Print this pageGAS HYDRATES IN THE NORTH MAKASSAR BASIN, INDONESIA
B.A. Jackson
Jackson Geophysical Consulting Pty Ltd, Nedlands, WA, Australia
A new gas hydrate province has been identified in the North Makassar Straits between the islands of Borneo and Sulawesi. The data base for the interpretation of the bottom simulating reflector (BSR) comprised 21,000 kilometres of newly acquired and reprocessed multi-client marine 2D seismic data encompassing an area of approximately 100,000 km2. The majority of BSRs have been identified within an 8000 km2 area in the deep-water West Sulawesi Foldbelt. Sediments in the West Sulawesi Foldbelt were sourced from the Mahakam Delta until the late Pliocene when a tectonic event in Sulawesi reversed the direction of sediment transport from eastward to westward. The same tectonic event was also responsible for the initiation of the west-verging fault propagation folds which created the foldbelt. The foldbelt is comprised of numerous thrust sheets creating long anticlinal structures and intervening mini-basins in which are observed numerous high amplitude reflection packages indicating the presence of coarse-clastic turbidite facies. The deposition of the fill and spill turbidite sequences had two significant effects for gas hydrate accumulation in the West Sulawesi Foldbelt. First, coarse grained turbidites provided an effective reservoir for hydrates in addition to a transport medium for migrating gas and a steady supply of water. Secondly, the turbidites were likely to transport terriginous plant matter into the deep water where biogenic processes produced the methane required for gas hydrate formation.
Most of the BSR anomalies are concentrated on the east side of the study area in the vicinity of the West Sulawesi Foldbelt approximately 300 milliseconds below the seafloor. The West Sulawesi fault propagation folds concentrate free gas below the hydrates resulting in a dramatic BSR which is almost continuously present within the foldbelt. Although the BSR fades in the intervening synclines, the hydrates are still likely to be present. On the eastern side of the foldbelt, the BSR can not be identified with certainty since there are many high amplitude turbidite slope sands and unconformities parallel to the seafloor that truncate reflectors in a manner similar to a BSR.
On the west side of the study area, in the vicinity of the toe-thrusts associated with gravity sliding in the Mahakam Delta, there appear to be fewer BSRs partially due to recent sedimentary processes that make identification difficult. The problem of BSR identification may be similar to that found in the Gulf of Mexico where, rather than the conventional through going reflector, the BSR is represented by a lineation of steeply dipping, high amplitude anomalies separated by a significant thickness of non-anomalous sediment.
BSRs are difficult to identify on the abyssal plain between the West Sulawesi Foldbelt and the toe-thrusts of the Mahakam Delta, due to the flat lying sediments near the seafloor. However, several BSRs can be identified where a through going reflector is observed in an area of pervasive block faulting near the seafloor. This observation provides evidence that there may be a more widespread distribution of gas hydrate in the abyssal plain than indicated by BSRs.
Unusual BSRs were observed within the study area in the form of a palaeo-BSR and a mud volcano. A possible palaeo-BSR was observed beneath the crest of an eroded anticline in the West Sulawesi Foldbelt. Erosion of the anticline would have caused the downward migration of the gas hydrate/free gas phase boundary thereby dropping the temperature at the palaeo-BSR and turning the free gas into high-velocity gas hydrate lens. Angular truncation of the back limb reflectors is observed at the seafloor providing strong evidence that significant erosion has taken place. As would be expected, the proposed palaeo-BSR exhibits the opposite seismic polarity compared to the underlying, present day BSR.
A mud volcano is located at the crest of a fault propagation fold in the West Sulawesi Fold Belt where significant fluids may have risen up the thrust fault from depth. The BSR is 250 milliseconds below seafloor at the crest of anticline directly below the mud volcano whilst on the flanks of the anticline, the BSR is at 300 milliseconds, suggesting that the geothermal gradient is higher in the vicinity of the mud volcano due hot fluid expulsion. The fluids rising from depth may also contain higher-end hydrocarbons and thermogenic gas. Massive hydrate mounds at the seafloor are often associated with such expulsion sites but have not been imaged by the relatively sparse areal coverage of the 2D surveys in this study.
Numerous debris flows and slump features have been observed in the study area which appear to be directly related to the gas hydrate stability zone (GHSZ). Elevated pore pressure at the base of the GHSZ caused by a periodic eustatic sea level drop and related gas hydrate dissociation are thought to be one cause of submarine slope failure. In addition, the North Makassar Basin is a tectonically active area with both east vergent and west vergent compression from the Mahakam Delta toe thrusts and West Sulawesi Fold Belt, respectively. Earthquakes associated with the compressional tectonics could have triggered numerous submarine slides within the North Makassar Basin resulting in massive debris flows up to 2400 km2 in area. An example is shown where the base of the debrite is approximately 300 milliseconds sub-bottom and is comparable to the sub-bottom depth of the adjacent BSR. The slope at the base of the debrite is 0.6 degrees.
There is also seismic evidence of rotated slump blocks that sole out at the base of the GHSZ and may be the precursors to a massive debris flow in which the bedding is disintegrated resulting in a typical chaotic internal seismic character. An example on the eastern side of the West Sulawesi Fold Belt shows a rotated fault block with an underlying listric fault that appears to sole out at the base of high amplitude reflectors at about 300 milliseconds sub-bottom. Although no definite BSR is present, the base of the GHSZ is inferred to be at about 300 milliseconds sub-bottom, similar to other areas within the West Sulawesi Fold Belt, and coincident with the change from shallow high frequency, high amplitude reflectors to deeper, low frequency, low amplitude reflectors. The average water bottom slope is approximately 1.3 degrees.
The geothermal gradient in the study area was determined from BSRs by estimating the pressure at the base of the GHSZ via a combination of hydrostatic (water column) and lithostatic (water bottom to BSR) pressure which yielded a temperature from the gas hydrate phase equilibrium relationship. The thermodynamic phase equilibrium curve was calculated with Heriot-Watt University’s HWHYD demo program (Tohidi, 2003) using structure I hydrates with a composition of 98% methane, 1% ethane and 1% carbon dioxide with 3.5% salt. A hydrostatic gradient of 9.8 kPa/m (0.433 psi/ft) was used for the water column and a pressure gradient of 18 kPa/m (0.795 psi/ft) was used in the sediment layer within the GHSZ to determine the total pressure at the BSR. Estimates of water bottom temperature were determined from empirical equations derived by Beardsmore & Cull (2001) using Bottom Water Temperature (BWT) data published in the Journal of Deep-Sea Research between 1984 and 1989 and which relate seafloor temperature, Tsf (°C), to water depth, z(m), and latitude, L(degrees). The geothermal gradient calculated from the BSR data appears highest in the centre of the Makassar Strait with a value of 6°C/100m and decreases to the east to about 3°C/100m in the southern part of the West Sulawesi Fold Belt and 2°C/100m in the northern area. Overall, the BSR derived geothermal gradients appear reasonable and show promising values for the maturation of hydrocarbons within the deeper parts of the West Sulawesi Fold Belt. The mean value for over 13,000 points at the BSR locations was 4.7°C/100m.
In order to calculate heat flow from the BSR derived geothermal gradients, an estimate of thermal conductivity was made based on a derived relationship between thermal conductivity and p-wave velocity (Horai, 1982). Data from the DSDP Leg 60 holes, Central America and the Nankai Trough (Yamano et al, 1982) were used to create a linear regression between p-wave velocity vs. thermal conductivity. An average p-wave velocity of 1.625 km/s was determined from several stacking velocity analyses on the 2D seismic data within the West Sulawesi Foldbelt and was used to estimate an average thermal conductivity of 1.016 W m-1 C-1 for the entire study area. The geothermal gradients were multiplied by the average thermal conductivity resulting in a BSR derived heat flow map for the study area. The map compared favourably with a heat flow map for SE Asia recently compiled from all available heat flow data by the SE Asia Research Group at the Department of Geology, Royal Holloway University of London (Hall, 2002).
Estimations of gas hydrate sourced methane depend on the areal extent, reservoir thickness and porosity, gas hydrate saturation and a hydrate gas yield volumetric factor that defines how gas hydrate converts to gas at standard pressure and temperature. A commonly used value for the hydrate gas yield volumetric factor is 164m3 of gas for every 1m3 of gas hydrate, assuming a 90% gas-filled hydrate lattice. A gas in place estimate was computed for the area encompassing ~8000 km2 contiguous BSR within the West Sulawesi Fold Belt. An estimated 40% gas hydrate saturation was used for this study based on a compromise between the very low pore space saturations of around 5-6% in the Blake Ridge to the high saturation of 80% established for the Nankai Trough. Using a porosity of 36% and a gas hydrate saturation of 40% of pore space, a 10m thick reservoir resulted in an estimated gas in place of 1.89 trillion m3 (67 TCF) at standard pressure and temperature.
Free gas zone estimates depend on similar parameters with the exception of the hydrate gas yield volumetric parameter which is replaced by the gas expansion factor for the expansion of gas to standard temperature and pressure. Once again, a 10 metre sand with 36% porosity is assumed as well as the same area. Pore space gas saturation is unknown but has been estimated in other areas of the world such as the Niger Delta (Hovland et al. 1997), to be quite low, in the 3-5% range. Average pressure and temperature at the BSR is around 21 MPa and 17.6°C, respectively, yielding a gas expansion factor of 263. Again, using a porosity of 36% and assuming a gas saturation of 5% for the entire area of 8000 km2, the estimated free gas in place within a 10m reservoir is 379 billion m3 (13 TCF) at standard temperature and pressure.
Although the hydrate and free gas volume estimations are very large for the West Sulawesi Foldbelt, the main problem is producing the deep water gas in an economic manner. Production of gas hydrates is the subject of ongoing research in several countries such as Japan, Canada and the United States.
In summary, the east side of the North Makassar Strait is a significant, new gas hydrate province containing considerable volumes of methane. The area exhibits several hydrate related phenomena such as mud volcanoes, palaeo-BSRs and submarine slides which, although not new, are rarely seen together. Furthermore, the gas hydrates overlie what may prove to be a prospective conventional hydrocarbon exploration area in the West Sulawesi Fold Belt. The information provided in this paper may provide an impetus to oil companies to investigate further gas hydrates whilst exploring for conventional hydrocarbons. Geothermal gradients derived from the BSR database averaged 4.7◦C/100m whilst derived heat flow values varied from 20 mW m-2 to 60 mW m-2, comparing favourably with regional heat flow data.
Reserve estimations for the most contiguous accumulation of gas hydrates (~8000 km2) yielded an estimated 1.89 trillion m3 (67 TCF) of methane gas. Reserves in the free gas zone beneath the gas hydrate in the same area were estimated to be 379 billion m3 (13 TCF).
The author would like to thank MIGAS, TGS-NOPEC and WesternGeco for making available the seismic data and for permission to publish this study.