| Popis: |
ABSTRACT Fluid loss during cementing operations has been suspected for a long time to have detrimental consequences. Among them, one can cite the reduction of slurry volume, the increase of slurry density and viscosity, the deposition of a cement cake which may lead to extra friction pressure drops during placement or the invasion of the near-wellbore reservoir by the filtrate. In this paper, all these consequences are briefly reviewed and quantified. Given a well geometry, a job design and the fluid loss properties of the cement slurry, it is shown how to predict what are the most critical phenomena susceptible to impair the cementing job and which are the most sensitive well or slurry parameters. Coupling these results with the degree of accuracy with which the well parameters are known, one can calculate the uncertainty of the predictions (their error bar) and deduce which are the key parameters which affect this error bar. 1 INTRODUCTION There is an abundant literature on the effect of fluid loss on well cementing and many phenomena have been suspected to play a role but no critical comparison exists concerning their relative importance. Two reasons of this plentiful literature are that (1) fluid loss agents for cements are expensive chemicals and (2) several operators believe that the mud cake is sufficient to provide a good fluid loss control1,2 The first objective of this paper is to gather what has been written on the importance of fluid loss in primary cementing and to establish ways of comparing the various means by which fluid loss may affect a primary cementing operation. Practically, it will result in predictive tools used as an aid to indicate whether a given cement slurry, for a given cementing job, needs to have a good fluid loss control. Most of the time, the prediction of requirements for wellbore fluids is restricted by the limited knowledge of downhole behaviors. For example, it has often been said that cement slurry filtration is limited by the mud cake but, until recently,3 there have been no field measurement in support of this. Thus, any prediction should account for the rather poor accuracy of input data. We will indicate one way of doing this. In this paper, we shall deal exclusively with primary cementing and will not consider other practical situations such as squeeze or lost circulation. This paper is organized as follows. In the next section, the way filtration properties are quantified is briefly reviewed. Then, published consequences of fluid loss and their quantification are described. At last, we consider the effect of input data uncertainty on predicted fluid loss control requirements. 2 FILTRATION REGIMES AND CHARACTERIZATION Two limiting filtration regimes correspond to static (or pseudostatic) filtration and dynamic filtration. Cross-over regime exists4 but we shall not consider it. Generally, the surface cake controls the behavior and any effect of an inner cake (formation damage) is ignored. Consequently, the filtration pressure is just the difference between slurry pressure and formation pressure. |
