Gringarten SPE16388R Type-Curve Analysis: What It Can and Cannot Do
Gringarten SPE16388R Type-Curve Analysis: What It Can and Cannot Do
Gringarten SPE16388R Type-Curve Analysis: What It Can and Cannot Do
Type-Curve Analysis:
What It Can and Cannot Do
Alain C. Gringarten, Scientific Software-Intercomp
Summary. Type-curve analysis has been used for more than 20 years in the oil industry and for more than
40 years in hydrogeology. Yet, what it Involves and how useful it is in practice is still being debated within the
oil industry. This paper attempts to answer some of the questions that are most commonly raised about type
curves and type-curve analysis.
'
----------------- ---------~
,
,,
WRI._~"
_ _ _OUII _ _
...111 .... ~ . .-
--
--....
~ /.----------------: --~ ... ,
,,
I.II ,~if:g~~~~~~~~~1
41;: ,...- ••• - .__ : l "
~ "-i.-...;IiNv"t..i
~_.~~ L I ~;].:':/.'=":':":':'_~":'==:-:':-=-=..:'·=r!_J
W
_.....
WIll _ . vDlfICAl........,.. COMDUCTIVITY
CJ ..... _ ..".Ul"Va.
_ _ . .OUII
~ou.CTUlll! . ._
Z
_tQ~ _... <
-. LOG of
....
~ES8 l:
0
W
Fig I-Example 01 graphic type-curve match a:
:::l
---------
Q. WIU _ .. ~ ITOOIAQI_
. .11110 . . . . . . . . , _
'0 _ ooo.&l! ~ _v_
CJ
0
....
.... --,", ,, . ,D€fW.TI'IE
,,
,- /
F,.
LOG of ELAPSED TIME
3-Examples 01 log-tog plots lor variOus interpreta-
.
LOG of ELAPSED raE• tion modets
Fig 2-Pressure·denvallVe characteristic shapes 01 In- presumed to ex ist, which mayor may not be the
terpretation model components case: an apparent straight line through a range of data
does not necessarily prove the existence of a specific
flow regime. An analysis based on an erroneous
How Does Type-Curve Analrajs Compare straight line will of course yield totally erroneous
With Conventional Analysis? results. The most efficient procedure is to start with
Type curves describe the entire behavio r of Ihc the identification of the interpretation model. to
interpretat ion model corresponding to thc well and thc evaluate all the appropriate well and reservoir
reservoir and include thc various flow regimes that parameters with type-curve analysis, and then to
successively dominate during thc test. As a result . confirm the resu lts with straight-line analyses if
type-curve analysis provides all thc well and reservoir applicable.
parameters that can be obtained from well testing.
Conventional a nalysis methods. on the other Can Type-Curve AnalysiS Provide a
hand-and morc generally. analysis methods based on Unique Answer?
straight lines- are valid only for a specific flow Assuming that the tesl has been well designed so thai
regime. As a result, they provide only well and the range of available test dalll is adequate , the
reservoir paramete rs characteristic of that flow interpretation model can be determined with
reg line. reasonable certai nty. Its mcaning in terms of rese rvoir
If a given fl ow regime ex ists during a test, the description, on the other hand, may not be unique.
correspondi ng straight-line and type-curve analyses This nonuniqueness is inherent to the testing process
must yield the same value for the parameters that and is independent of the analysis technique used .
characterize that flow regime. In such a case, The interpretation model only indicates how many
straight-li ne methods are easier to use if the straight different media arc contributing to the flow process
lines arc defined with little ambiguity. and how these media interact. But the analysis usually
The difficulty with conve ntional methods is cannot tell how these med ia are distributed. For
determining the existence of the particular flow example, well test data can indicate that a reservoir
regime. In type-curve analysis. this is performed as exhibits a double-porosity behavior, but they usually
part of the identification of the applicable cannot tell if the reservoir is fissured or multilayered
interpretation model. In the conveTltional use of without additional nontesting information.
straight-line analysis methods, on the other hand, Assuming that the interpretation has been identified
there is no such preliminary step. The flow regime is correctly and that an adequate range of test data is
12 Journal of Petroleum Technology. January 1981
SPE 16388
available, type-curve analysis should yield a unique drawn is assumed a priori to represent the required
match. In practice, however, the lack of resolution of flow regime.
the log-log representation used for most type curves The same applies to the validation part because no
may create a problem. This can be resolved by validation is possible with conventional analysis. As
combining type-curve and straight-line analyses, as mentioned before, a first level of validation consists
mentioned before. of checking type-curve analysis results with straight-
If the range of data is inadequate, and in particular line analysis techniques for the flow regimes
if semilog radial flow has not been reached during the identified in log-log analysis. A second, more
test, a unique answer cannot be obtained with type- stringent check consists of comparing a Horner plot
curve analysis without a validation process (of course, of the data with one predicted from the matched type
no analysis is possible with conventional techniques, curve. As' a final check of the consistency of the
either). analysis, measurements taken during the entire test
(not just the period being analyzed) can be compared
with those calculated frpm the matched type curve.
Does Type-Curve Analysis Really Improve
To date, validation is not widely used with type-
the Analysis Process?
curve analysis. It is expected that this will change
A complete interpretation process should include (1) because the simple validation techniques described
the identification of the interpretation model, (2) the earlier significantly improve the value of an analysis
validation of the interpretation model, and (3) the by allowing the interpretation of data that could not
calculation of the model parameters. Improvements in be uniquely analyzed otherwise and by providing a
the interpretation process can essentially come from measurement of the confidence in the analysis results.
improvements in identification and validation.
For the identification part, type-curve analysis JPT
represents a drastic improvement, especially with This paper is SPE 16388, Technology Today Series articles provide useful summary
information on bOlh classic and emerging concepts In petroleum engineering, Pur·
pressure derivatives, because there is no identification pose: To provide the general reader with a basic understanding of a significant con·
in conventional analysis: whatever straight line is cept. technique. or development within a specific area of technology,