Abstract

An underground blowout in Norwegian North Sea exploration well 2/4-14 was successfully killed by pumping heavy mud into the well from a relief well. pumping heavy mud into the well from a relief well. As part of the kill planning, simulations by use of the dynamic two-phase pipe flow simulator OLGA were performed. The code, originally developed for two-phase hydrocarbon flow in pipelines and risers, was modified to handle injection of kill fluids and inflow from the reservoir. This paper presents the principles for the kill simulator, the results of the principles for the kill simulator, the results of the simulations and how these results were used in the design of the kill operation. Finally a comparison with observed data are presented.

Introduction

In January 1989 the shear-ram in Norwegian North Sea exploration well 2/4-14 was activated after a high pressure kick hit the surface equipment during coiled tubing operations in the drillpipe. A relief well was spudded shortly after to be used in case the well should start to blow out. After reentering the well three months later with snubbing equipment, it was concluded that there was an underground blowout in the well.

A simplified sketch of the configuration of the wells, after the 7" liner had been set in the relief well, is shown in Figure 1. The well was blowing oil and gas from the top of the reservoir formation lo cated at a depth of 4707 m (15433 ft) to a sandstone formation at approximately 830 m. The reservoir is highly overpressured with an initial static reservoir pressure estimated at 980 bar (14200 psi). pressure estimated at 980 bar (14200 psi). During a time consuming operation to fish coiled tubing out of the well, the expected blowout scenario and the proposed kill scenarios were modified several times as more data from the well became available. The main kill scenarios ranged from mechanical kill methods based on reentry into the blow out well with a specially designed casing packer -to dynamic kill methods either by injecting kill fluids internally into the blowout well or by injecting into the relief well.

The dynamic kill method by injecting into the relief well was finally chosen as the preferred kill method. This decision was mainly based on a high degree of confidence that it would be possible to achieve good downhole flow communication between the two wells and the results of the dynamic kill simulations which confirmed the feasibility of this kill scenario.

The theoretical and operational aspects of dynamic kill have been extensively covered by Blount and Soeiinah and others. They describe dynamic kill as a technique for terminating a blowout utilizing flowing fractional pressure to supplement the static pressure of the kill fluid being injected up the blowing well. For instance, water flowing up the blowing well at a given rate can cause the same total pressure difference between the reservoir level and the wellhead as a static mud column balancing the reservoir pressure. As the dynamic kill condition is an interim condition only, the usually light initial dynamic kill fluid will finally have to be replaced by a heavier fluid of sufficient density to bring the well under static control.

P. 315