The Gas Storage
Optimization Program
GASTOP
is a development of MARACO, Incorporated
The
team from Maraco, Inc. is proud to announce the release of GasTop, their new gas
storage design program. GasTop is a
Windows based software application that helps determine the optimal economic
design model of gas storage facilities. The
model presently used in GASTOP corresponds closely to the way that a gas storage
operator in California prices its service.
In designing a storage facility engineers try to find the optimal
economic balance of three required constituent elements:
1. amount
of cushion gas for the reservoir, or each reservoir, hosting the facility,
2. number
of wells in the reservoir, which in turn determines the number of wells to be
drilled,
3.
total compressor HP, and the optimal location(s) thereof.
Using
a gas reservoir simulator, an engineer can manually test different combinations
of these elements and select the combination that gives the best result. GASTOP
automates this selection process.
Defining a Gas Storage Cycle
A
gas storage cycle (displayed in the drawing to the left) in GASTOP consists of
one injection period (segment A&B),
followed by a withdrawal period (segment B&C).
Gas In Place (GIP) at the beginning of a cycle (point A) must be the same
as at the end of the cycle (point C).
The
time span of the injection phase is fixed, though the injection process can (and
must) stop earlier if the pressure in the reservoir reaches the maximum allowed
value or the injection rate falls below a prescribed cutoff rate.
In
the drawing to the left, gas is injected during time segment A&B
causing GIP to increase steadily. Injection is stopped at B so during time
segment B&C
GIP remains constant. In GASTOP,
withdrawal of gas from the reservoir stops if one of three conditions occurs:
1.
time span of the withdrawal phase reaches the maximum allowed,
2.
withdrawal rate falls below a prescribed cutoff rate, or
3.
GIP falls to the prescribed cushion gas volume
If
the withdrawal phase of a cycle stops because of conditions 1 or 2, and GIP is
greater than the prescribed cushion gas volume, the cycle is considered invalid.
In
GasTop, a gas storage facility may contain up to 5 reservoirs, which require the
following variables for each reservoir; (a) amount
of cushion gas for the reservoir(s) hosting the facility, (b) number of wells in
the reservoir, which determines the number of wells to be dilled, and (c) amount
of 1st & 2nd stage horsepower for up to 5 compressors
in the surface network.
To allow rapid screening of the potentially large number of variables, a
simplistic tank-type reservoir model is used to determine gas flow rates for a
selected set of variables in a design case. These rates are combined with the
revenue functions to make a full economic evaluation of the case.
The results of which include; investments in
wells & compressors, make-up cushion gas and surface processing equipment,
and operating costs of all equipment and activities.
A
sequence of increasingly sophisticated search methods is used to find the
variable set that maximizes NPV. The
revenue function used in GASTOP corresponds closely to the way US based gas
storage operators price their services. The
revenue stream is derived from charges for three services, which are: storage
capacity, injection capacity and production capacity.
GasTop also allows sensitivity studies (incremental economic
evaluation runs) for key variables and parameters.
The
GasTop Interface
GasTop
has been designed with a relatively short learning curve and an interactive Help
system that guides the user through some of the more complicated aspects of the
program. Easy-to use screens are
associated to all functionality, plus all input sections have graphical plots
that appear at the click of a button.
GASTOP
currently operates as a stand-alone application, but plans are underway to
interface it directly to the GMAN simulator and, other industry standard
simulators. This first commercial
release has been designed for simplistic modeling of the reservoirs and an
upgraded version is currently in development, which will handle much more
complex field and reservoir configurations.
Additionally, changes are also underway to include a standardized
European economic model.
For
GMAN Users
As
a complement to GasTop, GMAN has been modified to allow simulations to be made using daily
time steps (a run is one year or less). Using
restart files, a sequence of annual cycles can be simulated to determine steady
state conditions. GMAN.HST provides graphs of pressure and flow rate vs time for
all significant locations in the reservoir/flowline model.
New Feature!
GASTOP’s
Oil Material Balance Equation (MBE)
Knowing the hydrocarbon pore
volume (HCPV) correctly is an essential requirement to properly design and
engineer the conversion of a depleted oil reservoir into a gas storage
reservoir. Verification of the accuracy of any value of HCPV obtained from the
reservoir files is almost certainly lacking. The literature gives many examples
of highly erroneous volumetric estimates of HCPV caused by errors in core
measurements – porosity, connate water saturation, rock compressibility. The
accuracy of performance estimates of HCPV is equally unknown given uncertainty
in reservoir parameters, analyst proficiency and techniques used.
Ergo, an MBE feature has been
added to GASTOP that provides a convenient and comprehensive means of making a
“best estimate” of HCPV. The feature treats an oil reservoir’s MBE as the
equation of a straight line. The analytical basis for this approach is explained
in L.P. Dake’s “fundamentals of
reservoir engineering”, Sections 3.3 – 3.7, pp 78 - 97. Given the range
of uncertainty in each of the seven parameters listed in the left panel of
Figure 1(Each entry is most likely (ML) value, % variation plus or minus)
GASTOP’s MBE feature uses an error-minimizing procedure (Powell Search) to
determine the ML value of each parameter (‘Estimated’) and the corresponding
ML HCPV. The procedure fits a
straight line to Y(=F), X(=E) pairs derived from the historical production and
reservoir pressure data shown in Figure 2 using test sets of the seven
parameters. The blue line in the
graph on the lower left is the ML straight line and the red circles are the
observed (F, E) points. (F is fluid expansion and E is total underground
withdrawal both measured in reservoir barrels.)
GASTOP’S
MBE feature provides several plot options as displayed in Figure 3 shows. The
plot of (F/E) vs (1/E) is a particularly useful prescriptive device. A
horizontal plot, as shown here, indicates no water influx; a non-horizontal plot
indicates some water influx occurred.
Figure 3
Coming
Soon!
The
next official release of GASTOP will include the capability
for storing and utilizing all pertinent data for the wells. This will include
items such as: spud date, surface and bottomhole location, details of the
completion string and perforations, all pressure and production data, fluid
composition and levels, lithology and pertinent formation characteristics
including fracturing and distance to faults.
These changes are tied to the latest release of the GMAN, which is now
capable of modeling the individual cells of the reservoir.
In addition, Maraco is working with an American storage
company to modify GASTOP, so that it can be used as a scheduling tool.
The combination of optimal economic design and scheduling will result in
a very powerful and versatile gas storage modeling tool. Details of this
new version can be found on this website under GasPal.
To download the above information in a PDF
format, right click here and save the file.
