The ability of the riser fluid used in a dual-density drilling system to retain sufficient solids suspension and cutting transport properties after heavy dilution of the well fluid with an unweighted liquid are both of critical importance. A practical system also requires that the mixed fluid that returns to the surface in the riser be separated into low-density and high-density streams for reuse.

The investigation of the liquid dilution alternative began by studying potential drilling fluid

Figure 1. Estimated trouble-free drilling time for conventional 17,000-ft (5,182-m) well.
formulations. The maximum drilling fluid density required for applying the dual-density concept to any of the actual deepwater wells studied was determined to be 17 lb/gal. Consequently, a number of lab samples of 17 lb/gal synthetic-based muds and a 7.6 lb/gal-dilution fluid with similar fluid makeup, but no weighting material, were formulated and evaluated for suitable properties individually and when combined to form a riser fluid.

These lab formulations provided good emulsion stability and ideal densities, but it was not possible to achieve appropriate rheologies in both the weighted fluid and the riser fluid. Earlier tests reported by de Boer (2003), however, gave more favorable properties when using a lower density weighted fluid. It was concluded there was potential for favorable properties to be achieved, and evaluation of separation methods was deemed necessary. Separation experiments were conducted with a laboratory centrifuge and a 2-in. hydrocyclone from a “clay ejector” system used in processing mud in the field.

The centrifuge was able to separate the riser mud into near ideal densities for dilution and drilling fluid. However, the dense slurry retained in the centrifuge had lower emulsion stability than the feed stream. Separation using the 2-in. hydrocyclone achieved less contrast in density between the low- and high-density discharges but consistently resulted in a beneficial increase in the stability of the mud emulsion and had more desirable rheological properties.

Considering these results along with those for a field-scale centrifuge reported by de Boer, it was concluded that separation of riser mud into fluid streams suitable for use as a dilution fluid and a drilling fluid is feasible and should be practical for field application. Subsequent recirculation tests on fluids from the 2-in. hydrocyclone indicated a density contrast between the low- and high-density streams similar to that reported by de Boer can potentially be achieved using hydrocyclones in lieu of centrifuges.

The experiments also provided insight into meeting the challenge encountered in the tests on laboratory muds when trying to simultaneously meet the rheology requirements for the riser and wellbore mud streams. Separation with the hydrocyclone and de Boer centrifuge deliver fluids that come close to meeting these requirements. Apparently, the separation overflows contain fine residual barite particles and have a finer emulsion than the feed mud. These factors may be the key to achieving the desired viscosity for the riser mud without excessive viscosity in the wellbore mud. Comprehensive testing of a more complete hydrocyclone-based system and field-scale centrifuges is necessary to confirm this preliminary conclusion.

Cost comparison

Past, current and future deepwater Gulf of Mexico activities were used to define the water and well depths and general geology of three hypothetical wells where dual-density systems might be applied. It was concluded that these wells:
• would be drilled with dynamically positioned rigs;
• would encounter abnormal pressures; and
• would not encounter long salt intervals.

The rigs would be capable of drilling in 10,000 ft (3,048 m) of water with a 21-in. riser and 183¼4-in. (10,000 psi) subsea blowout preventer (BOP) system. The two development wells would be in a water depth of 6,000 ft (1,829 m) with total depths of 17,000 ft (5,183 m) and 23,400 ft (7,134 m). The exploratory well would be in a water depth of 10,000 ft with a total depth of 20,500 ft (6,250 m). It should be noted that about 50% of all future deepwater Gulf of Mexico wells are likely to encounter salt and therefore are not well represented by these wells. Subsalt wells are less likely to benefit from the advantages of a dual-density system and therefore were excluded from consideration.

The data collected also was used as a basis for generic deepwater pore pressure and fracture pressure estimates. The average depth of the top of abnormal pressure for deepwater Gulf of Mexico wells was concluded to be about 2,000 ft (609) below the mudline and is independent of water depth. The average local pore pressure gradient in the abnormally pressured section of deepwater Gulf of Mexico wells was found to be about 15.5 lb/gal equivalent. The average fracture pressure in deepwater Gulf of Mexico wells can be estimated using these pore pressures, a compaction-based overburden stress model and a stress ratio of 0.7. Estimates of pore and fracture pressures based on these averages were used when selecting the mud weights, casing sizes and setting depths as well as hole sizes for the three representative well designs. The resulting well descriptions are reasonably similar to the descriptions of the actual and planned wells in similar water depths that were included in the data industry sources provided. Relatively small kick and trip margins were assumed in the shallower, lower risk section of the well to allow 9 5/8-in. production casing to be used.

A “trouble-free” drilling cost estimate was made for each of the three example well designs. The cost estimates assume each well is productive, production casing is run and set, and well operations are suspended pending later completion. Completion costs should not be impacted by dual-density drilling operations and were not included in these cost estimates.
The estimate of recurring daily costs assumes the rig day rate is US $160,000 for all cases. The additional recurring costs for rentals, services and other support are assumed to be $190,000 per day. Additional point-in-time costs such as the subsea wellhead, casing, cement, minimal open-hole logging and building an initial mud volume generally represented about 11% of overall well costs. A “trouble-free” drilling time estimate for the single-density well design is shown in Figure 1. The total time on location is estimated to be 50 days. The corresponding total well-cost estimate is $19.7 million.

The dual-density design for this same example development well represents a significantly smaller trouble risk from a well control and lost returns perspective than the single-density design for the same well because of larger kick and trip margins.

The same rigs will initially drill dual-density wells as drill conventional wells. The major
Figure 2. Estimated trouble-free drilling time for a 17,000-ft (5,182-m) well drilled with a riser gas lift.
components required to implement a riser gas-lift system would be nitrogen generation units (10 Mscf/min, 5,000 psi), a rotating control head at the surface on the riser, a choke dedicated to the riser and a high capacity mud-gas separation system (300 psi, 30 Mscf/min, 2,000 gpm). An additional cost of $60,000 per day is anticipated for these gas-lifting services and equipment rentals during dual-density operations.

A trouble-free drilling time estimate for the total time on location for the dual-density well design of the 17,000-ft (5,182 m) example well is shown in Figure 2. The total time on location is estimated to be 41 days. This drilling time results in an estimated total well cost of nearly $17 million. The apparent cost reduction realized by using a gas-lift, dual-density well design for this example is about $2.8 million. This would represent a 14% cost savings per well compared with the $19.7 million well cost using conventional single-density methods.
A similar cost estimate was performed for a dual-density well drilled with a liquid dilution system. The casing program was less optimum; additional equipment rentals were $20,000 per day compared with $60,000 per day. Trip and kick margins were only slightly more conservative than for a conventional well, and the cost savings were smaller than for riser gas lift but are still significant at 7%. Savings from using a riser gas-lift system to drill representative deepwater wells ranged from $2.1 million to $3.3 million or between 9% and 16% without considering the effect of reduced trouble costs.

Larger savings would be possible if smaller risers were available to extend the water depth limits of smaller, less expensive rigs. Larger savings also could be realized on wells whose depths or pressure profiles would otherwise require more casing strings and from using a wellhead pressure less than the hydrostatic pressure of a column of seawater, which would allow further optimization of the casing program for a dual-density well.

Finally, trouble costs are often a significant portion of deepwater well costs. Although a dual-density system might increase equipment-related downtime and trouble cost, the larger trip and kick margins should eliminate much of the trouble time associated with lost circulation, kicks and extra casing strings often encountered in deepwater drilling. A recent exploratory well drilled in 10,000 ft (3,050 m) of water exemplified these kinds of problems. Lost returns, flows and related problems consumed 11 days of rig time and increased the well cost by about 15% over the original estimate. An unplanned extra string of casing also was required to reach total depth. Another case history showed a similar experience.

The opportunity for the larger kick and trip margins provided by a dual-density system to reduce these kinds of trouble would reduce deepwater drilling costs. If only half of the trouble described in these examples were eliminated, the well costs would go down by about 8%, bringing the expected overall savings between 17% and 24%. Considering the other potential beneficial impacts of riser gas-lift operations described above, a dual-density well’s costs could be as much as 50% less than those for conventional methods to drill the same wells.

Conclusions

Additional research into and development of dual-density systems is needed, particularly for the riser gas-lift approach.

Riser gas-lift was projected to provide a cost reduction of at least 9%, and more likely between 17% and 24%, of the trouble-free cost for conventional drilling practices for the three example wells. A riser gas-lift approach also increases the feasibility of drilling deep wells in deep water that might otherwise be impossible with current rigs and equipment.
Additional cost reduction can be realized by using smaller risers and rigs.

The combined effect could be reductions in actual costs equal to as much as 50% of the estimated trouble-free cost of a well. Riser dilution using liquids is not expected to be as effective for reducing well costs and would reduce costs vs. conventional operations by 7% for the example studied.

Well control with riser gas-lift was studied and found to be feasible using methods generally analogous to conventional operations and those using a subsea mud lift pump. Kick removal should use returns up a gas-lifted choke line.

Results from this and work done by de Boer indicate that mud separation for a liquid-based riser dilution system is possible. A complete system to accomplish fluid processing with hydrocyclones was not defined but seems feasible.

Average pore and fracture pressure trends in the deepwater sediments of the Gulf of Mexico have been developed. The average depth of the top of abnormal pressure is 2,000 ft (610 m)below the mudline and is independent of water depth. The average local pore pressure gradient below this depth is equivalent to 15.5 lb/gal. The average fracture pressure can be estimated using these pore pressures, an overburden stress model and a stress ratio of 0.7.