Abstract

Techniques based upon proportion curves and event analysis can greatly aid rapid detection of possible high permeability conduits and/or barriers in reservoirs. The techniques can use both dynamic data such as production rates and monitoring logs as well as static data such as open hole well logs. A recently created interactive PC software package allows effective application of these techniques. An example is presented of how the techniques greatly accelerated the history match of a heterogeneous carbonate reservoir by improving the calibration of open hole permeability log predictors by dynamic data, and allowing the conduit and barrier structure to be largely understood before full field history matching stage.

Suggestions are made on when and how to generally apply these techniques. It will be suggested that such techniques can supply inputs for distribution methods, such as Indicator Kriging and Truncated Gaussian Simulation, and thereby significantly enhance them. Other unconventional, yet useful applications of proportion curves will be mentioned, such as monitoring the changes being made in large flow simulation models during history matching, and determining areas where further data acquisition might be useful.

Introduction

Most would agree that when performing either quick scoping studies to identify candidate wells for work over and stimulation, or more detailed reservoir characterization studies, that the analysis should respect multi-disciplinary data to the level required by the study, and yet ideally, be rapid to perform. Several analysis techniques already exist, such as the Moving Domain Analysis (1), which are designed for rapid analysis of production data. But a drawback of these techniques is that they do not usually faciliate integration of multi-disciplinary data. For example, application of the Moving Domain technique requires only the production data and areal coordinates of the well heads On the other hand, the typical workflow followed in major reservoir studies, which involve history matching usually does ensure strong multi-disciplinary integration, but is generally time consuming to follow, and so should be made more efficient if at all possible.

The techniques to be described here are different from previous ones in that they facilitate analysis of large amounts of dynamic data, such as that from well production histories, production logs and cased hole neutron logs, as well as static data, such as that from open hole logs, to rapidly gain useful knowledge about reservoir flow architecture. To allow them to be efficiently exploited they have been implemented into a userfriendly, flexible PC package which is described later in the paper. The techniques do not rely on solving any flow transport equations (and so do not directly involve flow simulation or pressure transient analysis). This makes these new techniques simple to apply. They involve proportion using curves and water-conduit tracking.

The discussion in this paper will be confined to oil reservoirs. However, as will be evident, it should be simple to adapt these techniques to gas reservoirs. The major emphasis will be on carbonate reservoirs, since these often possess high permeability heterogeneities which are difficult to capture by normal means. A vertical proportion curve (VPC) is a histogram which describes the frequency of occurrence of a categorical variable at every stratigraphic level within a given region of the reservoir (eg see figures 2 and 3). Generally, the categorical variables are determined at the well bores. The water-conduit tracking (WCT) technique involves creating a network of possible high permeability conduits from analysis of start times of water injection (if relevant), and water breakthrough times in historical well-production data.

This content is only available via PDF.
You can access this article if you purchase or spend a download.