Karst and Early Fracture Networks in Carbonates, Turks and Caicos Islands, British West Indies

2007 
Abstract Historically, studies of Quaternary carbonates have not adequately addressed the influence of early fracture networks on diagenesis. Because of this lack of detail, understanding and predicting fracture-related diagenetic heterogeneities and preferential fluid flow pathways in ancient carbonate successions is particularly challenging. The Pleistocene oolitic grainstones of the Caicos platform provide an excellent opportunity to evaluate the relative importance of fractures on early diagenetic alteration styles, and are a suitable analog for subsurface carbonate reservoirs. Detailed analyses of fractures (e.g., orientation, aperture, spacing, fill material) from Caicos outcrops combined with high-resolution, three-dimensional ground-penetrating radar (3D GPR), assisted in exploring the causality and distribution of fractures and relationship to karst. Four models were evaluated to explain the observed distribution of dolines: (1) gravitational, fractured-margin controlled, (2) tectonic-fracture controlled, (3) antecedent-topography controlled, and (4) a hybrid model. Based on observations of numerous fractures (n = 306) on the western Caicos platform, early fractures are abundant and dominantly margin-parallel. These fracture networks are well established in limestones prior to mineralogical stabilization, thereby indicating that diagenetic heterogeneities evolve very early in carbonate diagenesis. The spatial distribution of dolines on Providenciales is likely the result of a complex interplay between the antecedent topography, margin-parallel fracture systems, and meteoric fluids. Resultant diagenetic alteration is far more complicated than simple, unconfined, meteoric lenses associated with topographic highs. Any attempts to model early diagenesis in carbonates should not dismiss the role of fractures as diagenetic facilitators and diagenetic anisotropy templates.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    32
    References
    33
    Citations
    NaN
    KQI
    []