Summary
This paper presents a framework for a strain-based design of tubular strings
in extreme-temperature or high pressure/ high temperature (HP/HT) wells. The
relevant concepts are illustrated by examples from analytical and experimental
investigation of a casing material considered for use in thermally stimulated
wells operated by Shell Canada Limited (Shell) in western Canada. Much of this
framework is also relevant to other applications where deformation-driven
loading mechanisms are present.
Strain-based design uses material capacity beyond its elastic range to
overcome a number of economic and technical hurdles encountered in conventional
load-based designs. It has been used successfully in field applications where
plastic deformations occur (e.g., thermal wells and pipelines). However,
current industry standards for material selection have their origins in
load-based design. More sophisticated material-characterization tools are
required for strain-based designs, in which post-yield material properties
govern much of the system response.
This paper describes the application of strain-based concepts to the design
of casing strings under combined loading where some load components are
deformation controlled. The paper emphasizes the need to address strain
localization, high-strain cyclic plastic loading, strain-rate-dependent
strength, and associated stress-relaxation effects.
Strain-based design is most effective if relevant and reliable post-yield
material properties are available. Experimental investigation of a candidate
material considered for Shell Canada’s thermal wells consisted of a series of
custom-designed coupon-scale tests. The tests were conducted to acquire data
describing the post-yield material response to monotonic and cyclic loading at
temperatures ranging from 20 to 350°C.
Conclusions of this paper summarize findings of the executed
material-evaluation program, outline options to minimize strain-localization
impacts, and provide recommendations for strain-based designs of
well-completion tubulars. Following these recommendations should result in
higher reliability and more cost-effective wells in completion programs using
strain-based strategies for designs of extreme service wells.
Introduction--Strain-Based Design Considerations
Conventional elastic load-based well design is based on limiting all loads
on the casing system to be within the initial yield strength of all components
of the well structure. In contrast, an elastic/plastic strain-based design
assumes that a certain portion of the structure will yield. Furthermore, the
strain-based approach assumes the post-yield deformation to be an independent
loading parameter that will govern the distribution of stresses throughout the
structure. Under combined loading, one or more load components may be
deformation-driven, while other deformation components will depend on applied
loads and instantaneous post-yield material properties (for example, axial load
in a pipe may be deformation-driven, but radial and hoop deformation will
depend also on internal pressure). Designs for such combined-loading scenarios
are sometimes referred to as hybrid designs.
Many thermal-well applications incorporate loads that bring the casing
system into large-scale post-yield deformation, at least in the initial heating
phase of the operation. Where cyclic loading is anticipated, a common design
practice has been to limit the amount of plastic deformation in the load cycles
following the initial heating. Minimizing that plastic deformation would
provide a rationale for employing the load-based design strategy.
The conventional design has great appeal because of an extensive experience
base, specifications, and design codes that are available. For example,
connection specifications are load-based; material properties used for design
are expressed in terms of stress; and stress-corrosion testing procedures
incorporate stress loads that are severe, relative to stress limits. Challenges
presented by thermal-well conditions include stresses that are higher than
those used in standard testing procedures [e.g., sulfide-stress-cracking (SSC)
tests according to the National Association of Corrosion Engineers
specifications]. Such challenges are often met with development of
application-specific test procedures, which are based on the standard
procedures but incorporate updated stresses consistent with those expected in
the application. Nonetheless, the testing is still based on stress
conditions.
When field conditions and detailed designs are considered, it becomes
apparent that elastic conditions are not likely to be maintained everywhere in
the system after the initial heating phase. Furthermore, detailed material
characterization of casing materials under cyclic loading reveals attributes
that are not consistent with the assumptions used in conventional cyclic
plasticity. Connections present regions where stress and strain concentrations
are known to occur, both in the thread roots and across the wall thickness
where primary load transfer occurs in the threads. Differences in pipe geometry
and strength among various joints of the casing string may lead to strain
localizations, causing incremental plastic deformation of the pipe in each
loading cycle. Geomechanical loading also has been observed in many
applications, where significant permanent plastic deformation is imposed on the
casing system. All of these facets point to cyclic plastic deformation that is
not accounted for currently in the design process.
In spite of the observation that most thermal wells endure cyclic
deformations beyond the elastic-design limits, the rate of failure in such
wells is remarkably low. Corrosion failures are infrequent in field
applications, even where laboratory tests simulating those field conditions
demonstrate a significant risk of stress-corrosion failure. Material failure
caused by cyclic plastic hardening in the field is not common, even though
localization is recognized to occur and to produce strains beyond the design
assumptions.
It can be said, therefore, that field experience has already demonstrated
that wells can be operated under conditions that produce cyclic plastic strain
and that current load-based design strategies essentially limit the amount of
such plastic strain to a modest amount and to limited regions of the well.
However, the extent of such strains and the parameters controlling them have
not been studied in sufficient detail to determine how much plastic deformation
can be tolerated and how it can be harnessed. Therefore, a comprehensive
strain-based design method is not yet available because many of the associated
design parameters required to support such a method are not characterized or
controlled. This represents a substantial opportunity for future design
refinement.
Nonetheless, many strain-based design concepts can be applied in modern
well-structure design. These concepts can be used to avoid implementation
pitfalls, optimize material selections, and provide tools for forensic
investigations of the few failures that do occur. Descriptions of current
technology status and possible incorporation of strain-based designs in HP/HT
and thermal-recovery wells can be found in literature (Hahn et al. 2005; Skeels
et al. 2002; Wooley et al. 1977; Maruyama et al. 1990; Schwall et al. 1996;
Slack et al. 2000; Dall’Acqua et al. 2005). This paper presents examples of how
the strain-based design concepts are employed in a challenging thermal recovery
field in western Canada.
© 2008. Society of Petroleum Engineers
View full textPDF
(
1,759 KB
)
History
- Original manuscript received:
14 November 2006
- Meeting paper published:
20 February 2007
- Revised manuscript received:
25 February 2008
- Manuscript approved:
25 February 2008
- Version of record:
10 December 2008
View Cart
