SPE/IADC-184683-MS

Horizontal Drilling with Dual Channel Drill Pipe

O. M. Vestavik and J. Thorogood, Reelwell; E. Bourdelet, Total; B. Schmalhorst and J. P. Roed, DEA Deutsche
Erdoel AG

Copyright 2017, SPE/IADC Drilling Conference and Exhibition

This paper was prepared for presentation at the SPE/IADC Drilling Conference and Exhibition held in The Hague, The Netherlands, 14–16 March 2017.

This paper was selected for presentation by an SPE/IADC 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 Society of Petroleum Engineers or the International Association of Drilling Contractors and are subject to correction
by the author(s). The material does not necessarily reflect any position of the Society of Petroleum Engineers or the International Association of Drilling Contractors,
its officers, or members. Electronic reproduction, distribution, or storage of any part of this paper without the written consent of the Society of Petroleum Engineers or
the International Association of Drilling Contractors 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 SPE/IADC copyright.

Abstract
Drilling of long horizontal wells is subjected to challenges, such as hole cleaning, drill string torque and drag,
and challenges related to the Equivalent Circulating Density (ECD). Where the window for the downhole
pressure is small, the length of the open hole section is limited by the dynamic ECD. The result is that
several liners or casing strings may be required to reach the target depth.
The Reelwell Drilling Method provides a new solution for above mentioned challenges, incorporating
a dual drill string, with a separate channel for the return fluid from the well. This arrangement enables
Managed Gradient Drilling, i.e. to drill with a constant downhole pressure gradient that can be controlled
to be nearly independent of the flow rate. The solution is similar to Managed Pressure Drilling, however,
differs in that the downhole pressure gradient is managed instead of the pressure at one depth in the well.
In addition, this new method of ECD control provides efficient hole cleaning at low flow rates, and enables
torque and drag reduction due to buoyancy of the drill string.
A shallow horizontal trial well was drilled onshore Alberta, Canada, March 2016, to verify the new
technology. New drilling equipment, including a newly designed aluminum dual drill string was used.
New drilling procedures were employed, including the new Heavy Over Light drilling fluid solution. The
equipment and procedures successfully verified the key benefits of the technology:
– Managed Gradient Drilling with ECD eliminated by near static well annulus fluid.
– Torque and drag reduction resulting from the drill string buoyancy effect.
– Efficient hole cleaning at low flow rate, due to return flow inside the dual drill string.
In the following sections, we describe the technology, how the field trial was organized and present
drilling data from the shallow horizontal drilling operation that confirmed the capabilities of the system.

Introduction
The technology for drilling and completion of long horizontal sections has improved significantly in the
recent years. However, there are still challenges, such as hole cleaning, ECD control, torque and drag
reduction. ECD is linked to hole cleaning, and is often a limiting factor. Long reach horizontal wells can
also be limited by the torque and drag of the drill string.
2 SPE/IADC-184683-MS

The dual drillstring configuration has specifically been developed to provide solutions to the above
challenges as follows:
– Hole cleaning: drill cuttings are removed from the bottom hole through a separate flow channel inside
the drill string.
– ECD control: Managed Gradient Drilling creates a constant downhole pressure gradient that is
independent of the flow rate.
– Torque and Drag: the Heavy Over Light fluid solution reduces torque and drag due to buoyancy.
Figure 1 presents a schematic of the drilling fluid flow. It differs from conventional drilling by utilizing a
dual drill string. The dual drillstring allows drilling mud to be pumped down the annulus inside the drillpipe,
with the return flow up the concentric inner pipe. The system can be installed on a standard drilling rig.
A top drive adapter connects the rig’s top drive to the top of the dual drillstring. It contains a swivel that
allows for pumping mud into the drillpipe annulus and for the return flow from the inner pipe to be directed
back to the surface mud system. The lower end of the dual drillstring connects to a standard Bottom Hole
Assembly by means of an Inner Pipe Valve, which contains entrance ports for the return fluid, and isolates
the well during pipe connections.

Figure 1—Schematic of the Reelwell Drilling Method (RDM)

As shown in figure 1, a flow control unit is a key element to control the drilling fluid flow. The flow
control unit is a skid on which are mounted chokes, valves and sensors for pressure, temeperature and flow.
The flow control unit is normally positioned close to the rig, and is remotely controlled from a screen unit
in the driller’s cabin. The flow control unit regulates the flow into and out of the well, in order to keep a
near constant downhole pressure gradient during the drilling operations. A rotating control device at the top
of the BOP is normally used for the well pressure control during operation.
The dual drillstring configuration allows two different types of fluid to be used in the well during the
drilling operation. A low density and low viscosity active drilling fluid is circulated inside the dual drill
string and around the BHA is optimized for efficient claning capability and ECD control. The annulus
outside the dual drill string is filled with a near static passive fluid preferably a fluid optimized for torque
and drag reduction and for downhole pressure control.
SPE/IADC-184683-MS 3

This heavy over light configuration results in the passive well annulus fluid having a greater density than
the active fluid inside the dual drill string. When using the heavy over light solution, the dual drill string is
exposed to positive buoyancy forces due to the differences in fluid densities inside and outside the string.
The buoyancy forces result in a reduction of wall contact forces, thereby leading to a reduction in the torque
and drag during drilling. Additionally, the heavy over light solution means that surface back pressure on
the well annulus can be kept low during operation, enabling Managed Gradient Drilling operations to be
performed with little or no wellhead casing pressure.

Goal for the drilling trial

An onshore wellsite in Alberta, Canada, was selected for a verification of the new technology. The main
goals of the trial well were as follows:

• Verify tools and procedures:

– Reliable tool performance including a modified design of aluminum Dual Drill String.
– Operating procedures, including well control.
– Validate the heavy over light concept in reducing annulus backpressure, torque and drag
– System’s capability to provide a constant downhole pressure gradient

• Verify predictions of modelling software:

– Drilling engineering software for hydraulics, torque and drag calculations.
– Simulator for operation training and procedure development.

Operation
Figure 2 presents the planned well trajectory. The target for the horizontal well was a zone between about
450 – 470 m TVD predicted to be a relatively homogenous sand with no hydrocarbons and therefore suitable
for the trial well.

Figure 2—Planned well trajectory

Figure 3 shows the drill site and the rig adapted to accommodate the Reelwell equipment used for drilling
the well. The rig was capable of handling range 3 drill pipes.
– Fig 3, left: Overview of rig and well site in Alberta, Canada.
– Fig 3, right: Installation of the Top Drive Adapter showing the separate swivel and hoses.
4 SPE/IADC-184683-MS

Figure 3—Drilling Rig

Figure 4 shows the installation of the flow control unit and its control panel in the driller’s cabin. The unit
enables remote control of system functions, data logging and sensor monitoring. The control unit software
provides automated procedures for pump startup, shut-down, pressure control on selected set points and
alarms with associated recommended actions.
– Fig 4, left: Flow control unit in inside container, close to the rig.
– Fig 4, right: Control panel for the flow control unit positioned in the driller’s cabin.

Figure 4—Flow Control Unit and the human machine interface

Figure 5 presents the two different types of dual channel drill pipes used during the operation.
– Left: 6 5/8" steel dual drill pipes, mainly for use in the vertical section of the well.
– Right: The new aluminium dual drill pipes, mainly for use in the horizontal section of the well.
SPE/IADC-184683-MS 5

Figure 5—Dual drill pipes

Figure 6 presents the tool joints of the dual channel drill pipes.
– Left: Box end of the dual drill pipe, containing the inner pipe pin and sealing element.
– Right: Pin end of the dual drill pipe, containing the inner pipe box end for the stab in connector.

Figure 6—Tool joints of the dual drill pipes

The intermediate 13.5" hole section was directionally drilled with conventional 5" drill pipe down to
803m MD, 465 m TVD. The 10¾" casing was cemented in place and a formation integrity tested.
The specialized dual drillstring combined with a conventional rotary steerable system bottomhole
assembly was used for drilling the 9 7/8" horizontal section. A 1.1 sg water-based drilling fluid was used
for both active and passive fluids down to 1100 m MD. The fluid in the annulus outside the dual drillstring
was then displaced to a 1.6 sg density fluid to continue drilling the horizontal section in heavy-over light
mode. The remaining section to the target 1510 m MD, 452 m TVD was drilled in this mode with active/
passive fluid densities maintained at 1.1 sg/1.6 sg. The effect of the heavy over light fluid configuration
on torque and drag was measured at TD. Torque and drag readings were obtained first with the 1.1sg/1.6sg
light/heavy fluid configuration. The well was then displaced to 1.6 sg, in the well annulus and both channels
of the drillstring, and the measurements were repeated.
6 SPE/IADC-184683-MS

The well control procedures with the dual drillstgring were tested as a last part of the trial. After setting
a bridge plug in the 10 ¾" casing at 783 m, nitrogen was displaced through the dual drill string and trapped
in the casing at the plug, to create a volume of 430 l of compressed gas. During the subsequent circulation,
the nitrogen was successfully detected and circulated out of the well, in accordance with the agreed well
control procedure. The well was then abandoned in accordance with local regulations.

Results
Figure 7 presents the time depth plot of the operation. The dashed black line shows the plan, prior to
operation and the red line the actual time taken. The green line represents the actual time, when subtracting
downtime from various problems with the rig. The rig had been cold-stacked prior to the operation and
there were some problems with the electrical systems, as well as with the pumps and mud handling system
that were not identified prior to start-up.

Figure 7—Time depth plot of the operation

Target formation at around 460m TVD was clay rather than sand, which caused problems with mud rings
and associated delays of the upper 13 ½" hole section down to 804m MD, 466m TVD. As can be seen in
the figure, the operation followed the plan closely after removing the rig-related downtime.
SPE/IADC-184683-MS 7

Figure 8, left hand plots, presents an example of surface and downhole data when drilling from 867 –
905 m MD in the single fluid mode, with the passive fluid in the well annulus outside the dual drillstring
having the same density as the active fluid circulating inside the dual drillstring. The following points can
be seen in the figure:
– The ROP is around 30 m/h for the first two joints. The ROP increases to around 40 m/h for the third
joint at 905 m MD. The increase probably due to the higher WOB for the third joint.
– The measurements from the PWD tool shows bottom hole pressure maintained within ±5 bar.

Figure 8—Example of surface and downhole logging results

Figure 8, right hand plots, gives the same surface and downhole logging results when drilling from 1246
– 1283 m MD in the heavy over light. The rotational speed had been increased to more than 100 rpm, the
ROP is more irregular than in the previous figure and ROP values up to 50 m/h can be observed.
Figure 9 presents results from the torque and hookload measurements when investigating the effect of
the heavy over light fluid configuration on the torque and drag of the drill string. The measurements were
performed with the bit off bottom after completing drilling of the horizontal section to 1510m MD. The
measurements were performed by slowly running into and out of hole, with and without drill string rotation.
1. The left hand side of the figure presents the measurements performed when the well annulus contains
a heavy fluid with a density of 1.60 sg and both the inside channels of the drillstring are filled with
the light drilling fluid of density 1.15 sg.
2. The right hand side of the figure presents the measurements performed when the well annulus contains
a heavy fluid with a density of 1.60 sg and both the inside channels of the drillstring are also displaced
to the heavy drilling fluid of density 1.60 sg.
8 SPE/IADC-184683-MS

Figure 9—Measurements to verify the effect of heavy over light, for torque and drag reduction

As can be observed from the figure the torque values (red) during rotation on the left are, on average
6.4 kNm, whereas the corresponding torque reading on the right is averaged to 8.7 kNm. This confirms
that the torque of the drill string in this case is reduced by around 30%, which is in agreement with the
model. Precise measurements of the effect of the heavy over light configuration on drag are more difficult
to interpret, due to the low overall loads and the hysteresis effects of by other friction elements in the rig’s
hoisting system rig. The figure indicates higher drag for the single fluid case, observed by the step changes
in hook-load (blue) values when comparing the heavy over light and single fluid case.
Figure 10 shows some of the larger cuttings found at the shaker during the drilling operation. Since the
formation was predominantly clay, the bulk of the cuttings were very small, less than 1 mm in size. However,
when drilling occasional stringers, some larger cuttings size were observed. The bottoms-up time for the
cuttings was estimated to be about 9 minutes at TD.

Figure 10—Example of large cuttings examples observed

SPE/IADC-184683-MS 9

Figure 11 presents a log of the density of the return fluid, measured by the Coriolis flow meter at the flow
control unit during the operation. As can be observed the density in the return fluid is increasing throughout
the operation, due to accumulation of fines from the drilled clay. The reason for the increase was that the
rig’s centrifuges and solids control equipment had limited capacity to remove the fines. The occasional
drops in the density were caused by dilution of the fluid. It also can be observed that there is regular small
increase in the fluid density as each joint is drilled, as expected due to the cuttings.

Figure 11—Density of the return fluid during the operation

Figure 12 presents the results from one of the well control tests when circulating nitrogen from the well.
About 0.43 m3 downhole volume of nitrogen was injected into the well and trapped at a bridge plug in the
cased hole at 783 m MD. A short bottomhole assembly was used when running the drill string into the
downhole gas volume while simultaneously circulating at 800 lpm at standard dual channel circulation. The
following main events are identified in the figure:

• 100s: Started running in

• 140s: Driller opened the BOP annular briefly by mistake, resulting in a temporary loss of well
pressure.
• 230s: Bit on bottom.

• 380s: Flow out is rapidly increasing

• 420s: Flow out 1300lpm. Well shut in.

10 SPE/IADC-184683-MS

• 2720s: Pressures regarded stable. Circulation started to clean inner string.

• 2900s: Pressure stable at circulation target values, drill string is clean.

• 3800s: Normal circulation restarted after one bottoms-up, without indication of further gas present.

Figure 12—Example of pressure and flow measurements recorded during the well control trial

The test confirmed the ability of the flow control unit Coriolis flowmeter to quickly detect gas in the
well. The test also confirmed the ability of the system to monitor downhole pressures in the well annulus
and to circulate out the gas in a controlled and safe manner. The second test gave similar results.
Figure 13 presents an example of the calculated pressure gradient in the trial well during the heavy
over light drilling operation with the pressure inside the well annulus and in the two channels of the dual
drillstring as a function of vertical depth. The following points can be observed:
– The solid blue line represents the pressure in the well annulus, created by the near static heavy fluid,
in this case with zero wellhead pressure.
– The red and brown line indicate the upper and lower pressure limits in the well, here as an example
of a small pressure window.
– The green line represents the pressure inside the return flow path up the drillstring. While the active
circulating fluid is subjected to ECD to make it flow up the drillstring, its pressure is lower than that
of the passive fluid in the well annulus. However, the pressures of the two fluids are equal at the
entrance ports of the inner pipe valve at the top of the BHA.
SPE/IADC-184683-MS 11

– The violet line to the right of the figure is the pressure in the annulus of the dual drillstring, the gap
to the green line is the pressure drop at the bit. The standpipe pressure is slightly above 100 bar, as
observed at 0 m TVD.

Figure 13—Pressure gradients when drilling the trial well using the dual channel circulation

The pressure gradient in the well annulus is a constant straight line because heavy fluid in the annulus is
almost static. The pressure gradient of the light active drilling fluid inside the return flow conduit is not a
straight line, due to the hydraulic friction inside the inner string. This example shows the advantage of the
dual string configuration over conventional circulation, where the hydraulic friction is in the well annulus
creates additional bottomhole pressure.

Discussion
The Time Depth Plot in figure 7 show that the drilling operation from spud to abandonment was completed
in 15 days, which including 1 day contingency was according to plan. The non-productive time on the rig
was caused by electrical problems, some rig adaptations not foreseen in the preparations, rectification of
a misalignment of the mast and some operational problems with the mud pumps and the mud handling
equipment during operation. There was no downtime on the downhole drilling tools and equipment during
the operations. All the planned activities were conducted as per the program.
The surface and downhole logging results in figures 8 and 9 show that high ROP was obtained for
relatively low flow rates, when compared to conventional drilling, which demonstrates the capability of the
dual drillpipe system to effectively drill and clean the hole. The results indicate that the downhole pressure
gradient is held constant to within 50 kPa during starting and stopping the mud pumps. Detailed analysis
of the downhole pressures indicated that the pressure variations during connections were in the range of 5
bar, when using start-stop times for the pump of 1-2 minutes when employing automated mud pump control
procedures. If necessary, smaller pressure variations can be achieved by taking longer to ramp up from
pump start-up to full circulation flow.
From figure 8, left hand plots, it can be noted that the time for a connection, from stopping drilling to
starting drilling, is around 20 minutes. In figure 8, right hand plots, the same connection time is reduced to
12 SPE/IADC-184683-MS

about 10 minutes. This can be observed from the graphs to be a result of omitting the short washing and
reaming operation at the end of each stand. The reaming operation is regarded as standard in conventional
drilling, and was therefore performed in the first part of the well but were later omitted to improve drilling
efficiency, since the reaming showed no noticeable effect on the hole cleaning.
The torque and drag measurements in figure 9 show that the torque of the drill string was reduced by
around 30%, which is in agreement with the model. The reduction in torque with the heavy over light
configuration seems to be propotional to the reduction of the effective weight of the pipe when comparing
pipe weights when containing the two different fluids. This observation agrees with basic friction theory.
The effect of heavy over light HOL was difficult to interpret, due to the disturbance caused by other friction
elements in the hoisting system of the rig. However, the effect can be interpreted when comparing the drop
in hookload from 5 kdaN at time 08:17 for heavy over light to 8 kdaN at time 10:25 for the single fluid.
The cuttings examples in figure 10 show the large cuttings observed at the shale shaker when drilling of
hard stringers when approaching the final well depth. Normally the cuttings were very fine and had particle
size less than 1 mm. The observed large cuttings may have been created as a result of bit vibrations when
drilling the hard stringers, due to relatively large bit cutter elements. It is interesting to note that the bottoms-
up transport time for the cuttings is estimated to be 9 minutes at TD. Hence, the large cuttings may be very
valuable for formation evaluation while drilling. On this point the dual drillstring differs from conventional
drilling, where much of the information from the cuttings is lost due to long transport times, and the grinding
and mixing during transport in the well annulus.
The measurements in figure 11 show a steady increase of the drilling fluid density, as a result of fines
accumulating during the drilling process. The solids control equipment on the rig had limited capacity to
remove the fines, which accumulated due to the formation clay and the high penetration rates. The active
fluid had therefore a slightly higher density than originally planned, the occasional drops in the density is
related to dilution of the active fluid. The small increase in the fluid density for every joint being drilled,
clearly show the effect of the cuttings on the return flow density, and also proves the efficient cuttings
transport capability. Laboratory analysis of the mud samples confirmed that the build up of fluid density
was due to drilled solids, and not due to intermixing of barite from the heavy fluid, when performing the
heavy over light operation in the last part of the well.
The pressure and flow measurement recordings during well control trial in figure 12 demonstrate the
ability to quickly detect gas in the well with the Coriolis flow meter. The ability to monitor the flow and
downhole pressures using the static conduit in the dual drillstring was also verified. The ability to circulate
out gas in a controlled and safe manner was proven by the two well control trials performed.
The pressure gradient plot example in figure 13 shows the principal difference between the dual drillpipe
circulating system, effectively Managed Gradient Drilling, and conventional Managed Pressure Drilling.
Conventional Managed Pressure Drilling imposes a dynamic pressure gradient on the well annulus that may
require installation of a casing or liner when the dynamic pressure exceeds the downhole pressure window.
This Managed Gradient Drilling, enabled by the dual conduit drillstring, can, by adjustment of flow rates
and surface choking, provide a constant pressure gradient at all times, independent of whether the pumps
are running or not. Managed Gradient Drilling is thus capable of overcoming section length limitations
associated with conventional Managed Pressure Drilling.

Conclusions
The main results from the drilling operations were:

• All relevant procedures were verified, including well control

• Model for hydraulics and reduced torque & drag verified

SPE/IADC-184683-MS 13

• After implementing the heavy over light fluid configuration with passive fluid of 1.6 sg in the well
annulus and active fluid 1.1 sg inside the dual drillstring:
– Wellhead pressure reduced from 25 bar to 3 bar
– String torque reduced from 8.4 kNm to 6.2 kNm

• Managed Gradient Drilling eliminated the dynamic ECD in open hole.

• Demonstrated effective hole cleaning at low flow, 800 – 1000 lpm.

• Formation evaluation from cuttings enhanced with bottoms up time at TD of 9 minutes.

• 100 % up-time of the dual drillpipe, downhole tools and surface equipment, no HSE incidents.

• Successful integration of system with third party RSS & MWD & LWD tools.

• Rate of Penetration typically 20 - 40 m/hr.

• Operation performed according to plan and budget, duration 15 days.

By validating the theoretical models the field trial results demonstrate how the dual drill string method
can be used to solve key torque, drag, hole cleaning and ECD control problems in horizontal wells. A novel
solution for Managed Gradient Drilling is presented, which can enable increased casing setting depth in
zones with weak formations.

Acknowledgement
The drilling trial and development of the above described solutions was supported by DEA Deutsche
Erdoel AG, Total, Petrobras, the Research Council of Norway and by the service partners Halliburton
and Alcoa Oil and Gas. Thanks to the above mentioned parties for the support and for the permission to
publish the results of the work. Special thanks also to Barlon Engineering, Calgary, for supporting the
engineering and operational management, to Ensign Drilling for their proactive attitude and providing the
rig and experienced operation personnel and to Twin Butte Energy for providing access to the field and for
administering the permissions to permit the operation.

References
1. Vestavik et al. "Reelwell Drilling Method", SPE/IADC 119491, Amsterdem, March 2009.
2. Vestavik et al. "RDM- A Unique Combination of MPD and Liner Drilling", SPE 124891,
Aberdeen, Sept. 2009.
3. Vestavik et al. "New Approach to Improve the Horizontal Drilling Reach", SPE 137821, Calgary,
Oct 2010.
4. Belarde M et al., "Deployment of RDM in a Shale Gas Field in Canada", SPE 145599, Aberdeen,
Sept. 2011
5. Vestavik et al. "Extended Reach Drilling – new solution with a unique potential", SPE 163463,
Amsterdam, Mar 2013.
6. "Extending Innovation", European Oil and Gas magazine, pp 30-31, June 2014
7. Vestavik O, "Cut, reach, clean and analyse", Oilfield Technology pp 31-34, July 2014.
8. Francis D, "ERD systems address downhole challenges", Offshore Magazine pp 54-57 Sept 2014.
9. Vestavik O, "Dual Drill String Technology improves ERD", American Oil and Gas Reporter, pp
51-57, Sept 2014.
10. Aleksandersen J et al. "Dual Drillpipe Method shows success in PMCD wells", JPT, pp 32-35,
Apr 2015.
14 SPE/IADC-184683-MS

11. Vestavik O, "Conveyance in the Conduit", Oilfield Technology pp 31-34, May 2015.
12. Vestavik O. "Managed Gradient Drilling for improved horizontal reach.", Barents Drilling, ONS,
Stavanger, Aug, 2016.
13. Vestavik O, et al., " Experience from Horizontal Drilling using Dual Drill Pipe", GeoEnergy-
Celle, Germany, Sept, 2016.
14. Stranden E. "RDM – Experience ERD Trial ", NPF- Drilling Conference, Kr.sand, Norway, Sept,
2016.