Oil Quality Prediction in Deep Waters: Gulf of Guinea Applications

Jacques Bickert, Denis Levaché, Gilles Nicolas, and Gilles Sermondadaz
TOTAL Exploration & Production, CSTJF Avenue Larribau 64000 PAU France

In E&P processes, fluid quality is essential for reserves evaluations, well productivities and economics.

Fluid quality prediction is a key parameter to rank, select and launch Exploration/Appraisal projects. This topic is critical in Deep offshore environment where viscosities and productivities strongly impact the economics. Moreover, reliable bulk and associated PVT properties are needed to populate reservoir models in appraisal phases.

Several methods can be used to estimate the fluids nature and their qualities before drilling. In all cases, it is highly recommended to:

be pragmatic
favour direct methods based on suitable data base
think “Geology” before any fluid prediction work.

In deep offshore environments, Biodegradation can be considered as the major alteration phenomenon. It is generally accepted that it may alter fluids in the reservoir if the temperature is lower than 80°C.

Currently, the different phenomena which lead, by alteration, to a specific fluid composition, are individually known but their modelling is still difficult and their relative contributions are also questionable.

Basic questions are still disputed (Biodegradation kinetics compared to reservoir charging, Impact of secondary gas produced by biodegradation etc.) To solve these problems, several approaches have been tested to better predict oil quality (API, GOR and specially viscosity). They can be divided into several types:

Prediction Method using Analogues
Rules of thumb coming from data bases
Prediction by use of the temperature impact
Prediction method using Modelling
Prediction method using Data Base and Statistical tools

Because multiple parameters (T°C history, Oil quality, Geometry, Salinity, etc.) can impact oil quality, it is obvious that a single one dimension approach is inadequate. So, two solutions can be proposed to improve the accuracy of the oil quality predictions:

A statistical approach based on a neuronal network method:
A combined approach based on:

An adequate data base (corrected T°C, decontaminated bulk properties, one Generative System .., same alteration phenomenon)
A preliminary analysis and identification of the phenomena/parameters which mainly impact the oil quality in the geologically defined context
A statistical approach where the geoscientist tries to define a function (transfer function) which enables the determination of the researched property (API, GOR etc…) with selected parameters which can be estimated before drilling

This last method is used in Total in the Gulf of Guinea for tertiary reservoired fluids in Exploration and also in Appraisal contexts to estimate the PVT properties of undrilled sedimentary bodies.

Several examples will be presented to illustrate the Total approach.

Unnumbered Figure. Many parameters can control the intensity of the Biodegradation phenomenon in reservoir.