Unformatted text preview: OTC 20939
Gulf of Mexico's First Application of Riserless Mud Recovery for Top-hole
Drilling - A Case Study
J.H. Cohen, SPE, and J. Kleppe, AGR; T. Grønås, T. Martin, and T. Tveit, Statoil; W.J. Gusler, SPE, and
C.F. Christian, SPE, Baker Hughes; S. Golden, Transocean Offshore Copyright 2010, Offshore Technology Conference
This paper was prepared for presentation at the 2010 Offshore Technology Conference held in Houston, Texas, USA, 3–6 May 2010.
This paper was selected for presentation by an OTC program committee following review of information contained in an abstract submitted by the author(s). Contents of the paper have not been
reviewed by the Offshore Technology Conference and are subject to correction by the author(s). The material does not necessarily reflect any position of the Offshore Technology Conference, its
officers, or members. Electronic reproduction, distribution, or storage of any part of this paper without the written consent of the Offshore Technology Conference is prohibited. Permission to
reproduce in print is restricted to an abstract of not more than 300 words; illustrations may not be copied. The abstract must contain conspicuous acknowledgment of OTC copyright. Abstract
This paper describes the preparation and operation for the first use of a riserless mud recovery (RMR) system on the top-hole
section of a well in the Gulf of Mexico. The material includes pre-well engineering and preparations including hydraulic
analysis, pre-job vessel inspection, construction of new equipment, installation, pre-well planning decisions, and rationale for
decisions. In addition also discussed are benefits including improved wellbore quality due to use of an engineered drilling
fluid, logistics savings from reduction of drilling fluids, and minimized environmental impact. The paper also includes
descriptions of equipment installation and testing onboard the drilling vessel operations during drilling, problems
encountered and lessons learned from the operation. A description of all equipment is included in the paper along with
specifications and operation parameters.
An RMR system has application in the top-hole drilling of oil and gas wells. Using conventional methods, drilling fluid
pumped down the drillstring during operations flows out onto the sea floor; this is often referred to as “Pump and Dump”.
RMR collects the mud at the mud line and pumps the fluid back to the rig where it is reconditioned and reused. It allows the
use of engineered drilling fluids and has possible applications for all offshore drilling.
RMR was deployed on a dynamically positioned vessel, and successfully used to drill the 26-in. hole section. Drilling
fluid recovered from the mud line back to the drillship was processed and reused, resulting in significant reduction in the
volume of mud required for this top-hole section. RMR reduced costs through savings in drilling fluid and improved well
RMR is applicable to the drilling of top-holes in the entire Gulf of Mexico. It has significant potential to reduce top-hole
drilling costs, eliminate casing stings, extend casing shoe depths, drill through and past problem formations and improve the
wellbore by eliminating washouts and shallow hazards.
Operators continue to explore and develop fields at increasing water depths. In certain offshore areas where younger
sedimentary rocks are deposited, there is often a very narrow margin between formation pore pressure and fracture pressure
that creates tremendous drilling challenges (Rocha and Bourgoyne, 1994). The solution to this narrow operating window is
to develop techniques that extend the casing setting depths more efficiently.
The use of a riserless drilling technique, dynamic kill drilling (DKD), has been instrumental in successfully pushing the
casing depths deeper in deepwater applications (Johnson and Rowden, 2001). The DKD methodology employs the dual
gradient drilling concept, consisting of the seawater hydrostatic above the mud line with the ability to vary the hydrostatic
below the mud line by drilling fluid weight variations. This functional control of the drilling fluid density is tremendously
advantageous while drilling shallow gas or water flows from over-pressured formations where large washouts, caves,
formation compaction, and collapse could occur (Pelletier et al., 1999). This technique has been repeatedly employed in 2 OTC 20939 challenging deepwater projects where the initial upper hole sections were extended to obtain the required leak-off tests
Pumping a weighted drilling fluid with returns to the mud line presents numerous challenges. A typical DKD operation
requires large volumes of fluid and the ability to logistically provide sufficient volumes of drilling fluid with satisfactory
properties is critical to reaching the deeper casing points. In conventional DKD, a dense mud (16 pounds per gallon (ppg)) is
cut on the fly to the desired weight as drilling of the top-hole proceeds. Most vessels do not have the capacity to mix enough
drilling fluid for these operations and typically two to three support vessels supply the fluid from onshore facilities. This
process is very expensive and in some operations can leave the vessel without sufficient fluid for drilling operations; this is
particularly a problem if drilling does not go as planned and can result in downtime, with drilling waiting on resupply. The
process described above is not always possible in areas where supply vessels or onshore facilities are insufficient to support
the operation. In these situations, drilling decisions are often dependent on the amount of drilling fluid the vessel can carry
and often result in casing being set at less than optimal depths.
RMR uses subsea pumps to recover the fluid exiting the well, and returns it back to the vessel for conditioning and re-use.
Figure 1Error! Reference source not found.A shows a conventional top-hole and an RMR top-hole in drilling
configuration in Figure 1B. Using RMR substantially reduces the drilling fluid requirement and resolves many of the
logistical challenges associated with traditional DKD.
A suction module (SMO) (Figure 2) attached to the low-pressure wellhead acts as a bucket to collect the drilling fluid
exiting the well. The suction port of the subsea pumps attach to the lower portion of the SMO; cameras, shown in the figure,
and a sensitive pressure gauge on the SMO near the suction port monitor and control the level of the drilling fluid in the
SMO. Adjusting the pump speed keeps the level of the drilling fluid in the SMO constant. Pump speed and power are
associated with fluid volumes pumped; changes in flow result in a change in pump performance. If the change in flow is due
to an influx or lost circulation, the mud engineer can adjust the drilling fluid countering the problem. Changes in pump
performance occur with minimal flow changes, so the operator can pick up the problems very early before they can become
too severe. Figure 1—Conventional vs. riserless mud recovery operations OTC 20939 3 Figure 2— RMR suction module Drilling using RMR establishes volume control, and drilling becomes similar to that with the blow-out preventer (BOP)
RMR enables continuous conditioning and re-use of the drilling fluid that allows customized engineering of the drilling
fluid for the specific drilling environment. This allows specific problems to be addressed and solved, and permits the setting
of casing at planned or optimal depths. In addition, volume control is available allowing detection of shallow hazards, such
as lost circulation and gas or water flows, without the need of a pilot hole. Geologic information of shallow formations
becomes available from analysis of the cuttings recovered with the fluid.
MC540 Krakatoa is located in the Mississippi Canyon area of Gulf of Mexico. Shallow formations are composed of the
rapidly deposited sediments from a paleo delta. Turbedite sands are often over-pressured and highly permeable, recognized
as shallow water flows. Shale formations are soft and unstable, recognized as gumbo.
Well planning showed that to reach planned TD at 25,000 ft ( 7620 m)TVD with the number of casing strings available,
the 22 in. needed to be set as deep as possible. The challenges related to this were the 3 ppg pore pressure ramp-up starting at
only 1,300 ft below mud line (BML) and very challenging shallow formations.
A maximum setting depth for the 22 in. was targeted, based on a seismic anomaly at 3,583 ft BML that would not be
penetrated before BOP and riser had been installed. It was decided to drill down to the pore pressure ramp with salt water
(SW) and gel sweeps and install the 28-in. casing to secure the well’s structural integrity. From this depth, the well had to be
drilled with heavier mud, either by means of DKD or by means of an inhibitive water-based mud system and riserless mud
return. Having used RMR on multiple prior occasions, the operator considered RMR for this 2,000-ft (610 M) water depth
Specifically for the Krakatoa well, the RMR system would provide the following advantages over DKD:
• • Environmentally: RMR would enable re-use of the mud when drilling the 26-in. hole section, resulting in
less mud volumes, less transport of mud and less mud discharge to sea.
Improved borehole quality: Due to lower volume requirements and lower costs, an inhibitive water-based
mud system could be used in the 26-in. hole section. The engineered inhibitive water-based mud system
planned to be used with RMR was expected to stabilize the formations better than the traditional DKD mud,
increasing the probability of reaching the planned casing setting depth without remedial work to the wellbore.
This potentially also would improve the cement jobs as the hole is expected to be more in gauge.
Well control: The kick detection would be significantly improved because of the following:
· The RMR system enables real-time monitoring of the seawater/mud interface (“mirror”) by means of
cameras and sensors at the wellhead.
· Increased return flow can be detected immediately by monitoring the subsea pump speed.
· Returns to surface enables volume gain detection in the surface pits. 4 OTC 20939 · Traditional mud logging sensors can be used for kick detection (gas-out measurements, fingerprinting
Improved evaluation: Cuttings samples can be taken in a hole section where traditionally cuttings are not
collected. • The benefits were obvious: RMR would improve the borehole stability in this challenging environment and give better
well control in an area where shallow hazards are abundant. Total cost savings could be documented for riserless mud
returns due to substantially reduced mud volumes and logistic cost savings. In addition, the operator would make an
important step in qualifying a tool for mitigating shallow hazards and reducing the total number of casing strings in the Gulf
RMR Equipment (Stave et al., 2005; Smith et al., 2010)
The RMR system consists of seven major components listed below, three subsea, and four surface.
• Suction module (SMO)
Subsea mud pump module(SPM)
Mud recovery line (MRL)
Winch and umbilical
Hose hang-off platform (HHP)
Control container (two required)
Office container Each component performs a specific function permitting pumping of drilling fluid from the well back to the drilling
vessel where it is reconditioned and reused. The subsea components consist of the SMO, SPM, and MRL. Figure 3 and
Figure 4 show the SPM and SM, respectively. The SPM has four stages of disk pumps. Two 800 hp (600-kW) motors with
a shaft extending from both ends power the pump stages, each motor powering two stages. The pumps and motors attach to a
frame along with necessary piping, flange connections, electronics, and pressure compensation. A flexible suction line
connects the inlet of the SPM to the SMO. Buoyancy material lightens the suction line so that a remotely operated vehicle
(ROV) can move and connect it subsea, using ROV-friendly flanges. A cable run from the SPM to the SMO with the ROV
provides power and communications to the video camera, lights, and pressure transducer located on the SMO. The MRL is a
six-inch diameter soft hose that connects to the SPM and returns the mud to the surface. The MRL in this deployment
partially supported the SPM during operations. The SMO acts as a bucket to collect the returns exiting the well, and attaches
to the low-pressure wellhead using an adaptor manufactured to mate to the specific wellhead being used. Adaptors have been
manufactured for most major wellhead designs. Figure 3— Subsea pump module (SPM) Figure 4— Suction module on running tool OTC 20939 5 A winch and umbilical (Figure 5) deploys the SPM to the sea bottom providing power and communications from the
surface to subsea. An HHP (Figure 6) deploys and supports the MRL. Figure 5— SMO winch Figure 6— Hose hang off platform (HHP) Two standard 20-foot shipping containers (Figure 7) supply power to the system, one for each of the two large motors.
Each control container houses variable frequency drive, filters, and a step-up transformer to provide clean power and speed
control to the SPM 3000-volt motors. Power for the control containers can come either from the rig or from portable
generators. In this case, two 1200-kW generators supplied the 680-volt power. An office and a small workshop built into
another standard 20-foot shipping container house the control computers and system monitoring equipment for the RMR
equipment. Figure 8 shows the office section with computers and monitors. Spare parts stored in this container permit onsite repairs to components, if needed. Figure 7— RMR control containers Figure 8— Office container Preparation and Installation
Preparations begin very early with a site visit to the rig. An experienced engineer and technician conducted an inspection of
the Discoverer Americas when it was still in the shipyard in Korea. Working with rig personnel, the inspection team
analyzed two different deployment methods during the audit, an over-the-side deployment, and a moon pool deployment.
Over-the-side deployments are used for shallower operations, and have the advantage of being done almost totally off-line
from the rig’s operations. However, a soft MRL used in this method of deployment limits the depth of operations. A moon
pool deployment extends operating depth (1,500 ft (460 m) maximum to date), but the MRL consists of jointed pipe and must
be run using the rig’s derrick and tubular make-up equipment. This consumes rig time, even in a dual activity system. It was
decided in this situation to go with an over the side deployment, even though this was approximately 328 ft (100 m) beyond
any previous over-the-side deployments. The analysis resulted in a preliminary equipment layout shown in Figure 9. The
winch was to be located on the starboard ROV deck. 6 OTC 20939 Figure 9—RMR equipment layout The HHP was to be located on the moon pool deck on the starboard side, while the control and office containers and
generator are located on the well test deck. Rig contractor engineering data showed that these locations were acceptable from
deck loading and other technical perspectives. Discussions with the rig contractor determined it was best for this job to
power the RMR system with portable generators, as opposed to connecting into rig power.
This plan and a hydraulic analysis showed the required power needed to move the programmed mud from the expected
depth required a new four-stage pump module. This new pump module was designed to use two double-ended 800 hp (600kW) motors and to be suspended above the bottom of the Gulf. After manufacture, the pump module and the remainder of
the RMR equipment was shipped to a shop facility in Houston, Texas where final preparations were completed.
Several factors limited mobilization to the rig, including space at the shore dock facility, and schedule for equipping the
new rig in the Gulf of Mexico. The winch and HHP foundations were among the first items shipped out. A certified welder
must attach these to the deck and then a third-party inspection conducted. The other equipment followed. After placement,
the cabling between the components was installed. This included an explosion proof computer, mirroring the control
computer, in the driller’s cabin. This computer is used during operations and ensures good communications with the driller
and the RMR operator. To complete the setup, the office container was connected to the rig systems for cooling,
communications and operation of emergency shutdown equipment and warning systems. The winch was load tested with
water bags. Mobilization and installation took 16 days, which was due to a reduced crew of two. This is not typical for a
first-time installation, which is approximately seven days. The reduced crew was needed because of other onboard ship
activities having precedence at the time of installation. Subsequent mobilizations to prepared rigs take approximately five
days. RMR operators report that mobilization of equipment onto Discoverer America went extremely well with no major
All of the equipment shipped in eight 20-ft open top containers. Figure 10 shows the final surface equipment layout.
Once surface preparations were completed, the RMR equipment was ready to commence operations. Figure 10— Final surface equipment layout OTC 20939 7 Hydraulics
A pre-job hydraulic analysis was performed to confirm the pump power requirements and an operating envelope of the
system including pumping volumes with expected mud densities. A proprietary hydraulics program was used to perform this
task. A three-stage pump module was first proposed for the job; however the analysis showed a four-stage pump module
provided a larger margin at the top end of the equipment operating envelope. Table 1 shows the input used in the analysis.
Water Depth Table 1— Hydraulic Analysis Input
2,034 ft (620 m) Pump Stages
MRL ID 3 and 4
6 in (152 mm) Pump Rate Varied Drill Mud Density Varied Formation Density 20.03 ppg (2.40 sg) Rate of Penetration 98 ft/hr (30 m/hr) Mud PV 0.0360 Pa*sec (36 cp) Mud Yp 25 lb/100 ft (11.97 Pa) 2 Table 2 shows the power requirement for different density drilling fluids at different flow rates. The three-stage pump
has a maximum 1200 hp (900-kW) while the four-stage is 1600 hp (1200-kW).
Table 2—Hydraulic Analysis: Power verses Flow rate and Mud Weight
Available shaft power 1200 hp 1600 hp Required power at 1200 gpm and 13.35 ppg 1169 hp Required power at 1300 gpm and 13.35 ppg 1310 hp Required power at 1200 gpm and 14.19 ppg 1306 hp Required power at 1300 gpm and 14.19 ppg 1089 hp Required power at 1200 gpm and 15.02 ppg 1443 hp Required power at 1300 gpm and 15.02 ppg 1611 hp Pump speed was the limiting factor, as shown in Figure 11; the three-stage pump would over-speed with a mud weight of
15 ppg (1.8 sg) at a rig pump rate of 1050 gpm (4000 lpm) while the 4-stage pump could pump over 1400 gpm (5300 lpm) at
this same mud weight. Figure 11—Hydraulic analysis -- pump speed vs. rig pump rate for several different mud weights 8 OTC 20939 Operations
Deployment of the system from the drilling vessel began with the RMR winch running a clump weight to the bottom that
keeps the SPM, umbilical and MRL from drifting too close and clashing with the drillstring. Metocean data describing
currents were used to calculate the drag force on the SPM and MRL to size the clump weight. The data used for these
calculations showed that the clump weight must weigh three tons submerged to keep the system in place. A cable attaching
the clump weight to the SPM is sufficient in length to accommodate vessel heave, while keeping the SPM in a watch circle
small enough to prevent clashing.
The soil data describing the thickness of the seabed silt layer at the MC540 location was unreliable so that calculating the
exact depth the anchor would sink was not possible. Three scenarios were calculated and the attachment sling was designed
with three attachment points to accommodate the uncertainty in these calculations. A float on the cable kept the cable
vertical for ease of attachment to the SPM.
Once the clump weight was in place, the umbilical was retrieved and the SPM prepared for deployment. To deploy the
SPM the vessel crane lifted it from deck and into the sea. Simultaneously, the crew attached the MRL to the SPM. The
SPM was surface tested to confirm proper operation. Figure 12 shows the SPM surface test. Figure 12 SPM Surface Test In an over-the-side deployment, as used for this operation, installation of the SPM with the attached MRL was done off
the critical path and did not consume rig time. The umbilical winch lowered the SPM, and at the same time hose lengths
were added to the MRL. A total of 32 MRL sections were deployed: 3166 ft (20-m) sections and 116 ft (5-m) section. It
took a total of 28 hours to deploy the SPM to depth, where it was connected to the clump weight by the ROV. This process
was not continuous, because it used the rig crane to lift the hose sections as they were connected at the HHP. The rig crane
was pulled for more urgent operations and the SMP running process suspended. The time for deployment was longer than
typical, but was done slowly, because it was the first deployment of this vessel, and it had to be a safe deployment. There
was ample time in the schedule due to other ongoing rig activities. As the SPM was being lowered into place, the rig was
jetting in the 36-inch conductor casing.
The drillstring was used to deploy and install the SMO onto the big bore low-pressure wellhead. The SMO was prepared
in the moonpool and a made-for-purpose J-slot running tool attached to the end of the drillstring. This was lowered into the
SMO and locked into the SMO J-slots; the SMO J-slots are located in the throat of the SMO. Figure 13 shows the SMO on
the running tool passing through the moonpool and splash zone. This installation used an internal low pressure wellhead
adaptor which is a simple adaptor configuration, but one that limits the system to drilling operations only because casing
cannot fit through the adaptor. The decision to run this adaptor and not use the RMR for casing operations was due to the
first-time deployment of the system. The SMO is held in place by its own weight, approximately 2.5 tons. The internal
adaptor is equipped with an O-ring to prevent leakage of drilling fluids while in operation. External adaptors available that
allow running casing and monitoring of casing cementing operations. In most operations, external adaptors are used. OTC 20939 9 Figure 13— SMO on running tool in the moonpool The suction hose connects the SMO to the SPM, and is a flexible steel hose (Figure 14). This hose was lowered to the
seabed by an auxiliary winch located on the main umbilical winch. The SPM and SMO were equipped with ROV friendly
flanges to aid in connection. In addition, buoyancy material was attached to the hose to make it neutrally buoyant so the
ROV could easily move the hose into place. A signal cable for the camera, lighting and pressure sensors was installed
between the SPM and the SMO. The cable was picked up at the SPM, by the ROV, and stabbed into a hot stab receptacle at
SMO. This cable can be seen coiled on the SPM in Figure 3. Figure 14 Flexible Steel Suction Hose After system deployment and SMO installation, drilling began. The bit was stabbed into the SMO and flow rate was set
at 800 to 1000 gpm (3030 to 3785 l/m). While pumping at 800 and 1000 gpm, the pumps were at 64.3 and 66.7% of full
operating power. This is a very good range for the RMR pumps to be operating in during regular operations.
During this first RMR deployment in the Gulf of Mexico, an inhibited water-based drilling fluid was used in combination
with the RMR. This drilling fluid was inhibited to control reactive shale formations.
When cementing the 28-in. casing, parts of the diverter tool on the inner string had been lost in-hole requiring a milling
operation while drilling out the 28-in. shoe. The milling operation was performed after the installation of the AGR RMR
equipment, while taking returns to surface. Whether the milling operation would have been performed without the AGR
equipment is difficult to answer, but what is obvious is that the milling operation would have required an additional 50 to
60,000 bbl of mud. The RMR equipment enabled the drilling contractor to monitor the process of the milling operation by
collecting and weighing the shavings from the fish.
The fish was successfully milled with returns to surface. Total weight of metal shavings in returns, collected on ditch
magnets and in the junk basket, was 59.5 lb (total weight of the lost diverter was 80 lb.)
An important feature of the RMR system is the ability to conduct a flow check. A flow check confirms that the well is
under control. While doing a flow check with the Krakatoa RMR setup, the rig pumps and the RMR pumps were stopped so
that the mud level inside the SMO could be held constant. If during the flow check mud is observed overflowing the SMO,
then the well is flowing and the well is taking a kick. Conversely, if the mud inside the SMO is dropping, then the well is
experiencing a lost circulation event. Flow checks are usually performed by monitoring the mud level inside the SMO for
approximately 15 minutes using the pressure sensor inside the SMO and visually by the mounted cameras on the SMO and
ROV. Flow checks were made periodically during the course of drilling. 10 OTC 20939 In one instance, while drilling the 26-in. hole section, the rig pump rate was 1100 gpm and the SPM rate was between
1400 and 1500 gpm, indicating a kick. Observations at the SMO indicated some gas associated with the shallow flow.
Drilling operations paused and 14 ppg kill mud was pumped, killing the flow. Another flow check was conducted to confirm
that no fluid gains or losses were being observed and that the well was under control.
Shortly after drilling commenced, the first water flow was encountered and the first gumbo attack was experienced. The
nature of the gumbo attack was very different from experiences in the North Sea or the Caspian Sea when drilling with the
RMR system. Figure 15 shows the gumbo exiting the SMO during the clearing procedure. To clear the SMO, the RMR
pumps were shut down and the rig mud pumps left on to push the gumbo from the SMO. Figure 15— Gumbo exiting the SMO The gumbo attack led to a blocking of the suction hose connecting the SMO and the SPM. The suction pressure increased
very rapidly causing the suction hose to collapse (Figure 16). The ROV disconnected the suction hose from the SPM and the
SMO and replaced it with a new section of hose. Figure 16— Collapsed suction hose During the gumbo attack, the RMR pressure sensor located on the subsea pump maxed out at 70 psi (5 bar). The hose
collapsed due to the outside pressure from the seawater and had to be replaced.
The average rate of penetration (ROP) when drilling the 26-in. section was 266 ft/day (81 m/day), with instantaneous
ROPs of ~90 ft/hr (27 m/hr). Caution due to shallow hazards and flow checks reduced the ROP. The true vertical depth was
5,327 ft (1625 m) when the 26-in. hole section was completed. A total of 1,852 ft ( 564 m) of 26-in. hole was drilled in 167.5
hours. The 22-in. casing was set at 3,148 ft (960 m) below mud line. This is among the deepest setting depths for the 22-in.
casing in the area.
A leak off test could not be performed on the 28-in. shoe, due to shallow unconsolidated sands with little integrity below
the shoe, which prevented drilling with the maximum mud weight of 13.4 ppg at the beginning of RMR operations. For this
reason, a conventional riserless weight up plan was selected to maintain a theoretical minimum overbalance of 65 psi (4.5 OTC 20939 11 bar) throughout the section without triggering a loss situation. The plan was to drill the 26-in. hole section with an inhibitive
WBM mud, with an initial mud weight of 11.0 ppg, with a progressive weight up to 13.4 ppg at TD.
The 26-in. hole section was drilled in the Mississippi Canyon area in the Gulf of Mexico with an inhibited water-based
mud (WBM) using the RMR system. Water-based drilling fluid was used to mill a fish and drill a total of 1,852 ft (564 m)of
soft laminated sand and shale sequences to the casing point. The exceptional inhibitive properties of the WBM were
demonstrated by good cuttings structure and integrity across the shakers while drilling the interval. After nine days of the
hole being open the 22-in. casing was run to bottom without any wellbore stability issues or problems. The RMR technology
with the appropriately engineered water-based fluid enabled the 22-in. casing to be set at one of the deepest casing points in
While drilling the 26-in. section, shallow hazards were encountered including a gas kick and water flows. The shallow
hazards were safely managed by weight-up of the drilling fluid and by the addition of properly sized lost circulation material
(LCM). When drilling the well, pore pressure was higher than expected, so the mud weight had to be increased more
agressively than what was originally planned. The final mud weight ended up higher than the predicted at 13.4 ppg (Figure
17). The mud weight was increased in four steps from 11.5 to 14.3 ppg, with an equivalent mud weight at TD of 11.84 ppg.
An influx of water from the shallow water flows was the likely culprit of the gumbo attacks. In one instance, a gumbo attack
caused collapse of the suction line leading to an influx of seawater which lessened inhibition of the drilling fluid. The diluted
WBM was displaced to the sea floor with new inhibited drilling fluid and to assist in the removal of the gumbo from the
SMO. A majority of the 8,000 bbl of WBM lost to the seafloor can be attributed to clearing the gumbo attacks from the
SMO. Figure 17— Planned vs. actual mud weights The exceptional benefits of the RMR system can be seen by direct comparison to the DKD method, shown in Table 3.
Without the additional benefits of a RMR system, the 1,852 ft (564 m) of 26-in. hole would have been an extreme challenge
to drill logistically with DKD. A conservative estimate to drill this 26-in. section of the well by DKD would have required
more than 77,000 bbl of DKD fluid and seven to eight supply boats to supply the rig with the necessary volume. Despite the
drilling challenges created by the shallow hazards, using RMR required at least 70% less WBM with 40% less washout of the
hole when compared to DKD. Drilling fluid volume and hole washout can be further reduced with some minor engineering 12 OTC 20939 modifications to the SMO to more efficiently clear the gumbo during an attack. Table 3 is a comparison of the drilling fluid
requirements and observed hole enlargement for the RMR system verses DKD.
Table 3—Drilling Fluid Requirements
Volume of Inhibitive WBM
(Total WBM used)
Volume of WBM Lost to Seafloor
Hole Washout (%)
60 Estimated DKD
(Minimum DKD required)
>100 Problem Areas
This operation went extremely well with minimal problems. Some minor problems such as a faulty air filter on the generator,
exhaust gases entering the office container and an under-sized flange receptacle on the pump were easily solved and did not
result in any lost time. There were two major problems during the well; these centered around the large amounts of gumbo
encountered while drilling the 26-in. hole section. It was very difficult to clear the gumbo from the well and SMO. In
previous operations in the North and Caspian Seas, the volume of gumbo was much less, allowing the MRL to U-tube back
through the SMO while running the rig mud pumps at a high rate flushing the gumbo from the well and SMO in these
operations. This technique did not work in this situation, and gumbo blocked the suction hose at the SMO leading to the
hose’s collapse. Not having a software alarm on the suction pressure exacerbated the problem.
Altering the shape of the SMO and improving the sensor detection of the suction pressure can help prevent this in the
future. In addition, an improved suction hose with a higher-pressure limit could also improve the situation; however, this
comes at the price of stiffness and weight of the hose, both important factors in the ability of the ROV to handle the hose.
This will limit the changes in the hose that can be made.
Results and Conclusions
An RMR system was assembled for operation in the Gulf of Mexico and was deployed on a dynamically positioned vessel.
The system was used to recover the drilling fluid from the sea bed while the rig drilled a 26-in. section of the well. The
system worked throughout the drilling process and permitted operations that could not have been performed had mud
recovery not taken place. The use of RMR saved significant amounts of drilling fluid and time. One estimate shows that
almost three times the drilling fluid would be necessary without the RMR system. Mud recovery also permitted the use of
much heavier weight engineered drilling fluids than those conventionally used, allowing this section to be driven deeper than
in offset wells. The engineered drilling fluid permitted the control of shallow water and gas flows, as well as providing
excellent wellbore stability. The system permitted deeper completion of this well section.
These operations led to the following conclusions:
7. An RMR can improve overall well construction
An RMR over-the-side deployment on a dynamically positioned vessel is possible in the Gulf of Mexico.
The depth for an over-the-side deployment can be increased to 2030 ft (620 m)
A four-stage pump model with two doubled-ended 800 hp (600 kW) motors work very well in an over- the-side
Using an engineered drilling fluid enhances drilling operations and provides excellent wellbore integrity.
Earlier detection of shallow hazards would have prevented many of the challenges experienced with this 26-in. hole
section, including kicks and gumbo attacks.
RMR allows for the use of customized engineered drilling fluids that would not be used in pump and dump
operations. OTC 20939 13 Acknowledgements
The authors would like to thank Statoil, Transocean Inc., Baker Hughes, and AGR Subsea Inc. for their support during this
operation and permission to publish these results.
Johnson, M. B., Rowden, M., 2001. Riserless Drilling Technique Saves Time and Money by Reducing Logistics and Maximizing Borehole
Stability. Paper SPE 71752 presented at the SPE Annual Technical Conference and Exhibition, New Orleans, Louisiana, 30
September -3 October.
Pelletier, J.P., Ostermeier, R.M., Winker, C.D., Nicholson, J.W., Rambow, F.H., 1999. Shallow Water Flow Sands in the Deepwater Gulf
of Mexico: Some Recent Shell Experience. Paper presented at the International forum on Shallow Water Flows held in League City,
Texas, 6-8 Oct.
Rocha, L.A. and Bougoyne, A.T., 1994. A New Simple Method to Estimate Fracture Pressure Gradient. Paper SPE 28710 presented at the
SPE Conference and Exhibition, Veracruz, Mexico, 10-13 Oct,.
Smith, D., Winters, W., Tarr B., Ziegler, R., Riza, I., Faisal, M., 2010. Deepwater Riserless Mud Return System for Dual Gradient Tophole
Drilling. Paper SPE/IADC 130308 presented at the Managed Pressure Drilling and Underbalanced Operations Conference and
Exhibition, Kuala, Lumpur, Malaysia, 24-25 February.
Stave, R., Farestveit, R., Hoyland, S., Rochmann, P., Rolland, N. 2005. Demonstration and Qualification of a Riserless Dual Gradient
System. Paper OTC 17665 presented at the Offshore Technology Conference, Houston, TX, 2-5 May. ...
View Full Document
This note was uploaded on 11/23/2010 for the course PETE 4XX taught by Professor Mehmetcebeci during the Spring '10 term at Middle East Technical University.
- Spring '10