Enhanced recovery

Solvent-Enhanced Steamdrive: Experiences From the First Field Pilot

The addition of a hydrocarbon condensate to steam operations in heavy-oil and bitumen reservoirs has emerged as a potential technology to improve not only oil recovery but also energy efficiency.

Getty Images

In recent years, the addition of a hydrocarbon condensate to steam operations in heavy-oil and bitumen reservoirs has emerged as potential technology to improve not only oil recovery but also energy efficiency. The idea of solvent addition to a steamdrive process has been extended and applied for the first time in the Peace River area in Canada. There, evidence was obtained of oil uplift in the patterns where solvent was injected. However, piloting this new technology in a brownfield had many challenges, especially when evaluating its main economic factors: production increase and solvent recovery.


Vertical-well steamdrive (VSD) is the ­selected process to recover bitumen from the Peace River Bluesky formation. Solvent coinjection has been identified as an economical method to improve the efficiency of this process. In an early phase of the steamdrive, a slug of hydrocarbon condensate (diluent) is coinjected with the steam. The solvent condenses at the cold steam/bitumen interface to form a solvent bank. This bank has the potential to accelerate bitumen production by viscosity reduction and to improve ultimate recovery.

The efficiency of the diluent coinjection in a steamdrive process is expected to be lower than that of liquid addition to steam for enhanced recovery; however, the solvent recovery factors are expected to be much higher. The solvent recovery, therefore, is a key factor in the economic viability of the process. The main objectives of the pilot were to obtain a positive response in bitumen production and accurate quantification of the diluent recovery. An accurate assessment of the bitumen-production increase was not expected because of the small size of the pilot and lack of control patterns; hence, the injection slug size and concentration were designed to obtain a significant and measurable bitumen response.

Pilot Design

The Peace River lease in Alberta, ­Canada, has been subject to many well- and recovery-technology trials in the last 30 years. One of the technologies tried is cyclic steam stimulation (CSS) with multilateral horizontal wells. Pad 19 has been developed with so-called “soak radial wells”—four horizontal laterals in a cross pattern. Over a 9-year time frame, bitumen has been produced in seven to eight CSS cycles. With a recovery of less than 20% from the initial design, a part of the pad has been converted to a pattern steamdrive to increase recovery to more than 50%. Vertical injectors and producers are drilled to complete the 5-acre inverted-five-spot patterns. Vertical producers are perforated over almost the complete interval, while steam injectors have been completed with five limited-entry perforations (LEPs) to distribute the steam evenly over the entire reservoir interval.

On the basis of the field-­development plan, the newly drilled vertical infill producers were subjected to two CSS cycles to create communication between injectors and producers. Infill injectors were not subjected to cyclic steam. After that, the pad was intended to be switched to a VSD with continuous steam injection of 100 m3/d (cold-water equivalent). The rate was to be tapered down to 50 m3/d over the life span of the steamflood.

The solvent-injection strategy that determined in which injectors solvent would be coinjected, for which ­period of time, and at what concentration was designed in a two-stage process. In the first stage, a simplified element-of-­symmetry model was used to screen a wide range of options with respect to solvent concentration and start and duration of solvent coinjection. It was found that, on the basis of the development scheme, bitumen rates in VSD would be low for a few months before mobilized bitumen would reach the producer, at which point the rates would show a strong increase followed by a gradual decline. Solvent addition increases bitumen mobilization and leads to a higher desaturation of the steam chamber because of the formation of a solvent bank. This leads to a more pronounced increase in bitumen rates once the mobilized bitumen reaches the producer.

The effect of different solvent concentrations on the early bitumen rates and, hence, on the bitumen-uplift signal is shown in Fig. 1. After this period, injection conformance in the LEP injectors would have stabilized so solvent would be injected over the full height of the pay zone.

Fig. 1—Net bitumen rates in VSD as a function of solvent concentration predicted by the element-of-symmetry model.


In Stage 2, this solvent-injection strategy was tested in a full-field model that had been history matched to the historical CSS cycles on a well-by-well basis. In addition, the temperature data gathered from logging the infill wells had been used to scale the effectiveness of each leg of the multilateral wells.

Pilot Operations: Steam and Solvent Injection

After the second CSS cycle in the vertical infill producers, the downhole pumps were installed and the steamdrive phase of the project started in June 2014. Injectors 3, 5, and 9 could be operated at the targeted injection rate of 100 t/d, but Injector 7 had a maximum injection rate of approximately 75 t/d at a tubinghead pressure (THP) of 12.5 MPa. Because of the high pressure drop over the LEPs, the injectivity could not be increased by fracturing or dilating the reservoir and it was decided to operate this injector at the maximum THP of 12.5 MPa. The continuous steam injection started 7 June 2014, and, after approximately 8 weeks of steam injection, 15 wt% (cold) solvent was injected directly into the steam at the wellhead in Injectors 7 and 9. The injection was continued for 4 months without major problems, achieving a total of steam and solvent injected of 19,600 t of steam and 3,400 t of diluent. As supported by pressure/volume/temperature calculations, the addition of the solvent did not lead to a reduction of the steam-injection capacity in the LEP-constrained wells.

Bitumen Production

One of the objectives of the pilot is to demonstrate a significant bitumen-­production increase as a response to solvent injection. The solvent-injection slug was designed to give a significant bitumen-production increase in wells surrounding the solvent injection. This was clearly observed in the multilateral Well P11, which has a horizontal leg that is close to solvent Injector I9. After 1 month of injection, the production rate of this well more than doubled compared with the baseline production of the steamdrive. After this initial peak, which coincided with solvent breakthrough, the rates declined but were sustained at a higher plateau for approximately 6 months.

Solvent Recovery

Much attention was given to the accuracy of the diluent recovery in this pilot because it is a key economic factor in the solvent-injection process. The fact that more than 75% of the recovered diluent was produced through the casing-vent-gas system helped because the measurement errors of the gas stream (e.g., rate and composition) are much smaller than those of the emulsion stream. The main uncertainty comes from the calculation method to allocate the hydrocarbons to bitumen and diluent.

On the basis of the recovery so far and the current rates, the project is expected to achieve or even exceed the predicted recovery factor of 86% after 2 years. At that mark, the incremental oil/lost-solvent ratio will be 5.2, which is economically favorable and compares well with other solvent-coinjection processes.


  • Conducting a recovery-technology pilot in a small brownfield infill development is challenging. It requires a robust design for expected signal and high-frequency and redundant data acquisition to obtain quantifiable results.
  • Well testing and water-cut metering are a large source of error and need to be checked thoroughly and validated with independent measurements.
  • New allocation algorithms were developed and validated to be able to allocate solvent and bitumen accurately in a steamdrive process.
  • Bitumen uplift could be observed positively in several wells; the oil/steam ratio in the solvent patterns on average was 0.1 higher than in the patterns without solvent during the first 10 months of the pilot.
  • Solvent recovery is faster than expected. More than 50% of the solvent had been recovered within 4 months of stopping solvent injection.

This article, written by Special Publications Editor Adam Wilson, contains highlights of paper SPE 175414, “Solvent-Enhanced Steamdrive: Experiences From the First Field Pilot in Canada,” by M.L. Verlaan, SPE, and ;R. Hedden, SPE, Shell Canada; and O. Castellanos Díaz, V. Lastovka, SPE, and C.A. Giraldo Sierra, Shell Chemicals Americas, prepared for the 2015 SPE Kuwait Oil and Gas Show, Mishref, Kuwait, 11–14 October. The paper has not been peer reviewed.