U.S. firm mulls building geothermal power plant from Sidoarjo mudflow

THE CAUSE - in search for an explanation

N E D E R L A N D S

Here follows an attempt to explain what happened.

The Porong district at East-Java, had for long time been an area with small villages, some factories, rice-fields, a relatively big river, and a highway connecting the big cities of Surabaya and Malang. It seemed to stay that way forever. But since 29may 2006 an unstoppable stream of volcanic mud flows into houses, factories, covering roads, a railway, paralyzing people, once living and working here in peace. Since 10 years, oilcompanies discovered gas in the Wunut Field nearby, and kept on drilling new holes in the fragile earth. But the difference between any other part of the world, and this particular location is, that this place is right above the socalled "Ring of Fire". A dangerously volcano-active area. Mount Arjuna (3339m), an active volcano, is just 26 kms away. Moreover; a crack in the earth, the Watukosek fault, runs right below the drilling site. In the past longterm mudflows had already occurred. But experts of oil companies decided in 2004, that there was a high probability of finding more gas reserves here in the Kujung formation, estimated to be at a depth of 10.000ft. It would be the first exploration well of its kind in this area. Despite that no gas had been found on similar depths (10.000ft) in other locations. But, production wells here were succesful at a more shallower depth. In the beginning of 2006, the drill-site was selected, not far from the highway. Vegetation was cut away. The drilling rig and the heavy equipment were installed. The drill entered the earth and started its journey down to yet unknown layers of prosperous gas or oil. More pipes were added on top of the string, at first the earth gave way easily. Than at a depth of 6145ft the drill hit a hard layer. From than on progress downward went on extremely slow. What could have been validated as a protective screen, the drilling company experienced it as a time consuming and annoying obstruction. After 3000ft of tough drilling operations, near the point were this hard layer was finally penetrated, strange effects were noticed, such as a H2S alarm, and total loss of circulation (of drillmud). Than the hard layer was penetrated through and more problems occurred, which were resolved by rapidly plugging off the hole. It was decided to quickly retracting the drillstring back till 3580ft. Than at 4241ft the well kicked, and later the mudspill erupted 120m away.



The Mud (Spill) Volcano near Porong, in the district of Sidoarjo, East-Java, Indonesia, is unprecedented in its magnitude, in recent years. How could this happen ? The Banjar Panji-1 exploration Well, which was meant to find a gas pocket in the Kujung Formation, at a depth 8000ft or more, had been the first of its kind. In the Wunut gas field nearby, more than 20 production wells have already been installed. These have been delivering gas since 1999. But there is no production Well delivering gas from the Kujung Formation.



fig.1

Certain studies have been published which conclude that the mud disaster must be considered as a natural phenomena. However, there are many aspects to it which prove different.

Some scientists believe that the mudflow was sparked by an earthquake, near Yogjakarta, on the 27th of may 05.54 am (distance 260km). But this seems highly improbable due to the long distance, to cause any impact in Sidoarjo. Most of the publications point to the gas/oil-drilling activities of Lapindo Brantas as being the main cause. - Agnes Tuti, a researcher and social scientist at the University of Surabaya, said the gas company had been warned of the risk but ignored the advice. "PT Lapindo Brantas was warned by an American drilling company not to drill. But it decided to drill anyway, to use a protective shield. They ignored the warnings and do not want to take responsibility." (source:Washington Times)

Other sources were saying that the Lapindo Brantas drilling company did not have enough experience to accomplish well operations in such a location, with complicated sediments and formations, and the risk of a mudvolcano development.

Note that many of the publications try to explain the mud & water flow, but very scarce information exists on where the big white cloud of hot volcanic gas is coming from.



fig.2

On the 18th of may, 10days before the event, Lapindo's shareholder PT Medco E & P Brantas, sent a letter to Lapindo, and stated that it reminded the firm to set casing till a depth of 8500 feet, as agreed in the drilling program. (source Jakarta Post). Media reports say that Lapindo Brantas did not install a steel pipe casing when drilling at 3,580 to 8,500 feet as stipulated in its contract. Without the support of the pipe casing to withstand the pressure created while drilling, water and mud freely gushed through the surface. (source: Asia Water Wire).

Simulation




fig. 3

The picture above, are typical for the BPJ-1 area, before the mud-break, as they show black spots scattered around, which are originated from small sized existing gas/mud-springs. These images are from 8 months before the event. But even today you will notice them on satelite images.



Below a partial map of Java, with the red dotted line ""Sesar Simo" being a crack on which more mudvolcanoes have been developed, such as the one near Purwodadi and Bledug Kuwu. However that is a different fault-line than Watukosek, which is in SW-NE direction over Porong.



fig. 5

ANALYSIS

The story has several aspects to it. Reports which are available on www.Sidoarjo.info explain a great deal of the background. And it appears that no single (only) root cause can be pointed out, but several aspects around the BanjarPanji-1 drilling operation led to this mudspill disaster.

To mention as follows;

(1) It is near to impossible that this disaster could have happened if the normal drilling procedure had been followed, which means with the installation of a complete casing till the final depth, such as has been done in the nearby Porong-1 Well. Note that the Banjar Panji-1 ExplorationWell was the first to drill onshore into the Kujung Formation supposed to be at 10.000ft.

(2) The existance of the Watukosek fault on the map above (fig.5) had caused problems earlier when drilling the Porong-1 well. This must have been known, and taken into account, prior to the BPJ-1 was spudded. Geologists and drilling specialists on-site could have known the riscs involved by drilling straight into the Watukosek Fault, at least into the formation very close to it.

(3) Information of the drilling procedures on Porong-1 Well (in the year 1993) and Wunut-2 Wells show a realistic comparison with the drilling procedures carried out on the BanjarPanji-1 well. The Porong-1 Well has been completed with casing till approx. final depth of 8659 ft.

(4) Mudvolcanoes are not uncommon in the Sidoarjo and BanjarPanji-1 well area. All precautions to prevent one should have been taken.

(5) This Report about the formation (page3) further explains that certain assumptions and calculations in 1994 probably contain mistakes. However, warnings were given in this report for potential mudvolcano development.

(6) Main Lusi Report about Casing provides reasons why casing was deemed unnecessary at certain depths. However, there is no information available about a risk analysis to validate and substantiate this decision.

(7) From this report it also appears that uncertainty existed about the depth of the Kujung Formation, which had been bothering the drilling specialists. The exact depth of the Kujung Formation had been unknown, as well as what to expect if penetration into it would happen.

(8) Total loss of circulation occurred on 9297ft depth.This problem was solved by pumping lost circulation materials into the borehole.After the well was considered in stable condition,the drill string was pulled out of the hole till 4241ft.

(9) A "kick incident" happened on the 28th of may 2006, with the drill-bit at a depth of 4241ft. This is 1 day after the earthquake near Yogjakarta. The earthquake happened at the 27th of may 05.54am. On the 28th of may, so 1 day after the earthquake, the "kick incident" happened, while the casing shoe was at 3580ft. However, the maximum depth of 9297ft had already been reached the previous day, 27th of may, 11.00hrs (5hrs after the earthquake). If some impact of the earthquake was noticeable why did drilling continue on that day till 11:00am? Conclusion: both the moment of reaching maximum depth, and the kickincident happened at such a time, that any impact by the eartquake is highly improbable.

(10) The "Kick Incident" (while at 4241 ft) is said to be quickly killed by drillmud-injection. At this moment the drille-hole between 4241ft till 9297ft was left unprotected. The kick incident could have been an indication of formation breakage in the existing drill-hole, but at which depth, remains unclear. No reasons given why there was such a rush to withdraw the drillbit between the 27th of may 11:00am, and the next day 28th of may. It means that it was decided to withdraw 9300ft of drillpipes, install anorther (wider) drillhead, and lower the new string to 3580ft, all in 1 day.

(11) Therafter a check of communication channel between the well-bore, and the mud outflow was carried out. However, the well-bore had been locked from above at 4241ft by drill-mud. What happened below 4241 ft could not be monitored nor confirmed.

(12) Certain Seismic Reports confirmed presence of gas in the KujungFormation, the Banjar Panji-1 Well supposed to be one of many others in the region. However, it is difficult to extract information which confirms that the BPJ-1 ExplorationWell had indeed provided evidence of methane. Anyway, the H2S alarm went off on 9230ft, but it was considered a nuisance, and was ignored. It remains unclear why the equipment issued a nuisance warning, and why it was ignored.

(13) In This BPJ-1 Report , dates are mentioned as follows: a. BPJ-1 well location determined on 8march06. b. Drilling commenced from 3595ft on the 6th of may down to 9297ft reached on the 27th of may. Drill-bit has been changed at 6145ft. Note: first the complete string of drillpipes have to be taken upwards, before the drillbit can be changed. Than drilling continued. Accuracy of dates and times seem not always consistent, and sometimes appear improbable. At a depth of 9297ft, total loss of circulation occurred. The well was plugged with 'lost circulation material', and cement. Than the Drillstring pulled back to 4241ft and the well-kick occurred, which means leakage of the well into the formation.

(14) Gas and mudpressure measurements on the 27th, 28th of may would be very interesting. Gas or mud pressure while retraction of drill string till 4241ft could explain more, and maybe the reason why the decision was taken to set casing after all. Details about the gas content at this day are missing, focus is given to the mud, however everybody nearby the mudvolcano will see the big white cloud first, which has been there from the first day on. Remarkable that this issue has been neglected.

(15) In This BPJ-1 Report , it is mentioned that casing is required at 9200ft-9283ft, however while drilling through hard volcanoclastic sandstone. How to set casing at 9200-9283ft while drillhole above not cased yet, is not mentioned. It is not mentioned either to set casing between 3580ft-6145ft (top of hard formation which seals the Kujung Formation). Nor the reason of decision why it was not done.

(16) The mud probably originates from a layer between 4000 - 6000ft depth. In the BPJ-1 Report p.3 it says flowing occurred already at 3595ft, however more details about the content this "flow" is not provided.

(17) It is mentioned that the drillbit got stuck on the 28th of may at 4241ft, the situation could have been worsened by the injection of more heavy drill-mud to stop the kick while withstanding high upward volcanic gas pressure.

(18) There is low probability that the source of volcanic mud lies in the Kujung Formation. The mud is considered to come from a layer between 4000ft and 6000ft deep. However, the gas, originating from the BPJ-1 well into the Kujung Formation, appears to be the main driving force.



(19) Assuming no pressurized gas in the formation between 4000 - 6000ft deep, the source of gas can be important to assess the expected time, how long the mudflow will continue in the future.

(20) The nearest working volcano is Mt. Penanggungan (1653m) at 14km and the second Mt. Arjuna (3339m), at 26km distance. Due to the high temperature of the gas outflow, the existance of a underground link with the volcanic system of these volcanoes may be possible. (probable even). The geographic line followed by the Watokosek Fault is exactly to both of these volcanoes.

(21) Many other sources mention the earthquake at Yogjakarta to be the main cause. There are many reasonable arguments against this theory: (a) It is very coincidential that the mudflow bursted so near the BPJ-1 well. If the earthquake is the cause, the mudwell could have surfaced anywhere along the Watukosek Fault (assumed that this is the weakest area). (b) The hardpan of vulcanoclastic stone above the Kujung Formation has been measured to be 2000ft thick. It is hard to believe a spontaneous & sudden leak happened just due to this quake at such far distance. See below the correlation between distance of earthquake, and impact. LUSI is well above the line.



This chart shows the relationship between the magnitude of an earthquake, its distance to a location, and its effect on that location


(22) From one of the reports of a big stakeholder in 2005: "Brantas: 2006 Beyond : Continue to aggressively explore for hydrocarbons within the contract area." It is clear that the need for more organic fuel sources (gas) has become extremely important, with gradually depleting oil wells in this country, but yet the result of all this digging for oil and gas can be seen here, where not all the necessary precautions were taken.

note: unfortunately some key information in the different reports is not consistent.



fig. 6

POSSIBLE SCENARIO

The drillhead reached the penetrationpoint at 9200 ft. Only till 3580ft the required casing had been installed. Suddenly total loss of circulation occurred. Gaspressure became immensely high. All efforts were focussed to kill the gaspressure, but with limited result. Decision was taken to pull back the whole drillstring. In less than 24hours 9300ft of drill string were brought back to surface, another drillhead installed and lowered till 3580ft. Drilling commenced to widen the well as it was decided to install the casing after all. The drillhead finally got stuck at 4241ft. Gaspressure could not be killed and random outflows occurred at last coming together at the 'center mudflow'.

Than additionally, several other gas/mudchannels developed along the Watukosek fault line (Porong, BanjarPanji-1, Kedungbendo). The other ones closed after a few days, while the Centre Mudflow increased in size. Today large volume of gas-outpooring as a big white rising cloud can be observed. Some of the measured gasses are invisible. So the extend of the visible white cloud is not an accurate indication of the real volume of gas-outpoor. Occasional new gas outbursts are happening in random locations in residencial areas. These occurences reach the local newspapers, meaning they probably did not happen in such extend and frequency, as before the WellDrilling.

Assuming that the hot gas (temperature higher than 200C) under high pressure travels (like a jetpipe) from the Kujung Formation, upwards thru the mud-diapir (6000ft deep), than water will start boiling, its volume increases, and mud will be blown upwards. I propose a model in which the hot gas only, under high pressure, is responsible for the continuous mud-outflow, and not the mud-diapir at 6000ft itself. The mud, and water are taken upwards, by the hot gas. The white cloud is an indication of steam. Even while the hot mudwater flows through the Spillway and is pumped into the Porong River, at that point it is still steamingly hot.