Lusi – The “man made” volcano?

Not all natural phenomena that carry the name “volcano” are caused by the eruption of some kind of magma. A mudvolcano is one of them. Since the 29th of may, 2006, mud and steam are spewing from the earth on the island of Java, Indonesia, leaving  thousands of people homeless. Today, it is still “erupting”. At that time, a drilling rig was in the direct vicinity, and the drilling rig had experienced some problems with the well the days before. On the 26th of may, a M6.3 earthquake occured 250 km to the southwest of the location, with several aftershocks, leading to those problems with the well. So….. Who did it? What caused the Lusi Mud Volcano to form? Was it the oil company, was it nature, are they both to blame?

Let’s start off by mentioning that despite the general idea you might get from watching the Discovery Channel oil and rig shows, drilling a well for oil and gas nowadays is in general a very delicate, well planned and carefully managed operation. Countless rules, regulations, industry standards, best practices and company regulations have to be followed in order to begin or continue drilling operations.

Two of the key things to understand what could have caused this disaster are the management of formation pressure and the design of a casing scheme.

When we are thinking about volcanism, we are in general talking about very high pressured ‘fluids’ that are stored in a magma reservoir, that crack, break, melt and blow their way through overlying formations from their source, often tens of kilometers down. Those fluids can be almost continuous (undeveloped hot fresh stuff, like basalt) or fluids in a sort of matrix of crystals (viscous mush, like rhyolite). When we talk about oil, gas or water reservoirs, we are talking about fluids and gasses that are stored inside the pore spaces of existing rock at some 2 to 3 kilometers down (in this case). Those pore spaces can be very small, like in shale/claystone, or even nonexistent, as in a homogeneous rock salt. Formations that have a large percentage of pore space are generally called reservoirs, because they can contain gas, oil or water that can flow through the rock. Sandstone and carbonate-rocks are the main types of reservoirs that are of interest for oil&gas, geothermal or drinkwater prospectors, despite the recent trends in shalegas.

When we drill down, all pore spaces we encounter have something in them, usually fresh or salt water, sometimes oil or gas, with a certain pressure. This pressure is caused by the weight of the overlying rocks pushing down on it, and sometimes by local geological stress. The deeper you go, the higher the pressure. This pressure can be expressed as an absolute measure, for example 400 bar at a depth of 3000 meters (let’s stick with metric for ease of understanding and calculation). We can also express this as a pressure gradient, a number that stays somewhat constant no matter how deep you go. We know that just fresh water (which has a specific density/gravity or SG of 1.00) will give you a hydrostatic pressure of 9.81 bars at 100 meters depth, so we can calculate that the 400 bar at 3000 meters is roughly equivalent to a hypothetical fluid which is 1.31 times heavier than water. If we would fill our hole from top to bottom with a fluid of 1.31 SG, we would nicely balance the pore pressure at 3000 meters depth to 400 bar, so that no fluids enter or exit the wellbore into the formation. This is the stable situation that is needed for drilling the formation. If the hydrostatic pressure caused by the column of fluid is increased too far above the pore pressure of the rock, the drilling fluid will leak into the pores of the formation, which is called “losses”. If the hydrostatic pressure is lowered too far, formation fluids and/or gas can start rushing into the wellbore, which is called a kick or influx. If a kick goes on without controlling the pressure down in the hole in some way, it can evolve into a uncontrolled flow or even a blowout at some point.

Pressure gradients in formations lie almost always between a hydrostatic gradient (1.00 SG) and a lithostatic gradient (+/- 2.6 SG). In rare cases, the pressure can exceed those gradients. A gas or oil reservoir that has been depleted will generally have a gradient lower than 1.00 SG, down to practically 0.00 SG in some places. On the other hand, some geological processes can cause a gradient greater than 2.6 SG in some extreme cases.

The area seen from the sky. This used to be a place where you could drive through villages and meet people.

It often happens that drilling proceeds through various formations with quite different formation pressure gradients from formation to formation. This poses a problem, because drilling cannot be resumed. If two different formations need fluids with very different specific gravities, there is no right number to choose. The stable condition that is required for continuing the drilling is not possible anymore.  If a formation is expected where the pressure gradient will be a lot higher or lower than the previously penetrated formations, preventative measures have to be taken. A steel pipe, called casing, with a diameter slightly smaller than the wellbore, is lowered into the well all the way to bottom. Cement is then pumped into the void between the casings outer wall and the wellbore to isolate and seal off all the formations that have been penetrated up (down?) to that point. If everything went according to plan, the casing and the cement will serve as a barrier to allow drilling to continue with a smaller drillbit and with a fluid that has the required density for the upcoming formation.

Before a well is drilled, there has to be a complete plan in place regarding the drilling fluids to be used and the casings that are planned to be run. Some wells need only 1 or 2 casing intervals, others need up to 10 different casing intervals because of the changing conditions down in the hole. If several wells have been drilled in the area already, things get a bit simpler, because you have a very good idea which formations you will encounter and you can look at the data from previous wells to see if they had problems with specific formations.

Now, with all of this information in mind, we can start looking at what actually happened in Indonesia.

A drilling spot was picked by the oil company to target a reservoir. The area is seismically active (as everything in Indonesia) and there is an active conventional volcano at 15 km from the drilling site. The area is particularly known for the existence of mud volcano’s, several can be found in the area, all getting their feed of water/mud from the same overpressured carbonate formation that stretches the area. Drilling started and a casing was cemented in place. Drilling commenced with a smaller drillbit and just below the casing, the strength of the formation and quality of the casing and cement was tested by performing a so-called leak-off test. The cement and the formation were confirmed to be strong enough to continue safe drilling.

Construction workers building a dam around the area to direct the flow of the mud.

On the 26th of may, a M6.3 earthquake shook Yogyjakarta, some 250 km away. Right after the quake, 7 minutes later, it was noted that some drilling fluid was lost to the formation, most likely into a fracture or fault. Two aftershocks followed, leading to the immediate total loss of all drilling fluid in the hole. The losses were cured using a special blend of chemicals to seal off the fracture or loss zone, which is a very common occurence in the industry. After this, drilling resumed. A day later, overpressured limestone was encountered, leading to the influx of formation fluid into the well. The influx was “killed” by increasing the density of the drilling fluid until the flow stopped. Barely a day later, mud, steam and gasses started spewing from the earth less than 200 meters away from the drilling site.

So, who or what is to blame? A recent statement of geologists and drilling engineers in the court case points to a human cause. In the statement it is mentioned that the drilling program for the well had not been followed. A casing was planned to be set some 250 meters above the formation that later on provided the catastrophic influx that cracked and broke through the overlying rocks that had limited strength. After breaking through the final sealing formation just above the high-pressured zone, the cap was off the shaken bottle.

Many of the other arguments can be found on the wikipedia page of this event: http://en.wikipedia.org/wiki/Sidoarjo_mud_flow but it goes beyond this post to go into them in detail. Have a read through if you are curious.

The main thing is, that it is very well possible for human beings to have a devastating impact on the surroundings if great care is not taken. This example is probably the best candidate for being a truly man made volcano, although it doesn’t erupt any real lava. As this shows, it doesn’t take magma at 1000 degrees Celsius oozing or blowing out of the earth to do great damage to a village, just a bit of water and mud will have a devastating impact as well when driven by anomalously high pressure.

The main “vent”. It certainly does remind of a shield volcano.

The eruption started off with flowrates in excess of 100.000 m3/day, or more than 0.1 m3 km in 3 years. That flowrate has gone down a bit now, but geologists expect that it can take up to 35 years longer before the flow ceases. Meanwhile, all this material is being extracted from just below the eruption site. Subsidence is already measurable and even the formation of a small caldera is expected the next couple of years. Some 40.000 people have already had to flee their homes, a staggering number that even most large conventional eruptions cannot match.

El Nathan