Summary
Equation of State (EOS) predictions for gas condensate systems require
extended analysis beyond the heptanes plus (C7+) fraction. In the
absence of experimental data, several schemes have been proposed to extend
these compositional data based on the observation that a single straightline
relationship exists between log of mole percent and molecular weight for these
pseudocomponents or single carbon number (SCN) fractions.
An examination of compositional analysis for gas condensate systems showed a
discontinuity in the relationship between mole percent and molecular weight at
C8 and C13. As a result, two straight lines are needed
for a more accurate description of SCN composition; one from C8 to
C12, and the other from C13 and beyond. When applied, this new
universal observation gives an improved prediction of SCN composition. An
average absolute deviation of less than 6.0% between the predicted and
experimental composition was obtained using parameters from two straight lines.
From a single straightline relationship, this difference was as high as
36.0%.
This new observation provides the basis for defining the partial
experimental analysis required for applying extended models for a more accurate
description of SCN composition. For the logarithmic distribution, a partial
analysis to C20+ is required to define the change in slope at
C13 and beyond. For the three parameter gamma distribution function,
a partial analysis is required up to C14 and splitting can be
applied from C14+ and beyond. These widely used models are not
suitable for extending the C7+ fraction.
Introduction
With the increasing emphasis on liquid natural gas (LNG), natural gas
liquids (NGLs) and liquid condensates during the last 15 years, gas condensate
reservoirs became increasingly important. A combination of laboratory studies,
such as Chromatographic; true boiling point (TBP); and pressure, volume,
temperature (PVT) analyses became necessary for characterizing these reservoir
fluids and evaluating their volumetric performance at various pressure
depletion stages.
An accurate description of pseudocomponent compositions is an integral part
of the reservoir fluids characterization process. For gas condensate systems,
these data are applied with Equations of State (EOS) to evaluate gas and
condensate reserves and production for field development and surface facility
design. The evaluations rely on a tuned EOS formulated from adjustment of SCN
compositions. Good quality compositional data require minimal adjustment for
obtaining the best match between predicted and experimental phase behavior
data.
Very often the required extended compositional data are unavailable
experimentally and are generated from mathematical relationships. Literature
(Ahmed 1989; Danesh 1998; Pedersen et el. 1989) has shown that a plot of SCN
composition against molecular weight produces a continuous exponential
relationship for gas condensate systems. This observation also led to a
generally accepted representation of a single straightline relationship between
log of mole percent and molecular weight for these SCN fractions.
Based on this observation, very useful functional approaches called
"splitting" schemes (Whitson 1983; Pedersen et al. 1984) were devised
to describe the composition of these SCN fractions in the absence of
experimental data. Although splitting schemes are applied from the
C7+ or last available plus fraction, a review by Danesh (1998)
stated that a partial analysis is first required followed by the application of
these schemes. To date, literature has not specified the SCN or last plus
fraction for terminating a partial analysis.
From an examination of compositional analysis for gas condensate systems,
this paper describes a different universal trend from the single straightline
relationship between log of mole percent and molecular weight. Also, the last
plus fraction is defined for terminating a partial analysis. A total of 22
compositional data sets to C20+ were examined. Six of these were
generated experimentally from separator samples taken in Trinidad (Hosein 2004)
and 16 were taken from PVT lab reports generated from samples taken
worldwide.
© 2009. Society of Petroleum Engineers
View full textPDF
(
1,064 KB
)
History
- Original manuscript received:
31 July 2007
- Meeting paper published:
11 November 2007
- Revised manuscript received:
8 September 2008
- Manuscript approved:
1 October 2008
- Published online:
2 March 2009
- Version of record:
26 February 2009