Development and applications of a new system to analyze field data and compare rate of penetration (ROP) models

Abstract

Improvements in data acquisition technology have enhanced rate of penetration (ROP) modeling capabilities. Modern logging tools are able to record more complete drilling datasets at a higher frequency, allowing for better understanding of the many variables that affect the drilling process. ROP models published in literature simplify drilling rate formulations by combining complex drilling factors into model coefficients. The lithology dependence of ROP model coefficients, as well as the model's performance evaluated based on different types of rocks, is a topic explored throughout this project. A data analysis software developed in Microsoft Excel VBA, named ROPPlotter, provides ROP field data visualization and comparison of different ROP models. Userforms offer great flexibility in selecting different sections of the well and in highlighting lithology changes. The program accomplishes data filtering by detecting data outliers in the original dataset and excluding them for a more meaningful analysis. Then, VBA coding is applied in order to produce neat-looking plots automatically, overcoming Excel’s poor standard plot formatting. Excel Solver is employed in determining coefficients of six ROP models: Bingham (1964), Bourgoyne & Young (1974), Winters-Warren-Onyia Roller Bit (1987), Hareland Drag Bit (1994), Hareland Roller Bit (2010) and Motahhari PDC Bit (2010). By studying how these coefficients change with varying rock formations, valuable information about each model's behavior is obtained. Plots containing field data and ROP models, in addition to parsed data utilized in model calculations, can be saved for future analysis with the click of a button. ROPPlotter is useful in conducting case studies for industry, such as slow ROP in a section of the well or slide drilling. Furthermore, it provides a systematic way to assess ROP model performance and aims to quantify the lithology dependence of ROP models and their coefficients. This exercise provides a means of determining which ROP model works best for a specific field application. Later, by using an average value of model coefficients calculated for a certain field, optimal values of parameters controlled at the rig floor (weight-on-bit, rotary speed, flow rate) are determined for a future well to be drilled on the same pad.