Flowing Gas Material Balance
Flowing Gas Material Balance
Flowing Gas Material Balance
Material Balance
L. Mattar, R. McNeil
Fekete Associates Inc.
Abstract
The traditional material balance (p/z) plot for gas pools
requires fully built-up reservoir pressures, obtained by shuttingin the wells. The procedure described in this presentation does
not require shut-in of wells. Instead, it utilizes information normally obtained but not usually used by reservoir engineers to
quantify the original gas-in-placethe daily gas production
rates and flowing pressures.
The classical pseudo-steady state analysis and its shortcomings are discussed. In addition, a new procedure called the
flowing material balance is introduced. This procedure consists of a p/z plot of the flowing pressure (as opposed to the
average shut-in reservoir pressure) versus cumulative production. A straight line drawn through the flowing pressure data and
then, a parallel line, drawn through the initial reservoir pressure
will give the original gas-in-place. A variation of this method,
using wellhead pressures (tubing and casing) is discussed and a
field example is included. The method is a very practical and
powerful tool for the early quantification of reserves.
Introduction
The determination of gas reserves is a fundamental calculation
in reservoir engineering. This information is crucial for the development of a production strategy, design of facilities, contracts and
valuation of the reserves. Classically, reserves are estimated in
three ways: volumetric, material balance and production decline.
The volumetric and material balance methods estimate original
gas-in-place whereas production decline yields an estimate of
recoverable gas.
Volumetrically determined reserves can be very imprecise,
because the method depends upon detailed knowledge of many
reservoir characteristics that are often unknown such as the areal
extent of the pool. The material balance method uses actual reservoir performance data and therefore is generally accepted as the
most accurate procedure for estimating original gas-in-place.
Once determined, the original gas-in-place can be used to reliably
forecast the recoverable raw gas reserves under various operating
scenarios. The production decline method also uses reservoir performance data but the result is an estimate of recoverable raw gas
reserves under the existing operating conditions. A change in
these operations, for example a lowering of the compressor suction pressure, can change the deliverability and the recoverable
raw gas reserves. The original gas-in-place is therefore difficult to
ascertain from production decline.
It has been understood for many years that estimates of original
gas-in-place are theoretically possible using measured gas volumes and flowing pressures. The goal of this paper is to take the
flowing material balance from a theoretical possibility to a practical reality.
52
Flowing Data
The flow of gases through porous media can be divided into
two major categoriestransient and stabilized. Transient flow
behaviour is dominated by reservoir characteristics such as permeability, skin, degree of heterogeneity, location of boundaries, etc.,
and a complex function of time (log time). Stabilized flow, on the
other hand, is dominated by reserves and a simple time function
(t).
High and medium permeability reservoirs reach stabilized flow
relatively quicklywithin a few weeksbut low permeability
reservoirs can take a year or more to stabilize. When a well is in
stabilized flow, its behaviour is represented by the pseudo-steady
state equations found in the well testing literature(2). At the heart
of all these equations lies the material balance equation, in one
form or another. The following discussion will describe several
procedures for determining the original gas-in-place. All these
methods are variations of the material balance equation, with different assumptions and approximations.
Because stabilized flow is dominated by reserves, it should be
possible to estimate the reserves in a pool, if gas rate, pressure and
time data are available during the period of stabilized flow. Thus
for medium and high permeability reservoirs, we have the potential to calculate the reserves from flowing pressure data without
having to shut-in the well and lose valuable production.
The Journal of Canadian Petroleum Technology
Practical ApplicationAlternative
Methods
The data set used in this paper was taken from a producing well
for which accurate wellhead pressure and gas volumes were
recorded for over five months. During the first two months of this
test the well flowed at a fairly constant gas rate of approximately
250 103m3/d. Tubing and casing pressures were measured daily
and the casing pressures were converted to sand-face pressures
using a multi-step Cullender and Smith(3) calculation.
A review of the well test data and the core analysis indicates
that this formation possesses a good permeability to gas (50 mD).
The time to stabilization (assuming a one section spacing unit) is
approximately two weeks indicating that the bulk of the production data represents stabilized (pseudo-steady state) flow.
The remainder of this paper makes use of the established performance based methods to estimate the original gas-in-place and
then presents several variations of the flowing material balance
procedure as practical alternatives.
2.929 10 2 q 10 3 p i 288
c s l o p e / 2 101.325
4
2 10 3 q p i 6 3
10 m
cslope
.....................................................................(1)
Figure 1 shows a plot of pressure squared (casing pressure converted to sandface) versus time in days. The slope of this line during the pseudo-steady state period is 780,000 kPa2/day.
Values for the gas rate, q, (250 103m3/d) and initial reservoir
pressure, p i, (12,900 kPa) are easily obtainable. However, a value
for the gas compressibility, c, is more problematic. In a gas reservoir, c varies as 1/p, but the pseudo-steady state analysis procedure
traditionally assumes the gas compressibility is constant at initial
conditions. This assumption is good for an undersaturated liquid
53
OGIP =
If the gas compressibility had been calculated at the initial pressure of 12,900 kPa, the OGIP would have been 94 106m3.
As indicated above, the classical pseudo-steady state analysis
has an inherent weakness with respect to the value of gas compressibility employed. This analysis procedure will estimate original gas-in-place optimistically unless a satisfactory method for
incorporating the impact of depletion on gas compressibility can
be developed.
Conclusion
The original gas-in-place determined from the flowing data
(pressure and production) is approximately 71 106m3. This is significantly different from the volumetric original gas-in-place calculation of 500 106m3 assuming a one section spacing and indicates the pool has a much smaller areal extent than one section.
The procedure presented in this paper provides a very practical
tool for estimating gas-in-place using data generally available in
normal production operations. In addition, production losses can
be minimized by not having to shut-in wells.
It is possible to determine original gas-in-place with reasonable
certainty when shut-in pressures are not available.
NOMENCLATURE
the original gas-in-place. For the data used in this example, the
approximate wellhead material balance plot is shown on Figure 6
and gives an OGIP = 71 106m3.
This is a simplified material balance plot based on wellhead
pressures (rather than reservoir pressures) and it ignores z factor.
Our experience with this type of analysis indicates that it is a very
practical and acceptable procedure.
Practical Observations
The practical application of this procedure requires that certain
conditions be met. First, flowing pressures in the reservoir must
have reached pseudo-steady state. Second, the well must be
choked such that production is independent of line pressure and it
must be flowing at relatively constant gas rates. Although pseudosteady state theory requires a constant gas rate, practice has shown
it does not necessarily have to be enforced when flowing pressure
is plotted versus cumulative production. Third, once the well
begins to track gathering system line pressures, it begins its terminal gas rate decline and the procedure is no longer valid.
Application of this procedure seems to work best for good permeability gas reservoirs not affected by external drive sources.
c
OGIP
pi
pwf
Q
q
slope
T
z
=
=
=
=
=
=
=
=
=
REFERENCES
1. H A V L E N A , D. and O D E H, A.S., The Material Balance as an
Equation of a Straight Line; Journal of Petroleum Technology,
August 1963.
2. EARLOUGHER, R.C., Advances in Well Test Analysis; Monograph
Vol. 5, SPE AIME, 1977.
3. CULLENDER, M.H. and R.V. SMITH, Practical Solution of Gas
Flow Equations for Well and Pipelines with Large Temperature
Gradients; Trans. AIME, Vol. 207, p. 281.
4. E.R.C.B, Gas Well TestingTheory and Practice; Energy Resources
Conservation Board, Alberta, Canada, 1979, Fourth Edition.
Authors Biographies
Louis Mattar is president of Fekete
Associates Inc. He has co-authored 28 publications including the ERCB guide G-3
(Gas Well TestingTheory and Practice).
He is a member of APEGGA, SPE and The
Petroleum Society. He graduated from the
University of Calgary with a M.Sc. degree
in 1973.
Summary
Several methods are presented for estimating the original gasin-place without shutting in a well. The calculated values for this
particular case are summarized below:
Method 1
Method 2
Method 3
Method 4
Average
55
OGIP
(106m3)
(PSS)
64
(FMB) Reservoir 71
(FMB) Wellhead 72
(Tank)
76
(AWMB)
71
71