K. Persad¹ and S. K. Addy²

¹Mora Oil Ventures Limited, La Romaine, Trinidad
²CGG Americas Inc, Houston, USA

ABSTRACT: The Sidetracking of Two Wells in the Mora Field, Trinidad using Seismic and Attribute Facies Maps – Choice of Locations and Results

The Mora field is located 9.5 miles offshore from the southeast coast of Trinidad in the structurally complex petroliferous Columbus basin, which contains other fields, such as, Teak, Poui, Galeota, Samaan and others. The oil and gas production in these fields comes from post Miocene deltaic, near shore and shallow marine and deep water submarine sands in prograding sequences in structural and stratigraphic traps. More than 25000 ft of post Miocene sediments were deposited in the Columbus basin fed by the Orinoco River in the south. In the Mora field, sand stones of Gros Morne and Mayaro Formations of Pliocene age are the reservoir rocks. These sediments were deposited in a tectonically active basin with transpressional features, diapirism and growth faulting. As a result, rapid facies changes are observed with reservoir thickness and quality varying over short distances.

The discovery well, Galeota Ridge (GR-3) was drilled in 1982 by Amoco Trinidad Oil Company. This well flowed 2500 BOPD of 44.3 degree API oil from an 80’ thick shaly sandstone section of Gros Morne Formation. This was followed by GR-4 discovery producing 3108 BOPD of 29.5 degree API oil. The Mora platform was installed to facilitate production and drilling of exploratory wells such as, Mora 4 and Mora 5 among others. In 1995, a 28 sq. mile 3D seismic survey was acquired by Western for Amoco. After drilling a few disappointing wells, failing to identify rapidly changing reservoir facies in a high dip situation and the existing production not meeting the economic criteria, Amoco sold the field to Mora Oil Ventures Limited in 1994.

This paper presents results of side tracking of Mora 4 and Mora 5. The choice to sidetrack Mora 4 and Mora 5 was based on an integrated analysis of the various facies maps along with geologic and engineering data. In both instances, the wells were successful, coming in on depth, with greater oil sand thicknesses and higher flow rates in the sidetracked wells. The net pay and production were improved from 10’ and no production in Mora 4 to 30’ and 240 BOPD in Mora 4X. In Mora 5X well production increased to 500 BOPD in 100 ft sand from 300 BPD from 50 ft sand in Mora 5 well.

The regular seismic data were interpreted using CGG’s Stratimagic (presently owned by Paradigm), a 3D seismic stratigraphic interpretation system. All horizons were picked using an advanced model-based autotracking tool. Several horizon attributes such as amplitude, dip, azimuth, curvature, near reflector geometry, roughness, etc. were calculated. Intervals were created from these horizons for facies classification. Seismic facies maps were generated by neural network process. This patented technology was licensed from Elf Aquitaine. The objective of this method was to classify the interval based upon the shapes of the seismic traces. The seismic facies map was actually a similarity map of the actual traces compared to a number of synthetic traces generated by the software. This was an un-piloted method and required no prior acoustic modeling or well input. Various volume attribute maps such as, average sample amplitude, standard deviation, maximum peak, maximum trough etc. were created for the interval along with horizon slices at every 4 ms. Attribute facies maps of the intervals were generated from combining various attribute maps within the interval using a similar neural net technology (Addy 1998). The objective here was to classify an area based on several different kinds of maps and not seismic traces as in the previous case. In general, sands were of low velocity compared to the adjacent shales. As a results, sands were correlated to troughs in this SEGY reverse polarity data and in our display troughs have been colored red.

AAPG Search and Discovery Article #90906©2001 AAPG Annual Convention, Denver, Colorado