The increasing discovery success rate of Carnarvon Basin gas fields holding significant exploitive gas and condensate reserves has not only made laminated sandshale sequences or "thin beds" a primary exploration target but has also necessitated the need for a standardised evaluation approach. This paper describes the workflow and methodologies developed for evaluating laminated shaly sand sequences in this region. The thin bed workflow results are compared to a number of independent net sand and net pore volume approaches using tools such as tri-axial resistivity anisotropy, nuclear magnetic resonance (NMR), high resolution borehole images as well as core net sand count results. The applicability, validation and limitations of each method are discussed. The reservoir quality of the Carnarvon Basin laminated sequences varies from discrete multidarcy deltaic quartz arenite deposits up to several meters in thickness down to a tight argillacous silty sandstone complex often cemented by carbonate, siderite and pyrite whose permeability is less than a millidarcy. The main limitation of the traditional deterministic approach is the limited vertical resolution of the log measurements that impacts the estimation of net sand and net pay pore-volume on the downside. This paper describes how this is overcome by the development of a consistent workflow for the evaluation of stacked shaly sand laminated formations (in particular the Upper Triassic Lower Brigadier and Mungaroo Deltaics formations). The basic methodology, value of information and pros and cons of each technique are documented and compared. This is followed by a discussion on the often overlooked question of whether or not a thin bed evaluation approach needs to be applied in the first place. The utilisation of a variety of log indicators as well as geological field knowledge including Facies Associations and Lithofacies descriptions that are used to answer this important question are described in detail.

You can access this article if you purchase or spend a download.