This is my first home made capillary seal experiment. The setup is a 1/2 in burette filled glass beads of different sizes. 4 mm diameter color less beads for reservoir, and 1.5 mm black beads for the seal. The blue beads (2 mm) are used to keep the small light black beads from falling down into the "reservoir" section and from being blown upwards by leaking air bubbles. Column is filled with water dyed with red food coloring. Air is injected slowly from bottom using a air bulb.
 |
| Figure 1. Setup of capillary experiment - burette tube filled with water colored red with food coloring, and different sized glass beads to represent reservoir, and seal facies (black). |
An air column of about 3 cm is reached when addition air injected start to leak above the seal. The air water contact fluctuated by about up to ±0.5 cm depending on rate of air coming up due to variable injection rate. Injection may also be pushing on the water below in the closed system. However, the final air-water-contact did did not change for 3 hours after injection stopped.
 |
| Figure 2, final steady state trap - 3 hours after injection stopped. |
Observations:
1) Capillary seal works even with a very porous rock - the sealing beads are 1.5 mm in diameter, and pore throats are probably at least 0.5 mm. I am guessing the porosity is 40% and peaceability is >100 Darcies. Yet, it is able to trap an finite air column below.
2) Leakage rate and injection rate are not quite same due to unsteady injection rate with the bulb. Injection may also be pushing the air-water-contact up in the closed system. This may have led to some fluctuation of the column height.
3) However, the column height did not change with time once injection stopped as expected of a capillary seal. The final column is about the average of the entire session. This disputes the idea that capillary seal capacity is reduced after initial leak.
4) Migrating air bubble/slugs often get trapped along the way (migration loss), and migration resumes when new air merges with trapped ones (Figure 2). As a result, the first bubble reached the trap after some volume has been injected. This is called "migration lag" (the time needed to fill these small accumulations / "saturate" the carrier). Continued migration requires continued supply.
Below is the video of one of the sessions. Sorry about the shaky video recording with my phone in one hand and the air bulb in the other.
