AAPG Bulletin, V. 83 (1999), No. 2 (February 1999), P. 211-229.

Dimensions of Paralic Sandstone Bodies¹

A. D. Reynolds²
 

©Copyright 1999.  The American Association of Petroleum Geologists.  All Rights Reserved
 

¹Manuscript received August 11, 1997; revised manuscript received June 16, 1998; final acceptance July 14, 1998.
²BP Exploration, BP Plaza, Westlake Park Boulevard, Houston, Texas 77079; e-mail: reynoltd @bp.com

I am grateful to BP Exploration for permission to publish this paper. Thanks are also extended to AAPG reviewers Lesley W. Evans, J. R. J. Studlick, and G. J. Moir, and to numerous colleagues, in particular Mike Mayall and Mark Eller, whose comments significantly improved this paper. 

ABSTRACT

Petroleum reservoirs in paralic successions commonly comprise a wide range of sand-body types. To estimate the volume of petroleum that is present and the percentage that is recoverable, and to optimize development and production schemes, it is vital to know the dimensions and orientation of each sandstone body. Although the sedimentary facies and the orientation of paralic sand bodies have been extensively studied, there is little quantitative data on sand-body dimensions. This paper addresses that data gap and reports an extensive database of widths, lengths, and thicknesses of paralic sandstones.

Following an approach that has been successful in fluvial successions, an attempt is made to develop predictive relationships between sand-body thickness, which is commonly known from well data, and the key unknowns of width and sand-body length. In pursuing this goal, in addition to classifying the data by sand-body type, grain size, basin type, etc., a key aim of this study has been to test the hypo thesis that sand-body dimensions are controlled by their sequence-stratigraphic setting.

Crossplots of sand-body width against thickness show that discrimination of the data by sand-body type produces the tightest set of clusters; therefore, this should be the first step in choosing a realistic range of dimensions for a given sandstone thickness. Further analysis reveals that sequence stratigraphic setting is a useful means of refining the choice of analog and reducing the range of dimensions.

Specific conclusions are (1) Valleys are much wider than channels. Crude averages show that channels (distributary channels, crevasse channels, etc.) are narrow, less than 1 km in width, with aspect ratios of 1:100, whereas valleys average around 10 km in width and have aspect ratios on the order of 1:1000. (2) Shoreline-shelf sands are huge sheets tens to hundreds of kilometers in length, with mean widths that range from 7 to 25 km and vary according to systems tract. Shoreline-shelf sands deposited in highstand systems tracts are, on average, twice as wide as those deposited in transgressive systems tracts. Thicknesses of shoreline-shelf sands relate strongly to their position within a sequence set. (3) Valleys and shoreline-shelf sands have areal extents comparable to giant oil fields. Flood tidal deltas and mouth bars are comparable in area to small fields. (4) Distributary channels, crevasse channels, tidal flats, and crevasse splays are small areally. (5) Some systematic trends occur (e.g., in flood tidal deltas), allowing the prediction of width and length from thickness. Other data sets show a high degree of scatter, but maximum, mean, and minimum values can be determined for width, length, and thickness. (6) There is a clear partitioning of certain sand-body types into certain systems tracts.