Abstract
Simulating the mechanical response of PDC drill bits contains a lot of uncertainties. Rock and fluid properties are generally poorly known, complex interactions occur downhole and physical models can hardly capture the full complexity of downhole phenomena. This paper presents a statistical approach that improves the reliability of the PDC bit design optimization process by ensuring that the expected directional behavior of the drill bit is robust over a well-defined range of drilling parameters.
It is first examined how uncertainty propagates through an accurate bit/rock interaction model which simulates numerically the interaction between a given PDC drill bit geometry and a given rock formation, both represented as 3D meshed surfaces. Series of simulations have been launched with simulation parameters defined as probability density functions. The focus has been set on directional drilling simulations where the drill bit is subjected to significant variations in contact loads on gage pads along its trajectory. A global sensitivity analysis has also been performed to identify the key parameters which control drilling performance.
Directional system parameters are critical in terms of steerability and tool face control, particularly in high dogleg severity applications. Based on these simulations, a statistical optimization strategy has then been implemented to ensure that the directional performance of the drill bit remains effective under a given uncertain drilling environment. Statistical analysis combined with drilling simulations indicated that ROP improvements could even be achieved without compromising steerability. A balanced bit design was selected and manufactured in an 8 1/2-in. model to drill a 714 ft section of a Kuwait field. The bit was run on a high dogleg rotary steerable system and directional assembly. The bit achieved the high steerability goals required by the application while showing a good compatibility with the directional tool. Moreover, ROP was increased by approximately 27% compared to offset wells, setting a record rate of penetration in the field.
Whereas statistical analyses are commonly conducted in the field of geosciences, it has rarely been applied in the field of drilling applications. The statistical bit design optimization strategy deployed in this work has allowed to improve both the drilling performance of the drill bit and its reliability.
