Some pessimists and alarmists preach that the end of oil will come soon and that its results will be severe. But news of oil’s death is greatly exaggerated, and many underestimate the power of thought and ingenuity in proving up new reserves. Advancing technology, expertise, and knowledge are expanding the vitality of aging oilfields, just as medical advances are lengthening human life spans. Oilfield extension success stories are as varied as episodes of “ER” and “House.” As nimble, small companies take a fresh look at the hand-me-down oilfields and land blocks dumped by major companies the world over, bypassed oil is being resuscitated into significant production streams.

I work in one of those hand-me-down oilfields in Kern County, California. Our operations are about 10 miles east of Bakersfield in the Edison Field. My company bought the properties from a major oil company in 1994 when the production rate had declined to 250 barrels per day. We now sustain the field at about 1,000 bpd after returning wells to production, drilling of “stepping out” and infill producers, and using cyclic steam injection which we call “huff ’n’ puff.” Productive horizons in our field include the Kern River formation, the same one that produced over one billion barrels of oil via steam flooding in the famous Kern River Field, 15 miles to the northwest.

Until this year, we had not considered steam flooding as a viable option for our field. The previous operator conducted an expensive, four-injector steam flood pilot in 1982-83 that was abruptly terminated and never expanded. After we acquired the field in 1994, a consultant reviewed the pilot results and deemed the reservoir unsuitable for a commercial steam flood. Since a major oil company with all of its resources and technical expertise had walked away from the Edison Field, and in light of the consultant’s report, we thought, “How could there ever be steam flood potential here?” Boy, could we ever have been more wrong!

Although the Edison Field’s surface area is about 25 percent of the Kern River Field, our production rate is 1,000 barrels per day compared to 82,000 for the Kern’s. Based on this ratio alone, we should have been taking a closer look at the 1983 pilot results. Recently, I came across some data that gave me “probable cause” for a further study of the failed steam flood pilot. While preparing a program to re-enter one of the abandoned pilot steam injectors and reactivate it as a producing well, I found, hidden in the “abandoned well” file, a nugget of data from an uninterpreted high-resolution, down-hole temperature and pressure survey. Each of the four steam injector files had similar data showing that only one of the injectors had wet steam going down-hole; the other three had hot water. Now my interest was piqued. It was time to re-open the files and reconstruct the steam flood pilot’s cause of death, just like they do on the television drama “Cold Case.”

I looked at flowline temperature data and found that the hottest production wells surrounded the only injector receiving wet steam and significant heat, which interestingly was located at the highest ground-level elevation. Based on the poor steam quality and areal heat distribution, there was enough information to cast doubt on the results of the 1983 pilot and begin evaluating a new steam flood pilot project. I shared my findings at a Society of Petroleum Engineers presentation in Bakersfield on September 11, 2007. While preparing the talk, I dug deeper into the steam flood pilot results.

Performing an autopsy on a failed project requires an awareness of the entire system’s pathology: wells, reservoir, and organization. Most organizations have a common weakness related to the project’s life-cycle management. They do a good job of documenting studies and plans prior to execution, but unfortunately, post-mortem studies and documentation are practically non-existent. The 1983 Edison Field flood pilot is no exception. I found several reports and summaries documenting simulation studies, plans, and meetings leading up to the project’s execution. However, I did not find a single word about anything after the project was started.

The reports dating from 1981 showed that a steam flood pilot was necessary to determine whether the Kern River Field successes could be duplicated in the Edison Field. The pilot’s design was based on computer simulation models developed by the major oil company’s experts in Texas. Since the company had no significant steam flood experience in California, nor any experts in the field office, they were at the bottom of the learning curve compared to other major oil companies, and the pilot suffered accordingly in three major areas. First, the simulation engineers modeled the steam flood as a strictly frontal displacement process, as shown in the figure on the next page. Gravity was not viewed as a driving force for oil recovery, and actions during the pilot showed that steam breakthrough was feared rather than encouraged. Second, steam distribution best practices involving “dead end” T’s published in 1978 were ignored, and the project’s steam distribution system involved vertical T’s to accommodate farm tractors (see photo on the next page). Third, the pilot did not include any observation wells, which would have been useful in observing the true physics of steam movement in the reservoir.

The simulation reports indicated that the only uncertainties that mattered for the pilot’s success were oil saturation and relative permeability. Due to arrogance and/or ignorance, organizations often delude themselves into believing that one unique solution explains production results, conveniently covered within the pre-existing computer simulation sensitivity runs. After 12 months of steam injection, with very poor production results, it appears the organization concluded that the oil saturation was too low. The project was terminated and its technical support vanished. Several months after the termination, the operations group obtained a key piece of data that the modeling group apparently never reviewed. The operations group was concerned about the very high water production rate in the pilot’s confined production well. A water entry survey showed a 400 bpd water entry from perforations a hundred feet above the steam flood injection interval. For me, this was the smoking gun.

It is clear now that most of the injected steam never processed the intended interval, instead breaking through a thin overlying siltstone and immediately pressuring up the overlying wet sand, as shown in the illustration below. Due to the low compressibility of water, the steam injection resulted in immediate increases in water production rates in the offset producing wells. The simulation engineers modeled the pilot drive interval as a fully confined closed system and concluded that the response in the producing wells was due to low oil saturation in the drive interval. Additionally, the updip-confined well had a nice temperature response. The casing on this well was kept shut in to discourage steam break-through. The project was managed as a frontal displacement process. The benefits of gravity were ignored even though six of the seven simulation cases showed oil/steam drainage efficiency of 80 to 90 percent. In effect, the pilot tested the wrong process in the wrong interval.

We are currently planning a new steam flood pilot in an updip location where there are no intermediate wet sands to act as heat sinks and where reservoir pressures in the sands are less than 30 pounds per square inch. Since expected oil/steam drainage efficiency is at least 80 to 90 percent, current oil saturation is not a key uncertainty. We now believe the key uncertainties are vertical and areal conformance of injected steam. The new pilot will operate in a manner to encourage steam breakthrough and gravity drainage and will use observation wells to help quantify conformance uncertainties. We hope we can put gravity to work in recovering millions of barrels of additional oil. John Lennon would be proud.

Don Nelson is a Stanford-trained petroleum engineer with more than 25 years of field experience.