Abstract

This paper described a case study involved an investigation in a field in Libya, where massive unexplained fill had been reported accompanying obstruction of production for majority of production wells since the onset of production, indicating possible sanding issues for this field.

To investigate this problem, relevant data from different sources and different domains (i.e., wireline logs, laboratory test data, drilling data, well data and field data) were integrated to generate a Mechanical Earth Model (MEM). This model provided the descriptions of the rock strengths and in-situ stresses in the reservoir formation. Somewhat surprisingly, the model, backed up by the core laboratory test data, observations from core inspection and thin section analyses, revealed the rocks to be extremely hard and strong, and therefore highly unlikely to sand. These findings contradicted with initial impression and previous expectation on this sandstone that it should have been sand-prone formation. Facing these apparent inconsistencies, the investigation moved beyond an initial focus of sanding risk evaluation and sandface completion optimization. The final results revealed that the problems facing the field were other than conventional sanding and formation failure, and that they involved some rather interesting and misleading phenomena, such as precipitation of salt from production, tubing scale, spalling of borehole wall with drawdown and cavings bridge (cavings might fall into and become wedged in the openhole, forming a bridge with no material beneath).

The investigation concluded that installing sand control facilities were unnecessary, which otherwise would have cost millions of dollars without correctly addressing the real problem that this field was facing. The study highlighted the importance of a thorough investigation of the mechanism and source of sanding rather than premature conclusions based initial, and potentially misleading evidence. It also highlighted how the integration of information from different sources and disciplines were able to correctly identify and address a particular borehole fill problem, allowing for optimizing field operations, field management and workover strategies.

Introduction

The consequence of sanding is well known: obstructing production, eroding downhole and surface facilities and demanding additional process and exposal cost. It will be quite straightforward to detect and identify sanding problem if massive sand is produced to surface. However, situation is quite different if it is not the case.

Since it first went on to production in the year 1969, the field has been reported to have been experiencing sanding problems. For example, the openhole completed well AA04 was reported to have fill up to 1094 ft from TD in September 1971 in a period of 18 months. Sanding in the cased and perforated well AA05 had been reported that since 1970 it got fill of 882 ft in four months and up to 906 ft in ten months. To date, no sand control or sand management techniques have been implemented on a field wide basis for the field. Due to fill in the borehole, the production for this field is obstructed to a lower level, which lead to significant economical loss.

A geomechanics and sanding study was initiated in the year 2004 to investigate sanding mechanism and to provide the contingency plan to properly address the problem and maximize economic production.

The study started with data acquisition. Various wireline logs such as microresistivity image log and dipole sonic log were acquired from some of the key study wells. Geomechanical laboratory tests were carried out on core plugs to determine rock mechanical properties, such as elastic properties and unconfined compressive strength (UCS).

With all required data available, a MEM1,2 was constructed to provide a comprehensive numerical description of rock mechanical properties. This model, backed up by the core laboratory test data, observation from core inspection, and thin section analyses, revealed the rock to be extremely hard and strong, and therefore highly unlikely to sand. This was confirmed in a more detailed analyses of the data contained within the MEMs that involved computing stresses around the boreholes and perforation tunnels during production and depletion, and evaluating the potential for rock failure (i.e., sanding risk during production and field depletion).

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