Deepwater E&P Auger's geology guidepoint to deepwater US Gulf potential

May 1, 1995
D. S. Pfeiffer D. T. McGee J. G. Crump Shell Offshore The Auger TLP is currently constraining production from the reservoir, and Shell is in the process of debottlenecking the facility. Production will be raised to 65,000-70,000 b/d of oil. Table: Average Reservoir properties from rock data The well paths from the Auger TLP, shown as square in Block 426. Geological profile of Auger's five sands is shown in this composite type log of the various discovery and development wells.

Auger facility limiting reservoir performance;sand permeability,
continuity much higher than expected

D. S. Pfeiffer
D. T. McGee
J. G. Crump

Shell Offshore
The Auger TLP is currently constraining production from the reservoir, and Shell is in the process of debottlenecking the facility. Production will be raised to 65,000-70,000 b/d of oil.

Shell Offshore's Auger Field currently holds the record for deepest water depth oil and gas production in the Gulf of Mexico and deepest water depth production from a tension leg platform (TIP) in the world.

Production began in April, 1994 from the first well and has exceeded expectations. Currently, six wells are on production from the deepest reservoir with an average daily production rate as high as 62,000 b/d of oil and 125 MMcf/d of natural gas. The field is located in 2,862 ft. of water in the Garden Banks OCS area, 214 miles southwest of New Orleans.

Initial discovery took place in 1987 with appraisal in 1988. Fabrication of the TLP began in 1990 and installation was completed in February 1994. During this time, ten wells were predrilled (drilled, cased, and temporarily abandoned). Following installation, nine of the predrilled wells were completed and brought on production. Currently, most likely recoverable reserve estimates are 220 million BOE.

The Auger TLP is Shell's first effort with this development system. Previously, Shell held Gulf of Mexico water depth production records with the Cognac and Bullwinkle fixed platforms. The Auger TLP is a 32-slot structure with a permanent fixed drilling rig. The TLP is positioned by a lateral mooring system over each wellhead on the seafloor for drilling or completion operations.

Stacked turbidites

Auger Field consists of five major, stacked-pay intervals ranging from upper Miocene to lower Pliocene in age, and from depths greater than 20,000 ft to 15,000 ft:

  • The deeper pays, S, R, and Q, are volatile oils with API gravities of 34-38°.
  • The shallower pays, N1 and O, are condensate-rich gases.

The S, R, and O reservoirs are dominated by sheet sands. Above the S reservoir, across Miocene-Pliocene boundary, the reservoir architecture shifts from sheet geometries to shingled amalgamated channel geometries in the N1 and O reservoirs. The change in depositional styles within the field reflects the progradational nature of the basin fill and a shift from active lobe-dominated sedimentation in a rapidly subsiding basin to channelized sedimentation in a slowly subsiding basin, with minimal accommodation space.

The reservoir quality of these five pay intervals is very good, but porosity and permeability decrease with depth, which is largely a result of compaction and burial depth.

Main pay

The S sandstone, the main pay of the field, is a 200 gross ft plus thick package of coarsening and thickening-upward sheet sandstones containing volatile oil. The hydrocarbons are trapped against the upthrown side of the down-to-the-north fault and pinch out to the northeast near the paleoridge.

The accumulation has a hydrocarbon column height of 1,200 ft and an average net-to-gross of 0.8. The porosity range is 22-26%. The permeability range is 40 to 500 millidarcies (md).

The S reservoir consists of two sandstone members, S1 and S2, that are separated by a 15-20 ft thick mudstone layer and are internally correlated over the entire extent of the reservoir. The well control lies mostly along the structural strike of the field, which is interpreted to be the primary depositional flow direction (northwest to southeast).

The depositional model for the S reservoir is onlapping, layered, and amalgamated sheet sandstones deposited with excellent lateral continuity, but separated by dominantly thin (inches to a few feet thick), continuous mudstones. The high degree of lateral continuity is a product of the actively subsiding basin providing sufficient accommodation space for the incoming turbidite sediments.

Sheet, channel sand

The Q and R reservoirs contain volatile oil and have the same trapping mechanisms as the S reservoir. Although the gross thickness of each of these packages is as much as 200-275 ft, the net-to-gross is at best .55 in the Q and .60 in the R. Porosity is in the range of 24-29%. Permeability is in the range of 40-500 md.

Within the Q and the R reservoirs, the bottom two sandstone members are layered and/or amalgamated sheet sandstones, and the top sandstone member is a nonleveed amalgamated channel sandstone. All of these sandstone members are separated by mudstones. This change from the S to R and Q reservoirs reflects the progradational nature of the basin fill.

In the Q and R reservoirs, there is less continuity in the sheet sandstones, indicating that there was less accommodation space in the basin at the time of their deposition. However, the channel sandstones are not topped by a fining and thinning-upward package like the N1 and O reservoirs to be described later, suggesting that enough accommodation space existed such that no waning flow deposits were deposited at the field locality.

Amalgamated channel sand

The N1 and O sands are condensate-rich gas reservoirs that are distinctly different from the sheet sandstone-dominated reservoirs discussed above. The trap for the N1 and O sandstones is a faulted nose bounded on the east by the major down-to-the-east fault system.

The down-to-the-north fault consists of a series of small offset fault segments that do not seal the reservoir into isolated fault blocks. In contrast to the general coarsening and thickening-upward nature of the sheet sandstone reservoirs, the N1 and O sandstones are fining and thinning-upward sections. These reservoirs each have a basal massive channel sandstone package 20-120 ft thick topped by a fining and thinning-upward overbank/levee package consisting of sandstone with interbedded mudstone 35-120 ft thick.

The average net-to-gross for the massive sandstone is 0.89, but varies from 0.75 to 0.99. The average net-to-gross for the thinly bedded member is 0.30, but varies from 0 to 0.50. Rock data indicate that the porosity has a range of 25-30%. Permeability is in the range of 110-1000 md.

There is a lower degree of continuity within the Nl and O reservoirs compared to the S, P, and Q reservoirs. Stratigraphic shingles are visible on the seismic data and are interpreted to represent distinct sandstone bodies. The primary depositional flow direction for the identified shingles is northwest to southeast, but the shingles along the eastern margin of the reservoirs appear oriented north-south, suggesting a change in dominant flow direction through time.

Mapping the relationships between shingles in 3D was the basis for interpretations and maps of flow barriers and baffles. Based on such maps, it is interpreted that the O sand reservoir has no complete barriers to fluid flow, but it has many baffles that will most likely affect drainage patterns of the producing wells. The shingles in the Nl reservoir appear more amalgamated than in the O reservoir and are interpreted as not forming drainage barriers or baffles.

Initial development strategy

Although reservoir quality and thickness appeared very good, the initial development plan did not take full advantage of this potential. The plan called for a total of 27 development wells, including eight water injection wells, although a final decision on the latter would not be made until production data confirmed the need for pressure maintenance.

Early reservoir simulations suggested the possibility of production rates exceeding 10,000 b/d of oil per well from the S sand. However, since no actual production testing had been done, confidence in these estimates was modest at best. Moreover, sustained production rates of this magnitude had not been demonstrated in the Gulf of Mexico, and concern existed regarding the long-term effects of such high production rates on gravel-packed completion integrity.

Consequently, the initial plan was not as aggressive in reducing the well count as the current plan. Given its average per well recovery figure of over 8 million BOE, however, it was by no means conservative.

The strategy evolves

The development drilling program began in May 1990 with the spudding of the A-1 well, the first S sand development well. As the program progressed, confidence in the continuity of the S reservoir increased. Ultimately, only six production wells were drilled.

Confidence in the potential production rates also increased when data from a second core revealed some better permeabilities than had been seen in the exploratory well core. Also, reservoir simulations showed good confidence that high rates could be obtained, even if the completion efficiency was poor.

The completion efficiency of the S sand wells remained a large uncertainty to the end, due to the very long gross intervals (200 ft) to be gravel packed, and the great depth (19,000 ft) of the interval.

Study of the Nl and O sands revealed much larger aquifers than previously believed, so development plans were altered to take advantage of this. Well locations, which had been mid-dip in these very rich gas/condensate sands, were moved updip. In addition, an effort was made to reduce the number of wells to the absolute minimum in each of these sands.

In the case of the N1, well A-8 was drilled very high on structure. Pay quality in A-8 is considered very good, and continuity of the sand appears be better than that of the O sand, so development of this 1,000-acre reservoir will be done with a single well equipped with 4 1/2-in. tubing instead of the standard 3 1/2-in. string used in most of the other wells. This should allow production rates of 50-70 MMcf/d of gas and 9,000-14,000 b/d of crude.

The O sand reservoir, with its apparent baffles to flow, is somewhat more complicated. As in the case of the N1, O was found to have a larger aquifer than originally mapped, and well locations were moved updip. Pay thickness, however, is low in updip areas of this reservoir.

In order to take advantage of high structural location and to obtain high well rates, a horizontal well was proposed for an updip location. This also offers the possibility of incremental reserve volume by connecting separate shingles of the reservoir. In order to realize the potential rates from this well, 4 1/2-in. tubing will also be installed here.

The new development program, which takes full advantage of the prolific rate potential and good continuity of these turbidite sands, has resulted in a dramatic reduction in well count and cost. Current estimate of well count is 11 if no water injectors are needed, and l3-l5, if water injection is required. The cost savings is $150-250 million.

Higher production

The hydrocarbons in the reservoirs contain volatile oils and condensate-rich gases. Currently, only the S reservoir is being produced with wells averaging about 9,000 b/d oil and 20 MMcf/d gas. The wells can produce at higher rates, but the facility is currently limiting production rates.

The A-7 well was tested at 12,980 b/d oil and 24.7 MMcf/d gas on a 50/64-in. choke. Production data from the other wells in this reservoir indicate that they could produce at similar levels if capacity existed.

Three other wells completed in the Nl, O, and R reservoirs have not been produced long due to capacity restraints, but the A-14 well completed in the O gas reservoir had peak production of 33.1 MMcf/d gas and 7,141 b/d crude on a 40/64-in. choke from 30 net ft of massive sand and seven net ft of thin bedded sand. The A-19 R reservoir completion tested at 7,981 b/d crude and 25 MMcf/d gas a 40/64-in. choke.

For surveillance purposes, two permanent bottom hole pressure gauges with surface readout were installed in S sand wells, one near the northern end of the reservoir (A-5), and one in the south (A-7). The pressure behavior of these two wells confirms the good reservoir continuity in this reservoir.

Buildup tests in both wells have indicated average permeabilities of 300-500 md, compared to the averages of 150-200 md found from core plug measurements. Completion efficiency, as measured by wellbore skins from the buildup tests, averages about 30 for the northern well and 18-20 for the southern well.

Acceptable drawdown

The northern well initially suffered serious impairment due to completion fluid incompatibilities, and may still be somewhat impaired. So, the southern well is considered a more reliable indication of what the average gravel-pack completion efficiency is in this reservoir.

The values of 18-20 are in line with the values expected prior to completion. Drawdowns in these wells while producing at 8,000-10,000 b/d appear to be 300-500 psi, which is considered acceptable. No sand production has been noted to date.

Average shut-in pressures are determined whenever one of the wells containing the pressure gauges in the S sand is shut in. These are used to calibrate the reservoir model in order to determine the effective aquifer extent. Results to date suggest that the aquifer is large relative to the hydrocarbon pore volume because the pressures have held up much better than expected, although the decline rate is still high (6-9 psi/day). If this performance continues, water injection will not be required.

Although the A-8 N1 well, equipped with 4-1/2-in. tubing, has not been produced extensively, indications are that it will reach its expected potential production rates.

Current, future activity

The Q sand development well, A-17, has just been drilled. The next and final development well will be the O sand horizontal well. This will complete the initial development of Auger.

Conclusions

  • Reservoir quality: The quality of the reservoir is very good in the five major, stacked-pay intervals, but porosity and permeability decrease with depth, which is largely a result of compaction and burial depth.
  • Plan improvement: The initial development plan was not as aggressive in reducing the well count as the current plan. The current plan takes full advantage of the prolific production rate potential and good continuity of these turbidite sands. The cost savings is $150-250 million.
  • Facility-constrained production: Currently, only the S reservoir is being produced with wells averaging about 9,000 b/d of oil and 20 MMcf/d of gas. The wells can produce at higher rates, but the facility is currently limiting production rates.
  • Debottlenecking: Because of the high production rates, work is proceeding to debottleneck the processing facilities and increase capacity to 65-70 million b/d of oil and 160-170 MMcf/d of gas.
Authors
D. S. (Deborah) Pfeiffer is a senior geological engineer with Shell Offshore. She is responsible for reservoir characterization of the Q, R, and S reservoirs in the Auger Field. She holds a BS from the University of Nebraska and a MA from the University of Texas at Austin.
D. T. (David) McGee is a senior geologist with Shell Offshore. He is responsible for reservoir characterization of the N1, O, and S reservoirs in the Auger Field. He holds a BA from the University of Montana and a MS from the University of Oklahoma.
J. G. (James) Crump is a staff reservoir engineer with Shell Offshore. He is responsible for reservoir engineering in the Auger basin. He holds a BS from Massachusetts Institute of Technology and a PhD from California Institute of Technology.

Copyright 1995 Offshore. All Rights Reserved.