Summary The development of rotary rock bits with jet nozzles required means of estimating pressure losses in the drilling fluid flowing throughout the well being drilled and through the associated equipment. The initial tabulations were based on Newtonian fluids. Subsequent authors developed descriptions of drilling fluids based on Bingham or power law non-Newtonian fluid models. Because optimum-hydraulics theories dictate that hydraulic horsepower, impact, or impact force must be maximized, we made the difficult decision to determine these pressure losses by actual tests. A total of 119 water-base drilling fluids were pumped through capillary tubes up to 2 in, in diameter and through six standard sizes of drillpipe tool joint combinations. Drilling fluids were flowed through jet bit nozzles and were flowed up two annulus-size combinations as well as an annulus with hole enlargements. The annular tests included cuttings, which aided in determining flow patterns. This paper includes development of friction factors and empirical corrections for current theories to model flow of highly non-Newtonian fluids more reasonably. Procedures and equations arc offered to help estimate pressure losses in a drilling operation, even with very limited fluid property information typical of our industry. Introduction Field tests of flocculants and other polymers added to water or clay water drilling fluids resulted in lower pump pressures and higher pump speeds than predicted with available tables. In some cases, these pressure reductions were dramatic and resulted in loss of rig time inspecting for equipment failure. Field pressure measurements showed that the pressure losses for actual muds were significantly different from those calculated by available methods. A laboratory test facility was constructed to measure the exact pressure losses for various types of drilling fluids in actual drillpipes and annuli. The initial concept was to develop more comprehensive tables for various types of muds in the range of pipes used. After initial tests it became apparent that the number of tables required would make their use unattractive. Because this work was to assist field personnel in calculating pressure losses so that hydraulics could be optimized, it had to be based on flow property measurements available in the field. The best available field data are often the Fann plastic-viscosity and yield point measurements based on the 600- and 300-rpm readings. Dodge and Metzner indicate that the powerlaw fluid model can be used to describe the flow of drilling fluid in pipes; therefore, these fluids were treated as power-law fluids with suitable corrections to be applied where required. Equipment Flow properties and pressure drop in pipes were measured by pumping the test fluids through 10 pipes ranging in ID from 0.187 to 3.826 in. and in annuli from 5.044 × 2.5 to 12.715 × 5.0 in. Nominal 20 in. welded pipe and six types of drillpipe were manifolded into the full-scale test system. This system included a 20-bbl mud tank, a 100-hp electrically driven centrifugal pump, and automatic diaphragm valves for flow and bypass controlled by the flowmeter. Flow rates were measured by )- to 500- and 0- to 50-gal/min flowmeters that demonstrated better than 0.25% full-scale accuracy. Differential pressure along the test sections was measured with 0- to 100-psi, 0- to 400-in. water, 0- to 100-in, water, and 0- to 20-in. water differential-pressure cells. An automatic continuous flow of city water from the cell to the pressure taps was maintained at a rate of about 20 cm /min to prevent mud from entering the test lines. JPT P. 1414^
Read full abstract