API-44-053

RESERVOIR PERFORMANCE j-

J. J. MULLANE " I ABSTRACT

It is shown that there are 3 fundamental types of I water production as factors of reservoir performance, perforniance or recovery ~nechanisnls by whieh petro- leum reservoirs nlay be operated, viz.: l, dissolved-gas drive, in which only the energy due to the gas dissolved in the oil is utilized to produce oil; 2, gas-cap drive, in '

which displacement of oil is effected by an expanding gas cap, coupled with gravitational segregation of oil and gas; and, 3, water drive, in which displacenlent of oil is effected by the controlled encroachn~ent of water. Of these ~nethods, the latter two are the most efficient types of operation and. by means of these, the highest recoveries of oil from reservoirs can be obtained. The characteristics of each type of perforn~ance are de- scribed and illustrated by nieans of actual field examples.

Particular attention is directed to the i~nportance of early identification of the type of drive to be employed as the recovery nlechatlisn~ in a petroleum reservoir, and it is stressed that the dissolved-gas-drive nlechanisn~ is to be avoided whenever possible. The use of geologic and engineering data to make this distinctiot~ is illustrated. The proper control of producit~g rates, and gas and

are discussed; and a nuniber of exan~ples of the appli- catiora of ~noder t~ engineering principles to problenas of these types are given.

The subordinate r6le played by well spacing in the modern concept of reservoir perforniance is discussed. The particular point stresbed is that well spacing alone is not the controlling influence on recovery efficiency; it is the control of the reservoir as a whole, in such a way that the most efficient n~echanisn~ of recovery is propcrl? utilized. that governs the recovery to be ob- tained. Specifically, in any field, a sufficient nu~nber of wells should be drilled so that, when the field as a whole is produced at an efficient rate, the rate of production of individual wells is not excessive; i.e., individual wells

should not be produced at such rates that there is an

excessive production of free gas or water, or that the rate of depletion of the sand in the neighborhood df the wells is materially ,greater than in the remainder of the reservoir.

INTRODUCTION I make a choice between them a t a n early stage of the

The necessity and desirability of operating petroleunl reservoirs a t the maximum efficiency practicable have occasioned the development of the growing science of reservoir engineering. The proper development and con- trol of oil reservoirs form the subject matter of this field of study. During recent years the work of a num- ber of petroleum technologists has.resulted in the estab- lishment of certain basic principles, upon which modern thought on oil reservoir performance is based. I t is the purpose of this paper to present these principles and to show, by actual examples, how they may be effectively utilized.

Reservoir Performance

Three definite modes of performance of petroleum reservoirs have been recognized: 1, dissolved-gas drive; 2, gas-cap drive; and, 3, water drive. These a re defined according to the source of energy, or drive, under which the oil is moved through the producing formation into the well. I t is important to know the outstanding fea- tures tha t characterize each of these types of operation because, in a great many instances, it is possible to

* T h e i::lrter Oil CI!.. T~11sa. 0kl:l. t P r e s e ~ ~ t w l a t slrrilly ~nret i l lg uf the 3Iirl Contillcut District.

rlivision of Pr41ductiul1. Tr~lsic, Okl;~.. 31ny 5 - 2 6 , 1944 : presiding, : Potro lvnt t~ ('n H : ~ r t l ~ s \ ~ i l l r . Okla. , A .. . . - ... . . , - .. . .-. - - . . -~

v Dri l l i l~g Contractor :11lr1 C'onsllltillg k-~;r.inerr.' i ) k l :~h , ,~ i l i~ Citv. Ok1a.l : nlld E. A. Stel!llellsull, Ulli-

development of a field. I n particular, what must be known a r e the criteria by which one kind of perform- ance may be distinguished from another, the conse- quences tha t result from a specific choice, and the operating procedures t h a t must be follou~ed to achieve the type of performance desired. The character of .each type of drive will be briefly outlined and illustrated with examples taken from field experience.

Dissolved-Gas Drive

When the sole source of energy available or utilized to produce the oil from the reservoir is in the gas dis- solved in the oil, the reservoir is defined a s being in operation under a dissolved-gas drive. This type of performance is characterized by continuous and rapid decline in reservoir pressure, paralleled by a steady decline in well potentials; and it.exhibits a distinctive gas-oil-ratio behavior, in tvhich the gas-oil ratio rises steadily to a maximum value several times the original ratio, and thereafter declines. The recovery to be ex- pected from this type of mechanism ranges from 15 to 40 per cent of the oil originally in place.

The conditions which make this method of operation mandatory a re :

a., F la t structure, usually with substantial stratification or low vertical permeability, obviating the possi- bility of appreciable gravitational segregation of gas released from solution.

b. Absence of a free gas cap or of a water body which could move into the oil reservoir.

c. High ra te of production, substantially exceeding the ability of any water present to advance into the reservoirs, o r the ability of a free gas cap, if present, to expand efficiently.

In Fig. 1 a r e shown the production statistics fo r the Gloyd-Mitchell zone of the Rodessa Field in Louisiana. This reservoir is a nearly flat extension of the main Rodessa structure. Production is from 2 intervals in the Gloyd zone, found a t 5,900 f t . The upper one of these, a sandy lilne, is known a s the Mitchell sand; and the lower section is predominantly oolitic lime, and is designated a s the lower Gloyd. The oil produced from this field has a surface gravity of 42 to 43 deg API, and the solution g a s was originally 627 cu f t per bbl under the original pressure of 2,400 psi. No f ree g a s was originally present in this section, and there has been no evidence of water drive noted. The wells com- pleted in this zone were produced a t high rates, and experienced a rapid decline in production. The availa- bility of reasonably accurate data on this field has made possible the presentation of a n excellent practical ex- ample of the salient features of this type of drive, and provides a satisfactory check on the theoretical expec- tancy. It will be noted t h a t the production and pres- sures both suffered a steady rapid decline, and the

Rodessa Field-Northwest Gloyd Estension- Graphical Statistics.

FIG. 1

gas-oil-ratio history has exhibited behavior typical of this type of drive.

The ultimate recovery from this zone has been esti- mated to be 20 per cen t of the original oil in place. This low recovery is in accord with t h e expectation f o r this type of drive. This conclusion is based on the study of a number of fields which have been wholly, or largely, produced under this mechanism, and includes fields ' producing from the Chester sands in Illinois and the Bartlesville sand in Oklahoma. Although recoveries a s high a s 40 per cent ]nay be anticipated under this drive, by f a r the majority of cases reviewed showed recoveries tha t ranged from 20 to 30 per cent:

Gas-Cap Drive

Under this type of drive or recovery mechanism, a distinct free-gas area-either originally present or created--espands, or is caused t o expand by injection of gas, thus encroaching downward into the oil zone and displacing oil downstructure. . Here the energy i n the system in the foinm of solution g a s is augmented by t h a t of the expanding g a s cap and, under proper condi- tions, will be fur ther enhanced by the force of gravity. In this type of drive the pressure may be either main- tained a t approximately its original value or maintained under a controlled decline. A t the same time the pro- ducing rates a re maintained a t Inore uniform levels than is possible under ordinary dissolved-gas-drive oper- ation. The free gas in fields of this kind is virtually all segregated in the gas-cap zone; and- the gas-oil ra t io of all wells, except those close t o the gas-oil contact, remains a t a low value. The gas-oil ratio of wells close to the gas-oil contact 'continues to rise until largely f ree gas is produced, a t which time such wells a r e closed to production. The ultimate recovery to be espected ffom this recovery mechanism will range from 40 per cent to possibly a s high a s 80 per cent of the original oil in place.

The conditions which favor this type of operation a re :

a. The sand must have high permeability; because, fo r practical purposes, the prevention of channeling and bypassing of gas, both t h a t evolved from solu- tion and tha t previously in the gas cap, would be very difficult if this condition were not fulfilled.

b. Pronounced structure is usually necessary to aid gravitational segregation.

c. The sands must be continuous and reasonably uni- form to permit thorough displacement of the oil a s the gas cap expands through the oil body.

d. Some ~estr ic t ion of oil-withdrawal ra te is necessary to prevent gas channeling and bypassing.

e. Careful gas conservation i s usually necessary if the full use of the displacing ability of the gas cap is to be enjoyed. In some cases this may require return of gas to the crest of the structure.

As an esample of this type of operation, t.he produc- tion statistics f o r the Mile Six Pool in Peru a r e shown in Fig. 2. This field, which produced a 40-deg-API- gravity oil from the Parinas sandstone, met the require-

ments for gas-cap-drive operation. I t has good porosity and permeability and pronounced structural relief, a s is seen on the structure map in Fig. 3. Throughout operations, there has been a high degree of gravita- tional segregation; and the volume of gas returned to the sand, up to 1940, amounted to 127 per cent of all the gas produced from the field. The effect of this on the downward migration of the gas-oil contact is graphi- cally illustrated on the map in Fig. 3. I t will be ob- served in Fig. 2 that the pressure has been maintained ali~lost a t the original level, and that gas-oil ratios have held remarkably constant. Practically all of the oil produced from this pool to date has been obtained by natural flow, and the indicated recovery efficiency is approsinlately 50. per cent. This relatively high effi- ciency is typical of this recovery mechanism.

Water Drive

When, in a petroleum reservoir, the principal means for the displacement of oil from a producing horizon into the wells is the encroachlnent of edge water or bottom water into the oil zone, under the influence of the pressure gradients created by the production from the wells, the .reservoir is said to be operating under a water drive. The chief operating characteristic of this

type of recovery mechanism are the maintenance of reservo&- pressure, a t relatively high levels throughout the course of productiop, the retention of productivity by the \veils, and operation a t low gas-oil ratios. The degree to which these ends are attained depends on the relation that exists for a particular reservoir between the rate a t which oil, gas, and water are removed from the reservoir and the rate of influx of water. The re- covery of oil under this mechanism will range from 40 per cent to as high as 80 per cent, depending on the properties of the reservoir and its contained fluids:

Water drive may come about in 2 ways: 1, simple artesian flow, which is conlparatively rare; and, 2, as a result of the espansion of the water in that part of the reservoir outside the oil zone. The requirements for a satisfactory water drive are as follows:

a. The oil body must be in good communication with a permeable and continuous water-bearing sand (usually the regional continuation of the oil sand itself) of considerable areal extent, or outcropping a t a reasonable distance.

b. The permeability of the sand must be fairly high, so that the rate of water advance can be sufficient to be of benefit within practical time limitations. Similarly, oil viscosity must be reasonably low.

Mile Six Pool-Operating Data;

FIG. 2

c. The sands in the oil reservoir proper must be rea- sonably continuous and uniform, so that effective flushing nlay be obtained. .

d. Rate of fluid withdrawals must be restricted to a level comn~ensurate with the ability of the water to advance evenly into the oil sands, if maximum effectiveness of the flushing action, through avoid- ance of channeling and bypassing by the encroach- ing water, is to be realized.

The behavior of the Dis Pool (Fig. 4) in Jefferson County, Illinois, furnishes a n excellent exanlple of a field operated under the water-drive recovery mecha- nism. The reservoir here is a small anticlinal dome in the Bethel sand, and the water-bearing par t is con- tinuohs in all-directions from the pool for a distance of a t least 20 miles. Throughout the life of the field to date, the pressure in the reservoir has been maintained above the saturation pressure (270 psi) of the oil. It is of interest to note that, in the past year, there has been a buildup in pressure of 19 psi. The water encroachment has been, and remains, uniform-the water contours following the structure very closely. I t is too early in the life of this operation to forecast accurately the recovery efficiency for the reservoir, but

Mile Six Oil Pool-Migration of Gas-Oil Contact.

FIG. 3

prel i~ninary estimates place i t in the neighborhood, of 40 per cent. This is to be compared with Eas t Tesas, which is a t the opposite extremity of the efficiency range. A t Eas t Tesas the recovery has been estimated, both from cores taken in the flooded zones and from careful calculations, to be approsinlately 80 per cent. Other water-drive fields, such a s Magnolia in Arkansas, producing from the Smackover lime, and North Searight in Oklahoma, producing from the Ordivician Wilcox sand, lie in between these extremes. I n general, there is to be found a relation between the recovery efficiency and the properties of the reservoir and i ts contained fluids. I n particular, recovery is affected by the permea- bility of the reservoir and the viscosity of the oil and, in a Inore general way, the recovery will be affected by the uniformity, or lack of it, in the reservoir. Further , there a r e capillary forces a t the oil-water interface which exert a definite, but a s yet not measured, effect on the recovery. At Dis the reservoir sand has a per- meability of 80 millidarcys, and the reservoir-oil vis- cosity is 2.5 centipoises; whereas a t Eas t Texas the permeability of the Woodbine sand averages 1,500 milli- darcys, and the reservoir viscosity is low, approximat- ing 1.5 centipoises. In general, however, although the recovery under water drive may vary from 40 to 80 per cent, depending on reservoir conditions, for any par- ticular reservoir water drive can be expected to yield a higher recovery than simple dissolved-gas drive. To return to the example given previously, normal re- coveries by simple g a s expansion in the Bethel and other Chester series sands in the Illinois Basin average 20 to 25 per cent, whereas recovery from this sand under water drive will approximate 40 per cent.

Con~parative Behavior

To sumnlarize the principles observed above, we may draw the following conclusions with regard to the com- parative behavior of these three fundamental recovery mechanisms :

1. The simple dissolved-gas-drive mechanism operates with a rapid depletion of pressure, whereas in the gas-cap drive or in the water drive pressure is maintained a t relatively high levels, if the proper operating practices suitable to each a r e observed:

2. I n the dissolved-gas-drive operation there is a n estrenlely rapid decline in the production rate , due to the rapid depletion of energy, compared to the retention of productivity a t higher levels in the other recovery mechanisms.

3. Each type of drive shows a distinctive gas-oil-ratio behavior. In Fig. 5 there is shown typical gas-oil- ratio behavior of the 2 types of gas-drive recovery. It will be noticed that, in the dissolved-gas drive, there is a very rapid depletion of the gas in solu- tion; whereas .in the gas-cap drive, particularly under those conditions favorable t o gravitational segregation, the g a s is conserved in such a manner a s to be of material assistance in promoting higher ultimate recovery, and the gas-oil ratio throughout

Pressure Decline and Fluid Withdrawal-Dis Pool, January 1944.

FIG. 4

RESERVOIR PERFORMANCE 57

the major part of the life of the operation is main- tained a t a low level. I t will be observed that, where gravitational segregation of oil and gas is not effective, the gas-cap-drive mechanism reverts to a behavior lying between that of simple dis- solved-gas drive and gas-cap drive, coupled with complete segregation. Lack of segregation will result in the production of excessive amounts of gas, rather than in its retention in the reservoir. Such a condition may be encountered in very flat structures that afford little opportunity for the control of the free gas, in tight sands, or in re-

pressuring operations, in which gas injection is made throughout the oil acreage. In water-drive operations, when properly controlled, the gas-oil ratio is also maintained a t low levels and, under some conditions, can be maintained a t solution gas-oil ratio over virtually the entire producing life of the pool. This is particularly true in such cases as the Dix Pool, cited previously, in which the reservoir oil was initially undersaturated. In those cases, when the oil is initially saturated a t the original bottom-hole pressure and a free gas cap is present, adequate control of gas-oil ratio

Typical Gas-Oil-Ratio Behavior.

FIG. 5

i8 PRODUCTION

and the avoidance of the production of g a s from the gas cap can usually be attained by proper well-completion and production techniques.

. With regard to ultimate recovery, these types of performance a r e again divided into two distinct

' groups: a dissolved-gas drive, with a generally low recovery; and the gas-cap drive and water drive, with generally higher recoveries. Esperi- ence and study have shown t h a t the dissolved-gas- drive mechanism is inherently the most inefficient means of recovering oil. Laboratory esperimental work '" on the espulsion of oil from sands through the medium of dissolved-gas drive h a s shown that, whenv the g a s saturation in the sand reaches ap- proxin~ately 10 per cent of the pore volume, the flow of g a s (hence, the gas-oil ratio) increases '

rapidly; and, when the gas saturation reaches approsinlately 20 to 35 per cent of the pore vol- ume, the production of oil becomes negligible. I n other words, the recovery of this type of drive, expressed a s per cent of pore volume, will be, generally, in the neighborhood of 25 per cent. I t is apparent, of course, t h a t the percentage re- covery, in terins of the original oil in place, will vary with the percentage of connate water in the sand. The recovery efficiencies, then, in dissolved- gas drive, depend only on the conlposition and properties of the reservoir and its contained fluids. In the case of gas-cap drive, coupled with gravita- tional segregation, the esamples cited previously illustrate the higher recoveries obtained under this mechanism. To capitalize to the greatest possible extent on this type of drive, i t is often desirable

- . Figures refer to REFEI<ESCES on p. 64.

FRACTION OF ORIGINAL OIL PRODUCED

when, even though the field is produced a t efficient rates, i t will be necessary to resort to pumping production. This will occur in fields where the oil

PRACTICE

to augment the free gas originally present by injection of a t least all of the produced gas, to maintain the pool energy a t a high level. The inherent efficiencies of gravitational segregation and gravity drainage alone a r e graphically illus- trated by the behavior of the Oklahoma City Field, which produces in the Ordivician Wilcos sand. A t the time the pressures had reached 100 psi, only 23 per cent of the original oil in place had been recovered. Although the pressure in this reservoir had been rapidly depleted, the recovery to January 1941 was 35 per cent of the estimated oil in place, and is expected to reach 50 to 55 per cent ulti- mately by primary means. The difference between the ultimate recovery and the recovery a t the time of pressure depletion can be attributed largely to the mechanism of gravitational segregation. In the updip par t of the s t ructure on the east flank of the pool, cores taken in t h a t p a r t of the sand taken over by g a s have measured residual-oil saturations a s low a s 20 per cent. It is to be noted, however, that , in this pool, where the energy was not maintained, as was the case i n the Mile Six Pool, much of the oil to be recovered will be produced a t relatively low rates. Further , the

sand quality in the Wilcos horizon definitely repre- sents the most favorable conditions t h a t could be encountered. I n general, in this type of field, both from an econonlic point of view and from the point of view of recovery efficiencies, the mainte- nance of reservoir energy through the injection of gas is to be preferred.

The low residual saturations obtainable under water drive have been determined by the analysis of cores taken from the Woodbine sand a t E a s t Texas, in tha t p a r t of the producing horizon al- ready flooded out by water. Fur ther testimony to this is obtained from the experience of water-flood ~ p e ~ a t i o n s in which, in the Oklahoma-Kansas shallo\v area, residual saturations slightly below 25 per cent of the total pore volunle actually flooded have been obtained in the Bartlesville sand. Although these figures definitely represent the maximuln efficiency obtainable, a n analysis of other water-drive op&ations definitely establish the superiority of this type of recovery mechanism over sinlple depletion-type operation.

5. In addition to the obvious advantages involved in greater ultimate recoveries, the gas-cap and water- drive mechanisms offer also definite advantages from an operating point of view. I n many cases operations under these kinds of drive can be con- ducted with most, if not all, of the wells in the field producing by natural flow throughout most of their life; whereas under dissolved-gas drive the flowing life of the wells is usually brief. Particularly in water-drive operations there will be some cases

is so greatly undersaturated t h a t there is insuffi- cient energy to flom'the wells. Further , i t i s also to be noted that, when water production in the wells reaches a range from 20 to 50 per cent of the fluid produced, i t will usually be necessary to resort to pumping.

From the foregoing points, the superiority of gas-cap drive and water drive over simple dissolved-gas drive is definitely established. I11 actual practice, of course, many reservoirs a r e found to be operating under more than one , form of drive; however, some one recovery mechanism usually dominates, or can be made to domi- nate, the performance of the field. The major problem, then, tha t confronts the operator and the reservoir engineer when a field is opened to development is to establish what type of drive is present, o r what type of drive is to be utilized in the operation of the reservoir. I t is apparent tha t only when no other alternatives a r e possible will dissolved-gas drive be employed, if the masilnuln efficiency is to be obtained.

Identificatiot~ of the Type of Drive

Often excellent indications of the type of drive t h a t may be anticipated in a given field can be obtained from a study of the geological data tha t a r e available con- cerning the general a rea in which development i s planned. The geologic information required for pre- liminary study directed to the identification of the type, o r types, of drive possible a r e given in the following statements :

1. The nature and extent of the horizon to be exploited should be known. I n this connection, it is desirable to know whether the prospective producing horizon is reasonably continuous over a large area such as, for example, the Smackover lime; or whether i t consists largely of more or less isolated sand bars, such a s the Waltersburg sand in Illinois o r the Bartlesville sand in Oklahoma. From these data i t can be deduced whether there exists a suffi- ciently large body of h a t e r in connection with the oil reservoir to sustain a water drive. I n general, what is required for this purpose is a continuous

,water-bearing horizon, extending from 10 to 20 miles from the edge of the pool.

2. The presence or absence of major faul ts should be noted. I f the accumulation is against a fault, o r . near a major fault, the degree of water drive attainable will be reduced. This will tend to limit the rate of production tha t the water drive can sustain.

3. The continuity of porosity and permeability in the horizon should be determined a s soon a s possible. This information affords fur ther indications of the possibilities of t.he presence or absence of a water drive.

From these prelinlinary data and the subsurface geo- logic data obtaiflable from early wells-which data define the nature and extent of the oil reservoir, the structural relief, the presence or absence of a n original

g a s cap, and the location of the water table, if one is found to be present-an excellent preliminary picture of the possibilities may be set up. This can be well illustrated by the following example. In Fig. 6 is shown a schematic diagram of a n actual reservoir. This oil accuinulation was in a sand bar extending f o r several miles on a northeast-southwest trend. The location of a major faul t running in a north-south direction t o the north and east of the field, a s shown i n the diagram, mas known. A few dry holes i n the shale zone were drilled, defining the edge of the reservoir and also the continuity of the permeability and porosity on either side of the pool. Porous and permeable sand was found only to the southwest of the pool. From these data, then, it can be immediately concluded tha t only a very limited water drive is possible in this field. If, f o r example, the sand had been continuous in all directions around the pool for a distance of several miles, such a s w a s the case in the Dis Pool, previously cited, the possibilities fo r a water drive would have looked favorable.

Imlnediately upon con~pletion of the first well, there should also be made a number of physical measurements which will be of value in making a n engineering study of the pool and in identifying the type of drive to be employed. Included in- these a r e electric logs, core

Scl~elllatic Diagram of Reservoir.

FIG. 6

analyses, and bottom-hole-pressure and bottom-hole- sample data. With these geologic data and the physical measure~nents made during the early development of the pool, i t will usually be possible to tell reasonably early in the life of the pool whether a satisfactory water drive exists. I f a water drive can definitely be ruled out on the early data; the possibilities fo r gas-cap drive can usually be determined a s soon a s sufficient wells have been drilled to define the properties of the reservoir, such a s structural relief, porosity, permeability, and degree of stratification, and the extent of any gas cap found to be present. I n many cases there will be found distinct possibilities fo r the avoidance of operation under dissolved-gas drive. I n general, until such time a s a decision on the type of drive to be employed has been made, conservative producing rates should be em- ployed in order to avoid doing harm to the reservoir through the loss of energy which may be difficult t o retrieve.

Reservoir Control

I t has been pointed out t h a t the first problem in the control of the performance of a new reservoir is the absolute avoidance of the use of a dissolved-gas drive a s the recovery mechanism, if i t is a t all possible. Further , i t has been stressed t h a t conservative produc- ing rates should be employed, until such time a s the field is adequately explored and defined and the type of recovery mechanism to be used is identified.

In a gas-cap drive, usually, the primary problem is the control of gas production. I n this case the r a t e of production must be consistent with the rate a t which effective segregation of the gas and oil can take place. Further , there must be maintained a uniform advance of the gas-oil contact downstructure a s the gas cap es- pands, and the production of gas from the gas cap must be held to a minimum. When pressure is being main- tained by gas injection, the r a t e of withdrawal must be balanced with the rate of injection of gas. Excessive rates of production by individual wells in a field of this kind create high g a s saturations in the sand adjacent to the well bore, and can create serious problems of gas-channeling, particularly in wells in the neighbor- hood of the gas-oil contact. Further , production a t ex- cessive rates in fields producing under this type of drive does not permit taking advantage of the benefits to be obtained from the gravitational segregation of oil and gas in the reservoir, with i ts attendant high liquid saturations and high well productivities in the down- structure par t of the reservoir. I n outline, then, the proper producing procedure fo r a reservoir of this type is the production of oil a t such rates a s to minimize free- gas production and gas-channeling and to maintain a uniform encroachment of the gas-oil contact downstruc- ture, highest degree-of segregation of oil and gas, and a maximum liquid saturation in the producing horizon downstructure. Whenever possible, the return of pro- duced gas to the reservoir should be considered to main- tain reservoir energy, low fluid viscosities, and more uniform producing rates.

In water-drive operation there appears every type of control problem likely to be' found in any reservoir. Pr imary problems encountered in fields of this kind a r e the control of producing rates a t such levels a s to main- tain a balance or a favorable ratio between withdrawals and water influx, the maintenance of a uniform en- croachment of the water, and the control of water and gas production.

In Fig. 7 a r e shown the production statistics fo r the North Searight Pool in Oklahoma, another typical water-drive operation. Production in this field is from the Wilcox sand. Here, through the medium of prora- tion, producing rates have been such a s t o maintain a balance between fluid withdrawals and influx of water into the reservoir, such t h a t pressure may be maintained a t a high level. Moreover, a s will be seen in Fig. 8, there' has been a very uniform movement of the water table upstructure. This feature is desirable in order to at ta in a uniform displacement of oil f rom the sand in t h a t par t of the reservoir flooded by water.

The inlportance of effective control of the ra te of production from a reservoir, particularly in the early stages of development, is graphically illustrated by the behavior of the Elk Basin Field, which produces from the Tensleep sand in II'yoming and Montana. From geologic considerations and from the behavior of another nearby Tensleep reservoir, there is reason to believe t h a t a water drive of some magnitude may be present. How- ever, a t the rates of production which prevailed in the past, there has been a very rapid decline i n pressure; and engineering study has shown that, a t these rates, i t is impossible to tell whether a water drive exists. The calculations shown in Fig. 9 illustrate this problem. I n the study of the behavior of this reservoir, the pressure history to be anticipated was calculated on the basis of two assumptions: first, tha t a water drive exists; and, second, t h a t no water drive exists and the production is due solely to the expansion of oil and i ts contained gas within the boundaries.

Important tools fo r the solution of problems of this kind have been developed by Muskat,"ruce: Schilthuis,' and Katz.TThe most interesting and important feature of the calculations shown in Fig. 9 is that, a t the high ra te of production of 600 bbl per well day, so f a r a s pressure behavior was concerned, there is no distinction between water drive and simple depletion operation. Under these conditions, a satisfactory decision a s to the type of drive possible in this field could not be made and, in the meanwhile, there was occurring a rapid depletion of reservoir energy. I t will be observed that , a t the lower rates of production, there is afforded a n opportunity to make the necessary distinction and also to maintain the energy of the reservoir a t a high level. This latter feature is, of course, of importance, whether a water drive materializes o r not. I f i t should fai l to materialize, this high energy level will be of value, in order t h a t the m a s i m u ~ n benefits of a gas-cap drive may be employed. This reservoir is a steeply folded structure, has good permeability, and the crude oil a t the present pressure levels has a relatively low viscosity.

R . 6 E.

North Searight Pool-Keokuk Area.

FIG. 8

20M I I I O ORIGINAL PRESSURE ESr lUATED

MEASURED PRESSURE - ...... CALCULATED PRESSURE-WATER DRlVE

1900 CALCULATED PRESSURE-NO WATER DPlYE

I I I I I DEC JUN DEC JUN DEC JUN

1943 I 1944 I 1945

Elk Basin Field-Comparative Pressure Behavior.

FIG. 9

Thus, i t is seen that, if the water drive fails to be of a commercial magnitude, there are good possibilities that a form of gas-cap drive, or pressure-maintenance through the medium of gas injection, can be employed to advantage. This pool study presents a striking ex- ample of the alternatives that are presented in the selection of a recovery lnechanis~n when an engineering study of the re'kervoir is made in its early life. In particular, the opportunities for the avoidance of the -

adoption of dissolved-gas drive are stressed. A problem of particular importance in water-drive

reservoirs is the control of water production. This is not alone important from the point of view of individual well performance, but also from the point of view of the performance of the reservoir a s a whole; for, inso- f a r a s behavior of the reservoir is concerned, the with- drawal of a barrel, of water has just a s great a n influence on the pressure behavior and other features of the operation a s the withdrawal of a barrel of oil. There are three methods by means of which the produc- tion of water can be controlled, viz.:

1. Mechanical methods, such a s plugging back and the use o f small-bore pumps to facilitate uniform withdrawals.

2. The injection of water produced back into the horizon from which i t was taken.

3. The control of producing rates.

Inasnluch as the mechanical methods referred to are not prop&ly a part of reservoir engineering, they will not be further discussed here. However, particular at- tention is directed to the inlportance and availability of these techniques for the control of water production.

Attention has been focused recently on the return of produced water to the horizon from which i t was taken. Application of this technique has already been made a t East Texas and, a t the present time, application of this method is under consideration in the Magnolia Pool in Arkansas. The return of unavoidably produced water to the formation can be, under suitable circumstances, a ineans of reducing effective reservoir withdrawals and thus maintaining reservoir pressure-through the medium of which gas is retained in solution, high liquid saturations are maintained in the reservoir, and the flowing life of the wells is extended. The results that may be anticipated a t Magnolia from this kind of con- trol are shown in Fig. 10. The calculated pressure pro- duction behavior is shown for 3 sets of conditions. Curve 1 shows the behavior to be expected if all water production is returned to the reservoir, the gas-oil ratio is maintained constant, and the production rate reduced 4 per cent per year. Curve 2 shows the behavior when all water is returned and the gas-oil ratio and producing rate are maintained constant. Curve 3 shows the ex- pected behavior with no return of water to the forma- tion, maintaining the same producing rate a s in 2. It will be observed that the return of all produced water to the formation, coupled with moderate reductions in producing rate over a period of years, will sustain the flowing life of the wells in this pool indefinitely.

An interesting experiment along this general line is already being conducted in the Midway Pool in Arkan- sas, which also produces from the Smackover lime. A review of this operation has been presented to the American Petroleum Institute by Horner and Snow." In this instance fresh water is being injected into

RESERVOIR

the Smackover through wells on the edge of the pro- ducing zone. This experimental project has shown some degree of success a s a means of controlling the I,ressure behavior of field. hi^ project, however, is to be distinguished froln tha t contemplated for the ~ ~ ~ ~ ~ l i ~ pool. the ~ ~ ~ ~ ~ ~ l i ~ pool there already esists a substantial water drive; injection of water into the ~ ~ ~ ~ ~ ~ k ~ ~ ~ ~ , ill this case, is an ausiliary l,roductioll control, as well as a lnealls of disposal of salt water; the id^^^ pool did not have a water drive collll,arable in with ~ ~ ~ ~ ~ ~ l i ~ , and fresh water was injected primarily to augment the water drive. The interesting feature of this experiment is the practical denlonstration of the fact that it is possible to influence the behavior of the reservoir by the injection of water.

The third lnethod by \\+,ich water I,roduction can be controlled is by the control of the r a t e of production. Much water produced- in water-drive fields, particularly where the structure is rather flat alld where the pro- ducing section is relatively thin, is due to water-coning. The theoretical treatment of this llrobleln, for the case of a holnogelleous has beell gi\,ell by ~ ~ ~ k ~ t ; Most petroleuln reservoirs, of course, are not homo- geneous, but have tight streaks and shale breaks and other discolltinuities distributed more or less at random t]lroughout the The problem is, then, reduced to establishillg lllasilnum differential at which the average wells in field can be produced and not make water. This can be done by test data, o r production data taken from selected wells scattered over the area of the field. These test data, with the tileoretical rates calculated for the same wells on the assunlption of a holnogeneous sand, establish a factor for the relation of theory to performance f o r wells in the proclucing section. This factor can then be ~llultiplied by the theoretical ra te for the average well; i.e., a well having a n average thickness of sectioil above the water table and the average penetration, and thus determine the optimum producing ra te a t which produc- tion of water can be minimized. A similar technique can be used for the control of producing rates when the coning of gas is a problem. It is to be understood, of course, tha t in fields of this kind every effort is made in the con11)letion of wells to take advalltage of every natural barrier to the intrusion of water or gas. The intention of this brief review of the problem has been to demonstrate tha t there a re methods available, or methods which can be adapted, for the solution of the problenl of the control of producing rates in fields where g a s and water production a r e problems.

It is evident in this discussion on the control of reservoir performance that there a r e already available much technical knowledge and esperience which may be brought to bear upon the problem. Further , there a r e many technical developments ltrhich a r e a t present under trial, and many more lying dormant awaiting trial. Considerable advancement has already been made, and a considerable understanding of the problems involved has been attained.

PERFORMANCE 63

Well Spaeinag

Well spacing alone has, in the light of the advances made in the understanding of reservoir performance, heen relegated to a position of secondary importance insofar a s the ultimate recovery to be obtained from a . reservoir is concerned. At the present time attention is directed to the operation of the reservoir a s a whole, with the idea of making i t yield the maximum amount of oil ~~ossible , ra ther than merely attempting to com- 111ete each individual well with the maximum potential

A thorough and complete analysis of field experience has shown that, a t least over the range of well densities e~nployed in this country, i.e., up to 40 acres per well, there is no significant increase in ultimate recovery to be obtained by increasing the well density. A theoretical investigation of this problem by Muskat8 has shown t h a t the physical ultilnate recovery does depend on the number of wells in the case of dissolved- gas-drive operation. With regard to economic ultimate ' recovery, i t was shown that , when the rates of ~ r o d u c - tion a t abandonment a r e low, say 10 bbl per well per day or less, or when the sand is permeable, there was no aljpreciable difference between the physical ulti- mate and the economic ultimate recovery. Although i t

not possible to assign ljractical significance the numerical results obtained in this investigation, the qualitative conclusions drawn a r e believed to be Significant'

A worthwhile observation with respect to reservoir drainage can be made fro111 a study of the behavior of the Dill Pool in Oklahoma, which produced by dissolved- gas drive from the Hunton lime. This field was devel- oped on 4 0 - a ~ ~ spacing. One well, however, the Dill NO. 1, was originally completed in the Crolnwell sand. After depletion of this sand lens, this well was deepened to the Hunton lime. The original bottom-hole pressure in the Hunton lime was 1,725 psi; and the initial pro- duction of the Hunton wells averaged, a f te r acid treat- ment, 1,500 bbl per day. When the Dill No. 1 was completed in the Hunton lime, i t had a n initial produc- tion of 101 bbl of oil per day, and the bottom-hole pressure was found to be 113 psi. This indicated stage of depletion was approxin~ately the same a s that in the offset wells, which had produced tllroughout their entire life from the Hunton lime. That the oil, a s well a s the gas, had been drailled from this area was demonstrated by the fact tha t the gas-oil ratio of this well was ~0111- parable to the gas-oil ratio fo r the renlainder of the field, indicating comparable oil and gas saturations throughout the reservoir. The u l t i~na te recovery to be espected fro111 this well is 40,000 bbl of oil, compared with approsimately 160,000 bbl of oil fo r the offset wells. From these and other similar observations of pools producing under dissolved-gas drive, i t is evident tha t the only co~lsideration in which well spacing is of primary importance is in the proper exploration and definition of the reservoir. Insofar a s the proper deple- tion of the pool is concerned, i t is only necessary t h a t

sufficient wells be drilled to develop adequately the reservoir and obtain production a t economic rates.

The primary problem in successfully draining a reser- voir operating under water drive or gas-cap drive will be the proper location of wells, ra ther than well spacing, a s such. I n general, i t is obvious that, in a gas-cap drive, i t would be desirable to locate most of the pro- ducing wells downstructure-in which direction the ex- panding gas cap would tend to displace the oil; and, conversely, in water-drive fields, development would be concentrated upstructure to avoid unnecessary produc- tion of water and the drilling of unnecessary wells which contribute little o r nothing to the ultimate re- covery of the field. A s a purely practical matter, how- ever, wells a r e usually drilled throughout the field on a somewhat regular pattern, and a r e not confined t o tha t portion of the reservoir where they would be most

- effective. The number of wells required in a water-drive or

. gas-cap-drive field is t h a t which will permit the proper control of the movement of gas and water to accomplish uniform displacement. Further , sufficient 'wells should be drilled so that, w11en the field is produced a t a n efficient rate, the production rate of individual wells should not be excessive; i.e., individual wells should not be produced a t such rates tha t there is a n excessive production of f ree g a s o r water, o r t h a t the rate of depletion of the sand in the immediate vicinity of the wells is materially greater than in the remainder of the reservoir. A well, o r group of wells, operating a t ex- cessive rates in a reservoir of this kind can create a localized dissolved-gas-drive condition which could affect the ultimate recovery from the reservoir a s a whole, and this is a condition to be avoided. The particular spacing to be used, then, in fields of these kinds, can only be properly determined a t such a time a s a n effi- cient rate fo r the reservoir a s a whole can be estab- lished. The implication of this statement is tha t i t is prudent to develop a field on a nloderately wide spacing until such time a s the reservoir is defined and the proper' producing ra te determined. I f , then, fur ther develop- ment is indicated, in order t h a t proper control of the reservoir can be maintained, such additional wells a s may be necessary can be drilled. By a procedure or plan of development of this kind, i t will be possible to obtain the maximum recovery from petroleunl reservoirs and to maintain efficient and economical operation.

Discussion

In the foregoing sections, the fundamental features of petroleunl reservoir performance have been outlined and illustrated. Attention has been directed to the importance of making a thorough engineering analysis of the reservoir in the early stages of its development, and reference has been made to a few of the important theoretical and experimental studies which, together with the experience of the past, form the. basis of the present knowledge of reservoir engineering.

Although the details of the analysis of any specific reservoir a r e often somewhat difficult-particularly

when i t is considered that, due to the vagaries of nature, each reservoir presents new problems-the basic prin- ciples a r e essentially simple and of considerable gener- ality. Variations in the nature of the reservoir appear to modify only the details of i ts history, rather than t h e fundamental character. The fundamental features of the performance are, in a large measure, in the hands of the operator and the engineer. A s has been shown, i t is possible to determine in advance or a t a n early stage of production in many, if not in most, fields just what type of recovery mechanism is to be ,utilized. Accordingly, the broad general outlines of the history of the operation can be laid out. Thereafter, wha t is required is the accumulation of accurate measurements of the properties of the system, and accurate records of the production and performance, in order to maintain

' a detailed and uniform control of the operation. It is to be stressed t h a t efficient operation of the reservoir means proper control of the entire reservoir, and not merely par t s of it. A s has been stated by S. E. Buckley," good practice on one lease and bad practice on another do not necessarily average out to give a reasonably good average performance.

Although there a r e unquestionably large gaps in our knowledge of the details of reservoir performance, a small capital of knowledge has been accumulated over the past decade which, if intelligently and diligently applied, will serve to inlprove our understanding of these problems in the future, and will be of economic advantage to all concerned in the development of the petroleum industry. Intensive research is being con- tinuously carried on to fur ther improve the methods of control and yield of oil reservoirs.

ACKNOWLEDGMENT

The author is particularly indebted to the " ~ o i n t Progress Report on Reservoir Efficiency and Well Spac- ing" by the Committees on Reservoir Development and Operation of the Standard Oil Company (New Jersey) Affiliated Companies and the Humble Oil and Refining Company, which was freely drawn upon in the prepara- tion of this paper. Thanks a re also due to the manage- ment of The Carter Oil Company for permission to publish the paper.

REFERENCES .

JI. C. Leverett a n d W. B. Lewis. "Ste;dy Flow of Gas-Oil J I is turcs through UnconsoliBate(1 Sands, Trans. Am. I t ~ s t . iUi#lil#g ;vet. f?,rnra la2 l n 7 i l ! U l i

JI. l\Iuska Co.. New Tor

3 nr. A. Brll d111. I~ rx t . i1Ii1ri1rg Xc t . E~ ig r s . 151,

R. J. Sc l i i l t l~u~r . "Active Oil nl d?n. Ins t . .IIi)li~zg Met. E~rg r s . 118, 83 (

Q. L. Katz. "A RIethotl of Estirnatin

>..>.. ". A--, -". .-..--,. t, Flow of Ho~aoyc~ i ro r t s Flrtitls, IIcGraw-Hill Book k (1937) . ce, "Pressrcre Pre(licti6n fo r Oil Reservoirs," T~CIILS.

73 (1943 ).. 111 K e s e r r ~ ~ i r Energy," Trcr~re.

1036) . a Oil aud Gas Reserves."

9111. Inat. iliin'i~tg Nct . Engrs. 118, 2G i1936). L. Horlier and D. R. Snow. "A New Application of Wate r

Injection fo r JInintaiuing Reservoir Pressure aud Increasing Natilral Water Drive." Urillirig n ~ l d Prodlcction Prncticc, 2S ( l C l i R i , . -. . , . ' II. JIuskat. "An Al~proximnte Tl?epry of Wate r Coning i n Oil Pro(111ction." T ra~ l s . Am. Irtst. ~ U r n r ~ r a -1Iet. E ~ l o r s . 114. 144

DISCUSSION

J. J. Arps (British American Oil Producing Com- pany, Tulsa, Okla.) : I should like to ask whether re- covery percentages here mentioned for the various types of drive were expressed a s a percentage of the original residual oil, o r the original reservoir oil?

Mr. Mullane: In all instances percentages given were sub~nit ted a s percentages of the original residual oil.

E. A. Stephenson (Department of Petroleum Engi- neering, University of Kansas, Lawrence, Kans.) : When a dissolved-gas drive fails, which sometimes occurs, why not inject water, o r s ta r t a water flood a s soon a s such failure is demonstrated? Water flooding has been and is successful, and why not utilize the water a s soon a s the basic laboratory work indicates its feasi- bility? By so doing, the ultimate recovery should be greatly increased over tha t due to dissolved-gas drive.

Mr. Mullane: I regret to say t h a t I do not have a n intimate knowledge of water flooding. However, if a n operation has produced a pool by g a s drive, and one has developed the pool in a n economic manner, and the recov- ery has paid out and made a profit, then any additional recovery obtainable by water drive nlay be profitable. I f , however, the sand thickness, the kind of oil, and t h e other factors t h a t go to make a successful flood have been misinterpreted, and we put water into the horizon a t a n early stage, then the operator may encounter difficulties. There is always the possibility tha t the sand nlay not flood successfully. I have in mind the Oklahoma City Field, where I a m told t h a t water encroachment flooded out something like 30 or 40 wells. So one might actually figure on the fact tha t we know so little about the factors tha t control the thing that, if we wait until later in the game, we cannot possibly get hurt.

George R. Elliott '(Phillips Petroleum Company, Bartlesville, Okla.) : I should like to compliment Dr. Mullane on his paper, and also his company on i ts publication, which has given to us a n excellent summary on which y e can base fu ture studies on the behavior of reservoirs.

A. W. Walker (Stanolind Oil and Gas Company, Tulsa, Okla.) : Dr. Mullane has discussed very nicely the question of reservoir performance, o r reservoir mechanisms. Where the reservoir performance begins and mechanisms leave off is a n extremely difficult point to consider., In other words, all the pressure differ- entials causing fluid movement a r e due to pressures within the reservoir itself, and a t the well bore. Now whether this pressure difference be due to encroach- ment of water, o r to gas expansion, o r to gravity, the net result is tha t fluids move toward the well; but a t the same time there must be a displacement. I think

the mechanism which is directly related to the displace- ment medium (whether i t be water o r gas) must be taken into consideration.

There a r e a few points which I should like to discuss, such a s whether the reservoir is under water drive, o r gas cap, o r gas gravity, o r whatever you might choose to call it. Conditions which Dr. Mullane has pointed out a r e recognized primarily a s ideal cases. Actually, we find few cases where fields a r e under one type of operation; and very likely there a r e fields which will change from water drive6to gas expansion, o r vice versa, depending upon operating practices. I n a great many cases we have larger fields wvllere all three types a r e silnultaneously in effect. So in reservoir performance i e have to take into account the degree of the ~nechanisnl and how the production of the fluid is affected.

There is one point I should like to bring out. I think there is quite a bit of confusion regarding the question of "solution-gas drive." No doubt Dr. Mullane and his associates will readily agree that gas does no driving while in solution, only a f te r coming out of solution. In other words, the drive is the pressure exerted by the expansion of gas, a f te r i t has been liberated from the oil and is more or less uniforn~ly distributed throughout the reservoir.

Getting to the question of gas-cap drive, the ideal case presented here is extremely interesting, and probably is the first o r only one of which we have a complete history. Actually, in operation we find in this country t h a t we do not have t rue gas-cap drive; i.e., if we have enough gas expansion or liberation, we have a large amount of f ree g a s with the oil, and then we have increasing gas-oil ratios with a limited accumulation of gas in the reservoir-whether it be on top or in snlall localized traps. Actually the mechanism of the gas-cap drive is essentially gravity drainage. In other words, with a complete segregation and stratification of the gas and the oil, i t is the force of the gravity tha t causes the stratification and brings the oil to the well bore, although i t is operating under a high static pressure.

So f a r a s gravity drainage is concerned, we have heard a g rea t deal about i t lately. For a good many years it was more or less neglected, and then i t came up again, af ter the Oklahoma City and other fields seemed to eshibit special characteristics. However, there is a tendency in our production practice to go in cycles, and I think there is a tendency now to over- emphasize gravity drainage. We know i t exists in all things 'and a t all times. I n certain cases where condi- tions a r e ideal, i t may be the predominant influence. As I see it, in order to have t rue gravity drainage, we

'

must have a relatively thick uniform sand section, with very little stratification; therefore, whether we like i t o r not, we do not find i t very frecluently. Those two factors necessary for gravity drainage apparently were present in the cases which have been noted recently.