CLARKE, DONALD D., City of Long Beach Department of Oil Properties, CA, USA, CHRISTOPHER C. PHILLIPS,Tidelands Oil Production Company, Long Beach, CA, USA

Abstract: At 68 Wilmington Still Has Life: New Technology Revitalizes The Old Field

The Wilmington oil field of Los Angeles County California, the third largest in the United States, was discovered in 1932 and has been on continuous production ever since. Cumulative oil production has exceeded 2.5 billion barrels. Production is from Pliocene and Miocene Age basinal turbidite sands. There are seven productive zones that have generally been subdivided into 52 subzones. Locally the subzones have undergone detailed reservoir characterization in an attempt to better define the actual hydrologic units. The asymmetrical anticline is highly faulted and development proceeded from west to east through each of the ten fault blocks. In the areas with the longer production history (68years+) water cuts exceed 96% and the reservoirs are near the economic limit. Several new technologies have been applied to specific areas to improve the production efficiencies and thus prolong the field life.

Tertiary and secondary recovery techniques utilizing steam have proved successful in lower Pliocene and upper Miocene heavy oil reservoirs in the oldest section of the Wilmington Field. Field redevelopment is due, in part, to the participation in two United States Department of Energy (DOE) cost share projects.The mid term DOE project involves detailed reservoir characterization and optimization of a steam flood in the Tar Zone, Fault Block II. Lessons learned from that project were successfully applied in the Tar Zone, Fault Block V (4000 meters to the East). The short term DOE project focuses on 3-D property and geological modeling to define and exploit remaining oil saturation. Steam is used to fuse the unconsolidated turbidite sands to allow selective perforated completions to optimize water flood recoveries.

Novel alkaline water/steam sand consolidation techniques have been applied to each of the horizontal wells. Basically decomposition of bicarbonate ions at the steam generator produces wet steam that consists of high pH (11-12) liquid phase and a low pH (5-6) vapor phase. When the wet steam is injected, the high pH liquid phase partially dissolves feldspars at the elevated temperatures near the well bore. As the temperature and pH decrease away from the well bore, the reaction products (zeolites) precipitate out and cement the formation grains. Because this process is taking place during injection the rock permeability near the well is maintained by the high injection from the well bore to the surrounding formation. This process reduces sanding problems and reduces well costs.

The geological data was modeled and ten horizontal wells were drilled. Prior to drilling, it was discovered that the intraformational compaction caused numerous structural misties that could adversely affect the horizontal placement. This problem could negatively impact most unconsolidated sandstone reservoirs.

CASE HISTORY 1

The field has a long history of development with well logs dating back as far as 1937. Traditional, detailed reservoir characterization work has been performed on over 600 wells within the Tar Zone, Fault Block II.Vertical and lateral extent has been defined for 16 horizons. The scattered data from these wells has been compiled and incorporated into a three-dimensional deterministic geological model. In July 1995 the drilling program started and five observation wells were drilled. Data inconsistencies were revealed, when core hole OB2-003 was drilled. During the mapping it was discovered that the newer wells did not match the structure as defined by the older wells. Ground subsidence from oil production has been a continuing problem for the Wilmington oil field and historically, corrections to the data have been applied. The target window for the horizontal wells was only four and one half meters (15 feet); therefore precision placement was critical. Because of this, the subsidence correction was reevaluated. It was found that there were three significant variables. Surface subsidence over time is a critical variable. Every year the well pad is at a different elevation (usually lower with time) and some operators had fabricated well head elevations to normalize geological information to a particular datum year. Intraformational compaction over time changes both the location of structural horizons with respect to sea level and the thickness between horizons. The necessary data corrections were made and the horizontal wells were successfully drilled.

Modern computer mapping software was used to model the Tar Zone in Fault Block II. When problems were encountered the software was used as a tool to find the data busts. An analysis of the problem data revealed that data from wells of the same vintage were compatible but wells of differing ages were inconsistent. Once the data was corrected the software was used to rapidly produce new models. Cross sections were constructed from the models and used to assist the geologist in geosteering the horizontal wells. The sections were scaled and graphically enhanced to match the logs and surveys produced by measurement while drilling and logging while drilling. The method worked so well that excellent communication between the geologist and drilling engineer permitted instantaneous drilling rates of up to 185 meters per hour (600 feet per hour) within a four and one half meter (15-foot) target window. The 3-D visualization of the geological model helped the geologist and reservoir engineers to place the wells with a far greater confidence than before.

CASE HISTORY 2

The Fault Block II SAGD steam flood expansion has not yet proved economic but the steam flood newly established in the Fault Block V,Tar Zone has exceeded our expectations. Three 30-year old active water flood wells remained in the steam project boundary as of March 1996. The wells averaged 16 barrels of oil per day and 200 barrels per day gross with an average water cut of 95%.This area had approximately five years of remaining economic life under water flood with recoverable reserves estimated at about 75,000 barrels. Five horizontal wells were drilled based on the careful characterization and 3-D geological modeling. The wells underwent cyclic steam stimulations prior to conversion to steam drive operations. The two steam injectors and three producers average over 200 barrels of oil per day at 70% cut. The other wells in the area have also realized an increase from 20 barrels of oil per day to 200 barrels of oil per day while the cut dropped from 95% to 70%. This is a ten-fold increase in production rate! The project has added 1.7 million barrels of recoverable oil to the Tar V pool.

Several valuable lessons were learned. Old well data is problematical and difficult to work with but it is very useful and should not be ignored. Intraformational compaction is a problem for many reasons. First, it can result in surface damage. Second, geological data must be adjusted for subsidence if accurate modeling is to be performed. Then horizontal wells can bedrilled with precision. Third, the compaction impacts the reservoir properties. Modern computer tools have enabled us to map and remap very quickly and accurately. In addition, these tools are excellent for finding data busts. Cross sections derived from the 3-D deterministic models were constructed down the proposed well course. The cross sections were then used to geosteer the well at the rig.All combined we were able to target windows of four and one half meters (15 feet) in thickness at depths of 700 meters (2300 feet) and accurately place horizontal wells.

CASE HISTORY 3

Under the Department of Energy's short-term cost share project, a horizontal well was drilled in thin heterogeneous Miocene age, Terminal zone turbidite sands. The challenge was to identify and exploit bypassed oil.Technology was used from the mid-term project to sidetrack an existing well bore with a horizontal lateral to capture hydrocarbon reserves not economically recoverable using conventional methods. Areas of high sand quality and high oil saturation were modeled in 3-D. An idle well was recompleted to test production from the target sands and to test the effectiveness of the novel alkaline steam consolidation process on sands deeper than the Tar zone. Based on the success of the first well, a second idle well was selected for sidetracking a horizontal lateral.

The horizontal well was porpoised into a "water flood" position and included two of the individual sands comprising the Hx0 interval. Several thin depositional units of the Hx0 interval are shown and three log curves are shown. With an average of 3.6 meters (12 feet) of net sand out of 5.2 meters (17 feet) of interval, it was a challenge to geosteer the horizontal lateral.As with the mid term project, the cross-sections extracted from the deterministic model were used for geosteering. A post-mortem analysis of the Logging while Drilling (LWD) data helped further define the geological structure.There were new techniques used for the LWD modeling. The heterogeneity of the sands were identified and modeled using the 32 different phase and amplitude measurements. Resistivities derived from the inversion processing and modeling agreed well with both the offset wireline data and the LWD log. These examples show how innovation can bring continued life out of an old oil field.

AAPG Search and Discovery Article #[email protected] International Conference and Exhibition, Birmingham, England