1. THE SKIN EFFECT AND ITS INFLUENCE ON THE PRODUCTIVE CAPACITY OF A WELL ‘A. F. VAN EVERDINGEN, SHELL OL CO., HOUSTON, TEXAS, MEMBER AIME ABSTRACT ‘The pressure drop in a well per unit rate of flow is con- trolled by the resistance of the formation, the viscosity of the fiuid, and the additional resistance concentrated around the well bore resulting from the drilling and completion technique employed and, perhaps, from the production practices sed, ‘The pressure drop caused by this additional resistance is defined in this paper a= the skin effect, denoted hy the sym- bol S. This skin effect considerably detracts from a wells ‘capacity to produce. Methods are given to determine quanti- tatively (a) the value of S, (b) the final build-up pressure, and (c) the product of average permeability times the thick ress of the producing formation, INTRODUCTION Equations which relate the pressure in a well producing from a homogeneous formation with pressures existing at var- fous distances around the well are generally used within the industry. The relation is quite simple when the fuid flowing is assumed to be incompressible. It becomes somewhat more ‘complicated when the flowing fluid is considered compressible so that the duration of the flow ean he considered. In each case the major portion of the pressure drop occurs’ close to the well bore. However analyses of pressure build-up curves indi cate that the pressure drop in the vicinity of the well bore is greater than that computed from these equations using the Known, physical characteristics of the formation and the fluids. In order to explain these excessive drops it is necessary to assume that permeability of the formation at and near the well bore is substantially reduced as a result of drilling, com: pletion and, perhaps, production practice. This possibility has been recognized in the literature." ‘A method to compute the pressure drop due to a reduction of the permeability of the formation neat the well hore, which is designated as the skin effect, , is given in the following paragraphs, To start, equations normally used to describe flow in the vicinity of a’ well are given without considering this ‘effect. These equations then are modified to include the effect. fof a skin on the pressure behavior, Finally a method is given to estimate the effect of the skin won the pressure and produc: tion behavior of a well PRESSURE EQUATIONS Incompressible Fluid Flow If pe is defined as the flowing pressure isi a well of radius re: the pressure at distance r from the well has been shown to hes! ey * Seth 7 ‘The total pressure drop between the drainage boundary, 1 and the well hore is given by Pen = Be = In (ra/r fi Pla) Pe = gle In (on/re) oe OD) stant fie ithe Beem Bran oe Sen Exrygig! he Fete * Vol. 198, 1953, PETROLEUM TRANSACTIONS, AIME These equations are valid only if the flow towards the well ‘occurs in a horizontal homogeneous medium and the fluids are incompressible. The assumptions imply that all fluid taken from the well enters the system at 7, a condition rarely en: countered in practice, Compressible Fluid Flow, Steady State ‘A more realistic equation is obtained if it is assumed that the compressibility, ¢, of the flowing fuids is small and has a constant value over the pressure range encountered. After the well has been producing for some time so that.its rate has become constant and steady state js reached, the pressures throughout the drainage area are falling by the same amount per unit of time, and the pressure differences between a point in the drainage area and the well are constant. When these conditions are met, the rate of production, gy from a well is equal to shr'cf(dp/de), where dp/dt isthe pressure drop per unit time. The fluid fowing at a distance r from the center ‘of the well js equal to g(r,—r)/ru From the last equation tnd from Darey's law it-can be shown that Pl) —Pe= le (Inlra/re) 14) « e) ‘The equation holds for a depletion-type reservoir of radius re. drained by a well located in its center, provided the compres: sibility of the fluid per unit pressure drop is small and con stant, and no fluid moves across the boundary’ ry Compressible Fluid Flow — Nonsteady State Table II of reference (5) shows the relationship between the pressure at the well bore and the reduced time, 4t/fuct's» The pressuresdrop function, py. represents the drop below the original reservoir pressure, pu, caused by unit rate of production. (gis, = 4H/2ekh = 1) for several values of R, the ratio of drainage boundary radius, ra to well radius, rq Tin most reservoirs the values of ra/re approach infinity, and under these conditions the values of pyr, shown in Table I of reference (5) can be used where pyr, then signifies the dif ference between the pressure in the well and the prevailing reservoir pressure per unit rate of flow. The total pressure drop below prevailing reservoir pressure amounts to py— (qn/Bekk) per. where the factor gu/2ekh converts the etm: lative pressure drop per unit rate of production to eurnulative pressure drop for actual rate, q. For values of T> 100 the ry function may he written (equation VELS of reference 3) Pin = (In T + 0.809), Using the time conversion T= kt/fucr'e, the difference in pressure between reservoir and well becomes Tn (ht / fuer ar ii Tn (ht/foer.) + 0:09] @) If values for the physical constants of the formation and the fluids are inserted, itis found that 7 exceeds 100 after a few ids f production (or elosed-in time), s0 that the approxi mation hecomes valid almost at once. A simple relation between the pressure in the well and in the reservoir ean also be derived by considering the well as @ point source" instead of a unit cirele source, that is, by using Lord Kelvi’s solution instead of the unit circle source m Pumpems 358 THE SKIN EFFECT AND ITS INFLUENCE ON THE PRODUCTIVE CAPACITY OF A WELL solution.’ The difference in pressure at the well bore and the prevailing reservoir pressure then is Pom p= = Me Bi (— fore /Mt) a and this expression also approaches Equativn (3) closely whenever the value of (aer/4kt) is smaller than QOL which as stated before is the ease a few seconds after the start of production (or shutin). In all of the equations given so far itis assumed that unit rate of production is obtained immediately upon opening the well and, alternatively, that upon shutting-in the well the rate of production from the formation ceases abruptly. The storage capacity of the casing and tubing prevents this ideal condi tion from being obtained immediately in wells: Hence it is mally observed that the pressure builds up gradually <0 that it takes from a few minutes up to several days. (depend: ing on the characteristics of the formation, the contained Mid, and the storage capacity of the casing) before the observed Dressure-time relationship assumes the logarithmic relation mentioned above. Effect of Storage Capacity of Casing and Tubing Two methods can be used to express the effect ofthe storage capacity of casing and tubing on the flow equations given above In both cases it is assumed that a well hat been elosed-in saliiently long for the pressure to attain substantially the prevailing reservoir pressure. The well then is opened and its ‘measured production is corrected for the fluid obtained from easing and tubing to obtain the cumulative production from the formation proper. These corrections are derived from observations of casing-head, tubing-head, and bottom-hole pres sures, and from a knowledge of the dimensions of casing and tubing, and the weight of the oil and gas columns Method 1: The cumnlative production from the formation is plotted versus time, measured from the instant of opening the well, and, in general. a graph is obtained similar to Fig. 1 iene PRODUCTION 7 1TiMe = FIG, 1 — DETERMINATION OF RATE OF FLUIO ENTRY INTO WELL BORE 72 PETROLEUM TRANSACTIONS, AIME uur RSTE OF FaOMI Mogg whieh shows that the rate of production from the formation ean be closely approximated by a formula of the type ane), the observati ‘here both g and a are constants evaluated from The rate of production, q, in ce/second at reservoir condi tions equals gu/2xkh reduced rates. The dimension of the factor a, appearing in the exponent, is T". Expressing the ime in redveed units T= kt/fuer', causes the value of « to change to sfucr'./k = 8. The numerical value of the product 5 remains equal to From reference (5) it is clear that if a unit rate of produc tion (que) = 4H/2ekh = 1) gives a pressure drawdown of pi atmospheres, then the rate (1—e®") during the pressure drawdown will result in & pressure drop. Bir which is given bs the relation Poo PY pan 6 tion poxy with Sa whore pi isthe dierent of the unit expel 10 tie Using H(-£i(—47) for the pressure drop eaused by anit rate f production, tie fund that Equation (8) has ae ite explicit edhaion Fos Be [tn b—2y + Ind + BILE) m (6) where = 057722 = Euler's comtant and Ex(AT) = fF (e"/ubdu, whose values are given in refer ence (7) To analyze the complications encountered in the pressure buiap curves of wells the point source solution ie adoption of this cou ouree solution (which i more difialt Sidered permissible as shown by the fll Pos computed from: p=0001 50 100 0.21247 21608 200 0.46799 0.7336 500 1.20915 121487 1000 219670 2.20112 2000 3.38670 3.3844 Vol. 198, 1953A. F. VAN EVERDINGEN mw Method 2: Observations on wells have aso shown thatthe amount of fisid, C, which can be withdrawn from (or store into) casing and tibing per atmosphere pressure diferenc Is ‘Constant whose value can be determined ith reasonable accuracy. These alues of € are expressed in ce/atmosphere at reservoir conditions. In the system of reduced nite aed throughout this work C = 2rfherxC. as previously discussed: If we denote the pressuredrop —time relationship which would result from the relation of rate of production and un: Toading of casing ax Pry. the rate of production in reduced nits from the formation will approach unity according to (1-Gahin/a) oh and by the superposition theorem’ the drawdown in a well alter opening will be equal to Fon =F 1-Clfer/aT vw?» B) which fs Equation VIILS appeating in reference (5). Pin Curves ate shown graphically as Fig. 8 ofthe same reference As shown on Fig. 2, both Fis, and Py when ploted versus the log of time show some of the lag in’ pressore buil-up 50 characters in all pressure observations ‘The Skin Effect Although the pressuredrop function modified for the vari- ble rate of production prevailing immediately after closing-in, or opening-up of a well shows some of the characteristics of the observed pressure buildup or drawdown curves encount- tered in practice, agreement between these modified functions and factual data leaves much to be desired. In general, the lilference between prevailing reservoir pressure and flowing pressure is larger than can be accounted for by allow: ing for the variable rate in the manners explained above Better agreement can he obtained if it i= assumed that the permeability of the formation at and near the well hore is ‘substantially reduced as a result of drilling, completion and, perhaps, production practices. Whatover may he the cause for this reduction mo reason can he found to assume that this reduction is present beyond 20 ft around the bore hole and probably not that far. The volume of the fuids contained in such cylinder is small compared to the volume of fluids within the drainage area of a well. It may therefore be con- cluded that any transient conditions set up in this eylinder are of short duration and can be neglected in the analysis. Hence the effect of a reduction in permeability in this eylinder ccan be taken into account as an additional pressure drop, proportional at all times to the rate of production from the formation. For this reason the additional. pressure drop (per unit rate of flow) near the well bore is considered to be caused by a skin and denoted by S. Under these conditions Equation (5) is modified further to give Pa Bor fe) paw dT’ + (1-6 which has the explicit solution Bon = pe + S—4e™ [In B27 +Ind + Ei(AT) + 25] De ee (10) ds (9) ca It is easy to see that for large times when e™” becomes zero, Equation (10) gives Pen = Pn +S, +o that the entire pressure drop equals pen Ltn (ht fuer Ap = 2b [pen + S$] = Lin (ht/ ner.) + 0.909 + 25) an which is the same as Equation (3) after allowing for the pres- sure drop caused by the skin, Vol. 198, 1953 PETROLEUM TRANSACTIONS, AIME 3581 Furthermore, the presence of a skin causes Equation (8) 10 be modified to Bos 3 L-Cldin/AT)] pwnd” + (1-Capins)/AT)S «ay With the help of LaPlace transformations jn, can be ex- pressed as the infinite intogeal PY ae (a3) VIET of reference which becomes identical with Equat (5) when $= 0, It can be shown that for large times, Equation (12) also gives Fin) = pen) +S: Both Equations (10) and (13) can be used to represent with reasonable accuracy the entire pressure build-up in a well, so that itis felt that all factors influencing the pressure rise (or drop) have been taken into consideration. However, it is not possible to determine the numerical value of the ‘ous parameters entering the equations; to be precise, it is rot possible to determine from a preseure build-up eurve, even if tt fite a theoretical curve neatly, the value of S and of the ie conversion k/fncr’s. Equation (11) gives, in a simpler ‘manner, all information useful in field operations which 10 date has een extracted from the more complex Equations (10) and (13) ANALYSIS OF BUILD-UP CURVES ‘The pressure build-up curve of a well is obtained by meas- uring the bottom-hole pressure in a fowing well, pz, together with the subsequent pressure increases during period of Sufficient duration following the shutting-n, Tt is assumed that the well has been producing at a constant rate, q, during 4 considerable time, t. The pressure increase upon closing-in is recorded as a function of the elosedin time, 8, and only those pressure increases are used after the effects of storage RATE OF PRODUCTION, @ FIG. 3.— SUPERPOSITION OF FLOW RATE FOR CLOSING IN A WELL. 173.THE SKIN EFFECT AND ITS INFLUENCE ON THE PRODUCTIVE CAPACITY OF A WELL in casing and tubing have died down. In the formulas the closingin of a well is taken into account by superposing a negative rate, 9, s0 that the rate of withdrawing fluids from the formation becomes zero, as shown on Fig. 3. From Equation (11), and as shown in reference (6), the pressure drop at any time (¢-+ 8) caused by the production is found to be oe . gly Um (+ 8) tm (e/ free’) + 0.809 + 25) and the pressure increase caused hy the superposed nega rate g Is = [Im + In (k/ ery) + 0.809 + 28 lg (Uns +n Ck fners) + 5} ‘The actual pressure decreate at time 8 after elosing-in is given by the sum of these two expressions em po= Ape = 3H, (ln (¢ + 8) —In3] In[(etay/o] ad) Equation (14) indicates that the pressure change is propor- ional to a Incfunction of (¢-+8)/3 and therefore forms a straight-line relationship when plotted on semilog paper. Using Equation (14) for the determination of the prevailing reservoir pressure requires that r4/re be essentially infinite. For the determination of the average permeability and the skin factor it is only necessary that the pressure build-up curves contain 4 straight-line portion. Some of this information has been presented in somewhat different forni™* and is included here to present a complete fanalysis of a pressure curve including the skin effect. Determination of Prevailing Reservoir Pressure For infinite closed-in time (8 = ce), In [(t + 8)/3] and Ap ecome zero. Therefore the pressure read at In [¢-+2)/8] 0 reflects the reservoir pressure, ps. which would prevail at that point had mo fluids ever been taken from the reser- ‘oir at that particular location. The relation affords a simple Imeans to determine reservoir pressures from build-up surveys where the actual measured pressures are still considerably below final shut-in presturas. It requires, however, that con tinuous measurements be made. Socalled spot readings are Tikely to he misleading, since the errors in such measurements commen ae T= be Table 1—Reperforating Job Perforated interval 6258.70. 6258-70 ft ‘producing interval 58 em 58cm umalative production 33500/D 5500 B/D. Production rate 6BID B/D Production time M21 days 91.67 days production time $501,000 sec 7,920,000 see Shrinkage factor 0595 0.708 igateeservoir conditions BRec/see 13D ee/see Pressure inereae per eyle Gopi Spt Pelowing pressure 2060 ps 2502 pe ‘a iacosits 65 ep 65 ep Reporesiy 9219) 219) id compressibility {00017 /stm 0.00017 /atm 630m 63cm ey well adios (especially when made with different recorders) are apt to exceed the relatively small presture increases actually occur ring and thereby give an erroneous slope to the straightline portion of the build-up curves Determination of Average Reservoir Permeability ‘The slope of the straightline portion of the build-up ‘curve according to (14) is equal to gu/4kh. When a plot on semiloglog paper is used, the pressure increase peru rate of flow (gir, = qu/2ekh = 1), per fold increase in (t+ 8) /8 is equal to 8 In 10 = 1.1513 atmospheres = 16.924 1h, Hence if the pressure increases 40 Ib per eyele, the well has produced at 40/1692 236 unit rates = qu/2ekh When q and 1 are known from other sources the average value ‘of kh can be found. If reliable value of h is also available fan average value of the permeability can be obtained, Determinal of Skin Factor Equation (11) shows the pressure drop at the time of shut in to be = (In t+ tn (k/ foes) +0809 4 28]. 1 AP ign = Ga (i/foer’) + 0809+ 28]. (1) = FIG, 4 PRESSURE BUILD-UP CURVE BEFORE REPERFORATING. 4 PETROLEUM TRANSACTIONS, AIME FIG, 5— PRESSURE BUMLDLUP CURVE AFTER REPERFORATING. Vol. 198, 1953A. F. VAN EVERDINGEN 3581 ce 7 and Teme Mr ee She Fo, 6 PRESSURE BUILDUP CURVE BEFORE ACIDIZING. FG, 7 —PRESSURE BUILD-UP CURVE AFTER ACIDIZING. Equation (14) shows the pressure drop at closed-in time 8 production before closing-in, by the rate prevailing at that tobe time. This approximation of ¢ becomes more reliable the longer the well is produced at the constant rate, q. The following a inf +9V/8) « (14) information ean be derived from the table and Figs. 4 and 5. OP Gakh For values of 8 small compared to t, In[(¢-+2)/8] is essen tially equal to Jn 2/2, so that Equation (14) ean be modified to =H UIntind Ma 0, = ier ltnt In) (ita) By subtracting Equation (14-4) from (11) = [In 8-+ In (h/ facts Pagan ~ BPigs =e Und + In (hfe?) + 0.809 + 25) a5) Since the lefthand side of this equation is the difference be- tween flowing pressure and the pressure on the straight line for time 8, and qu/2rkh is the slope, a value of In (k/fucr’.) ++ 0.809 + 25 (a6) be found. If the radius of the well bore, the compress of the fluid, and the porosity are inserted in the equati ‘value for $ is obtained To obtain the above-mentioned objectives a definite str Line portion of the build-up curve should be available. This is further shown by a consideration of Fig. 2 which shows that the Bir» functions have a tendency to show a linear relation ship with Jn T before actually coinciding with the pre, fune- tions, ‘Therefore the duration of the pressure survey should bbe considerably longer than the time required for the effects ‘of storage in casing and tubing to die down. FIELD APPLICATIONS Field experience shows that the productive capacity of a well ean be increased considerably by reducing the value of (4 Ink/fuer's +S). Two examples are given to illustrate this statement. Reperforating ‘The effect of reperforating a well on the value of S is given in Figs. 4 and 5; production, well, and PVT data are sum- marized in Table 1. ‘As a well is seldom produced at a constant rate, the time, 1, before closingsn is approximated by dividing the cumulative Vol. 198, 1953, PETROLEUM TRANSACTIONS, AIME Afr Reperforating Pe 2554 psi Pa-Pr 52 psi 187 48 [In (k/fucr) +8] 454 psi i STA YG In (k/fucr’n) +8] $54 ps 24 ps MAT a (349 tn (k/jucr’e) 13 psi 8 psi SSE (In (k/ freee) psi si 1487 gu 491 psi i sree Ps 16 psi s we 46 k 118 ma Al md ‘The curve shown is typical for a formation having high values of 2rkh/1, as usually found in the Miocene sands around the Gulf Coast. Under these conditions the pressure increases per eyele are small, which stresses the necessity of obtaining accurate data for determining the straight-line por- tion of the buildup curve, The determination of $, also, is highly sensitive to variations in the small values of gq/2zkh. In extreme cases the straight-line portion of the build-up curve approaches a horizontal line. and the entire pressure increace after shutin of the well is then an indication of the ‘Table 2— Acidizing Job Belore 2103 cm 250B/D Production time se 4,908,000 see Shrinkage factor 088 at reeervoir conditions 523 ce/see Prese The pst Pol 3534 psi iscsi O80» Pporesiy ‘039 ld compressibility 1.0001 stm Pes well radius £503 emTP. 3581 ‘THE SKIN EFFECT AND ITS INFLUENCE ON THE PRODUCTIVE CAPACITY OF A WELL presence of a considerable skin effect, whose actual value, however, cannot be determined, Tt is evident that in the case diseussed above the fist per- foration job was not eficient and that reperforation essentially removed a large resistance which existed near the well bore- Before reperforating, the entire pressure drop was 494 psi out ‘of which 441 psi or 89 per cent was caused by the skin eflect, After reperforating, these figures were 52 psi, 16 psi and 31 per cent respectively. Since the rate before reperforating was 96 B/D against 60 B/D after reperforating, the last set of pressure figures given should be multiplied by 1.6 in order to obtain a valid comparison. Acidizing An example of a well before and after acidization is given in Figs. 6 and 7; the production data are assembled in Table 2, Pertinent conclusions follow. Before After Acidizing ps 4600 psi Dempe 1066 psi [Yn (k/uer's) +S]. 519 psi ek TAT [46 In (h/ fuer STO [Yn (K/foers)] 147 gu 188 psi 381 psi 56 69 md 70 md ‘The presence of a negative skin may be questioned and is shown here merely to illustrate the change resulting. from acidizing the well. It is probable that a negative value for S reflects an increased effective well radius, 7 ‘The above analysis shows that the value of S, is not the most significant figure that can he obtained from this type of information. It seems reasonable to suppose that acidizing increases the permeability of the formation immediately sur- rounding the well bore to such an extent thet it becomes tremely large compared to its original value. This inerease in permeability for some distance around the well bore can be regarded as increasing the effective radius of the well several fold (without increasing the size of the bore hole). Hence, acidization will eause not only the numerical value of S to decrease but also the value of % In (k/fuer’.) to decrease in the same operation, due to an increase in r.- Hence th ‘effectiveness of acidization can best be judged by comparing the combined values of 3 In (k/fucr’s) -+ S before and after acidization, without considering which of these two factors ‘contributes to the improvement. From this reasoning it follows that acidization should be repeated whenever a sizeable de- crease in the sum of these two factors can be expected. NOMENCLATURE Used in this Paper Used in Ref.5 Production time, seconds ' T Production time in reduced unite rT ' Time well has been shutin, seconds 3 - Permeability, dareys k K Porosity, a fraction i f Viscosity of reservoir uid, centipoises » . 176 PETROLEUM TRANSACTIONS, AIME Compressibility of the reservoir fuid/atm « © Radius of well, em. t R Drainage radius of well, em n RR, Rate of fluid flow (ce/see., reservoir conditions) Formation thickness, em. Rate of flow in reduced units Prevailing reservoir pres sure, Le, well pressure alter infinite shutin time py - Pressure at drainage boundary of flowing well Flowing bottom-hole pres- sure in the well at time of shutin Increase or decrease in Dottomhole pressure at time 2 after shutting-in or opening-up a well Pressure decrease caused by unit rate of flow Pressure drop per unit rate of flow caused by the ‘kin, dimensionless s - ‘The natural (Naperian) log of a number H en = a8) 2k guy = gn/ 27K Pla) = op oP Pen Pov ny - Note: All terms in the equations used in this paper are in the system of units associated with Darcy's law, and the reader is referred to suitable conversion tables such as shown in reference (6) whenever production data are expressed in a different system. ACKNOWLEDGMENT ‘The author is indebted to William Hurst who, while an ‘employee of the Shell Oil Co., was associated with him in the development of much of the’ work presented in this paper. Hurst, however, has no responsibility for the form and con- tent of this publication. REFERENCES J. Ex: “Permeability as Re- API Dril. and Prod. Prac. (1943) 66:81. 2 Miller, C. Cy, Dyes, A. B., and Hutchincon, C. A. Je: “The Estimation of Permeability and Reservoir Pressure from Bottom Hole Pressure Build-up Characteristics,” Trans AIME, (1950) 189, 91 3, Johnston, N, and Van Wingen, N.: “Evaluation of Zonal Damage,” World Oil, (Dec., 1950) 156-164. 4. Muskat, Mz The Flow of Homogeneous Fluids Through Porous Media, McGraw-Hill Book Co., (1987). 5. van Everdingen, A. F., and Hurst, W.: “The Application of the LaPlace Transformation to Flow Problems in Reser- voirs.” Trans, AIME, (1949) 186, 305. 6, Horner, D. R.: “Pressure Build-up in Wells” Proc. Third World Petroleum Congress, Section Il, E. J. Brill, Leiden, (1951). 1. Tables of Sine, Cosine and Exponential Integrals, Federal Works ‘Agency, Works Projects Administration’ for the City of New York, (1940), 8. Carslaw, H. S, and Jaeger, J. C.: Operational Methods in Applied Mathematics (Second Edition), Oxford University Press, (1947). +e Vol. 198, 1953