The Luanda Chapter of the Society of Petrophysicists & Well Log Analysts (SPWLA), in April 15, 2009, organised a forum title “Addressing the Formation Evaluation Challenges of the Future — Deeper Water, Deeper Targets, and Cost Management”.

During the Forum, the Angolan Petroleum system issues were addressed.

Excerpts from one of the papers presented:

A Paper presented by Bambang Gumilar Stephen Van Daleri’, Sarah D. Springer of Chevron, Aberdeen, UK and Chevron, Houston, TX respectively.

Predicting reservoir rock permeability from well logs in uncored intervals or wells is an important yet difficult task in reservoir characterization. This is particularly the case with certain thin-bedded reservoirs where these reservoirs are at or below the resolution of the wireline logs. Under these circumstances, it has been a major challenge for study teams to properly apply existing permeability equations into the dynamic reservoir modelling workflows. This concern necessitated work to generate a new approach to overcome the problem of the cyclicity and magnitude of the reservoir permeabilities by generating new transforms that more closely resemble the thin bedded reservoir.

The new workflow as in most permeability estimation models for clastic reservoirs includes shaliness as its second variable. It used porosity as the main predictor as suggested by core data. Available core permeability shows a strong correlation with porosity log in a semi-log domain. The core permeability also indicates a moderate relationship with shaliness and geological facies. Using the combination of the relationships as a multi-regression equation should give better results than using them separately.

This study tailored an empirical approach to quantify the phi-k-shaliness relationship and to capture high resolution data into the relationship by utilizing the Enhanced Vertical Resolution, EVR processed neutron - density logs. A method of deconvolving the far-spaced detector response of a dual-spaced neutron logging tool, coupled with dynamic matching of the near-spaced detector response, improves the vertical resolution of the tool’s formation porosity measurements. Near/far ratio processing and the standard calibration procedure are retained to continue to provide accurate, borehole compensated formation porosity measurements. These logs (ERHO and ENPH) provide critical role in characterizing the thinly bedded sand packages.

The proposed workflow with regard to this subject has three major analysis steps excluding the data preparation work. Firstly, normalizing and rescaling an EVR neutron and density separation based porosity log, ENDIFF. This normalized curve represents both high frequency data points and shaliness which was to replace Gamma Ray derived Volume of Shale, VSH_GR which was impacted by the occurrence of dolomite/carbonate cemented streaks. Secondly, generating the e-facies flags. The Formation was divided into three main electrofacies named Shelf, Transgressive and “Nose” based on its distinctive ENDIFF and Resistivity curves characteristics. Lastly, is to assign the proper permeability equations to the respective e-facies.