Analysis of Surface and Wellbore Hydraulics Provides Key to Efficient Blowout Control

Abstract Assessment of the blowout rate and composition is a key issue for decisions on the blowout control method, design of the kill operations, determination of direct (e.g. environmental) and indirect (loss of reserves) damage etc. Usually loss of control over the outflow of the well implies that the blowout rate can not be established with conventional means. Experience shows that, due to the large reservoir decompression, production well performance can not be extrapolated reliably to blowout conditions. Hence (simple) surface observations, such as wellhead/drill pipe pressure and temperature, effluent shape and size, noise field, results of previous kill attempts and pressure response of nearby wells, have to be interpreted in terms of the blowout rate. This paper gives two examples of blowouts where the combination of such data led to a consistent and accurate picture of blowout well performance and describes how such estimates contributed to decisions on the kill strategy, leading to success over relatively short periods of time. Introduction With the advent of modern drilling equipment such as MWD and kick detection systems, the frequency of blowouts, i.e. complete loss of control over the outflow of wells, tends to decrease. Nevertheless, unfortunate combinations of equipment failure, human error, geological uncertainty etc. still give rise regularly to incidents which may lead to loss of wells, equipment and even human life. In order to keep these risks to a minimum, decisions on the appropriate strategy to regain control over a blowing well have to be taken rapidly. This implies that the information on which these decisions are based is usually limited to the drilling and production history, experience with offset wells in the area and simple observations made prior to and in the early stages of the blowout. It will be demonstrated later on that extrapolation of production well performance gives unreliable results under blowout conditions. Hence field observations must be interpreted correctly to define an optimal control strategy. Examples of such observations are:–Pit gain and pressure build-up following a kick.–Size and shape of surface outflow, e.g. plumes, vents, craters.–Type of outflow, gas, oil, water, steam, solids–Wellhead pressure (temperature).–Wellhead noise. For an ignited well, which can not be approached, some of these parameters can not be determined. In this case the temperature field around the well may give a clue on well performance. In a later stage, the selected control strategy can be refined on the basis of additional data, including:–PLT logs.–Pressure response of adjacent wells.–Results of early kill attempts. For a number of these observations, methods to interprete quantitatively and translate to a kill strategy have been discussed (refs. 1-8). In this paper it will be demonstrated that by combination of the various methods a consistent picture of blowout well performance over time can be obtained even though direct measurement is impossible under the circumstances. For this purpose, two case histories will be discussed; a steam injector blowout and a high GOR oil well blowout. It will be shown that in both cases, owing to a good understanding of well performance, the methods selected to regain control over the well were successful over a relatively short period. Case I: Steam injector blowout The incident occurred when this 800 m deep steam injector was suspended temporarily during a workover to repair a leaking tubing. P. 711