Equilibrium-ratio correlations

The equations for gas-liquid equilibria calculations and the definition of

equilibrium ratio (K-factor) were discussed in chapter 12. This chapter will
examine k-factor correlations.
The equation which defines k-factor:

j=

yi
xi

And the equation resulting from the combination of Raoults and Daltons
equations:

y i Pvj
=
xi p
Can be combined to give an equation for calculating the factor of a
component, j, in a mixture which behaves like an ideal solution.

K j=

P vj
p

The vapor pressure of component j, pvj, is solely a function of temperature.

Thus equation 14-1 shows that if ideal solution behavior exists, the k-factor of
each component simply depends on pressure and temperature.
Remember that one the principal properties used to define an ideal solution is
that the intermolecular forces of attraction and repulsion are the same between
unlike as between like molecules. This property does not exist in real solutions.
Molecular behavior in a real solution depends on the types and sizes of the
molecules which are interacting.

Therefore, the k-factor for a component of a real solution depends not only on
pressure and temperature but also on the types and quantities of other
substances present. This means that any correlation of k-factor must be based
on at least three quantities: pressure, temperature and a third property which
describes nonideal solution behavior. This property represent both the types of
molecules present and their quantities in the gas and liquid.
Attempts have been made to define this third property in several different
ways. We will look only a convergence pressure, which appears to be the most
convenient for the types of calculations required of petroleum engineers.
A large number of k-factor correlations have been proposed. We will present
only two: one graphical, the other in equation form. A partial set of k-factor
curves is given in appendix A. only 14 charts out of a total of 90 of this
correlation are given. These should suffice for illustrative exercises in this text.
Notice that the k-factor charts in appendix A apply to petroleum mixtures which
have convergence pressure of 5000 psia. The other charts of this correlations
(not reproduced in appendix A) are for use with mixtures with other
convergence pressure. A value of convergence pressure applicable to the fluid
of interest must be estimated in order to select the correct set of graphs.

Convergence pressure
Experimentally determined k-factors normally are plotted against pressure on a
log-log sacale. Figures 14-1 and 14-2 show equilibrium ratios of two typical
petroleum mixture for several temperatures.
The shapes of the curves in these figures are characteristic of most
multicomponent mixtures. At low pressures, the slope of each curve is
approximately 1.0. A slope of 1.0 for ideal k-factors is predicted by equation
14-1. Each curve passes throught a value of unity at a pressure very close to
the vapor pressure of the component at the temperature of the graph. This also
is predicted for ideal-solution behavior.
At higher pressures, the effects of nonideality are seen. The curve for each
component departs from a slope of 1.0. The curves tend to converge toward a
k-factor of 1.0.

Definition of convergence pressure

The values of pressure for which the k-factors appear to converge to unity is
known as the convergence pressure. If a mixture is at its true critical
temperature, the curves in reality will converge to a value of 1.0 at the critical
pressure. At any temperature other than the critical temperature, the
equilibrium-ratio curves actually do not extend past the bubble-point or dewpoint pressure of the mixture. However, the curves can be extrapolated to
determine the point of apparent convergence.

Sometimes convergence pressure is called apparent convergence pressure.

Although the convergence pressure is not equal to the critical pressure of the
mixture, to a certain extent it does characterize the properties of mixture.
Convergence pressure is useful in the correlation of k-factor data.

Estimation of convergence pressure

Several methods of estimating convergence pressure have been proposed.
These methods have been evaluated using laboratory data of petroleum
reservoir samples. The method described below is as accurate as any and is
the easiest to apply.
Convergence pressure of binary hydrocarbon mixtures may be estimated from
the critical locus curves given in figure 2-16. A similar curve which includes
multicomponent mixtures is presented in figure 14-3.

This method of estimating convergence pressure involves a trial and error

procedure in which a first trial value of convergence pressure is used to obtain
k-factors. A reasonable estimate of convergence pressure for a first trial value
may be obtained from:

C 7+4200
p k =60 M
Where Mc7+ is the molecular weight of the heptanes plus fraction.
The compositions and quantities of the equilibrium gas and liquid are computed
in the usual manner. Then the liquid is equated to a pseudobinary mixture
consisting of the lightest component of the liquid and a hypothetical heavy
component.
The lightest component of petroleum mixtures always is taken to be methane.
The hypothetical heavy component is represented by a critical temperature
and a critical pressure calculated as the weighted average of the true critical
properties of all components of the liquid except the lightest.
The critical properties of methane and the weighted-average critical properties
of the hypothetical heavy component are plotted on figure 14-3, and a locus of
convergence pressures for that pair is interpolated using the adjacent critical
loci as guides.

The estimated of convergence pressure is taken from the point this locus
crosses the temperature at which calculations are desired. If this convergence
pressure is not reasonably close to the first trial value of convergence pressure,
the procedure should be repeated. The estimated value to convergence
pressure can be used as the new trail value.

EXAMPLE 14-1: the compositions of the gas and liquid from a black oil at 1300
psia and 160F calculated with equation 12-17 are given in the table below. A
convergence pressure of 5000 psia was used to obtain the k-factors. What
value of convergence pressure should have been used for this mixture at
160F?

SOLUTION
First, the liquid composition must be expressed in weight fraction.

Second, delete methane and adjust the weight fraction of the hypothetical
heavy component. Then calculate weight-average critical properties.

Third, plot weight-averaged critical point on figure 14-3, interpolate a locus of

convergence pressures, and read pk at 160F.
A convergence pressure of 10,000 psia should have been used for this mixture.
See figure 14-4. Obtain k-factors at pk=10,000 and repeat the gas-liquid
equilibrium calculation.

When the operating pressure is considerably lees than the convergence

prssure, an error in the estimate of convergence pressure has little effect on
the resulting calculations. As operating pressure approaches convergence
pressure, however, equilibrium ratios become very sensitive to the
convergence pressure used and care must be taken in the selection of the
correct value.
The correlations from which the equilibrium ratio data in appendix A were taken
include charts for convergence pressures of 800, 1000, 1500, 2000, 3000,
5000, and 10,000 psia. When the convergence pressure for the mixture is
between the values for which charts are provided, interpolate between charts.
Interpolation is necessary when the operating pressure is near the convergence
pressure. At low pressure, simply use the chart with convergence pressure
nearest the value for the mixture.
Bloack oil usually have convergence pressure of about 10,000 psia, retrograde
gases about 5000 psia and volatiles oils about 7000 psia.