Академический Документы
Профессиональный Документы
Культура Документы
Figure 5 (below): The two architectures defined at the feasibility study. On the left is
The two architectures
architecture 1 retained as primary. On the right is architecture 2 retained as contingent.
In order to tackle this drilling challenge,
Contingency the 2004 feasibility study identified two
First option
Borehole Strengthening
Run expandable -
Penetrate reservoir different architectures. They are shown
then reservoir at low
MW.
below frac P - Drilling
liner - reservoir at low in Figure 5.
MW
7" TOL @ 4 1/2" TOL The first one involves the use of a spe-
5040m @ 5015m
cially designed mud loaded preventively
6-5/8" TOL
@ 5065m with specific loss circulation materials
9 7/8" @ Exp TOL @
5115 m
(LCM). The technique is known as “bore-
5141m 9 7/8" @
5141m Mw = 2.00 to hole strengthening.” Once reservoir has
2.15 sg
Mw = 1.86 sg 8-½" Drilling
been penetrated, the RPTZ is cased off
Cap Rock
4-1/2" TOL
The second one involves the use of an
@ 5400 m Top reservoir
@ 5433 m
Mw = 1.86 sg Liner drilling
7.375" drill shoe
expandable liner to cover most of the
7" @ 5453m +30 / -45 m 6-5/8" @
5453m cap rock. The RTPZ is then drilled with
a lower mud weight below the FPG. The
Reservoir
5-5/8" Drilling
5-5/8" Drilling
4 1/2" liner
4 1/2" liner
remaining opened cap rock would be
Mw = 1.15 to
1.35 sg short enough so that borehole instabili-
4 1/2" @
5675m
4 1/2" @
5675m ties can be managed. The RPTZ is then
4 1/2" liner 18.9 ppf 4 1/2" liner covered with a cemented 6 5/8-in. flush
25%Cr set in 7"
production liner
25%Cr to cover 6-5/8"
intermediate liner and connection drilling liner. Once the reser-
expandable
voir has been drilled, the 6 5/8-in. liner is
covered by a 4 ½-in. production liner 18.9
The other information requested from c) Top reservoir depth uncertainty ppf, 25% Cr with premium connections.
rock mechanics was the borehole stabil-
Depending on the architecture selected, Once RPTZ is covered, the mud weight
ity of the cap rock. This allows definition
the top reservoir depth prediction was can be decreased to a value just enough
of the minimum mud weight one can
critical to maximize the success of the to safely drill the reservoir.
use to cross the RPTZ without suffering
RPTZ drilling. Extensive geophysical
unmanageable borehole instability of the As the main difficulty was lying across
techniques were used to minimize the
open hole above. The amount of informa- the RPTZ crossing, the top architecture
uncertainty attached to this test predic-
tion on the cap rock is even more limited of the well was taken as a standard
tion. These included, among others,
than on the reservoir. The uncertainty Elgin/Franklin architecture as designed
thorough examination of the seismic
on the defined value is consequently for virgin pressure. Production casing
data and uncertainty studies on depth
higher. set at the top of the cap rock (above
conversion. The prediction was given to
PTZ) is 10 ¾-in. and 9 7/8-in. casing. This
+30m / -45 m.
allowed focus on the RPTZ crossing and Figure 6 shows efficiency of plugging at one was planned to be installed in front
gave a comfort factor, because this top various concentrations. of the previous casing to act as a top
architecture is designed to hold well seal and hanger mean.
full of gas at virgin reservoir pressure, Expandable liner
regardless of circumstances. Expandable liners are becoming more Implementation Strategy
Before being implemented, both architec- widely used in the industry. They are Primary and back-up architecture
tures needed further development work, usually used to seal off weak zones to
thorough testing, qualification and engi- allow for increasing the mud weight to Out of the two architectures, it was
neering calculations. The main ones are penetrate deeper zones at higher pres- decided to implement architecture 1 as
detailed in the next section. sure. This makes them work under a the primary. The main reasons were:
burst mode. In our case, the expandable
liner had to work in a collapse mode. • In case of failure of the borehole-
Special Developments Expanded pipe has low collapse capac- strengthening method, it allowed plug-
Specially designed mud ity, and this is one of the limitations of ging of the hole with a cement plug
the technique. The standard off-the-shelf and side track from the 9 7/8-in. shoe,
The borehole-strengthening technique expandable for this application was starting with a fresh hole to implement
was already used in the industry. The 75/8-in. x 9 5/8-in.- 29.7 ppf, with a claimed the contingent architecture.
principle is to create a fracture by using collapse capacity of 218 bar. To increase
a mud weight higher than the FIP and • Also, in case of failure of the borehole-
the collapse capacity to the required strengthening method when reaching
plug it on creation, to prevent its further value, a development programme was
development. The plug is formed by the reservoir, the depth of the top
undertaken with the selected provider. reservoir would then be accurately
LCM continuously present in the mud.
The created fracture increases the rock A heavier expandable pipe of 38 ppf known, and the placement of the
stress locally, enhancing the hole’s ability (0.5-in. wall thickness) was selected, and expandable liner would be optimized.
to support high mud weight. The process the associated connection was designed. • It was quicker.
is easier when filtration exists behind An expansion test of the system was car-
the plug in the fracture, because the ried out in the provider facility during Architecture 2 was kept as contingent
pressure inside the fracture decreases summer 2005, including expansion of an and preparatory work was done to make
and allows it to close back on the plug. elastomer-bonded section inside a joint it ready for deployment.
It is therefore more commonly used with of base pipe. This test proved that such
water-based mud, which exhibit higher thick pipe and connections could be suc- Two situations were identified when
filtration values. The high temperatures cessfully expanded. The expanded pipe the contingent architecture would be
experienced on Elgin/Franklin dictate was then cut into sections and pressure- deployed prior to entering the reservoir:
the use of oil-based mud. As filtration of tested to burst, collapse and tension con- • If high-pressure gas-bearing limestone
oil-based mud is very tight, the filtration ditions. The minimum collapse pressure layers were encountered in the cap
from the fracture faces into the formation was 366 bar. So the operator confidently rock. This would require a high mud
is virtually nil. The consequence is that considered the system capable of sus- weight to be controlled safely during
the plug at the fracture mouth needs to taining 345 bar collapse pressure. a trip, and therefore the probability of
be tightly sealing as soon as it is created. success of the borehole-strengthening
Although expandable liner has been run
The other difficulty is to define the width previously in high-temperature wells, the method would be lowered to an unac-
of the fracture to seal. Rock mechan- operator thought it prudent to test the ceptable value.
ics calculations show that there is a expansion system with high mud weight • If the defined mud weight was found
direct relationship between geometry of and high temperatures. This was done too low to manage borehole instability
the fracture, width and length, and the in a shallow well in a special facility in in the cap rock, necessitating again an
amount of overpressure. The higher the Dallas, Texas. The system was left soak- increase in the mud weight and com-
pressure, the wider the fracture, at a ing with 2.15 sg oil-based mud at 176°C promise the borehole-strengthening
given length. For a given overpressure, for 18 hr before initiating expansion and technique success.
the equation has two unknowns, so one expanding 18 m of pipe. Examination of
has to fix a parameter. In our case, it was the seals and parts of the system after- Mud weight definition
estimated that the mud could be designed wards showed no significant degrada-
to form an efficient plug of 1-mm width, tion. The system was declared qualified The mud weight necessary to implement
while keeping reasonable rheological for the application. the primary architecture had to be a com-
properties despite a high solid content. promise: Using a low mud weight maxi-
It was decided to not cement the expand- mizes the chances for a successful bore-
The mud was therefore designed and able liner on deployment in the well. The hole strengthening, but it increases the
tested to be able to seal a 1-mm gap. The volumes of cement involved were too chances of borehole instabilities in the cap
LCM additives consist of sized grounded small to guarantee good-quality cement rock. Although instabilities can be man-
marble and sized graphitic material. at 5,000 m depth. Any problem encoun- aged to a certain level, there is a thresh-
The graphitic material comprises some tered during the expansion process old from which they cannot be managed
very coarse sections. Laboratory testing would have left the cement setting and without raising the mud weight. Using
took place to adjust LCM additives rela- prevented the expansion process to take high mud weight reverses the problems.
tive concentrations, to achieve a quick place. Instead, some joints of the liner
and efficient plugging of 1-mm slots. As were bonded with elastomer. Some were Given the uncertainties on the rock
well as these in house tests, tests were planned to be installed in the open hole mechanic properties in the cap rock and
performed in a third-party laboratory. to seal against the formation. Another in the reservoir, a probabilistic approach
was taken to define a theoretical opti- practice to deal with losses is to pump If losses happened, the first thing we
mum mud weight. For each mud weight, LCM pills with increasing concentration wanted to achieve was to pull back the
a probability of success of the borehole- of LCM material, and if that’s not suc- BHA into the previous shoe, set approxi-
strengthening technique was defined, cessful, keep the hole filled with mud and mately 300 m above, in order to remove
as well as a probability of managing the then revert to a lighter fluid, firstly base it from the cap rock formation most
borehole instability. The two curves were oil then water, when available volume of susceptible to collapsing under loss of
plotted on the same graph (Figure 7). The mud has been used. hydrostatic. The level in the well would
mud weight value at which they cross is be allowed to drop. The potential small
the optimum mud weight. In our case, a This was not applicable in this case: gas ingress from the limestone layers in
1.86 sg mud was deemed optimum, yet the mud is already largely charged with the wellbore would be kept at bottom by
the minimum mud weight the cap rock LCM material; therefore chances of pumping mud at a low rate of 100 l/min.
had been drilled so far was 2.15 sg. success with LCM pills would be low.
Additionally, the potential for collapsing As soon as the BHA is safely pulled
Drilled in 6 5/
8-in. liner of cap rock around the BHA (bottomhole to shoe, the level in the well would be
assembly), as well as the potential for assessed by the use of an echo meter.
The liner drilling technique was becom- small gas layers to flow into the bore- This would allow the definition of the
ing available at the time development hole, under a decrease of the hydrostatic mud weight required to control losses.
work was ongoing. This technique column, were to be considered. The well would then be displaced to the
presented two main advantages. First, required mud weight, and a cement plug
there is no need to trip out of hole to The expected rate of losses also set at bottom to abandon the bottom of
run the liner, leaving the cap rock in an had to be taken into consideration: the hole and revert to the contingent
underbalance condition for a long time. Geomechanical calculations indicated architecture. This was the ideal case.
On these wells, it is typically four days that, given the amount of expected over-
balance, any fracture opening would However, so much uncertainty was lying in
between when the bit is picked off bot-
propagate so quickly that very little time the expected pressures and well behaviour
tom and when the liner arrives at bot-
would be allowed to prepare a pill. that all cases had to be considered care-
tom, applying HPHT procedures.
fully and a way of regaining full control of
Second, it allows using a higher mud Keeping the well full with lighter fluid, the well found for each permutation.
weight during drilling. If heavy losses typically water, once all available mud The benefits of the plan were explained
were experienced when entering the res- had been pumped into the well, would to the offshore crew, who then agreed
ervoir, and the hydrostatic applied on the have resulted with a mixture of incom- to the proposed losses response plan.
cap rock dropped, the formation may col- patible fluids in the wellbore. This would The fact that the top well architecture
lapse. With a drilled-in liner, it would col- have created more difficulties in regain- was able to sustain a well full of gas
lapse around the liner, leaving the hole ing control of the well. increased their confidence in applying it.
cased off. Isolation behind the liner might In our well, the chance of important gas As will be developed in the following sec-
be already achieved by the collapsed for- ingress following losses was considered tions, this losses response plan did not
mation. In case further isolation is need- low. In the reservoir itself, the fracture need to be applied. However, it had to be
ed, an integral liner top packer would propagation pressure was higher than revisited several times to incorporate the
seal off the top of the liner, and the shoe the pore pressure. Above it, the only new information acquired while drilling
could be tacked by a cement squeeze. source of gas ingress was the potential and to adapt it to each drilling situation.
Losses response plan small carbonate layers. They would bleed
gas into the wellbore rather than gener- Actual Drilling
The main risk associated with the prima- ate a genuine kick. Therefore, the usual
ry architecture was encountering total requirement of keeping the well full at Herring anomaly
losses when penetrating the reservoir any price was challenged very strongly. Top sections were drilled as planned.
with a high overbalance. The common The 13 3/8-in. casing was called 100 m
8069 Total: Depleted Well Testing; Test 1; Pressure Response 8069 Total: Depleted Well Testing; Test 3; Pressure Response
Temperature: 150 o C; Fracture Profile: 1mm - 0.2mm; Injection Pressure: 160 bar Temperature: 150°C; Fracture Profile: 1mm Parallel; Injection Pressure: 140 bar
0.45 180
160
0.40 160
140
0.35 140
120
0.30 120
Injection Pressure (bar)
100
Pressure (bar)
Pressure (bar)
0.25 100
80
0.20 80
60
0.15 60
40
0.10 40
20 0.05 20
0 0.00 0
0 100 200 300 400 500 600 700 800 900 1000 0 1000 2000 3000 4000 5000 6000 7000 8000
Time (seconds) Time (seconds)
Figure 6: Different mud formulations used for sealing a 1-mm gap. On the left, a plug forms but breaks out. On the right, the plug forms
and does not transmit any pressure past it.
350 400 450 500 550 600 650 700 750 800 850 900 950 1000 1050 1100 1150 1200 Once the expandable shoe was cleared,
-50 and the stability of the well assessed, the
6 5/8-in. drilled in liner was run. The well
0 was displaced to the specially designed
Top C mud loaded with LCM originally designed
50
for the borehole-strengthening technique.
Case uniform F3
The running string was spaced out so
Top Up B
100 Case uniform F6 that continuous drilling could be achieved
FIA Case uniform (average B) on a full 28-m stand, hoping to penetrate
Depth below top fulmar TVDss
150 FIA Kv/Kh dow ngraded in C sand the reservoir in this interval. The idea
FIA B/C Multi-layer behaviour Top Lw B was to keep a high ECD while drilling,
FIA Tigth top C
200 therefore keep the open hole high-
Initial
FIA Base case
pressure shale open. A connection would
Top A
250 Measure FMT make the downhole pressure decrease to
Average BHSIP derived from WHSIP equivalent static density and increase the
chances of borehole collapsing.
300
As seen above, the reservoir was deeper,
Top Up Pent
350 so a connection had to be performed. The
hole remained stable when circulation
400 was stopped to connect the next stand
and drilling could resume. The cutting
450 Top Lw Pent
structure of the shoe was damaged on a
hard limestone stringer before encoun-
tering the reservoir, and drilling stopped.
500
The decision was made to pull out to
Figure 8: Comparison of actual pressures versus planned pressures within the reser- change the cutting structure. As some
voir. In thick red are the actual pressures. The red dot in the C sands is the pressure cutters were left in the hole, a clean-out
measurement taken at 1004 bar just below the shoe of the drilled-in liner. run was undertaken with a bit and a
junk basket. Then the drilled in liner was
the Valhall high gas layer, and one was launcher element. Once initiated, the run again.
installed at the top to seal against the expansion process went smoothly, and During all these operations, the hole
existing casing. There were lengthy dis- the liner was installed as planned. remained perfectly in gauge without
cussions on the thickness of the elasto- any fill found at bottom. This was a
mer. They needed to be thick enough so Injectivity test at expandable liner
surprise as rock mechanic calculations
to ensure sealing against the formation shoe
had accounted for manageable borehole
in the open hole, but thin enough to mini- Prior to continuing to drill, the well was instabilities with this mud weight in the
mize the piston effect when run inside displaced to the planned 1.86 sg mud high-pressure cap rock.
the casing and to ensure that they would weight. It was checked that the hanger
not prevent the expansion process. The liner drilled the remaining of the
joint was sealing properly, then the shoe cap rock and penetrated the reservoir
was drilled with a special drill bit. As at 5,451 m. No more than 500 l of losses
A consensus was reached with 0.19-in.
soon as the shoe was drilled, an influx were noticed when penetrating the res-
(4.8-mm) thickness. The specially bonded
was recorded from the well. Multiple ervoir. These very low levels of losses
joints had to be flown from the US to pro-
flow checks and circulations showed a lead to questioning the top reservoir
vide the required elastomer thickness.
decrease in the flowing trend. At some depletion. Drilling was continued until
The liner reached TD without problem. stage, it was believed that the elastomer 5,484 m, where the liner hanger was
The dart was dropped, but the circulat- around the liner was not properly isolat- about to hang on top of the expandable
ing rate was limited to 300 l/min to mini- ing the Valhall formation. liner. The liner was then cemented.
mize ECD on bottom and prevent losses At the end of the cement job, the liner
In provision of a cement squeeze at the
in the Plenus Marls fracture. With this top packer could not be energized. When
shoe, an injectivity test was performed,
low flow rate, the dart was not displaced the running tool was pulled to surface,
which gave an injection pressure of 2.17
properly and stopped half-way down the it was found that the packer energizing
sg EMW. This value was a surprise, as
string. Pumping was stopped after theo- dogs had been prevented to expand by
the formation had always been drilled
retical displacement volume plus some jammed LCM material from the designer
with a 2.23 sg ECD on the non-depleted
margin. After investigation, including mud. The debris barrier, plugged by
wells. Further flow checks and cir-
wireline work to locate the dart in the the same material, had collapsed and
culations were performed with still a
string at 600 m, a decision was made to failed under the increase of pressure
decreasing trend, the last one being per-
displace the dart at 600 l/min. This flow downhole. A separate run had to be per-
fectly stable. This showed that the seal
rate achieved dart displacement to bot- formed to run a back-up liner top packer.
was effectively achieved. It is believed
tom, but with only 5% mud returns.
that the influx – experienced after drill- Reservoir drilling
The expansion process was initiated at ing the shoe – was created by the mud,
Once achieved, the successful isolation
a much higher pressure than expected, which had been squeezed when displac-
of the transition between cap rock and
close to the burst pressure of the cone ing the dart.
reservoir, one could believe that the mud
weight could be decreased to drill the could not create enough drawdown rela- • Despite efforts to reduce geological
remaining of the reservoir with a mini- tive to the hydrostatic pressure of the and reservoir uncertainties, surprises
mum overbalance. mud to give actual formation pressure. were found – the biggest in the over-
The indication gained was only that the burden, long before reaching the res-
Mud weight was decreased down to 1.5
pressure was lower than 1.42 sg EMW. ervoir. Most are believed to be a conse-
sg prior to drilling the shoe with a special
PDC bit. High levels of gas were experi- A wiper trip was performed with a quence of the reservoir depletion.
enced as soon as the shoe was drilled. formation evaluation tool in the string. • High permeability sands up to 100 mil-
Mud weight was raised to 1.60 sg, and There was no blockage of the tool by the lidarcy were drilled with 600 bar over-
drilling continued on 40 m, in the believed LCM within the designer mud. balance without any significant losses.
poor productivity C sands, still with high
Final 4 ½-in. liner was then run. It hung
gas levels. These high gas levels could • The formation damage created by the
up 40 m above TD. It was decided to
have come from the cap rock via a leak designer mud, if any, was bypassed by
cement it in place and drill out inside
path through a deficient cement sheet of the perforations.
cement afterwards to give access to the
the liner, or from a higher-than-expected
lowest planned perforation intervals. Aknowledgements: The Elgin / Franklin
pressure in the drilled reservoir section.
project is a development operated by Elf
Completion and perforation
Drilling stopped, and wireline logging Exploration UK PLC on behalf of itself and
was performed, including pressure mea- For completion, the well was displaced its co-venturers, whose support is acknowl-
edged: Elf Exploration UK PLC (operator);
surements. The well was displaced back to water and checked for integrity. The
E.F. Oil and Gas Ltd; Eni Elgin/Franklin Ltd;
to 1.86 sg designer mud to allow a safe completion was installed without notice- BG International (CNS) Ltd; E.ON Ruhrgas
trip out of hole. able problem. Perforation was started on UK Exploration & Production Ltdl; Esso
wireline. Because the formation pressure Exploration and Production UK Ltd; Chevrom
A resistivity/gamma ray log was then
was not yet ascertained, pressure was North Sea Ltd; Dyas UK Ltd; Oranje-Nassau
run. These logs showed the presence (UK) Ltd.
applied at surface to make sure the perfo-
of good sands at the top of the reser-
ration would not take place largely under- The authors want to thank the TOTAL man-
voir, covered partially by the drilled-in
balance. There were problems during agement who supported this publication, and.
liner. A pressure measurement logging The authors want to thank M. Gregoire, S.
perforations operations, including misfire,
tool indicated a pressure in this layer Robertson, A. Humphreys, V. Neillo, C. Tindle,
stuck guns and coiled tubing fishing. The
of 1,004 bar, which was an equivalent C. Onu, G. Holm, Total Exploration UK Ltd,
last part of the perforation was performed A. Onaisi, TOTAL, for their important contri-
mud weight of 1.92 sg. Nevertheless, the
on coiled tubing. The Top C layer perfo- bution to this project. The authors want to
Geoscience Department remained con-
ration was aborted for fear of potential thank the offshore team that contributed to
vinced that the best reservoir sands still
problems generated by the high-pressure the successful drilling of this well. This would
to come were heavily depleted. not have been achieved without their strong
difference between this layer and already-
involvement and commitment.
The situation was as follows: All tech- perforated low-pressure sands.
niques planned to isolate the high-pres- References
Clean-up, bottom pressure
sure formations from the low-pressure Aston, M.S., Alberty, M.W., McLean, M.R., de
ones had been used, and still the open The well was lifted with coiled tubing by Jong, H.J. and Armagost, K., BP explora-
hole was showing high pressure. The pumping nitrogen and cleaned up. The tion. 2004. Drilling Fluids for Wellbore
only remaining option was to drill to wellhead shut-in pressure by the end Strengthening. SPE 87130
final depth with the designer mud. The of the clean-up indicated a bottomhole Alberty, M.W., McLean, M.R., BP exporation.
losses response plan was updated, and shut-in pressure of 440 bar: 0.82 sg 2004. A physical model for stress cages. SPE
drilling continued at a reduced penetra- EMW. This indicated that the best sands 90493
tion rate into the remaining C sands, the had been drilled with 600 bar (8,700 psi) Demong, K., Halliburton Energy Services,
B sands and 55 m into the A sands. The overbalance. Figure 8 illustrates the Rivenbark, M., Enventure Global Technology
plan was to stay away from the virgin comparison between pressure prediction and Mason, D., Shell Technology ventures.
2004. Breakthroughs using Solid Expandable
pressure pentland sands. and actual pressure at the time drilling Tubulars to Construct Extended Reach Wells.
the reservoir was performed. SPE 87209
TD was reached at 5,678 m with only a
seepage losses rate noticed around 200 Moreover, the indications were that the Dupriest, F. E., ExxonMobil. 2005. Fracture
l/hr for a small period. ECD during this well was delivering its expected potential Closure Stress (FCS) and Lost returns
drilling was computed at around 1.93 sg, and that the formation damage created Practices. SPE 92192
with peaks up to 2.0 sg EMW. It is dif- by the designer mud, even under this Hemphill, T., Halliburton Baroid. 2005.
ficult to ascertain which of the following large overbalance, was either insignifi- Integrated management of the safe operating
happened: cant or largely bypassed by the perfora- window: Wellbore stability is more than just
fluid density. SPE 94732
• The borehole was fractured, and the tion tunnels. Since then, the well has
been delivering 17,000 bbl/day. Marchina, P., SPE, and Onaisi, A., SPE,
borehole-strengthening technique Total. 2004. Reservoir-geomechanics Coupled
worked. Simulations: a Powerful Tool for Well Design
Conclusions and operations in an HPHT Environment.
• The designer mud increased the frac- SPE 92546
ture initiation pressure by creating a • An infill well was successfully drilled,
perfect seal between the borehole and completed and perforated in an HPHT Sinor, L.A., Hughes Christensen Co., Tybero,
reservoir after depletion of 660 bar P., and Eide, O., Amoco Norway Oil Co, and
the formation. Wenande, B.C., Consultant. 1998. Rotary Liner
had occurred.
Drilling for depleted Reservoirs. SPE 39399.
• The mud column pressure was below
the initial fracture initiation pressure. • Two architectures were designed to This article is based on IADC/SPE 112708,
achieve this goal, providing a fallback “Successful Development Drilling of an HP/
Wireline logging was performed, which HT Infill Well in a Highly Depleted Reservoir:
solution for any geological or reservoir
included pressure measurement points. Case Study,” 2008 IADC/SPE Drilling
surprises encountered. Conference, held in Orlando, Fla., 4-6 March.
However, the pressure measurement tool