Dynamic Fluid-Loss Study Of Fracturing Fluids

Abstract

Abstract The fluid-loss behaviour of a number of aqueous fracturing fluids was evaluated under dynamic conditions. The effects of core permeability, fluid velocity, differential pressure and fluid-loss additive were measured. Most of the data presented was generated with non-crosslinked hydroxypropylguar fluids. Examination of the data revealed that on low-permeability cores, there was essentially no difference between the results of static and dynamic tests. Plots of fluid-loss volume vs (time)1/2 were linear for both tests. Fluid velocity in the range from 0 to 0.6 m/sec did not affect fluid loss. Increasing the differential pressure across the core increased thef/uid-Ioss rate. Increasing the c.ore permeability by a factor of 100 drastically increased spurt volumes but slightly decreased Cw values. The addition of diesel oil improved the fluid-loss properties of the non-crosslinked fluid. Tests on crosslinked fluids indicated that the fluid-loss rate decreased for a period of time, but eventually became constant. The length of time necessary to reach steady-state fluid loss depended on the fluid and on experimental conditions. Introduction The fluid-loss characteristics of a fracturing fluid greatly affect its performance. The lower the leakoff rate of the fluid, the more effective the fluid will be at creating a fracture. Consequently, the fluid-loss properties of fracturing fluids have become a major concern of many oil and service companies. The most commonly used method for studying fluid loss is a static test in which fluid does not move past the core face. There has been concern that this test does not accurately simulate the fluid-loss process. A flowing fluid under shear could have a significantly different rate of filter-cake deposition and equilibrium filter-cake thickness than a static fluid. Several studies on dynamic fluid-loss testing of fracturing fluids are available in the literature(1–5). The earliest reports were by Hall and Dollarhide(1–2), who studied both oil and water-base systems. While fluid loss from crude oil systems as found to follow a square root of time dependence, fluid loss from kerosene and water-base systems showed a straight time dependence. Thus, the crude oil behaved in a dynamic test as it would in a static test, but the kerosene and water-base systems did not. The authors suggested that once a filter cake is established, flow past the filter cake inhibits further buildup. All but one of these early studies dealt with oil-base or non-crosslinked and water-base fluids. In addition, the test flow velocities and/or shear rates were often higher than currently believed to exist in the fracture. This study was undertaken to learn more about fluid-loss control of non-crosslinked and crosslinked aqueous fracturing fluids under flow conditions similar to those found in the fracture. Experiments These dynamic fluid-loss tests were run on a fluid-loss model previously used for cement testing. The equipment design and general test procedure are described elsewhere(6). All tests were run at ambient temperature. All of the fluids were prepared with the same base material-hydroxypropylguar (4,790 g/m3), silica flour (420 g/m3) and buffers to produce a pH of 9 to 10.