Abstract:
Although the Amazon Fan is regarded as a typical "mud rich" submarine fan, systematic continuous coring of the various architectural elements of the fan at 17 drill sites during ODP Leg 155 in 1994 revealed much more extensive distribution of thick sands than conventional models predict. Prior to ODP drilling, high-resolution seismic, GLORIA side-scan, SeaBeam swath-mapping and piston-core data showed that the aggradational channel-levee system is the basic depositional unit of the upper and middle fan. High-amplitude reflections (termed HARs) beneath the channel axis were predicted to be laterally restricted, coarse-grained channel-fill deposits. In addition, more laterally extensive high-amplitude reflection packets (termed HARPs) at the bases of channel-levee systems were interpreted as coarse sediment deposited from turbidity currents and related flows issuing through a crevasse in a levee during an avulsion event (i.e. similar to crevasse splays) and then progressively building laterally outward and down slope from the channel mouth as the newly forming channel-levee system prograded progressively down fan. When the channel stops growing down fan, the terminal end of the HARP deposit on the lower fan becomes a lobe deposit. ODP Leg 155 cored >4 km of sediment from 17 drill sites on Amazon Fan which included channel-fills (HARs), HARP units from the bases of a variety of channel-levee systems, levee/overbank deposits, lower fan lobes and surficial and buried regional mass-transport deposits. Levee/overbank deposits are mainly muds with thin silt and fine sand laminae and beds. Sediments cored from channel axes (HARs) are predominantly thick-bedded, coarse facies. The most prevalent facies is disorganized structureless to chaotic, poorly sorted, fine-to-coarse sand; large mud clasts are common. The coarsest and thickest (10-18 m) sand beds were recovered from the thick (>100 m) HARP units and from lower-fan lobe deposits, which presumably are equivalent to HARP units up fan. Wire-line logs in combination with Formation MicroScannerTM (FMS) images and rare recovered pebbles suggest that some intervals of no core recovery in HARPs and lobes are thick sandy gravels. The cores show that HARPs and lobes predominantly contain medium-to-thick (10-20 m) beds of disorganized, structureless to chaotic, medium-to-coarse sands, commonly with mud clasts. These facies support the previous seismic interpretations for formation of HARP units. The Leg 155 cores show that despite its classification as a "mud-rich" submarine fan, many elements of the Amazon Fan (HARs, HARPs, lower-fan lobes) actually contain very thick, laterally extensive sand deposits, which potentially could form good reservoir sands. In particular, the recognition that HARP units are laterally extensive sand deposits, which are 10's to 100's of meters thick and >100 km in length, suggests that in addition to the well known sand-rich channel fills and lower-fan lobes, major additional, yet unrecognized, reservoir potential exists in HARP units of ancient "mud-rich" fans. These HARP-unit reservoirs would be enclosed in muddy levee/overbank and mass-transport deposits, which should provide good seals.