Frac Home Lab: Instructions and Results for Porosity & Permeability Experiments

Here is the written guide for the experiments I discussed in the Porosity and Permeability Video, with an opportunity to nerd out on details. These experiments can be done at home or in your garage.

Porosity in rock provides storage for water, oil & gas. Oil is not produced from some massive downhole “cave” or “lake”. An oil “reservoir” does not look anything like Cherry Creek reservoir. An oil reservoir consists of a vast number of small interconnected pores like the exaggerated pore space in the sponge on the left. 

The thin section on the right shows a see-through slice of sandstone, showing sand grains in white and pore space in blue. Sedimentary rock was made millions of years ago from eroded particles that do not fit well together. They leave some open spaces when packed. 

If the grains are well sorted, like in a beach or dune sand where wind or water has helped select grains by size and eroded them to be well-rounded, porosity can be 25 – 40%. When Mother Nature covers this sand with 2 miles of rock, maybe a half or a third of the porosity remains after consolidation. 

Porosity is a key parameter to know because it has a direct tie to how much oil is stored in a reservoir. For example, the Middle Bakken formation in North Dakota has a porosity of about 8%. That may not sound like a lot until you consider the large area and thickness. The 8% Bakken porosity is about 50% water filled and about 50% filled with oil. The Bakken is 40 ft thick. 4% oil out of 40 ft represents an oil “column” of 1.6 ft. But here is the kicker – this knee-deep level of oil from all these connected pore spaces extends over an area of about 150 x 150 miles. That represents about 180 billion barrels of oil in place, of which we can produce about 25 – 30 billion barrels in primary production. That is enough oil to supply the world for nearly a year! 

Enough about background - time to do a few experiments with different materials.

Porosity experiment #1

Needed: Two identical glasses (ideally two long-drink glasses that are cylindrical in shape), water, marbles or beads or your mother’s pearl necklace. We recommend you ask for permission. 

Fill one glass to the top with the marbles and fill the other glass with water. Pour the water into the glass with marbles. Almost half the volume should enter the pore space, as randomly stacked marbles fill only about half of all space.

Porosity experiment #2

Needed: same as above, but replace marbles with dry sand, for example from a sandbox. Also, use a tape measure to measure the height of the column of water before and after filling the pore space (I did not have long drink glasses, so I used a scale to measure weight changes).

Pour the water into the glass with sand – but slowly, until all open space is filled with water. Measure the remaining column height after pouring contents in glass with sample. Calculate porosity using this equation:

((original water column height – final water column height)/(height of sand column)) * 100 = % porosity

Sand for a sandbox is not as well sorted and more angular than marbles. As a result, porosity is reduced as smaller particles and sharp angles fill up some of the open spaces.

You will also see that it takes a lot longer for the water to distribute in the pore space – due to lower ability of flow through pores, or lower permeability.

Porosity experiment #3

Needed: same as above, but replace dry sand with dry dirt or soil from the garden. It is important that the sample is dry, as water already in the sample will lower your porosity measurement. Therefore, if needed let the sample dry; then pack it is the glass.

This material is even more poorly sorted than the sand. 

There are a lot of additional complicating factors to porosity that these simple tests ignore, but these experiments cover the basics.

 The same is true for the next parameter we cover with these critical concepts: permeability.

Permeability in rock is the ability to flow liquids and gasses through pores. This first picture is a representation of pores in dark colors around grains in white. The picture on the right shows flow velocities. Note that higher velocities are possible in wider channels.

Permeability experiment #1

Needed: funnel, coffee filter or wire mesh, grainy materials like marbles, sand.

Place the coffee filter or wire mesh in the funnel. This helps keep the grainy material in the funnel and prevents it from washing down. There are more experiments out there that show this, and it works for smaller-grain materials. 

However, there are some problems with this test if you want to observe a reasonable correlation between permeability and flow rate or duration of a column test. Two problems with the funnel are the small exit diameter the coffee filter, which may drive the duration of the flow test due to convergence. 

Our Lab Manager Joel Siegel originally wanted to use a beer tap for this experiment. For high-permeability samples, we could not exceed a certain flow rate due to flow restrictions at the tap. For this reason, we opted for the wide “full-bore” gate valve.

Permeability experiment #2

Needed: gate valve(s), plexiglass tubes. Sieved 30 and 80 mesh proppant; steelshot; marbles; and wire mesh materials matching these material sizes.

We conducted a series of four tests (Test Tube #1 through #4 in the table below) with the materials highlighted in yellow. We selected materials with grain diameters that reduced by a factor of ~3.2, resulting in permeability changes of about a factor of 10 (one order of magnitude), as permeability is proportional to the square of the grain diameter.

We then measured the time required to drop water levels by 4 inches for tests 1 through 4. We observed how these column test durations increased from about 0.3 s to 3 s, about 30 s and then about 300 s – roughly proportional to the permeability changes. The flow rates are therefore roughly proportional to the changes in permeability.

Assuming this proportionality remains, we can extrapolate to additional tests in sands and shales with the same test conditions. Going all the way to imaginary test #16 to the lowest-permeability shales, we would need millions of years for the 4” water column to disappear. Of course, that water column would evaporate in only about a month.

On average during the flow test, the 4” water column above the test pack and water in the test pack itself provide about 5” of hydrostatic pressure on the middle of the pack. That’s about 0.18 psi differential pressure from the weight of the column. In a producing shale well that differential pressure (between far field reservoir pressure and producing well pressure) is ~3-4 orders of magnitude higher than in my garage column, getting us back to test sample #13 for a flow example, where it would still take 10,000 years for the column to flow through the pack.

In addition, fracture treatments in shale focus on creating surface area. When we create a highly complex fracture system with a fracture surface area of millions of square feet (~7-8 orders of magnitude higher than the 0.05 ft^2 area of the gate valve in my setup), we move farther to the left toward sample #5 in the table above. A reasonable shale well produces about a tablespoon of oil daily from a 1 ft^2 tile. The large fracture surface area combines all these daily tablespoons of oil into ~1,000 bbl of daily oil production we see for the first few months out of a shale well.

Orders of magnitude lower shale rock permeability is traded for orders of magnitude higher fracture surface area. That’s shale frac’ing in a nutshell.