Reservoir Heterogeneity and Characterization in Deltaic Depositional Systems- Outcrop Analogs for Heavy Oil and Oil Sand Developments

Grant D. Wach¹ and Hasley Vincent^1,2
¹Dalhousie University, Halifax, Nova Scotia, Canada
²Ministry of Energy and Energy Industries of Trinidad and Tobago

Introduction

Permeability baffles and barriers create reservoir heterogeneities that can result in significant bypassed hydrocarbons if the geometry and architecture of the channel bodies are incorrectly identified and not correlated within a rigorous sequence stratigraphic framework. 3-D seismic data improves our understanding of the breadth of channel systems developed within these mixed deltaic settings, but cannot resolve the internal heterogeneity of the reservoirs formed with these systems. We need analog models from outcrops and field developments with similar depositional characteristics to input channel reservoir heterogeneity, model fluid flow and predict reservoir performance.

The numerous exposures of the Pliocene Morne L’Enfer Formation of Trinidad (Figures 1a and 1b) allowed for its detail study. The formation represents a late stage of deltaic, shallow water sedimentation within the Southern Basin of Trinidad (Barr et al., 1958; Bower 1968) and overlies other deltaic sediments of the Forest and Cruse formations (Table 1). Several heavy oil and oil sand occurrences are known at surface in the Morne L'Enfer Formation. These are not processed or upgraded for their oil but some are now quarried for road material. The exposures within one of these excavations, the Stollmeyer oil sand quarry, afforded an opportunity to study the distribution of heavy oil within reservoir and non-reservoir intervals. This bed-scale knowledge, combined with a regional correlation of Morne L'Enfer parasequences, affords an ideal analog for similar subsurface heavy oil reservoirs (Wach et al., 2004).

Stratigraphy of the Morne L'Enfer Formation

The Morne L'Enfer Formation consists of five members (Table 1): Upper Forest Clay, Morne L'Enfer Silt, Lower Morne L'Enfer Sandstone, Lot 7 Silt and Upper Morne L'Enfer Sandstone members. The “lower” Morne L'Enfer is further subdivided into at least three progradational parasequences, bounded by the lower Morne L'Enfer Sequence Boundary with overlying lowstand-fluvial and transgressive-estuarine sands. The maximum flooding surface is represented by the Lot 7 Silt Member and this is overlain by at least five parasequences of the upper Morne L'Enfer Highstand Systems Tract.

Facies and Depositional Environments

Five facies are recognized in the quarry: delta plain mudstones, basal lenticular sands (fluvial channel); conglomeratic mud clasts (channel lag); amalgamated sigmoidal sands (tidal channel fill) and slumped silts (abandonment plug) (Figure 2). The oil sands are quarried from the amalgamated, sigmoidal cross bedded sands of the transgressive systems tract. The sands comprise approximately 25 metres of amalgamated, sigmoidal and trough cross-stratified sands. This sand interval is targeted because of the relatively even distribution of oil saturation and high net to gross. The underlying mud clast conglomerate and lenticular sands form the economic base of the quarry on account of their lower sand/ shale ratio.

Oil Compartmentalization

Reservoir compartmentalization occurs at several scales and is controlled by lithofacies and faulting across the quarry. From a lithofacies perspective, there is uniform distribution within the lenticular and amalgamated sands. Local baffles to flow include mud drapes on sigmoidal toesets and ‘spillage’ zones adjacent to bedding planes caused by relative bed impermeability. The high proportion of silt and clay within the lenticular sands do not appear to be effective barriers to fluid flow. This contrasts with the slumped silts which were impervious to oil flow. There is also evidence for both sealing and transmissive faults within the amalgamated sands.

Ongoing Research

We estimate the field size represented by the Stollmeyer quarry as equivalent to 800,000 bbls OOIP. Channel and bedform dimensions, lithofacies, net to gross, and reservoir heterogeneity data have been collected for input into a reservoir model being developed using Petrel and Eclipse software. Detailed mapping of the quarry exposures continues as finer scale permeability barriers and baffles are defined and correlated using high resolution digital photography, outcrop gamma logs, LiDAR and Ground Penetrating Radar. Initial reservoir modeling is controlled by the stratigraphy and architecture elements of the systems, followed by fault input that further compartmentalizes the reservoir. Additional quarries, plus well-exposed coastal outcrops provide considerably more detail to the preliminary results presented by Wach et al. (2004).

Conclusions

Accurate analysis of seismic, well log, core and paleontological data reduces risk of both planned and ongoing of oil sand and heavy oil developments. More effective reservoir development and depletion strategies can reduce the risk associated with field development, decrease the number of wells required (Wach et al., 2004); and provide more accurate planning for mining and extraction schemes. Incorporating outcrops of analogous depositional systems minimizes risk associated with reservoir modelling. This provides a greater degree of confidence to assumptions that go into reservoir models and helps to establish the size and distribution of the reservoir and the internal architecture and geometry that is so important in determining and controlling fluid flow.

References

Barr, K.W., Waite, S.T., and Wilson, C.C., 1958, The mode of oil occurrence in the Miocene of southern Trinidad, B.W.I; in Weeks, L. G., (ed.), Habitat of oil: A Symposium of the American Association of Petroleum Geologists. p.533-549.

Bower, T.H., 1968, Geology of Texaco Forest Reserve field, Trinidad, W.I: Transactions of the Fourth Caribbean Geological Conference, Port of Spain, Trinidad, p. 75-86.

Wach, G.D., C.S. Lolley, D.S. Mims and C.A. Sellers, 2004, Well placement, cost reduction and increased production using reservoir models based on outcrop, core, well logs, seismic data and modern analogs- Onshore and Offshore Western Trinidad; In: “Integration of Outcrop and Modern Analogs in Reservoir Modeling”(G.M Grammer, P.M. Harris and G.P. Eberli, eds.) American Association of Petroleum Geologists, Memoir No. 80, 279-307.