SPE Member

Abstract

Recently completed field experiments using an instrumented string and test well have verified that the presence of annulus water in injection well annuli introduces a real and economically important heat loss mechanism—wellbore refluxing—for surface-generated steam. Measured boiling heat transfer rates at local hot spots are sufficient to maintain steady refluxing over a realistic range of injection conditions and overburden properties. It is shown that with the wellbore wet, the thermal properties of the cement and overburden, not those properties of the cement and overburden, not those of the insulated tubulars, control the magnitude of the wellbore heat loss.

Steady refluxing without venting at the wellhead has been observed; hence, oil field operators should not automatically assume that a nonventing annulus is dry. Insulated couplings have been demonstrated to reduce coupling heat loss by up to 35% in a dry well, and in wet wellbores to prevent significant boiling at the couplings, typically the primary source of refluxing.

Heat loss estimates for refluxing or flooded wellbores indicate large incremental economic penalties, of the order of $100 million annually in penalties, of the order of $100 million annually in the U.S. alone. The successful demonstration of simple, easily implemented, insulated coupling designs, and the confirmation that wellbore heat loss in refluxing annuli scales with annulus pressure in a predictable manner, suggests that much pressure in a predictable manner, suggests that much of this loss is preventable.

Introduction

It has been widely believed that the efficiency of steam drive for enhanced oil recovery could be substantially improved with the use of insulated tubing. To our knowledge, the first data indicating that the benefits of insulated tubing might be minimal if water is present in the well annulus were 1 and 2. Those papers summarized the results of a series of insulated tubular field tests in the Aberfeldy field of western Saskatchewan. The primary purposes of those tests, a joint program involving Sandia National oil Company, Ltd., were to evaluate commercially available insulated tubulars for their field performance and reliability, and to provide field data for use in the development of advanced wellbore heat loss computer codes. In addition to meeting those objectives, it was unexpectedly found that for all tests, the casing temperature stabilized within a day or two at the local boiling point, opposite the lowest and highest quality insulated tubulars alike. The overall wellbore heat loss using 4.500-in. OD, double-wall, insulated tubulars was only 30%-40% less than for 2.375-in. OD bare tubing, and was 3-6 times higher than predictions based on the range of tubing insulation properties. It was concluded that if the wellbore is wet, there is no advantage to using insulated tubing whose conduction heat loss is any less than that dictated by the cement and overburden thermal properties, with the casing fixed at the local boiling point.

These observations led us to propose the existence of wellbore refluxing in wet wellbores. Further, it was suggested that the presence of wellbore refluxing in injection well annuli was probably a typical situation, as a result of probably a typical situation, as a result of accumulated water due to even minor steam leaks downhole, or to water being in the well when the packer is set. Several techniques were suggested packer is set. Several techniques were suggested for preventing refluxing, or lessening its effects. These included drying out the annulus prior to setting the packer and operating the wellbore at reduced pressure to lower the steam saturation temperature. The development of insulated couplings was also recommended.

The statements summarized above created a controversy when presented. One criticism was that the tests were conducted on only a limited portion (the upper 300 feet) of a single, 1700-ft well, with inadequate control of the conditions within the annulus.

P. 631

This content is only available via PDF.
You can access this article if you purchase or spend a download.