Diagenesis Driven by Geothermal Convection in Lacustrine Carbonates: Developing Reservoir Quality Predictive Concepts with Reactive Transport Models
Gareth D. Jones and Yitian Xiao
ExxonMobil Exploration Company, Houston, TX, USA
World class deepwater hydrocarbon discoveries in the pre-salt basins of Brazil exemplified by the Lula field in the Santos Basin, and exploration potential in the conjugate African basins have created a business imperative to predict carbonate reservoir presence and quality in similar tectonostratigraphic settings. What are the key geological processes that resulted in the generation and preservation of reservoir quality in these deeply buried, often greater than 5 km, pre-salt carbonates? In contrast to their marine counterparts relatively little is known about carbonate facies deposited in non-marine environments. Lacustrine carbonates, including microbialites, can have a diverse range of primary porosity and pore types but analog studies and sparse subsurface data suggest that diagenesis is often a significant control on reservoir quality in addition to depositional texture.
Tectonic and climate influenced variations in hinterland drainage, lake level, lake basin configuration and lake hydrochemical evolution impact near surface early diagenesis in lacustrine environments. But to what extent can collective early diagenetic processes generate permeability, and its preservation, to explain prolific extended well test flow rates publicly reported from the Santos Basin? Indeed, observations from modern and ancient lacustrine carbonates often demonstrate that early diagenesis results in a dramatic net reduction in primary porosity and permeability.
In contrast, the potential to enhance reservoir quality by dissolution in the burial environment is poorly understood. Geothermal convection is a style of groundwater flow, driven by temperature induced variations in fluid density, capable of diagenetically modifying reservoir quality. Based on a Santos Basin half-graben conceptual model, we used Basin2 a two dimensional numerical basin model with the capability to simulate variable density fluid flow coupled with thermodynamic reactions (reactive transport) necessary to evaluate the potential for diagenesis driven by convection. Our stratigraphic framework includes a volcanic basement, a rift section composed of volcanicalstic sediments and an overlying “Sag” section of lacustrine carbonates vertically partitioned between sublittoral microbialites and coquinas that transition to proximal profundal mudstones and argillaceous limestone down stratigraphic dip. Extensional faults variably penetrate the stratigraphy up to the base of salt. Petrophysical rock properties including porosity and permeability are specified as a function of depositional rock type using relationships constrained by both modern lacustrine carbonate sediments and present day pre-salt carbonate reservoirs. We simulated lacustrine carbonates deposited on an intra-lake fault controlled high isolated from a rift hinterland. Specifically we tested the effect of variations in basal heat flux, lake temperature including depth and magnitude of the thermocline, permeability, faults, stratigraphic architecture (depositional facies, thickness and geometry), salt thickness, salt rugosity, salt minerals and pore fluid composition.
Diagenetic rates are critically controlled by temperature gradient and fluid flux following the principles of retrograde solubility. Simulations predict that the temperature contrast between cool sub thermocline lake water and warmer near surface waters drive convective groundwater flow, at rates of up to 0.15 m/yr, prior to salt deposition. Temperature gradients along flow paths are weak because the principal flow direction is concordant with stratigraphy and sub parallels isotherms. Consequently, associated thermodynamically controlled rates of dissolution in the reservoir prone interval are generally low being on the order of 10-2 to 10-4 volume % / M.y. and thus insignificant with respect to reservoir diagenesis. Permeable vertical extensional fault zones are the exception. Convective driven ascending flow in faults amplifies dissolution rates up to 8 volume % / M.y in response to faster flow and steeper thermal gradients. Conversely descending convective driven flow focused in faults results in enhanced local cementation. Such early, pre-salt burial modification of fault zone permeability would strongly influence the magnitude and extent of any later dissolution resulting from burial fluids ascending via faults.
Post-burial simulations demonstrate the critical role of salt rugosity and the presence of salt withdrawal basins. Contrasts in thermal conductivity between the salt and less conductive sediments that fill the withdrawal basin create local subsurface temperature and fluid density anomalies capable of driving flow. The greatest potential for dissolution at rates of 0.4 – 1.2 volume % / M.y occurs where salt welds thin to less than 300 m. Dissolution is greatest around the edge of the withdrawal basin in the top of the reservoir. Cementation is focused along the base of the reservoir due to underlying volcaniclastics that are an aquitard to convective flow. With 10’s of M.y. available residence time, when the above conditions are satisfied, convection could locally result in porosity changes of 1-10 % and potentially an order of magnitude in reservoir permeability. In contrast, rates of diagenesis are too low, even with a long residence time, to impact reservoir quality in regions where salt withdrawal basins are absent or poorly developed.
By integrating our model driven predictive diagenetic concepts for geothermal convection and other diagenetic processes with traditional subsurface datasets and applicable outcrop and modern geological process analogs we are able to further refine our exploration to production scale understanding and hypotheses of lacustrine reservoir presence and quality in the South Atlantic basins and elsewhere.
AAPG Search and Discovery Article #120034©2012 AAPG Hedberg Conference Fundamental Controls on Flow in Carbonates, Saint-Cyr Sur Mer, Provence, France, July 8-13, 2012