Managing Circulation Losses in a Harsh Drilling Environment: Conventional Solution vs.CHCD Through a Risk
Автор работы: Пользователь скрыл имя, 11 Марта 2013 в 12:27, реферат
Краткое описание
The presence of extensive karst/fractures along with severe circu-
lation losses prevents successful drilling using the conventional
circulation technique. For this scenario, the pressurized-mud-cap
closed-hole circulation-drilling (CHCD) technique was applied,
and it proved to be the most effective method to drill through the
Lower Carboniferous carbonate.
Прикрепленные файлы: 1 файл
2011 SPE Drilling & Completion
1
Managing Circulation Losses in a Harsh
Drilling Environment: Conventional
Solution vs.CHCD Through a Risk
Assessment
S. Masi, C. Molaschi, SPE, and F. Zausa, Eni E&P; and J. Michelez, Kwantis
Copyright © 2011 Society of Petroleum Engineers
This paper (SPE 128225) was accepted for presentation at the IADC/SPE Drilling
Conference & Exhibition, New Orleans, Louisiana, 2–4 February 2010, and revised for
publication. Original manuscript received for review 24 May 2010. Revised manuscript
received for review 20 January 2011. Paper peer approved 24 January 2011.
Summary
Conventional drilling techniques used in harsh drilling environ-
ments are sometimes impractical or uneconomical. This was
experienced while drilling through the reservoir section being
investigated. This section is characterized by high pressure and
formation fluids estimated to contain 18–20% H
2
S and 4–6% CO
2
.
The presence of extensive karst/fractures along with severe circu-
lation losses prevents successful drilling using the conventional
circulation technique. For this scenario, the pressurized-mud-cap
closed-hole circulation-drilling (CHCD) technique was applied,
and it proved to be the most effective method to drill through the
Lower Carboniferous carbonate.
In conventional drilling through such lost circulation zones, the
typical method is to isolate the thief formation [lost circulation-
materials (LCMs) pills, gunks, and cement plugs] to maintain an
overbalanced condition. These methods have their disadvantages:
uncertain results, time and cost increase, and damages on well
productivity. Furthermore, it may be a long, ineffective process
because of multiple leakoff zones or massively fractured zones.
Where unsustainable losses occur and conventional circulation is
no longer possible, the CHCD technique may be used to allow the
continuation of drilling. This technique is not underbalanced drilling.
The annulus is closed, and no returns are circulated to the surface.
The CHCD is not a simple or an inexpensive system. It is used
only when other systems for controlling losses have proved to be
ineffective. Large quantities of water and mud are required as well
as additional equipment and specialized personnel. An accurate
evaluation of several factors has to be performed. Drilling hazards
are the major issue because the primary barrier to well influx is
jeopardized, but also rig time, material availability, and consump-
tion have to be considered.
The aforementioned uncertainties can be addressed through
risk evaluations by comparing two drilling scenarios of a typical
development well where one uses CHCD technology and the other
uses conventional drilling techniques.
This paper will discuss the operational details of conducting
CHCD as well as the risk-management approach, which involves
the identification, evaluation, and mapping of all risks involved in
each scenario (qualitative risk assessment). A probabilistic model
is, therefore, developed to combine all the risks identified and to
address their consequences within operational time (quantitative
risk assessment).
Introduction
Drilling the reservoir section is the most challenging part of the
drilling activity that is being carried out at the oil field under inves-
tigation. The reservoir is a carbonate platform, and it is well known
that carbonate formations (limestone/dolomite) are characterized
by the presence of fractures and/or large vugs, and consequently
circulation losses while drilling occur frequently. Sometimes losses
are minimal, but at other times they are large enough to become
total losses. This situation has often been experienced in this
reservoir. In fact, in certain areas, unsustainable losses occur and
conventional circulation is no longer possible. Therefore, one of
the main concerns is the magnitude of losses that could be faced
while drilling the reservoir zone.
It must be emphasized that when using conventional drilling
techniques through lost circulation zones, the standard practice is
to minimize losses by attempting to seal off the thief zone before
continuing drilling. This is usually achieved by pumping LCMs
pills and setting cement plugs such that the mud weight will over-
balance formation pressure at all times. However, this is a long and
often unsuccessful process because multiple leakoff zones may be
encountered in sequence or because massively fractured zones can-
not be cured. Furthermore, these methods have several disadvan-
tages such as offering uncertain results and being time consuming
(and, therefore, expensive), and in hydrocarbon reservoirs, these
may seriously reduce well productivity. In this scenario, the CHCD
technique allows drilling in circulation-loss zones by controlling
the annulus pressure and injecting a sacrificial fluid into the thief
formations, thereby minimizing the nonproductive time (NPT) and
costs (large amounts of lost drilling fluids and nonproductive rig
time) typically required to implement the traditional methods of
circulation-loss control.
For this field, the CHCD was determined to be the most effec-
tive method to drill through the carbonate reservoir.
The opportunity to use the CHCD system is subject to the res-
ervoir conditions. These conditions are strictly correlated through
geological interpretation of the reservoir scenario. Despite the
results of the geological analysis, the possibility of needing to
use the CHCD technique is always considered during the plan-
ning of the drilling activities. Therefore, before beginning drilling
of the reservoir, all the required mitigation actions and tools are
identified and planned properly. Implementing mitigation actions
before performing CHCD guarantees the same level of safety as
in conventional drilling. Furthermore, it is demonstrated that this
method is the most convenient one (time and, therefore, cost)
when drilling this type of reservoir. To asses this, two different
scenarios have been compared: (a) conventional overpressure drill-
ing by attempting to cure losses with traditional methods and keep
CHCD technique as a backup plan in case conventional drilling
is no longer possible (reaction); (b) switch immediately to CHCD
technique anywhere severe losses occur without any attempt to
stop them (anticipation). Both strategies can be addressed in a
stochastic analysis to consider not only planned events but also
uncertain events. Such an analysis is performed through project
risk-management process.
CHCD Technique
The conventional drilling technique is based on
• Maintaining circulation
• Using the mud column as the primary well-control barrier
• Transporting the drilling cuttings to surface
In a fractured reservoir of significant thickness, it is possible that
a drilling fluid properly balancing the reservoir pressure at top,
2
2011 SPE Drilling & Completion
increasingly overbalances the reservoir pressure with depth, result-
ing in circulation losses (Urselmann et al. 1999).
This is because of the differential specific gravity (sg) between
the drilling fluid and the formation fluid. When circulation losses
are experienced if balance is achieved at the depth at which the
losses occur, above this depth, there will be underbalanced condi-
tions. In the reservoir under investigation, which has light reservoir
fluids, high reservoir pressure, significant reservoir thickness, and
a fracture system, the previously described scenario is likely. Fig. 1
summarizes the concept.
The continuous green line is the pressure exerted by a column
of mud at 1.94 sg, properly balancing the formation pressure (red
line) at its top. The blue line represents the under-/overbalance
condition whenever a circulation loss is experienced and balanced
by the mud in hole at 4350 m true vertical depth (TVD) in a res-
ervoir bearing fluids at 0.61 sg.
CHCD (Sweep et al. 2003) is a technique that allows safe drill-
ing with no returns to the surface. For this method, a sacrificial fluid
is pumped through the drillstring and is intentionally lost through
the formation while the annulus is filled with a dual-gradient fluid
system {i.e., the sacrificial fluid on the lower portion of the hole,
and the mud cap in the upper portion [light annular mud (LAM)]}
(Colbert and Medley 2002). Although CHCD is not underbalanced
drilling in the sense that formation fluids are not allowed to flow
to surface, the equivalent annular dual-system density is such that
it equals (or it is slightly lower than) the formation gradient.
The annulus is closed off at the surface, and drilling fluid is
continuously injected into the drillstring and through the drill bit
in order to clean the hole by injecting cuttings into the fractured
formation (Fig. 2). In practice, bottomhole pressure will be con-
trolled by the drilling mud in the annulus and the sacrificial fluid
in the drillpipe.
The sacrificial fluid will be lost in the lost-circulation zone, so
an inexpensive and lightweight fluid is normally used. The vis-
cosity of the sacrificial fluid should be low to minimize frictional
pressure loss. Because the sacrificial fluid has a lower density than
either the equivalent density of the reservoir pressure or the density
of the annular fluid, float valves are placed in the drillstring to
prevent backflow of the injected fluid when the injection pumps
are shut down.
The density of the annular fluid (LAM) added to any surface
pressure on the annulus is equal to the formation pressure at the
top of the lost-circulation zone. The fluid density should be only
slightly less than the equivalent density, which would exactly bal-
ance the reservoir pressure. This slight hydrostatic underbalance
will facilitate monitoring hydrocarbon migration in the annulus.
In fact, because of specific-gravity differences, hydrocarbons
from the reservoir migrate into the annulus and up through the mud
column that is being used to help balance the formation pressure.
When this occurs, the annular pressure will begin to slowly
increase. Once a predetermined casing pressure is reached, LAM
is bullheaded down the annulus until the casing pressure reaches
the datum value. This will push the migrating hydrocarbons back
into the reservoir.
As the well is drilled deeper, additional lost-circulation zones
may be encountered. If these zones are in pressure communication
with the original lost-circulation zone, the pore pressure will be
higher by an amount equal to the reservoir-fluid gradient multiplied
by the increased depth. Because the reservoir-fluid gradient will
be lower than the conventional-drilling-mud fluid gradient, the
equivalent mud density in these deeper zones will be less than that
of the original zone. However, because the actual absolute pressure
will be higher, the column of fluid in the annulus will continue to
be supported by the pressure of the original lost-circulation zone.
Fig. 3 depicts this operation.
CHCD Feasibility and Required Equipment
CHCD has been used for rim wells and some transition wells of the
field, which have been proved to have a high likelihood of experienc-
ing massive circulation losses. In fact, minor circulation losses are
mitigated with traditional remedies, but none of these solutions is
fully successful when catastrophic losses occur (Sweep et al. 2003).
In addition to incurable circulation losses, other conditions for
which the CHCD technique is essential for efficient drilling are
• Low injectivity pressure into the formation—high formation
capability of receiving the drilling cuttings
• Safety concerns for the surface facilities and personnel
because of an unmanageable flow of sour reservoir fluids (high
percentages of H
2
S and/or CO
2
) into the wellbore
The injectivity pressure into the major fractures of Unit 1 is so
low (high conductivity) that the pumping rate for the sacrificial
fluid is not important. However, this is also a characteristic of
formations that will flow at a high rate. Thus, the combination of
sour gas, high bottomhole pressure, and severe circulation losses
(with low injection rate) adds an extra risk to continue convention-
ally drilling the reservoir section.
For CHCD to be effective, a number of essential items are
required before starting such an operation:
• A complete feasibility study to establish the reservoir-pres-
sure gradient and reservoir-fluid characteristics (i.e., above or
below bubblepoint)
• Rig-personnel training
Res pressure and mud weight
4000
4100
4200
4300
4400
4500
4600
4700
4800
700
800
900
1000
Pressure [bar]
TVD [m]
Under-overbal Mud
Res Pr (Fluid Grad 0.61sg)
1.94sg Mud Grad
Fig. 1—Reservoir pressure and mud weight.
SAC fluid and cutting
pumped into Karst Formation
Fig. 2—Reservoir scenario while drilling using CHCD.
2011 SPE Drilling & Completion
3
• An adequate supply of low-cost sacrificial fluid
• Capability to increase mud volume fast and effectively (mud
supply)
• Adequate additional surface equipment [rotating blowout
preventer (BOP), high-pressure pumping system]
The implementation of the CHCD system requires a large
amount of sacrificial fluid. Drilling blind (no fluid returns to sur-
face) by simultaneously pumping oil-based mud (or water-based
mud) into the drillstring and annulus is clearly too expensive. Thus,
water is used as sacrificial fluid. The sacrificial fluid is pumped
through the drillstring and displaced in the karsts/fractures while
carrying the drill cuttings. Meanwhile, a pressurized LAM column
is applied in the annulus. It is preferred that the last casing shoe
be set as close as possible to the top of the fractured zone, and it
is recommended that the total loss zone be drilled with a reduced
hole size to assist hole cleaning and to reduce the volume of sac-
rificial fluid required.
The CHCD technique uses blind drilling with the choke com-
pletely closed while the annulus is sealed in by the rotating control
device (RCD). One or more rubber elements are mounted on a bearing
assembly inside the rotating control head to seal around the drillstring.
Drillstring rotation is still accomplished because of the bearing assem-
bly in the RCD. The RCD is the only additional piece of well-control
equipment required at the surface (see Figs. 4 and 5).
Well control is maintained by closing the annulus and not
allowing any reservoir fluid to flow to the surface. However, before
A
B
C
D
E
A. Fractured level is encountered. Annulus mud
level falls until leak off pressure is equalized. LAM
weight determined
B. Resume drilling: more fractured levels encountered
C. Fluid from highest formation pressure zones can
go only to a lower pressure level. It would not go up
to the annulus as far as the annulus is shut in (BOP)
D. & E. Fluid from highest formation pressure zones
cannot leak downward because formation pressure
increases with depth. Migration of fluid into
wellbore is controlled by csg shut in pressure.
Fig. 3—Wellbore condition while drilling with the CHCD technique.
Rig Floor
Rotating Control Device
Shaffer Annular 10k psi
Kill Line
Pipe Rams 7”
Blind / Shear Rams
Pipe Rams 5”
Pipe Rams 5”
Section D Casing Spool
1.844 m
± 1.21 m
1.538 m
1.718 m
1.718 m
Landing Point
Choke Line
8.028 m
Fig. 4—BOP stack with RCD.
Fig. 5—Well-control equipment: BOP stack with RCD.
4
2011 SPE Drilling & Completion
starting CHCD, supplementary considerations are required. In fact,
CHCD requires the following additional equipment:
• RCD.
• High-pressure fluid-circulation system (standpipe, rotary hose,
top drive). This (most likely a 7,500-psi working pressure system)
is required for water injection; injection pressure in certain cases
exceeds 5,000–6,000 psi.
• High-pressure pumps (grizzly pumps), other than those typi-
cally provided by the drilling rigs.
• Drillstring floats.
• Casing deployment valve to maintain double-barrier condi-
tion while tripping the bottomhole assembly (BHA).
Sour-Gas Consideration. The reservoir hydrocarbon is 45°API
oil with associated gas. The gas contains 25 mol% H
2
S and 2 to
4 mol% CO
2
. The presence of H
2
S at this concentration requires
rigorous personnel safety and environmental protection. This
means that no formation fluids are allowed at the surface unless
in a strictly controlled manner (e.g., while well testing or during
production operations).
At reservoir conditions for most parts of the wellbore, the
range of possible pressures does not allow the sour oil to release
gas. Bubblepoint is near 4,000 psi, and gas starts to separate from
oil at approximately 1,400–1,500 m TVD below the wellhead, if
1.90- to 2.00-sg mud is being used. Therefore, there will not be
free H
2
S gas downhole.
The reservoir temperature is approximately 95–110°C. Suscep-
tibility of carbon steel to sulfide stress cracking is greatly reduced
once the temperature exceeds 65.5°C, and hydrogen embrittlement
(of which sulfide stress cracking is a form) is minimized at tem-
peratures above approximately 85°C. In normal conditions, this
should mitigate the negative effects of the H
2
S sour fluids on the
drillstring. However, because of the continuous pumping of treated
water, especially during the winter season, the bottomhole tem-
perature may be greatly reduced—approaching 65°C. In this case,
the bottomhole temperature may not be high enough to mitigate
H
2
S corrosion; therefore, extra care has to be taken to protect the
drillstring from cracking. For this reason, all fluid pumped down-
hole while using CHCD, including the sacrificial fluid, is treated to
prevent corrosion. Along with these preventive measures, drilling
mud is adjusted to optimize its pH.
Risk-Assessment Approach
It has been stated already that CHCD-technique application is
subject to several conditions. However, all the reservoir plans that
were considered for this field were investigated as potential CHCD
scenarios regardless of any geological considerations. For this
reason, all the required mitigation actions as well as any neces-
sary drilling equipment were thoroughly identified and organized
before the start of drilling for the reservoir section. The likely
drilling scenarios were assessed using a stochastic analysis that
would consider all planned events and permit the evaluation of any
uncertainties. This was performed by comparing two situations:
• A—Conventional + CHCD: Plan for conventional overpres-
sure drilling by attempting to cure losses with traditional methods
and keep CHCD technique as a backup plan in case conventional
drilling is no longer possible (reaction).
• B—Anticipated CHCD: Switch immediately to CHCD tech-
nique anywhere severe losses occur, without any attempt to stop
them (anticipation).
A pure statistical comparison of these two strategies would
have been possible if data from a large number of reference wells
reflecting both options were available. However, at this time, the
number of CHCD applications in analogous conditions is rather
limited and not sufficient to make any kind of reliable statistical
conclusion.
Therefore, the evaluation of the opportunity to anticipate the use
of CHCD in the drilling conditions just described was addressed
through two different reference wells:
• A—Conventional + CHCD: It was attempted to drill the res-
ervoir section by using the conventional drilling technique. Losses
occurred, and despite pumping LCMs and setting gunk plugs, they
could not be stopped. CHCD was applied when the losses became
unsustainable and continuing with conventional drilling techniques
was clearly unproductive and time consuming. Liner running and
cementing did not result in any NPT, despite other previous experi-
ence in the field.
• B—Anticipated CHCD: From the start, it was planned that
this reservoir section would be drilled using the CHCD technique
as soon as massive drilling losses were recorded. Therefore, the
top part of the reservoir section (tight zone) was isolated with a
liner string that was run and cemented without any problems, just
above the severe-losses zone. This was performed to set the last
casing shoe as close as possible to the top of the fractured zone,
and to continue drilling using a reduced hole size. Drilling was
thus switched to CHCD as per plan. Drilling operations continued
efficiently down to section total depth without any problems.
A slotted-liner string was then run in hole and set, allowing opti-
mization of the subsequent completion activity.
The comparison between the aforementioned two strategies was
performed through operational productive time and related NPT.
All assumptions were reviewed from a risk-assessment point of
view to address any specific uncertainties. Evaluations were based
on offset data as a starting point (basis for evaluation), and then
reviewed using technical criteria (expert review) because avail-
able offset data were minimal (because of the limited number of
reservoir sections drilled in the potential lost-circulation zone) and,
therefore, were insufficient for any kind of generalization.
Eni E&P has implemented a risk-management process to be
applied to any investment evaluation, from drilling to facilities. This
is a common approach in the industry, as most oil and gas invest-
ments are now evaluated on a probabilistic basis. All investment
assumptions (capital expenditures, operating expenditures, revenues)
are by nature uncertain and, therefore, require risk-assessment
techniques to develop accurate estimates. For that reason, the
evaluation of well programming within uncertain conditions is
made through risk analysis.
In this particular case of CHCD strategies evaluation, all related
uncertainties can be assessed through risk evaluation.
Methodology
The analysis was performed using a risk-management process
(Grey 1995), which is summarized in Table 1.
Qualitative Risk Study. The analysis consisted of the identifica-
tion of all the required data to perform the risk assessment. The first
step was to accomplish the time-breakdown analysis for the two
reference wells. Time-breakdown analysis involves the application
to any interval of timing as defined in the daily drilling report by
a distinct code. This code will be used as a comparison tool for
statistical purposes. The time-breakdown coding system does not
aim to translate into codes all the operations that were carried
out in the well, but it allows the identification of both the main
operations and the major downtimes. Deterministic accounts were
derived using historical well data given in Table 2.
Actually, these deterministic durations in Table 2 are based
on reference wells that have been drilled in similar conditions.
These values reflect normal operating time, where the benefits of
Anticipated CHCD are already visible (193 hours). Nonetheless,
the objective of this study was also to address the uncertainties
related to both scenarios.
Identification and Evaluation. Several events were identified as
potential risk issues for the wells while drilling the reservoir sec-
tion being investigated. These are described in Table 3.
Quantitative Risk Study. The identified events were further
assessed quantitatively (probability and time impact) for both
strategies and were introduced into the stochastic model as vari-
ables. Uncertainties were built as variables, using discrete distri-
bution (P1/X1, P2/X2) for the risk occurrence and then triangular
(TRIGEN) distributions (minimum, most likely, maximum) for
2011 SPE Drilling & Completion
5
performance factors and uniform distributions (minimum, maxi-
mum) for risks time impact (Fig. 6). Some correlation groups were
defined to represent that some input variables occur to some degree
in tandem (e.g., circulation losses and BHA stuck).
Results
Once the stochastic model was built, it was run using a commercial
risk-analysis software for more than 5,000 iterations. This number of
iterations gave a sufficient level of convergence for the results (less
than 1.5% variation). The main results are summarized in Table 4.
It is clear that should large losses occur while drilling through
the reservoir section being investigated, it is more economical if all
the requirements for the CHCD technique are in place and ready to
be put into service rapidly. With a P50 level of confidence, drilling
time can be 25% less than when attempting to continue drilling
with a conventional drilling technique. It is expected that the cost
will also be reduced by approximately 25%. It is possible that some
uncertainties will affect drilling cost and not drilling time. When
this occurs, the reduction in cost when compared to that of time
will likely be different.
Fig. 7 shows the normal (productive) time distributions on
a noncumulative basis. These values do not include any risk at
that time, but only variations of planned activities caused by rig
performance, crew capabilities, and rate of penetration/tripping
speed. These variations are built from performance records for
offset wells when available and from extrapolated ratios when no
offset reference was available to keep consistency throughout the
assumptions.
Fig. 7 illustrates the reduction of overall uncertainty (lower
standard deviation) in Plan B.
Fig. 8 is a cumulative probability curve showing the total time
distribution for both strategies. This includes all normal and trouble
time. The y-axis gives the confidence level for any value on the
x-axis (e.g., total time at P50 for Plan B is 1,208 hours). Therefore,
Fig. 8 validates the aforementioned conclusions: Drilling time,
and therefore costs, can be reduced through the anticipation of
TABLE 1—RISK-MANAGEMENT-PROCESS METHODOLOGY
Steps
Objective
Input
Process
Output
e
vi
ta
til
a
u
Q
Identification
Better
understand
project
uncertainties
Brainstorming
ideas, checklists,
offset data, well
operation
personnel
interviews
Identify all the
internal/external
risks and
opportunities
and list them
Risk Register
Evaluation
Asses and
categorize
each
individual risk
Statistical
records,
interviews, expert
estimates
Evaluate every
identified risk in
terms probability
of occurrence
and impact
(including time
for mitigation or
remedial actions)
Single Risk Evaluation
e
vi
ta
tit
n
a
u
Q
Duration
Risk
Analysis
Secure
duration
accuracies
and
contingencies
for one well or
a sequence of
wells
Deterministic
duration variables
and correlations
(as from Risk
Register and
expert
assumptions)
Build a model
including
variables and
correlations for
each type of
well. Combine
variables for all
wells using
Monte Carlo.
Duration risk model
that provides a range
of possible results:
probabilistic curve,
results at several
confidence levels
(e.g., P10, P50, P90),
and probability of
reaching a specific
target
Sensitivity
Analysis
Identify main
drivers for
duration
Duration risk
model (as
developed in the
previous step)
Simulate several
assumptions for
the variables
from the existing
model
Monitor duration
changing factors
through tornado
diagrams and spider
plot
TABLE 2—TIME-BREAKDOWN ANALYSIS
Phase
Operation
Plan A: Conventional +
CHCD (hours)
Plan B: Anticipated CHCD
(hours)
8½ in.
BOP stack (installation
and test)
5.
5
2
0
0.
3
1
0
5.
2
1
4
4
1.
4
2
7
g
ni
lli
r
D
.n
i
½
8
0
5.
6
2
9
7.
1
6
g
ni
tt
e
s
g
ni
s
a
C
.n
i
½
8
0
0.
7
1
0
0.
8
3
g
ni
tn
e
m
e
C
.n
i
½
8
0
0.
4
2
0
0.
8
6
1
n
oi
ta
ra
p
er
p
ll
e
W
.n
i
½
8
0
0.
6
9
1
—
g
ni
lli
r
D
.n
i
⅞
5
0
0.
0
1
1
—
g
ni
tt
e
s
g
ni
s
a
C
.n
i
⅞
5
0
5.
1
1
8
3
9.
4
0
0
1
L
A
T
O
T
6
2011 SPE Drilling & Completion
TABLE 3—POTENTIAL RISK ISSUES
NPT Type
s
e
s
u
a
C
Source Used for
Uncertainty Identification
Well problems
Circulation losses in the reservoir section
From reference well data
Fluid influx into the wellbore
From reference well data
Reaming and washing of the wellbore
From reference well data
at
a
d
ll
e
w
e
c
n
er
ef
er
m
or
F
s
kr
o
w
w
ar
D
er
ul
ia
f
gi
R
Mud pumps
From reference well data
Top drive system
From reference well data
at
a
d
ll
e
w
e
c
n
er
ef
er
m
or
F
P
O
B
Downhole
equipment failure
Bottomhole assembly, drillstring (H
2
S
cracking effect on downhole equipment)
From reference well data
Wireline equipment
From reference well data
Drill bits
From reference well data
Surface equipment
failure
Wellhead
From reference well data
Contractor’s surface equipment
From reference well data
at
a
d
ll
e
w
e
c
n
er
ef
er
m
or
F
re
ht
a
e
W
e
m
it
g
ni
ti
a
W
Contractor’s equipment
From reference well data
Shortage of LAM/OBM fluids and scavenger.
Critical consumable supplies/high volume
requirement (high cost impact)
From experts review
Shortness of cement and additives. Issue in
conventional drilling due to the high number
of gunk plug, this can cause a stop of the
drilling activity
From experts review
Lack of sacrificial
fluid
Weather (ice season)
From experts review
Pump failure
From experts review
Liner problems
Critical liner running with one barrier only (in
CHCD) and lower running speed
From experts review
Poor quality cement job
From experts review
Potential liner top cement jobs
From experts review
BHA stuck
Hole cleaning difficulty (pump acid fluid)
From experts review
Opportunity—
Easier completion
job
Reduced problems while performing
completion job
It is not the scope of the
quantitative study
Kick event
Loss of a safety barrier in CHCD = This is a
safety concern and needs further
investigation through fault tree analysis
It is not the scope of the
quantitative study
the need for CHCD. But the main indication is the reduction of
the time spread distribution when CHCD activities are anticipated.
Actually, this spread reduction is visible on normal-time analysis
(Fig. 7) and is highlighted further when considering the trouble-
time scenario (Fig. 8).
In fact, trouble time can be mitigated better through early
CHCD anticipation, thus the overall uncertainty in well duration
decreases. As a consequence, increasing the level of confidence
increases the time difference between both scenarios. At a P10
confidence level, the benefit of anticipating CHCD is about saving
336 hours, to be compared to the 193 hours in the deterministic
calculation presented in Table 2. At a P90 confidence level, the
benefit increases to 504 hours (21 days). These results are clearly
shown in the cumulative distribution for total time (Fig. 8).
Distribution
Graphic
Values Definition
Related Risks
TRIGEN
Minimum at a determinate
percentile (normally P10), Most
Likely (normally the
deterministic),
Maximum at a determinate
percentile (normally P90)
This is a simple representation of a normal type risk. There is a
Most Likely situation, with a higher probability with Minimum and
Maximum situations with decreasing continuous probabilities.
Use of the Risk Trigen function avoids the problem of the
minimum and maximum values not actually being possible
occurrences in the standard Risk Triangular function
DISCRETE
Value 1, Probability 1,
Value 2, Probability 2,
…
Value n, Probability n
Distribution to reflect some independent situations, each one
having its own probability of occurrence (e.g.. success vs.
failure)
UNIFORM
Minimum, Maximum
Distribution to assume a possible range of continuous situations
having all the same probability of occurrence between a
Minimum and a Maximum
Fig. 6—Basic distributions used to define related uncertainties.
2011 SPE Drilling & Completion
7
Distribution for Normal Time
0
1
2
3
4
5
6
7
8
650 700 750 800 850 900 950 1000 1050 1100 1150 1200
Hours
Frequency
A - Conventional + CHCD
B - Anticipated CHCD
Fig. 7—Normal (productive) time on a noncumulative basis.
Distribution for Total Time
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
900 1,000 1,100 1,200 1,300 1,400 1,500 1,600 1,700 1,800 1,900 2,000 2,100 2,200
Hours
Cumulated probabilities
A - Conventional + CHCD
B - Anticipated CHCD
Fig. 8—Cumulative probability curve of total time distribution (productive and trouble time).
Sensitivity
The sensitivity analysis is automatically provided by the Monte
Carlo simulation tool. It gives a tornado representation of the
most-correlated uncertainties for the results. The tornado diagram
highlights the variables on which efforts must be focused to change
the spread or the values of the output. Attention must be paid to the
fact that Standard B coefficients lower that 0.5 have very limited
impact on the results if taken individually.
The sensitivity regarding the impact of each variable on total
time (including NPT) is given in Figs. 9 and 10. The greatest
uncertainty for Plan A is circulation losses in terms of magnitude
of losses and successful cures. Previous experience highlighted that
attempts to cure losses to continue drilling with the conventional
technique, apart from being time consuming, have increased drilling
and completion costs and completion time, and also highlighted
that these efforts were not successful because at a certain point
drilling was switched to CHCD. For Plan B, there are no major
uncertainty impacts because of the fact that everything is properly
planned in case unmanageable losses should occur, specifically
that there will not be any loss of time before switching to the
CHCD technique. The required equipment is already in place, and
thus circulation-losses uncertainty has only a minimal impact.
TABLE 4—TIME-BREAKDOWN ANALYSIS
Min
P10
P50
P90
Max
Time
(hours)
A-Conventional + CHCD
1040
1384
1624
1904
2312
vs P50
64%
85%
100%
117%
142%
B-Anticiapted CHCD
808
1048
1208
1400
1776
vs P50
67%
87%
100%
116%
147%
8
2011 SPE Drilling & Completion
Regression Sensitivity for TOTAL A - Conventional + CHCD
0.47
0.38
0.38
0.27
0.25
0.25
0.19
0.14
0.10
-1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8
1
Circulation losses
Ream & Wash
Drilling 8 1/2
BHA, Drillstring,…
Hole cleaning difficulty
Shortness of LAM/OBM fluids
Shortness of cement
Critical liner running
BOP
Std b Coefficients
Fig. 9—Sensitivity analysis for Plan A.
Regression Sensitivity for TOTAL B - Anticipated CHCD
0.34
0.32
0.28
0.27
0.20
0.18
0.16
0.15
0.12
0.12
Contractors equipment
Drilling 5 7/8
BHA, Drillstring,…
Shortness of LAM/OBM fluids
Shortness of cement
BHA, Drillstring,…
Circulation losses
Drilling 8 1/2
Weather
Fluid Influx
Std b Coefficients
-1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8
1
Fig. 10—Sensitivity analysis for Plan B.
Conclusions
The CHCD technique was envisaged as the most effective method
to drill through the Lower Carboniferous carbonates of the reser-
voir when extensive fractures/karts were encountered. It allows
drilling to continue safely with minimum NPT and lower drilling
costs—in particular, when drilling rim wells and a certain number
of transition wells. In these types of wells, CHCD also reduces
formation damage and lower-completion timing and cost. How-
ever, before planning the CHCD technique, several considerations
have to be taken into account (nature of fractures and their extent,
knowledge of reservoir petrophysical characteristics, nature of
reservoir fluids). In particular, to ensure sufficient hole cleaning,
the presence of large fractures or karsts capable of receiving the
drill cuttings is required. Additional equipment and specialized
personnel are required, and mitigation actions have to be put in
place (e.g., rig-crew training).
The possible reservoir-drilling plans can be assessed using a
stochastic analysis. This allows the identification of the uncertain-
ties that will have an impact on the reservoir-drilling activity. The
analysis confirmed that the highest uncertainty when attempting
to drill conventionally is the magnitude of losses. A reduction of
the overall uncertainty is achieved when CHCD is anticipated.
This reduction demonstrates that the CHCD technique has been
planned and executed properly. Furthermore, it was shown that this
technique will allow a reduction in drilling time of approximately
25%. Costs are expected to be reduced by approximately 25%.
However, it is possible that some uncertainties will affect drilling
cost and not drilling time. When this occurs, the reduction in cost
when compared to that of time will likely be different.
Further Work To Determine CHCD Blowout Risk
CHCD operations should be considered very carefully at the
health, safety, and environment level because the primary barrier to
well influx is jeopardized. As such, it is recommended to perform
a dedicated blowout probability analysis to address the proper level
of safety risk and whether any further procedures and/or additional
equipment is needed.
The time impact because of a blowout cannot be accounted
for in the preceding operational-time risk assessment. Therefore,
such a probability calculation is beyond the scope of this paper.
A blowout probability analysis should include the following steps
found in Fig. 11.
2011 SPE Drilling & Completion
9
Fig. 12—Fault tree for a blowout event.
All calculations are made using the fault tree, combining the
probabilities of single elements and logical links between events.
An illustration of this process is given in Fig. 12.
The fault tree is a rigorous technique used to calculate a blow-
out probability for a complex well-drilling system (Masi et al.
2010). The accuracy of the input data strongly impacts the analysis
results. Therefore, the results must be interpreted considering the
limitations on the input data. Sensitivity analysis can be used to
identify the input values that should be reconsidered to increase
the accuracy of the predictions.
The following are recommended to improve the accuracy of a
blowout probability analysis:
• Involve all the relevant drilling and completion experts to iden-
tify the most critical phases in terms of possible blowout and
related risks.
• Find, if possible, more than one value of failure rate for each
piece of equipment, comparing different information sources.
• In case of different database values for the same equipment-
failure rate, build a probabilistic distribution to reflect the
discrepancy.
• If a failure rate is not available, seek advice from the most-
competent experts to make the best estimation of the failure
rate. Expert opinions can also be taken into account through
probabilistic distributions.
• A fault tree may be tailored to its top event (blowout), and shall
include only those faults that contribute to the top event in a
realistic manner (avoiding overcomplexity with very marginal/
remote probability for side events).
Acknowledgments
The authors would like to thank the management of Eni E&P PERF
department for their support and permission to publish this paper.
A special acknowledgment goes to Giorgio Girola and Graeme
Cuffy for their assistance when preparing this paper.
Identify the drilling
and completion
phases when may
initiate a blowout
Identify the
blowout scenarios
(including
flowpath &
barriers)
Build a fault tree
for each scenario
Estimate the
reliability/error
probability of all
Basic Events
Estimate the
probability of
blowout during
drilling and
completion
OPTIONAL STEP
Sensitivity analysis
IDENTIFICATION
OF D&C CRITICAL
PHASES
IDENTIFICATION
OF THE
BLOWOUT
SCENARIOS
FAULT TREE
CONSTRUCTION
EVALUATION
OF BASIC
EVENTS
PROBABILITY
OF BLOWOUT
SENSITIVITY
ANALYSIS
1.
2.
3.
4.
5.
6.
Fig. 11—Blowout probability analysis.
10
2011 SPE Drilling & Completion
References
Colbert, J.W. and Medley, G. 2002. Light Annular MudCap Drilling—
A Well Control Technique for Naturally Fractured Formations. Paper
SPE 77352 presented at the SPE Annual Technical Conference and
Exhibition, San Antonio, Texas, USA, 29 September–2 October. doi:
10.2118/77352-MS.
Grey, S. 1995. Practical Risk Assessment for Project Management. West
Sussex, England: Wiley Series in Software Engineering Practice, John
Wiley & Sons.
Masi, S., Molaschi, C., Zausa, F., Michelez, J., and Rossi, N. 2010. Key
Factors Sensitivity Analysis on Blowout Probability in Dangerous
Drilling Conditions Applying Different Technical Solutions. Paper SPE
133027 presented at the SPE Annual Technical Conference and Exhibi-
tion, Florence, Italy, 19–22 September. doi: 10.2118/133027-MS.
Sweep, M.N., Bailey, J.M., and Stone, C.R. 2003. Close Hole Circulation
Drilling: Case Study of Drilling a High-Pressure Fractured Reser-
voir—Tengiz Field, Tengiz, Republic of Kazakhstan. Paper SPE 79850
presented at the SPE/IADC Drilling Conference, Amsterdam, 19–21
February. doi: 10.2118/79850-MS.
Urselmann, R., Cummins, J., Worral, R.N., and House, G. 1999. Pressur-
ized Mud Cap Drilling: Efficient Drilling of High-Pressure Fractured
Reservoirs. Paper SPE 52828 presented at the SPE/IADC Drilling
Conference, Amsterdam, 9–11 March. doi: 10.2118/52828-MS.
Silvia Masi is currently the Field Operation Interfacing Manager
for Eni Zubair’s Zubair (Iraq) Project. Previously, she held the
position of Head of Well Engineering Department for Eni Italian
District in Ravenna. Masi has also worked as the Junior Drilling
and Completion Supervisor in Venezuela, the Operational
Drilling Engineer in Kazakhstan for the Kashagan Project, and
as the Senior Drilling Engineer at Eni headquarters in Milan.She
holds an MS degree in civil hydraulic engineering from UNICAL,
Cosenza, Italy, and an MS degree in petroleum engineer-
ing from Imperial College, London, UK. Claudio Molaschi has
been working with Eni E&P from 1981. Currently, he is technical
leader for Eni New Technology development and applications.
Molaschi has covered the positions of drilling supervisor on
land and offshore rigs and drilling engineer from 1991, work-
ing on high-pressure/high-temperature wells and on relief-well
operations,drain holes,and deepwater activities.He has been
drilling and completion manager for Eni activities in the Eastern
countries. Molaschi was one of Eni’s representatives for ISO
standards,and he is a teacher for directional drilling,horizontal
wells, and kick- and blowout-control methods. He has worked
on the development of Eni patent technologies such as LEAN
profile, compact wellhead, dual casing running, E-CD, and
ENBD.Fabrizio Zausa is technical leader of drilling technologies
with Eni E&P. He joined Eni E&P (formerly AGIP) in 1990. Zausa
spent 9 years working in rock-mechanics laboratories focusing
on rock-mechanics aspects applied to drilling activities,mainly
wellbore stability and rock/bit interaction. In the late 1990s, he
moved into the drilling department, working on drilling tech-
nologies and contributing to the development of many of the
new Eni drilling technologies such as lean and cross-lean pro-
file, ENBD, and ceramic centralizers. In the last 3 years, Zausa
started developing and managing a risk-management system
focused on drilling, completion, and production-optimization
activities. Jean Michelez is cofounder and managing partner
of Kwantis, a company dedicated to support decision mak-
ers to control future uncertainties on their investments. He has
been involved in project risk-management studies for more
than 15 years in various sectors (energy and utilities, infrastruc-
tures,transportation).Michelez is currently working for Eni E&P in
implementing an integrated risk-management system to sup-
port the dissemination of risk-management awareness and the
execution of complex risk assessments.
Информация о работе Managing Circulation Losses in a Harsh Drilling Environment: Conventional Solution vs.CHCD Through a Risk