Summary
Brigham’s model has been used extensively in the petroleum industry for the
design and interpretation of interwell tracer tests. The model is based on
correlation and has included physical dispersion as an input parameter. In
spite of its limitations, the model is useful in estimating layer
heterogeneity, layer distribution, permeability contrast, and dispersion in the
reservoir. However, the model can only handle nonpartitioning tracers that
have no solubility in oil. With the advancement in partitioning tracer
technology and interpretation technique, interwell partitioning tracer testing
has gained its popularity, especially in China,1 for determining residual oil
saturation, Sorw, between wells. Partitioning tracer testing also finds
its application in environmental protection, where tests are routinely run to
determine the amount of nonaqueous liquid phase nonaqueous phase liquid (NAPL)
trapped underground due to spill or seepage. While sophisticated streamline or
finite difference simulators have been increasingly used to determine Sorw
distribution from the tracer production data, the simple semiquantitative model
still has its merits in providing a direct, unambiguous estimate of average
Sorw along the tracer flow path. This paper broadens the scope of the
original Brigham’s model by incorporating partitioning tracers into the model
using a chromatographic transformation technique. By matching the partitioning
and nonpartitioning tracer curves, Sorw can be determined by layers. The
extended Brigham model was applied to the Ranger oilfield multiple tracer test,
and the residual oil saturation determined compared favorably with those
obtained by chromatographic transformation method and numerical simulation.
Introduction
Interwell tracer testing has been recognized as a reliable method for
determining residual oil saturation between wells. The method 2--5 involves the
injection of a slug of partitioning and nonpartitioning tracers at an injector
and production of the tracers from nearby producers. Partitioning between
phases slows down the partitioning tracers by a delay factor of 1+ GREEKbeta in
a phenomenon known as chromatographic retardation, from which the residual oil
saturation can be determined. Oil distribution between wells is derived by
matching the tracer production profiles using a 3D finite difference simulator
such as UTCHEM 6,7 or by a streamline model. 8,9 To circumvent the technical
problems encountered in simulation, an analytical chromatographic
transformation method was proposed by Tang 2--4 and Wood et al. 5 and a
moment-analysis method 10--13 to calculate S or.directly by comparing the
relative separation of tracers. Chromatographic transformation was employed by
Tang 2--4 and Wood et al. 5 to determine S orw for the Golden Spike, Judy
Creek, and Leduc interwell partitioning tracer tests. The moment analysis
method was applied mainly to laboratory tests or small-scale NAPL tests where
the complete profiles could be generated within days. In a reservoir-scale
tracer test, the tracer monitoring program is usually stopped for cost-saving
purposes once the peak is observed; therefore, complete production is seldom
attained. Extrapolation of tracer production curve to completion will introduce
a large uncertainty in moment calculation, 14 and therefore in S orw
determination. The semiempirical Brigham's homogeneous five-spot model, 15--17
which is a common tool in the industry for tracer design 18,19 and data
interpretation for its simplicity, lies between these two extremes. However, it
should be noted that in spite of its frequent use, the applicability of the
model to patterns with significant deviation from the aforementioned
assumptions has not been rigorously proved. The original Brigham model can only
handle nonpartitioning tracer, and its functionality is expanded in this paper
to include partitioning tracer for S orw determination. The extended Brigham
model was used to interpret the Ranger field tracer test data, and the results
were compared with those obtained by chromatographic transformation and other
numerical means.
© 2005. Society of Petroleum Engineers
View full textPDF
(
258 KB
)
History
- Original manuscript received:
16 March 2004
- Revised manuscript received:
14 February 2005
- Manuscript approved:
28 February 2005
- Version of record:
15 June 2005
View Cart
