recuntcau wore 373 PREDICTION of OIL RECOVERY by WATER FLOOD — A SIMPLIFIED GRAPHICAL TREATMENT of the DYKSTRA-PARSONS METHOD CCAUFORNIA RESEARCH CORP. UA HABA, CAUF (CARL E JOHNSON, 18 INTRODUCTION ‘A. method for predicting water- flood oil recovery was reported by H, Dykstra and R. L. Parsons’ in 1950. It is now generally known as the Dykstra-Parsons method and is widely used by petroleum engi- The method is semiempiricel and consists of a correlation of four fundamental variables. These are: V, vertical permeability variation;’ «, mobility ratio; S., initial water sat uration; and 2, fractional recovery of oil in place at a given producing wwater-oil ratio. The correlation ex- tends over a wide range in each of these variables, and may be applied to all formations with inital oil sat ae cw Ho ll Keone turations of 45 per cent or greater. yg. |._PeRMEABILITY VARIATION PLOTTED AcaINst Mowitity Ravio The method of calculation origin- ‘SHOWING LivEs oF CosTANT R(L ~ Sy) FOR A PRODUCING ally outlined by Dykstra and Parsons WareR-On, Rarto oF 1 closely follows their derivation, And ase) this is a convenient way of illustrat- ing the development of some rather ‘complicated ideas. However, in ac- tual use it is more cumbersome than necessary. The purpose of this note, therefore, is to provide a simplified method for making Dykstra-Parsons predictions. ‘A SIMPLIFIED GRAPHICAL ‘SOLUTION ‘The correlation between V, a, S.. and R, corresponding to a given pro- ducing water-oil ratio, can be shown con a single graph. This is best done by plotting V against @ to show lines of constant RC ~ Sy . WOR™). Figs. 1, 2, 3, and 4 show these plots row Fig, 2—Perweanitity VaRTION PLotrep AcaINst Monitity Ratio Sianscrpt secind in Petrol Branch Suowine Lives oF Constant RU ~ 0.72.$.) FOR & PRODUCING Eeftnce vt atch of paper. WareR-On, Ratio oF 8 SPE 733-Gfor water-oil ratios (WOR) of 1, 5, 25, and 100, respectively. In each figure WOR"* has been reduced to its numerical value, Use of the figures requires assign- ment of definite values to V, a, and Sq. Both V and S_ can be obtained from core analysis data.* The mobil- ity ratio, @, depends upon four sub- Sidiary variables. It is given by the expression: where 1, and jr» are the viscosities of the oil and the flood water under reservoir conditions, ky. is the rela- tive permeability to water in the res- ervoir when only water is flowing, and ky, is the relative permeability to oil when only oil is flowing. Meas- turement of the {wo viscosities is sel- dom a problem, And the two relative permeabilities can be measured ex- perimentally or estimated from the data of Leverett and Lewis’ for un- ‘consolidated sand packs. To find R the appropriate figure is entered at V and a. For example, suppose we wish to predict the frac- tional oil recovery when the pro- ducing water-oil ratio reaches five. Fig. 2, which shows the correlation at a water-oil ratio of five, must be used. Suppose V and a are 0.50 and respectively. Enter Fig. 2 at these values. The point of intersec- tion shows that R(1 ~ 0.72 Sy) is 0.25. If the water saturation, Sv, at the beginning of the flood is 21 per cent (0.21), then R is found to be 0.29, This means that 29 per cent of the oil in place will have been re- covered when the producing water- cil ratio reaches five, ‘The R value obtained above ap- plies to a linear flood. When deal- ing with pattem floods, R must be ‘multiplied by @ correction factor Which takes into account the fact, the areal sweep efficiency of the pat= tern is less than one. Furthermore, to convert R to barrels of oil recov- ered, the volume of oil in place must be known. If this number is ex- pressed in stock tank barrels, thea its product with R will be the recovery in stock tank barrels. Kew e Kone Fic, 3—PERMEADILITY VARIATION PLOTTED AGAINST Mositity Ranio Suowinc Lines oF Constant R(I ~ 0.52 5,) FOR 4 PRODUCING WATER- ‘Ow Rano oF 25, Me i Keno Kom Fic. 4—PeRMeaniLity VaRiavion PLortep AGAINST Mositity Ratio SnowiNe Lines oF Constant R(1 = 0.40 Sy) FoR 4 Propucine WATER On. Ratio oF 100. ACKNOWLEDGMENT ‘Second Edition, API, New York, ‘The author expresses his appreciae N+ Y: (1950) tion to Herman, Dykstra who pro. 2. For an explanation of the deter- vided the basic information neces- mination of V from core sample sary to the construction of Figs data, see Ref. 1, p. 171, Single and 3 ‘copies of Ref. 1 may be obtained by writing to the author at Cali REFERENCES fornia Research Corp., P. 0. Box 1, Dykstra, H., and Parsons, R. L.: 446, La Habra, Calif “The Prediction of Oil Recovery 3. Leverett, M. C., and Lewis, W. by Water Flood” Secondary Re | -B.: Trans. AIME (1941), 142, ‘covery of Oil in the United States, 107. ee