Selection of the particle size distribution of granular lost circulation materials for use in naturally fractured thief zones

Abstract

Granular lost circulation materials (LCMs) are used to control fluid loss in naturally fractured thief zones. Rapid and automated identification of their optimal particle size distribution (PSD) is highly desirable. This paper proposes a PSD selection method for LCMs based on a PSD representation algorithm and the Monte Carlo method. The PSD representation algorithm describes the cumulative and compositional PSD curves of LCMs according to a piecewise cubic Hermite interpolation algorithm. Comparison with measured PSD data shows that the representation algorithm is more accurate than other commonly used PSD models in describing both homogeneous and mixed LCMs, and the relative errors are all within 10%. The Monte Carlo method is adopted to simulate LCMs with various compositions and select ones with PSDs that meet a proposed selection criterion. Experiments show that the PSD curves of the LCMs were all within the upper and lower bounds specified by the selection criterion. Plugging experiments show that LCMs with the recommended compositions can effectively enter fractures, and the fluid loss is less than 200 mL for half an hour under the pressure difference of 7 MPa. The proposed selection method can be used for the rapid and automatic design of granular LCMs used in fractured thief zones.