- - ~ -
r
(j
()
Ninth Printing
1981
During the past twenty years petroleum engi- behavior of hydrocarbo n systems, sampling of oil
neers have developed many specialized methods of reservoirs, and material balance calculations.
calculating the pressure-pr oduction characteristics Conflict in the nomenclature of fluid flow and
of oil and gas reservoirs. Part of the basic informa- thermodyn amics was encountered early in the prep-
tion required in these calculations pertains to the aration of the book. As a result, it was necessary to
physical properties of the reservoir fluids and to the readjust some of the symbols from their usual mean-
manner in which the properties change as the reser- ing in order not to have duplication. A complete
voir pressure changes. Naturally, many relevant list of the nomenclatu re and definition of terms used
papers have appeared in the technical literature. in this book follows Chapter 8.
Some have pointed out the manner by which the
fluid properties have been measured in the labora- The author has attempted to present as maQ
tory. Others have pointed out the manner by charts as possible to assist in reservoir engineering
which the fluid property data have been used to calculations. In this connection, the Appendix
calculate reservoir behavior. Yet, only a few contains three calculating charts that are helpful
authors have coordinated the two and, what is in estimating properties of naturally occurring
most important, have shown how the measured hydrocarbo n systems at reservoir conditions from
laboratory data can be altered to fit specific reser- the field parameters of gas-oil ratio, gas gravity,
voir conditions. and oil gravity. While the data from which the
charts were constructed are exclusively "Cali-
The types of hydrocarbo n systems encountered in
fornia data," there are no obvious reasons why the
oil field operations range from almost pure methane
charts will not apply to systems from other locations.
gas to heavy bituminous material having the nature
of road tar or asphalt. In an effort to systematize The author wishes to acknowledge the kind inter-
the discussion of the phase behavior of these com- est of the Manageme nt of the California Research
plex systems, they have been divided into three Corporatio n and others during the processing of this
categories: (a) gases, (b) condensate systems, and book. Mr. A. L. Vitter, Jr., of The California Com-
(c) dissolved gas systems. Just where the dividing pany and Messrs. R. L. Parsons and R. E. Loecu
line between gases and condensate systems lies is California Research Corporatio n acted as a review-
difficult to say, although condensate systems and ing committee during the preparation of the book
dissolved gas systems can be differentiated easily and made valuable suggestions toward the final
by the critical temperatur e of the system. Sup- form. Mrs. B. Hansen, Mrs. V. Cox, and Mrs.
porting framework to the discussion of these three E. K. LaRue were particularl y helpful in preparing
systems is given in the chapters on general phase charts and in proofreading.
M. B. STANDIN G
July, 1952
La Habra, California
Contents
Preface iii
u
CHAPTER 1
Hydrocarbon systems found in petroleum reser- temperature can be defined as the temperature
voirs are known to exhibit multiphase behavior at above which two phases cannot exist, regardless of
temperatures as low as -250F and as high as the pressure upon the system. Likewise, the critical
1000F. Many systems exist as two phases at pres- pressure is that pressure above which vapor and
sures as high as 15,000 psia. Because the molecular liquid cannot exist in equilibrium.
composition of the gas and liquid phases in equi-
80
librium with each other in a reservoir rock is dif-
ferent, the physical properties of the phases differ.
This difference in phase properties, coupled with the
ability of reservoir rock to transmit gas and liquid oC
700
with different degrees of ease, causes the complex
behavior of oil reservoirs.
A few of the basic concepts of phase behavior and
L/'11110 I
fluid flow in porous media will be presented in this 0
/
chapter. The phase behavior discussions will start
with the behavior of a pure compound and will pro-
ceed from there to discussions of binary and multi-
component system behavior. Following the phase 500
I VAPOR
To illustrate the use of Figures 1 and 2, let us The behavior of other pure compounds is similar
assume that one pound of ethane is contained in 0.25 to that outlined for ethane. A vapor pressure chart
cu ft of space at 60F. At the specified tempera- for the lighter hydrocarbons is shown in Figure 3.
ture and volume conditions, ethane will be in vapor Table 1 lists important physical properties of the
form and exert a pressure of 465 psia. This is light paraffin hydrocarbons, as well as those of several
shown in Figure 2. When the ethane vapor is com- miscellaneous compounds.
pressed at a constant temperature of 60.F, liquid
will condense first at 495 psia. Point B, the state Binary Systems
at which the system is entirely in the vapor form The phase behavior of two-component systems is
except for an infinitesimal amount of equilibrium slightly more complex than that of pure compounds.
liquid, is, by definition, the dew point. Further The main differences are: (1) the pressure-tempera-
900..---r-r-.--...,.....--r---,---,
BOOl---1-+----1----".---l----+---I
a-- l'r--a
8001---+----h,l---'r+--~ u
.
SINGLE
PHASE!?
REGION j
~
II)
~ 600 ,__ _,____ t1,L-1------1-+-----1
I
SINGLE
I
I
TWO PHASE
... PHASE
RE6/0N
~
LIQUID I
I REGION
/
Figure 2-Pressure-Volume Diagram of Ethane
I
900
800, 800
'
Tool 700
6,
z
.....
!'-
I
....I 0"'
... 0
....~
C
~
("\
.....
"I
.... 0
"I
"'
-<
It:
~ ...
It:
("\
'
0
....~ ....... z
,,
("\
m
.....
Vl
C
Figure 3-Vapor Pressure Chart for Hydrocarbons
...
C
....
TABLE I
NNONAN E 128.3 303.4 335 I 073 o. 7211 45.02 337. 95 I 7 .78 15.11 <:)
NDECANE 14 2.3 345.2 312 I 11 5 0.7333 44.78 374.91 11 .30 14.30 z
(-- -
'
INTRODUCTORY CONCEPTS 5
temperature diagram takes the form of an envelope dentherm, and maximum pressure value are func-
rather than thil" single vapor pressure curve of pure tions of the composition of the particular mixture.
compounds. The area within the envelope defines For example, in the ethane-n-heptane system the
the temperature and pressure range wherein vapor variables of pressure, temperature, and composition
and liquid can exist in equilibrium. Outside the are as shown in Figure 5. In Figure 4, it will be
envelope only one phase can exist. The dashed noted that the critical state of the 96.83 mole per
line aa is the maximum pressure at which two phases cent ethane mixture is well to the left and below the
can exist. The line bb, defining the maximum tem- maximum two-phase pressure. In Figure 5, it can
perature at which two phases can exist in equi- be seen that as the per cent of n-heptane in the
librium, is called the "cricondentherm." mixture increases, the critical state moves toward
the right side of the phase envelope, until at 75
mole per cent n-heptane it lies well to the right and
below the maximum pressure.
Multicomponent Systems
The phase behavior of multicomponent systems is
much the same as that of the ethane-n-heptaneo
system. As will be shown later, knowledge of the
critical temperature and pressure of multicompo-
nent systems is useful in describing and correlating
the phase behavior of such systems. As a matter
of fact, from. the. critical temperature alone, one
can make reasonable estimates regarding PVT
behavior of any reservoir mixture.
- - - - CRITICAL L OCIJS As an example of how knowledge of the critical
Figure Ci-Pressure-Temperature-Composition Diagram of the
state can be used, let us assume the particular mix-
Ethane-n-Heptane System ture of natural gas and gasoline studied by Katz and
Kurata 2 to be representative of the hydrocarbon
Point C in Figure 4 is the critical pressure and system existing in a particular reservoir,: For the
temperature of this particular mixture. It is ap- moment we shall not consider how we know it
parent from inspection of Figure 4 that the defini- represents the reservoir material, but will assume
tion of the critical temperature and pressure used that it does, and that we have a small quantity of
for a pure compound cannot apply to the binary the mixture isolated in a laboratory pressure cell in
mixture, as multiple phases exist at pressures and which we can observe the phase behavior at various
temperatures greater than the critical values. A pressures and temperatures. This behavior i\.)
better general definition of the critical state is that shown in Figure 6.
state at which all intensive* properties of the liquid The first thing to be noted is that the general
and vapor phase become equal. This definition shape of the two-phase envelope in Figure 6 is simi-
applies equally well to pure compounds, but is lar to the envelopes exhibited by binary systems.
usually neglected in favor of the simpler definition. As in binary systems, the loci of bubble points and
The critical pressure, critical temperature, cricon- dew points meet and join as a smooth curve at the
critical state. In fact, the only major difference
"'Intensive properties are those which are independent of the between the natural gas-natural gasoline and ethane-
amount of material; for example, density, pressure, temperature,
viscosity, chemical potential, and surface tension. Extensive
n-heptane phase diagrams is the higher critical
properties depend on the amount of material; for example, area, pressure exhibited by the multicomponent system.
weight, inertia, and volume. The product of an intensive and In addition to assuming that the data shown in
extensive property is a measure of energy; for example, pressure X
volume = mechanical energy. Figure 6 represent the phase behavior of the reser-
6 BEHAVIOR OF OIL FIELD HYDROCARBON SYSTEMS
voir fluid, let us also assume original reservoir con- An easy way of thinking of retrograde behavior is
ditions of 3000 psia and 190F. Referring to to use the word retrograde as meaning "in opposite
Figure 6; it would be expected that when the reser- direction to that usually observed." Thus, the for-
voir pressure had fallen to 2690 psia, a liquid mation and growth of a liquid phase as a result of
phase would appear throughout the reservoir. The isobaric increase in temperature are opposite from
liquid would continue to increase to a maximum the usual behavior, as pure compounds always
vaporize (decrease liquid phase) upon isobaric
increase in temperature. Likewise, our knowledge
of the behavior of pure compounds indicates that
an increase in pressure, at constant temperature,
will result in condensation of vapor into liquid
(Figure 2); therefore, the opposite beha Yior of
binary and multicomponent systems causing a liquid
phase to decrease in extent while undergoing an in-
crease in pressure is called a retrograde phenomenon.
Referring again to Figure 6, it will be noted that
had the reservoir temperature been 70F rather than
190F, a bubble point, rather than a dell" poiry 1
would have been reached at 2690.psia. This ,rnuh:r..,
have resulted in the appearance of gas in a sand
saturated with liquid, rather than the appearance
of liquid in a sand saturated with vapor. Conse-
quently, one would expect some difference in the
pressure-production behavior of the resenoir and
some difference in the manner by which the reser-
voir should be produced to obtain the maximum
recovery of hydrocarbons.
TEMPERATURE-F Two important facts are to be learned from the
Figure 6-Phase Behavior of a Natural Gas-Natural Gasoline above:
Mixture (After Katz and Kurata, Ind. Eng. Chem.,
1940)
(1) The composition of a hydrocarbon system
of 11 per cent at 2200 psia, after which further defines the critical pressure and temperature
pressure reduction would result in the revaporiza- of the system. (See Figure 5.) As the pro-
tion of the liquid phase.* portion of light hydrocarbons (methane,
The formation of liquid from a vapor as a result of ethane) in the system increases, the critical
decrease in pressure is termed "retrograde condensa-
tion." The opposite process of vaporization of a
temperature of the system decreases.
u
(2) The relation of the critical temperature of the
liquid as a result of an increase in pressure is called
hydrocarbon system tothe reservoir tempera-
"retrograde vaporization." There can exist, of
ture will determine ,vhether retrograde be-
course, the comparable isobaric processes of retro- havior will be encountered. Whether the
grade vaporization and condensation as, for exam- system exhibits a dew point or a bubble
ple, at pressures above 27,50 psia (the critical point at reservoir temperature determines
pressure) in Figure 6. the flow mechanism in the reservoir sand.
"'The data of Figure 6 were determined for a condition of con-
stant system composition; i.e., the reduction of pressure was Rigid application of these principles to oil reser-
obtained by increasing the volume of the container in which the
system resided. This obviously docs not conform to reservoir voir systems is difficult, as few data are available
behavior wherein the pressure reduction is a result of removing regarding the critical temperature of crude oil-
material from the system (variable composition). However, the
assumed process is satisfactory for the present discussion. natural gas systems as a function of composition.
INTRODUCTORY CONCEPTS 7
The fact that oil field usage recognizes only two where q,, q,, q,, are volumetric rates of flow of gas,
components, oil and gas, each of which is a multi- oil, and water at reservoir conditions; k., k,, kw are
component system, makes generalizations difficult; effective permeabilities to the three phases; 0 , ,, w
nevertheless, the following qualitative statements are viscosities of the three phases at reservoir con-
can be made: ditions; and (dP/dx)., (dP/dx)o, (dP/dx)w are pres-
sure gradients within the three phases.
( 1) Systems existing in one phase and having gas- A reasonable representation of the relative flow
oil ratios greater than 10,000 cu ft per bbl are rates of gas phase and liquid oil phase can be made
likely to have high concentrations of methane by neglecting the behavior of the water phase and
and the other light hydrocarbon components. assuming equal pressure gradients in the two hydro-
Therefore, the critical temperature of the sys- carbon phases. Dividing the first of Equation (1)
tems will be low (perhaps of the order of that by the second, and rearranging gives,
of methane, -116F). Consequently, at
reservoir temp'eratures. of 150 to 300F, one qg kg o
may expect that such a system will exhibit -
q.,
<X -
ko
-
/J.u
(2)
isothermal retrograde phenomena.
Equation (2) expresses the relative flow rates of
(2) Systems existing in one phase and having gas and liquid oil phases under reservoir conditions
gas-oil ratios less than 2000 cu ft per bbl are
likely to have critical temperatures higher
of pressure, temperature, and fluid. sa;turation. U
Two generalizations can be made, even though
than the reservoir temperature. These sys-
phase properties and reservoir rock properties vary_
tems will have bubble points, and initial pro-
duction will be by dissolved gas drive. between reservoirs and between areas within: a : .
single reservoir.
(3) Systems in the intermediate range of gas-oil
ratio may exhibit either bubble point or dew (1) The viscosity ratio, ,/,,is generally greater
point behavior. It is necessary to make lab- than 10 and can be as high as 106 This is be-
oratory tests to determine the actual be- cause reservoir gas viscosities are of the order
havior of the system at reservoir conditions. of 0.01 to 0.07 centipoise 3 and oil phase vis-
cosities vary from 0.2 to 10,000 ceritipoises. 4
Multiphase Fluid Flow
(2) The permeability ratio, k, /k,, may vary from
When pores in reservoir rocks contain more than zero to infinity because, over par.ticular ranges
one fluid, the permeability of the rock is different to of gas and liquid oil saturation, the permea-
each fluid. The permeability to water or gas that a bility to either phase may be zero.
given rock matrix will have is primarily a function The gas and oil permeal;,ility curves of a Gulf
of the water or gas saturation within the rock Coast sandstone shown in Figure 7 are typical of
pores. The permeability of the rock to oil is data that have been determined for a large number ..
affected by the fraction of the pore occupied by oil of sandstones. The curves indicate the generalU
(oil saturation), as well as by the relative propor- ranges of fluid saturation wherein the permeability
tions of water and gas present. When pressure to one or the other of the hydrocarbon phases can
gradients exist in the fluids, the rates of flow of the be zero.
individual phases will be as follows: Let us speculate on the probable phase and fluid
q
II
k,, (dP)
o:-.
fl
-
dX O
flow behavior in the region of a well, using in Equa-
tion (2) the information given by Figure 7. Let us
110:-.-
" ,,
k,, (dP)
dx ,,
(I) consider also the phase behavior of the original
reservoir fluid to be represented by the data of
q,.. k,, (dP)
o:;-.
u
-d
X w
Figure 6, first taking the case of the reservoir fluid
originally existing at 2800 psia and 70F, and sec-
* This statement assumes that water is the fluid wetting the ond, the fluid originally existing at 2800 psia and
rock material and that gas is the fluid residing near the center of
the pore. Liquid oil would be the intermediate phase. 190F.
8 BEHAVIOR OF OIL FIELD HYDROCARBON SYSTEMS
(4) During the time the gas saturati on is increas-
ing in the region of the well bore, there will be
some impairm ent of the well's capacity to
produce oil because of the decrease in liquid
phase permeab ility.
zero. Thus, the 7.1 per cent condensate measured the phase volumes at several indicated
formed by retrograde action will be an im- pressures and temperatures. Because the com-
mobile phase and will be lost insofar as flow position of the system was not changed during the
into the well is concerned. However, this tests, the results are referred to as flash equilibrium
statement must be qualified as covering the data. An example of a dynamic flash equilibrium
behavior occurring in the major portion of process is the separation of gas and oil in a field
the reservoir; the effect of converging flow
separator or trap. Over relatively long time periods
lines in the area adjacent to the well will
make it possible for some condensate to flow the composition of the fluid entering the trap can
into the well bore. be considered to be constant. Consequently, the
composition and proportions of gas and liquid
Flash and Differential Equilibrium Processes phases in the trap at any specified time are func-
tions of only pressure, temperature, and the over-all
In the discussions to this point, two concepts have system composition.
been implied as to the type of equilibrium process An example of a differential process can be visual-
that occurs between hydrocarbon phases within the ized by using Figure 6 as a reference. Suppose the
reservoir matrix. The data shown in Figure 6, system studied was originally at 2800 psia and
which have been used to picture the phase behavior 70F. Reduction in pressure to 2690 psia would
of fluids within reservoirs, were determined for a
system of constant composition. In contrast to
cause the system to be at its bubble point condi- Q
tion; further reduction to 2689 psia would result in
the constant system composition viewpoint, Equa- equilibrium gas and liquid phases. If it were pos-
tion (2) and the data of Figure 7 indicate that in the sible to remove the free gas phase from the cell
great majority of cases the hydrocarbon liquid and without disturbing the liquid phase, a new system
gas phases in a reservoir move at different rates. having a bubble point of 2689 psia would remain.
Because the gas and liquid phases in a multicom- Continued stepwise removal of small increments of
ponent system are always of different composition, equilibrium gas would eventually bring the cell to
the preferential movement of any one phase across atmospheric pressure via a series of systems having
the boundary of the system (reservoir) always different compositions and bubble point pressures.
results in a change in over-all composition of the To summarize the difference between flash and
remaining material. The two processes have been differential processes, it might be said that the
termed the "flash" process and the "differential" latter requires a knowledge of the exact route by
process. A flash process is one in which the over-all which the process has progressed to a certain point,
composition of a system remains constant; a differ- whereas the history of past events is immaterial in
ential process is one in which the over-all composi- a flash process. In a flash process it is sufficient
tion of a system varies as the process proceeds. The to state the original system composition and the
difference between these processes can be made desired pressure and temperature conditions; in a
clear by an example of.each. differential process the manner by which the originalU
In obtaining the data of Figure 6, Katz and system degenerates to the desired pressure and
Kurata made use of a variable volume PVT cell and temperature also must be defined.
Except at the critical state. Composition is one of the inten-
More will be said about flash and differential
sive factors in which the vapor and liquid are equal at the critical. processes in subsequent chapters.
CHAPTER 2
The object of sampling is to obtain oil and gas for (1) Completion of the well may result in simul-
laboratory- studies of physical and chemical prop- taneous production of fluids into the tubing
erties of fluids present in underground reservoirs. from two or more zones. Thus, the sample
Whereas the desired type of result determines obtained may be an unknown mixture of two
methods of performing the laboratory tests, the fluids and will not be representative of either.
characteristics of the reservoir fluids and the
mechanical equipment used in completing the well (2) Where both liquid and gas phases exist in int9
mate contact in the same zone, the flow rates
determine the method to be used in obtaining th~ of the two phases into the well bore are almost
samples. Needless to say, obtaining a proper always different. Consequently, it is for-
sample is as important as subsequent laboratory tuitous if the flow rates are such as to cause
tests; yet few engineers understand the advantages the tubing fluid to have the same composition
and limitations of the several methods that are as the reservoir fluid as a whole.
commonly used in sampling.
Before discussing their merits and limitations it The third factor involves the accuracy of sampl_ing
may be well to consider several general facts in- the fluid flowing in the tubing. While the field engi-
volved in getting samples of oil and gas that are neer has little control over the first two factors, he
representative of reservoir fluids. In the first can do something about the third.
place, there is no assurance that any sample The greatest use of PVT data lies in. calculating
obtained from one well Is representative of the reserves of oil and gas in place in the reservoir and
fluid throughout the reservoir. Theoretically, the the effect of field operational methods on the recov-
effect of the earth's gravity field should cause differ- ery of oil and gas. Both calculations make use of
ences in composition of oil lying at different eleva- the change with pressure of the physical properties
tions within a reservoir. Also, the reservoir fluid of the reservoir fluid. Because it is difficult t-"'
may vary in composition between locations having evaluate the over-all composition and physiJ.J
the same structural elevation because of movement properties of a reservoir fluid (except possibly in the
of the rock strata comprising the reservoir during case of an under-saturated fluid), most PVT data
geologic time. Both kinds of compositional varia- are designed purposely to. give only the properties
tion have been observed in oil fields. of the individual phases. In fact, many tests may
The second factor involves the question: To what. not evaluate directly the properties of the indi-
extent is the fluid in the tubing representative of the vidual phases believed to be in the reservoir, but
reservoir fluid in the region of the well being sam- give data from which the desired property can be
pled? This factor is important because all the obtained by extrapolation. In either event, the
methods make use, either directly or indirectly, of over-all properties of the reservoir fluid are calcu-
fluids obtained from the well tubing. Two reasons lated from the individual phase properties and an
why the fluid flowing in the tubing might not be estimate of the relative phase volumes present in
representative of the reservoir fluid are: the reservoir rock. The best possible estimates of
SAMPLING METHODS AND APPARATUS 11
phase properties during the entire life of a reservoir point of both dissolved gas and condensate system><.
calls for the testing of more than one sample. This Flowline samples will he considered entirely in con-
is brought out by the following three points: nection "'ith condensate systems.
(1) .-\ sufficient number of samples should be
Bottom-hole Sampling Equipment
taken to give the properties of the fluids in
different parts of the reservoir. It may not be Most samplers are designed to he run inside 2-inch
necessar, to run complete PVT analyses on tubing on a steel wire of the type used in the opera-
all samples; but key determinations, together tion of pressure recorders. The samplers are of the
with producing gas-oil ratio surveys, tem- order of 6 feet in length and can hold approximately
perature surveys, tank oil gravity surveys, 600 cc of fluid.
and, in the case of an undersaturated fluid
In order to get a sample of the fluids near the bot-
re:-:ervoir, a ~aturation pressure survey, can
be used to define the "average" fluid tom of the well, a trip valve mechanism is provided
properties. for opening and closing the fluid ports of the
sampler. Samplers differ mainly in the type of
(2) Samples should be taken during the early life valve mechanism used for trapping the sample.
of the reservoir. It is not too difficult to esti- The Humble Type A sampler operates when a
mate the properties of the original reservoir sudden movement of the sampler causes a suspended
fluid from samples taken while the reservoir weight to drive a hardened steel point through a
pressure is still above 80 per cent of its original thin steel disk, allowing the fluid in the well to
V
value. It is almost impossible to do so accu-
enter the previously evacuated sample chamber.
rately with samples taken after the reservoir
pressure has declined to 40 per cent of its The Humble Type B sampler also allows well
original value. fluids to enter a previously evacuated chamber, but
the operation is carried out by the temperature
(3) Samples should be re-run after the reservoir and pressure at the sampling depth, which cause
pressure has changed materially. Labora- rupture of a tin disk. In both types of Humble
tory tests are made on the basis of specific samplers a second check valve seals the fluid entry
test processes and the assumption that
port when the sampler is withdrawn from the well.
thermodynamic equilibrium exists between
phases at all times. The estimation of the The Bureau of Mines, the Perco, and the Wofford
fluid properties in a reservoir at, say, 1000 psi type samplers differ from the Humble samplers in
can be made more accurately from samples that a port in each end of the sampler remains open
obtained when the reservoir is near 1000 psi while the sampler is lowered into the well.* Thus,
than from samples taken when the reservoir fluid can flow through the sampler while it is being
pressure is 3000 psi. run into the well and until the fluid ports are caused
to close. The valves closing the fluid ports can be
Three types of samples are commonly used in de-
set in operation by either a clock-electromagnetic
termining properties of reservoir fluids: first, bot- 1 1
time latch or a shear pin arrangement that operates V
tom-hole samples from near the bottom of the
when a sudden movement is imparted to the sam-
tubing, taken either while the well is flowing at a
pler. Clock-operated mechanisms seem to he the
slow rate or shut in; second, recombined samples of
most successful, the chief disadvantage being that
the liquid and gas from a surface trap; and third,
the clock almost doubles the cost of the unit.
surface flow-line samples ordinarily obtained at the
From the above brief descriptions, it can be ap-
well tubing head. The remainder of this chapter
preciated that a bottom-hole sampler might fail to
will touch on several points in favor of, or limiting,
operate for one or more of the following reasons:
the accuracy of these three types of samples and
will outline methods by which the samples are "' More detailed description of the Humble type sampler has
taken. The discussion of bottom-hole sampling been published by Schilthuis. 6 Grandstone and Cook 6 have
methods pertains to dissolved gas systems; recom- described the Bureau of Mines sampler. Catalogs of the Engineer-
ing Laboratories, Inc. describe the Perea and Wofford type
bination samples will be discussed from the view- samplers.
12 BEHA YIOR OF OIL FIELD HYDROCARBON SYSTEMS
(1) The thermo-pressure disk type may not rup- reasonable to assume that the proportion of gas
ture, or may rupture prematurely if the disk is and liquid in the sampler is unknown at the instant
not of the proper thickness for the pressure the valve mechanism operates to close the sampler.
and temperature conditions in the well. Consequently, many bottom-hole samples should
(2) Mechanisms operating on the principle of a not be considered as representative of any specific
sudden movement of the sampler may func- fluid in the well or the reservoir. Rather, they are
tion prematurely when running the sampler better considered to be only "samples" that have
to depth. This often is caused by striking a limited chance of being a replica of the reservoir
a liquid surface or running into a paraffin fluid.
accumulation in the tubing. Fortunately, one thing can be done toward limit-
(3) Sudden movement mechanisms may fail to ing the range of variation of fluid composition
operate when used in wells containing a high obtained in the sampler. By sampling the well
viscosity oil or in deep wells. This is because shortly after it. has been shut in, there exists a
sufficient movement cannot be imparted to smaller chance of forming the sample from inde-
the sampler by jerking the supporting line at terminate amounts of gas and liquid. If low pres-
the surface. sure draw-down conditions are established in the
well prior to sampling, and if after shutting in the
In addition to the bottom-hole sampler, con- well sufficient time elapses for any free gas to segre( )
siderable surface equipment is necessary to make gate in the top of the tubing, the sampler usually'""'
spot field tests on the sample obtained in the well obtains a completely liquid sample, saturated at
and for transferring the sample from the bottom previously flowing bottom-hole pressure and tem-
hole unit to containers for transportation to the perature. Such a sample is reproducible, and a
laboratory.
second or third sample will exhibit bubble point
pressures and other properties in good agreement
Plan of the Bottom-hole Sample Method
with those of the first sample. As will be shown in
Several reasons were listed early in this chapter Chapter 7, it is possible to estimate the properties
that cast doubt on our ability to reconstruct in the oi the reservoir liquid phase by extrapolating the
laboratory an exact duplicate of a reservoir fluid. measured properties of the bottom-hole liquid
Of course, a bottom-hole sample obtained in the sample.
tubing under a pressure such that only a single fluid
phase could exist would be representative of the Bottom-hole Sampling Methods
fluid in the tubing and probably representative of In this section a short description will be given of
the fluid in the vicinity of the well. the field operations of taking bottom-hole samples.
When a well is flowing and both gas and liquid The purpose is not to present a detailed description
phases exist at the point of sampling, it is difficult to of the various steps necessary for the preparation
say whether the gas and liquid flowing past the of the sampler and the sampling itself, but rath(.)
sampler is mixed homogeneously. If the gas is dis- to discuss the reasons for the various tests made in
persed homogeneously throughout the liquid, then the field.
the sampler should obtain a representative sample An obvious precaution, but one frequently over-
of the material flowing in the tubing. However, looked, is to determine whether free water exists in
the well might "head," which can result in periods the well at the desired sampling depth. The
during which the material flowing past the sampler presence of both water and hydrocarbon liquid in
may be either predominantly gas or liquid. Cer- the sampler cannot be determin~d easily without
tainly, a sampler of the Humble type is filled within taking a chance of destroying the sample. As it
a few seconds after the disk is ruptured and samples takes considerable time to run the sampler to depth,
whatever happens to be flowing past it during that take the sample, bring it to the surface, and deter-
short interval. While it is more difficult to picture mine the probable contents, the simple precaution
the actions within, say,. the Perco sampler, it is of running a tubing bailer to depth and looking at
SAMPLING METHODS AND APPARATUS 13
the liquid brought to the surface is well worth the instrume nt closed. After the sampler is removed
extra effort. from the well, a transfer head assembly is . con-
Between the time that the sampler traps a sample nected to one end and mercury from the mercury
of fluid and the time the sample is studied in the injection pump is forced against the sampler valve.
laborator y, several simple tests can be made in the When the mercury pressure is raised to the point
field to identify partially the material obtained in where it is greater than the pressure of the material
the sampler and to determin e whether fluid loss within the sampler, the valve is forced open. Mer-
occurs while the sample is in transit to the labora- cury or sample material can then be injected into or
tory. These tests are: withdraw n from the sampler, as shown in Figure 8.
(1) To determin e the pressure existing within the The pressure existing inside the sampler when the
sampler when it is received at the surface. valve first opens is roughly indicative of the mate-
rial within the sampler. For example, if the sam-
(2) To determin e the compressibility of the mate-
. rial within the sampler. pler were completely filled with liquid saturated at
bottom-hole conditions of, say, 2500 psi and 200F,
(3) To determin e the bubble point pressure of the the pressure existing in the sampler at 80F would
material in the sampler. be near 1650 psi. The reason for this is that con-
In Figure 8 is shown the arrangem ent of a cali- traction of the liquid when cooled to 80F causes
brated mercury injection pump, pressure gauge, and a reduction in pressure, which is partially offset byQ
gas coming out of solution from the liquid. Had a
CALIBRATED McRCIIHY sample of gas been obtained at 2500 psi and 200F,
PIIW'
the pressure in the sampler at 80F again would be
near 1650 psi. However, had the sampler obtained
a sample of water at the bottom-h ole conditions,
the low solubility of gas in water would cause the
PRCSSVRE LINE 0,.
SUFFICIENT LEN6TH
sample pressure at 80F to be much lower, probably
TO PERMIT ROTATION close to atmospheric. Thus, the pressure in the
OF SAMPLER
sampler measured when the sampler valve first
opens gives some indication of a water sample.
SAMPLE/f WHILE INJECTING OR WITHDRAW/NG MERCURY
By itself, it usually cannot be used to differentiate
between a gas sample and a saturated liquid sample,
,;
or combinations of gas, saturated liquid, and water
samples.
MERCURY LEI-FL The compressibility of the material in the sampler
also is roughly indicative of the material sampled.
For example, if the sampler obtained 600 cc of satu-
rated liquid at 2500 psi and 200F, the 600 cc o(j
PRESSIIRE-APfflOX.
IOOO PSI ABOVE
FULL PReSSVRE
TO THIS VAL VE
liquid would form approxim ately 50 cc of gas and
BtJlll1t. POINT 550 cc of liquid when cooled 'to 80F. If the gas
PRESSURE j_
were forced back into solution at 80F by injecting
POSITION OF SAMPLER WHILE TRANSFER/NO SAMPLE TO mercury into the sampler, the pressure in the sam-
SHIPPING CONTAINER
pler would increase at about 40 psi per cc of mer-
Figure 8-Arrangem ent of Sampler and Auxiliary Equipment
While Detennining Probable Contents and Transfer-
cury injected. After the bubble point pressure .at
ring Sample 80F was reached, further injection of mercury
would cause an apparent compressibility of about
sampler used in field testing the sample and in trans- 300 psi per cc of mercury injected. If the sampler
ferring the sample to shipping containers. As the had obtained a sample of gas, the greater compressi-
sampler is brought to the surface, the pressure ex- bility of gas would result in a rate of pressure in-
erted by the fluid inside keeps the valves in the crease of about 3 psi per cc of mercury injected.
14 BEHAVIOR OF OIL FIELD HYDROCARBON SYSTEMS
- .
+- -r .. ,.+
t -1,'~
t~ -~~
bubble point pressure higher than the bottom-hole ,:-
pressure indicates that both gas and liquid existed 1800
in the sampler at the time the sampler valves I
closed, or that liquid phase has leaked from the 1600
-
sampler.
,, -
Unfortunately, it is impossible to list nnmbers or
values for the three tests which will identify accu-
1400.50 $5 6$
MERCI/RY Pl/MP READING-,:,:
6()
u
rately the material obtained in the sampler or indi-
cate whether the sample is a desirable one for labora-
tory work. Variations in the properties of crude oil
and in the quantity of gas in solution. under the wide 3200
ranges of pressure and temperature encountered in +
oil fields cause differences in the values of the num- 3000 'Jn-l. 1.. 33
bers indicated above. Consequently, the best cri-
terion as to whether a "good" sample has been 2800 - +-~i---1---b-- ~
obtained is if a second or third sample taken from INITIAL . .r.:t~-....
2600 SAMPLER
the well immediately afterward has the same meas- PRfSS,UR~-
.-+. ,-
ured properties as the first. An illustration of this ' SAMPLE I 117!1 PSIS <
. r
2400
criterion is shown in Figure 9, in which duplicate
samples taken in two adjacent wells producing 2200 ~-
undersaturated liquid samples gave comparable +-
-1-+
bubble point pressures. Roughly the same amount
2000 tt ~ti
of mercury was injected to reach the bubble point
pressure, and the initial pressures in the sampler
ttt-'
1800 _,,_ 1
" ~
+.
-/78!1
''.
--l-1- - ' _._
'.;/)
-:-!l'.,,
were equivalent. These data would indicate that :t::(. ::_)'' 1T:fit
any of the four samples could be used for laboratory /600
:f Tc-, '
tests. --- i+H- - +. J~Jltl
/40~ $5 60 6$
Plan of the Recombined Surface Samples Method MERCI/RY Pl/MP READING-,:,:
By recombining samples of all surface gases and Figure 9-Results of Field Test on Four Bottom-hole Samples
liquids in the proportions in which they are pro-
duced from a well, a hydrocarbon system is ob- in the well tubing, the fluid compositions are the
tained that is representative of the fluid produced at same at the well head and at the bottom of the
the well head. Under steady state floll' conditions tubing. Thus, a true reproduction of the fluid
entering the well tubing can he obtained hy proper
* Bubble points inrrease at about 500 p!:ii per I00F inrrcasc in
temperature. This is not true of all systems, ho\,ever. recombination of surface samples. This, of course,
SAMPLING METHODS AND APPARATUS 15
doe.s not necessarily give a true reproduction of the ized PVT relations for use where specific data are
reservoir fluid as a whole, but rather only a hydro- not availabk'' Secondly, data obtained on other
carbon system that can be used in estimating the than produced ratio system., often can be used in
reservoir fluid properties. calculations pertaining to past or future behavior of
As a matter of laboratory operational con- fluid from a particular well or field. For example,
venience, samples used in recombination methods for economic reasons laboratory tests are often de-
are usually from the primary trap. The recombina- ferred until the reservoir pressure has declined from
tion of these samples to reproduce the well effluent its original value and an indication of the reservoir
hinges on accurately metering the two streams pro- size has been obtained. Over limited reservoir
duced from the pri'mary trap. :.\1ost field installa- pressure and producing gas-oil ratio ranges, gas and
tions do not provide for metering trap liquicl liquid sampled at the primary trap will not differ
dir'.ectly, and, as a consequence, the trap liquid flow greatly from the gas and liquid originally produced
rate must be calculated from the tank oil production from the trap. Consequently, a reasonable recon-
rate and from a shrinkage factor to compensate for struction of the original produced system can be
the decrease in volume of liquid as it progresses made by recombining the later date samples in the
from the trap to the tank. While metering of original produced ratio.
primary trap gas and the gauging of tank oil are
field operations, the determination of the approxi- Surface Sampling Methods
mate shrinkage factor is usually a laboratory Field trap installations seldom permit a choice of()
measurement. sampling points from which to obtain gas and liquid
Accurate gas and oil production rates are impor- samples. Ordinarily a valve for sampling gas is pro-
tant in picturing the state of the fluids in a reservoir vided at the location of the gas meter. Liquid sam-
on the basis of recombined sample tes.ts. For ex- ples as a rule must be taken from the bottom of
ample, a bubble point pressure-gas-oil ratio relation the liquid level glass, although this location is less
can be determined in the laboratory with probable desirable than at an outlet directly ahead of the oil
errors of a few pounds per square inch and one or dump valve.
two cubic feet per barrel. The use of such a rela- A simple gas sampling scheme is shown in the
tion, however, to decide whether the well effluent upper part of Figure 10. A H-inch copper tube
exists as one or two phases under reservoir pressure connects a gas sampling point to the bottom of a
and temperature conditions depends on field gas-oil drying tube containing 4 to 6-mesh calcium chloride.
ratio measurement. Unfortunately, field measure- The sample container for collecting and storing the
ments are often in error by more than 5 per cent. sample is connected to the top of the drying tube.
An error of this magnitude can easily cause a 150 psi The drying tube is maintained under full trap pres-
error in defining the bubble point pressure of the sure, and the quantity of gas admitted to the sample
well effluent. Thus, the recombination methocl is container is controlled by means of the valve at the
inferior to the bottom-hole method, primarily when top of the dryer. After purging the sample con-o
an accurate measure is desired of the bubble point tainer several times to remove air, the container is
pressure of an undersaturated reservoir fluid. filled to the desired pressure for shipment to the
Before passing to a discussion of surface sampling laboratory. This method works well for relatively
methods, a word should be said regarding the re- dry gases and usually can be relied upon to give a
combination of surface trap samples in ratios other representative gas sample where the trap pressure
than their produced ratio. The purpose of such is less than 500 psi and the trap temperature is near
tests are two-fold: First, data can be obtained on the or below atmospheric temperature. At higher trap
physical properties of a wide variety of systems. pressures, or where the trap temperature is con-
Bubble point pressures, dew point pressures, forma- siderably higher than atmospheric temperature, the
tion volumes, and partial phase diagrams can be gas may contain a large proportion of heavy com-
correlated in terms of gas-oil ratio, gas gravity, oil pounds. The sudden reduction in pressure when
gravity, pressure, and temperature to give general- the gas first enters the container may cause a retro-
16 BEHAVIOR OF OIL FIELD HYDROCARBON SYSTEMS
grade liquid to form. The accumulati on of any tainer. The container is filled to full trap
retrograde liquid in the container during the purging pressure. This step must be done slowly so
process will give a nonreprese ntative sample. that gas is not caused to come out of solution
Where this condition might exist, one should dis- in the oil in the drier.
pense with the drying tube and take the sample (3) The air originally in the sample container,
by displacing water from the container under the any gas that came out of solution as the liquid
full trap pressure. However, this second method is first entered the container, and the flashed
not too satisfactory if the trap gas contains a quan- liquid are bled out through the top valve of
tity of water-soluble constituent s, such as H,S and the container. At least three container
CO,. volumes of new liquid are then bled through
The sampling of trap liquid is illustrated in the the sample container in order to displace all
lower part of Figure 10. Liquid flows upward liquid that was altered by the original flashing
process.
(4) Both valves of the sample containers are
closed, and the container is removed and
shipped full of liquid.
WFLL - -
lmVENT A second method of obtaining a liquid sample is .
to displace water from the sample container undei..)
full trap pressure. The drying tube is dispensed
with in this method. Only about 90 per cent of the
water should be displaced by the oil sample if the
METHOD OF OBTAIMN6 TRAP GA$ temperatur e of the sample is likely to increase
before reaching the laboratory. The final 10 per
J$zo' CO#Ett n.. cent of water should be removed after the valve
DOWN WINO[
between the trap and the container is closed. This
creates a small gas cap within the container which
will prevent large increases in pressure for unavoid-
WELf-. able increase in temperatur e.
tn'LUNT
The design of satisfactory sample containers is
A$ l.ASTR$OltT,
SAIIM.ES Motl' 11
oarA1NED FlfOfl
shown in Figure 11. It has been found that 500-cc
UIIE&t.ASS
I Jt'ELD ANO MACH/NC l"LIISH
.Eoo.-1//l"lO. SHELar
SEA1111.SS rua,1116
METH()() OF OBTAINING TRAP LIQUID
(1) The sample container is connected to the top Figure 11-Diagram of Liquid and Gas Sample Containers
of the drying tube, leaving the top valve of gas containers and 50- to 100-cc liquid containers are
the container closed and the bottom valve the most useful sizes for laboratory work.
open.
The best rate at which to flow the well prior to
(2) The valve at the top of the drier is opened and during the sampling period has to be considered
slowly to admit liquid into the sample con- individuall y for each well. In general, the best pro-
SAMPLING METHODS AND APPARATUS 17
ducing rate is the one which results in the smoothest usually is adequate, provided the flow rate and pro-
well operation and the best measureme nt of surface ducing gas-oil ratio remain relatively constant during
products, but which is still consistent with the that time. Tubing pressure, casing pressure, and
normal field operation of the well. The factors gas and oil production rates should be gauged at
usually considered when selecting a producing rate 8-hour intervals prior to the sampling period, and
are primarily of a mechanical nature; they include the readings should be made at hourly intervals
whether the well flows by heads when producing during the sampling period. The flowing bottom-
at certain rates, the size of surface-gas metering hole pressure should be measured immediately after
facilities, and the size of oil gauge tanks. These obtaining the samples.
factors influence the rate selection in the following
manner. Flowline Samples
Flow Characteristics of the Well. The alternate Samples of the gas and liquid flowing in the tubing
slug flow of liquid and gas into the surface traps at the well head or in the flowlines connecting the
may cause minor changes .in the composition of the well with the surface traps are sometimes used for
trap samples, but a more important consideration PVT tests. This type of sample is used primarily
is the difficulty in determining an accurate gas rate where a number of condensate wells are producing
during the observation period. For this reason, the into a common field collection line or into a single
well should always be produced at a rate that will trap installation. 9 10 To sample the productio n()
result in a smooth flow of gas and liquid from the from an individual well, it is necessary to sample
surface trap. upstream of the gathering system. By experience
Surface Meter Installations. When an effort is it has been found easier to use a small laboratory-
made to flow the well at low rates, the quantity of scale trap and sample only a small portion of the
primary trap gas may be so small that the gas can- well effluent than to move a full-scale trap to the
not be metered accurately. One important factor well under test and connect it into the flowline.
in the selection of the producing rate is the size of The basic sampling device, called a tip, consists of
meter runs installed at the well and whether the a straight section of small tubing inserted into the
available orifice plates will give good, reliable meter well tubing or flowline with its open end pointed up~
charts. In some instances, it may be necessary to stream to the flow of fluid. The sample entering
lower the trap pressure in order to increase the ac- the tip flows through the tubing to a small-scale
curacy of the gas metering. However, this proce- trap. An indication of the composition of. the sam-
dure is recommended only as a last resort, because ple is obtained by measuring the amount of liquid
changing the primary trap pressure will, in itself, formed when the sample is brought to a predeter-
result in a different produced gas-oil-ratio. Even mined pressure and temperatur e in the trap.
though the PVT tests may be extensive enough to Many complex problems are involved in obtaining
evaluate the effect on producing gas-oil ratio of a representative sample of two-phase well produc-
changing the trap pressure, the changing of trap- tion from the well tubing or surface lines. Any()
ping conditions from the normal field conditions liquid present in the stream wets the sides of the
unnecessarily complicates the use of the laboratory flowline and forms a film that builds up on the pipe
data. wall and flows along with the gas phase. The gas
Stock Tank Installations. Field stock tank in- flowing past this film tends to ripple its surface and
stallations are not always the best size for determin- tear off droplets; consequently, the liquid phase
ing short time production rates because they are moves both as film on the pipe wall and entrained
built primarily for accumulati ng oil for shipment. liquid in the gas stream. The distribution of the
However, plotting the cumulative oil production vs. liquid is an unknown function of the turbulence in
time usually will give a satisfactory production rate the stream.
figure. When the sampling tip is placed in the center of a
It is advisable to flow a well for an extended flow stream, it cannot by itself sample aliquot parts
period before taking samples. Three to four days of the film liquid, entrained liquid, and gas. How-
18 BEHA YIOR OF OIL FIELD HYDROCARBON SYSTEMS
ever, the velocity of the sample stream entering the make the sample composition agree with the flow
tip can be controlled by regulating the back pressure stream composition, a curve is obtained similar to
at the downstream end of the sampling tube, and the one illustrated in the l.ower part of Figure 12.
can be. made either greater or less than the fluid An ideal sampling device would be insensitive to
velocity in the body of the flow stream. As illus- velocity ratio. freferably, it also would have a
trated in the upper portion of Figure 12, the sample unity correction factor. However, it cannot exist
for a flow stream in which one of the phases prefer-
.
"" h-
+
.. Cl:.
t
1.U :pt .!:,.'t .IC,
+1-n ;-~-1.:: .(fr : :r.r
:r1I.,
.:..;:-4..:
entially wets the walls of the pipe. Therefore, the
problem of obtaining samples from flowlines is pri-
/,. marily that of calibrating and using a particular tip
t under controlled conditions of tip location, sample
velocity, and flowline velocity.
!
~
I.
Obtaining a representative sample from the well
8'=---:
"
~
is not, of course, the whole problem in determining
~ -1
" L: _i:r: TABLE 2. SUMMARY OF ADVANTAGES AND DISADVANTAGES OF
i I.O
VARIOUS SAMPLING METHODS
ADVANTAGES DISADVANTAGES
Bottom Hole Samples
't'j':!' ' 1:ti:; . .
' :.1
.,T .. - ,_ -A i:~~~:L .Etr I. Often useful to gauge the 1. Chance of leakage heforeu '
T
widest applicability. A summary of advantages While the final laborator, PVT report seldom
and disadvantages of the three methods is presented includes the results of the tests made in the field
in Table 2. to identify the sampler contents, it is in order for
Because of the complex fluid behavior encoun- the engineer to request these data to assure himself
tered in reservoir rocks and because of the many that the fluid whose properties have been measured
ways that reservoir fluid can differ in composition in the laboratory is the same fluid sampled at the
within different areas within a reservoir, one is well bottom.
seldom justified in directly applying the results of The recombined surface sample method at times
a single PVT determination to the reservoir fluid suffers inaccuracies because of its dependence on
as a whole. Rather, it should be recognized that field measured gas-oil ratios to reconstruct the well
the limitations imposed by sampling in the well bore effluent. In many instances the uncertainty in the
or at the surface usually require extrapolation of use of the PVT data caused by the measured field
the PVT data to reservoir fluid conditions. gas-oil ratio applies to an equal d.egree to bottom-
Two or three trips with a bottom-hole sampler hole samples. For example, the evaluation of the
should always be made before any one of the sam- fraction of produced surface gas that exists as free
ples obtained at the surface can be assumed to be gas in the reservoir rock involves the difference be-
representative of the fluid at the well bottom. The tween the field measured producing gas-oil ratio
fluid remaining in the sampler after part of the and the laboratory measured dissolved gas-oil ratio. 0
original fluid has leaked out while the sampler was The laboratory measured solubility can be equally
being withdrawn from the well is of little use in accurate whether determined from bottom-hole or
estimating the properties of the reservoir fluid. recombined surface samples.
(.)
CHAPTER 3
BEHAVIOR OF GASES
Almost all petroleum engineering problems use air or 16.04 pounds of methane is confined at 14.7
the PVT behavior of gases in one form or another. psia and 60F (called standard pressure and tem-
Gas reserves, gas pressure gradients in well bores, perature conditions) , it is found that it occupies
and the calculation of gas flow rates at the surface very close to 379 cu ft of space. In fact, if they
are three example problems requiring a knowledge were "ideal" gases, they would occupy 379.4 cu ft
of gas behavior. per lb mole at 14.7 psia and 60F. ( '\
The basic equation expressing the PVT behavior In practice, it has been found that no gas obe)Y
of gases is the ideal gas law, this simple law over all pressures and temperatur es.
PV _ RT Hydrocarbo ns, in particular, deviate widely from
(3)
M the ideal gas law at the elevated pressures and tem-
where, in usual engineering units, peratures of petroleum reservoirs. In order to
P = pressure, absolute (psia) express a more exact relationship between the
V = specific volume of the gas (cu ft per Jb.)
T = temperature, absolute, R = (460 + F)
0 variables, P, V, and T, a correction factor must be
R, arbitrary constant, = 10.73 for all gases introduced into Equation (4). Either oftwo meth-
M = molecular weight of the gas
ods is normally used; the first method introduces a
Thus, as expressed by Equation (3), the volume of a multiplying factor into the right side of Equation
unit weight of an ideal gas at any equilibrium tem- (4), and the second method makes use of an adding
perature and pressure can be evaluated solely from a factor. The multiplying factor goes _by several
knowledge of the molecular weight of the gas. names; three of the most common are compres-
If Equation (3) is changed to represent the PVT sibility factor, deviation factor, and super-ex-
state of m pounds of gas by multiplying both sides pansibility factor. The compressibility factor and
by m, we have deviation factor are synonymous, and the super-
PV = mR7' expansibili ty factor is the reciprocal of these two
4
M ( l factors. The adding factor is termed the residuu
where V = total volume occupied by m pounds of gas (cu ft) volume. Both methods are accurate to the same
degree, but the first is easier to use and for this
The number of unit pound molecular weights, N,
reason is the most popular.
comprising the system is represented by the term
m/M in Equation (4). In practice, N, referred to
as the pound moles or moles of gas comprising the The Compressibility Factor
system, can be evaluated from the relation N = Placing the compressibility factor, Z, in the right
m / M, or it can be evaluated in terms of a volume side of Equation (4) and at the same time substi-
.of ideal gas measured at a reference pressure and tuting N for m/M gives
temperatur e. An example of the foregoing is that
PV - NZRT (5)
28.97 pounds of air represents one pound mole of
air, just as 16.04 pounds of methane represents one Equation (5) is the basic equation for gas behavior
pound mole of methane. If either 28.97 pounds of calculations. An illustration of its use is given in
BEHAVIOR OF GASES
21
Table 3. The compressibility factor of an ideal gas
is unity at all pressures and tempe rature s. The /. 0
compressibility factor of a dry natura l gas at atmos - I
~
Aft"TH-4,t,rl
lo#O F'
"'"-a..
--=- - 5
compounds. Because of the small quantities of
'
111 1
heavy components, the calculated pseudo-critical
o.
'\.
--- 1--
.
G as gra d 1ent
9 91
= -144 = 0 .0 6.9 psi. per f t
0
"' C,+ fraction has been assigned properties of octane.
With the exception of the critical pressure of meth- Pseudo-critical constants-gas gravity correlations
ane and minor discrepancies caused by the isomeric similar to those shown in Figure 16 have been pre-
compounds, there is good correlation between the sented by other authors. Mathews, Roland, and
critical constants and molecular weight of the paraf- Katz" have presented data on 2,5 Oklahoma City
fin series of hydrocarbons found in petroleum (Wilcox) gases that included high-pressure equilib-
reservoirs. As the pseudo-critical pressure, the rium vapors, trap gases, plant intakes, and plant
p~eudo-critical temperature, and the average molec- residues. The relations obtained for these gases are
ula:r weight of a gas are always obtained by multi- shown in Figure 17. In addition, these authors
plying the appropriate constants of the components found a relation for 150 miscellaneous gases that
by their mole fractions and adding the resultant falls almost exactly on the California gas relation
products, it would be expected that a correlation shown in Figure 17. Standing and Katz 13 ob-
of the pseudo-critical constants with the average tained a slightly different relation; however, their
molecular weight of the gas would exist for gases of gases differed from the California gases in that the
similar makeup. high gravities were caused by quantities of hexane U
A test of this reasoning is shown in Figure 16. and heptanes plus material in the gas, rather than
The pseudo-critical constants of 71 California quantities of ethane and propane. Consequently,
natural gases, calculated by the method illustrated their gravity-pseudo-critical relations would not be
in Table 4, have been plotted against their calcu- expected to be the same as those of the California
lated gravity (ratio of the molecular weight of the gases.
gas to the molecular weight of air: 28.97). These Before discussing in detail the use of Figure 17,
71 gases consist of high-pressure trap gases (200- the reason for the favorable correlations should be
500 psi), low-pressure trap gases (30-100 psi), and emphasized. All the gases in the correlations had
tank vapors, although a few condensate well nitrogen contents of less than 2 per cent. The car-
effluents and laboratory mixtures have been in- bon dioxide content of the gases was, with few ex-
cluded. As can be seen, a good correlation exists ceptions, less than one per cent. Finally, their
between the variables for these particular gases. hydrocarbon compositions followed a definite trend,
24 BEHAVIOR OF OIL FIELD HYDROCARBON SYSTEMS
1.0 1.0
- . - ~ .....o.s
~~
11
.2 i .. - b !. i,. _ l'I.~
~
..;. - ''
.....;: . ~
-, ' . f-Ll-
'
.. -
'
r-,
- ' . I ,- Jt:
0.9 '
~
, r . ' .
i\\',N I.' - "s.;l:
i1
~.
'
' " - }( \
_, ' ' . ~ ~ r
~i
'
~
er ~ .
.
'
,-,1.e- f- 1-~ ... -- '
.!~L';"
.:.,..:_
'4/
,.
0.8 .., 1.7
-!--,.... ' - 2ft T
' '
1.45 ~
' 'C't"' - ii~
.j,. .. ''
' ' ,
N 0.7
h ' 1.6 L)
' 1.3!>
a:: t t- . ':rt
-t-+
,,;
0
I- ,_. ' ' .
u ... -0 ,,
~ 0.6
)-
I- '
- , 1.25- -.
~
"':'j
~/
4 '
i_._ .. ,.
it+-'
'11
-t- .-f -
..;
- -~,.,,.,
-~ \,,
1.5
:J -
iii
II) 0.
,4:J..Li".~' ~t:J- ~i
1-.,!
II)
.r ~-, ,;l.'..r-~ .. 1.4 N
It.I
a:: I.I.!> 71,+f' I
.I.~ -
.:,
. " ,.to;-
1-,1.~
a::
0
Q. Y,
' 'rLt I-
-~
:l + ., ~ + ~-~
u
0
+ _1:17 -<t-,i.
..,, ... ~~
u 0. -. ~
L
~ -
1.3
rl
7 r' 1 .. 2., -
4
~ ~
'-' 1:"'. - ~ - :J
;~ iii
0.3 ' iii
' 1.2 U)
~
..
I.I~.
~
.o ,..,. w
0.2 " a::
Q. ()
13,0 :l
C: 0
u
?.-
I.I
.'f
1.0 ,.
l) ,'l ~
~IF'-
;-1
,., . .
1COMPRESS1BILI 1.0
'
..... ....
NATURAL GASES
1-0~ '
. T
- T
I T
0.9 0.9
7 8 9 10 11 12 13 14 15
PSEUDO REDUCED PRESSURE
Figure 16-Comp ressibility Factors for Natural Gases (After Standing and Katz, Trans.
AIME, 1942)
BEHAVIOR OF GASES 25
unknow n gas can, of course, be greater than 3 per
cent, because of the possibility that the unknow n
gas contains nonhydr ocarbon compon ents or a dis-
tributio n of hydroca rbons different from those used
in the correlation. However, as the relation s used
to obtain Figures 18 to 20 gave good agreeme nt
with Mathew s, Roland, and Katz's 150 miscellaneous
gases, their 25 Oklahom a City gases, and the 71
California gases, some confidence can be placed in
the general use of the charts. An illustrat ion of
such general use is given in Table 5.
in error by less than 3 per cent. The error to be z = compressibility factor at flowing pressure and temperatu re
P = flowing pressure (psia)
expected when applying Figures 18 to 20 to an h = orifice differentia l (inches of water)
26 BEHAVIOR OF OIL FIELD HYDROCARBON SYSTEMS
700 !
l t j -11-,.L/-.j.-J..;..+ -f-+++++
_1---f-j--l-
1- :
- --!-+- t- -!-+--t-
f- -:
+-++-+-H-+-r-1-S- -+++-f+IH-l+.JA.+IH-1-
-- - -i--1- - J- -1- J--t-1--H-+
1
1-H,-+f+l-l-l-.1-l-+-l-+I+ -.IHl-1-l-++++A.-!.+1-1-1..J
675
-H . I
-,LI
L . :
. .,..I
".'.~ca'H:7r:H+1-Hi=t.t++1;:t++i
Tri
........
r-H-,
-i - j- --
', 11., ~
c-l -
1
. , - C,9tftr ....'l. ,A- 1- r -
""""" t- -
1
.. -:- L1-+-f-l-+-I-I , L
;...
-H--H-1-1-1-+++-fA-1-I-S-++L-l-i-
+4- -f-Li-
650
625
I-
U-H-l-lL~H+tc
-I--_;__++-
'-j
I --t:I~=~
J._ a-r-~ --- --
I
lli'cssuh"il-~+++,++++_-+_,'.::+11-.. _IH-H-1'4-+++:++-H+H-1
~+
-+I--
L-.
-l+l+-f+-1,-1-H
j : I
. -++-H-+-f-'1-+..-++-'..J
1+++-1-11.+1+++-i-e-
++++++++.,_ l-+++-f+H-+++++l-+-l-l-++++-H-l-H-++.A-+-i-H-'--'-'--'-+1-1-1.+.. 1 -1-Hl-+,+-f-1-1-1-H
I
- L +-f-1-1-1-il-H
... J--1-1-'l-l-l-+.J-+-f-!-I-Hl-+-l--i--l-l-.l
++++J.-HI-H-+-++++++-N++++++H~~~"' C.)
-HH'-+IH-I-H-1- f-+ ,.~t_'.*M+t++-+,+l-1-H-t+t+++++-H'-i:-/-H-t+t++++I
... f-'I--H,-1-1-,Lf-.J.41-.j.-J.l~V LI---H-I-H-H-+-f+-1-1
1-J..+-f-Hf-- "-~- +-.i-.\-+-"t~'.rJf--J4-l-l-1-l-!---L.
H-+l-+H-+++-/+J.-1- 1
- --~
+++->-I--Ll-H--l-l---l-1-f-,_,_.,_ ;,-1-1 . .1:t- "J,V.i,_"M-+Hl-H-++ Hl--9++f-lH- I ' H-++-1-1-+l+++++-I
-1-f.++++++-L1-- -
l-+l-h-+-l+-l+-l-l-i-++++++l->,4\
,,,/J"
l+...J-.J-1-1++++-l-i-++++-l-hv~
1-L-l......;.-i-1-...1-1-1-i--i-1-~. r,~
++4-+.:,~
I ~+++H-J-++-f+i-+l-J-+-i--f-1
,
,.
1
,--
I
- - - CALIFORNIA GASES -
- - - - OKLAHOMA CITY GASEs/14):
400
- -- - CONDENSATE FLUIOS(t:S) :
I -
- !
- I ' ! I ' I I i I
375
:1-~ ~~ : -4-
+-'-f-.f-1-,0
! I
-- ! I 1
-- - 1
,,
1-~ 4 -l-H-++--1--H-H-i--
I _,_ 1--- - I - 1 I
I .,-
1++y+-f-i-l-J-f++-f-1-1-J-)-I++-f-1
350 II
f- -
.l-l-1-t--1-t-+t-+-H+-H-+++-f+l-+l-++-+++IJ ..L
_j
1- -1-t-l--
1
-1
++..-'+o+++-H- .
I
T _!_j..... L ~
+-t J_
I '
+-Hl-+++-f-1-1-l-l-++++l+IH-H-++++-+H-+++,-1-I..J1-J-H
I I- ,-11_-!-f-
j_j_ -1-~
-i:-t:-;= ~ L
325
i-l-1tf-tttij'jjtttttt+:tt~-+ tt+-~l- - :
_j__
- I
'I I .;h-
!
+f+l-l-l-+~L+4~+.:,+:..::.t-.:.1+4~~tt-t+:++-+1- I 11 ~I --i- 1++
300 '
0.60 0.80 1.00 1.20 1.40 1.60 1.80
GAS GRAVITY-1:!RI.O
Figure 17-Psuedo-critical Constants of Gases and Condensate Well Fluids
15
11111111111111
.....
"'~
~ "'
-
9 l'TI
I
.., >
8 <
I 0
),.
..... "'0..,
ii; G)
:t:
>
~ V,
l'TI
V,
II)
"'....
C C
"'"'
:wt
15 It
r
14
/3
12
"'
rn
:r
)>
I I <
....... JO
6
"'
~ T ..... j._.:.-t 0
.,,
~
II:
.
Ill
9
8
t 0
r
.,,
rn
r
0
I 7 :r
),,, -'-.....is:::
.... -<
0
~ 6 "'
0
r,
~ 5 .,. )>
II)
'"I
"'0"'
Co 4
z
V,
-<
V,
J.0 --<
3 rn
3::
V,
,0.5
'l.f
~
I 0
!JO
-s:-t~
00 400 800 1200 1600 2000 2400 2800 3200 3600 4000 4400
PRESSURE -PS/A
Figure 19-Pressure-Volume-Temperature Relations of 0.70 Gravity Natural Gas
C C
l
15
/4
...
I..
~ /0
~ co
rn
I
IQ
.... >
~
I 0
""
,,
0
C)
>
V>
rn
V,
0 ''--" 1
a Je.,. 'jjcti;jj
---;.::: i$t:;::;.Jt i
0 400 800 1200 1600 2000 4000 4400
PRES SURE - PS/A
Figure 20--Pre ssure-V olume-T emperat ure Relation s of O 80
Gravity Naturn! i.;.as
K)
,0
C C
30 BEHA YIOR OF OIL FIELD HYDROCARBON SYSTEMS
as a function of flowing temperature and pressure. that volumes of natural gas and nitrogen are not
An illustration of this is shown in Figure 21. quite additive. Values of c, presented by the au-
....o thors in chart form, indicate that the correction
,ii'! . factor, c, is between 1:00 and 1.04 at temperatures
pTt: 382R abo.ve 70F, and for nitrogen concentrations less
uo pP~ 669 PS/A than 20 per cent.
The above method is, of course, complicated by
the requirement for c charts as well as compressi-
... bility charts both for the hydrocarbon gases and for
l'' '40 .
,io-'
nitrogen. Use of critical constants of nitrogen,
492 psia and 227R, in the theorem of corresponding
states yields a much easier way of calculating a
compressibility factor. At nitrogen concentrations
of less than 20 per cent, the compressibility factor ..
1.10
of the entire gas e,.9Juated by the theorem of corre- U
sponding states is usually in error less than four per
1.00 cent.
0 !100 /000 /$(}() 2000 JOOO
PRESSURE-PS/A
Olds, Sage, and Lacey 17 investigated the behav-
Figure 21-Meter Factors. Murphy-Coyote 0.67 Gravity Gas ior of carbon dioxide-methane mixtures. Reamer,
Olds, Sage, and Lacey18 determined compressibility
Effect of Nonhydrocarbon Components on factors for the ethane-carbon dioxide system.
Compressibility Factors Application of the theorem of corresponding states
Nitrogen and carbon dioxide are often present in will yield satisfactory results for natural gases hav-
appreciable concentrations in natural gases. As ing 2 per cent carbon dioxide or less. Olds, Sage,
can be seen in Table 1, the critical constant-gravity and Lacey indicated a possible error of 5 per cent
relationship of these compounds does not fit in with in the compressibility factor at a concentration of
that of the paraffinic hydrocarbons. It is to be 4 per cent carbon dioxide.
expected, therefore, that correlations based solely
on gas gravity can be in error when applied to gases Residual Volume Methods
containing large quantities of the nonhydrocarbons. Residual volumes present the second or "addi-
Eilerts, Mullens, and Carlson 16 have studied the tive" method of calculating PVT behavior of gases.
compressibility of nitrogen and natural gas mix- The residual specific volume of an individual gas ic\' 1
tures. They proposed that the compressibility of a defined by the equation: V
nitrogen bearing natural gas be determined by
separately evaluating the compressibility factors V=V -V _RT -V
(perfect) (<:tual) MP (10)
of the hydrocarbon portion and the nitrogen portion
where
of the gas mixture and then obtaining an "addi-
tive" compressibility factor as follows: V - residual specific volume, cu ft per lb
V - specific volume, cu ft per lb
z.. = nZ,. + (1 - n)Z,. (8)
R - gas constant, per mole
where
Z,. = compressibility factor of the nitrogen in the gas sample M - molecular weight of the gas
Zh = compressibility factor of the mixture of hydrocarbons in T - temperature, 0 R
the gas sam pie
n = mole fraction nitrogen in the sample P - pressure, PSIA
The computed value of the additive compressi- Thus, the residual volume is the difference between
bility factor, z.,
then has to be corrected to the the volume of a gas computed as an ideal gas and
BEHAVIOR OF GASES
31
the actual gas volume. The residual becomes at tempera tures between 70F and 250F. Since
negative in the region of tempera ture and pressure then, partial volume data (from which residual
where the actual gas occupies a volume larger than partial volumes can be calculated) have been ex-
the corresponding volume of an ideal gas. (This tended to 10,000 psia and 460F. 21 Additio nal
is comparable to the region of Figure 15 wherein Z systems composed of methane-decane, and meth-
is greater than unity.) ane-n-butane-decane have been investigated.
The use of residual volumes becomes more com- Let us consider the use of residual partial specific
plicated when a multico mponen t gas is considered. volume data in computing the specific volume of a
Just as in the case of the compressibility factor gas at an elevated pressure and tempera ture. In
method, it is necessary to take into account the order to use the published residual data, 19 it must be
composition of the gas when calculating the devia- assumed that the residual partial volume of the
tion of the gas behavior from the ideal behavior. material that is heavier than isobutan e can be
In calculating a compressibility factor, the weight- represented by that of n-butan e. If we let (V1) 1,,;
ing is accomplished by multiply ing the mole frac- (Yrha, and (Y,)u represent the residual p;rtial
tion of each compon ent by the critical constants of specific volume of methane in the methane -ethane,
the component and summin g the products. In the methane-propane, and methane -n-butan e systems,
residual volume method the weighting is accom- and give weight to the individual methane values
plished as shown by the following equation s: in direct proportion to the concent ration of the
heavier component in the three binary systems, the
(11) weighted residual partial volume of methane may
be established from the following relation :
or as can be seen from Equatio n (10),
- RT - -
- (v,)
V1 - n,+(v,) n,+(v,) n,
2- ---V- 1n 1
1:2- - 1:1- - I:~-
V=---V 1n,-V2n
M.P -- -- --
(12)
- = l-!!_1 (13)
where
M. - average molecular weight of the In absence of similar data to calculate ,rnighted
whole gas
residual partial volumes of ethane, propane , and
n, - weight fraction of a component in
- the gas
butane, it is necessary to assume that the value
determined in combination with methane is appli-
V- residual partial specific volume of a
- component in the. gas, cu ft per IL
cable to any multico mponen t system.
Substitu ting the value of '{1 in. Equatio n (13)
The first term in Equatio n (12) represents the spe- into the second term in Equatio n (12) gives the
cific volume of the whole gas if it follows the ideal following equation :
gas law, whereas the succeeding terms are the
deviation from this idealized behavio r of each of the
components in the gas. In Equatio n (12), the
residual partial volume of each component is a u
function of pressure, tempera ture, and the nature (14)
and amount of all the other components present
in the gas.* The calculation of the specific volume of 0.67/i
Sage and Lacey and their co-workers have studied gravity gas at 3000 psia and 250F, shown in
systema tically the behavio r of binary systems com- Table 6, illustrat es the use of Equatio n (14) and
posed of the lighter components found in natural residual partial volume data. The length of the
gases. By 1939, residual partial specific volume calculations would appear to be about equal to the
data had been reported " on methane, ethane, pro- length using compressibility factors. However.
pane, and n-butan e at pressures to 3000 psia and residual partial volumes are not interpol ated easily
between tabulate d pressures and tempera tures, so
The method of determini ng residual partial specific volumes
and their mathemat ical significance has been given by Sage and the usual calculation is actually more laborious
Lacey.10,20 The discussion will not be repeated here. than indicate d here. When a problem calls for a
32 BEHAVIOR OF OIL FIELD HYDROCARBON SYSTEMS
TABLE 6. Ex,UIPL E CALCUL ATION OF THE SPECIFI C
Vol,UM E 01'' A GAS USING RP;8IDU AL Vot,UME
DATA
Pressure 3000 PSIA Temperature 250 F
1 2 3 4 5 6 7 8
Molecu lar -
Mole Weight of Weight
Fractio n ComKI1nent Fraction
Component n nM n <'?,k- y
C, 0.8602 16.0 13.80 0.7128
C, 0.0770' 30.1 12.82 0.1201 - -
C, 0.0426 44.1 0.00648 0.0352
1.88 0.0973 0.0066 0.0320
i-C, 0.0057 58.1 0.33 0.0171
n-C, 0.0087 58.1 0.51 0.0264
i-Cs 0.0011 72.2 0.08
n-C~ 0.0014 0.0041 Weight Fractio n 0.0037 0.0315
72.2 0.10
c,+ 0.0033 100. 0.33
0.0052 i-C,+- 0.0699
0.0177
1.0000 19.32 1. 0000 I
Values for f 1, (y' 1) 1 : 2 etc. at 3000 PSIA and 250F are from data tabulate d in Reference 19
Specific Volume, V = 1~:~T -[ ~~ ] [ c'?1h: 2n2+(~'i)1, l n3+(''1
1 1 h:, n, J-Y2!h -Ya!h -f,n, =
10.73 710 0.7143
19.32. 3000 (1-0.7143) [ (0.0064 8)(0.12 01)+(0. 0066)(0 .0973)+ (0.0037 )(0.069 9) ] - (0 0352)(0 1201)
. .
-(0.0320)(0.0973) -(0. 0315)(0 .0683) 0
Specific Volume, V-0.116 8 cu ft per lb
solution of pressure, with the tempe rature , volume, Summ ary of Princ ipal Point s
and composition of the gas given, the calculations
may become excessively complicated because the The equat ion, PV = NZRT , and the two con-
pressure must be roughly known in order to evalu ate stants , 10.73 and 379, are the most impor tant factor s
the residual partia l volumes. In other words, it is in makin g gas calculations. Reme mberi ng these
necessary to resort to a trial and error calculation three factors will allow one to solve any gas PVT
method of determining the pressure. problem.
The accuracy of the residual volume metho d is at When gravi ty with respec t to air is the only
present no better than that of the compressibility _known param eter of a gas, the PVT behav ior of the
factor method. For example, specific volumes of gas is best obtain ed by use of Figur es 18 to 20. At
the gas, shown in Table 6 _calculated at 3000 psia gas gravit ies greate r than 0.8, the pseud o-crit ical-
and 250F by the three metho ds outlin ed in this gas gravit y relatio ns shown in Figur e 17 can be
chapter, gave the following agree ments with the used to evalu ate a comp ressib ility factor at the
experimentally determ ined specific volume: desired pressure and tempe rature .
In gases conta ining only 1 or 2 per cent of heav yu-
Specific Volume
(cu ft per lb)
Per Cent
Error
fractions, the true critical const ants of the next
Experim ental 0.1217 higher comp ound shoul d be used for the "plus "
B\ Figures 18 and 19 0.1~03 -1.15
B) pseudo-critical method 0.1197 -1.65
fraction. Thus, a hexanes plus fracti on uses the
By residual volume method 0.1168 -4.40 critical const ants of n-hep tane.
CHAPTER 4
BEHAVIOR OF LIQUIDS
The PVT relation of liquids als~ is import ant in volume of a liquid mixture at 14.7 psia and 60F
calculations pertain ing to petrole um reservoir be- from the composition of the liquid and the densities
havior. The isother mal compressibility of under- of the individual components. The atmosp heric
saturat ed liquid, the shrinkage in volume that a volume of the mixture can then be corrected for
barrel of reservoir liquid undergoes while progress- compressibility and therma l expansion to obtain its
ing from the reservoir to the lease tank, and the
in-place density of the reservoir liquid are three
volume at an elevated pressure and temper ature.
When applyin g the ideal solution method to
0
examples that require a knowledge of hydrocarbon reservoir liquid systems that contain large quantit ies
liquid behavior. of dissolved gases, it is appare nt immed iately that it
There are several method s of calculating the vol- is impossible for such systems to continue to be all
ume that a given mass of liquid occupies under an liquid when brough t to atmosp heric pressure and
elevate d pressure and temper ature. When the temper ature. However, this physical limitati on
composition of a reservoir liquid is known, its does not impair the mathem atical use of atmos-
volume at any pressure and temper ature can be pheric liquid densities of methan e and ethane in an
calculated either by applyin g ideal solution prin- ideal solution calculation, because the "pseud o-
ciples or by using partial volume methods. When liquid density " of the system at atmosp heric condi-
the composition of the liquid is unknown, but the tions is only a steppin g stone to the evaluat ion of
properties and quantit ies of surface gas and oil de- the density at an elevate d pressure and temper ature.
rived from the reservoir liquid are known, empirical Standin g and Katz" studied measured data on
methods are available to calculate the liquid volume systems contain ing methan e plus other compounds
under reservoir conditions. This chapte r will out- and ethane plus other compounds. By taking the
line first the method s that can be used when com- experimentally determ ined system density at nu-
position of the liquid is known. Following this, merous elevate d pressures and temper atures and U, '
two empirical method s will be present ed that are correcting this density to 14.7 psia and 60F, a
based on surface gas-oil ratio measurements. so-called pseudo-liquid density of the system was
calculated. Subtra cting the 'volume and weight
Method of Calculating Liquid Densit y from contrib uted by the "plus" compound from the
Composition Using Ideal Solution Principles pseudo-density left an "appar ent liquid density "
of the methan e or ethane. In other words, the
The behavi or of mixtures of liquid hydrocarbons appare nt liquid densities of methan e and ethane
at atmosp heric pressure and temper ature ap- are fictitious densities and are the values which,
proaches the behavi or of ideal solutions. That is, used in an ideal solution calculation with the true
a unit volume of one liquid, such as pentan e, added liquid densities of other compounds, give an
to a unit volume of anothe r liquid, such as heptane, atmospheric system density that can be corrected
will result in two volumes of mixture. This ideal to elevated pressures and temper atures by suitabl e
solution behavi or makes it possible to calculate the compressibility and t,hermal expansion factors.
34 BEHAVIOR OF OIL FIELD HYDROCARBON SYSTEMS
For convenience, the appare nt liquid densities are three values the pseudo-liquid densit y of the system
calculated at conditions of 14.7 psia and 60F. at 14. 7 psia and 60F is determined directly from
When the above procedure was carried out, the the chart. This procedure is illustr ated by the
appare nt liquid densities were found to vary with example problem given in Table 7.
the densit y of the system as a whole, as shown in
Temperature-specific gravit y corrections for liq-
Figure 22. Direct use of the data in this form
uid hydrocarbons having specific gravities in the
.... o.
range 0.60-1.0 are given in the Nation al Standa rd
Petroleum Oil Tables. 23 The NGAA Stand ard Cor-
...~ rection Facto rs" cover a lower tempe rature range
,.~ .,,- than the NSPO Tables, in additio n to lower gravi-
..,.. .. ~-
O.!J
-,1- ties. Sage, Lacey, and co-workers on API Projec t
~ 37 have measured the effect of tempe rature and
~
I
.t ...- +
.
~
0.4
-- --r
pressure on the density for a numbe r of multicom-
ponent liquids. A survey of these data indicates:
I
:.
0 ..1
(1) The compressibility of hydroc arbon liquids at
tempe rature s below 300F can be assumed
to be a function of the liquid densit y at 60F. u
Minor differences exist in the liquid compress-
I
0.4
',!J--._ ibility between systems of the same density,
but the differences are not regular enough to
~
,. ::
0 . .1
. .. "-" ., . . V " take into account.
i
~.
' I
0.1
y "'
(2) The thermal expansion of hydrocarbon liq-
uids is affected little by pressure. Thus, it
.
~ 0.1
V
.
...... "44 ~
+,
if
ETHAN - N BUTANE
ETHANE- HEPTANE
ETHANE- CRY.STAL OIL
is possible to use the tempe rature corrections
listed in the Nation al Stand ard Petrol eum
Oil Table s" at elevated pressures, as well as
Q: MTHANf:-C YCLO HEXANE
c? M[THAN 8tNZEN
at atmospheric pressure.
:.
~ METHANEPENTAN!
o METHANE- HEXANE
"9' MTHAN[H IEPTAN
METHANE PROPANE
An additive correction factor to the atmospheric
M[THAN-C RTSTAL OtL
i METHANE CRUD[ OIL
density of a liquid for the effect of pressure has been
I I presented by Standi ng and Katz. 22 The range of
0 ..1 0.4 0.S o. 0.7 o. the factor was later extended by Hanson, Kuist,
DENSIT Y OF SYSTEM .aF a ATMOSPHERIC PRESS/
IRE and Brown.'" This correction factor, presen ted in
(lllt/C~
'
60
,
~ "'v I
,.
-
,.,~~
"'!I'~ '
'<I "
50 ~ 0 ....
/ifs
~~
I/ ,
,, '
40
f
~,
~ /;j
f~ "". ~1, r;; I.I
,er-,,
~ I I ,_
I I
-
o~ 1-1-~
60
~
It:
t
0
t I/ r,
- I I
I '
-\0
...,
QI
~~
I/
[7
~+~ ', I
-~\t~
30
v~~\-fJ
~q.. q..~ ',
;' ,,
T
T
-0
I
..,
..:
~ ~\ z.O i
~~
"i "
:;.~'(:, ,~ - "'
,t_ , 40 ~
'"
I/ ~
~
" '
'!JO !C
L ~
a
C
30
'.
r:
I,
~ 0
- , ,_
~
..:
, ~
2 0 c.,
~
).
....
ii;
..... ~
Q
Figure 23-Pseudo Liquid Density of Systems Containing Methane and Ethane
36 BEHAV IOR OF OIL FIELD HYDROCARBON SYSTEMS
10 - - -
-H- - f
- -- - -C --~
- - - - -
-
g -H- - + C - -
-
' _, ' -
U\ - -~
0 _J-.;_
~
- t
' -
~ - -C-
'
-
-
~~ ~"
<J) '
-
- -
' - '
I '
\$< -
r
cc
0
~ -
',.0 ~
'..s, '
~ ~' ;s,G,
~ ~-
"- ~
~,
I>~
' ~
?.
"i O
'
~~~- -C-
0
I
.
'
~..:
~~ooo
0
I
:r -
0 - ' ''
a SO S 40 45 50 55 60
DENS ITY AT 60 F ANO 14.'J' PS/A -LB PER CU FT
Figure 24-Dens ity Correction for Compressibility of Liquids
BEHAVIOR OF LIQUIDS 37
10
!J
0
s0,..
0
5
1P.0 '
"'6'o
".96
2
'<'q
10()
I
80
0 60.
25 .JO .J5 40 45 SO 55 60
DENSITY AT soF ANO PRESSURE p - LB PER Cl/ FT
Figure 26-Density Correction for Thermal Expansion of Liquids
38 BEHAVIOR OF OIL FIELD HYDROCARBON SYSTEMS
TABLE 7. CALCULATION OF LIQUID DENSITY USING IDEAL SOLUTION PRINCIPLES AND APPARENT METHANE AND ETHANE
DENSITIES
Problem: To calculate the density of a crude oil-natural gas system at its bubble point of 3280 psia at 2!8F, given the com-
position of the system.
Basis: One-pound mole.
1 2 3 4 5 6
Licid Volume
of omponents
Liquid Density at 60F & 14.7
Molecular Weight of of Components PSIA
Mole Weight of Components at 60F & 14. 7 PSIA cu ft
Fraction Components lb lb per cu ft nM
Component n M nM p p
C, 0.4404 16.0 7.046
C, 0.0432 30.1 1.296
C, 0.0405 44.1 1.786 31.66 0.0564
C, 0.0284 58.1 1.650 35.87' 0.0460
C, 0.0174 72.2 1.256 39.16' 0.0321
C, 0.0290 86.2 2.500 41.43 0.0603
c,+ 0.4011 287 115.1 56.6 2.032
1.0000 130.69 2.227
factor of -3.57 lb per cu ft should be applied to can be calculated by summing the products of the
the 3280 psia-60F density value to account for individual component weight fractions and their
the thermal expansion of the compressed liquid in partial liquid volumes. This can be expressed
going from 60F to 218F. The density at 3280 mathematically as
psia and 218F is, therefore, 51.7 - 3.57 = 48.1 lb
per cu ft. (15)
The method just outlined and illustrated by Table
7 usually will give densities less than 3 per cent in As in the gas calculations, the partial volumes of
error. Standing and Katz" compared experimen- the individual components are a function of prQ
tally determined densities of fifteen liquids saturated sure, temperature, and the kind of material com-
with gas at pressures up to 8200 psi with calculated prising the rest of the system. This third variable,
density values and found an arithmetic average in the case of hydrocarbon liquids found in petro-
error of only 1.16 per cent. The greatest error was leum reservoirs, is expressed by the average molecu-
four per cent. Checks with various data of Sage lar weight of the other components in the system.
and Lacey have indicated comparable accuracy. Therefore, in using the partial volume method, it is
necessary to evaluate both the weight fraction of
Method of Calculating Liquid Density from every component in the system and the average
Composition and Partial Liquid Volumes molecular weight of all other components except
the one under consideration. It can be appreci-
The specific volume of a liquid (reciprocal of den- ated that with three variables to consider, extensive
sity) at an elevated pressure and temperature state cross-plotting of the data often is required.
BEHAVIOR OF LIQUIDS
39
TABLE 8. CALCULATIO N OF LIQUID DENSITY BY THE METHOD OF PARTIAL VOLUMES
Problem: To calculate the density of a crude oil-natural gas system at its bubble point 3280 psia at 2!8F,
given the com-
position of the system.
Basis: One pound mole.
1 2 3 4 5 6 7 8
-
Average Partial
Molecular Weight of. Molecular Li'buid Volume
Mole Volume
of omponents
Weight of Components Weight Weight of cu ft per lb at 3280
Fraction Components lb Fraction PSIA & 218"F
Other v
Component n M nM !! Components y!!
C, 0.4404 16.0 7.046 0.0539 221
- - -
0.09
-
C, 0.0432 0.00486
30.1 1.296 0.0099 135 0.0347 0.00034
C, 0.0405 44.1 1.786 0.0137 134 0.0347 0.00048
C, 0.0284 58.1 1.650 0.0126 133 0.0303 0.00038
C, 0.0174 72.2 1.256 0.0096 131 0.0252 0.00024
C, 0.0290 86.2 2.500 0.0191
c,+ 0.4011 287 115.1 0.8812 ( 23 0.0175 0.01575
1.0000 130.69 1.0000 0.02204
~!
r .. +i
+- 't-1 l f-f
-, -t- +
f- - '-.-t ""T .,. 1-; -
r1J.- ---;-+
T"; ; 1 "T""
--t++t r ,-1-;-1--j-
_.
.
- .. ~- -I- -,
, ,
I
--+- 'f-~
I ' 1
rl+-1-1-++-i-H,--,-+, ..1,... T -1,-T I -t .. - - ' 1 T, -~---t-.....;- t-
# -H+ -- -H-++++-H-1-+
t--H'-++'+++,-+++--+-:-t-'--':++++-1-++, -
t-.
.::p- LH
, +~ uoE -1--l-'
! r:ifl.~P. 1A-
t++ Hi++- .d::!'. -
.-J..i<
c~~ ~:t -+1J:1--:-.;:::t ::;;- ..'-1- 1_
3 5 -HH--1-++-l--l--'-++ -- ~ I- . rb -- -1- Tl - -I - ' , - + -H-1-
1--H-++++_++-_.+-~1:...;~-+-+j-+-+-
'
+ zO
O
-~W-
, i- :!Ji- }-1-
-!--L '.t
-
.J-j
+ - ; ~: ~-~
T"
-I
-:-- ++--1-,-++~H-+-I--I--H--H--+-+ t :312 '
:ff-~+-l-+++--1 --
-
'- -lt- .
l--1--
+-+--~ '
- _J - ...
- 1;.;I,.,~ r 60 +
t H-+-+-H-l--;-
t-
,_
H
I--H~+-l-+++l--1--l--H I--H-++-++++-f--l--H--H-1-1--Hh+ 1
'
'
.. =.-rtH +-++-++-i--l--1--H -i--l--l--+-1-1-+--f-+++-H
' '
/- li=+-- ----
f-.. L.1.i
-I--:tt
1 - -+ -i- __
' + --~-
-H- '-'i:f-
--:.,--. +,
-
f+-l++H-t- ~
1--[ M+L
---1- ---l -
'
+ 1--1+t--H-+-r-M'--:L..ri-c--:__.-rt--i--1--i--t--1--t---t-f+t--Hl--1
' -H
15
t- 1-- ------ -+- +: -1 . --1+
0.6 0.7 0.8 0.9 1.0 I.I l2 1.3 1.4
GAS GRAVITY - AIR : I
Figure 26-Apparent Liquid Densities of Natural Gases (After Katz, AP! Drilling and Production Practice, 1942)
tion volume factor of the liquid used in the previous liquid, rn; the dissolved gas gravity, 'Yo; the tank
illustrations. The calculated formation factor of oil specific gravity, o,; and temperature, t. The
1.279 bbl per bbl of tank oil compares favorably pressure is not used directly in this c01Telation, as it
with the experimentally determined value of 1.302 is an implicit function of the other variables. The
bbl per bbl of tank oil. formation volumes correlated in Figure 27 check,iQ
with an error less than 0.5 per cent, 4,'i per cent of
Method of Calculating Liquid Density from the 105 bubble point liquids comprising the correla-
Generalized Data of California Crudes tion. The arithmetic average of all errnr., ll"as 1.17
per cent. Although extensive checks ha,e not been
A second empirical method that can be used to made with crudes from outside California, such
calculate a liquid density employs a correlation crudes should not yield errors appreriabl,' larp;er
developed by Standing' for California crude oils and than those indicated abo.ve.
natural gases. This empirical correlation of the It will be noticed in the example shown in Table 9
formation volume of a natural gas-crude oil system that the calculation of the density and weight of a
at its bubble point condition is shown in Figure 27 crude oil-natural gas system at its bubble point
as a function of gas-oil ratio of the bubble point allows one to evaluate the formation volume at that
BEHAVIOR OF LIQUIDS 41
condition. This relation ran be used in reverse tion volume of the liquid at the desired pressme
direction to calculate the density of a liquid from and temperature. An example of this method is
the weight of material in the system and the forma- shown in Table 10.
TABLE 0.' C.U.LTLATJO:-; OF SHRl;'\KA<iE AND FOR~IATfO:-; VOLUME FACTOR 0~' A CRu.tH; 011,-'.'\ATURAL GAH MrxTURJ; lT:-;rN(i AP!'AHJ<:,','1'
GAS DENSITY :\1ETIIOD
Prnblem: To calc11late the shrinkage that a rescnoir liquid undergoe., in being produced to the lease tank, given the surface
production of gas and oil, the resenoir pressure and temperature, and assum1,;ll,g the liquid in the reservoir to be at
its bubble point.
Pri1nary Secondary
Trap Trap Tank Total
Produred gas-oil rntio, ('U ft p<'r hhl tank oil. 414 00 25 52\l
Produced gas graYity, air = I. 0.640 0.807 1.54
Tank oil gravity. . ........ 27 4 AP! - 0 Sill sp gr
S0l11tion: Basis: One barrrl tank oil
414 0.640 +!JO 0.807 + 2.5 1.54 ~84.2
A\erage produced gas gravity
414 + \JO + 25 = 52\) = 0.721i
5211
u
netght
dissohed gas= ~ _ 0./20 28.0i = 2!l.3 II l
3 /\)
Weight tank oil = 0.801 5.615 62.4 = 312 lb
From Figure 26,
Apparent density of dissolv<'<l gas = 24.0 lh p<'r cu ft nt 14.7 psia and 60F 0
. . . ~.3
Pseudo-ltqmd volume of dissolved ,:i;as = .H .17i cu ft
24
Volume Weight
Dissolved gas. l.177cuft 29.3 lb
Tank oil. .1.615 cu ft 312 lb
u. 792 cu ft 341.3 lb
,,,J (-f:- ) 0 11
+ I IS I
, ..
' ' ,'
The information in Figures 24, 25, and 27 can be
used to determine the compressibility and thermal
expansion of reservoir liquids when direct measure-Q
ments are lacking. For example, using the method
Figure 27-Formation Volume of Bubble Point Liquids from Gas-
Oil Ratio, Dissolved Gas Gravity, Tank Oil Specific outlined in Table 10, a density of 46.0 lb per cu ft
Gravity, and Temperature was obtained for the liquid at its bubble point. By
applying the data of Figure 24, the bubble point
liquid is calculated to have an undersaturated
Summary of Principal Points density of 46.62 lb per cu ft at 5000 psia and 218F.
The average compressibility of the liquid over the
Four methods have been outlined of calculating
indicated pressure interval is
the density of a liquid at its bubble point condition
of 3280 psia and 218F. For the single liquid used 1/46.62 - 1/46.0)
46 31 ( .5000 - 3280 =- 7.3 (rn-)
in the illustration, the following answers were
obtained. volume per volume per psi.
0
CHAPTER 5
VAPOR-LIQUID EQUILIBRIA
Combining the equations of these two laws, we have other.* By using fugacities to define equilibrium
constants, the departure of the vapor from ideal
xp=yP (1!))
gas behavior and the effect of the total system
pressure on the vapor pressure of the components
which can be arranged to give
are taken into account.
(20)
PR$$Ult~ -
<IOO
I'S/A (.)
important because ideal equilibrium constants and Figure 28-Equilibrium Constant Data at 120F as Determined in
Various Systems
empirical constants far more accurately describe
multicomponent system behavior. In an ideal solution the fugacity of any compound
in the system is equal to fugacity of the compound
Ideal Equilibrium Constants in the pure state times its mole fraction. Thus, if
/'., and/'., are the fugacities of a compound in the
A better approximation of actual equilibrium con- liquid and vapor phases, and/., and/, are fugacities
stants is obtained by defining the constants in terms in the pure states,
of the fugacities of the compounds. 29 30 The fu-
(21)
gacity may be looked upon as a vapor pressure
modified to represent correctly the escaping tend- of Hydrocarbons
Refer to Sage and Lacey's "Volumetric and Phase Behavior
O
(Reference 8) for more details regarding use of
ency of the molecules from one phase into the fugacities in determining ideal equilibrium constants.
VAPOR-LIQUID EQUILIBRIA 45
At equilibrium conditions the fugacities of the com- been studied to define the manner in which the
pound are equal in the liquid and vapor phase, so composition influences the values of the constants.
that The shape of the log K-log P curves shown in Fig-
y /kl ure 28 is characteristic of all multicomponent system
-X - -,, = K ciJe"l> (22)
1 curves. At low pressures the slope of the curves is
Ideal constants are limited by one important fact. near minus one. As pressure increases, all the equi-
At a given temperature a compound can exist only librium constants decrease and, with the exception
as a single phase at any pressure other than the of methane, all pass through a minimum value.
vapor pressure, so that the fugacity of either the The constants tend to converge to unity at some
vapor or liquid (depending upon whether the pres- high pressure.
sure is greater or less than the vapor pressure of the The apparent convergence of the equilibrium con-
compound) must be obtained by extrapolation of stants to unity at some pressure is often misinter-
the fugacity-pressure relation. As the extrapola- preted to mean that the system is near its critical
tion is somewhat arbitrary, the farther the pressure state. Unless the temperature at which the con-
is from the vapor pressure, the more uncertain the stants are determined is the critical temperature
equilibrium constant becomes. Usually the ideal of the system, it is impossible for the equi-
equilibrium constants are considered undependable librium constants actually to converge, because
at pressures more than twice the vapor pressure of it is only at the critical pressure and temperature o
the compound. For this reason, they have limited of the system that the phase compositions are the
usefulness for calculations dealing with petroleum same. At any temperature other than the critical
reservoir conditions. temperature, the equilibrium constants become
meaningless after the bubble point or dew point
Empirical Equilibrium Constants pressure of the system is reached. Nevertheless
isothermal log K-log P curves are often extra~
The third method of evaluating equilibrium con- polated to pressure regions wherein two phases
stants is to determine experimentally the equilib- cannot possibly exist. Quite often the curves have
rium compositions 'of vapor and liquid phases at an appearance of converging. Therefore, in order
various temperatures and pressures. The equi- to forestall confusion between actual behavior of
librium constant of each component can then be systems at their critical state and their behavior pos-
evaluated by calculating the ratio of the mole tulated from extrapolated equilibrium constant
fractions in the vapor and liquid phases. data, it should be recognized that in the great
When done correctly, this procedure results in a majority of instances equilibrium constants con-
set of equilibrium constants representing a specific verge to an "apparent convergence pressure," and
system. Figure 25 shows two sets of such constants that this pressure actually may be quite different
at 120F. The data of Roland, Smith, and from the critical pressure of the system.
Kaveler 31 were determined for a Gulf Coast con- The equilibrium constants of the plus fractior,Q
densate system containing roughly 70 mole per cent often behave in a manner different from the other
of methane and 21 per cent of hexanes plus material. components of a system. This is because the plus
The system studied by Standing and Katz" also fraction in itself is a mixture of compounds, and both
was a condensate system. It contained about 83 its vapor and liquid phase composition can vary
mole per cent of methane and 6 mole per cent of with pressure.
hexanes plus. The constants of both systems are In Figure 29 are plotted the specific gravities and
in good agreement at 1000 psia and, on the basis of molecular weights of the heptanes plus fraction 32
other data, are expected to continue in agreement at whose equilibrium constants are shown in Figure 25.
lower pressures. However, the wide differences in At 4000 psia, for example, t.he equilibrium was be-
the constants of the two systems at pressures ap- tween 130 molecular weight material in the vapor
proaching 3000 psia reflect the effect of system com- phase and 255 molecular material in the liquid phase.
position. Unfortunately, too few systems have Increasing the pressure to 6000 psia caused the
46 BEHAVIOR OF OIL FIELD HYDROCARBON SYSTEMS
200
i.i 0
~ ~
:i::
100 '
..
::: ~ 80
~
~
C(
i.i ~
~ ~
~ i.i
15:s: !!;
~ l:l ~
~ 'I.
~
ct:
~
i
~
~
. - 8000
0
6000
,f()OO
2000
+
. ,H '
r
. ,f
I
I-
~ .....
~ ~
t:l Cl~
a:
u
~ ~
~
Iii
Q:
\t.!
'
i._
Cl
'41
4
.
~ ~
u
C/)
.... ~
~
:I:
Q:
~
~ ~
~ ~
st
r- _
2000
000
0.05 0.005 0.002 aoo,
MINIMUM K
Figure:32--Correlation of Minimum Equilibrium Constant with the Ratio of Apparent Convergence Pressure to the Vapor Pressure of
the Component
50 BEHAVIOR OF OIL FIELD HYDROCARBON SYSTEMS
plotted at the experimentally deter- two phases can exist. The mathematical equations
mined bubble point pressure. Curves used in making these calculations are derived in the
are sketched in through these values to following section.
give an approximate apparent conver- Flash Vaporization of a System. In all calcula-
gence pressure. tions it is assumed that the system is in thermo-
(c) The locations of the minimum K values dynamic equilibrium at the given temperature and
are determined by use of Figures 31 and pressure. The equations used for the calculation
32, the estimated apparent convergence of the relative proportions of vapor and liquid are
pressure, and vapor pressure values of the obtained from a material balance, as follows:
individual components.
If we let
(d) Smooth curves are drawn that connect
F = total moles in the system
the apparent convergence pressure, the V - total moles of gas at equilibrium
bubble point pressure, the minimum K L = total moles of liquid at equilibrium
values, and the vapor-pressure points. x = mole fraction of any component in the liquid
y = mole fraction of any component in the gas
In the low pressure region the curves are n = mole fraction of any component in the system
given slopes of 45 degrees.
K = 11. = equilibrium constant for any component
X
(e) If the curves do not look right, the ap-
parent convergence is changed slightly
and another set of curves is constructed.
A material balance of the system gives: 0
F=L+V (23)
This is continued until a set is obtained
that has the appearance of those illus- A material balance of any component, k, gives:
trated in Figure 30.
n,F' = x,L + y,V (24)
(f) The methane equilibrium constant curve
is constructed parallel to the unmodified Substituting x.K. for Y in Equation (24) and
curve at pressures below 1000 psia. solving for xk gives the mole fraction of the com-
The curve is brought into the apparent ponent in the liquid phase:
convergence pressure, maintaining the Fn, (25)
form illustrated in Figure 30. Xk = L + VKk
(2) When PVT data are not available to give the At equilibrium, the mole fractions of the compo-
bubble point pressure, apparent convergence nents in both phases must sum to unity, or:
pressures are assumed roughly as follows: k-j
Molecular Weight
of C1-r Fraction
Apparent Convergence
Pressure (psia) l x, = 1 (26)
120
170
3000
6000
-1
k-i
220 9000 l
-1
y, =1
Admittedly, the above method has many faults.
However, in the region of pressures wherein the Applying the requirements of Equation (26) to
system composition influences the equilibrium con- Equation (25),
stants, it is better to use. such a method of taking k? k? Fn,
into account the system composition than to neglect L Xk = L L + VKk = l (28)
k-1 k 1
the effect completely.
A similar equation can be obtained by solving for
Calculation of Vapor-Liquid Equilibrium
the composition of the vapor phase. The equation
The use of equilibrium constants makes possible is
the calculation of bubble point pressures, dew point
(29)
pressures, and the proportions of vapor and liquid in
equilibrium at pressures and temperatures where
VAPOR-LIQUID EQUILIBRIA 51
Equations (28) and (29) must be solved by trial involves the application of Equation (28) to the
and error methods. The calculations are simplified material entering that stage. From experience it
by letting F equal 1, in which case the sum of Land has been found easiest to treat each stage separately
V must equal 1. The procedure is to assume a on the basis of one mole input to that stage. The
value of L or V and calculate values of L +nVK," actual quantities of vapor and liquid are calculated
later from the results of the individual computations.
By plotting the summatioj of L +n}K. against the Bubble Point Pressure of a System. The bubble
point of a system has been defined as the state at
assumed L or V and usi1'g the resulting curve to which an infinitesimal quantity of gas is in equi-
guide the next assumptio*, a satisfactory solution librium with a large quantity of fluid. This defini-
I
TABLE 11.1 EXAMPLE CALCULATION OF EQUILIBRIUM QUANTITIES OF GAS AND LIQUID
I
Pressure = 2000 psia Temperature = 218F
'
1 2 3 4 I 5 I 6 7
I Equilibrum Assume
i Constant at L-0.790
Mole 2000 PSIA& v-0.210
Fraction Com~sition
218'F n of apor
Component n K VK L+VK L+VK y-Kx
C,
C,
C,
C,
0.4404
0.0432
0.0405
0.0284
2.85
1.17
0.680
0.598
0.246
0.143
1.388
1.036
0.933
0.3173
0.0413
0.0434
u
0.410 0.0860 0.876 0.0324
c. I 0.0174 0.244 0.0512 0.8412 0.0207
C,+ 0.0290 0.150 0.0315 0.8215 0.0353
C, 0.4011 0.0160 0.00336 0.7934 0.5055
1.0000 0.9959
Assume
L-0.776 V-0.224
C, 0.638 1.414 0.3114
C, 0.262 1.038 0.0416
C, 0.153 0.929 0.0436
C, 0.0918 0.8678 0.0327
C, 0.0546 0.8306 0.0210
C, 0.0336 0.8096 0.0358
c,+ i
0.00359 0.7796 0.5145
'
1.0006
I Assume
L-0.778 V-0.222
c, 0.632 1.410 0.3123 0.8896
C, 0.260 1.038 0.0416 0.0487
C, 0.151 0.929 0.0436 0.0296
C, 0.0910 0.869 0.0327 0.0134
C, 0.0542 0.8322 0.0209 0.0051
C, 0.0333 0.8118 0.0357 0.0054
c,+ 0.00355 0.7815 0.5132 0.0082
1.0000 1.0000
can be obtained, usually l:Jy the third trial. This tion leads to an equation for calculating the bubble
procedure is illustrated in table 11 by the example point pressure of a system. Starting with Equa-
of calculation of the quant ties and composition of tion (29), it can be seen that as the vapor quantity
vapor and liquid in equili rium at 2000 psia and becomes infinitesimal,
218F. wi i
----
PRIMARY TRAP 1
TANK VAPORS
GAS TRAP ~AS
o.suo 0./9ti.,
0.11991
0. 0$60
I
' O.l4JO 0.111115 The accuracy of calcula ted vapor and liquid quan-
ETHAJIE o.,ozz
tities, using equilib rium constan ts and the over-all
I 0.1207
0.()118
O.OOH 0.0504 0.177$
t1UTA11S
system composition, is depend ent on the composi-
o.oon I 0.0121 O.OS/.J
P!/ITANS O.O,J96
NEXANES 0.0019 0.0094
0.04,U
tion of the system, the relative propor tions of vapor
I
HPTAltES PL/15 0.0oz, 0.0/06
T:oooif' Tooix) T7iiiiio
and liquid in equilibrium, and errors in the indi-
/1'0/JMO klOLE.
l$-OC. RAnD-CFII ,,.
0.4111 0.091
90
0.897
0.025
,.,.
z,
vidual equilib rium constan ts. Because of the sev-
$,,45 t;IU~-AJl t I O.ti40
eral variables, it is impossible to make a general
statem ent regardi ng accuracy, but examples of
several individ ual cases will illustra te a method of
evaluat ing the effects of the variables.
The calculationsin Table 14 show the effect that
an error in any one equilib rium constan t can have
on the quantit y of liquid phase compu ted by use of
ELL
EFFLUNT
PRIMARY
TRAP L/0/110 TANK OIL Equati on (28). The dissolved gas system having
al:"".L'IE 0.,1404 0.Jl/0 0.0/09 0.0007 the composition shown in column 2 has been shown
o.o,:n O.OOJII
in Table 11 to form 0.778 mole fraction liquid an 11 \
O.O-OZ 0.01.JS
!......,,Wf" O,o,$5 0.0210
0.0405 0.0#9
-1< O.O,J,o
0.222 mole fraction vapor at 2000 psia and 218F_.1
0.02114 0.0419 0.040J
1lJ."~S
0.DZIII 0.0,0S 0.0299
~*:'VTA/1/ES O.lJl7'4
H:u.s 0.()290 O.O<MS
0.6877
O.OS57
O.llt:JZ
0.0561
0.11$$0 The equilib rium constan ts used to calculate these
quantit ies are shown in column 3. Colum ns 4, 5,
w:.~T,USt+ / ..!..!!_
1.()()()0 1.0000 TJiiiio T.oiiiii,
---
1 2 3 4 I 5 1
- - - -Tank
--
- 0.491 , 0.051 - 0.025
2 3
Pound Mole
4 u
Primary
Trap Equilibrium Assume Oil Per Gallon
Liquid Constant at L=0.844 V =0.156 Mole at 14. 7
Mole 65 PSIA & Compo- Fraction PSIA& Gallons per Mole
Compo- Fraction 120F n nent n 60F Tank Oil
nent n K L+VK y=Kx C, 0.0007 0.129 -
-C,- - 0.1110 60.0 0.0109 0.6530 C, 0.0038 0.104 0.04
C, 0.0339 10.6 0.0135 0.1430 C, 0.0210 0.0958 0.22
C, 0.0489 3.4 0.0355 0.1207 C, 0.0330 0.0821 0.40
C, 0.0419 1.25 0.0403 0.0504 C, 0.0299 0.0724 0.41
C, 0.0281 0.415 0.0309 0.0128 C, 0.0566 0.0641 0.88
C, 0.0485 0.170 0.0557 0.0095 c,+ 0.8550 0.0263 32.53
c,+ 0.6877 0.013* 0.8132 0.0106
1.0000 1.0000 1.0000 1.0000 34.48
Barrels tank oil per pound mole well effluent -
Moles secondary trap liquid per mole well effluent
- 0.582 , 0.844 - 0.491 0.466 ~2 34.48 - 0.383
Moles secondary trap gas per mole well effluent Molecular \Veight c1+- 287
-O.S82 , 0.156 -0.091 Specific Gravity C1+ - 0.9071
* Equilibrium constant of Heptanes( +) assumed to
be 0.15 times that of Heptane.IHI
psia and 218F, a + 1.$ per cent error in the constant. For the particular conditions unde,U
methane constant would. cause a -1.0 per cent study in Figure 36, an error of - 2.5 per cent in the
error in the calculated liquid quantity. To cause value of the methane equilibrium constant results
the same - 1.0 per cent error in the calculated liquid in an error of +40 per cent in the quantity of cal-
requires that the pentane equilibrium constant be culated liquid. An error of +s per cent in the
in error by + 195.1 per cent. same constant results in the calculation of a dew
Results of calculations similar to those presented point state, rather than the correct 0.051 mole
in Table 14 are shown in Figures 34, 35, and 36. fraction liquid content. Errors in the heptanes
The curves of Figure 36.are of interest because they plus constant, while not as drastic in their effect
show that calculations near the dew point pressure on calculated liquid volumes as errors in the
of a condensate system can be influenced greatly by methane constant, also cause large errors in the
both the methane and heptanes plus equilibrium calculated phase quantities.
54 BEHAVIOR OF OIL FIELD HYDROCARBON SYSTEMS
EnEcT o~ ERRORS IN EQUILIBRIUM CONSTANTS ON THE CAI,CULATED QUANTITY OF LIQUID PHASE IN A D1sSOLVED GAS
T.\BLE 14.
SYSTEM
I 2 3 4
I 6
I 8 I 7
I 8
I 9 I 10 11
u
ASSUMED CONDITIONS AT .
2000 PS/A ANO 218F
I/OLE EQUILI/IRIUI/,
w
~
COMPONENT
0.8
FRACTION CONSTANT ,
~
. ~s :i,
.
METHANE 0.4404 I!. BIS
... ETHANE 0.04/SZ / ./7
'
.
~
I PROPANE 0.04011 0.680 . ~
it
...'
.i:
HANE BUTANES o.ol!B4 0.410
~ 0.80 E HA PENTANES 0.0174 0.144 1 '-.J
PR 'ANE
HEXANE$ o.oe,o O.l/10
' L-2 ~
...., ......
; I,;
HEPTANE$ PLUS 0,4011 0.016
HE 1.0000
~ 'ANE
::i
o.;,.. p"' l'1 "!: ~
0 ...
_, ...
0
...'
=t
HEXANE$
Me \)
-2 ~
0
j:: 0.i'
\) _., It
~ ... l)
"'.... O.i'4
_., ...
0
~ IS\)
~
0.1
0.114
f
~o
T'
ASSUM0 CONDITIONS AT
65 PS/A ANO tzor .. f
-111---i-
MOL COU/1.IBRIUM
COMPONENT FRACTION CONSTANT
j:
0.1142 METHANE 0./110 60.0
THANE o:ong 10.6
.j.
-az
.L PROPANE 0.'0489 S.4
tt BUTANES
~PENTANS
0,0419
0:0281
1.2!1
0.4/!I
0.841 TI HEXANES 0.048!1 0./70
HCPTANES PLUS o.,6877 0,01'
1.0000
" .. ' '. I I 1,,1,,,,,,, o, ' .. +
C .
0,840 :i e:
-so -40 -!lo -20 -,o 0 /0 20 !10 40 !10
PER CENT RROR IN EOUILIBRIUII CONSTANT
Figure 36~Effect of Errors in Equilibrium Constants on the Calculated Quantity of Liquid Phase in a Dissolved Gas System
q
;:
t:t~
f ~ ...
ASSUtlED CONO/nONS AT
2650 PS/A ANO /9!Jf" ~
::,
0.10 MOLE EOUILISRIIJII ....
.... COMPONENT
FRACTION
~
CONSTANT
GAS-CONDENSATE SYSTEMS
For purposes of this disc1,1Ssion, a gas-condensate and surface gases, quantity of liquid formed at
system, is one which will exhibit isothermal retro- reservoir and atmospheric conditions as a result of
grade behavior in the temperature range that is of retrograde action, and data from which the decline
interest to petroleum engineering studies. Such a in reservoir pressure resulting from production
system has a critical temperature lower than the of reservoir vapor can be calculated directly. Less
reservoir temperature or the surface temperature at is known of the composition of reservoir vapor and
which it is separated into g11s and liquid. Produc- retrograde liquid during pressure decline than of
tion of a condensate system through surface traps their volumetric behavior. Laboratory experiments
0
usually results in gas-oil ratios greater than 5,000 to determine the composition of the two phases are
cu ft per bbl and tank oil gravity between 45 and laborious and time-consuming and have received
70 APL The tank oils usually are white or only little attention. Calculated compositions and phase
slightly colored. A definition of the. upper value volumes based solely on equilibrium constant data
of gas-oil ratio of a gas-condensate system is arbi- can be greatly in error. Probably the best estimate
trary; usually a system having a ratio greater than of phase composition can be obtained by combining
100,000 cu ft per bbl (0.43 gal per MCF) is con- laboratory-determined volumetric data with pub-
sidered a dry gas. lished equilibrium constant data. The phase com-
Pressure decline in gas-condensate reservoirs re- position so computed will not be greatly in error and
sults in formation of a liquid phase within the reser- will indicate the trend that the reservoir phases will
voir rock pores. During the last fifteen years, the follow.
operation of processing production from gas-conden-
sate reservoirs for the heavier hydrocarbon compo- Laboratory Tests
nents and returning the light components to the The laboratory apparatus used in condensate
reservoir (cycling) has become increasingly popular. studies differs from the apparatus used in dissolved
It often is stated that the purpose of cycling is to gas studies for two reasons:
maintain the reservoir pressure near its original pres-
(1) The dew point pressure of most condensate
0
sure in order to remove the heavier hydrocarbon
systems cannot be detected by a sharp change
components which otherwise would form the retro-
in the pressure-volume relation of the system.
grade liquid. (Bubble point pressure is determined most
The PVT behavior of gas-condensate systems easily by this means.)
that are representative of reservoir fluids can be
determined easily by laboratory tests. Information (2) The liquid phase constitutes only a small part
that usually is determined in the laboratory includes of the total volume in the laboratory testing
equipment. It is necessary therefore to have
compressibility factors of the reservoir vapor phase methods of precisely measuring small liquid
* The figure 5.000 cu ft per bbl ass.umes production of only vapor quantities.
from the reservoir. The production of both vapor and liquid from
a reservoir to give 5,000 cu ft per bbl (or more-) does not necessarily
One of the most satisfactory PVT cells is one with
indicate a condensate system. glass windows to permit the visual observation of
.
"
BEHAVIOR OF OIL FIELD HYDROCARBON SYSTEMS
.:2 , ::,i.:::s i.e.-.: :::e quantity of retrograde liquid at of 250 psia and still have sufficient liquid phase in
::-:"o"":.:.,s :e"" :'::;,n the dew point pressures. The the cell to measure accurately its volume. There-
; .. ,:,, :i :,<. ::: -:r:iich the liquid phase volume is fore, in many condensate systems, it is necessary to
-~~r.-::::c-.: ::: :0c:1ting the interface between vapor use equilibrium calculation methods to evaluate the
, . , ::.:.:.:,: ': ::.ermal methods* is equally satis- effect of surface trapping conditions on the quantity
~;:~:,~ x::e:: ::.~c.id rnlumes greater than two cubic and gravity of tank oil and trap gases. Because of
.'c:c::~e:.;cs i.-,, 'nrnlved. However, the second the inaccuracies of the calculation methods, the
:.,:,, :i :< :s :.::.:e:-ior to a glass cell when an accur- absolute values obtained by this process may not
;.; :c:ce."'~"' :s ,iesired of a system's dew point agree with field measurements, although the trend
--~-,.:.~. ec:1::...-s> the dew point pressure must be shown by the calculation usually is correct.
:~:=c-.: ::. ar::cipolation of a liquid phase volume- Differential Tests. By definition, a differential
:::"o""'.:.~ :~:2 '.J zero liquid volume. process is one in which the over-all composition of
T 1;;. :: ::;:s. F11>h tests are carried out on gas- . the system varies as the process proceeds. The dif-
_,:::,ie::;:.;:.; s:,s:e:n.s to determine volumetric be- ferential tests made on gas-condensate systems are
:-.r~c :{ :::.e s::s:ems under surface trap conditions, based on the assumption that the retrograde liquid
-~ic .-,:.:. :~:::::g conditions, and under reservoir formed in the reservoir rock is immobile. In line
.':c:-:::c:c:s. _'..; pre,-iously defined, a flash test with this assumption, the laboratory test procedure
,.:__-...c::cs ::c:s:.s.::: system composition during the
::;"-""'"":::::Ce :est. The tests are made most often
usually is as follows:
(1) A system having a composition as close as
u
:-:: ~--;:s= ~:-s';:'<J.red from surface samples of trap possible to that of the original reservoir fluid
:...c.:cs :-s-:::c::::::ei in their produced ratios. is constructed in the laboratory PVT cell.
:: ~ :.c: ::.:.=,.:,ilt to measure phase volumes in a The system is brought to the original reser-
:.;;::,:"-':::: ?\--:- cell at the higher pressures corre- voir pressure and temperature or to the
s...::.-==-: ::: -;\--=~ tubing pressures arid to reservoir system's dew point pressure at the reservoir
::~---,c.:.,~. :-::.e ,eight of hydrocarbon material 'in temperature.
::y :< .::.:.:', :::Ce tests can be determined accur- (2) Starting at the dew point or reservoir pres-
;:,:-- l'. :::e :cr::pletion of the tests by measuring sure, equilibrium vapor phase material is
::Y ===::: .s.::,i gra,-ity of the gas and liquid re- removed from the top of the cell. At several
:::,:-;: ::-:::: ::.e c-ell. Often the gas and liquid are points during the process, measurements are
:-,,::,:--~i ':-:r:: :;:e cell under the pressure and tem- made of the cell pressure, the. volume of
~:-:;:-.:..-e ::::,i:'.::ons existing in the field traps. retrograde liquid formed in the cell, and the
':::e:: .:.:s:,; i,:::e. the volumetric data can be given cumulative volume of vapor removed from
.::: :a=.s :i :::=tion volumes, or in terms of vol- the cell. This step simulates the production
="" 2 ,:e: '."i<',.-oir conditions per MCF of surface of vapor phase from the reservoir rock.
:..-s:- ;:,..."c':i. ? :: many purposes a specific volume (3) During the vapor removal process, the tot :.L
:: ~,,;_-s:: :,,:;c -Jbrained from the measured weight volume occupied by the system is kept cu(_)
.i :::.c ,-,:,,=. 1.::d iis measured volume at specified stant. This is equivalent to assuming that
-.-,--,-- , . , :emperatures is more valuable in the reservoir volume does not change during
;---------
.._--=-r ___ ,., .........,Tations than the formation volume its producing life because of water encroach-
::- ~s:::--":C .olume. ment. The assumption, of course, may be
>cs:: .,:,:1::s ieasible to determine experimen- invalid for niany reservoirs, but its invalidity
,': :~'c ,,'\:: oi surface trap pressures and tem-
does not detract from the usefulness of' the
laboratory data because adjustments can be
~:-:;::..-s :r. :':e properties and volumes of surface
made to the laboratory data where the true
: ....--- -,. 3c:e '.imitations of the usual laboratory
reservoir condition is known.
:,..,--:- ~=:..:-c:::e::c oiten make i.t impossible to expand
'.i: ;;..~::: :-l'.:o systems to pressures of the order The fundamental assumption of the differential
tests concerning immobility of retrograde reservoir
:-~.~ : ::-,- ."::i: .i.:!.C. i:s o~ra.tion are described in more detail in
~.::::':'- .:..::.i -..,T 3-.1~ .:1nd Lacey.u liquid is believed reasonable in view of the facts that
GAS-CONDENSATE SYSTEMS 59
most condensate systems reported in the literature pressure equilibrium vapors. The pseudo-critical
yield less than 10 per cent retrograde liquid by pressures were found to be roughly 50 psi higher
volume during pressure decline at reservoir tempera- than true critical pressures of equivalent molecular
tures, and that the effective permeability of reser- weight normal paraffin hydrocarbon. The pseudo-
voir rocks to this phase usually is zero in the satura- critical temperatures were found to be about 30F
tion range of interest. higher than the true critical temperatures of
There is good reason to believe, however, that the equivalent molecular weight paraffin hydrocarbons.
above assumption is not valid in the region of the Mathews, Roland, and Katz 14 found that volu-
well bore. A higher liquid saturation probably metric calculations could be made with increased
exists near the. well bore than out in the reservoir accuracy if the specific gravity, as well as the
because of the larger pressure gradient and the molecular weight, were used in evaluating the
increased fluid flux density near the well bore. pseudo-critical constants of heptanes plus fractions.
Over a period of time the liquid saturation near the The correlation presented by these authors is
well bore may increase to the point where the pressure reproduced in Figure 37 and is recommended for
gradient may overcome the capillary forces retaining use where both the specific gravity and molecular
the liquid in the rock. The liquid phase will then be weight of the heptanes plus fraction are known.
produced in the well concurrently with vapor mate- Where neither specific gravity nor molecular
rial. This local action in the region of the well bore weight is known, it is necessary to assume pseudo-
is of interest primarily to the problem of recombining critical values. A reasonable assumption is that 0
surface samples to obtain a system that is representa- the molecular weight of the heptanes plus is equiva-
tive of the system in the major portion of the lent to that of nonane, which is 128. By adding
reservoir. the 50 psi and 30F found by Standing and Katz
to nonane's true criticals, the pseudo-critical values
Condensate System Behavior in the Single-Phase are 385 psia and 1100F.
Region The compressibility factor, Z, as normally used in
:!
tI 500
' -++
:-H++-1'++++++_,t_H-H--1--f++-
-!-
H-+-+-+-+-H-H-H-~H-H-"f"t-
I I - .;
'
-i--
+-;--r
--!---4--,1---l----,s,-i-,-1-
-1-1--1-1--H-H-ll--l +-++-+--'---++I -
~ -f-+~
!
: I
' ' ,-1.
:J
~ -f-L.. -r-
~
Q.
400
..... o>o -- ~.._.>~
o
+
~ 1--+-H--H-H) '
+I
i::: I
'
(t ~0
(.j FHH'-e-i-l--l-++--1----1--'I . . t+ ~ -+'--!-++L
<:) i-+'1--'H-4-+,+1 l::i,:L ++ ~ _ - -+ 'i
~
...
-1--+--1~~1---1' 1+ ---l-+ +!--LI
,- -'---. '--'. ttt
1--'--'---',-+-+--L
~ -
H-i"+ I '-+++--1-- ~+++++W--~'
. -i+
~l+1-+-1--'--+-H--l--l--l-+_: -
~ 200 : - .H--1 I I : - t+ L . ~-'-+-
I. 5 I I I I I I I
I! I I l l , .I I I ! I 1-fJI
1-+-+-+-+-+-+-+-+-1 RELATION OF MOLECULAR WEIGHT 0.60
t+++++++-H-HAND GRAVITY OF CONDENSATE
._1~'q..+-,1
.. l
GAS GR.
'
0
)
A':>o
'
I. 4t+++-+"t..: 150
~,t
~le,
H-!-Hl;;;iil~ 1 1
1-+-+-+-+~ - . ioo
t+++-+~-~1--f- 4$"'1- $0. _1-1--$5-1-- . . . 60
H-+++l+H-+++++l-l-r':-.A~P,;:,l~-1-=1.-=1.~~ttttL-+....-+,, ,
.
.
I. 3 :1,i
l-++-++-++-+-+-++-++-++-++-++-++-1-1--1-~rJ,.~\:,{,
I
,'
~o~
~,,. :\~ '
,O~ .
I ". v,,/
0
,.,
I '
I.I
/.
,.I-
0
~'
~99(),0,.
20
BBL.
40
!f'f'PP
0
CONDENSATE PER MMCF
I
'
I
84
~1,0 0
/00
CFS
Figure 38-Effect of Condensate Volume on the Ratio of Surface Gas Gravity to Well Fluid Gravity
62 BEHAVIOR OF OIL FIELD HYDROCARBON SYSTEMS
TABLE 15. ESTIMATION OF GAS-IN-PLACE AT ELEVATED PRESSURE equation:
AND TEMPERATURE USING WELL GRAVITY TO EVALUATE
PSEUDO-CRITICAL CONSTANTS
Problem: To estimate the standard cubic feet of gas in place in a (33)
reservoir that has the following properties:
Given: Reservoir pressure 3000 psia
Reservoir temperature 240F where
Rock porosity 30 per cent 'Y~ = gravity of well fluid (air= 1)
Interstitial water 27 per cent
Tank oil production 323 bbl per day
Tank oil gravity 45 API Well fluid gravities calculated by this equation do
Primary trap gas rate 3,765 MCF per day
Primary trap gas gravity 0.65 (air = 1) not differ materially from those calculated by use of
Tank vapor rate 169.2 MCF per day Equation (32) or Figure 38. In the example given
Tank vapor gravity 1.25 (air = 1)
Solution: Basis: One acre-foot sand volume
in Table 15, a well fluid gravity of 0.896 was ob-
Average separator gas gravity tained by Vitter's equation compared to 0.917 ob-
_ 3,765 0.65 + 169 1.25 = O tained by use of Figure 38. Better comparisons are
677
- 3,765 + 169
323 obtained at high gas-oil ratios and high O API con-
Bbl condensate per MMCF = 3.934 = 82.1 densate gravities.
. well fluid gravity Empirical Behavior in the Single-Phase Region.
From Figure 38, t rap gas grav1'ty = 1.35
Sage and Olds' correlated the volumetric behavior
Well fluid gravity - 1.35, 0.677 = 0.917
From Figure 17, pTc = 430 R, pPc = 650 psia
offive condensate systems produced from CaliforniaQ
460 + 240 700 3000 fields, The producing gas-oil ratios of the systems
pT, = 430 - 430 = 1. 63 pP, = 650 = 4 62 were in excess of 5,000 cu ft per bbl, and the tank oil
From Figure 15, Z = 0.840 gravities were between 52 and 63 APL An em-
The reservoir hydrocarbon space per acre-foot of rock is 43560
0.30 (1 - 0.27) = 9,540 cu ft pirical relation was determined which expressed the
Total moles of hydrocarbons per acre-foot at reservoir conditions, formation volume of the system as a function of
PV 3000 9540 pressure, temperature, and gas-oil ratio. The
N = ZRT - 0.840 . 10.73 . 700 = 4540 lb moles
relation is
Now, if all this quantity of hydrocarbons were produced as gas
at the surface, it would represent (34)
4540, 379 = 1730 MCF per acre-foot
However, as a portion will be produced as tank oil, it is necessary where
to make allowance for this fact. One barrel of 45 API tank A = empirical coefficient, as tabulated in Table 16
oil (0.802 specific gravity) having a molecular weight of 140 P ... pressure (psia)
would correspond to T = temperature ( R) 0
5.615 62.37 0.802 = ,O1 lb mo les r = gas oil ratio 1 cu ft per bbl of tank oil
2 JI, = formation volume (bbl per bbl of tank oil)
140
At the present producing ratio of 12, l 70 cu ft per bbl, the total
surface gas amounts to TABLE 16. VALUES OF COEFFIC,ENT A USED TO CALCULATE
FORMATION VOLUME OF GAS-CONDENSATE SYSTEMS,
12,170
379 = 32.1 moles METHOD OF SAGE AND 0Los 8 *
Theoretically, A is not a function of pressure and volume calculation in the example shown in Table
temperature only, as indicated by Table 16. The 18 normally is carried out by use of the calculation
gas-oil ratio, the gravity of the gas, oil gravity, and chart.
molecular weight are factors which cause minor
variatibn in the value of A. Notwithstanding the
apparent neglect of the above factors, the method
gives formation value factors that follow the volu-
metric behavior of the California systems with
,
errors no larger than those obtained when -0ther ,, , . ,
methods ,are used. An example of use of this
method is shown in Table 17.
#) (It ..: 1// l?,
.,,.__
An empirical correlation of the gas plus liquid
~
phases formation volume has been made by Stand-
ing. 7 This correlation is satisfactory for dissolved
TABLE 17. ESTIMATION OF GAS-tN-PLACE AT ELEVATED PRESSURE
y j':, ',',
. / #000 ,$IA
""
'-60001'#A
...
AND TEMPERATURE USING METHOD OF SAGE AND 0LDS 8 ,.,.,.
Problem: To estimate the standard cubic feet of gas in a reservoir
having the following properties:
,. ,. .,. - .
Given: Reservoir pressure
Reservoir temperature
3000 psia
240'F
',.
,.
w /
0
Porosity 30 per cent I
Interstitial water 27 per cent
Tank oil production
Primary trap gas rate
Tank vapor rate
323 bbl per day
3,765 M CF per day
169.2 MCF per day
,. ... l
,..
r llLu
...
t4_1"10-0.0001rr
,f $ .,.~
psia
iI
4S
Reservoir temperature
Rock porosity
240
30
F
per cent ~
Interstitial water 27 per cent ~ 40
Tank oil production 323 bbl per day iiiI\
Tank oil gravity
Primary trap gas rate
45
3,765
AP!
MCF per day ... ssoo
Primary trap gas gravity 0.65 (air = I} ~
Tank vapor rate 169.2 MCF per day ~
Tank vapor gravity 1.25 (air = 1) lit !1000
Solution: Basis: One acre-foot sand volume
AveragEl separator gas graVity
i!!
3,765 0.65 + 169 1.25
- 3765 + 169 . = 0 677 I0,000 l!0,000 .JOp(JO
Producing gas-oil ratio 6AS-O/L RATIO - CU FT PCR BBL
= (3765 +323169)(10') = 12,170cuftperbbl Figure 40-Influence of Gas-Oil Ratio and Tank Oil Gravity upon
Retrograde Dew Point Pressure at 160F (After Sage
(t)U.S -0.00027r and Olds, Trans. AIME, 1947)
r - - (li1)2.9X10
(-y~)0,3
(240)o.6 -00002712170 the properties of the 8000 cu ft per bbl system. "j
- 12li0 (0.677)'' (0.8017)',x" .
On the other hand, it is not inconceivable that the\
1~5
12170 0.889 0.999 = 2.125(10 5) actual reservoir vapor might be the equivalent
of the 12,000 cu ft per bbl system and that the
From Figure 39, formation volume at 3000 psia = 13.5 bbl per
bbl of tank oil 8000 ratio was a fortuitous result of simultaneous
One acre-foot reservoir rock volume has production of vapor and retrograde reservoir liquid
l into the well bore. Under such circumstances, the
4:J560 ,0.30 , (I - 0.27) , . - liOO bbl of hydrocarbon
5 615
use of the phase behavior of the 8000 cu ft per bbl
space
1700 system to predict the reservoir fluid beha,dor could
Therefore, tank oil volume = . = 127 bbl per acre-foot result in errors in both the correct formation
13 5
surface gas volume = 127 12170 = 1534 MCF per acre-foot volume of the reservoir system and the quantity of
retrograde liquid that would be formed as a result
the correlation should be applied with caution to of reservoir pressure decline.
other fields. Because of the domelike shape of the gas-oil ratio-
The basis for the 160F data, presented in Table dew point pressure curve, one should not place as
19, is shown in Figure 40. The curves illustrate a much emphasis on laboratory dew point pressure
behavior of condensate systems, the understanding measurements as is usually placed on bubble point
of which is important in the correct application of pressures. As will be shown in Chapter 7, there
laboratory results to field conditions. The fact that is a monotonic relation of bubble point pressure wit!()
the curves show maximum dew point pressure gas-oil ratio, so that simultaneous flow of free gas
Yalues makes it possible for a laboratory system to and saturated liquid into the well bore will always
have a dew point pressure equal to the reservoir cause the bubble point pressure of the recombined
pressure, but not necessarily to have the same com- dissolved gas system to be higher than the bottom-
position as the reservoir vapor. For example, the hole pressure. On the other hand, simultaneous
uppermost curve shown in Figure 40 exhibits a dew flow of retrograde reservoir liquid and saturated
point pressure of 4700 psia over a gas-oil ratio vapor into the well bore may result in a condensate
range from 8,000 to 12,000 cu ft per bbl. If the system having a dew point pressure lower than the
producin.; gas-oil ratio at the time of sampling were bottom-hole pressure and thus cause one to believe
8000 cu ft per bbl and the reservoir pressure were the reservoir fluid to be undersaturated.
4700 psia, the data could be interpreted as indicating Liquid Phase 'Behavior. The most characteristic
the reservoir fluid to be at its dew point and to have curve of gas-condensate systems relates the quantity
GAS-CONDENSATE SYSTEMS 65
TABLE 19. RELATION OF DEW POINT PRESSURE OF CALIFORNIA CONDENSATE SYSTEMS. DATA OF SAGE AND 0LDSs
1 2 I 3 I 4 I 5 I 6 I 7
Tank
Oil Gas-Oil Ratio, Cubic Feet per Barrel
Gravity
0
API 15,000 I 20,000 I 25,000 I 30,000 I 35,000 I 40,000
!OO"F
52 4,440 4,140 3,880 3,680 3,530 3,420
54 4,190 3,920 3,710 3,540 3,410 3,310
56 3,970 3,730 3,540 3,390 3,280 3,180
58 3,720 3,540 3,380 3,250 3,140 3,060
60 3,460 3,340 3,220 3,100 3,010 2,930
62 3,290 3,190 3,070 2,970 2,880 2,800
64 3,080 3,010 2,920 2,840 2,770 2,700
160"F
52 4,760 4,530 4,270 4,060 3,890 3,650
54 4,400 4,170 3,950 3,760 3,610 3,490
56 4,090 3,890 3,690 3,520 3,380 3,270
58 3,840 3,650 3,470 3,320 3,200 3,110
60 3,610 3,430 3,280 3,150 3,040 2,960
62 3,390 3,240 3,100 2,990 2,890 2,810
64 3,190 3,060 2,930 2,820 2,740 2,670
220F
0
54 4,410 4,230 4,050 3,890 3,750 3,620
56 3,990 3,780 3,600 3,440 3,300 3,180
58 3,700 3,480 3,280 3,110 2,970 2,850
60 3,430 3,210 3,030 2,880 2,760 2,660
62 3,150 2,970 2,800 2,670 2,570 2,480
64 2,900 2,740 2,590 2,470 2,380 2,300
"0
@
tween volume per cent liquid and pressure for a flash
process; the approach to the origin is merely an ex-
-7 / ~
tl
~
....
pression of the large total volume necessary to reach ,
/ @-
~ \ 'Xi> ~
;I; )
a low pressure for a system of constant over-all /
\\ t
""
composition. /
The curves illustrated by B and C are charac-
,,/ ~ [:- \
teristic of differential processes in which the total
volume of the system is maintained constant during
the pressure reduction. Temperature and the
PRESSURE
initial composition of the system will affect the dew Figure 41-Curves Illustrating Several Types of Liquid Behavior
of Condensate Systems
point pressure and the quantity of liquid at any
pressure. In general, a system that has a surface whereas a 50,000 cu ft per bbl system ordinarily
producing a gas-oil ratio of 15,000 cu ft per bbl will will give less than 1 per cent liquid. Curves B and
give a maximum liquid content of 4 to 6 per cent C illustrate differences in liquid volume behavior
by volume at reservoir temperatures near 200F, near the dew point pressure. While insufficient
66 BEHAVIOR OF OIL FIELD HYDROCARBON SYSTEMS
data are available in the literature to explain these curve represents the instantaneous value of the
differences, it is believed that the asymptotic ap- vapor phase gravity. As will be shown in the next
proach to the dew point pressure of curve C is section, the calculated quantity of butanes plus
caused by a system that has a wider range of
heavier components than a system that gives the B
curve.
Where actual volumes of liquid formed during
....:i!
I.DO
~'
ferential process proceeds. In natural systems, the
maximum amount of liquid occurs in the pressure
,o
range 1200 to 2500 psi for both the flash and differ- ...
ential processes. At lower temperatures larger tl ~
amounts of liquid are formed and maximum liquid
occurs at a lower pressure.
Ii o . ~o ~~
!S .,.R.O. a,.
Vapor Phase Behavior. Just as the volume of
!~ 40 .~~
liquid formed by a condensate system goes through '"Ii! Ill>
~~
a maximum near 2000 psi, the quantity of heavy
components held in the vapor phase goes through
a~.I OINT I'll SSV
+
I~~
~
. l,IO P$IA AT JgsF
a minimum, usually at about 1000 psia. This
~
. I
!
behavior of condensate systems is illustrated quite 0
0 ' !100 1000 l/100
~
l!OOO 1!100
0'
clearly by the results of th.e differential tests shown l'R$$UIIE -I'S/A
in Figure 42. The percentage of the system that Figure 42-Behavior of 2960 Psia Dew Point System at 196F
"""'
has been produced is related to the resulting decline Caused by Differentia~ly Removing Equilibrium Vapor ,
in system pressure, as shown by curve A of Figure material in the. vapor phase decreases from 5.2 gal
42. For example, 20 per cent of the system avail- per MCF at the dew point to 2.6 gal per MCF at
able in tl)e vapor form, between the dew point 1000 psia.
pressure of 2960 and 14.7 psia, is produced during
pressure decline from 2960 to 2370 psia. The Phase Compositions
relation presented in curve A can be used directly Laboratory tests on gas-condensate systems usu-
with field production data to calculate reserves ally are made to determine the volumetric behavio{)
where it is known that the reservoir hydrocarbon of the systems at reservoir and surface trapping
space does not change because of reservoir pressure conditions. For special engineering studies, how-
decline. Curves of the type illustrated are char- ever, one may need to know the phase compositions
acteristically S-shaped because of changes that at various pressures during the depletion of a con-
occur in the compressibility factor and system com- densate reservoir. The composition of each phase
position as pressure declines. can be determined experimentally, but to do so
Curve B of Figure 42 shows the amount of retro- requires more work than often is believed justified.
grade liquid formed during the differential test. Methods of calculating phase compositions and
An indication of the richness of the vapor phase at volumes by use of equilibrium constants and the
any pressure is given by curve G, Figure 42. The composition of the system already have been out-
horizontal line segments represent measured average lined in Chapters 3, 4, and 5. In gas-condensate
gravities of the vapor removed from the pressure cell systems, however, the calculated phase compositions
in the indicated pressure intervals. The smooth and volumes, in particular the liquid phase volume,
GAS-CONDENSATE SYSTEMS 67
can be greatly in error through use of present known TABLE 20. CALCULATION OF THE ORIGINAL AND F1NAL
QUANTITIES OF MATERIAL REMAINING IN THE UNIT
equilibrium constants. It becomes advisable, there- RESERVOIR VoLUME
fore, when making such calculations, to control Basis: 100 cu ft hydrocarbon space
them by use of experimentally determined volume Initial dew point pressure 2960 psia
Final pressure 14.7 psia
measurements. The remaining portion of this Temperature 195F
chapter will discuss the steps required in such cal- Initial condition at 2960 psia:
culations. The example given pertains to an actual 1 2 3 I 4 5 I 6
gas-condensate system that has a dew point pressure
P'l',udo
of 2960 psia at 195F and the composition shown in Critical Cntical
Constant of Temperature
Table 20. The results of laboratory differential Mole
Compo- Fraction Components and Pressure
tests to determine the volume of liquid phase and nent n Tc Pc nTc nPc
the pressure reduction-vapor withdrawal relation C, 0 7527 344 673 258.7 506.0
have been given previously in Figure 42. c, 0.0766 550 709 42.1 54.3
c, 0.0441 666 618 29.4 27.2
As a starting point in the calculation, let it be C, 0.0309 750 540 23.2 16.7
assumed that a reservoir hydrocarbo n space of 100 c; 0.0221 838 484 18.5 10.7
C, 0.0206 915 434 18.8 8.9
cu ft represents the unit working volume of the sys- C,+ 0.0530' 1042 418 55.3 22.2
tem in which we are interested. It will be assumed 1.0000 446.0 646.0
that this volume remains constant throughout the "' C 7+ molecular weight = 114
calculation process. The calculations in Table 20 c?+ specific gravity = 0.755 (.)
determine that 54.0 moles constitute the original 2960 = 4.58
Pp R - 646.0 pTR = 655 = 1.47
446.0
system at 2960 psia and 195F. The calculations Z = 0.78 (Figure 15)
also show that 2.1 lb moles of material remain in the N PV 2960 100 lb l
= ZRT = 0.78. 10.73. 655 = 540 mo es
100 cu ft volume at 14.7 psia and 195F. The dif- Final condition at 14. 7 psia:
ference between this value and the original (51.9 lb Specific gravity of cell residue = 0.753 60/60 (Fig. 42)
moles) represents the quantity of material removed From0.753 Figure 25, density of cell residue at 14.7 psia and 195F =
62.37 - 3.6 - 43.4 lb per cu ft
during the differential process. From Figure 38, assumed molecular weight of residue = 113
The second step is to determine a set of equilib- From Figure 42, volume of residue liquid at 14.7 psia and 195F =
4.9 cu ft
rium constants with which to calculate phase com- 4 9 43 4 1
positions. An immediate test of the applicability
Quan ftI y of res1"d ue l"1qm"d = .
113
= .88 lb moles
of any set of constants is that the sum of n /K must Quantity of residue gas = (l00 ; 4.9) ::~ = 0.20 lb moles
3 9
be equal to unity (Equation 31) at the observed Total quantity of residue gas and liquid =
dew point pressure of 2960 psia. In Table 21 are 1.88 + 0.20 = 2.08 lb moles
presented three dew point pressure calculations Material removed from the unit volume in going from 2960 to
14.7 psia = 54.0 - 2.1 - 51.9 lb moles
TABLE 21. DETERMINATIO N OF RELATIVE EQUILIBRIUM CONSTANTS TO COMPUTE DEW POINT CONDITIONS
Basis: Roland, Smith, and Kaveler's Data 200F
k-j
\' n,1; = 1
L, K,
k-1
1 2 3 4 5
I I 6 7
I 8
Pressure Pressure Pressure
Mole 2960 PSIA 3500 PSIA 3600 PSIA
Fraction n n n
Component n K R K R K R
C, 0.7527 1.67 0.4500 1.35 0.5575 1.30 0.5795
C, 0.0766 0.92 0.0830 0.96 0.0799 0.97 0.0790
C, 0.0441 0.69 0.0649 0.82 0.0538 0.85 0.0519
c. 0.0309 0.52 0.0594 0.71 0.0435 0.76 0.0407
C, 0.0221 0.34 0.0650 0.54 0.0410 0.60 0.0879
C, 0.0206 0.23 0.0896 0.44 0.0468 0.50 0.0412
c,+ 0.0530 0.095 0.5580 0.28 0.1892 0.35 0.1514
1.0000 1.3690 1.0117 0.9816
Dew Point Pressure = 3540 psia
68 BEHAVIOR OF OIL FIELD HYDROCAR60N SYSTEMS
usmg Roland, Smith, and Kaveler's constants for been drawn to connect the pressures at which the
200F. (It is not justified at this point to work individual constants are equal to unity (vapor
out constants corrected to 195F.) These constants pressure), the minimum constants, the constants at
satisfy the condition that the summation of nk /Kk the experimental dew point pressure, and a con-
is equal to unity only at 3540 psia. As the system vergence pressure of 3500 psia. Roland, Smith,
under study gave an observed dew point at 2960 and Kaveler's curves at 200F also have been
psia, Roland's 3540 psia values must be used at plotted in Figure 43 to aid in drafting the 195F
2960 psia pressure. Thus, the published constants curves. The curve constructed through the solid
are used here primarily for their relative values and points will be explained later.
not for their absolute values. The fourth step is to establish a calculation proce-
IThe methods presented in Chapter 5 are used dure to simulate the differential process of producing
next to estimate equilibrium constants at pressures equilibrium vapor from the unit volume. A simple
below the dew point pressure. Assuming an ap- stepwise calculation process is diagrammed in Fig-
parent convergence pressure of the system several ure 44. This calculation process can be visualized
hundred psi above the dew point pressure makes it
possible to calculate by use of Figures 31 and 32 I II III .N ]t
-~
~I!; . I 0 HA
pressure reduction, rather than before. The com- ~ I
position of the vapor and liquid phases calculated "':i:
HA
by this second process does not differ materially
from that calculated by the process just described. ~i 1000
500 1000 /500 2000
PR/ISSI/Re-PS/A
2500 3000
Of course, the greater the number of steps, the
closer the two methods will agree with each other
;;
b~
and, for an infinite number of steps, each process
becomes a true differential process. I!; 1 HA
()
The PVT data presented in Figure 42 are in terms
"'~
It.. j
a75
of volumes, whereas equilibrium constant calcula-
tions make use of molal phase quantities. To con-
vert the cubic feet of liquid phase calculated from
Figure 42 to pound moles of liquid, it is necessary
. a7o
0 500 K)(iJ() /500
PR/ISSI/Re-PS/A
2000 2500 3000
to know the molecular weight and density of the Figure 46-Assumed Relationship of Molecular Weight and
liquids as they exist at the elevated pressures and Specific Gravity of Heptanes Plus to Pressure
195F. These values can be determined by the
methods presented in Chapter 4, provided the in vaporization calculation work: the behavior of
pressure-density and pressure-molecular weight re- the heptanes plus fraction at different pressures.
lations of the heptanes plus fraction are known. The composition of the dew point liquid at 2960
The only direct laboratory data available are the psia and 195F is shown in Table 22. The method
molecular weight and specific gravity of the total of obtaining the equilibrium constants listed in col-
heptanes plus in the original system and the specific umn 3 has been discussed previously in connection
gravity of the residual liquid remaining in the lab- with Figure 43. The properties of the heptanes plus
oratory cell at the completion of the differential fraction required to calculate its volume and weight ..
process. These data have been used with the trend were obtained from Figure 45. Volumes of the indi- U
of the curves in Figure 29 to prepare the assumed vidual components listed in column 6 were obtained
pressure relations shown in Figure 45. The meas- by use of the atmospheric liquid densities of Table 1.
ured values of 114 and 0.755 are shown as corre- An arithmetic average of the liquid densities of iso-
sponding to the dew point vapor. Inasmuch as and normal butane and pentane was used in the cal-
the residual cell liquid at 14. 7 psia and 195F is culation. It is interesting to note that almost 20
of the order of 85 per cent heptanes plus material, weight per cent of the dew point liquid is methane.
its measured specific gravity of 0.753 has been cor- The calculation of vapor and liquid cdmpositions
rected upward to correspond to a heptanes plus at 2650 psia is shown in Table 23. In part B the
specific gravity of 0.760. From this gravity, the number of moles of liquid and vapor in equilibrium
molecular weight of the heptanes plus in the residue is calculated from the total moles (48.81) comprising
liquid is estimated to be 117. Unfortunately, these the first modified system and the quantity of liquid
are all the data available from which to draw the phase determined by the. laboratory tests. To con-
70 BEHAVIOR OF OIL FIELD HYDROCARBON SYSTEMS
vert the measured liquid volume to pound moles, it (5) The mole fractions of the components in the
is assumed that the factor, pound moles per cubic vapor phase are made to add to unity by
foot, evaluated at 2960 psia (Table 22) also will minor adjustments in the mole fraction of the
apply at 2650 psia. This assumption is checked methane.
TABLE 22. CALCULATION OF PROPERTIES OF DEW POINT LIQUID AT 2960 PstA AND 195F
Basis: One pound mole of liquid
1 2 3 4 5 6
Licid Volume
Weight of of omponents
System Equilibrium Liquid Components at 14.7 PSIA
Composition Constant at Phase lb per lb & 60"F
Mole 2960PSIA& Composition mole liquid cu ft
Fraction 195F n nM nM
Component n K K K Kp
C, 0.. 7527 1.32 0.5720 9.15
C, 0.0766 0.96 0.0799 2.40
C, 0.0441 0.83 0.0531 2.84 0.074
C, 0.0309 0.74 0.0418 2.43 0.068
C, 0.0221 0.57 0.0388 2.80 0.071
C, 0.0206 0.47 0.0438 3.77 0.091
c,+ 0.0530 0.31 0.1706 23.55' 0.472'
1.0000 1.0000 46.44 0.776
"'Assumed molecular weight of C1+ = 138 (Figure 45)
r._)
Assumed specific gravity of C7+ = 0.800 (Figure 45)
Density propane plus at 14.7 psia and 60F = ~~;~! - 45.7 lb per cu ft
2.40
Weight per cent ethane in ethane plus = .
37 63
= 6.3 per cent
. . 9.15
Weight per cent methane in system = 46 .44 = 19.5 per cent
From Figure 23, pseudoliquid density at 14.7 psia and 60F = 31.3 lb per cu ft
From Figure 24 1 density tit 2960 psia and 60F = 31.3 + 2.7 = 34.0 lb per cu ft
From Figure 25, density at 2960 psia and 195F = 34.0 - 9.3 = 24.7 lb per cu ft
and, if necessary, is corrected after the composition The procedure just outlined assumes that all equi-
of 2650 psia liquid has been calculated in part B. librium constants are correct, except that of the
Part B of Table 23 outlines the vaporization cal- heptanes plus fraction. This assumption, of course,
culation at 2650 psia, using the relative values of L may not be strictly valid, but it presents a workable
and V determined in part A. The following proce- method of calculating the phase compositions.
dure has been followed in making these calculations: The net effect of throwing all equilibrium constant, ..
errors into the heptanes plus constant is shown bL)
(1) With the exception of heptanes plus, the mole the solid points in Figure 43. If the methane con-
fraction of each component in the liquid stant curve had been drawn slightly higher, so that
phase is calculated by the usual methods.
its value at 2650 psia was 1.60 rather than 1.48, the
(2) The mole fraction of heptanes plus in the calculated equilibrium constant of the heptanes plus
liquid phase (0.2059) is set at the value fraction would have been 0.173 rather than 0.217.
necessary for 2:x = 1. This would have caused better agreement between
the calculated heptanes plus curve and the curve
(3) Working backward from the values of x,+
previously obtained by consideration of the ap-
and n,+, the values of L + VK 1 are calculated.
parent convergence pressure. However, unless a
(4) The mole fraction of heptanes plus in the much wider discrepancy is obtained than that illus-
vapor phase, y,+, is calculated from the values trated in Figure 43, it usually does not pay to
of K,+ and x,+, using the relation y = Kx. undertake further refinement of the curve.
GAS-CONDENSATE SYSTEMS 71
The calculations in part C of Table 23 are re- Figure 44. The remaining moles of
material con-
quired to check the assumed value of 0.53 lb moles stitute the second modified system
that will be
per cu ft of liquid used in part A. The agreement expanded to 2370 psia.
of 0.53 with 0.536 is considered satisfactory. Had a The results of the ten-step calculation process are
check not been obtained, it would have been neces- shown graphically in Figures 46 and 47. The
com-
sary to recalculate parts A, B, and C. position, density, and molecular weight of the liquid
The calculations in part D of Table 23 correspond phase at various pressures are shown in Figure
46.
to the removal of 5.19 lb moles of 2650 psia equi- As might be expected, the quantity of
dissolved
librium vapor, as pictured by condition IV of methane continuously decreases as the
pressure on
TABLE 23. CALCULATIO N OF VAPOR AND L1Qurn COMPOSITIO N AT 2650 PsIA
Part A: Estimation of Vapor to Liquid Ratio
Basis: 48.81 lb moles of system
From calculations in Table 22, assume that the liquid phase will have the equivalent of 0.53 lb moles
per cu ft.
From Figure 42, volume of liquid at 2650 psia is 5.2 cu ft. Therefore, moles of liquid = .5.2 0.53
Mole fraction liquid, L = !:.: = 0.056: Mole fraction vapor, V
Part B: Calculation of Vapor and Liquid Composition
= 0.944
= 2.76.
System
2 3
Equilibrium
4
L - 0.056
5 6
Liquid Volume
of Component s
specific gravity C 7+ = 0.790
. and
Density propane plus at 14.7 ps1a
cu ft
60F = 39.86
_
0 890
= 44.81 b per u
Liquid Phase Weight of at 14.7 psia.
Compositio n
Mole
Compo-
nents
&60'F
cu ft
Weight per cent ethane in ethane plus = 422.42
_ = 5.7
28
Compo- Fraction lb xM . .
nent X xM -p Weight per cent methane m system = &31 _ = 16.4
50 59
From Figure 23, pseudo-liquid density at 14.7 psia and 60F =
c, 0.5180 8.31 33.2 lb per cu ft
C, 0.0804 2.42 From Figure 241 density at 2650 psia and 60F = 33.2 + 2.0 =
C, 0.0531 2.34 0.074 35.2 lb per cu ft
C, 0.0475 2.76 0.076 From Figure 25, density at 2650 psia and 195F = 35.2 - 7.9 ==
C, 0.0442 3.19 0.082 27.3 lb per cu ft
C, 0.0509 4.39 0.106
c,+ 0.2059 27 .18' 0.552 Pound moles liquid per cu ft liquid ""' ~: = 0.539
1.0000 50.59 5 9
0.890 (Assumed value of 0.53 in Part A is satisfactory .)
72 BEHAVIOR OF OIL FIELD HYDROCARBON SYSTEMS
T.\BLE 23-(f'ontinwd)
l)arl I>: ('11/r-11/11/io11 of .l!11diJil'd Sy.~t,m :it 2:{70 Psia
Basis: .\8.81 lh mo[Ps of matPrial nt 2H:i0 psia
;).1\1 !Ii mo!Ps of 21\50 wda \a11nr n'llHl\'('cl
I 2 :1 I 5 n 7
-
'
--- I I
-- ----
i - .
u
~
..~
i
I
~
~ <>
..i
I ~
I
I ~
I
!!
I"
~
~
......
i'I
""'o
- /000 /$00
Pll$$IHl-l'!flA
- "" !1000
Ito
...
~
~
~
I:
!.
I
,o '
I'
'
II ~
~
I
J.1 : ! J
+
I
j r
I
.~ t: ~
~
ilt ''
i:r
. i
o. ,f
~
Ii , I ~ ,1
,,c'", ~
.
I!
o.
~
l
- - I
..,. I
! ....i
:
100
/000
"""
fli/lE$$UIIC-P$IA """ ""
JOOO 0.10 !JOO 1000 /!JO() 2000 2!100 :,000
PRESSURE-PS/A
Figure 46-Calculated Composition and Properties of Liquid Phase Figure 47-Calculated Composition and Properties of Vapor Phase
,
GAS-CONDENSATE SYSTEMS 73
the system is reduced. Likewise, the proportio n of relation of the critical temperat ure of the system to
heavy compone nts increases with decreasing pres- the reservoir temperat ure.
sure. Intermed iate compone nts tend to have maxi- The volumetr ic behavior of gas-cond ensate sys-
mum concentra tion in the liquid phase at pressures tems can be calculated easily by using the pseudo-
near the pressure of their minimum equilibriu m critical constant method proposed by W. B. Kay. 12
constant. For example, the maximum butane In many instances the pseudo-critical constants of a
concentra tion in the liquid is near 850 psia, and the series of gas-condensate systems can be plotted
equilibriu m constant of butanes (Figure 43) is a against the gravity of the system; calculatio n charts
minimum at that pressure. The behavior of the then can be construct ed that will give the volume
vapor phase is shown in Figure 47. It will be noted of the system as a function of pressure, temperat ure,
that the calculate d minimum gravity of 0.780 at and system gravity. Figure 38 presents an easy
1000 psia does not check the laborator y test value way of calculatin g the gravity of a well fluid from
of 0. 75 shown in Figure 42. This may be caused in the quantitie s and gravities of the resulting _surface
part by the high value of the heptanes plus equi- gases and tank oil.
librium constant. A second factor that may con- Naturally occurring gas-cond ensate systems, in
tribute to the discrepancy is the assumed molecular general, contain greater than 75 mole per cent
weight of the heptanes plus vapor. methane. The volume of the liquid phase in equi-
It can be appreciat ed from the foregoing section
that calculate d compositions of the vapor and liquid
librium with vapor under reservoir conditions sel-
dom is greater than 10 per cent of the total volume
u
phases in a gas-cond ensate system may be in error occupied by. the system. At pressures of the order
because of any one of the numerous assumptio ns of 3000 psia or greater, the retrograd e liquid in a
that have to be made in the calculatio n process. reservoir may contain up to 40 per cent liquid
The method outlined above, where the volumetr ic methane by volume. The density of the liquid
behavior, of the system determin ed experime ntally under these conditions often is as low as 25 lb per
in the laborator y is used as a control in the calcula- cu ft.
tion process, is believed to be the most satisfacto ry The quantity of butane plus material held in the
one available. The accuracy of these calculations vapor phase usually is at a minimum near 1000 psia.
can be expected to improve as additiona l data be- This means that productio n from high pressure con-
come available on the behavior of condensate densate fields should follow an increasing gas-oil
systems. ratio trend until the reservoir pressure declines to
near 1000 psia. The producing gas-oil ratio nor-
Summary of Principal Points mally will decrease at pressures less than 1000 psia.
Calculati on of phase compositions is made best by
Gas-cond ensate systems are identified by their using experime ntally determin ed phase volumes to
high gas-oil ratios and lightly colored, high-grav ity control the equilibrium calculations. However, it
tank oils. The critical temperat ure of the naturally
occurring condensa te systems is lower than the
should be recognized that the calculate d composi-
tions can be in error beca1Jse of using nonrepre senta-
\J
temperat ure of the reservoir in which they are tive equilibriu m constants and incorrect assumpti ons
found. The systems normally exhibit retrograd e regarding the pressure-density and pressure-molecular
behavior at pressures above 2000 psia because of the weight relations of the heptanes plus fractions.
CHAPTER 7
Most naturally occurring hydrocarbon systems various pressures and temperatures, the shrinkage
behave in a manner that involves bubble points. In of the liquid hydrocarbon phase when gas is removed
terms of multi-component system behavior dis- from solution under both reservoir and surface
cussed in Chapter 1, such a system has a critical trapping conditions are determined routinely in
temperature greater than the temperature at which many laboratories.
it exists in a reservoir. To have a critical tempera- This chapter will describe first the apparatus anclu
,
ture of 200F or higher, it is necessary for the system methods used in making laboratory tests. Dat
to contain a large amount of the heavier hydrocar- characteristic of the behavior of dissolved gas sys-
bons. Such systems, when produced through sur- .terns under surface trapping conditions will be pre-
face trap settings, will give gas-oil ratios less than sented in the second section of the chapter. The
2500 cu ft per bbl.* The tank oil normally is third section deals with the behavior of dissolved
deeply colored, and the gravity may be as high as gas systems at reservoir pressures and temperatures.
45API. The last section will describe methods of adjusting
Pressure decline in dissolved gas-type reservoirs the laboratory data to apply more fully to field
causes a gas phase to be formed within the reservoir conditions.
rock pores. The mechanism of oil production
caU'sed by gas coming out of solution and expanding Laboratory Tests
is termed a "dissolved" gas drive, or a "solution"
gas drive, and usually is thought of in connection Apparatus. The essential pieces of laboratory
with reservoirs having no original free gas phase. A equipment consist of a pressure cell in which equi-
related producing mechanism exists in the oil zone of librium can be obtained between gas and liquid
a reservoir initially containing free gas phase ac- phases, a controlled temperature bath with a range
eumulated in a gas cap. While the pressure- from 60 to 300F, a pressure-measuring system, and
production behavior of a reservoir under a dissolved a calibrated pump for injecting mercury into thP
1
gas drive is different from that resulting from an ex- cell in order to change the volume occupied by tliJ
panding gas cap, the thermodynamic behavior of the hydrocarbon system.
reservoir liquid phase during pressure decline is the Many different forms of pressure cells have been
same under the two mechanisms. used. Acceptable work has been done with a cell
Laboratory tests to determine the PVT, gas solu- constructed from two bull plugs welded together and
bility, and viscometric behavior of dissolved gas suspended in a water bath. While this describes
systems are standardized to a greater degree than the crudest type of cell, it has the basic features of
tests performed on gas-condensate systems. Data many cells in use today. The main operational
on the compressibility and thermal expansion of feature of this type of cell is that only the total
the system, the quantity of gas in solution in oil at volume of the hydrocarbon system in the cell can
be determined. When two phases are present, the
The indicated 2500 cu ft per bbl solution gas-oil ratio repre-
sents approximately the maximum amount of gas that can be
volume of either phase cannot be determined sepa-
dissolved in 'oil at pressures up to about 10000 psi. rately without removing the other phase from the
DISSOLYED GAS SYSTEMS 75
cell. For discussion purposes, a cell like this will mined by locating the mercury-hydrocarbon liquid
be called a Type I cell. An illustration of such a interface within the cell with an electrically en-
cell is in Figure 48. More elaborate pressure cells ergized needle. The interface detector head 1s
.L.U$ TIIAl',0
#NIM'fr IA$
AMO LHHIIO CONTltfN.LED TD/PERATI/RE
SATN
...
,...
., ""
THI$ LIIII IIUD TD CHA/f'II IIILL
'IIOII IIOTTOII HOLi $AM/II.Ill
l
IIIIICUll'I I.I/ti
rnroM$
IIIAM/ill
IIAOIJATI TO
IIIA#UltE LIQUID
DETAILS OF
INTERFACE
DETECTOR NEAD
IJSED ON Ol$SOtYD
SA$ Sr5TEM$
CALlltATEO lllltt;Ultr
'fllll'. ALL 'IOWl!lffltlO
IIEA0fll6$ MADI NIU
... PREO!IURE
3f..2.
F.,r...,ti, ... ~
T,mp.
o.
TOT & I U\ /.
r Off Oc:rcc.t M\. e,.... w
N,
Ditt- Ml.
. u.. .,,i.i :r.. P.~I. KN, fl,,.P.fl 1., REMARKS TotI L'11111.1i4
SIJBBLI POIN AT 185.F K 4-.837 I( ,o-4
185.3 0830 0920 G012.E: Slo.'37 3(:)70 2.9 :3'=>5'!:,
<;o22:, 86.85
'03'Z.O 87. !lZ
3309 -
-
3294-
-.....
Z.9'-5 2950
"041,0 87,76 Z&BS 2.&20 ~~ 'v
-
18~.2 0935 "048. 7 SS.14 2~10 2501 ~~~~.-ct
<>944 l'o058.0 88.!>9 Z.4-79
- Z.4'=>3
--
3.02,91
18~-Z 09!!o3 G.0&~-4 89.1.5 2437 2.4'2:. I 2.9Z49
1ooe, ~083..0 89,81
1014,. "-102.7 90.77
238b
zazs
z:ei.,o Z94-!,/ )
,-
s.o Z309 2. 0195
1033 "123'.I 91. 77 2.Z'='3 ...... 224-7 2.7750
h/53.b 93.2.4 2'118 ..- 21'-2. Z.72"--5
10!!,.3, 1oe.1 '-194.0 95.2'3 Z.077 ...--- 2.0"' I 2.1o7eb
--
I 104-- hZ3+.Z 91.zo 1992. \976 Z.<o085
--
II I :5 <o'Zeb.8 99.76 1892. 187b Z.5492
11%:b ~34Z.I 1024' 1803 J 787 2:.4804
1152. <,894,4 105.02 "'392.4 1002.0 81.{)0 17Zb 1710 2.43!3~
185.~ &.~0?,.6 110.36 ~~23.tl 980.8 eo.~ 1589 3.1 I !!>72 2 3'=>29
1237 67Z3.e 121.10 5115.b 9+8.Z 78.97 1378 .3 13bJ Z.2!3.b6
12.~I 702Z.7 135.70 6102.9 919.8 77.58 \ Ibo- 3.4- 114-8 Z. 1991
1306 7586.3 IE:.3.29 E:,7090 877-3 "15.51 943
"' 7
926 z..ooe1
1325 84-"19 Zot,.'21 7627'.I 840.8 73.73 725 4-.J 708 ,\.9013
185.3 134-0 96192 Z7Z.Z5 9017.6 601.6 "71.81 54-8 4-.6 !:>32 1.7624-
14-00 11070.3 333.34 10291.5 778.8 70.70 4-50 4-.5 4-34- 1.7-Z.19
14-20 IZ44U, 400.30 ll'=:.80.4 7"11.2 69.84- 377 6.0 362 l.(c,778
TEM' . CHA GE'. TC l&SF K= 4.84-7 KI0- 4
165.0 14-4-0 1500 5990< 8'5.27 2.9
--
3'"8" 3671
l )
lb5.Z hOOZ.J 85.8& 3187 0172.
lb5.Z "010:a 86:Z'=i. 'ZS90 2875
1"5.1 Co019.4, 8<:..71 Z'!)S:5 25a.6
-
~
1 (88.50 - 86.05) 0
a -
87 .25 205
_
165
- 7.02(10-) vol/vol/ F
'
78 BEHAVIOR OF OIL FIELD HYDROCARBON SYSTEMS
tern having a relative volume of 1.099 bbl per bbl stance, it is J')resupposed by the use of a gas-oil ratio
of saturated oil. value that one knows the method whereby the gas
The volume of most dissolved gas systems at pres- and oil are separated. As the gas-oil ratio of a
sures greater than their bubble point pressures system is affected by the pressures, temperatur es,
varies linearly with pressure. As a consequence, it and number of equilibrium steps_ used to effect the
usually is not difficult to evaluate with good pre- separation, it is necessary to define explicitly the
cision the change in volume of the system caused by process by which the gas-oil ratio value was ob-
a pressure change of a few psi. At pressures less tained, as well as the particular gas-oil ratio value.
than the bubble point pressure, the P-V relation of The laboratory method of determinin g the gas-
the system takes a form similar to that of a hyper- oil ratio of a particular system under surface trap-
1-iola. Under these conditions, there are several ping conditions is essentially as follows:
ways of smoothing the experimental data in order
to evaluate the small system volume changes often (1) 'L'he system is brought t6 equilibrium at the
required in material balance calculations. Prob- pressure and temperatur e condition of the
ably the best method of smoothing P-V data at primary trap in the field and the volumes of
pressures below the bubble point pressure is by use the liquid and gas phases measured. The re-
sulting gas and liquid in the laboratory cell
of an. empirical relation termed a " Y" correlation.
are comparable to the fluids leaving the pri-
This empirical relation can be expressed as follows: mary trap, provided that thermodyn amicLJ
equilibrium is established in the trap.
(37)
where (2) The equilibrium gas phase is removed from
PB ..,, bubble point pressure (psia) the top of the cell, maintainin g the cell pres-
P = pressure at any point sure constant during the process by inject-
VB = bubble point volume, consistent units
V = two-phase volume at pressure P, consistent units ing mercury into the bottom of the cell. The
Y = empirical value volume of the gas phase is determined by
collecting the gas in previously evacuated bot-
Systems composed almost entirely of hydrocarbons tles and noting the change in pressure within
show a linear relation of Y and pressure. This fact the bottle. By means of the gas law equa-
is illustrated by the lower curve in Figure 50. Sys- tion, the observed gas volume is corrected to
tems that contain appreciable quantities of nonhy- the standard reference condition of 14. 7 psia
drocarbons usually give a curved relation of Y and and 60F.
pressure. The sensitiveness of the Y correlation is
(3) The gravity of the gas is determined by re-
greatest close to the bubble point of the system.
moving a portion of the gas from the bottles
An example is the plotted point at the right end of into a previously evacuated glass flask. This
the curve shown in Figure 50. To bring this 2463 flask is light in weight and can be weighed on
psia point onto the line requires a correction in the an analytical balance before and after filling
system volume from 88.59 cc to 88.62 cc, a change with gas. The molecular weight of the gai\_) 1 1
of less than 0.04 per cent of the volume reading. can be calculated from the increase in weight
The properties of a fluid being produced from a of the flask per unit change in flask pressure.
well are evaluated from results of laboratory tests by This is shown by the following equations:
reference to the system's bubble point pressure or From the gas law,
gas-oil ratio. Thus, one often speaks of "the com- PV - NZR1' (:lS)
pressibility of the 2520 psia saturated oil" or "the m
formation volume factor of the 525 cu ft per bbl N--
M (39)
Letting tl.m and tl.P indicate the change in into a single phase* and a portion of it forced into
weight and pressure of the previously evacu- the small glass trap. Pressure corresponding to
ated flask, the field primary trap pressure is maintained bv the
ll.m pressure regulator. After a desired amou~t of
M - il.P C (41)
liquid has accumulated in the trap, the valve at the
In Equation 41, C is a constant which is pri- top of the cell is closed and the trap brought to
marily a function of temperature (the volume atmospheric pressure. The liquid remaining in the
of the flask is constant; Z does not vary ap- trap corresponds to tank oil, and the total gas
preciably from unity, as the measurements measured by the wet test meter corresponds to the
are always carried out at less than atmos-
sum of the trap gas and tank vapors.
pheric pressure). The factor C can be
determined for any particular flask by going Differential Vaporization Tests. Differential va-
through the same process, but substituting porization tests are meant to represent the behavior
air for gas. of the. portion of original liquid phase that is not
produced to the surface during pressure depletion
(4) The liquid remaining in the cell from the first
separation is brought to equilibrium at the of a reservoir. As discussed in Chapter 2, the
pressure and temperature of the second-stage mobility of the gas phase (k, /,) becomes much
trap, and the procedures outlined above in 2 greater than the mobility of the liquid phase (k, /,)
and 3 are repeated. The procedures again ~fter an. appreciable gas saturation has been esta~ \
are repeated at the stock tank pressure and hshed w1thm a rock pore. Under these conditionV
temperature. a portion of the reservoir liquid phase can be thought'
(5) The volume of liquid remaining in the cell of as being relatively immobile within the rock pore.
(tank oil) is measured and its volume cor- Although the liquid may be immobile, it contin-
rected to 60F by use of the N a.tional Petro- uously contributes a mobile gas phase to the well
leum Tables. 23 A portion of the liquid is production. This process constitutes a differential
removed and its density determined. process; the composition of the system within the
The laboratory data obtained while determining rock pore is changing with time because of the
the surface gas-oil ratio of the Douglas sample movement of the gas phase out of the pore.
under a two-stage separation process are shown in The above word picture attempts to explain be-
Table 25. Part A of Table 25 lists the measured havior of gas and liquid within an individual pore.
pressure and volume data during the separation of A different picture is obtained if we consider phase
gas and liquid at the primary trap and tank con- behavior in numerous pores. As gas particle A,
ditions. The data reported here show that 87 .92 having just become separated from liquid particle A,
volume units of liquid saturated at 2520 psia and travels toward the well bore, it comes in contact
185F will shrink to 70.02 volume units of liquid with liquid B. Liquid B is very similar to liquid A,
in the primary trap and then to 63.60 liquid volume having just released gas B at only a slightly lower
units in the tank. Part B shows the method of pressure than the release at A. The combinat()
calculating the quantities and gravities of the sur- of gas A and liquid B produces a system very close
face gases and tank oil. in composition to the original A system, but at a
If the primary surface trap is operating at rela- slightly lower .pressure. The net result is a process
tively low pressures, or if the well production is comparable in many ways to a flash process, i.e.,
going directly to the stock tank, it often is impossible one of constant composition.
to simulate the field conditions by expanding the Actually, the vaporization process in a reservoir
volume of the laboratory system to the point where rock probably lies somewhere between the differen-
the cell pressure is equal to the field primary trap tial and flash vaporization processes. In many
pressure. An alternate method of determining the dissolved gas systems it makes little difference as to
gas-oil ratio and formation volume data under these * The two phases shown in the cell in Figure 48 are required for
a discussion of the differential test method. For the present dis-
conditions is illustrated in the upper right corner cussion, assume that the gas phase shown has been forced into
of Figure 48. The system in the cell first is brought solution.
80 BEHAVIOR OF OIL FIELD HYDROCARBON SYSTEMS
TABLE 25. RESULTS OF FLASH TESTS ON DOUGLAS SATURATED 0IL AT SURFACE TRAP CONDITIONS
Primary Trap 385 psig, 80F
Tank O psig, 80F
Saturated Oil Volume 87.92 cc at 2520 psia and 185F
Barometer 761.7 at 75F
PART A: Measured Data Primary Trap Tank
Initial Final Initial Final
Cell Pressure, psig. 385 0
Cell Temperature, F. 80 80
Total Volume, cc .... 261.55 70.05 63.63
Liquid Volume, cc .. 70.02 63.60
Gas Volume, cc. 191.53
Gas Removed
Volume of Collection Bottle, cc .. 6275 2413
Pressure, mm Hg. l.5 644.5 l.5 751.5
Temperature, F. 65 66
Gas Gravity
Flask Pressure, mm Hg. 1.0 613.5 1.0 719.0
Weight, gm ... 17.4158 17.5495 17.4158 17.6598
Temperature, F. 76 76
PART B: Calculated Data
Primary Gas
520 664.5 - 1.5
Gas Volume
Gas Gravity,
= 6275 525
(6m)
AP " =
_
760 0
= 5420 cc at 60F and 760 mm
17.5495 - 17.4158 0.1337
= 612 _5 =0.0002182
u
613 _5 _ l.O
(~;),,, at 76F = 0.0003157
0.0002182
Gas Gravity = 0 _0003157 = 0.691
Tank Vapors
520 751.5 - 1.5
Gas Volume = 2413 526 760
= 2370 cc at 60F and 760 mm
(t.m)
Gas Gravity, .1.P " =
17.6598 - 17.4158
719 .0 _ l.O ""' 0.000340
6
( .:i
mp) .
at 76F = 0.0003157
a,r
0.0003400
Gas Gravity = 0.
0003157
= 1.077
Tank Oil
Weight Pycnometer + Oil = 16.2169 gm Temperature ,.. 70F
No. 5 Pycnometer Volume = 9.811 cc
Weight Pycnometer - 7.9923 gm
Oil - 8.2246 gm
8.2246
Density = O.Sl l = 0.8383 gm per cc at 70F
""' 0.8420 gm per cc at 60F
Specific Gravity 60/60 - 0.8428 - 36.35API
Tank Oil Volume - 63.60 0.9910 = 63.03 cc at 60F
Gas-Oil Ratio
. . 5420
Primary Gas to Tank Oil Ratio = 63 .03 5.615 = 483 cu ft per bbl
. D- . 2370
TankVaporto T an k 0 11 .1~t10 = . ,5.615 =.211cuftperbbl
63 03
PART C: Summary
Solution gas~il ratio = 483 + 211 = 694 cu ft per bbl tank oil
. . (483 0.691) + (211 1.077)
Average solution gas gravity =
483
+ 211 = 0.810
Tank oil gravity = 36.35API
Bubble point formation volume at 2520 psia and 185F = :~:~! =- 1.395 bbl per bbl tank oil
--------
DISSOLVED GAS SYSTEMS
81
which process is considered to predominate, as both
A portio n of the P-V behavior of the original sys-
processes give comparable P-V behavior over a wide tem was determined prior to removal of the first gas
range of pressure. increment at 2203 psia. This behavior is shown by
It is not feasible for labor atory tests to correspond curve A. After bringing the original syste m to
to true differential vapo rizati on processes. To do equilibrium at 2203 psia and measuring the total
so requires that each infinitesimal gas bubb le be volume of the system (92.48 cc), the gas phase was
removed from the cell as it is formed. Rathe r, in removed from the top of the cell and its quan tity
the labor atory , a series of stepwise flash vaporiza- measured by collecting it in previously evacu ated
tion conditions is imposed upon the original syste m bottles. The cell pressure was main taine d as
and the gas formed at each step is removed as a unit. close as possible to 2203 psia during the gas removal
This stepwise process is called a differential process process by injecting mercury into the botto m of the
in pract ically all PVT work concerned with dis- cell. In Figure 51, the process of removing the
solved gas systems. gas from the cell is designated by the vertical arrow
Figure 51 illust rates the process of conducting a connecting syste m A to system B.
differential vapo rizati on test at 185F. In this par- After removal of the first gas increment, a porti on
of the P-V curve of system B was determined. Sys-
tem B was found to have a bubble point pressure of
2230 psia and a bubble point volume of 85.25 cc.Q
The fact that the bubble point pressure of syste m B
was great er than 2203 psia indicates that a small
amou nt of gas was left in the cell at the completion
of the previous gas withdrawal step. Ordinarily,
all the gas phase is removed from the cell at each
step, so that the bubble point pressure of the new
syste m agrees with the pressure at which the gas
withdrawal process is carried out.
The data in Table 26 show that 1017 cc of gas,
measured unde r stand ard conditions, was removed
in changing from syste m A to system B. The 6. 75
cc volume that the 1017 cc of gas occupied in the
cell at 2203 psia and 185F is the differential be-
tween the bubble point volume of syste m B (85.25
cc) and the volume of syste m A at 2230 psia (92.00
cc). The compressibility factor of the gas which
came out of solution between 2520 and 2230 psia .
is calculated at 2230 psia as follows: U
(42)
l'/16$$U ll~-P$/A where subscr ipts land 2 indicat e standa rd conditi ons
and reservo ir
Figure 61-Pre ssure- Volum e Relations of Douglas Satura conditions, respectively.
ted Oil.
Differential Vaporization at 186F
Solving Equa tion (42) for Z, gives
ticula r test, gas was removed from the system in
nine steps, the first step occurring at 2203 psia and (43)
the last step occurring at 14. 7 psia. Ordinarily, no
effort is made to remove the gases at partic ular As Z, is very close to 0.998 at 14.7 psia and 60F,
pressures, altho ugh the pressure increment between
steps normally is limited to 300 psi. z _ 2230 . 6. 75 520 0.998 _ 0.810
' 14.7 1017 645
82 BEHAVIOR OF OIL FIELD HYDROCARBON SYSTEMS
The gravity of the gas which came out of solu- The gravity of the gas which came out of solution
tion between 2520 and 2230 psia is shown in column at each step is listed in Table 26 and shown graph-
6 of Table 26. These gravities were determined by ically by curve A of Figure 52. As the measured
the method outlined in the previous section. gravities represent an average of the gases coming
TABLE 26. RESULTS 01'' DIFFERENTIAL VAPORIZATION TESTS AT 185F
Initial Saturated Oil Volume = 87.92 cc at 185F
1 2 3 4 5
--- 6 7
Volume of Gas Removed Gas Phase
Pressure ccat14.7
Compressibility
ccin Liquid Phase Volume Gas Gravity Factor
Step PSIA PSIA and 60F Cell cc Air==l z
0 2520 0 0 87.92
1 2230 1017 6.75 85.25 0.743 0.810
2 1930 964 7.50 82.87 0.740 0.823
3 1617 940 9.21 80.54 0.742 0.868
4 1310 984 10.86 78.34 0.752 0.872
5 907 1172 20.21 75.34 0.766 0.858
6 588 922 25.77 72.93 0.822 0.899
7 300 903 52.5 70.40 0.938 0.955
8 147 596 .... 68.46 1.159 . ....
9 14.7
14. 7 PSIA & 60F
1545 .... 61.89 1.73 . ....
58.26
--
I TOTAL 8953 1,_)
Residual Oil Specific Gravity - 0.8528 60/&J - 34.4 API
The shrinkage that the liquid phase undergoes be- from solution over a pressure interval, they are
cause of gas removal is indicated by the curve con- plotted as horizontal line segments. For example,
necting the several bubble points shown in Figure in Table 26, the measured gas gravity of 0.743 of the
51. System J was the final liquid remaining in the 1017 cc removed during step one is the average
cell at 14.7 psia and 185F. System K is the same /.60
material as system J, but measured at a tempera-
ture of 80F. The volume of the residual liquid left
l,IO
in the cell at 14.7 psia and 60F was obtained by
correcting the 80F volume for the effect of tem-
perature. A residual oil volume of 58.26 cc was
.
;;
obtained by this means. .
!!:
1.40
I.OD :a.I""'
'l
Table 26 presents a tabulation of results of the I:
s: /,JIO o.,o ~
differential test at 185F. The cumulative weight
of hydrocarbo ns in the cell at each pressure can be ..
~ ~
.
calculated from data presented in this table. The
weights are as shown in column 5 of Table 27.
.. I.OD O.IO t::
~u
;;;
.
At 60F
Notes on Calculation Method
Column 3, Weight of gas - 0.001222 cc , gravity
Each compressibility factor listed in Table 26 is Table 25 shows that a comparable 87.92 cc of
0
that of a different composition gas, although above bubble point liquid produced only 7790 cc of gas
1500 psia the gas gravity data indicate only minor and 63.03 cc of liquid when separated by a two-stage
differences in composition of the gases. As the vol- surface trapping arrangement. Thus, the gas-oil
umes of the gas phases within the cell were measured ratio of the 2520 psia bubble point liquid can be
at specific pressures, the compressibility factors 8953
must be plotted at the pressure at which the meas- expressed as 58 _26 5.615 = 864 cu ft per bbl or
urement was made. For example, the average of 7790
the gases coming out of solution between 2520 and 63 _03 5.615 = 694 cu ft per bbl, depending on
2230 psia has a calculated compressibility factor of which separation is being considered. Obviously,
0.810 at 2230 psia and 185F. This fact should be the exact processes by which the gas and oil separa-
kept in mind when comparing a smooth curve such tions are made must be known when speaking of
as curve B in Figure 52 with compressibility factor gas-oil ratio and liquid volume shrinkage. This
curves of constant composition gases. can be done by using the term "residual oil" to indi-
cate the liquid at 14.7 psia and 60F that results
Flash Vaporization of Dissolved Gas Systems at
from a differential process carried out at, or near,
Surface Conditions
the reservoir temperature. "Tank oil" is the teri( ")
It is easy to see from the data in Tables 25 and 26 for the liquid at 14.7 psia and 60F that results front""'
that a barrel of reservoir oil undergoing differential a flash process at, or near, surface temperatures.
vaporization at reservoir temperatures near 200F The terms "saturated oil" and "bubble point
will give different quantities of gas and oil than wilf liquid" are common to both processes and serve as
the same barrel undergoing flash vaporizations at the common basis for conversions between differ-
the lower pressures and temperatures existing at ential and flash processes. The ratio of the volume
surface traps. As an example, the da~a in Table 26 of a system to the volume of residual oil, or the
show that, under the reservoir conditions of 185F, volume of saturated oil, is called the "relative
87.92 cc of 2520 psia bubble point liquid produced volume" of the system. The ratio of system vol-
a total of 8953 cc of gas and 58.26 cc of residual ume to tank oil volume is called the "formation
liquid on undergoing differential vaporization at volume." Thus, in the example outlined above,
185F from 2520 psia to atmospheric pressure. the relative volume of the bubble point liquid at
84 BEHAVIOR OF OIL FIELD HYDROCARBON SYSTEMS
TABU.: 28. EFFECT OF SURJo'ACE TRAP l'R1':S1-iURio; 0~ QUANTITY AND PttoPJ<:RTH::-i (H' 8URt',\CJo: PI-UH>!'C"l'IO:'\
1 2 I 3 I 4 5 I 6 7 8 9
Surface Gas Released Surface Gas Gravities Tank Oil Produced Total Gas-Oil
Trap cu ft at 14. 7 PSIA and 60'F Air=l.O Ratio
Pressure at 14. 7 PSIA and 60'F
cu ft ~er bbl
PSIG Trap Tank Total Trap Tank bbl 'AP! tan oil
386 346 151 497 0.691 1.077 0.717 36.4 694
285 387 109 496 0.709 1.142 0.721 36.6 688
185 416 87 503 0.732 1.24 0.721 36.4 697
0 558 0 658 0.977 ..... 0.680 34.4 820
TABLE 2H. EFFECT OJ,' StTRFAGJo: TRAP Tf:MPERATURE ON QUANTl'rY AND PROPERTIES OF SURFACE PRODUCTIO:-f
1 2 3 4 5 I 6
Fonnation Volume Surface Gas Richness
Separation of Saturated Tank Oil
Temperatme Total Gas-Oil Ratio Reservoir Oil Gal C, + per MCF
'F cu ft per bbl tank oil bbl per bbl tank oil
Gravity
'AP! Trap I Tank 0
Primary Trap at 3 8 5 PS IG
100 710 1.425 35.6 0.99 8.1
80 694 1.396 36.4 0.71 5.9
60 679 1.369 37.2 0.45 3.8
Primary Trap at 186 P SIG
100 713 1.415 35.6 1.40 9.2
80 697 1.387 36.4 1.08 6.9
60 682 1.362 37.2 0.78 4.6
I
"""
1111111
1111
""" L.
I
....... : J
STAGE SEPARATION
l. I
I /1
-1-++>H+Hl+l+l+"lll-
'
II '
1111111
.t-t-+-r
:tt-t-ttiiitttt+Htt[ffijtttttttt
!JO PSIG PRIMA R~RAP
JG RJMARY A
'"
BASIS: SEPAR ATIIT i:F
30 40 50 60 80 JOO 200 300 500
SECON DARY TRAP PRESS URE - PS/A
6 .
- -t
- -+
5
ill
-
T t.
..
~II
IHI/HI
4 . .SE.
ti:4, - - . .
1111~
FOUR S :4GE
- .-
'AftoN I
IHIIHI
II
-
3 I
-
L,:
2
0
II
0
~ ~-
I
-
BA IS."
I II
- 00 P 16 P /MARY 1r/A1
2!10 PSJG SECONDARY TRAP'
SEPARATION AT 80F
- .
1111111
-
0
10 20 30 40 50 60 80 100 200 300 500
THIRD TRAP PRES SURE -PS/A
Figure ti3-Effec t of Stage Separation on Tank Oil Recovery
BEHAVIOR OF OIL FIELD HYDROCARBON SYSTEMS
_,_ :::C.-.: faetor which can affect the quantity of psig. Three-stage separatio n often calls for in-
=-'i,'"= ;:m:..:ucts is the number of traps used in creasing the first stage pressure to 400 or 500 psig
"'"""' :~ sep-,rate the gas and oil. Increasing the with the second stage operating in the region of 60
.:.=~-c ,,f srnges will result in more tank oil of psig.
::;;:':~, i::::is:ty and less surface gases. For any A precise study of the pressure-production be-
,.,,:.-:c.C.:~ xell effiuent and temperatu re, there are havior of a reservoir requires the gas solubility and
',:-.:::::=. ;:::,,ssures at which the traps can be oper- formation volume data used in the stud, to be
.,_:2.::.:: 2r--::e: to yield the maximum tank oil or, if evaluated under conditions represent ative of the
~es...""'!.:. :::e maximum amount of gas. An ex- field operating conditions. This usually requires
=:;e :i :::e effect of stage separation on tank oil specific laborator y PVT tests to determine the
:,;._,~ -,,,:,- :.,; shown in Figure 53. * The upper half appropria te properties of the reservoir fluids.
:i ?ir.:...--e ,53 shows that with a primary trap oper- However, in many instances, inaccuracies of several
;_::::,,i ,: 515 psia, the optimum secondary trap per cent are allowable in the values of bubble point
::--,~""'! wculd be i5 psia. Operating with this pressure, gas solubility, or formation volumes and,
:-,- .:,::;;::c:: of trap pressures would result in as a result, use can be made of general correlations
:,;-~:--"2.,,; l-!.2 per cent more tank oil than would of these properties. A correlation of the formation
:-,, ,,,._,:-.-e::,,,i ii the well production were flashed volume of bubble point liquids already has been
given in Figure 27, and a correlation of the total
.::..""'!c'::.- :0 :he tank. The chart also shows that
=.:.:::i.:::,- :he same recovery would be obtained gas plus liquid phase formation volume is shown in 0
-.'. :.:e ,r'.;:::;,ry trap pressure were reduced to 265 Figure 39. In Figure 54 7 is shown a correlation
""'" .1.::c: :::e secondary trap operated at 55 psia. 10000
,ooo
'::-:e _ewe: ;:,art of Figure 53 shows the effect of ooo
700D
:::s-::~ ~ s 500 psig primary and 250 psig second- 6DOD
--
1: ::-s.:: s.::-i varying the pressure or' the third-stag e
SD DD
-
:~:. :: =
=i:
be seen that for the particula r fluid
~c'.:...-sion the addition of a third trap yields
ooD
,ooo
- -- - -- f-
I-- -
I
u
;)~~c.~ ~e3iderations, ho'1-tever, may warrant four- 100
10' ~ S 4 $ 6 7 I ,,oz
,:~ se~,,.-,cion. The trap pressures yielding the /.!:L-)0_.I 100.0DO,I f
:,c:-,u.:,;_ =ount of tank oil are dependent upon f "f, 100,f/l.4 ,,,,
:c,, :-:;::;:,:,sition of the system being separated and Figure 64-Correla tion of Bubble Point Pressure with Dissolved
Gas-Oil Ratio, Dissolved Gas Gravity, Tank Oil
"-""" :erer-,ined best by laboratory experimentation Gravity, and Temperatur e
:c '::c e-s:.~ulation. In a two-stage separation of
.::,,.._---s:,,ec: g-iB systems, the maximum quantity of of the bubble point pressure of a dissolved gas sys-
:,-:, oi'. ;;;,ually is obtained when operating the tem with dissolved gas-oil ratio, dissolved gas
::..--sc-,ta.,:,> ;rap in the pressure range 150 to 250 gravity, temperat ure, and tank oil gravity. A
calculating chart based on this correlation is given
.. Y..i, :'.;,~a.s production but a. dissolved gas system having
":t-:~~i~ ~'-'~ to those of the Douglas system, in the Appendix.
DISSOLVED GAS SYSTEMS 87
The correlation shown in Figure 54 only indirectly 1000
PR$$1/Rffl$IA
1000
Gas solubility determined by differential test
methods is expressed most often in terms of cubic Figure 66-Relative Oil Volumes vs Pressure
feet per barrel of residual oil or of initially saturated
Douglas Saturated Oil
Differential Liberation at 186F 0
oil. . Relative oil volumes also are commonly based
on a unit volume of residual oil or of saturated oil. It often is necessary to compare the results of dif-
Data on the solubility and relative oil volumes of ferential tests on similar samples tested by different
the Douglas saturated oil at 185F are shown in laboratories. It is quite possible that the com-
Figures 55 and 56. The values shown in these parison will be complicated by a difference in the
figures are calculated from the data given in columns bubble point pressures of the two samples. When
2 and 6 of Table 27. It will be appreciated that in such a difference does occur, it is obviously incor-
each figure the values of the two curves at any pres- rect to compare the gas solubilities and relative
sure are in the ratio of 58.26 to 87.92, which are the volumes on the basis of unit volumes of saturated
respective volumes of the residual oil at 60F and oils. Rather, one should make comparisons on the
the 2520 psia saturated oil at 185F. basis of a unit of residual oil or some saturated oil
88 BEHA VlOR OF OIL FIELD HYDROCARBON SYSTEMS
that is common to both systems. When this is solubility-pressure and formation volume factor-
done, one usually will find that the solubility and pressure relations of the reservoir liquid phase, both
relative volume of the two samples will be in determined under surface separating conditions.
agreement. All these factors can be obtained by modifying
At ti.mes the solubility and relative volume curves the data presented thus far.
of two similar systems will not be in agreement when Pressure-Total Volume Relations. An experi-
based on a barrel of residual oil, but both sets of mental method of developing a system having a
curves will be displaced from each other by a con- saturation or bubble point pressure of 2750 psia
stant percentage. When this occurs, it is usually at 185F is to add sufficient gas to the bottom hole
because of an error in the measurement of one of sample to raise its bubble point pressure from
the residual oil volumes. If there is good agreement 2520 psia to the desired value. The added gas
between the relative oil volume curves but poor should have, of course, a composition as close as
agreement in the solubility curves, the possible possible to the gas that originally came out of solu-
errors are indicated to be in the measurement of tion in the reservoir liquid between 2750 and 2520.
the liberated gas quantities. One source of such a gas is casing gas taken at the
Fortunately, differential gas solubility curves and time the bottom hole sample is taken. A second
relative oil volume curves are not sensitive to the choice of gas would be from the primary trap.
pressure at which the differential. gas withdrawal However, in many instances, the composition of the Q
process was initiated. As pointed out in Chapter 2, primary trap gas is different from the composition
bottom-hole liquid samples or recombined surface of the free reservoir gas and consequently has
samples often are quite different in composition limited applicability.
from the liquid phase within the reservoir rock. The mathematical approach to the problem is to
Nevertheless, good estimates of the properties of use the data shown in Figures 52, 55, and 56 to
the reservoir liquid can be made from extrapolated obtain rough values for the new P-V curve having
properties of the laboratory samples. As a matter the desired bubble point conditions, and then to
of fact, the extrapolation may not involve going smooth the rough values by use of the previously
just from well bore conditions to average reservoir mentioned " Y" correlation. The steps are as
conditions, but often involves extrapolating from follows:
the present day well fluid properties to original
(1) On the basis of a unit ba.rrel of residual oil,
reservoir conditions. This will be illustrated in the liquid phase relative volume curve of sys-
the following section wherein the properties of the tem A is drawn as in Figure 57. The curve is
original Douglas reservoir oil at 2750 psia and 185F extrapolated to 2750 psia, giving a value of
are calculated from the measured properties of the 1.548 bbl per bbl of residual oil at this pres-
2520 psia bubble point liquid obtained in a bottom sure. The extrapolation also is shown in
hole sampler. Figure 56.
Adjusting Laboratory Data to Field Conditions (2) The P-V curve of system A (2520 psia bub{_)
ble point) is constructed in terms of the unit
The method of applying the properties of dis- barrel of residual oil. This is done by divid-
solved gas systems to reservoir behavior calcula- ing the total volumes of system A, listed in
tions will be discussed in the next chapter. For the Table 24, by the volume of the residual oil,
moment, it is sufficient to know that one of the 58.26 cc. For example, at 1710 psia, the
methods to be discussed requires a P-V curve at . vo lume of system A.1s 105.02
reservoir temperature of the original liquid phase
relative .
58 26
= 1.803.
in the reservoir, the solubility of gas in the original This is shown in Figure 57.
reservoir liquid when separated under the field (3) Extrapolation of the upper solubility curve
trapping conditions, and the formation volume in Figure 55 from 2520 to 2750 psia indicates
factor of the original reservoir liquid. The second that system B will contain an additional 92
material balance method requires the flash gas cu ft of gas per bbl of residual oil. (995 -
--
DISSOL YED GAS SYSTEMS 89
,l~ c"
The above procedure is based on the assumption s
that the relative liquid volume curves of systems A
_,.., I
and B are the same and that the compressibility
factor of the 92 incrementa l cu ft of gas can be rep-
resented by the curve shown in Figure 52. The first
assumption generally is true when there is not too
large a difference in the bubble point pressures of the
two systems. The second assumption is good over
the range of pressure in which the composition of
the gas coming from solution is close to the composi-
tion of the incremental gas. In this particular in-
stance, the gravity of the gas (which reflects the
composition of the gas) remains relatively constant
1.0
-~}
/nF
+
,ooo
down to about 1000 psia.
/000 1~00 1000 A second method of calculating a P-V curve from
/l6S$UIIC-PSIA
differential vaporizatio n data is illustrated in Table
Figure 67-Construct ion of P-V Curve for System B from Data of
System A 31. The volume of free gas present at any pressure
is calculated from the solubility curve and a com-
863 = 92). These additional 92 cu ft of gas,
when compressed to pressure P and 185F,
pressibility factor determined from the weighted gas
gravity. The free gas volume when added to the
0
will occupy
liquid phase relative volumes gives the desired two-
92 14.7 645 Z phase P-V curve. The P-V curve then can be
5_615 p 520 Z = 298.5 p bbls of space.
smoothed by the usual " Y" correlation. This
Thif\ incrementa l volume of gas added to sys- method is not as accurate as the previous one as it
tem A's relative volume curve yields the de- depends in part on a gas phase compressibility
sired P-V curve for system B. The P-V factor determined from reduced pressure and tem-
curve of system B then is smoothed by a perature correlations. The example shown :in Table
" Y" correlation. These calculations are 31 has been limited purposely to calculating relative
shown in detail in Table 30. volumes of system A in order that a comparison
TABLE 30. METHOD OF CALCULATING P-V CURVE FOR SYSTEM B FROM DATA ON SYSTEM A
1 2 3 4 5 6 7 8
Relative Free Gas Relative Smoothed Relative
Volume Volume Volwne Volume
System A Free Gas
*298.5p
z System B System B
Pressure bbl.f,er bbl Com~ressibility bblJ:bbl Smoothed
PSIA resi ual oil actor bblJ:bbl
bbl resi ual oil "Y" "Y" resi ual oil
2750
2520
2247
1.5091
1.5751
0.804
0.812
0.0952
0.1079
1.6480
1.6043
1.6830
2.5102
2.5671
2.7365
2.6100
1.5480
1.5996
1.6808
u
2061 1.6345 0.818 0.1185 1. 7530 2.5245 2.5245 1. 7530
1876 1. 7123 0.826 0.1315 1.8438 2.4381 2.4381 1.8438
1710 1.8026 0.833 0.1455 1.9481 2.3532 2.3532 1.9481
1572 1.8942 0.840 0.1595 2.0537 2.2939 2.2939 2.0537
1361 2.0786 0.852 0.1870 2.2656 2.2016 2.2016 2.2656
1148 2.3292 0.865 0.2247 2.5539 2.1476 2.1040 2.5747
926 2.8027 0.880 0.2838 3.0863 1.9822 2.0025 3.0999
1 2 3 4 5 6 7 I 8
,s:i :'<' :,ia,i, ,,f ,akulated ~nd experimental values. smoothed values of the two-phase volumes relative
T',:, ,,,,n;,.:is,n is shown m columns 7 and 8. to the bubble point volume. The data of column 6
r::, :/,i:,i m,'t h,,d of calculating the P-V curve of have been referred to a unit of residual oil in column
a ,,,:,:n ,, t,y u"'' of the generalized correlations 7 to facilitate comparison with the relative volume
..,,:.. , ,.1 }'~nn', :,11 and 5-1, or by use of the respec- data of system B given in Table 30. This third
;,w
l,,' \. ' ' " '-:
_.a:.-,::a:in~ ,brt in the Appendix .. I~ the ex- method of calculating a P-V curve of a particular
:.::<, ;i, ,:: in T:ible 32, the gas plus hqmd P?ase system has considerable merit, as it can be used
:,.,:-~~~~~:,.,:~ \~'lU!llt" data in c?lumn 2 were obta1~ed when direct laboratory data are not available.
',::\\ , ~<' :w,,..,,bse format10n volume calculatmg Pressure-Solubility Relations. When using the
.:, .... "'><' ~ahil rntio of the particular system Schilthuis" form of material balance, it is necessary
:::;:",;, "<",,: 1 t,y making use of the calculating to know the solution gas-oil ratio of only the original
liquid phase in the reservoir. Consequently, in
. ... ,., : , "u:-t,le point pressure of 2750 psia at
'~ ' ,,.,,., ,11c ...
',
~ ::~~-... '
~ .~:~1..,tved gas
' ..... , . ' ""
::~.'.~: ..s:,,i :.e _tank oil gr~vity was taken as
gravity was assumed to.
...,
many routine PVT tests a solution gas-oil ratio
(flash) of only one system ordinarily is determine( 1
;:,q-' .\:''.. l'i't' ,,r the ealculatmg. chart with these In many instances, the gas-oil ratio reported is thW
:,.':'<':::.s ,,i :::Ce ,urfaee gas and 011 gave a solut10n of the bottom-hole sample. It remains for the
~~~:'. ,-,,:',' ,,f 77\1 eu ft per bbl o( tank oil. Column engineer to convert this ratio to a value representa-
:, .., ".,'-\' ;::c Ct?'.<'ri' the format10n volume data to tive of the original reservoir liquid phase. More
,, ... ...,., "t' ,.'\.":':.ultlOll
, .,,.--., " Of ro
.... i 0 . The da t a 0 f
ps1a. recently, Tarner" suggested a form of material
-~,.......
'"''"" . . .
, , ""'' i ~tve the calculated values of Y
" " balance that requires the solution gas-oil ratio of
,'\ ~---'' ' -
'"-'"" ::c .,<::::,:: 4. The smoothed Y data m
'' '' . the reservoir liquid phase at all pressures below the
--:
... ....... , ; "~~ ,;:-mined by constructing a straight
,,.. " " .
original reservoir pressure. Thus, to use Tamer's
:,:,., :>,.,_,:;:\ :2e ,::i,ta oi column 4, less weight being method of material balance, it is necessary to be
....,
~'"'" :,' :~~ ,':Lc'ulated values of " y " near t he able to convert solubilities determined by differ-
:_ ...,,:c, "-'i:t: ,, :he system than at pressures be- ential methods to flash solubilities.
:,:~~:.,, ~:.:,_'I.' 3::,.: 1300 psia. Column 6 presents It is not too difficult to determine experimentally
DISSOLVED GAS SYSTEMS 91
TABLE 32. METHOD OF CALCULATING P-V CuavE FROM GENERALIZED Two-PHASE FoaMATION VourME CoRREJ,ATION1
1 2 3 4 5 6 7
Smoothed Smoothed
Forma'tion Volume Volume Relative
Volume Relative to Relative to Volume
Pressure bbl per bbl Volume at Smoothed Volume at bbl per bbl
PSIA tank oil 2750 PSIA ''Y" "Y" 2750 PSIA residual oil
2750
2520
1.47
1.51
1.000
1.027
--
3.380
--
2.865
1.0000
1.0318
1.5480
1.5972
2247 1.58 1.075 2.985 2.718 1.0824 1.6756
2061 1.66 1.129 2.592 2.613 1.1279 1. 7460
1876 1.75 1.191 2.439 2.510 1.1856 1.8353
1710 1.84 1.252 2.413 2.420 1.2513 1.9870
1572 1.93 1.313 2.394 2.343 1.3198 2.0431
1361 2.14 1.456 2.238 2.230 1.4576 2.2564
1148 2.43 1.653 2.137 2.110 1.6613 2.5717
926 2.83 1.925 2.129 1.990 1.9898 8.0802
::c::: reservoir oil saturated at 2520 psia and 185F. cu ft per bbl and 0.830, 0.862, and 0.920 gravity
\\:::. this assumption, Equation (45) can be s1m- respectively. These calculated values, together
cccd further to with the 2750 psia saturated oil values, are shown
694 ( ,, ) (2750)' _l_
83 (46) as the solid black squares in Figures 58 and 59.
" - 0.810 2520
rs = 952-Yu (47)
T':e ~ssumption that tank oil gravity remains rela-
::.e:v constant for reservoir pressures above 1000
Dll'f!RNTIAL LIIIClfATIDN
:s:-, \ reasonable, and is in agreement with field AT IH,:
:,ct> :raps and causes an increase in tank oil gravity.) AT 110, TN/ti/ TWO
STAR SPARATl0li
Ju-o n/6
.\ :-econd equation relating produced solution gas-
:,: cs.:io with solution gas gravity can be set up by
=Ilg use of the results of differential vaporization
:cs:s. Referring to Figure 52, it is reasonable to 0
a
CXHlflllNTAL VALllt:S
CALCIILATt:0 VALi/$
...., 0
ar.:et that had we been able to sample the original
:-,,,'e::rnir oil, the gravity of gas coming out of solu- 1000 l!IOO
::c:. at 1S5F between 2750 and 2520 psia would OIICSSURC-PSIA
::~:e been 0.745. In terms of surface gas produc- Figure 58-Gas Solubility vs Pressure Obtained by Differential
and Flash Processes
::cr:. the 2i50 psia saturated oil can be thought of as
,rn:J.i.ning (r, - 694) units of 0.745 gravity gas, in 1.80
"{~ =
694. 0.81 + (rs
r,
- 694)0.745
(48) "-; l . r
v:::-='!"c' 69-l = 50lution gas-oil ratio of the 2520 psia saturated oil
.
~
1:~
+ SL INGASG.iMTY
DIFFERENT/AL LIBEfl.AT N
-, AT JS:SF '
,).81 = tots.I gra\'ity of surface gases produced from 2520 /,
psia saturated oil
= solution gas-oil ratio of 2750 psia saturated oil !
r8
"{
1
= total gravity of surface gases produced from 2750
psia saturated oil
.
:: /,
).';"-l:5 = average gravity of reservoir gas that would be lib-
ers.ted between bubble point pressures of 2750 and
2520 psia. Value obtained by extrapolation of curve
A, Figure 52.
0
a
S:Jiving simultaneously Equations (47) and (48) INSTAN "ANE, S
GRAVITY BY OTFFEHE
A
LIBERATION AT 18$.--,
f_:e:- the following quadratic equation:
a o /000 /SOO 2000 2500
ra 1 - 70nra - 42{145 =0 (49) PRESSURE-P S/A
T::e particular solution of Equation (49) is that Figure 59-Gravity of Dissolved Gases Obtained by Differential
and Flash Processes
, = i65 cu ft per bbl of tank oil and that the
.sciution gas gravity is 0.804. The solid black circles in these figures are the experi-
The solution gas-oil ratios and solution gas gravi- mentally determined values.
::es of 2000, 1500, and 1000 psia saturated oil cal- The lower curve in Figure 58 represents the sur-
~i:,.:ed by similar equations are 538, 395, and 250 face gas-oil ratio obtained when only saturated
DISSOLVED GAS SYSTEMS
93
reservoir oil is produced from the rese
rvoir. The reservoir oil was flashed through the
curve has purposely been extrapolate trap s. This
d to zero gas- last value should not be confused with
oil rati o at a pressure nea r 100 psia. the surface
This has been grav ity, which might be obta ined as
done to illu stra te the poin t tha t a rese a result of
rvoir oil of simultaneous production of free rese
low satu rati on pressure may not prod rvoir gas and
uce trap gas satu rate d reservoir oil.
or tank vap ors when sep arat ed und
er the lower Pressure-Formation Volume Relatio
surface tem per atur e. ns. Aft er
The gas gravities shown as functions of the qua ntit y of gas in solution at any
pressure in part icul ar
Fig ure 59 illu stra te the wide vari atio pressure has been determined, it is a
n tha t can be simple mat ter
obta ined when sep arat ing gas and liquid to calculate the formation volume of
by different the satu rate d
processes. The lower curve, previous oil as a function of pressure. This
ly shown in calculation is
illustrated in Table 33 and makes use
of the app ar-
TABL E 33. MET HOD OF CALC ULAT ING
FORM ATIO N VOLU ME OF SATU
RATE D 0IL AS A FUNC TION
USIN G EXPE RIME NTAL DATA OF PRES SURE AND GAS SOLU
TO GuID E CALC ULAT IONS BILIT Y
1 2
3 4 5 6
Item Saturation Pres sure -PSI A
Description
2520 2000 1500 1000
1. Solubility, cu ft per bbl tank oil ..... ...
2. Solution gas gravity, air= l .O ..... ..... 694* 538 395 250
Weight of :\.as, lb ..... ..... ..... ..... ..... ..... ....
3.
4.
5.
Weight tan oil, per bbl, lb ..... .....
Total system weight, Jb ..... ..... .....
..... ..... .....
..... ..... ..... .
0.810*
42.9
295.0
0.830
34.1
0.862
26.0
0.92 0
17.6
0
..... ..... .... 295.0 295. 0 295.0
6. Appa rent gas density, 14. 7 PSIA and 60'F 337:9 329.1 321. 0 312. 6
7. Pseudo Jiquid volume of dissolved gas, cu , lb per cu ft .. 23.9 24.3 24.9 26.0
8. Pseudo liquid volume of system at 14. 7 ft ..... ..... .. 1.79 5 1.40 3 1.04 8 0.67 6
~ft .... .... .... .... .... .... PSIA and 60F,
9. Pseudo liquid density of system at .... .... .... .... ...
14. PSIA and 60F,
7.410 7.01 8 6.658 6.291
lb per cu ft ..... ..... .. : ..... .....7.....
10. Pressure correction to saturation ..... ..... 45.60 46.9 5 48.2 0 49.7 5
11. Tem perat ure correction from 60'Fpress ure, lb per cu ft ...
to 185'F, lb per cu ft ..
+0.8 5 +0.6 5 +0.4 5 +0.3 0
12. Density of system at saturation pressure, lb -3.8 5 -8.2 5
13. Volume of system at saturation r:r_
cu ft .... 43.10 44.3 5
-3.1 0
45.5 5
-3.0 0
47.0 5
14. Fonnation volume of system atpress ure, cu t ..... .....
saturation pressure, bbl
7.83 3 7.42 5 7.05 0 6.645
per bbl tank oil. ..... ..... ..... ..... ..... .....
... 1.39 5' 1.32 3 1.25 5 1.18 2
*Experimental data
Basis : One barre l tank oil
Nates on Calculation Meth od
Item I.From Figu re 58
Item 2.From Figure 59 Item 7. Item 3 divided by item 6
Item 8. Item 7 plus 5.615
Item 3. Weig ht = c; ~t 28.97 grav Item 9. Item 5 divided by item 8
7 ity
Item 4. From Tabl e 25, tank oil Item 10. From Figure 24
= 36.35' AP!, or specific Item 11. From Figu re 25
grav ity = 0.8428 60/6 0 Item 12. Item 9 plus item 10 minu
Item 5. Item 3 plus item 4 s item 11
Item 13. Item 5 divided by item 12
Item 6. From Figure 26
Item 14. Item 13 divid ed by 5.615 0
Figure 52, shows the grav ity of the gas
just coming ent gas density met hod outlined in Cha
out of solution in the reservoir oil at pter 4 and
185F. The a formation volume determined experim
upp er curve represents the gra vity of all entally at
the gas still some reference pressure.
in solution at any pressure when a 185
F differential Referring to Tab le 33, items 1, 2,
vaporization process is followed from and 14 of
tha t pressure column 3 were determined experimenta
to 14.7 psia. The values shown here lly by flash-
were obtained ing a sample of the 2520 psia satu rate
by dividing the weight of gas in solution d oil thro ugh
, shown. in a part icul ar surface trap arrangement.
column 3 of Tab le 27, by the volume of The deriva-
gas and the tion of these dat a is presented in Tab
con stan t 0.001222. The inte rme diat e le 25. The
curve repre- calculation of item s 1 thro ugh 5 is stra
sents the tota l gra vity of the solution gas ightforward,
tha t would being an expression of the pounds of
be obta ined at the surface when only material pro-
satu rate d duced at the surface per uni t barrel
of tank oil.
94 BEHAVIOR OF OIL FIELD HYDROCARBON SYSTEMS
The apparent density that the solution gas exhibits A comparison is shown in Figure 60 of the relative
when in solution at the reservoir temperature can oil volume obtained by differential liberation at
be evaluated for the experimental system (column 185F with the formation volume of the saturated
3) by working backward from the measured forma- oil as determined from surface trap sets. The solid
tion volume. For example, the reference system black circle represents the value of formation vol-
consisting of 337.9 lb of material occupies 5.615 ume factor determined experimentally. The solid
1.395 = 7.833 cu ft of reservoir space at 2520 psia
/40
and 185F. The density of the system at this
pressure and temperature is therefore 43.10 lb per 0 EJtH/fll/E#TAI. VALU~I
CALCULATED HAU/II
cu ft. Taking into account the compressibility and
thermal expansion of the reference system (Figures "D n L aEAION
AT-.114F
24 and 25) results in the evaluation of a pseudo-
liquid density of 45.60 lb per cu ft at 14.7 psi~ and
60F. This pseudo-liquid density, when divided
into the weight of the system, yields a pseudo-
liquid volume of!!~~~ = 7.410 cu ft, which is com-
-
posed of 5.615 cu ft of tank oil and 7.410 - 5.615
= 1.795 cu ft of dissolved gas. The apparent
- xr . ;tt ,:i
/000 -
IOOO 1400
0
density is obtained from the weight of gas, 42.9 lb, PR$$IJRE-PS1A
divided by its apparent volume, 1. 795 cu ft, or Figure 60-Relative Volumes and Formation Volumes of SaturM
ated Oil vs Pressure
23.9 lb per cu ft at 14.7 psia and 60F.
The calculations thus far have been toward black squares of the lower curve represent the data
evaluating an apparent density for 'the solution gas calculated in Table 33. The left end of the for-
obtained from the separation of the reference system mation volume curve is extrapolated to a value of
by the desired trap operation. After the apparent 1.063 at 14.7 psia and 185F because of the thermal
gas density for the reference system has been cal- contraction of 14.7 psia saturated oil in cooling
culated, the data can be used to calculate the forma- from 185 to 60F.
tion volume of other saturated oils. For example,
to obtain the formation volume of the 1500 psia Summary of Principal Points
saturated oil, the 23.9 lb per cu ft gas density of the
reference system first is plbtted on Figure 26, and a A dissolved gas system has been defined as one
curve parallel to the tank oil gravity lines is con- that has a critical temperature greater than the tem-
structed through the point. The curve so con- perature 3<t which the system is studied. Because
structed shows that the 0.862 gravity gas obtained of the particular shape of pressure-temperature
from the 1500 psia satu~ated oil would have an phase diagrams of naturally occurring systems, tbQ
apparent density of 24.9 lb per cu ft.* The volume relation of critical temperature and reservoir tem-
occupied by the 1500 psia saturated oil at reservoir perature is such that only bubble points are en-
conditions is calculated by the method described countered at the phase boundary.
in Chapter 4, which makes use of the weight of solu- The behavior of dissolved gas systems ordinarily
tion gas and its apparent density. Example cal- is described in terms of the pressure-volume relation
culations are shown in columns 4, 5, and 6 of Table of the total system, the pressure-liquid phase volume
33. relation, and the solubility of gas in liquid. Volu-
metric relation may be expressed either in terms of
Note that the apparent density values correspond to a tank
formation volumeor relative volumes. The forma-
oil gravity of about 45 API, rather than the 35 API obtained in tion volume defines the volume occupied by the
the laboratory test. Thus, Figure 26 is used only for the relation system at elevated pressures and temperatures
of apparent density to gas gravity and not for the absolute values
of the apparent density. (formation conditions) per unit volume of 60F
- i
DISSOLVED GAS SYSTEMS
95
surface tank oil. The relative volume defines the
process, also is 76.05 cc. Thus, in this parti cular in-
volume of the syste m at some elevated pressure and
stance, there is no difference in t.he volumetric be-
temp eratu re per unit volume of satur ated oil, or
havior of the liquid phase down to 1000 psia, re-
unit volume of 60F residual oil. Residual oil is
gardless of which process is followed.
the term for the liquid phase that results from
a Common PVT usage results in some confusion in
differential vaporization at reservoir temperature.
the meaning of the terms "bub ble poin t press ure"
The solubility of gas in oil ordinarily is expressed
and "satu ratio n pressure." The bubble poin t has
in units of cu ft of gas, measured at 14. 7 psia and
been defined as the state at which an infinitesimal
60F, per barrel of oil. The barrel of oil may be
quan tity of gas is in equilibrium with a large amou nt
60F tank oil, 60F residual oil, or oil satur ated at
of liquid. As commonly used, the satur ation pres-
a specified pressure and temp eratu re.
sure is analogous to bubble poin t pressure, and its
In most instances, it is of small consequence use implies the existence of only a small quan tity
whether one considers a differential or 'flash process
of gas phase. However, on occasion, the term
to prevail in the reservoir. For example, in Figure
"satu ratio n press ure" is used to identify the pres-
50, the volume of liquid phase obtained under flash
sure at which vapo r and liquid phases ai:,e in equi-
conditions at 1000 psia and 185F s shown to be
librium, regardless of the proportion of the phases.
76.05 cc per 87.92 cc of 2520 psia satur ated liquid.
The. term "bub ble poin t press ure" has a precise
In Figure 51, the volume of liquid at the same pres-
meaning and should be used to identify the pres surQ
sure and temp eratu re, but obtained by a different
at which only an infinitesimal gas phase is present.
0
CHAPTER 8
MATERIAL BALANCES
.\n important use of PVT data is for evaluating constant-volume tank, initially containing liquid
,,it ,1t1d gas reserves. The evaluations can vary in and gas phases. To simplify the problem, it can be
,,,n1pJ,,xity from the simple evaluation of the stand- assumed that
:<:\l c'lll,i1 feet of gas per acre-foot of sand to the more
( 1) The gas is insoluble in the liquid at all pres-
smpJ,,x evaluations that involve the pressure-pro- sures.
,,,.di,m performance of reservoirs. It is the pur-
"'"' ,,f this chapter to discuss the use of PVT data in (2) The gas obeys the ideal gas law, PV = Cat allu
pressures.
; ~''"'' more complex types of calculations. It is not
;'>''l''"""d to discuss in detail the material balance (3) The tank liquid is incompressible.
:,,,,th,,d. other than to point out a few of the assump- (4) The vapor pressure of the liquid is zero.
'.>,'llS that are made in deriving material balance
(5) The tank temperature is equal to the outside
,,;uMi,,ns and to show how these assumptions, in
atmospheric temperature.
,,n,,ral. will a.ffect the calculated reserves.
rtw term "material balance," when applied to A schematic representation of the tank and the
~'-~cr"'kum reservoir engineering, has come to mean assumed solubility and volumetric relations is
: ;arti,ular type of calculation. In fact, to many shown in Figure 61.
;"<mlenm engineers, it connotes a particular equa-
:.,,'n pn,posed by Schilthuis in 1936. 41 As the two
'"'i,ls 'material". and "balance" suggest, the 6A$MITElf
Q, m(p, - p)
" Q, = r + p (51)
8
ti Equa tion (50) is a basic type of mate rial balance
:lj " O equa tion that evaluates the original quan tity of liq-
~
.,<:> ' uid in the tank as a function of two pressures, the
6 cumulative produced gas-liquid ratio, and the ratio
~
"',o,.,
of gas and liquid volumes originally prese nt in the
"'I container. Two unknowns, Q, and m, and only
a
... 4 single equa tion require that a second equa tion be
~
.,.,
developed to obtain specific values of Q,.
~ A plot of Equa tion (51) is shown in Figure 62 .
~
Q, I!
~
'.; ... i'""t,
The fraction of the liquid recovered at any pressure
is shown for several values of m and r.
(1) Solubility of the gas in the liquid is propor- As the cumulative produced gas is the difference
tional to the absolute pressure (Henry's between the original gas present and the gas remain-
Law). ing at time O,
~
(2) Free gas in the bottle obeys the ideal gas law, rQ, -
5.615
. P,mQ,(l + acP,) + Q cP 0 -
14 7 0 14 7
PV = C.
P[Q,(I + acP,)(1 + m) - (Q, - Q,)(l + acP)] - (Q, - Q,)cP
(3) The liquid phase is incompressible. which simplifies to
(4) Vapor pressure of the liquid is zero. Q, -
Q,[r + (0.382 - c)P + 0.382acPJ
(5) The bottle temperatur e and the atmospheric 0.382m[(l +acP,)(P, -P)J + cP, - c(l+ 0.382aP,)P + 0.382acP
temperatur e are the same (assumed 60F in (52)
example). or, in terms of the fraction of original liquid that has been pro-
(6) Solution of each increment of gas causes an duced, &
equal volumetric increase of the liquid phase. Q,
Q, =
The linear solubility-pressure relation can be ex- 0.382m[(l + acP,)(P, - P)] + cP, - c(l + 0.382aP,)P + 0.382acP
pressed in petroleum engineering units as follows: r + (0.382 - c)P + 0.382acP
(53)
S ... cP
1/000
Volume (bbl) of liquid phase - Q,(l + acP,)
Volume (bbl) of gas phase = mQ,(l + acP,)
Total original volume (bbl) - Q,(l+ acP,)(1 + m)
The standard cubic feet of free gas originally present 1 1/SOO
5.615
= !4. 7 P,mQ,(I + acP,)
...
~
The cubic feet of dissolved gas originally present - Q,,cP o
Total cubic feet of gas originally in the system
5.615
i
.... /000 .
= 14.7 P,mQ,(l + acP,) + Q.d',
At any subsequent time, 0, the volume of the /SOO
phases can be expressed as follows:
Volume (bbl) occupied by liquid phase = (Q, - Q,)(l + acP)
Volume (bbl) occupied by free gas = total volume - volume oc-
cupied by liquid phase 0 1.0
0 02 0.4 0.6 0.8
Volume (bbl) occupied by free gas
- Q,(l + acP,)(I + m) - (Q, - Q,)(1 + acP) FRACTION OF ORIGINAL SURFACE OIL PRODUCED
Standard cu ft of free gas
Figure 64-Curves Illustrating the Effect that Size of Gas Cap and
5.615 Producing Gas-Oil Ratio Have on the Quantity of
- 14. P[Q,(l + acP,)(l + m) - (Q, - Q,)(l + acP)J
7 Produced Oil. Calculations Made by Pop Bottle
Cu ft of dissolved gas - (Q, - Q,)cP Material Balance Method
MATERIAL BALANCES
TAB LE 34. ExA MPL E APPLICATION 99
OF Pop HOT TLE MAT
Problem: To calc ul~t e reservoir ERIA L BAL ANC E MET HOD TO RES ERV OIR
pressure as a function of the BEH AVI OR
pro pert ies: qua ntit y of produced surface oil, mak ing use of the
following assumed fluid
Original pressure
Original gas solubility 2750 psia
Formation volume of origi~ 765 cu ft per bbl tank oil
Solution: nal saturated liquid phase
1.420 bbl per bbl tank oil
Fro m Equ atio n (53)
Q, 0.382m[(l + acP ,)(P , -
P)) + cP, - c(l + 0.382oP
Q, - r + (0.382 - c)P + 0.382acP
,J + 0.382acP
765
2750 - 0.278
C -
cPo =- 765
0.420
a -
765 - 5.49(10-J aP, - 5.49(10-)2750 - 1.51
acPo =- 0.420
Case 1. No free gas at star t. Producing ratio equal to orig
inal solution ratio.
m =0 r "'"765
Q, 765 - 0.438P + 0.58 -)P
Q, = 765 + O. l04 P + _ 2(10 .
Case 2. No free gas at start. 0 582 (1o-) P Curve 1, Figure 64 1s a plot of . .
Producing ratio equal to twice this equ atio n.
original solution ratio.
m =0 r =- 1530
Q, 765 - 0.438P + 0.58 2(10 -)P
Qo = l530 + O.l04P + _ ' . .
Case 3. One~third free gas at 0 582 (I0- 4)P' Curve 2, Figure 64 is a plot of . .
start. Producing ratio equal this equation.
to original solution ratio.
Q, - 1510 - 0.709P + 0.58 2(10m =- 0.5 r = 765
-)P '
765 + 0 . 104p + 0 _582 ( 10 _4)P'
Q,, .
Case 4. .10 % free gas at star Cur ve 3, Figure 64 1s a plot of this
t. Producing ratio equal to orig equation.
inal ratio.
931 - 0.500P + 0.582(10-)P= 0.111
Q, m r - 765
- =-
Q, 765 + 0.104P + 0.582(10-)P Curve 4, Figure 64 is a plot of
this equation.
r,.Q, = mu'oQo_!_
Vo
+ raQo - !V [u'oQo(l + m) - u 1 (Qo - Q,)
- (W - w)] - (Q. - Q,)r
which simplifies to
Q,[u' + (r. - r)v] - (W - w)
Qo=---"""---'--'-''---.-'--'-~'-----"-'~ (54)
Figure 66-Scbematic Representation of Idealized Reservoir and u' - u'o +mu'o(~ -1) + (r,, - r)v
0
Assumed Solubility and Volumetric Relations Used in
Reservoir Material Balance I Derivation where r,. - cumulative net produced gas-oil ratio, cu ft per bbl of
tank oil
r ""' dissolved gas-oil ratio, cu ft per bbl of tank oil
When the same approach is used as in the pop
bottle derivation, the original phase volumes and It is important to note that the terms Q,, Q., r,
gas quantities at pressure P, are as.follows: and u' are results of the type of flash separation
Liquid phase, volume (bbl) ""' u' oQo conditions employed at the surface to separate gas
Gas phase volume (bbl) = mu' oOo. and oil. The last two factors would be obtained
----- from curves similar to the lower curves in Figures
Total original volume (bbl) = u'c.Qo(l + m)
58 and 60. Further, it should be realized that to
where u' 0 == formation volume of original liquid phase, bbl per bbl
of tank oil say a reservoir originally contains Q, barrels of tank
Q bbl of tank oil originally present
0 - oil implies that all the reservoir fluids will be proc-
Standard cu ft of free gas originally present = mu'Q o o-
1 essed through a particular surface trapping arrange-
Standard cu ft of dissolved gas originally
ment. If, at some time during the depletion of a
present reservoir, the surface trapping conditions are
changed, it then becomes necessary to reevaluate
Total gas {standard cu ft) = mu, oQo -1 + roQ,, Q., u', and r in terms of the new separation process .
where v0 = volume, in bbl, of one cubic foot of free gas when Equation (54) considers the saturated liquid arr,,
under reservoir pressure and temperature conditions free gas as two separate hydrocarbon quantities \,.,;
1 14.7 T the reservoir. It does not differentiate completely
Vo - 5.615 Po 520 Zo between produced gas originating from the gas cap
- 5.03(10-') r:. and produced gas that was originally dissolved in
the liquid phase. The term (r, - r) in the denomi-
r0 = original dissolved gas-oil ratio, cu ft of surface gas per
bbl of tank oil nator of Equation (54) represents the volume of
free gas at pressure P which was originally dissolved
At any subsequent time, IJ, and at reservoir pres-
in oil at pressure P,. The v in this term should be
sure, P, the volume of the phases can be expressed as
obtained by using a compressibility factor, Z, that
follows:
is representative of gas coming out of solution.
Liquid phase volume (bbl) - u'(Q. - Q.)
Free gas phase volume = total volume - liquid phase volume -
net water encroachment
The middle term in the denominator, mu',(~ - 1).
MATERIAL BALANCES
101
clearly represents the expansion of gas cap material.
jected freely into or with draw n from the botto m
If the gas in the gas cap is of a composition different
of the reservoir.
from gas originating in the oil band , or if the pres-
At the original pressure P,, the volumes of the two
sure in the gas cap is materially different from that
hydrocarbon phases are expressed as
in the oil band , the v of the middle term will not
be the same as the v in the third term. However, Volume of liquid phase (bbl) = u',,Q,,
Volume of free gas phase (bbl) = mu Q 1
as the middle term v's appe ar as a ratio, the effec 0 0
t If a quan tity
of differences in values caused by composition is of mercury is with draw n from the seg-
minimized. The v appearing in the num erato ment, causing the pressure to drop from P, to P,
r
cann ot be defined rigorously because gas in the the volumes occupied by the original phases will be
net produced gas-oil ratio, r., is not specified as to Volume of original liquid phase (bbl) = uQ,,
origin. Volume of free gas phase (bbl) = mu',,Q,,.!
where u is the two-p hase forma tion volum e of what
Reservoir Material Balance, II satura ted liquid. (Note the difference herein the origin ally was
use of u and u'
of the previous metho d.)
A much simpler derivation of a material balance
equa tion can be made by evalu ating the expansion The sum of the expansions of the original free gas
of the original liquid phase and the free gas phase as phase and the original liquid phase plus the volum
e
reservoir pressure is changed. Referring to Figure of any encroached water, W, caused to flow into the
segment by the reduction in pressure is, of course,
Q
66, let it be assumed that we are observing the
equal to the volume of mercury withdrawn.
Mathematically, this can be represented as
Total expansion =s mu'oQ,, (t - 1) + Q,,(u - u',,) +W
Let us now assume that gas, oil, and wate r are
s.MWU '' produced from the segment while the pressure re-
,...,..
.... ,,,....
110,11111, - - , - I
WAr11,r
mains cons tant at P. This can be done by rein-
jecting the mercury removed previously. Ob-
o.
viously, to keep the pressure constant, the volum
e
of reinjected mercury will be equal to the sum of the
l'ltOOtlOo .,,.,,.... , r produced gas, oil, and wate r volumes, all measured
, lllfllf/lJM IIIII
ro, , ,,., at pressure P. These quan tities are expressed as
follows:
Volume of produ ced water (bbl) = w
Volume of produ ced tank oil and gas original1y in
solutio n at P,,
in that tank oil (bbl) = uQ,
Volume of free gas produ ced is-equal to the differe
nce
the net produ ced gas and the gas originally dissolv between t, )
ed
produ ced tank oil. At pressu re P, volume of produ in the V
gas = Q,(r,. - r,,)u ced free
Thus, the volum etric contra ction of the system
- Q,(r., - r,,)u
+ uQ, + w
As the expansion of the idealized reservoir in the
Figure 66-Sc bema tic Repre sentat ion of Idealize
Reservoir and first stage of the process must be equal to the con-
Assum ed Solub ility and Volum etric Relati ons Used traction in volume in the second,
in
Reser voir Mater ial Balan ce II Deriva tion
Equation (55) is the familiar Schilthuis material petroleum reservoir coupled with the movements of
balance equation. This equation also can be ob- these or similar phases into or out of the reservoir.
tained from Equation (54) by substituting u - The equations say nothing regarding movement of
(r0 - r)v for the liquid phase formation volume, u', fluid volumes within the limits of the reservoir.
of Equation (54). The difficulty of rigorously To put it another way, in developing Equation (55),
defining the v terms is the same as that expressed an imaginary boundary was placed around a volume
in the previous section. of earth and an attempt made to estimate the
For many purposes, it is advantageous to separate quantity and type of fluid contained within that
the net produced gas-oil, r., into its two components: volume by observing the manner in which pressure
the produced gas-oil ratio, r"' and the injected gas- changes within the confines of the boundary as
oil ratio, r,, giving fluids are moved across it. Unfortunately, it is
Q,{u + (rp - r 0 }v- r,v,] - (W - w)
possible to observe and measure only the fluids
(56) flowing across the boundary face at the earth's
Q. -
u - n' 0 + mu'"
(
t- 1
)
surface; fluids flowing across it at other locations
can be evaluated only by indirect calculations.
If G is the cubic feet of surface gas injected into the
Equation (55) contains three unknown fluid quan-
reservoir during a gas injection program, G can be
tities: the original tank oil in place, Q.; the volume
expressed in terms of an injected gas-oil ratio.
of water entering the reservoir, W; and the volurrw"'\
Thus, of the original gas cap in the reservoir mu'0 Q.._J
G = Q.r;
then Obviously, it is impossible to arrive at a unique
Q Q,[u + (r, - r 0 )v] - Gv, - (W - w) (
57
) solution to the material balance equation without
1)
0
..... 0
I
~
G.0 HCO '-
i;j t
IYXJ tti 6 ij
1200'1:
~ .,
II ~
I;: q
1100,;,
4~
(>
I.)
I
IGW~ !!t
t
900
ii!
2~
=::!
q ;:::
.,. R .;, ~
g
~
194 4 194 6 l!J4" 194 8 194 9
l'OO !
0 I.>
Figure 67-Pressure-Production Hist
ory of the Douglas Field
and shown gra phi call y in Fig ure
67. These dat a the net volume of san d above the
show tha t from April 1, 1943, to
the field pro duc ed 19.5 million bbl
Jan uar y 1, 1950, was available to hydrocarbons
wat er tab le tha t
. Core analysis
Q
of tan k oil, 27.0 results were also of assistance in
billion cu ft of gas, and 4.5 mil eva lua ting the net
lion bbl of water. san d volume.
In so doing, the ave rag e reservoir
pressure dropped The results of por osit y, fluid satu
from 2750 to 1720 psia, alth oug rati ons , and
h the rate of pres- inte rsti tial wat er test s made on core
sure decline was lowered ma teri s from selected
ally afte r April, wells were used with the above san
1948, by the inje ctio n of 1250 MC d volume to de-
F of gas a day. termine the volumes of hydrocarb
Estimation of Original Hydrocarb ons initially in the
ons in Place. field. Fro m an average por osit y
To eva lua te the wat er influx term of the sand, 21.4
, W, in Equ atio n per cent, and an average inte rsti tial
(57), it is necessary to eva lua te wat er of 32 per
ind epe nde ntly the cent, the tota l hyd roc arb on volum
original tan k oil in place, Q and e was calculated
0, the rati o of free to be
gas phase volume to liquid pha
se volume, m.
These values ord ina rily are obt ain 96,o oo. 43,5 60. 0.214(1 - o.:i21
ed from a stu dy of 5.615
_ ( , bbl
- 108.5 IO) .
104 BEHAVIOR OF OIL FIELD HYDROCARBON SYSTEMS
Core nnnlysis results indicated that 7.0 per cent of 1-1.7 (i.ifl
0. i07 = 0.000!11 hhl 1wr 1u ft
5.fWl 2750 fi20
the h,drocarhon volume was occupied by a free
gas phase; therefore, the volume of saturated oil (compressibility factor from Figure f:i2). The quan-
amounted to 0.93 108.5(10') = 100.9(10 6 ) bbl. tity of free gas in the reservoir, measured under
The volume occupied by free gas amounted to surface conditions, therefore amounted to
0.07 10~ ..'i(lO') = 7.G(lO') bbl.
7.6(10")
From the data presented in Figure GO, the forma- - - - 8 O<J( Io~) standard cu ft
0.000!)4 . ' .
tion volume of the original saturated liquid is 1.422
bbl per bbl of tank oil. Therefore, the original tank Properties of the Original Hydrocarbons in Place.
100.9(10 6 ) Results of PVT tests on a bottom-hole liquid sam-
oil in place amounted to 1. = 71.0(10') bbl.
422 ple that had a bubble point pressure of 2,520 psia
If it is assumed that the compressibility of the orig- at 18.5F are given in Chapter 7. Methods were
inal free gas is equivalent to that of gas coming out o'utlined by which the data could be ext.ended to give
of solution at 2750 psia, the gas volume factor, 1.,, the properties of a liquid having a saturation pres-
relating the barrels of reservoir space occupied by sure of 27,50 psia at 18.5F, the original reservoir
one cubic foot of surface is liquid in the present example.
1943
June 2750 0 0 765 1.4220 0.940
Sept. 2735 0.220 0.168 765 1.4247 0.947
Dec. 2720 0.310 0.235 765 1.4275 0.950
1944
March 2690 0.652 0.499 765 1.4333 0.966
June 2655 1.202 0.926 770 1.4402 0.980
Sept. 2620 1.952 1.534 786 1.4474 0.990
Dec. 2585 2.500 2.012 806 1.4562 1.010
1945
March 2550 3.152 2.633 835 1.4626 1.025
June 2475 3.778 3.266 865 1.4803 1.065
Sept. 2420 4.652 4.245 0 910 1.4935 1.082
Dec. 2360 5.357 5.157 0.051 962 1.510 1.110
1946
March 2275 6.205 6.257 0.097 1007 1.536 1.155
June 2225 7.030 7.415 0.152 1055 1.553 1.180
Sept. 2150 7.950 8.891 0.227 1117 1.577 1.225
Dec.
1947
March
2085
2000
8.750
9.502
10.375
11.873
0.302
0.452
1185
1248
1.602
1.635
1.270
1.326
0
June 1920 10.350 13.570 0.827 1310 1.670 1.390
Sept. 1860 11.278 15.235 1.072 1338 1.699 1.440
Dec. 1810 12.158 16.322 1.563 1343 1.726 1.486
1948
March 1770 12.862 17.345 1.802 1350 1.750 1.520
April 13.205 17.870 0 1352
June 1750 13.780 18.615 2.354 0.71 1353 1.763 1.540
Aug. 1745 14.329 10.356 2.502 1.35 1363 1.765 1.547
Oct. 1745 14.995 20.371 2.675 2.10 1358 1.765 1.547
Dec. 1740 15.652 21.310 2.773 2.82 1362 1.769 1.550
1949
Feb. 1740 16. 150 22.075 3. 182 3.60 1366 1.769 1.550
April 1735 16.850 23.078 3.557 4.38 1370 1.772 1.553
June 1735 17.495 23.980 3.852 5.15 1371 1.772 1.553
Aug. 1730 18.150 24.975 4.073 5.95 1375 1.775 1.560
Oct. 1725 18.853 26.003 4.253 6.72 1378 1.778 1.565
Dec. 1720 19.554 27.000 4.473 7.48 1381 I. 781 1.670
MA TERI AL BALANCES
105
T\\'o- phase fonna tion \'olume faetorn of the origi-
the values of ,, shown in colum n 10, Table :i:;.
nal reser\ 'oir liquid are listed in colum n 8, Table :l5,
The gas volume facto rs in colum n 12 have been
and are sholl'n as a funct ion of press ure in Figur e GS.
calcu lated on the basis of 0.60 gravi ty injec tion gas.
These value s were obtai ned by calcu lation from the
In this instan ce the comp ressib ility facto rs were
smoo thed "Y" curve funct ion prese nted in colum n
obtai ned throu gh use of pseud o-crit ical const ants for
7, Table 30, and show n in Figur e 68. Use of the
a 0.60 gravi ty gas and the comp ressib ility facto r
" Y" funct ion great ly simplifies the work of tabu-
chart show n in Figur e 15.
lating value s of u at press ures less than 2750 psia.
For exam ple, at 25,50 psia, Relative Phase Volumes. The ratio of the
reserv oir volum e origin ally occup ied by free gas
V = 2.i-17; 1.-122 (2i50 - 2550) phase to the volum e occup ied by liquid can be
II = J.422 + i7.n
2550 = 1.4626, evalu ated by sever al meth ods. Core analy sis re-
sults show ed that 7.0 per cent of the reserv oir sand
The gas volum e factor s, ,,, show n in Figur e 68
were calcu lated by using the comp ressib ility facto rs volume conta ined free gas. Cons equen tly, the
.
show n in Figur e 52. It was assum ed that the gas value of m 1s _0.07_ = 0.07,5.
cap mate rial \\'as of comp arabl e comp ositio n to the 1 0 07
When a free gas phase does not exist in the reser-
gas comin g from soluti on in the reserv oir oil when
voir, the value of min Equa tion (57) is zero. How-
the comp ressib ility facto rs were used to calcu late
ever, the fact that unde rsatu rated fluid is produ ced
Q
TABLE 35-(Co ntinue d)
10 11 12 13 14 15 16 17
Intcted Gas
u+(r"-r )v0
Q,[u+(r p-r o)v]
olume (u.u'0 )+mi.(; 1) Q{u-d.+md{ Gross Water Net Water
bbl per bbl
Factor, V;
bbl per cu ft Gv, bbl per bbl
0
~. -1) J Encr-0achment Encroachment
tank oil Obi (IO-~) (103) bbl (lO-S) tank oil bbl (IO-)
w ~W-,w)
bbl (10-e) bb (10-S)
1.4220 0 0
1.4247 0.313 0
1.4275 0.0035 0.248 0.065 0.065
0.443 0.0066 0.469 -0.026 0.026
1.4333 0.935 0.0142
1.4451 1.737 1.008 -0.073 -0.073
1.4672 0.0227 1.612 0.125 0.125
2.864 0.0311 2.208
1.4966 3.742 0.0421 0.656 0.656
2.989 0.753 0.753
1. 5344 4.836 0.0502
1.5858 5.991 3.564 1.272 1.272
1.6504 0.0713 5.062 0.929 0.929
7.678 0.0876 6.220
1. 729 9.262 1.458 1.458
0.1073 7.618 1.695 1.644
1.816 11.268 0.1384
1.895 13.322 9.826 1.539 1.442
2.008 0.1582 11.232 2.242 2.090
15.964 0.1873 13.298
2.135 18.681 2.893 2.666
0.2174 15.435 3.548 3.246
2.275 21.617 0.2567
2.414 24.985 18.226 3.843 3.391
2.524 0.2991 21.236 4.576 3.749
28.466 0.3337 23.693
2.584 31.420 5.825 4.753
0.3658 25.972 7 .011 5 44!>
2.639 33.916 0.3939 27.960 7.758 f. <)56
2.669 0 .....
2.675
2.683
36.779
38.330
1.665
1.670
1.182
2.254
0.4091
0.4119
29 046
29.245
s9.333
903 6.549
6.831
40.231 1.670 3.507 0.4119 29.245
2.694 42.166 1.675 4 .723 10.154 7.479
0.4162 29.550 10.666 7.893
2.701 43.621 1.675 6.030 0 .4162
2.712 45.697 1.680 29.550 11.223 8.041
2. 713 7.358 0.4195 29.785 12.111 8.554
47.464 1.680 8.652 0.4195 29.785
2.727 49.495 1.685 10.026 12.879 9.027
2.737 0.4233 30.054 13.488 9.415
51.601 1.690 11.357 0.4269 30.310
2.748 53.734 1.695 12.679 14.187 9.934
0.4305 30.565 14.963 10.490
106 BEHAVIOR OF OIL FIELD HYDROCARBON SYSTEMS
"
culated pressure curve and the true pressure curve, of the liquid 500 feet below the interface could be
such that the standard deviation of the difference as low as 2920 psia. Because of the static pressure
was only 3 psi, could be obtained by: exerted by the liquid column, the reservoir pressure
was 3160 psia at 500 feet below the gas-oil interface.
(a) An aquifer permeability of 46 millidarcys and
Thus, the saturation pressure at this depth could
an original reserve of Q. barrels, or
vary anywhere between 2920 and 3160 psia.
(b) An aquifer permeability of zero and original The importance of correctly weighting PVT data
reserve of 2.6 Q0 barrels. can be appreciated from the previous discussion of
material balance methods. For example, a reser-
Thus, when one considers the difficulty of evaluating voir originally having a liquid undersaturated
an" average pressure" during the early life of a field throughout, but in which the liquid phase is pro-
and the properties of the aquifer surrounding the gressively more undersaturated at lower points in
field, it would appear that original reserves calcu- the reservoir, should not be represented by a P-V .
lated on the basis just discussed are open to curve that has a distinct "break" at a particular
skepticism. saturation pressure. Insofar as the reservoir be-
havior is concerned, the liquid elements at different
Effect of Reservoir Structur.e on Material Balance
Calculations depths go through their bubble points at different
times;. conse.quently, any P-V curve weighted tur
Reservoirs that have thick sands, or reservoirs represent the gross reservoir should exhibit
with thin sands but with steep dips, present their gradual change over the pressure range of all the
own problems in material balance calculations. In bubble points.
addition to the problem of how best to average the A second point to be considered in regs.:u to mate-
pressures throughout the reservoir, the problem rial balance calculations on high relief pools is that
arises of how best to represent the "average" PVT of possible gravity drainage effect$, Burtchaell 50
behavior of the hydrocarbon in the reservoir. Not points out that the material balance approach in-
only may the properties of the reservoir fluid differ volves an assumption that the liquid phase within
from place to place, but the fluids can, and often the sand will not materially change its location
do, vary at different elevations within the reservoir. during the life of the field. Where steep dips and
Sage and Lacey 49 have presented data from which high permeabilities are involved, so that the liquid
the change in composition in a static oil column and gas phases can migrate toward top and bottom
can be calculated as a function of distance below a of the reservoir, it becomes necessary to take this
gas-oil interface. Such differences in fluid com- factor in to account in weighting the PVT properties
position have been observed in several California of the reservoir material.
fields by noting that the producing gas-oil ratio of
wells high on the structure (fluid still undersatu- Summary of Princi1,>al Points
rated) is greater than the gas 0 oil ratio from wells The primary object of this chapter was not to c( )
located structurally lower. Also, the OAPI gravity cuss in detail the many ways of making mater'nrr'
of the produced tank oil often decreases with depth. balance calculations. Rather, it was to point out
Apparently there is no manner by which one can how laboratory PVT data are used in several of the
forecast the saturation conditions that exist through- more common methods.
out a reservoir. The method of Sage and Lacey For discussion purposes, material balance equa-
indicates only the limiting case of the saturation- tions can be placed into one of two types. The first
depth behavior, and is based on the existence of type, which is exemplified by the work of Tarner,
gravitational equilibrium between the many hydro- treats the reservoir liquid phase and reservoir gas
carbon components. For example, calculations for phase as separate identities. Consequently, one
one California reservoir, in which the liquid phase requires the pressure-gas solubility and the pressure-
was saturated at 3000 psia and 185F at the gas- formation volume factor of saturated liquid at pres-
oil interface, indicated that the saturation pressure sures over the complete range of the study, as well
MATERIAL BALANCES
as the volumetric be ha vio 109
r of the free reservoir gas.
Th e seC'ond typ e is exemp ve rt these surface volumes
lified by the ma ter ial to "in pla ce" volumes
ba lan ce eq ua tio n pu bli she mu st be ob tai ne d from
d by Schilthuis. He re, flash tes ts. Nowhere in
the original liquid an d the the ma ter ial ba lan ce eq
original gas in the reser- ua tio ns exemplified by
voir are the tw o main flu Eq ua tio ns (54) an d (57)
ids whose physical pro p- do differential da ta occur
ert ies mu st be known. dir ect ly. Tr ue , a pic tur e
Th us, when using the of the va po riz ati on proc-
second me tho d, it is nec ess occurring in the res erv
essary to know on ly the oir ma y en tai l differential
sol uti on gas-oil rat io, for ma vaporization, bu t the cal
tio n volume factor, an d cu lat ed volumes are of
the pressure-volume relati necessity in ter ms of flash
on of the original reser- va po riz ati on s co nd uc ted
vo ir liquid phase, an d the un de r surface conditions.
vo lum etr ic be ha vio r of the Th is requires tha t for-
original reservoir gas phase ma tio n volumes an d sol
. ubilities be de ter mi ne d
Because the "v olu me tri c ba un de r the tra pp ing condition
lan ce " mu st be ba sed s tha t exist in the field.
in part. on the volumes of If at some tim e du rin g the de
liquid an d gas wi thd raw n ple tio n of the res erv oir
from the reservoir an d me the surface tra p condition
asu red at the surface, the s are changed, it the n
sol ub ilit y an d volumetric becomes necessary to tak
factors necessary to con- e thi s int o acc ou nt when
ev alu ati ng the pro pe rtie s
of fluids.
C)
0
u
GLOSSARY RELATING TO HYDROC
ARBON BEHAVIOR
nt an elevated pressure and temperature state (res- properties. Those which are iridependen t of the
ervoir). Formation volume factors are further quantity of material and therefore describe its con-
specified as to liquid phase (single phase) or gas dition at a particular state are called intensive
plus liquid phase (two phase) formation volumes. properties.
Gas Gravity: Ratio of the molecular weight of the Pseudo-critical Pressure and Temperature: Fic-
p;as to the molecular weight of air, 28.97. titious critical pressure and temperature values
Ideal Gas (Perfect Gas): A fictitious gas conforming ascribed to a multicomponent system in order that
to the equation of state: the reduced pressure-volume-temperature states of
the system conform to the reduced states of pure
PV = RT gases.
M
Reduced Pressure and Temperature: The ratio of
in which P is absolute pressure, V specific volume,
the pressure in a system to the critical pressure (or
M molecular weight, T absolute temperature, and R
pseudo-critical pressure) of the system. Reduced
the universal gas constant. An ideal gas has a com-
temperature is the ratio of the temperature of a sys-
pressibility factor of unity under all conditions (see
tem to the system's critical temperature.
"Compressibility Factor"). Actual gases never
assume all of the properties of an ideal gas, although Reservoir Rock: Any rock which contains a com-,
some gases approximate the ideal behavior at low mercially exploitable concentration of hydrocarbm(j
pressures.
Residual Oil: The liquid remaining in a PVT cell
Injection Gas: Gas injected into a reservoir. at the completion of a differential process at or near
the reservoir temperature. Also, the liquid remain-
Liquid Saturation: The extent to which pores in a
reservoir rock are filled with liquid. Saturation ing in the reservoir at the end of a specified process.
ordinarily is expressed as a per cent of pore volume. Saturated Liquid: A liquid which is in equilibrium
with vapor at a given pressure and temperature
Permeability: A measure of the capacity of a porous
state.
material to transmit fluid. The unit of permea-
bility is the darcy. A material has the permea- Saturated Vapor: A vapor which is in equilibrium
bility of one darcy when one atmosphere pressure with a liquid at a given pressure and temperature
differential across one centimeter length causes a state.
viscous flow of one cubic centimeter per second of a
Saturation Pressure: The pressure at which vapor
fluid of one centipose viscosity through a cross sec-
and liquid are in equilibrium. Saturation pressure
tion of one square centimeter.
often is used synonymously with bubble point pres-
Phase: A homogeneous body of material which dif- sure or dew point pressure.
fers in its intensive properties from its neighboring
Shrinkage: The decrease in volume of liquid pet 1
phases. leum caused by the release of solution gas and'r,,t'
Production (Well Effluent) : The material which is by the thermal contraction of the liquid. The
produced in liquid and /or gaseous phases at the well shrinkage may be expressed. '
head. (a) as a percentage of the stock tank volume, or
(b) as a percentage of the reservoir oil volume.
Production Gas-Oil Ratio: The ratio of natural gas
production rate to crude oil production rate, ex- Shrinkage Factor: The reciprocal of formation vol-
pressed as cubic feet per barrel measured under ume factor, expressed as barrels of stock tank oil
standard conditions. per barrel of reservoir oil.
Properties, Extensive and Intensive : Properties Solution Gas-Oil Ratio (Dissolved Gas-Oil Ratio) :
which are directly proportional to the amount of The volume ratio obtained by dividing the amount
material making up the system are termed extensive of gas which would be liberated from a system exist-
GLOSSARY RELATING TO HYDROCARB
ON BEHAVIOR
113
ing originally as a liquid by the liqu
id remaining geneous syst em the intensiv
afte r the syst em is reduced to stan dar d e properties vary only
atmospheric in a continuous man ner with
conditions. respect to the exte nt
of the system. A heterogeneous syst em
is made up
Stock Tank Oil: Crude oil in equilib of a num ber of homogeneous par ts,
rium with a cha and abr upt
por tion of its evolved gases at stan dar d nges in the intensive properties occ
atmospheric surf ur at the
conditions. ace of con tact of the homogeneous par
ts of the
system.
Systems, Homogeneous and Hetero
geneous: A Undersaturated Fluid
bod y of mat ter with finite boundaries : A liquid or vap or capable of
tha t repre- holding additional gaseous
sents the mat eria l und er consideration. or liquid components in
. In a homo- solution at the specified
pressure and tem per atur e.
\_)
/
u
LIST OF SYMBOLS
,:
!I
' '
116 BEHAVIOR OF OIL FIELD HYDROCARBON SYSTEMS
V1(oV)
n Net
ume per volume per F, oT , p Produced
a (alpha) Proportional to B Bubble point condition or prope rty
/j (beta) Coefficient of compressibility, volume Dew point condition or prope rty
u
REFERENCES
i
118 BEHAVIOR OF OIL FIELD HYDROCARBON SYSTEMS
(35) White, R.R. and Brown, G. G.; "Phase Equilibrium at High (43) Coleman, S., Wilde, H. D., and Moore, T. W.; "Quantitati ve
Temperatur es," Ind. Eng. Chem., 34, 1162 (1942) Effect of Gas-Oil Ratios on Decline of Average Rock Pressure,"
Trans. AIME, 17 4 (1930)
(36) Sage, B. H., Hicks, B. L., and Lacey, W. N.; "Tentative
Equilibrium Constants for Light Hydrocarbo ns," API Drilling (44) Cook, A. B.; Part II, Magnolia Oil Field, Columbia County,
and Production Practice, 386 (1938) Arkansas, "Derivation and Application of Material-Ba lance
Equations," U.S. Bureau of Mines, R. I. 3720 (1943)
(37) Dodson, C. R. and Standing, M. B.; "Prediction of Volu-
metric and Phase Behavior of Naturally Occurring Hydrocarbo n (45) Woods, R. W. and Muskat, M.; uAn Analysis of Material
Systems,'' API Drilling and Production Practice, 326 (1941) Balance Calculations," Trans. AIME, 160, .124 (1945)
(46) Hurst, W.; "Water Influx Into a Reservoir and Its Applica~
(38) Sage, B. H. and Lacey, W. N.; 11 Apparatus for Determinat ion tion to the Equation of Volumetric Balance," Trans. AIME, 101,
of Volumetric Behavior of Fluids/' Trans. AIME, 174,102 (1948) 57 (1943)
(39) Rzasa, M. J. and Katz, D. L.; ' 1 Calculation of Static Pressure (47) Old, R. E., Jr.; 11 Analysis of Reservoir Performanc e,"
Gradients in Gas Wells," Trans. AIME, 160, 100 (1945) Trans. AIME, 161, 26 (1943)
(40) Vitter, A. L., Jr; 11 Back Pressure Tests on Gas--Condensate (48) Brownscombe, E. R. and Collins, F.; "Estimation of Re~
Wells," API Drilling and Production Practice, 79 (1942) serves and Water Drive from Pressure and Production History,"
Trans. AIME, 186, 92 (1949)
(41) Schilthuis, R. I.; "Active Oil and Reservoir Energy," Trans.
AIME, 118, 33 (1936) (49) Sage, B. H. and Lacey, W. N.; "Gravitatio nal Concentrati on
Gradients in Static Columns of Hydrocarbo n Fluids/' Trans.
(42) Tarner, J.; "How Different Size Gas Caps and Pressure AIME, 132, 120 (1939)
Maintenanc e Programs Affect Amount of Recoverable Oil," Oil (50) Burtchaell, E. P.; "Reservoir Performanc e of a High Relief
Weekly, 114, No. 2 (1944) 32 Pool," Trans. AIME, 186, 171 (1949)
u
l)
APPENDIX I
sures. Chart 2 presents the bubble point pressure Chart 3 allows one to calculate the formation volume
of a dissolved gas system as a function of gas-oil of bubble point liquids. From these three calcu-
ratio, gas gravity, tank oil gravity, and temperat ure. lating charts, it is possible to make engineeri
ng
The chart also may be used in the reverse direction estimates of the physical properties of multi-com
po-
to estimate the solution gas-oil ratio of a system nent systems under conditions existing in
petroleum
from its saturatio n pressure and temperat ure. reservoirs.
u
.i
!
APPEN DIX II
CHA RT EQUATIONS
At the time this book was prepared (1950), two Figure 15, Page 24. Compressibility Factors for
important tools of the petroleum enginee r were a Natural Gases
JO-inch slide rule, and a compilation of charts, ta-
bles, and miscellaneous information. In fact, the The Standing-Katz 1 compressibility factor chart
has been set up for compute r solution by a number of
large size of the book was selected specifically to
permit the use of large-size charts.'
individuals. Takacs 2 recently published a compar- 0
Today, slide rules have pretty much been replaced
ison of eight comput er program s to calcula te
.
I
;
Z-factors as a function of pseudoreduced pressure ''
by small electronic calculators and by large comput-
and temperature. 3 11 His comparison of 180 chart
ers. This reduces the need for charts but opens up the
and calculated values showed that the Dranchuk and
need for mathematical expressions of the charts. The
reprinting of the book by the Society of Petroleum Abou-Kassam equations 11 had the smallest absolute
error, 0.316 percent. A FORTR AN program for this
Engineers of AIME offered the opportunity to "up-
calculation is given in the original article and will not
date'' it by adding this Appendix.
be repeated here.
Most of the equations that follow were developed
by the author by simple curve-fit procedu res. They A modification of the equations published by Brill
are not the most precise equations that can be de- and Beggs 9 yield compressibility factor values accu-
veloped but are conside red to be adequate for their rate enough for many engineering calculations. They
intended use. The goodness of fit of the equations to have the additional advantage of being program-
the originating chart is indicated by the average mable for the HP-67 and TI-SR52 calculators. The
percent difference of samples of chart and equation equations are:
values and the standard deviation of the percent Z =A + (1-A)/d' + C p,,,." ( l)
differences. Note that these criteria apply to values where
read from charts and do not indicate the degree of A= l.39(T,,,. -0.92) 0 5 -0.36T,, ,. -0.101 (2)
agreeme nt between original data values and equation
B = (0.62-0. 23 T,,,.)p,,,. + [ 0.066 - 0.037] P,,,.'
values. (T,,,.-0.8 6)
The format of the Appendix is simple. Equations
will be given in the order of the corresponding
+ 0 .32
10 9(7',,,. 1)
p,,,.' (3)
chart's appearance in the book - the gas compres- C = (0.132 - 0.32 logT,,,.) (4)
sibility factor chart on Page 24 being the first. In D = antilog (0.3106 - 0.49 T,,,. + 0.1824 T,,,.2) (5)
addition , equations are given at the end of the Ap-
pendix with which to calculate viscosity of reservoir Agreem ent between calculated Z values and chart
oils and gases.
Z values was tested at T,,,. values of I .2, I .3, I .4,
1.5, 1.7, 2.0, and 2.4andp ,,,. values of 0, I, 2, 3, 4,
122 BEHAV IOR OF OIL FIELD HYDRO CARBO N SYSTE MS
5, 7, 9, 11, and 13. Results were as follows. of Wic~ert and Aziz. 12 If YoM is the gravity of the
whole gas mixture, the gravity of the hydrocarbon
T,,,., Pvr Range
portion is
1.3 <Tm< 1.7 I .2 < Tw < 2 .4
0 <p,,,. < 5 0 <p,,,. < 13 - y,,.,, - 0.967yN, - 1.52Jco , - l.l8yH2S
'YuHC -
( 10)
Number of points 24 70 } - YN2 - Ye<>2 - Y112S
c.,percent 0.19 0.02 The pseudocriticals of the whole gas mixture are
S(C.), percent 0.96 1.18 then
Llmnx percent +2.63 -3.93
The above relationships are not valid outside the
PvcM = (1-yN -yco2 -Ytt 2s) Pvc11c + 493 YN 2
2
Naturally occurring reservoir and surface oils do Figure 25; Page 37. Density Correction for
not have large weight percents of dissolved ethane Thermal Expansion of Liquids
and methane. The values of w2 = 1.05; w1 = 5.4
The units of density (lb/cu ft) to be subtracted
shown in Table 7, Page 38, are typical of a reservoir from the pseudo-liquid density value at pressure p
oil having a bubble-point pressure of about 3,500
and 60F when the liquid is heated to temperature T
psia. Thus, there is limited use for a large portion is given by the equation:
of Figure 23. Furthermore, the form of Equations
(16 and 17) is not the best for solution for P,+ and !:,.p1, = [0.0133 + 152.4 (P,,,. +t,.p,.)- 2"] (T-60)
Pi+ Therefore, values obtained from Equations (16 6
- [ 8. l( 10- )-0.0622 10-o.0764(Pp1. +ap,J ](T-60)'
and 17) were curve fit in the region 40<p 3 + <60;
(2 l)
O<w 2 <10; O<w 1 <l6, with the following results:
where (Pt, +/:;.pr) = pseudo-liquid density at pres-
P,+ = P,+ (1-0.01386 w2 -0.000082 w2 2 ) surep and 60F.
+ 0.379 w2 + 0.0042 w,2 (18)
The agreement between calculated values of /:;.p,,
P,+ = P,+ (l-0.012 w 1 -0.000l58 wi') and those read from Figure 25 at temperatures of
+ 0.0133 w, + 0.00058 wi' ( l 9) 80, 100, ... , 240F is shown in the following
tabulation:
f' 1nsities calculated from Equations (18 and 19)
~ree with densities calculated from Equations (16 !:,. S(!:,.)
and 17) within 0.05 lb/cu ft and are more precise than (P,,,. +!:,.Pr) (eercent) (eercent)
values read from Figure 23. 60 0.58 2.04
55 -0._68 0.75
Figure 24, Page 36. Density Correction for 50 -0.65 l.53
Compressibility of Liquids 45 -0.27 3.42
40 -l.45 l.74
The units of density (lb/cu ft) to be added to the 40 to 60 -0.50 2.10 (50 values)
14.7 psia/60F pseudo-liquid density (Figure 23) to
obtain a pseudo-liquid density at pressure p and 60F Figure 26, Page 40. Apparent Liquid Densities
is given by the relationship: of Natural Gases
The chart by Katz 14 showing the apparent liquid
!:,.p =/0.[67 +[6.[8[[Q-0.0425PpL)/__r__\ density of a natural gas of gravity Yn when dissolved
" \ \),000/ in oil of AP! degrees gravity has the relationship:
- 0 0 [ lo 299 + 263 lQ-0.0603 Pp,\
\ J
(__f!__)
1,000
2
This relationship is the basis for the bubble-point The fit of the above equations and Beal's Figure 8
pressure chart discussed in Appendix I and located in is indicated by the average difference (25 values at
0
the back of the book. API of 15, 25, ... , 55) of -1.58 percent and a
Reservoir Oil Viscosity Correlations of Beal and
standard deviation of differences of 9.4 percent. The
of Chew and Connally large standard deviation results primarily from the
15 AP! values. If these are left out, the equations fit
Estimating reservoir oil viscosity has been done the remaining 20 values with is:= 1.62 percent; S(Li)
primarily by means of the correlations of Beal 17 and = 4.92 percent.
Chew and Connally 18 until recently. Beal's correla- Chew and Connally's equation for the viscosity of
tion is used to obtain the gas-free crude oil viscosity saturated oil (bubble-point oil) is
(stock-tank oil or "dead oil") at reservoir tempera- . 00 = A (.,o)' (33)
ture as a function of its AP! gravity. This' 'dead oil''
is then adjusted for the amount of dissolved gas it where A and bare functions of solution gas-oil ratio.
contains at reservoir pressure by means of Chew and Fit of A and b values given in their Table 3 are
Connally's correlation. Should the reservoir pres- A= antilog {R,[2.2(10- 7)R,-7.4(10- 4)]} (34)
sure be greater than the bubble-point pressure of the
b= 0.68 + 0.25
hydrocarbon system, a further adjustment is made to I0 8 62( IO-') R., 101.1 (10 3 ) R.,
the bubble-point liquid viscosity to account for the
0 degree of undersaturation in the reservoir.
0.062
+ ~lo"'a.cc,4:-C(c-'IOc=_-;sc,)----:R:-., (35)
The more recent work of Beggs and Robinson 19 Equations (34 and 35) reproduce the 12 A and b
apparently gives a better estimate of saturated reser- values in Table 3 with the following precision:
voir oil viscosity. However, it should be pointed out
i"i" S(~)
that neither Beal's nor Beggs and Robinson's corre- Coefficient (percent) (percent)
lations take into account the chemical nature of the A -0.21 2.79
hydrocarbons that make up the crude oil part of the b -0. IO 0.36
reservoir oil. Chemical nature is an important Beal's Figure 11 shows a correlation of the slope
parameter in correlating liquid hydrocarbon viscos- of the viscosity-pressure curves in the undersatu-
ity behavior, and is usually indicated by the UOP rated oil region as a function of bubble-point viscos-
Characterization Factor. Until some factor that ex- ity. A fit of his curve is
presses chemical nature of the crude oil is included in
/J-o - .,, = 0.024 . 1., + 0.038 . 0. 56 (36)
reservoir oil viscosity correlations, it is doubtful that 0.001 (p-p,) oh ob
Errata
M.B. Standing, author of Volumetric and Phase
0 Behavior of Oil Field Hydrocarbon Systems, SPE,
Dallas (1977), notes that Eqs. 13 and 35, which appear
on Pages 122 and 125, respectively, contained
typographical errors. The corrected equations appear
below.
0.68 0.25
b=~~~~~+~~~~~
108.62(10-s)R, 101.1(10- 3 )R,
0.062
+ ...................... (35)
!03.74(10- 3 )R,
INDEX
A Ethane:
apparent density in liquid solutions, 34
Apparent convergence pressure, 45, 111 compressibility factor at 140 F, 21
as function of C1+ molecular weight, 50 P-T diagram, 1
Apparent liquid density, 33, 111 P-V diagram, 2
physical properties, 4
B vapor pressure, 3
Ethane-n. heptane system, 2, 5
Binary systems, 2
Bubble point, 1, 51, 111
F
Bubble point pressure, 111
Douglas bottom hole sample, 77
Factor:
maximum in ethane-n. heptane system, 5
compressibility, 20
Butane:
deviation, 20
physical properties, 4
super expansibility, 20
vapor pressure, 3
Flash gas liberation, 111
Flash process, 9, 111
C Flash vaporization tests:
Composition:
dissolved gas system, 77
critical locus of ethane-n. heptane system, 5
gas condensate system, 58
Compressibility:
Formation volume factor, 83, 111
Douglas bottom hole sample, 77
of condensate systems
factor of gaseous mixtures, 20, 111
-empirical relation of Sage and Olds, 62
of methane, ethane, propane, 21
-empirical relation of Standing, 63, 119
Condensate, 57, 111 '
of dissolved gas systems
0 system, relation of critical temperature and operating tempera-
ture, 57
Conditions, sta.ndS.rd surface, 116
. -empirical relation of Standing, 42
Fugacity, 44
Constants:
equilibrium, 43 G
behavior of "plus" fraction constant, 45
empirical, 45 Gas:
from Raoult'.s and Dalton's laws, 44 constant per mole, 20, 116
ideal, 45 gravity, 112
Corresponding states, theorem or, 21 meter factor, 25
Cricondentherm, 5 Natural gas-compressibility, 106
Critical pressure, 1, 5, 111 -viscosity, 7
of paraffin hydrocarbons, air, N;, 0 1, CO 2, H 2S, H 20, 4 natural gas-crude oil mixtures
Critical state, 5, 111 -bubble point pressures, 86, 119
ethane-n. heptane system, 5 -composition of gas phase, 51, 53, 72
Critical temperature, 1, 5, 111 -composition of liquid phase, 5, 53, 72
or paraffin hydrocarbons, air, N 2, 0 2, CO 2, H 2S, H,O, 4 ..,....formation volume, 42, 63, 119
Crude oil: natural gas-natural gasoline mixture, P-T diagram, 6
compressibility, 34 perfect, 112
thermal expansion, 34 Gas cap, effect on material balance calculations, 106
Gas-condensate system, 57
D Ga.see:
sampling from surface traps, 16
Dalton's law, 43 sampling from flow lines, 17
Decane, properties, 4
H
Deviation factor, gases, 20
0 Dew point, 1, 111
pressure, 111
Heptane:
physical properties, 4
Differential ga:s liberation, 111 P-T diagram, ethane-n. heptane system, 2
Differential process, 9, 111 P-T composition diagram, ethane-n. heptane system, 5
Differential vaporization tests: vapor pressure, 3
gas condensate systems, 58 Hexane:
dissolved gaa systems, 79 physical properties, 4
Dissolved gas, 111 vapor pressure, 3
Dissolved gas drive, 7 4, 111
description of process, 8 I
Dodecane, vapor pressure, 3
Ideal gas law, 20, 112
Ideal solutions, principle of, 33
E
L
Equilibrium constant, 43
behavior of "plus" fraction constant, 45 Liquids:
empirical, 45 compressibility', 34
from Raoult's and Dalton's laws, 44 sampling from surface traps, 16
ideal, 43 sampling with bottom hole sampler, 12
use in phase calculations, 50, 51, 52 thermal expansion, 34
viscosity, 7
130 INDEX
Liquid density: ----<iriterion of good sample, 12
calculation using ideal solqtion principles, 33 fl.ow line, 17
calculation using partial liquid volumes, 38 -advantages and disadvantages of, 18
calculation using apparent gas densities, 39 surface recombined, 14
calculation using California crude correlations, 40 -advantages and disadvantages of, 18
pseudo-liquid density, 33 Sample containers, plans, 16
Liq1.lid saturation, 112 Sample tips, 17
Samplers, bottom hole, 11
M Saturated liquid, 1, 112
Saturated vapor, 1, 112
Material balance, example of, Douglas Field, 104, 105
Saturation pressure, 95, 112
Material balance, methods,
Shrinkage, 112
pop bottle, 97
Shrinkage factor, 112
reservoir I, 99
Solubility,
reservoir II, 101
Douglas saturated oil, 80, 87
tank, 96
Henry's law, 98
Methane: Solution gas drive, 74
compressibility factor at 140 F, 21
Solution gas-oil ratio, 112
physical properties, 4
Stock tank oil, 113
-apparent density in liquid solutions, 34
Super-expansibility factor, 20
Moles, pound, 20 Systems:
N heterogeneous, 113
homogeneous, 113
Nonane:
physical properties, 4 T
s
Samples:
bottom hole, 11
-advantages and disadvantages of, 18
PR0PR T!S OF NATURA L HYDROC ARBON MIXTUR ES OF GAS AND L!QU/0
EXAMP LE
REQUIR ED:
Forma.fio n vofum, of llitt gaJ plus liquid
pha.Jtts of a. 1500 CFS mi'x.lure, ga.s grnvily
0.80 1 tank oil gra.vily ,-4o'"AP I, a.f i?OO F
and I 000 PSIA
PROCED URE:
S#a.rfing a.f !he lttf'I Jldtt of fhe charf,
proceed horiZonfa .!ly a.long lhtt 1500 CFB line
lo lhtt 0.80 gaJ gra.vily !init. From lhii poinf
drop nrfica.ll y lo !hit 40 AP/ fii,e. Proceed
horizonf ally lo 200 "F. and f'rom Iha.I poi"nl
drop fo the 1000 PS/A prttuun fir,e.
requirttd forma.lion volumit 1$ f'ound lo 6e
Thtt .8
<
,!!
5.0 6arrtt/J per ba.rrttl of lank oil.
'-0
'
t
"''"
.
l
/
~
rn-r
~
~s
l'
5
,". \
~Ji'~ 0
r,l
Ilf ~o\; CHAHr I
'7.,o~
k'::..:.l[,,,
,ii
,oi
COMPLIM ENTS or
CALIF'OHNIA RESEARC H CORPORATION I
illll i Copyri ght 1947
Califor nia Researc h Corpora tion
I
/
PROPERT IES OF NATURAL HYOROCA li'BON MIXTURES OF GAS ANO LIQUID ,,.~pf!q;""-'
BUBBLE POINT PRESSUR E
EKAMPLE
REQUIRED :.
Bubb/fl poif'lf prnsurt! a.I 200 F
or a. I/quid ho.vt"ng ct ga.s -oil ro.fiO of"
JSO CFB, a. gas gro.vify or O. 75, a.nd
a. fo..nk oil gro.vify of" 30 AP!.
PROCEDU RE:
Sfa.rfil"lg a.I !ht! lt!f"f s1"dt! of" fhe
cha.rf, proct!ed horizonfa!ly a.lol"lg fhe 350
CFB h'rle lo a 90.s gra.vify of" o. 75. From
fhi's po,nf drop verHca.lly fo fhe 30"API
lint!. Proceed horizonla.lly f"rom fhe
/Ql"lk oil gra.vify sea.le io fhe 200 p
line. Thtt requir~ prttssure 1$
f"oul"ld fo be 19.30 PS/A .
CHART 2
COMPLIM NTS OF
CALIFORN IA RESEARCH CORPORATION
BUBBLE POINT PRESSUR E -
Copyrig ht 1947
Californi a Research Corporati on
PROPERTIES OF NATURAL 1-fYOROCARBON MIXTURES OF GAS AND LIQUID
EXAMPLE
REQUIRED:
Formafion volume o.f 2ooF of a
bubble poinf liquid hoving a gas-oll
ro.fio or JSO CFB, o gas gra.vify or 0.7S7
and a fa.nk oil gravify oF 30 AP/.
PROCEDURE:
Sfarfing o.f fhe lerf side or fhe cho.rf,
proceed horizonfally along fhe JSO CFS
li'ne fo o. 90.s gravify or o. 75. From fhis
poinf drop verfic:o.lly fo fhe .30 ~Pl line.
Proceed horizonfo.lly from fhe fank oil
gro.vify scale fo fhe .?ooF line. The
required -f"ormo.fion volume 1$ fiHJnd to btT
1.22 barrel per ha.rrel of fa.nk oil.
CHA/Ir ~
I.!.IJ'T
2EJ
COMPLIIIENTS OF
CAL/FD/IN/A RESEARCH COHPOIIATION
EXAMPLE
REOIJIREO:
Density of 200F of o bubble point liquid having o
gos-017 ratio of 350 CFB, a gos gravity of 075 and
a tonk oil qrovity of 30 AP/.
PROCEDURE:
From Chart 3 determine formation volume of 1.22 barrels
per barrel of tank oil. Starling a /ell side of chorl
proceed hor/zontol/y along the 350 CFB line too' gas
gravity of 0.75. From this point drop vertically to the
30 AP! line. Proceed horizontally from I/le tt:mk oil
gravity scale to the formation volume of 1.22. The re-
quired density is found to be 47.6 pounds per cubic
foot.
IIIIIIIII
rm
,=w::p:+ ! 4-i ,r I i i
'
C:
tr ...
CHART 4
COMPLIMENTS OF
CALiFORNIA RESEARCH CORPORATION
' I I !-: , -t-tt-rn
-I I I ; I
-+ CT' ' ., '
-,--,"CTH-trr,:i~
.;..+-f--r~.c! , . . :. -:
4*~' SPEC'lf-lG' 6/f~ .. ,,
:;:+["C-'-1'' I,' '--~~
1.:,': ,,,,,, '1, . .,,.-, ..... , ,
,., .,.,.. ov
""L"
r-,_v ~"
'"'"~c1.:,:eL'...:.1__
=-, [
f ', COPYRIGHT @ 1959
California Research Corporation