Effects of Completion Design on Thermal Efficiency in SAGD

You have access to this full article to experience the outstanding content available to SPE members and JPT subscribers.

To ensure continued access to JPT's content, please Sign In, JOIN SPE, or Subscribe to JPT

During the circulation (startup) phase of steam-assisted gravity drainage (SAGD), high-quality steam injected through the injector and producer wells heats the reservoir between the wells. The viscosity is thus lowered, making fluids mobile at approximately 50 to 100°C and creating interwell fluid communication. This paper uses a simulation model to evaluate and compare the thermal efficiency of five different completion design cases during the SAGD circulation phase in the Lloydminster formation in the Lindbergh area in Alberta, Canada. The results show that completion-design configuration affects the heat transfer and thermal efficiency of the circulation process.


The SAGD process is the most commonly used thermal method of in-situ recovery for extracting heavy oil and bitumen resources in Alberta and can yield recovery factors of greater than 60%. The technique requires two parallel horizontal wells, a producer and an injector, known as a well pair. The horizontal producer well is placed approximately 3 m above the oil/water contact (OWC) or above the bottom of the reservoir, and the injector well is placed above the horizontal producer well. Most SAGD well pairs are placed 5 m apart, which corresponds to approximately 50 kPa of hydrostatic head from the injector to the producer. The lateral section of the well pair is approximately 700–1200 m in length. The SAGD thermal method usually consists of two phases, the circulation phase and the full SAGD, or production, phase.

After the circulation phase, the well pair is converted to the production phase, during which steam is injected through both tubing strings of the injector well in a dual-completion design while bitumen or heavy oil and condensed steam are produced through the producer well using natural or mechanical lifting. Constant steam injection causes the steam chamber to grow and expand in the reservoir.

Variations in thermal efficiency during the circulation phase result from factors such as tubing size, well trajectory, well length, completion configuration, reservoir properties, and operating parameters. To the authors’ knowledge, no published study has determined the thermal efficiency of different completion designs during the circulation phase of a SAGD well pair in a Lloydminster formation. A previous study used a discretized thermal reservoir-wellbore modeling simulator to history match field data obtained from a SAGD well pair in the Lloydminster area.

Operating Strategies and Reservoir Model

The complete paper discusses SAGD operating strategies and development of well and completion designs in pilot and commercial operations.

The thermal reservoir simulator used to simulate the wellbore hydraulics of five different completion designs allows the modeling of different tubing strings, casing, and slotted liners within one wellbore and the modeling temperature and pressure changes along the length of the well. A 3D, multicomponent, fully implicit model was created using this thermal simulator, in which the wellbore grid cells are fully coupled with the reservoir grid cells, allowing the calculation of temperature, pressure, and saturation separate from the reservoir grid cells. The relative flow rates within the well are solved using the Beggs and Brill multiphase correlation, which takes into account liquid holdup, frictional pressure losses, and flow patterns in deviated wells. The reservoir simulator calculates the transmissibility between the casing-annulus wellbore cells and the reservoir cells. The model on which this paper is based was created with a rectangular Cartesian grid, which included the vertical, build, and reservoir sections of an existing well in the Lindbergh area.

Reservoir geology and rock-­fluid properties were obtained using core data from delineation wells drilled in the Lindbergh development area lease. Because the reservoir properties are consistent through the Lloydminster sand, a homogeneous model was built.

The directional surveys from an actual horizontal well were imported into the model. Both the injector and producer well were represented by a discretized wellbore, each containing different completion tubing strings based on each completion-design modeling case. Details of the well-constraint configurations used in all the cases are outlined in the complete paper.

Completion-Design Cases

Case 1—Dual-String, Bare Tubing. This case was simulated using a dual tubing-string configuration consisting of a long tubing string to the toe of the well and a short tubing string to the heel. This completion design has been used in all of the featured SAGD well pairs during the circulation phase. The circulating phase was initiated by injecting steam through both injector and producer long tubing strings at high pressure and temperature. Steam flowed to the end of the long tubing string and back through the annulus toward the heel, and back to the surface through the short tubing string.

Cases 2 and 3—Vacuum Insulated Tubing (VIT). VIT technologies can minimize heat loss that runs radially from the inner tubing to the wellbore annular space. VIT has been used in projects during the SAGD circulation phase. VIT consists of two lengths of tubing aligned concentrically and welded at each end. The annular space between the inner and outer tubing is either filled with inert gas or sealed with low to medium vacuum and also contains aluminum foil wraps separated by a scrim to minimize heat transmission by radiation. Most VIT designs try to preserve a high ­vacuum by using a getter to remove hydrogen and other gases. The connection of the VIT is not thermally insulated, and 50–90% of heat loss is through the connection area and around welds. However, some customized designs include an additional covering that isolates the thread connection to reduce the heat loss on the coupling.

Case 2 was simulated using a dual ­tubing-string configuration in both the injector and producer wells, similar to Case 1. However, the long tubing string was run using VIT from the surface to the heel and bare tubing from the heel to the toe to reduce wellbore heat loss to the surroundings and to maintain the injected steam at high steam temperature and quality. Circulation also was similar to that of Case 1.

Case 3 used a single tapered tubing-string configuration in both the injector and producer wells, consisting of only one long tubing string to the toe of the well. The string was run using VIT from the surface to the heel and bare tubing from the heel to the toe. Circulation was similar to that seen in Cases 1 and 2.

Cases 4 and 5—Gas Blanket. When not using VIT in the completion design, the casing-tubing annular space fills with dry air, natural gas, or nitrogen. In Canada, most wells completed with dual concentric tubing inject methane or nitrogen in the annulus between the tieback and the long tubing string, minimizing heat loss around the vertical and build section of the wellbore from the steam injected to the fluid returning through the intermediate casing annulus.

Case 4 used concentric tubing in both the injector and producer wells consisting of a 177.8-mm tieback string to the heel and a 114.3-mm tubing string inside the tieback to the toe. Case 5 used concentric tubing consisting of a 139.7‑mm tieback string to the heel and an 88.9‑mm tubing string inside the tieback to the toe.

Some sensitivity analysis has shown that the steam-injection rate is the ­second-most-influential factor in SAGD thermal efficiency after the completion design. The steam-injection rate for all five cases for this study was set at 103.0 m3/d for the entire circulation phase. On the basis of each completion-design case, the injection-steam temperature and pressure changed during the circulation period set to 115.0 days. This change resulted mainly from differences in the inner diameters and the low-heat-conductivity material used in the vertical and build sections of the wells.

All completion-design cases were simulated injecting steam at a quality of 98%. Also, it was assumed there was a static fluid with water saturation of 1 in each lateral wellbore. The paper contains detailed discussions of modeling-case analyses and results.


  • The best completion designs to deliver steam to the toe of the well in a short period of time were Cases 2, 3, 4, and 5. These completion designs consist of VIT in the long tubing string or methane in the annulus space between the tieback and the long tubing in the vertical and build section of the well.
  • Of the two cases using methane as an alternative insulated material, Case 4 delivered much more energy and higher steam quality to the lateral section than Case 5.
  • Case 4 showed the best delivery of high-quality steam to the toe of the well during the entire circulation phase, with 86% of the initial steam-injection quality.
  • Case 4 also delivered more heat energy inside the long tubing from the heel to the toe in both early and late circulation, with 264.0 GJ/D.
  • In summary, the larger the inner diameter, the higher the energy and steam quality delivered to the lateral section of the well and the less energy delivered from the return fluid to the vertical and build section.

For a limited time, the complete paper SPE 193357 is free to SPE members.

This article, written by JPT Technology Editor Judy Feder, contains highlights of paper SPE 193357, “SAGD Circulation Phase: Thermal Efficiency Evaluation of Five Wellbore Completion Designs in Lloydminster Reservoir,” by Daniel Ayala Rivas, SPE, and Ian Gates, SPE, University of Calgary, prepared for the 2018 SPE Thermal Well Integrity and Design Symposium, Banff, Alberta, Canada, 27–29 November. The paper has not been peer reviewed.

Effects of Completion Design on Thermal Efficiency in SAGD

01 April 2019

Volume: 71 | Issue: 4



Don't miss out on the latest technology delivered to your email weekly.  Sign up for the JPT newsletter.  If you are not logged in, you will receive a confirmation email that you will need to click on to confirm you want to receive the newsletter.