[First Hit]

AAPG Annual Convention and Exhibition

Datapages, Inc.Print this page

Challenges and Opportunities in a Large Oil Field With Poor Quality Data: Integrated Modelling Solution for the Barrancas Field, Cuyana Basin, Argentina


In Argentina, a number of fields have declined significantly while applying historical development strategies. These fields typically have a long production history and a large number of wells, often with incomplete and inconsistent data sets. Redevelopment strategies based on integrated models are needed to extract remaining oil, overcoming contradictory electric log signatures, old lithostratigraphic correlation and poor understanding of dynamic behavior. The fault-bounded Barrancas anticline is one of the aforementioned fields, with more than 400 wells and over 60 years of oil production history. Early water-flood showed an excellent response in the field; yet production has strongly declined and significant oil remains in the subsurface (RF 26%). The primary objective of this study was to quantify remaining opportunities and obtain an optimized development plan, justifying further investment in this old field. Facies associations coming from core and outcrop analyses interpreted the Barrancas Formation as a North-South prograding alluvial-ephemeral system. Changes between progradational and retrogradational periods were used as chronostratigraphic correlation surfaces. Areal trends in reservoir quality, resulting from this stratigraphic model, were clearly reflected in uneven production distribution across the field, but were not as clearly shown in the outdated and contradictory electric logs responses. Log data was in fact found to be inconsistent, due to low resolution Spontaneous Potential profiles and volcanic-affected Gamma Ray logs. Since production history was the most consistent reservoir response, it was decided to use dynamic data upfront, as the main constrain for reservoir characterization and modelling. By doing so, different geological scenarios were tested and compared, letting dynamic data guide the appropriate hierarchy, porosity and connectivity of reservoir bodies. To aid integration and guarantee consistency between static and dynamic models, the geological model was deliberately built at a scale that could be simulated. Dynamic models built this way managed to overturn existing preconceptions about the field (reservoir quality distribution, distinct OWCs, aquifer impact) and to successfully predict the existence of unexploited flank oil and unswept central zones. Even when the field was supposed to be mature, the study proved existence of enough opportunities to support additional investments.