Formation of Accretionary Prism and Associated Fluid Flow:
A Research Strategy
Toshihiko Shimamoto*, Wataru Tanikawa, Hiroko Kitajima,
Yasutaka Aizawa and Manabu Komizo
Division of Earth and Planetary Sciences, Graduate School of Science,KyotoUniversity
Kyoto 606-8502, Japan(*corresponding author: )
Abstract
Fluid flow in accretionary prisms is receiving increasing attention in relation to IODP project on “drilling into the seismogenic zone” in Nankai Trough, planned in a few years using Chikyu. To understand the role of fluids in a variety of behavior of subducting plate boundaries from slow slip to great thrust-type earthquakes, one needs to know pore-pressure distribution and permeability structure of accretionary prism. We have thus worked in the last few years on Tertiary to Quarternary sediments in Ashigara Group, the focal area of the 1999 Taiwan Chi-Chi earthquake, Miyazaki Group, the the focal area of the 1999 Taiwan Chi-Chi earthquake, NiigNiigataBasin and Miura Group, trying to develop methodology to determine permeability structures and pore-pressure distribution from surface samples and/or core samples. Sedimentary basins were selected not only because accretionary prism is one of our main target, but also because their history both in space and time is much simpler than that of basement rocks. Our basic strategy has been three-fold: (1) measure transport properties (porosity, permeability and storage capacity) under deep conditions to estimate those properties at depths, (2) measure those properties for samples collected from all stratigraphic horizons to evaluate long-term cementation empirically, and (3) analyze interacting processes of sedimentation, fluid flow and compaction using measured properties to model the basin development. Those approaches have been successful and we show representative examples in our presentation. Also, each work will be displayed in more detail at three posters by the co-authors at this symposium. A notable example is the prediction of the development of abnormal pore-pressure in the focal area of Chi-Chi earthquake by Tanikawa which agrees nicely with drill hole data by CPC in the oil field immediately to the north of the area. Also, Kitajima showed how porosity of sedimentary rocks of Miyazaki Group decreases with time and the data can be used to test prediction of the cementation via solution/precipitation processes in the future. Another important problem is the effect of faults and fractures on fluid flow which are commonly regarded as passage of fluid flow. Recent data on fault zone permeability have been revealing that fault zones in porous sediments are more impermeable than host sediments and that presence of fractures in porous sedimentary rocks has little effect on overall permeability. We will show in the presentation at what stage of compaction, fractures and fault zones begins to work as fluid passages. The results will highlight the advantage of sedimentary rocks as waste disposal sites. We plan to apply all of those analyses to ODP cores from Nankai Trough in the next two years as a US-Japan cooperative project.
1. Introduction
Fluids and earthquakes have long been fashionable research topics, and yet the exact role of fluids on the earthquake generation is still in the realm of speculation. Even permeability structure and pore pressure (Pp) distribution at depths, needed to calculate fluid flow, are not known.Thus we initiated a series of work to develop methods to determine transport properties at depths and Pp distribution through laboratory measurements using surface samples. The focal region of the Chi-Chi earthquake is an ideal place for such studies because deep geological structures are very well known by seismic reflection studies and pore pressures have been measured at nearly 10 deep drill holes for petroleum exploration and production(Suppe and Wittke, 1977).Predictions can be tested against observation in the focal area as attempted by Tanikawaet al. (2003). Niigata and Miyazaki are also good places for testing prediction from drilling data for oil and gas (work by Aizawa and Shimamoto and by Kitajimaet al.: both in this symposium). Miura Group has been studied by Komizo and Shimamoto (this symposium) to gain insight on deformation and fluid flow in the shallow portion of accretionary prism. Work on the Shimanto Belt by Tsutsumi and Nishino (this symposium) is a study on a deep portion of an accretionary prism by similar approaches.
In our presentation, we will show how our work progressed using examples from those five areas and will propose a strategy to study compaction, cementation and fluid flow in accretionary prism primarily using basin analyses. Fluid flow in fault zones and along fractures needs to be evaluated to model the fluid flow in the accretionary prism. Fluids can flow easily along fractures and fault zones in well-cemented basement rocks which have very permeable damage zone, consisting of fault breccia and fractured host rock (Caine et al., 1996; Evans et al., 1997). On the other hand, fault zones in porous sedimentary rocks do not have very permeable zone, and porous rocks with fractures are not necessarily permeable. An important question is at what stage of diagenesis (compaction and cementation) do fault zones and fractures in sedimentary rocks become similar to those in basement rocks. We will show our preliminary data on the effects of faults and fractures in porous sedimentary rocks on fluid flow.
Studies on transport properties of sedimentary rocks have important implications for waste isolation and for the Earth’s environmental issues. Sedimentary rocks become more impermeable with increasing compaction and cementation, thus increasing seal capability. Thus optimum seal capability should exist in between where sediments are compacted enough to reduce permeability, but not enough for easy fluid flow along faults and fractures. Natural oil and gas must have made full use of these sediment properties in their accumulation. One issue we want to learn in this symposium is the migration mechanisms of gas such as methane and dissolved ions in porous sediments. Understanding those mechanisms will lead to further insight into material circulation in the Earth (e.g., subduction zones), waste isolation and spreading of pollutions.
2. Permeability structure of the Ashigara Group
A simple approach to estimate transport properties at depths is (1) to measure permeability, porosity, and storage capacity at elevated pressures and temperatures, and (2) to do the same measurements for samples from all stratigraphic horizons to evaluate long-term cementation empirically. Figure 1 is such an attempt for Ashigara Group which Tanikawa has done for his graduation thesis in 2001. A complexity about using surface samples is how to evaluate the effects of burial and uplifting/erosion history. As a first approximation, we have regarded burial and subsequent uplifting and erosion as a pressure cycling process as shown for two cases A and B in Figure 1a. In view of our many pressure cycling tests, sample A collected at a surface outcrop would show a permeability or porosity vs. depth curve starting from “sample A” and joining the normal burial curve (SABD) at about the maximum burial depth A. Thus the normal burial curve can be estimated by using samples with different burial depths (samples from different stratigraphic horizons). Above the maximum burial depth the permeability depends highly on the history of the sample, and permeability measured at low effective pressures must be interpreted with caution.
We thus constructed porosity-depth curves for the Ashigara group, which constitutes a small basin east of Mt.Fuji, central Japan (Figure 1b). This group thrust underTanzawaMountains in
the north across Kan-nawa Fault, commonly regarded as a plate boundary fault. Samples were collected from several stratigraphic horizons, shown as the burial depths in the inset table, and porosity on the vertical axis was estimated from measured permeability using the relationship by Chilinger et al. (1963). Results are similar to logging data from various oil fields except at low effective pressures where permeability and porosity are affected by previous history of samples. At greater depths, porosity from samples with deeper burial depths (older samples) clearly exhibits smaller porosity than those from shallower burial depths (younger samples). This difference is probably due to the effects of time dependent compaction, chemical cementation and tectonic deformation. Such a trend was not recognized for siltstone from Ashigara group.
3. Abnormal pore pressure in the focal area of the 1999 Chi-Chi earthquake
The above method is simple and quick to predict transport properties at depths, but it is not precise because sediments cannot compact unless interstitial fluids move out. Thus, so-called basis analysis has to be carried out to predict sedimentation, fluid flow and compaction. Basin analyses are extension of consolidation problem and theories exists for a long time (e.g., Gibson, 1957). But an important aspect is that transport properties of most sediments change even by a few orders of magnitudes with overburden pressures and with long-term cementation, and those have not been taken into account in most basis analyses. Thus Tanikawa et al. (2003) performed one-dimensional basin analysis for the focal area of the 1999 Taiwan Chi-Chi earthquake, using measured permeability and storage capacity and geological data for sedimentation rate, with good agreement with measured pore-pressure (Pp) distribution (Figure 2). The analysis predicts very nicely the development of abnormal Pp at a depth of 5 to 6 km (present depth of around 4 km taking into account the eroded sediments); calculated curve is not a curve-fitting, but is a prediction based on measured properties. Prediction assuming constant permeability disagrees completely with the Pp data (not shown here).
Physically, fluids escape to the surface from the top sediments and thus sediments compact from top as fluids move out resulting in reduction in porosity and permeability. Long-term cementation enhances this process. In the focal area of Taiwan, permeability of shale and siltstone decreases down to the order of 10-18 m2 at which fluids cannot move much even in the million- year time scale. We interpret that below this depth, excess pore fluids could not escape to form abnormal Pp zone. It is quite interesting that gas and oil are accumulated in this abnormal Pp zone, so gas and oil could not escape either. Basis analysis with three phase flow might be able to predict the accumulation of hydrocarbons as well. Another important finding is that the simple method in the last section is applicable to hydrostatic Pp zone above the abnormal Pp zone although data comparisons are not shown here (see Tanikawa et al., 2003). Hence the simple method can be applicable to most cases where abnormal Pp is not developed.
4. Long-term cementation estimated for the Miyazaki Group
H. Kitajima performed similar measurements for the Miyazaki Group (see her poster at this symposium). Her data shows porosity and permeability distribution beneath Miyazaki where methane-rich water and hot spring water have been pumped out. Her measurements yield data of practical use in geo-engineering applications. Her prediction is based on measurements on surface samples and will be tested soon against measurements on drill cores.
An important outcome from such a study is the estimation of long-term compaction. Figure 3 shows porosity values for sandstones and siltstones with different ages at effective pressures of 20 and 100 MPa. Microstructures of sandstones exhibit abundant evidence of solution/precipitation processes, so that reduction in porosity is likely due to long-term cementation. Although cementation by such processes cannot be predicted accurately at present, theoretical prediction can be tested against such data in the future. At least, cementation processes can be traced by measuring porosity at pressures for samples from all stratigraphic horizons.
5. Effects offaults and fractures on fluid flow in porous sediments
Fluid flow in fault zones has received much attention recently in relation to fault-seal potential to hydrocarbons (see Takahashi, 2003; and references quoted therein). If one can predict whether a fault can seal hydrocarbons without drilling, enormous amount of money can be saved in petroleum exploration. Recent data on fault-zone permeability in porous sediments have been revealing that a fault zone is more impermeable than host rock by more than one order of magnitude because clastic grains in porous sandstones are comminuted to make a fine-grained fault zone with low permeability. Permeability drop can be by more than two orders of magnitudes when shale is trapped into a fault zone. Thus fault zones in porous rocks do not necessarily act as fluid conduit. We have not determined as yet at what stages of diagenesis do fault zones in sedimentary rocks change their characters to those in basement rocks which have permeable zone consisting of fault breccia and fractured host rocks.
Fluid flow along fractures can be even more important than the flow in fault zones, but it still remains unexplored particularly for fractures in porous sediments. Although still preliminary, Komizo and Shimamoto (this symposium, Fig. 5 in their paper) have measured permeability of fractured and intact specimens during pressure cycling tests using siltstone from the Miura Group. They found that fractures have almost no effects on permeability when permeability of host rock is greater than 10-16 to 10-17 m2, but afractured rock has permeability larger than that of host rock by a few orders of magnitude when host-rock permeability is less than about 10-17 m2.
Another interesting finding is Kitajima’s observation that joints in Miyazaki Group begins to develop only when porosity reduces down to 15 to 20 % (Kitajima et al., this symposium). This was rather surprising since we have had impression that joints form in almost any rocks and sediments. We tried to collect samples with joints for laboratory measurements of fracture permeability, but could not find them even upon close search in the field. So pore spaces between clastic grains must have hindered crack propagation. This observation and her preliminary experiments suggest that fluid flow along fractures is not important for porous rocks.
We plan to perform more detailed work on fluid flow along faults and fractures not only to understand fluid flow in the accretionary prism, but also to obtain basic data relevant to waste isolation and the Earth’s environmental issues.
5. Suggestions for future studies
Techniques we have developed and outlined above can be applied to accretionary prism, and indeed we plan to measure transport properties of ODP cores from Nankai Trough from this September as a US-Japan cooperative project. This work will be our preparation for deeper drilling into the seismogenic zone with Chikyu. In addition to the measurements of transport properties, basin analysis in Taiwan needs to be extended to at least 2D analysis incorporating the effects of deformation (folding, faulting and fracturing). One surprising thing from the basin analysis in Taiwan is that geological structures in the focal area (i.e., fold-and-thrust belt) are not totally different from those in accretionary prisms. Thus good correlation of 1D result in Figure 2 with observed abnormal pore pressure suggests that fluids (water, gas and oil) have not leaked much along faults and fractures. But this needs to be shown by more detailed work.
So far we have been concerned only with transport properties and fluid flow in sedimentary basins. However, methane and methane hydrates aroused our interest in material transport in fluids. For instance, water in shallow, very permeable formations in the Miyazaki Group contain lots of methane. But how can the methane be maintained? We look forward to gain some insight on mechanisms of gas migration from this “gas hydrates”symposium.
References
Bethke, C. M. and Corbet, T. F., Linear and nonlinear solutions for one-dimensional compaction flow in sedimentary basins. Water Resources Res., 24, 461-467, 1988.
Caine, J. S., Evans, J. P. and Forster, C. B., Fault zone architecture and permeability structure. Geology,24, 1025-1028, 1996.
Chilinger, G. V., Main, R. and Sinnokrot, A., Relationship between porosity, permeability, and surface areas of sediments, Jour. Sediment. Petrol.,33, 759-765, 1963.
Evans, J. P., Forster, C. B. and Goddard, J. V., Permeability of fault-related rocks, and implications for hydraulic structure of fault zones. J. Struct. Geol., 19, 1393-1404, 1997.
Gibson, R. D., The progress of consolidation in a clay layer increasing in thickness with time. Geotechnique, 8, 171-182, 1958.
Soh, W., Nakayama, K. and Kimura, T., Arc-arc collision in the Izu collision zone, central Japan, deduced from the Ashigara Basin and adjacent Tanzawa Mountains, Island Arc, 7, 330-341, 1998.
Suppe, J. and Wittke, J. H., Abnormal pore-fluid pressure in relation to stratigraphy and structure in the active fold-and-thrust belt of northwestern Taiwan. Petroleum Geology of Taiwan, 14, 11-24, 1977.
Takahashi, M., Continuity and transport properties of shale smear in sedimental basin (in Japanese with English abstract), Jour. Geography, 112, 926-938, 2003.
Tanikawa, W., Shimamoto, T., Wey, W.-K., Wu, W.-Y., Lin, C.-W. and Lai, W.-C., Permeability structures and prediction of the development of anomalous fluid pressure at depths in the focal area of 1999 Chi-Chi earthquake (in Japanese with English abstract), Jour. Geography, 112, 939-952, 2003.