Over the years, the oil and gas industry has played a key role in the US energy sector. Oil companies have continually worked to develop technology to maximize performance and reduce the cost per barrel of crude produced, lately with the optimized performance of horizontal wells. Horizontal wells allow for more resources to be produced per well; however, they are more expensive since the horizontal section substantially increases length and drilling time. In drilling, now more than ever, optimization of the drilling parameters needs improvement to assist these companies in reducing the drilling time which can potentially save millions of dollars. Numerous individuals have analyzed theoretical rate of penetration (ROP) equations by finding the optimum value for constant drilling parameters for single bit runs, however, since the formation drilling variables change throughout the drilling process, so should the drilling controllable variables. Therefore, the idea of constant drilling parameters potentially results in wasted time and dollars for operators, which could be vastly improved through use of dynamic variables. This leads to a new research approach, one that attempts to optimize drilling parameters through the use of a swarm algorithm with the goal to reduce the drilling time through measurable improvement in rate of penetration, and therefore, reduced total drilling time. In the work presented herein a particle swarm optimization technique (PSO) has been incorporated to find the optimum combination of the drilling parameters, weight on bit (WOB) and revolutions per minute (RPM) for every foot of formation to be drilled. The integration of this new swarm algorithm into the optimization process allows for ROP equations to dynamically change, which can better adjust to the formation environmental variations throughout the drilling process. The results of this new area of research could change the way future horizontal wells are planned. Implementation of this algorithm can be applied in a multitude of ways; incorporating it as an artificial intelligence module in an existing drilling optimization simulator program, and/or directly integrated in the planning process by the drilling engineers to optimize drilling parameters for future planned wells.
Oil well drilling is a complex procedure and in order to optimize it, the physical phenomenon must be modeled as represented by the primary governing equation (Kerkar et al., 2014). Before the early 1980s, most oil wells were mainly drilled vertically (Helms, 2008). These vertical wells were influenced by many physical control variables including: weight on bit (WOB), revolutions per minute of the bit (RPM), drilling fluid type, drilling fluid viscosity, bit type, bit wear, etc. For drill bits, there are many sub segments of this area, however, most all of them can be classified into three categories: Natural Diamond Bits (NDB), Polycrystalline Diamond Compact Bits (PDC) and roller cone bits. NDB's are bits that have natural diamonds that are set along the surface of the bit face and grind the rock. PDC bits have polycrystalline diamond cutters set in the blades at the bit face and scrape or shear the rock. Roller cone bits are bits that have cones that roll along the rock face which crushes and gouges the rock as the bit teeth crush and penetrate into the rock. In this research, PDC bits will be incorporated since they are the most common.