yield, and finally failed after a relatively small amount of overall squeezing. They showed that classic pattern of initial volumetric contraction followed by dilation before failure.

But the DVC analysis of the neutron images let us see *where* this deformation was happening inside. Before the rock failed, the volumetric contraction (getting smaller) was more pronounced in the areas that were initially *more* porous. Conversely, the areas that were initially *less* porous tended to show more dilation (getting bigger). This is a cool finding – the empty spaces you start with dictate how you squeeze and expand locally!

As we squeezed harder, the deformation started to concentrate. Eventually, clear shear bands formed – localized zones where most of the deformation (both dilation and shear) was happening. After the shear band formed, the link between initial porosity and deformation wasn’t as clear; the shear band dominated the strain pattern.

Comparing the two samples, IG16 (with the more vertical porosity pattern) seemed a bit stiffer and stronger but failed earlier than IG24 (with the inclined pattern). This suggests the way the initial heterogeneity is arranged really impacts the rock’s overall strength and how it deforms. The pronounced strain concentration we saw in IG24 might be why it was weaker.

The brief higher pressure experience for IG16 might also have played a role, potentially pre-compacting it slightly and affecting its later behavior, but that needs more digging.

Where the Water Went: Tracking Fluid Flow

Watching the fluid flow with neutron tomography was fascinating! Initially, the water definitely preferred the paths through the high-porosity regions, especially those with larger pores and less tortuosity. Makes sense – water takes the path of least resistance!

As we squeezed the sample (IG24, since we had flow data for it), the flow patterns started to change. After the first couple of squeezing steps, the overall bulk permeability increased significantly. The neutron images showed that the water was accessing a larger volume of the sample – the flow area increased. This suggests that the deformation, particularly the dilation we saw happening in some areas, was opening up pathways that weren’t initially used by the fluid.

Interestingly, the flow path also seemed to shift and become more inclined as the shear band started to form. After the sample failed and the shear band was fully developed, the higher flow speeds were concentrated within that band. It acted like a preferential highway for the fluid, similar to how a fracture might.

However, even with the shear band acting as a fast lane, the initial structure still had a strong influence. The main flow path still seemed tied to the original high-porosity zones. This tells us that while deformation *does* change things, the rock’s starting point (its initial heterogeneity) is a major player in determining where fluids will go.

We also saw some hints of sand production (tiny grains breaking off) within the shear band after failure. This could potentially block some flow paths and make the remaining ones more twisty, which might counteract the permeability increase you’d expect from dilation in the band. It’s a complex interplay!

The Big Picture: Structure, Squeeze, and Flow

So, what’s the big takeaway from all this high-tech peeking and squeezing? It’s clear that for these heterogeneous sandstones, the initial structure – things like how the porosity is distributed, maybe the sedimentary bedding, and even the orientation of the grains – has a huge impact on *everything*.

• It influences where the rock deforms and how those damaging shear bands form.
• It dictates where the fluid wants to go initially.

Deformation *does* matter too! The squeezing and expanding (strain), especially the dilation, can open up new flow paths and increase the overall permeability. But in our tests, the initial unevenness seemed to have a stronger hand in guiding the main fluid flow paths than the changes caused by deformation alone.

Seeing all this happen in 4D inside the rock, rather than just measuring at the edges, gives us a much richer picture of these complex coupled hydromechanical processes. It confirms that you can’t just treat these rocks as uniform blocks; their internal personality matters!

Why This Matters

This kind of detailed understanding is crucial for making better predictions in all those geoengineering applications we talked about. If you’re trying to store CO2 underground, you need to know exactly where it might go and how the rock’s behavior under pressure might change the storage site over time. Same for geothermal energy or cleaning up contaminated water – knowing the precise flow paths and how they evolve is key.

This work provides a solid foundation, but there’s always more to do! We need to look at how these rocks behave under different pressures and keep pushing the limits of imaging resolution and analysis techniques. The datasets and methods we developed here are also super valuable for building and testing more accurate computer models of these complex rock behaviors.

It’s a challenging but incredibly rewarding field, and seeing inside the rock with neutrons and X-rays is definitely one of the coolest parts!

Source: Springer