High temperature steaming of wells can cause plastic yielding of the production casing. However, when properly designed and installed these casing strings will last the full life of a thermal project. Poor designs or installations an cause significant well problems ranging from wellhead growth and settling to buckled or parted casing.
Since 1979, Shell Canada Limited has been operating a thermal project in the Peace River/Cadotte Lake area. Steam injection temperatures have been as high as 335 < C. Only three of the 250 steamed wells have required significant remedial action. Some ten to fifteen other wells have experienced wellhead settling (casing contraction). Several wellheads have gone down 25 to 50 cm, requiring wellhead and piping modifications.
This paper discusses the casing and cement designs used historically, field observations and explanations for the casing "shrinking" phenomena and the field testing of K-55 grade casing. Shell Canada's current design utilizes K55 production casing and slag cement. This design was recently utilized on four horizontal wells.
Shell Canada Limited has operated a steam injection project at Peace River since 1979. The production zone is an unconsolidated sandstone of the Lower Cretaceous Bluesky and Gething formations. This bitumen reservoir is 550 m from surface with an average pay thickness of 25 m. The bitumen gravity is 8 - 100 API with a iscosity at reservoir conditions of 200 Pa.s. The bitumen saturation decreases with depth in the zone. At bitumen saturations below 65%, the water phase is mobile and allows steam injection below fracture pressures. Most, but not all the wells drilled have a mobile bottom water zone.
The recovery process to date has been based on a pattern steam drive. Dedicated injection wells were utilized, however all wells were initially steamed and produced to establish bitumen mobility and communication between wells. In 1993 four horizontal wells were drilled to evaluate the Enhanced Steam Assisted Gravity Drainage Process (ESAGD). Steaming operations al the horizontal wells commenced November 1993.
Casing designs have varied from NM-80 LT&C, L-80 YAM, L-80 OBTC (oversize buttress threaded and coupled), and most recently to K-55 NS-CC premium connection. Between 1979 and 1993 there were three casing failures at the Peace River Complex (PRC). One was due to buckling of the casing, another due to a parted connection and the third due to C02 corrosion. Ten to fifteen wells have had the casing "contract", causing the wellhead to settle 10 to 50 cm. This does not cause a failure but can impose additional operating expenses and difficulty. In general, the casing performance has been satisfactory. The above failures can be explained and continued reliability from the existing casing strings is anticipated.
An understanding of the stresses the casing goes through during thermal operations is needed to properly design the casing and evaluate its performance. Historically casing strings were designed to avoid tensile yielding. Parted connections were the common mode of failure. In the late 60's and 70's most thermal well casing designs were based on Wellhite and Dietrich model1 and subsequent paper by Holliday2.