Horizontal Steam Injectors in the Kern River Field

Fig. 1—Scale deposition on the low side of the 3½-in. tubing in the heel region of Injector A.

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

A horizontal-steam-injection pilot project has been under way for the last 4 years in the Kern River heavy-oil field in the southern San Joaquin Valley of California. This project was designed to address a series of learning objectives for horizontal steam injection in a mobile heavy-oil reservoir. These objectives included mechanical integrity and operability, steam conformance control, steam-flow profiling, and reservoir response and long-term operability.


While documented cases of continuous and cyclic horizontal steam injection in mobile heavy oil exist, these have not led to large-scale economic development of thermal heavy oil. Most thermal, mobile, heavy oil is currently produced with vertical steam injectors. The Kern River Field is a good example of this recovery technique.

The primary heavy-oil reservoir in the field is 300 to 1,300 ft deep, with a reservoir thickness of up to 900 ft. During the last 50 years, most of the field has been developed with continuous vertical steam injectors, with typical pattern spacing being 2.5 acres.

The pilot area is approximately 12 acres and is on the northwest flank of the field. The pilot injected steam into sand that has an average thickness of 22 ft across the pilot area. The pilot’s original configuration included two horizontal injectors, nine vertical producers, and 12 temperature-observation wells (TOWs). Of the 23 wells in the pilot area, five were existing wells and the remainder were new drills. Two types of TOW were used in the pilot: (1) injector TOWs, which have a short offset of less than 40 ft from the horizontal injectors and are intended to assist in understanding injector steam conformance, and (2) reservoir TOWs, which are more than 50 ft from injectors and producers and are intended to help understand far-field reservoir response.

The pilot used horizontal continuous steam injection but is different from a bitumen steam-assisted-gravity-drainage application in that the heavy oil is mobile and will flow at reservoir conditions. Another important difference is that a horizontal producing well has not been placed 15 to 25 ft under the horizontal injector. In fact, vertical producing wells were used in the pilot area offset 140 ft from the horizontal injectors.  

Mechanical Integrity and Operability

Precautions were taken to ensure that learning objectives were captured and not cut short because of the potential mechanical loss of the injectors. A 7-in. slotted-liner completion was installed in the openhole horizontal section of both injectors. A tubing-deployed flow-control-device (FCD) and cup-packer design philosophy was used for conformance control.  

Influx of sand into the liner, scale deposition, corrosion, and liner damage caused by installation and thermomechanical loads were recognized as threats to this conformance-control system. More than 800 infill horizontal producers have been drilled in the Kern River Field to capture warm oil not being captured effectively by vertical wells. Most of these infill horizontals are drilled with toe-high, heel-low trajectories to gravity drain oil and water to a rod pump landed in the heel region. Both horizontal injectors were designed with the same toe-high, heel-low trajectories for future conversion to production. Some scale deposition and corrosion failures occurred in the heel region of both injectors. Fig. 1 above shows salt scale that formed on the 3½‑in. tubing on the low side of the wellbore in Injector A.  

Conformance Control

A key conformance-control uncertainty was the annular condition between the outside of the slotted liner and the open hole. If the slotted liner/openhole annulus remained an open conduit, then the effectiveness of both tubing and liner-deployed FCDs to manage steam conformance would be compromised. In Injector A, steam was injected for 45 days. The injection assembly was pulled, the well was killed with water, and an acoustic-based isolation scanner and flexural-attenuation log was run with a tractor through the liner lap to the midpoint of the horizontal section. Because an acoustic-based log was used, a liquid-filled borehole was required, and creating a liquid-filled environment to the toe of the well was not expected to be possible.  

Water was pumped slowly into the well while logging, and no evidence of gas in the wellbore was seen. The log identified solids and preferential reservoir consolidation outside the slotted portions of the liner where steam would be expected to exit the liner into the reservoir. Fewer solids and less consolidation of the open hole were seen around the liner in the blank portions. The 7×9⅝‑in. liner lap was largely liquid-filled, as would be expected. The openhole consolidation around the 7-in. liner was clearly less in the 12-in. openhole section relative to the 8½-in. openhole section. Preferential high-side and low-side reservoir consolidation was clearly evident on the basis of the relative bearing. The conclusion from this log interpretation is that steam injection assists in creating reservoir consolidation around the slotted liner and open-conduit annular flow paths do not exist outside the slotted liner.  

The authors conclude that tubing-deployed FCDs alone were not effective in redirecting steam or providing conformance control in the horizontal section. The use of cup packers in combination with FCDs was effective in redirecting steam and improving conformance.

Steam-Flow Profiling

A multiphysics approach with Euclidian convergence analysis was used to develop fiber-optic steam-flow profile interpretations with good statistical confidence in both injectors. Fiber-optic flow profiling and analysis confirmed the following:

  • Countercurrent steam flow in the tubing-liner annulus occurred and steam flowed back preferentially to the heel of Injector B when cup packers were not installed.
  • All steam injection in Injector A was moving toward the toe.
  • Liquid was pooling in the heel region for some completions that had packer leaks.
  • Cup packers installed for conformance control can hydraulically isolate portions of the wellbore effectively, preventing steam injection into selected intervals of the horizontal section.
  • High heat loss can occur in the heel region when liquid is pooled.
  • Well-calibrated distributed-temperature-sensing data can be used to calculate steam pressure from the steam tables.

Reservoir Performance

A reservoir-simulation model was constructed to assist in the pilot interpretation of well and reservoir heating patterns and to gain experience in predicting oil-production response by use of simulation models. A reservoir simulation work flow was used that included a multisegmented-well (MSW) feature, which couples a wellbore thermohydraulic model with the thermal reservoir-simulation solution. The MSW model calculates friction in the flowing segments (e.g., tubing, annulus, or liner), heat transfer between the wellbore and formation, and fluid properties in the wellbore. The geologic model was constructed with geostatistical tools to populate fluvial facies, saturations, and permeability.  

In the pilot area, oil-saturation variations in high-permeability channels were observed. Undeveloped areas of the field have been pressure-depleted because of production; water influx has occurred; and gas has evolved from the depressured oil, causing a distribution of small gas-saturation pockets. These gas-­saturation pockets are not in long-term equilibrium and are trapped under high-viscosity heavy oil or tar mats.  

The TOWs and model both exhibited heel-biased heating in the first few months, followed by Injector A moving to uniform heating at approximately 6 months, then transitioning to toe-biased heating. Injector B had strong heel heating. This heel-heating bias was moderated when cup packers were introduced in combination with the FCDs.

Reservoir areal heating was affected by gas saturations present at the start of steam injection, dip, injector design, and producing-well voidage. Reservoir mobility was seen to be the strongest effect governing initial heating patterns and was evident in the reservoir model, fiber-optic flow profiling, and TOW data. This observation can be generalized to conclude that in-situ fluid mobility, fluid pressure gradients in the wellbore and reservoir, and permeability heterogeneity affected steam conformance strongly.  

The steam-injection-flow profile along the lateral was also affected by the toe-high, heel-low injector architecture. While relatively uniform heating was achieved along the entire length of Injector A, steam conformance was biased toward the toe after a few months of injection.   

The center of the pilot area remained cold at 3 years of steam injection. Lack of heating was influenced by lack of fluid drainage from the vertical producers in this area. Limited drainage in the center of the pilot area caused by impaired slotted-liner completions resulted in heat and fluid ultimately migrating outside the pilot area.   

The MSW model predicted that steam would exit the liner where the highest transmissibility and differential pressures exist in the horizontal section of the wellbore. For the subject injectors, without cup packers in the annulus, the only force available to direct steam is the annular friction pressure in the 3½-in. tubing by 7-in. liner annulus. Fiber-optic flow profiling and well-work activities confirmed that wellbore thermohydraulics was further complicated by liquid collecting under gravity forces at the low point in the heel of the injectors.

The MSW simulation model was updated periodically during the first 4 years of the pilot life. The model was very useful in understanding heating patterns and thermal maturity, was used to support recompletion decisions, and provided physics-based insights. The actual production response from horizontal steam injection in the pilot area was within the range of outcomes predicted from the simulation model.

This article, written by JPT Technology Editor Chris Carpenter, contains highlights of paper SPE 187258, “Design and Operational Experience With Horizontal Steam Injectors in the Kern River Field, California, USA,” by R.S. Buell, SPE, R. Gurton, J. Sims, SPE, M. Wells, G.P. Adnyana, M. Shirdel, SPE, C. Muharam, T. Gorham, SPE, E. Riege, and G.B. Dulac, Chevron, prepared for the 2017 SPE Annual Technical Conference and Exhibition, San Antonio, Texas, USA, 9–11 October. The paper has not been peer reviewed.

Horizontal Steam Injectors in the Kern River Field

01 June 2018

Volume: 70 | Issue: 6


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.