Summary
A new generation of sampling technology is introduced that allows a wireline
formation tester (WFT) to sample reservoir fluids in open hole with levels of
filtrate contamination that are, in many cases, below measurable limits. Also,
the time required on station to clean up before sampling is significantly
reduced in comparison to conventional sampling methods.
Formation-fluid sampling has always been adversely affected by mud-filtrate
contamination, which introduces errors into the laboratory measurements of
fluid properties and requires analytical methods to back-calculate the measured
properties to approximate the uncontaminated reservoir fluid. The ability to
secure a totally clean sample of formation fluid at reservoir conditions is a
significant advance that provides accurate fluid information for
characterization of the reservoir, flow assurance, facility design, production
strategies, and defining reserves.
The application of this new focused sampling technology is presented in four
case studies from wells drilled on the Norwegian continental shelf. A wide
range of formation fluids and permeabilities are examined, in both oil-based
and water-based drilling fluids. Results from focused sampling are compared
directly with conventional sampling in the same reservoir zones. This study
also gives insight into the cleanup dynamics of invaded filtrate and explores
the different factors that affect performance of the focused sampling
technique.
An important consequence of achieving negligible contamination is the
ability to accurately measure fluid properties in-situ. Reduced cleanup time
allows for efficient reservoir fluid profiling, whereby multiple zones can be
scanned sequentially in real time to quantify the fluid properties at a much
higher resolution than traditional sampling methods. Downhole fluid analysis
(DFA) can thus provide an additional source of information in the process of
revealing complex reservoir architectures.
Introduction
An accurate description of reservoir fluid properties is critical in all
stages in the life of an oil or gas field. It is required in exploration to
ascertain the true nature of a discovery and to assist in defining reserves to
value the economic potential. In appraisal phase, it is used to determine layer
connectivity and field structure as well as the optimization of well completion
and production tests. For development of the field, fluid composition is
crucial for material selection of well completion and surface flowlines, flow
assurance, design of process control, and production facilities. Later during
the exploitation of the reserves, it is necessary to understand fluid behavior
during the production and life of the asset.
Reservoir engineering and production strategies are crucially dependent on
knowledge of phase behavior and multiphase fluid flow, and they rely
increasingly on numerical simulators tuned to pressure/volume/temperature (PVT)
laboratory measurements. The presence of compositional gradients because of
fluid migrations or fluids showing near-critical behavior at reservoir
temperature must be understood to develop a valid model of the reservoir
(Fujisawa et al. 2008).
Understanding the nature and composition of formation water is also critical
to the economics of field development. Chemical analysis of formation or
connate water determines the scaling and corrosion potential of produced fluid
required for the design of completion and processing facilities (Raghuraman et
al. 2007). It also establishes the salinity for petrophysical evaluation and
fingerprints the aquifer for studies on basin hydrology. Water composition is
important for production strategies involving inhibitor injection, wellstream
mixing, process sharing, and enhanced-oil-recovery (EOR) injection.
Fluid sampling operations are continuously under pressure from cost control,
operational limitations, and, sometimes, the lack of understanding of their
true value in downstream processes. The risk of financial loss attached to poor
fluid characterization, although difficult to quantify, can be enormous. This
risk is magnified in deepwater projects, where the development can be extremely
expensive and decisions on facility design must be made early in the
project.
The type of reservoir fluid to be sampled has a considerable influence on
the challenges encountered when sampling with WFTs. Near-critical systems are
notorious for phase changes, which yield large proportions of both liquid and
gas with only small reductions in pressure below the saturation curve. Because
these proportions can exceed critical saturations, both fluids will be mobile
to some extent, and the resulting production and samples may be difficult to
interpret. If variations in sample properties are caused by contamination,
these phenomena may not be identified until several wells have been
appraised.
At the other extreme of the fluid range, hydrocarbons with very low gas/oil
ratio (GOR) may cause separation and measurement problems or prevent proper
collection of samples. Viscous oils can lead to high drawdown and plugging of
sample lines, which require specialized sampling techniques and equipment.
Emulsions may be formed, which make it difficult to collect a representative
sample. Solids production, such as asphaltene deposition or wax formation, also
affect fluid analysis with consequences on flow assurance in completions,
pipelines, and surface facilities.
Water in hydrocarbon samples has traditionally been regarded as a
contaminant, yet most hydrocarbon fluids naturally contain small quantities of
connate water. Water concentrations can rise above 5% in
high-pressure/high-temperature reservoir fluids. Understanding the effect of
this water on phase behavior and fluid production becomes significant, but the
issue is compounded by poor knowledge of actual concentrations in most
reservoir fluids.
The presence of miscible contamination from the invaded filtrate of drilling
fluids represents the largest obstacle to obtaining valid reservoir fluid
samples. This occurs when sampling hydrocarbons in oil-based mud (OBM) and
synthetic-oil-based mud (SOBM), or when sampling formation water in water-based
mud (WBM). Miscible filtrate can severely affect the PVT fluid characteristics
measured in the laboratory, and it results in inaccurate data for field
development and reservoir modeling (Gozalpour et al. 1999). Mathematical
techniques exist to back-calculate the uncontaminated fluid properties;
however, these methods introduce additional uncertainties even when
contamination is relatively low.
Fluid samples from WFT have been adversely affected by mud-filtrate
contamination since their introduction. The first generation sampling tools
consisted simply of a sample chamber connected by a flowline to the probe,
which could be opened to collect a limited volume of near-wellbore fluid.
Extremely high contaminations were common, and laboratory PVT analysis was not
a viable option. This situation improved dramatically in the 1990s with the
introduction of second-generation sampling tools, whereby a downhole pump was
integrated into the flowline. This enabled fluid to be pumped from the
formation to the wellbore during the “cleanup” while measuring contamination in
real time with optical fluid analyzers to determine when flow should be
diverted to the sample bottle. The configuration of modules in this tool string
evolved over time to enable better control of filling sample chambers; however,
the basic limitation regarding the interface between the tool and the formation
still existed.
The rate of contamination decrease slows during the pump-out operation, and
the level of contamination in the incoming fluid will not reach zero in any
practical length of time (Mullins et al. 2001). The inability to reach zero
contamination by invaded mud filtrate continually feeding the sampling zone
from around the probe interface, and, in some cases, is caused by reinvasion of
the sampling zone through the surrounding mudcake (see the section titled
“’Discussion”). To achieve very low contamination levels, the WFT must usually
engage in pump-out operations for extended periods of time, which can be
expensive and risky in an openhole environment. In many situations, the only
method to achieve uncontaminated samples has been through the large-volume
cleanup associated with expensive drillstem tests (DSTs), and experience has
shown that even these are not guaranteed to provide contamination-free fluid
samples.
© 2008. Society of Petroleum Engineers
View full textPDF
(
4,936 KB
)
History
- Original manuscript received:
20 June 2006
- Meeting paper published:
11 September 2006
- Revised manuscript received:
20 August 2007
- Manuscript approved:
18 September 2007
- Version of record:
25 April 2008
View Cart
