A Geyser Made of Glass

Brent Cullimore

Thanks to fun headlines like “Geysers erupt because they’re all bendy inside,” this paper caught my attention in 2014:

“Geyser preplay and eruption in a laboratory model with a bubble trap,” Esther Adelstein, Aaron Tran, Carolina Muñoz Saez, Alexander Shteinberg, Michael Manga; Journal of Volcanology and Geothermal Research, 285 (2014) 129–135. A pre-print is available here.

These fine people built a desktop geyser! That’s right, a model of a geyser that would erupt over and over in front of them, built out of glass so they could see what is going on inside. So of course, I just had to make a model of the model.

When you’re an engineer, you’re used to getting ribbed for the reading material that sits on your nightstand. Oscillating Heat Pipes. Advances in Automotive Engineering. Hot Bearings I Have Known.

But geysers fascinate everyone. Here’s one in Iceland that I had the pleasure to watch explode slowly several times, with a rush of bubbles visible just under the plug of uplifted liquid water as it hits the surface.

Why do geysers cycle like this? Is it because it takes some superheating of the hydrostatically pressurized water that just came rushing back into the hot spot from cooler springs nearby? Is it because a slow bubbling becomes a runaway process as the rising bubbles expand quickly? (This is also what happens with magna rising in a volcano, enhanced by gasses coming out of solution. But that will have to wait for a future model.)

Or, is it because geysers are just plain bendy inside?

Each geyser has its own unique behavior, and they rarely exhibit a simple repeating cycle. They can’t all be called Old Faithful, after all.

And even that geyser does not exhibit a simple repeating cycle. It is bimodal. According to Wikipedia,

“The time between eruptions has a bimodal distribution, with the mean interval being either 65 or 91 minutes, and it is dependent on the length of the prior eruption. Within a margin of error of ±10 minutes, Old Faithful will erupt either 65 minutes after an eruption lasting less than ​2 1⁄2 minutes, or 91 minutes after an eruption lasting more than ​2 1⁄2 minutes.”

Since each geyser has such a unique pattern, volcanologists (and aspiring volcanologists) wonder, “What the heck is going on beneath the surface that can explain this behavior?”

The hypothesis explored by the Glass Geyser People was: What if the main path that the water and steam follow on their way up is not straight? What if it has some reversals in elevation … some ups and downs … that can trap steam? A picture from their paper shows the desktop geyser and some typical points in its operation. They did find at least two distinct cycles: a short one and a long one. (Plus a few medium duration ones, but those just obscure an otherwise good story.)

It took me a few years to find enough nights and weekends to generate a decent thermal/fluid model. Part of the delay involved the challenges in representing the eruptions in the very small (40cc) lower reservoir: the complex ways in which the liquid and steam enter the conduit above. In the meantime, I was able to take advantage of new features that had been introduced each year in the CRTech tool suite.

The result isn’t exactly a validation case, and I don’t expect a sudden rush of calls from interested volcanologists. (OK, I dream of such calls. “Please fly to X to help us study Geyser Y!” But I don’t expect them.) There were two sizes of eruptions, but I'm not ready to declare it to be bimodal.

I will have to settle for a more realistic hope: that engineers trying to figure out what thermohydraulic programs like FloCAD can do will get a flavor of this specialized field, without having to learn the complex terminology or design concerns of a more typical application in the aerospace, energy, or automotive industries.

And maybe they’ll have a little fun in the process. Remember, geysers fascinate everyone. Even different types of engineers.