November 13 Class Summary
TodayÕs class was a talk by Max Rudolph, about his work on the Lusi mud volcano in Indonesia.
Relevent paper: Rudolph, M.L., Karlstrom, L., Manga, M., 2011. A prediction of the longevity of the Lusi mud eruption, Indonesia. Earth Plan. Sci. Lett. 308, 124-130.
á We spent some time discussing the history of the Lusi volcano. It began erupting about 200 m from a natural gas exploration well, which probably triggered the initial eruption after the blowout cap was put in.
á Three models currently attempt to explain the cause of the eruption. The first, Davies, et al. (2011), postulates that a deep aquifer of CO2-saturated water is flowing up through the drill hole into a clay layer and mobilizing the mud.
á The second model, by RudolphÕs group, postulates that the water content of the clay layer is sufficient to fuel the eruption, except for perhaps its initial stages. The water content of the clay is estimated to be about 30%.
á The third model suggests that a deep hydrothermal system related to a nearby silicic volcano is providing the necessary fluid for the Lusi mud volcano.
á Rudolph et al.Õs model is conceptually similar to the one from Monday (Karlstrom et al., 2012) for caldera formation of a mud layer. The mud chamber at depth is assumed to be symmetrical and cylindrical, and it grows as deviatoric stress mobilizes increased volumes of mud on its margins.
á The main difference is that in the magma model, there is a limit to the amount of magma that can be mobilized, because the modeled locked magma chamber is finite. The mucd layer is assumed to be infinite in extent, however, in the Lusi model.
á The surrounding rocks are assumed to behave elastically.
á The conduit to the surface is assumed to be a single conduit (there are actually two right now), and is assumed to be cylindrical and symmetrical. There is also no modeled drift velocity of bubbles relative to the surrounding mud. Simple conservation of mass an momentum equations governed flow in the conduit.
á Changing whether the mud is CO2 of CH4 saturated does not affect the results of the model.
á The model predicts an end to the eruption within 21 years with 33% probability, within 40 years with 50% probability, and within 84 years with 67% probability.
á The model ended eruptions in three ways: there was no longer overpressure in the mud chamber relative to lithostatic pressure, caldera collapse occurred, or time-out at 100 years duration (stopped due to computing time constraints).
á Finally, he discussed new improvements on the model since the 2011 paper. Notably, output is now assumed to decrease exponentially with time, with output decreasing to 104 cubic meters per day by 2017, but this is poorly constrained due to the great difficulty of measuring discharge rates.