Gravity Drainage of Oil Into Large Horizontal Fractures

Abstract
Published in Petroleum Transactions, AIME, Volume 219, 1960, pages 7–15. Abstract An investigation has been made of the production of oil through horizontal fractures of high capacity and large radius placed at the base of producing formations. The specific aims of the study were todetermine by model studies the reservoir performance when gravity drainage is the only producing mechanism, anddevelop a method of predicting such performance from commonly measured reservoir properties. Although the theoretical approach failed to develop a rigorous analytic solution to the transient problem, it did lead to a practical method of predicting performance. Predictions are in good agreement with rates and recoveries observed in scaled-model experiments. Capillary effects and the influence of relative permeability are taken into account. The experimental program wad mainly concerned with the influence of fracture radius, fracture capacity and capillary hold-up on the total recovery and the rate of recovery. Much of the experimental work was done on a triangular model which represented one-eighth of a square well pattern, so constructed that horizontal fractures of varying radii and flow capacity could be studied. An electrolytic model was used to relate the initial production rates to fracture radius, formation thickness and the drainage radius of the well pattern. The conclusions reached as a result of the work are as follows.The production of oil under gravity force alone into a well-propped horizontal fracture, whose radius is of the order of hundreds of feet, is a recovery method which combines high recoveries with very acceptable production rates.The mechanism is effective in "dead oil" reservoirs as well as in reservoirs which possess the energy of solution gas.A method of predicting performance has been developed which gives good agreement with experimental behavior in cases of practical interest. Even for very thick beds and large fracture radii, where actual conditions do not conform to the assumptions made, the method gives order of magnitude predictions which would not otherwise be possible.