You are on page 1of 9

V INGEPET 2005 (EXPL-2-MF-132)

MANAGING WELLBORE STABILITY: FROM THE MODEL


TO DRILLING MODELING

M. Frydman, J. Palacio, D. Lee ─ Schlumberger; J. P. Cassanelli ─ Pluspetrol

Abstract
Different oil operators have faced technical and economical challenges while drilling wells in the Andean
foothills. Drilling in a tectonic active setting creates problems ranging from influxes and stuck pipe incidents to
circulation losses and reservoir damage. Operations in Camisea experienced similar events, especially
wellbore instability. This instability was responsible for undesired lost time associated with excessive
reaming, hole-cleaning problems, expensive stuck pipe incidents and sidetracks.

To mitigate drilling incidents, a Mechanical Earth Model was generated at the beginning of the drilling
campaign. This model was used in several ways including, understanding problems that occurred in offset
wells, predicting potential problems in new wells, quickly reacting with informed decisions while drilling, and
optimizing drilling practices to minimize the economical impact. However, application of the model goes
beyond current drilling practices. By understanding factors that affect the stability of a formation, well design
can be optimized based on engineering and geological decisions while minimizing the geomechanical risks.

The concept of the Mechanical Earth Model (MEM) will be presented, and the factors and conditions that
affect wellbore stability described. The paper will explain the importance of calibrating the model with drilling
events to ensure consistency. Real time monitoring and updating adds tremendous value to the model by
reducing uncertainty. Examples of the MEM comparison with actual events will illustrate this validation. The
relationship between the mechanical earth model and pre-drill decisions will be shown, and how those
decisions can improve wellbore conditions and stability.

It is clear from the Camisea experience that this process can help in cost reduction of high-risk projects. The
next challenge is to apply a similar process to reduce the technical limit of low to medium-risk wells or
increase the production in certain reservoirs.

Introduction
Drilling wells in Camisea is technically and economically challenging. Wellbore instability, in this area, has
been responsible for costly stuck pipe incidents and in some cases multiple sidetracks. Stuck pipe is
responsible for lost bottom hole assembly (BHA) and considerable non-productive time (NPT) spent freeing
pipe, requiring additional wiper trips and hole cleaning. Minimizing non-productive time associated with
wellbore instability reduces the risk of major and catastrophic incidents and is required to complete wells on
time and within budget.

Pluspetrol, Schlumberger and other service companies worked together during planning and execution to drill
successful deviated development wells in this remote location in the tectonically active Peruvian Andes. This
team approach contributed to the reduction of drilling time while improving hole quality. The major drilling
problems avoided or resolved in achieving the savings included wellbore instability and poor bit and bottom
hole assembly (BHA) performance. The collaboration between the operator and service companies also
helped reduce other problems resulting from drilling fluid loss, hole cleaning, and reactive clays. All were
common problems in the field and had required multiple sidetracks to TD in previous wells.

Location
The Camisea (Lee et al., 2003) blocks (38/42) are located in the tectonic active foothills of the Peruvian
Andes (Figure 1). Initial discovery of the San Martin structures and drilling of exploratory wells occurred in the
mid 1980s and early 1990s.

The location within the rain forest, coupled with the economics of the region, makes access difficult, and
complicates any potential plans for road building, development of petroleum storage, transport, construction
of housing and other infrastructure.
In this remote jungle area, environmental issues and logistics constrain multiple pad locations for
development wells. Drilling of multiple deviated wells is done from a single pad, similar to offshore platform
drilling. Early in the planning, it was realized that directional wells would be more costly and incur additional
risks.
V INGEPET 2005 (EXPL-2-MF-132) 2

Figure 1. Camisea blocks located in the tectonic active foothills of the Peruvian Andes.

Description of the Three Phase Process


Initially, a Data Audit (Phase I) was conducted. The audit identified what information existed to characterize
previous drilling problems, diagnose the root cause of drilling problems and construct a MEM (Mechanical
Earth Model) needed for advanced well planning (see MEM section for details). After completing the Data
Audit and due to the high risk of the drilling plan, the second phase, a risk management process was
implemented for the San Martin field development. The combined results from the drilling hazards assessed
and the MEM were used to accelerate learning about the Camisea asset while helping reduce drilling risk
and cost. A mechanical earth model created from drilling, wireline, geological, and seismic data was used to
help determine root causes of the drilling problems in offset wells. Every problem identified was examined to
determine its impact on drilling the planned trajectory, and steps were established to reduce or eliminate
NPT.

To reduce the uncertainties of the initial model, real-time monitoring and model updating was performed
while drilling (Phase III). New knowledge obtained during drilling is captured in the MEM and helped to
optimize the Camisea field development plan. The MEM was verified and updated in real time using drilling
data (rig parameters, mud properties, drilling events, cavings type, extended leak off test, etc.), and LWD
tools. With this methodology, the look-ahead wellbore stability prediction was improved. As part of a wellbore
stability management, the mechanical earth model was used not only to determine the risks of a possible well
trajectory but also to design a directional trajectory that minimized wellbore stability problems. Predictability
accuracy of the mechanical earth model was improved by implementing inventive interpretation techniques
during MEM construction.

The Team
To evaluate the potential risk and plan for mitigation, a core team was designated to analyze previous drilling
problems, create a plan for reduced risk drilling, and monitor and update the plan while drilling. The team was
integrated by experts in geomechanics, drilling, geology, petrophysics, and seismic from both Pluspetrol and
Schlumberger. Synergy was created so the team worked together to generate Drilling Programs, specific
workflows and contingency procedures. In addition to the office team, dedicated engineers at rig site
monitored real time drilling signals and provide continual model updates. The office and rigsite combination
allowed the team to focus on risk management. A multidisciplinary end of well review allowed the team to
understand the drilling events from multiple perspectives, drilling and geomechanics.
V INGEPET 2005 (EXPL-2-MF-132) 3

As members of the team were geographically separated, a real-time monitoring system (Figure 2) was used
to keep everyone informed and to disseminate information. This system allowed secure posting of reports
and documents to share with other team members along with remote real-time monitoring of rig parameters
and downhole sensors via the Internet.

Figure 2. Real-time monitoring system used to keep the drilling


team informed and to disseminate information.

Drilling Risk Management


Prior to drilling the first well, the team analyzed previous drilling problems in the area and created a plan for
reducing drilling risk. Risk Management is key for the consistent delivery of performance where every
problem identified was examined to determine its impact on drilling the planned trajectory, and steps were
established to reduce the related non-productive time (NPT).

During the Data Audit, three clear goals were established:

• To gather and organize data information: Obtaining a catalog of drilling information, create an easy way
to identify the events in tables and graphically, and to store all information in a events ledger
• To perform a basic analysis focused in time distribution and drilling events: To visualize NPT distribution,
determine the drilling events impact (probability and severity) and establish the immediate suspected
causes.
• To establish a strategy for risk analysis, based upon recommendations from a multidisciplinary team.

It is important for Risk Management to count drilling events with an adequate Risk / Events ledger so all the
events can be properly identified, the root cause diagnosed, the risk categorized properly and the
corresponding mitigations documented. In several cases, a comprehensive Decision Analysis process was
needed to mitigate the risk. Key decision points identified and decision trees as well as scenario comparisons
were used to improve the selection process. The decisions points were specific depths or intervals; certain
moments when the real measurements differed from drilling models, or when those measurements exceeded
a preset threshold value. Detailed procedures or flowcharts were generated to determine the best decision to
take based on the specific well conditions (scenario identification). Contingencies were discussed during this
process. Management of Change procedures were implemented to ensure any alternative proved was
properly assessed.

A graphical risk assessment was displayed at the rigsite integrating the risk ledger information with the
Wellbore Stability Profile obtained from the MEM. This assessment was provided at the beginning of the well
to keep a big picture view of all the risks during the well. Every day, or at the beginning of special operations,
a risk forecast report was provided to remind the entire field team on the upcoming hazards and the specific
preventative and contingency actions for its mitigation

Two field engineers supported continuously to the Company Man in the risk management process and field
engineering tasks. These engineers are experts utilizing Drilling Models, and interpreting measurements
related to Drilling Mechanics, Pore Pressure and Wellbore Stability. Using the assessments, these engineers
communicate risks to the drill team and alert the team of concerns. They also provide recommendations
based on the planned actions and contingencies
V INGEPET 2005 (EXPL-2-MF-132) 4

The Mechanical Earth Model (MEM)


The Mechanical Earth Model is a numerical representation of the state of stress and rock mechanical
properties for a specific stratigraphic section in a field or basin (Plumb et al; 2000). Figure 3 is a schematic
representation of the MEM, showing creation starting with the framework model, to mechanical stratigraphy
through rock properties with the stress profile. From left to right, the profiles include: Poisson’s ratio (PR),
Young’s modulus (YME), unconfined compressive strength (UCS), friction angle (FANG), pore pressure (Pp),
minimum horizontal stress (Sh), maximum horizontal stress (SH), vertical stress (Sv), and the direction of
horizontal stress axes.

The framework model includes formation tops along with other key geological events such as faults and
unconformities. Both drilling events and the stress state must be viewed in the correct geological setting to
ensure a consistent MEM.

Structure

Mechanical
Stratigraphy
Rock Mechanical Parameters

Elastic Strength Earth Stress& Pore Pressure


Young’s
10
Modulus 100 0 Friction Angle 70
Stress Direction S h
0 Poisson’s 1 UCS 400 Stress 200 W N E
20 0
Ratio MPa

•Formation Tops 1.0

•Unconformities
•Faults

• Grain-Support fault ?

Facies PR Ε UCS Φ

• Clay-Support
Facies Pp Sh SH SV
Regional
Trend

• Deformation
Mechanisms
Figure 3. Schematic MEM, showing a framework model (left), mechanical stratigraphy (center) and rock
mechanical and earth stress profiles (right).

The second step in building the MEM is identifying the mechanical stratigraphy. Mechanical stratigraphy is
used to differentiate grain support rock from clay support rock. To determine the mechanical stratigraphy, a
petrophysical model was developed to identify matrix, clay and fluids from surface to reservoir. The model
was calibrated with analysis of clay minerals from core in offset wells. Principal clay minerals identified
included illite and smectite. This analysis together with a drilling event root cause analysis of previous drilling
events from offset wells identified the major causes of non-productive time along with possible failure
mechanics. In San Martin, the main wellbore stability concerns are reactive clays, plane of weakness failure,
fluid losses, unstable shales, and differential sticking. Table 1 illustrates the relationship between
components of the model and drilling planning and execution decisions.

Frydman et al. (2003) and Lee et al. (2003) present an extensive analysis on the MEM creation for San
Martin. Figure 4 shows the workflow used to build the complete MEM.
V INGEPET 2005 (EXPL-2-MF-132) 5

Table 1. Relationship between drilling decisions and the mechanical earth model (Plumb et al; 2000).

Framework
1 2 Model 3 4 5
Geomechanics & Mechanical Overburden Pore
Audit Drilling Stratigraphy Stress Pressure
Hazards

6 7 8 9
Stress Minimum Maximum Stability
Rock Strength Direction Stress, Sh Stress, SH Analysis

DrilMap

Figure 4. Pre-drill workflow.

Real Time Update of the MEM


To reduce the uncertainties of the initial model, real-time monitoring and model updating was performed
while drilling. The MEM was calibrated in real time using drilling data (kicks, losses, cavings type, extended
leakoff test, etc.), and LWD data. With this methodology, the look-ahead wellbore stability prediction was
continuously improved and uncertainty reduced.

While drilling, every new data measurement was used to update the MEM. Extended leak off tests were
performed at each casing shoe. From this test it is possible to extract a direct measurement of the minimum
horizontal stress (Sh) and upper limit estimation of the maximum horizontal stress (SH). As each hole section
was completed, LWD and wireline logs were processed and used to refine the MEM. After running an image
log, estimation for the maximum horizontal stress (Plumb et al, 1998) was computed at depths where
breakouts and induced fractures were identified. Figure 5 shows an example of this calibration. A
minimization method was used to define the stress field based on the available constraints.
V INGEPET 2005 (EXPL-2-MF-132) 6

Figure 5. Stress field calibration using: leakoff tests, image log (FMI) and DSI.

Real Time Verification and Decisions


The reliability of the mud weight forecast was validated in several cases where instability developed after
mud weights fell below the recommended range. Figure 6 shows a comparison of the drilling events with the
MEM prediction of minimum mud weight required for stability. In this figure, light blue represents the mud
weight required to avoid breakout while the actual mud weight used is shown in red. When the mud weight is
less than the collapse gradient, one can expect cavings. This ability to predict and verify failure gives much
more confidence to real time decisions. To mitigate risk in one section, a shallower casing point was
identified if the predicted instability between 1200 and 1300m could not be controlled with hole cleaning.
When that zone was reached, cavings and high torque required setting the casing shallower, as per
discussed several days before. This allowed reducing wellbore instability with higher mud weight in the
following section.

This example illustrates the value of model monitoring while drilling. Careful monitoring of actual vs. predicted
wellbore stability enables uncertainty in the model to be reduced, promoting accelerated learning and
providing confidence in the model when the well performs as the model has predicted. An accurate model
results in fewer unexpected wellbore instability incidents and lower costs related to non-productive time.
V INGEPET 2005 (EXPL-2-MF-132) 7

Figure 6. Comparison of the events vs. prediction.

During one of the LOT’s, a natural fracture behavior was identified from the time-pressure data. This was
immediately communicated to the drilling team and the Geophysics group was able to confirm the fracture
after reviewing the 3D seismic. Identification of the fracture helped the team determine the cause of cement
loses in this section and take appropriate actions to prevent opening other existing fractures.
To reduce tripping NPT, the BHA and bits were configured to drill each hole section in a single run. A rotary
steerable system helped to improve hole cleaning and hydraulics, while maintaining full control of the well
trajectory in hard formations.
During drilling, rig parameters and downhole sensor measurements were monitored in real time on a secure
Internet distribution site by team members in located worldwide. This capability enabled the team to make
fast, informed decisions based on shared information.
Team communication in both office and rig is fundamental for real time decisions. While developing a new
field, it is very common to find differences between the expected formation depth and the actual top. By
monitoring LWD logs while drilling, a sandstone section was identified 40m higher than expected. With input
from the Geology group, a decision was made to move the casing shoe. If the original design had been used,
the sandstone would have been exposed to a mud weight higher than its fracture gradient, with the risk of
total mud loss. The decision to change mud weight during drilling was made only after the effect on wellbore
stability had been evaluated. For the case when mud weight was reduced, mud loggers were advised to
report any increase in cavings and the rig crew was advised to improve hole cleaning. Drilling data always
confirms the accuracy of the wellbore stability forecast when cavings appear where predicted (Figure 7).
V INGEPET 2005 (EXPL-2-MF-132) 8

Figure 7. Comparison of the events vs. prediction.

Results
The teamwork approach based on risk management and the continue use of a Mechanical Earth Model,
LWD and other technologies helped reducing drilling time by improved performance and reduced NPT.

Lessons learned while drilling were used to improve drilling performance and reduce costs in the following
wells. This process resulted in an accelerated learning curve that reduced the time to drill a well to TD by half
(Figure 8) while reducing the well drilling cost in about 20% in the six well campaign.

Figure 8. Drilling to TD (Drilling/100m).


V INGEPET 2005 (EXPL-2-MF-132) 9

Conclusions
The mechanical earth model concept has been defined. In this challenging environment, building an accurate
pre-drill model and updating in real time can significantly reduce financial risk.

In economically challenging fields, timely access to fit-for-purpose information is vital for optimizing real-time
decisions on the rig. Information from the MEM can be delivered on short notice to support real-time
decisions on the rig using state of the art communication and data management techniques.

Mechanical earth models have the greatest value to well construction and field development when they are
integrated into a risk management and continuous updating processes. The MEM has been used to forecast
mud weights on subsequent development wells, accelerating the drilling learning curve.

Drilling in a difficult area requires geomechanics drilling expertise, teamwork, data management and
excellent communications among service companies and their client. This team effort in Camisea
represented a considerable reduction in cost compared to a previously drilled offset well.

References
Frydman, M., Palacio, J., Lee, D., Pidcock, G., Delgado, R.; and Cassanelli, J.P.; “No Drilling Surprises
Process and the value of the real time update, a case history in Camisea/Peru”; VIII Simposio Bolivariano
Petroleum Exploration in the Subandean Basins, Cartagena, Colombia, September 21-24, 2003.

Lee, D.; Cassanelli, J.P.; Frydman, M; Palacio, J; Delgado, R, Collins, B.; 2003; “Using a Dynamic
Mechanical Earth Model and Integrated Drilling Team to Reduce Well Costs and Drilling Risks in San Martin
Field”; SPE 84557; SPE Annual Technical Conference and Exhibition; Denver, Colorado, U.S.A., 5 – 8
October.

Plumb, R.A., Papanastasiou, P., Last, N.; 1998; “Constraining the state of stress in tectonically active
settings”; Eurock; 179 – 189, SPE/ISRM 47240.

Plumb, R.A. et al.; 2000; The Mechanical Earth Model Concept and Its Application to High-Risk Well
Construction Projects; paper IADC/SPE 59128 presented at the 2000 IADC/SPE Drilling Conference, New
Orleans, Feb. 23–25.

Prats, M.; 1981; “Effect of Burial History on the Subsurface Horizontal Stresses of Formations Having
Different Material Properties”, SPEJ, 658-62.

Warpinski, N.R.; 1986;. “Elastic and Viscoelastic Calculations of Stresses in Sedimentary Basins”, SPE
paper 15243 prepared for presentation al the Unconventional Gas Technology Symposium of the Society of
Petroleum Engineers held in Louisville, KY, May 18-21.

Acknowledgements
The authors thank Pluspetrol and all partners of Block 88 Upstream for their contribution and permission to
publish this paper.

You might also like