Testing Cement Static Tensile Behavior Under Downhole Conditions

Abstract

Abstract Today's analytical models can calculate stresses imposed uponthe annular cement sheaths of wells, and it has becomeapparent that induced stresses can often be tensile in nature. These models often predict the cement failing in tension aswell. This understanding has caused the industry to stepbeyond API tests that only recognize methods for testingcement under compressional loads. Instead, the determinationof cement tensile mechanical parameters should be consideredas important as cement compressive strength. Unfortunately, with no API guidelines, most oilfield cement tensile behaviortesting typically uses ASTM construction concrete testmethodology. With oilfield cements, most ASTM tests sufferfrom various shortcomings. These can be traced to the fact thatthey are designed primarily to test construction cements thatare not usually placed more than a few meters underground. These tests do not incorporate procedures to replicate thecuring of the cement in downhole environments, and do notconduct the actual static tensile testing under conditionssimilar to those occurring in oil and gas wells. When oilfieldcements are prepared and cured using API procedure in HTHPcuring chambers, they must still be subjected to significantinduced stresses as cooling and de-pressurization back toambient conditions occur before they can be subjected toASTM tests. Samples prepared and tested in such a mannermay undergo sufficiently induced stress to exhibit physicalsigns of initial mechanical failure, prior to even being placedin ASTM testing fixtures. To produce cement tensile behavior data that is morereflective of actual downhole performance, the authorsdeveloped an automated, microprocessor controlled testingdevice that cures and mechanically tests the cement undersimulated downhole conditions. Once slurry is placed in thetesting device, and the temperature and pressure is ramped upto simulate downhole curing conditions, the samples never seeambient conditions again until the conclusion of testing. Themicroprocessor controls and automated data acquisition unitalso allow for the determination of tensile stress-strainrelationships, prior to testing the sample to ultimate(mechanical) failure in tension. In this work, the authors present detailed descriptions of theoperational capabilities of the new testing device, along withdata developed with the device. Using the tensile datadeveloped under more realistic downhole conditions, inducedstress models can now generate even more accuratepredictions about fit-for-purpose cement designs for wells ofall depths and applications. Introduction When new technology is introduced to an industry, it oftenrequires supporting data that is completely different from thedata previously required for older technology. Such is the casewith the ability to calculate induced stress on the cementedannulus of oil and gas wells1. Most of the newer induced stressmodels currently in use show that normal well operations cancause the cement sheath in the annulus of a well to be placedunder various compressive and tensile stresses. In order toverify if a subject cement system is capable of withstandingsuch induced compressive stresses, those induced stresses canbe compared to the known compressive strength of the subjectcement system. The testing methodology for the determinationof cement compressive strength is well documented by theAmerican Petroleum Institute (API), and such tests areroutinely carried out every day in oilfield cementing labsaround the world. However, such is not the case when itcomes to the determination of cement tensile strength tocompare to induced tensile stresses. Different researchers haveused different methods to determine tensile strength. The APIhas not yet developed standards for testing the tensile strengthof oilfield cement and this has created two major issues thatthe authors of this work felt the need to address.