Casing Wear During Drilling-Simulation, Prediction, and Control

Abstract

Summary. Four case studies show that laboratory simulations of casing wear caused by rotating tooljoint hardfacings correspond very well to field-measured casing wear. Moreover, the work reveals concentrations of casing wear at the kickoff point of a deviated well and near casing collars. Drillpipe/casing protectors do not always appear effective in avoiding casing wear. It is concluded that prediction of casing wear is possible with laboratory simulations, and that casing wear can be controlled by sufficiently smooth hardfacings, weighted mud, and moderate tooljoint/casing contact forces. An inspection technique can be used to qualify hardfacings for use in casing. Hardfacings that are machined to meet the smoothness requirement are commercially available. Introduction Internal wear of the casing during drilling and workovers is an ongoing concern, especially in deep and/or deviated wells. The reduced pressure integrity of the casing represents a safety hazard and can lead to well-control problems and even to blowouts. Additionally, the costs of repairing worn casing are always high. Occasionally, severe casing wear could even cause an almost-completed well to be abandoned. The tensioned drillstring is the main cause of casing wear, because it is pulled firmly against the casing in the dog legs of the well while rotating to drive the bit. The tooljoints of the drillstring are of particular interest because they are in contact with the casing and bear the side forces. Often the wear penetration into the casing is acceptable, but in some instances severely worn casing has been reported. This casing wear is caused mostly by new tungsten-carbide hardfacings - often of poor welding quality - on the drillpipe tooljoints. The problem arises from the initial surface roughness. The work presented here will highlight the prediction and the prevention of casing wear caused by rough hardfacings, which are known to be responsible for most of the wear problems. Prediction of casing wear must rely on laboratory simulations, because the roughness of the hardfacings is too complex to deal with on a mathematical basis. Case studies will show that field casing wear can be simulated in the laboratory to a sufficient degree of accuracy. Casing wear can be controlled under normal drilling conditions, leaving laboratory simulations for special circumstances. For control to occur, three requirements that specify the hardfacing roughness, the mud, and the tooljoint/casing contact force have to be met. Background Bradley and Fontenot published an extensive study on the prediction of casing wear. The investigation included casing wear caused by tooljoints, by drillpipe bodies, by drillpipe/casing protectors, and by wireline cables. They also differentiated between drillstring rotation and round trips. They concluded from laboratory tests that rotation of tooljoints is the most dominant factor in casing wear. The proposed prediction method was checked with field-measured wear and showed that the agreement was reasonable in one-half of the cases investigated. The work concentrated mainly on wear caused by smooth tooljoints. Many laboratory tests have simulated casing wear caused by hard-facings. Unfortunately, the test setups used were not validated with field-measured casing wear. The test results can be valuable, however, as a relative comparison of the different types of hard-facings. Refs. 1 through 5 recommend smooth hardfacing, but no quantification of the smoothness is given. Best and Bol discovered that a strong film of mud-weighting solids (e.g., barites) between the tooljoint and the casing reduces casing wear efficiently for smooth tooljoints. Simulation and Validation Four wells in which the worn casing-wall thickness has been measured are reviewed to validate laboratory simulations. The wear penetration into the casing has been measured either downhole with a casing caliper or at surface if the casing had been recovered. The accuracy of the wear measurements depends on the caliper, but is estimated to be better than 0.5 mm [0.02 in.]. The simulation in the laboratory is done on a full-scale test rig as described by Best. In the test facility, successive passages of a rotating tooljoint over a casing sample are simulated; this closely resembles field drilling conditions. The wear penetration into the casing sample is measured after every passage of the tooljoint and plotted to yield the wear curve. Each wear curve is valid for one combination of the following parameters: tooljoint/casing contact force, rotational speed, drilling penetration rate, mud, casing, and hardfaced tooljoint. During a test, all these parameters are close to the field conditions under which the wear occurred. An important parameter in the laboratory tests is the tooljoint/casing contact force, which varies in the field with depth. It is computed with the well survey data and the buoyed weight of the drillstring. The number of tooljoint passages to be simulated is based on entries in the daily drilling reports. The various rotational speeds and drilling penetration rates in the field are included by use of a linear relationship for the rotational speed and a reciprocal relationship for the drilling rate. Four Case Studies Four case studies will be discussed separately, field measurement and laboratory simulation will be compared, and every case study will highlight a special aspect of field casing wear. The test conditions, simulation results, and field measurements are summarized in Table 1. Case Study 1. A casing caliper recorded casing wear of 0.7 mm [0.03 in.] in one spot and 2.4 mm [0.09 in.] in another. The main reason for the wear was the pronounced doglegs and the high drillstring tension. The tooljoints contacted the casing with a lateral force of 10 and 15 kN [2.2 and 3.4 kip], respectively, for the two wear spots mentioned. The number of tooljoint passages through the particular casing was 187 (at a rotational speed of 220 rev/min and a drilling rate of 9 m/h [30 ft/hr]). The tooljoints were smooth. The wear curve obtained in the laboratory for a tooljoint/casing contact force of 15 kN [3.4 kip] results in 2.3 mm [0.09 in.] of wear after 187 tooljoint passages. This is very close to the caliper recording. The same good agreement is found for the wear curve at 10-kN [2.2-kip] contact force, where the simulated wear is 0.6 mm [0.024 in.] (the caliper recorded 0.7 mm [0.028 in.]). The measured wear curves (Fig. 1) show that the running-in wear (which is the high initial wear) can contribute significantly to the final wear. SPEDE P. 375^