SPE-175594-MS

Creating Variance Cube from Seismic Data and Converting to

Transmissibility Multiplier Pattern to Assist History Match
John Gao, and Bart Schrijver, Wintershall Holding GmbH

Copyright 2015, Society of Petroleum Engineers

This paper was prepared for presentation at the SPE Reservoir Characterisation and Simulation Conference and Exhibition held in Abu Dhabi, UAE, 14 –16 September
2015.

This paper was selected for presentation by an SPE program committee following review of information contained in an abstract submitted by the author(s). Contents
of the paper have not been reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material does not necessarily reflect
any position of the Society of Petroleum Engineers, its officers, or members. Electronic reproduction, distribution, or storage of any part of this paper without the written
consent of the Society of Petroleum Engineers is prohibited. Permission to reproduce in print is restricted to an abstract of not more than 300 words; illustrations may
not be copied. The abstract must contain conspicuous acknowledgment of SPE copyright.

Abstract
This paper demonstrates how to utilize seismic data to create a variance cube with different values and
convert it into a transmissibility multiplier pattern to assist the history match in dynamic simulation.
Lateral variations of the seismic signal can be related to various reasons, such as faults respectively
associated fault damage zones and sedimentological bodies like channel features. Dissimilarities within
the seismic data can be emphasized by calculating a variance attribute. The different value ranges within
the variance cube can in the chosen examples be associated to the fault plane itself and different intensities
of fault related rock alterations or channel characteristics, e.g. erosional planes. These features can result
in improved or reduced lateral transmissibility as poor or good rock properties coincide with poor or good
transmissibility for fluid flow. The variance value patterns can be analyzed and categorized according to
the discrete value ranges. Then the categories are converted to reference numbers, e.g. variance reference
number 1 represents a good transmissibility (e.g. a channel body), 2 represents reasonable and 4 represents
the poorest class (e.g. an intense faulted zone). Subsequently the variance reference number is converted
to a transmissibility multiplier value. During the history match process, these multipliers are adjusted in
order to improve the history match.
The advantage of this approach is that it does not only change the transmissibility multiplier in one cell,
but changes the transmissibility multiplier value of each cell that belongs to one type of rock quality (one
variance reference number or one transmissibility group) in one calculation. This is especially useful
where there is limited history data and the variance data indicates similar patterns throughout the variance
cube. This method can speed up the history matching as it is a batch change of transmissibility multipliers,
similar to the fault transmissibility multiplier method in ECLIPSE but with a wider application. In a case
study a comparison was conducted with using the variance cube / transmissibility multiplier patterns and
confirmed the method workable.
In conclusion using the variance cube to create transmissibility multiplier patterns is an efficient way
to assist and improve history matching, especially in fields with limited well control.
This paper presents a new way of creating transmissibility multipliers based on seismic data (variance
attribute) and assisting and improving the history match process
2 SPE-175594-MS

Introduction
Lateral variations of the seismic signal can be related to various causes; such as faults/fault damage zones
or sedimentological bodies such as channels. Dissimilarities within the seismic data can be emphasized
by calculating a variance attribute. The different value ranges within the variance cube can potentially be
associated to a fault plane itself with different intensities associated to the surrounding fault zone or other
features such as channels. These features can be translated into transmissibility multipliers to improve or
reduce lateral transmissibility for fluid flow during dynamic model history match.
The advantage of this approach is that it does not only change the transmissibility multiplier in one cell,
but changes the transmissibility multiplier value of each cell that belongs to one type of rock quality (one
variance reference number or one transmissibility group) in one calculation. Therefore the time spend for
history matching can be optimised as it is a batch change of transmissibility multipliers, similar to the fault
transmissibility multiplier method in ECLIPSE but with wider application. It is especially useful in the
areas where there is limited history data and the variance data indicates similar patterns throughout the
variance cube.

Creation of Variance Cube and Transmissibility Multiplier Cube

The workflow of translating the seismic data into transmissibility multiplier data is shown in Figure 1.
From the seismic data the variance attribute is calculated in three dimensions, representing the

Figure 1—Workflow of translating seismic data to transmissibility multiplier data

trace-to-trace variability over a defined area. The outcome is an interpretable lateral change pattern in
acoustic impedance. Similar traces produce low variance amplitude in an unbroken reflection layer,
whereas discontinuities are characterized by high variance amplitudes. Variance attributes provide a direct
measurement of dissimilarity rather than the inferred similarity (coherency) of seismic data, producing
sharper, more distinct results. Figure 2 (T. Randen, et al, 2005) shows a variance cube (Figure 2 (b))
created from seismic data (Figure 2 (a)), highlighting faults or sedimentological bodies as these features
create discontinuities.
SPE-175594-MS 3

Figure 2—(a) Seismic cube (b) Variance cube

Figure 3 shows the workflow of a case study. Starting in Figure 3(a) with the variance cube of the
study area. The variance attribute in this example was subdivided in four groups (x1 to x4). The groups
can be associated to variations in reservoir quality and/or tetonic overprint (direct fault zone or fault
related damage zone).

Figure 3—The conversion from a seismic variance cube (a) to transmissibility multiplier cube(c)

Thes variance cube is then transformed into a variance pattern cube (Figure 3(b)) in which group 1
categorises the unaltered rock matrix and 4 the severe damaged fault zone. The next step is to convert the
variance pattern to fluid flow abilites respectively transmissibility multipliers (e.g.1.0, 0.8, 0.4, 0.1) or a
transmissibility multiplier cube (Figure 3 (c)). Red color represents a normal transmissibility multiplier
without modification (multiplier value 1.0), dark blue characterizes the damaged fault zones with reduced
tranmissibility (multiplier value 0.1). During history matching these transmissibility multiplier values are
adjusted to mimic the reservoir behavior and to deliver a satisfactory match of the production history.
The variance cube can be generated using the built-in “variance cube” functions of, e.g. the Petrel
software, and subsequently be converted into the variance pattern cube by applying the Petrel property
calculator. There are different ways to convert a variance pattern cube into a variance transmissibility
multiplier cube with a keyword MULTX and transmissibility multifplier values. One way would be to
create a new property in Petrel representing the transmissibility multipliers. Another way is using a
manual approach where the keyword MULTX is included with which the transmissibility multiplier
values replace the numbers in the variance pattern cube file, latter imported in the dynamic model as input
data.
4 SPE-175594-MS

Application in History Match

Using the above described workflow can assist and optimise history matching. Figure 4 shows the
improved match of bottom hole pressures of two wells of a field study, testing the variance cube approach.
In both plots (a) and (b) the dotted lines show the measured pressures and the solid lines the simulated
pressures. The green lines show that in the beginning of the history match process the simulated pressures
did not match the measured pressures. After adjustment of the transmissibility multiplier values the history
match could be improved (red lines). Since the chosen workflow allows a batch change of transmissibility
multipliers (iterative process of adjusting and testing transmissibility multipliers) the time spend for
history matching can be optimized.

Figure 4 —Well bottom hole pressures, (a) well W1, (b) well W2

Discussions
Besides time optimization, one clear advantage of the described workflow lies outside well controlled areas.
The areal distribution of transmissibility multipliers based on varaiance data provides a meaningful geologi-
cally controlled approach compared to a pure statictical method or the traditional way by subdividing a study
area in a number of subareas with manually assigned transmissibility multiplier values.
The variance cube can also be used for other property adjustments, e.g. a permeability multiplier cube,
which is one of the most frequently adjustment parameter during the history match process.
Conclusion
Using the variance cube to create transmissibility multiplier patterns is an efficient way to assist and
improve history matching, especially for the fields with limited well control. It also provides a geoloci-
cally meaningfull way to assign transmissibility values outside well control areas. In addition the variance
data can also assist for other property matching like permeability multipliers.

Acknowledgment
The authors would like to thank the management of Wintershall for the permission to publish this paper
and their support.

References
T. Randen, et al, 2005, Mathematical Methods and Modelling in Hydrocarbon Exploration and
Production, Atlas of 3D Seismic Attributes, pp23
Schlumberger, 2006, Fault Resolution Improved for Norsk Hydro’s Varg Field, Case study: Variance
Cube software enhances imaging and reduces processing time