Drilling offshore oilwells is a very expensive and complex job, in which all efforts must be spent to keep the annular pressure between a minimum (pore pressure) and a maximum pressure (frac pressure). A lot has been invested in downhole sensors and data transmission systems but there is a common sense in the industry that very little is available in real time data interpretation. The correct interpretation of PWD (Pressure While Drilling) data is a very powerful tool to identify undesired phenomena. The anticipation and remediation of potential hole problems is an ultimate goal of most real time measurement devices installed on drilling rigs. The accurate quantification of the phenomena that impact bottom hole pressures is a must on anticipation operational problems. This paper presents a comprehensive methodology to accurately quantify some of the most common phenomena that occur during the drilling job. Transient modeling, pattern recognition and AI strategies were considered. The final goal is the development of a real time computational tool to interpret the operational parameters, identify and remediate problems, reducing risks, time and costs.
Examples of problems detection (such as cuttings excessive accumulation, kicks, wash outs, hydraulic packers, instability related issues, etc) will be addressed. Besides, the impact of remedial actions on downhole pressures and solids concentration profiles will be also detailed.
The anticipation and remediation of potential hole problems is an ultimate goal of most real time measurement devices installed on drilling rigs. Among the several sensors available, PWD (pressure while drilling) measurements gained popularity due to its potential for problem diagnosis (Aragão et al, Teixeira et al, Rommetveit et al, Green et al). The complete understanding of the physical phenomena governing downhole pressure is, however, far from being spread among the drilling teams at rigsites. Figure 1 shows a typical PWD log. These aspects motivated PETROBRAS to start the development of a computational tool to interpret PWD and mudlogging real time data while drilling. The tool (called PWDaTM - Pressure While Drilling Analyzer) should identify undesirable phenomena such as deficient hole cleaning, pressure increments due to gel breaking when circulation is resumed, downhole pressure variations due to drillstring movement (surge and swab), undesirable formation fluid influxes (kicks), etc and alert the operators. Nowadays, an expert monitors the operational parameters (bottomhole pressure and other surface sensors data) and tries to identify and/or anticipate problems. Thus, the identification of potential problems is a subjective process and can vary depending on the expert interpretation. The proposal of this work is to provide a tool to help the operators to take important decisions rapidly in an objective way, optimizing drilling job (reducing time and operational costs).
The tool should receive real time PWD (ECD, ESD, internal column pressure and temperature) and mudlogging (pump pressure, rate of penetration, flow rate, drillstring rotation, torque, drag, bit and hole depth, etc) data during drilling job and predict ECD (equivalent circulation density), pump pressure and solids concentration (Gerhard et al). PWD data and real pump pressure are compared to the predicted parameters. Differences between real and predicted ECD curves (along the time) as well between real and predicted pump pressure curves indicate some unexpected phenomenon is happening. The different tendencies of real and predicted curves are interpreted to identify potential problems. Once a problem is identified, the software proposes preventive and/or corrective actions to be taken.