Вы находитесь на странице: 1из 14

SPE 129126

Real Time Optimization of Drilling Parameters During Drilling Operations


Tuna Eren, SPE, Eni E&P, and M. Evren Ozbayoglu,* SPE, Middle East Technical University

* Now with The University of Tulsa

Copyright 2010, Society of Petroleum Engineers

This paper was prepared for presentation at the SPE Oil and Gas India Conference and Exhibition held in Mumbai, India, 2022 January 2010.

This paper was selected for presentation by an SPE program committee following review of information contained in an abstract submitted by the author(s). Contents of the paper have not been
reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material does not necessarily reflect any position of the Society of Petroleum Engineers, its
officers, or members. Electronic reproduction, distribution, or storage of any part of this paper without the written consent of the Society of Petroleum Engineers is prohibited. Permission to
reproduce in print is restricted to an abstract of not more than 300 words; illustrations may not be copied. The abstract must contain conspicuous acknowledgment of SPE copyright.

Abstract
Real time optimization of drilling parameters during drilling operations aims to optimize weight on bit, bit rotation speed for
obtaining maximum drilling rate as well as minimizing the drilling cost. The process is formation specific. A Statistical
method namely multiple linear regression technique is used for the drilling optimization methodology. An extensive literature
survey on drilling optimization was conducted for this research study. A model is developed for this purpose using actual field
data collected through modern well monitoring and data recording systems, which will be predicting the rate of drilling
penetration as a function of available parameters. The rate of penetration general equation is optimized for effective functions
at each data point. In order to optimize the parameters in the field, a computer network is required to be developed. The
computer network will keep the piped data directly from the data source, and continuously be collecting the new data to be
piped. A database present at the central computer will be continuously calculating the developed model parameters by means
of multiple regression technique and inform the team at the field. The field engineer will transmit the current drilling
parameters back to the central computer, and the headquarters will determine the new model parameters and optimum drilling
parameters by including the recently received information. Therefore, there will be a real-time-optimization process. It is
considered that this technique is going to be widely used in future drilling activities since it could reduce drilling costs and
minimize probability of encountering problems due to working with optimized parameters.

It has been found that drilling rate of penetration could be modeled in real-time environment as function of independent
drilling variables such as weight on bit, rotation speed of the string, mud weight, and formation characteristics. The ability to
have the drilling rate of penetration with respect to depth characteristically with certain parameters for specific formations on
real-time basis could bring new insights to the nature of drilling operations. Any departure from the trend could have
significant reasons. The study has also achieved one of its objectives, giving the optimized independent drilling parameters
found following statistical synthesis.

Introduction
The objective of optimizing drilling parameters in real-time is to arrive to a methodology that considers past drilling data and
predicts drilling trend advising optimum drilling parameters in order to save drilling costs and reduce the probability of
encountering problems. In order to achieve an effective optimization an extensive literature survey was conducted and findings
of previous researches were taken in consideration on how to develop a drilling optimization methodology in real-time
environment. The linear drilling rate of penetration model previously introduced by Bourgoyne and Young (1974) that is based
on multiple regression analysis has been utilized in real-time to:

o Achieve coefficients of multiple regression specific to formation,


o Have a rate of penetration vs. depth prediction as a function of certain drilling parameters.
o Determine optimum drilling parameters specific to the formation being drilled,

The following assumptions also as given by (Miska 1988) are considered to be satisfied so that the equations given in this
study are functioning properly:

o Bottom hole cleaning is effectively achieved.


2 SPE 129126

o The bit and Bottom Hole Assembly (BHA) assembly combination in use is one properly selected for the
formation being drilled.
o The formation interval being drilled is considered to be homogeneous.
o The rig and auxiliary equipment are efficiently functioning.

Data of three directional offshore wells located in the Mediterranean area were used to test the introduced methodology. Data
belonging to one well has been presented only in this paper.

Literature Survey
Numerous detailed research studies have been performed for the optimization of drilling activities, having an objective to
maximize footage drilled in time and minimize drilling costs. Drilling optimization could be achieved by pre-selecting the
magnitudes of the controllable drilling parameters. Authors believe that the research studies on drilling optimization are going
to continue to be an area of interest for scientists.

Most of the early studies performed in the literature have foreseen static drilling optimization process. The drilling parameters
were required to be investigated off-site due to lacking the opportunity of transferring data in real-time. Recent studies are
observed to do drilling optimization in real-time, however among the investigated references there is no work working with
statistical correlations in real-time environment.

Figure 1 gives the timeline of some important achievements in drilling and optimization history. In 1950s the scientific period
took place with expansion in drilling research, better understanding of the hydraulic principles, significant improvements in bit
technology, improved drilling fluid technology and most important of all optimized drilling. After 1970s rigs with full
automation systems, closed-loop computer systems, with ability to control the drilling variables started to operate in oil and
gas fields. In the mid 1980s operator companies developed techniques of drilling optimization in which their field personnel
could perform optimization at the site referring to the graph templates and equations. In 1990s different drilling planning
approaches were brought to surface to identify the best possible well construction performances (Bond et al. 1998). Later on
Drilling the Limit optimization techniques were also introduced (Schreuder and Sharpe 1999). Towards the end of the
millennium real-time monitoring techniques started to take place, e.g. drilling parameters started to be monitored from off
locations. A few years later real-time operations/support centers started to be constructed. Some operators proposed advanced
techniques in monitoring of drilling parameters at the rig site. In recent years drilling parameters are easily acquired, stored
and also transferred in real-time basis. Following the invent of the sophisticated and automated rig data acquisition
microelectronic systems linked to computers, a range of drilling optimization and control services started to take place
(Bradley 1987). With advanced smart computer systems drilling penetration rate and bit lives are optimized with performing
drill-off tests (Devereux 1998). Currently state-of-the-art, high-speed IP communication systems are developed functioning
with microwave broadband network, useful tool for oil and gas operations, enabling deployment of faster, more efficient
networks to the fields (Nagel 2009).

Figure 1: Drilling optimization timeline.


SPE 129126 3

One of the first attempts for the drilling optimization purpose was presented in the study of Graham and Muench (1959). They
analytically evaluated the weight on bit and rotary speed combinations to derive empirical mathematical expressions for bit life
expectancy and for drilling rate as a function of depth, rotary speed, and bit weight.

Maurer (1962) derived rate of penetration equation for roller-cone type of bits considering the rock cratering mechanisms. The
equation was based perfect cleaning condition where all of the rock debris is considered to be removed between tooth
impacts. A working relation between drilling rate, weight on bit and string speed was achieved assuming that the hole was
subject perfect hole cleaning circumstances. It was also mentioned that the obtained relationships were a function of the
drilling depth.

Galle and Woods (1963) are also one of the first researches who investigated the effect of best constant bit weight and rotary
speed for lowest cost; developing mathematical relations. Graphs and procedures for field applications were produced to
determine: best combinations of constant weight and rotary speed. They assumed a relation for the wear rate as a function of
time in relation to inverse ratio of bit weight to bit diameter. The given equation was limited with a load application of 10,000
lbf/in of bit diameter. They also published an equation giving a relation in between the tooth wear rate and rotary speed for
only milled tooth bits designed for soft formations. In their graphs drilling cost, footage, drilling hours and condition of teeth
and bearing of the dull bit may be calculated. The drilling costs were demonstrated to be reduced using the recommended
combinations of the drilling parameters.

Eckel (1968) performed microbit studies expressing the drilling rate exponentially as a function of pseudo bottom hole or near
bit-nozzle Reynolds number. The relation introduced was reported to be independent of bit weight and speed, and differential
borehole pressure, formation.

Reed (1972) developed a variable weight-speed optimal drilling model which was being solved using a Monte Carlo scheme
for the least cost per foot drilling. Linear (using the least squares technique) and curvilinear smoothing techniques were
developed to re-order and smooth the paths as well as decrease the cost per foot at intermediate states of the Monte Carlo
calculations. The developed method was proved valuable because of the rigor and the remarkable ease of constraint inclusion.
It was reported that optimal drilling was desirable not only for weight and speed in short, but also for the hydraulics, bit
selection, well design, mud treatment, solids separation.

One of the most important early studies performed in regards to optimal drilling detection was by Bourgoyne and Young
(1974). They constructed a linear penetration rate model and performed a multiple regression analysis of drilling data in order
to select the optimum bit weight rotary speed, and bit hydraulics. Minimum drilling cost was tried to be achieved, by attaining
maximum ROP which coincides with minimum cost approach if technical limitations were ignored. In their analysis they
included the effects of formation strength, formation depth, formation compaction, pressure differential across the hole bottom,
bit diameter, bit weight, rotary speed, bit wear and bit hydraulics. They found that regressional analysis procedure can be used
to systematically evaluate many of the constants in the penetration rate equation. They mentioned that multi-well data sourcing
should be obtained for the regression constant evaluations. They concluded that the use of relatively simple drilling
optimization equations can reduce drilling costs by about 10%.

One of the first real-time drilling optimization studies was performed by Simmons (1986). His findings were based on
application of all possible means at the time their study was published in order to get the drilling parameters optimized at the
rig site. It should be acknowledged that data transmission from drilling locations, usually remotely allocated has been started
by the beginning of the new millennium. He concluded that the combination of current technology and engineering, coupled
with real-time drilling optimization, will nearly always save on rotating hours, improve drilling efficiency, reduce possible
formation damaging affects and ultimately save on overall drilling expenditures.

Warren (1987) presented an ROP model that includes the effect of both the initial chip generation and cuttings-removal
process. The rate of penetration equation they derived is formed of two terms, working only with perfect hole cleaning
assumption. The first term defined the maximum rate supporting the WOB effect without tooth penetration rate, the second
term on the other hand considering tooth penetration into the formation. The equation was found to fit the experimental data
for both steel tooth and insert bit types.

Miska (1988) presented three governing differential equations; rate of penetration, rate of teeth wear, and rate of bearings
wear. It was concluded that the given equations could have been successfully used for predicting and optimization purposes
provided that some major conditions were satisfied. Three major conditions could be listed as: bottom hole cleaning is
adequate, rock-bit is properly selected to the formation drilled, and formation can be considered macroscope-homogeneous.

Maidla and Ohara (1991) tested a drilling model on offshore drilling data and compared the findings with the Bourgoyne and
Youngs model. Their objective was to be able to select bit, bit bearing, WOB and drillstring rotation to minimize drilling
4 SPE 129126

costs. Their finding was that ROP for successive wellbores in the same area could be predicted based on the coefficients
calculated from the past drilling data, resulting in cost savings. They concluded that the drilling model performances depended
on the quality of the data used to conduct the syntheses. They presented isocost, iso-ROP graphs for cost effective drilling
activities.

Ursem et al, (2003) demonstrated how an operator and a service company implemented the use of latest technology within the
scope of Real Time Operations Center (RTOC). They performed some pilot tests on exploration wells which revealed
communications improved interventions and made the advices much more clear limiting downtimes. They proved that critical
decisions are multidisciplinary in origin and common ground is needed for all parties involved. They concluded that it was
possible to influence unexpected outcomes in real-time instead of relying on an expensive lesson to be learned.

Rommetveit et al. (2004) developed a new and innovative drilling automation and monitoring system. The project was named
as Drilltronics. All available surface and subsurface drilling data was utilized to optimize the drilling process. One of the
introduced modules was Bit Load Optimization Module which modulated rotary speed and weight on bit and observed the
how respective changes effected the rate of penetration. They mentioned that preventing stick-slip occurrences by means of
activating one of the introduced algorithms increased rate of penetration by 15% to 30%.

Dupriest and Koederitz (2005) effectively used Mechanical Specific Energy (MSE) concept in evaluating drilling efficiency of
bits in real-time basis. They developed a system allowing the driller to continuously monitor MSE calculated through surface
measurements alongside with other normal mechanical drilling logs. Bit balling type of occurrences was easily identifiable
with the analysis.

Milter et al. (2006) worked on real-time data transfer from offshore to land support centers for drilling, well intervention and
production operations. The piped data was focused on quality by multi-disciplinary relevant personnel, not necessarily from a
fixed location, but anywhere with high speed internet communication. The optimization was conducted based on the judgment
of expert involved in the process based on his/her experience. They concluded that by means of the automatic surveillance by
means of real-time data transmission the number of unforeseen events was reduced. The number of well shut-ins was
decreased increasing the regularity in operations.

Monden and Chia, (2007) showed that decision making point could be moved from the data acquisition point to the Operation
Support Centres (OSC). It was mentioned that real-time connectivity from the rigs to the offices was becoming the norm for
many operations being performed recently. They concluded that, a data-centric approach to drilling optimization can
improve drilling efficiency. It was shown that in many cases the value of the OSCs can be measured, which would have direct
impact on time and cost considerations.

Iversen et al. (2008) adapted a system into the rig control mechanism in order to transfer signal from both surface and
downhole sensors in real-time basis to and from the rig in the North Sea. The introduced system worked dynamically for well
flow, drillstring mechanics, thermo-physical, solids transport, torque and drag models. Their test study proved that it is
possible to achieve a system which can calculate parameters and verify the quality of the safeguard calculations. For instance
the diagnosed wellbore stability issues and cuttings transport through trend analysis of friction between wellbore and
drillstring using a torque and drag model. They concluded that the system may alleviate some of the challenges like fluid loss,
stuck pipe, and pack-off tendencies, and suggestions for behaviour to be followed could be made in assisting to avoid them.
They also mentioned that the calculation functionality of the system is a function of data quality as well as it is a function of
correct system setup.

Drilling Rate of Penetration Model Theory


The drilling rate of penetration model adapted for this study is a function of eight independent variables as has been introduced
by (Bourgoyne and Young 1974). The reason behind selecting this model is due to the reason that this model is one of the
complete mathematical drilling rate of penetration model which has found greater acceptance within the drilling projects
among the models in use of the industry. Equation 1 gives the general linear rate of penetration equation, which is a function
of controllable and uncontrollable drilling parameters.

dF 8
= exp a1 + a j x j (1)
dt j =2

Schematically represented general linear rate of penetration equation is given in Figure 2, giving the relation of each function.
Coefficient a1 represents the effect formation strength. In other words it should mean that the less the magnitude for this
coefficient, the less the penetration rate and vice versa. It should be noted that this coefficient also includes the effects of
SPE 129126 5

drilling parameters not included in the mathematical model. Examples to those parameters not included could be drilled
cuttings, efficiency of the equipment/material, crew experience.

f1 function is defined as the formation strength, and it should have the same unit as rate of penetration, which is also known as
drillability of the formation. x1 is the dummy variable which is equal to 1 for every observation of rate of penetration. The
effect of formation compaction on rate of penetration is represented with two functions. The primary effect is normal
compaction, f2, and is given with an exponentially decreasing response with increasing depth. In other means this function
assumes increasing rock strength with depth due to the normal compaction. The secondary effect of normal compaction is
represented by f3. This function considers the effect of under-compaction in abnormally pressured formations. Within over-
pressured formations rate of penetration is going to end up with increased magnitudes. There is an exponential increase in
penetration with increased pore pressure gradients. f4 represents the function for pressure differential of bottom hole. The less
the pressure differential at the hole bottom, the less the penetration rate to be observed. True Vertical Depth corresponding
magnitudes have been used in calculations of the latter three functions. f5 represents the function for bit diameter and weight
applied onto the bit. Rate of penetration general equation is directly linked with the weight applied over the hole diameter.
This function is normalized for 4000 lb per bit diameter. (W/d)t is the threshold bit weight, which is known as the force at
which rock fracturing begins. f6 represents the function for rotary speed. Likewise the direct defined relation of bit weight on
penetration rate, the rotary speed is also set to have a direct effect.

Figure 2: General rate of penetration equation.

f7 represents the tooth wear function, and is calculated by means of determining the fractional tooth height at any instant
during the drilling process. In order to calculate the instantaneous respective tooth height, a bit record for similar bit type that
has been used within the same formation is useful to give the abrasiveness constant. Appendix A is dedicated to the
calculation of abrasiveness constant, and determination of fractional tooth height worn away. f8 represents the hydraulics
function taking into account the effects of bit hydraulics. Reynolds number function has been selected to be in the function
which includes the apparent viscosity term.

Two modifications to the methodology defined above are applied to account for weight normalization due to wellbore
inclination and additional bit rotation due to the existence of motors. First modification is to normalize the weight applied.
Since the well trajectories of deviated wells are going to be having varying inclinations, the applied weight can be normalized
to account for the deviation. Considering the geometrical relations the applied weight on the bit (W) could be normalized to its
vertical component (Wvert). Equation 2 gives the vertical weight on bit component.

W
Wvert = (2)
Cos ( )

The second modification is to account for additional bit rotation due to additional generated by the motor. The magnitudes are
determined based on the motor manufacturers definition which is a function of flow rate. The efficiency of the motors should
be considered and additional bit rotation calculation be adequately performed as a function of motor efficiency.
6 SPE 129126

Optimization of the Drilling Parameters


Drilling rate of penetration model is defined in order to accomplish the unique objective of this study that is to conduct real-
time analysis for drilling rate of penetration optimization. The study aims to optimize applied weight on bit and string rotation.
Optimization is to be formation specific. Multiple linear regression technique is used for the methodology of optimization,
which is a statistical approach. Multiple regression is used to find the parameters of an equation which causes the equation to
best represent the data (Mitchell 1995). A computer program has been prepared to find the coefficients of the model;
mathematically correlating rate of penetration with the controllable drilling and uncontrollable parameters. The task is to
acquire drilling data at a rig site network, pipe the acquired data to the operation center, and run the analysis and send feed
back to the rig site. Figure 3 gives the data transmission methodology of the process.

The data process technique is performed to the drilling data set to achieve the general linear rate of penetration equation
constants. The determined equation constants are used in the general equation to predict drilling rate of penetration as a
function of input drilling parameters.

Figure 3: Drilling optimization data transmission process.

Figure 4 gives the schematically represented workflow for multiple regression. The minimum number of dataset for a multiple
regression to solve a matrix of 8x8 is 5. The coefficients for the first 5 data set are first solved, and then the loop is repeated in
order to solve coefficients each time including one more set of data. The loop is continued until the number of requested data
sets is processed. A computer program has been written in order to do the necessary calculations.

The multiple regression technique is based on a regression model that contains more than one regressor variable (Montgomery
and Runger 2003). Multivariate data analysis is characterization of an observation unit by several variables (Davis 2002).
Multivariate analysis methods get affected for the changes in the magnitude of the several properties simultaneously. Multiple
regression considers all possible interactions within combinations of variables as well as the variables themselves.

Mathematical model for the penetration rate could be written as a function of drilling parameters as given in Equation Figure
8. The optimized weight on bit and bit rotation speed should lie within the operation window of their respective applicable
range.

dF W
= f vert , N , h (3)
dt db
SPE 129126 7

Figure 4: Multiple regression workflow.

Drilling cost per foot equation is defined to account for daily rig rate, bit cost, and timings required in the course of bit runs.
Equation 4 is the drilling cost formula.

C b + C r (t t + t c + t b )
Cf = (4)
F
After the necessary calculus the optimized equation for the vertical weight component for each diameter of bit size is as given
in Equation 5, after (Bourgoyne et al. 1986).

W W
a 5 H 1 + a 6
Wvert d b max db t
= (5)
d b Opt a5 H 1 + a6

In a similar manner the optimum bit speed can be expressed in the form as given in Equation 6 after (Bourgoyne et al. 1986).

W W
vert
db max d b Opt
[N ]Opt = 60 H (6)
t W
b 4
db max

The optimization process flow applied in this study was composed of using drilling parameters data collected at the mud
logging units. Figure 5 gives the flow chart of the process flow applied. The data was prepared accordingly in order to provide
suitable means for the multiple regression application based on the defined general rate of penetration equation. Pore pressure,
wellbore inclination, equivalent circulating density, mud rheological properties, and additional rotation of the bit due to the
motor for each data point was calculated and included in the database. Once the required dataset was stored, multiple
regression was run, and regression coefficients were determined. The determined coefficients are used in calculating the
predicted rate of penetration and determine the optimum parameters.

The same process flow performed in the scope of this study could be performed in real-time environment. The data could be
received using high tech communication means and data processing could be performed using todays high tech computers.
The output could be transferred back to the rig site. The role of driller is at utmost importance for the implementation of the
optimum drilling parameters determined following the process (Reinhold and Close 1997).
8 SPE 129126

Figure 5: Process flow Chart.

Representation of the results


Table 1 gives the minimum, maximum and average data input ranges of three important drilling parameters for the wellbore
section results of which are presented. The well was being directionally drilled in the target formation which was dominated
mainly with shale and sand in 12 of hole diameter.

Table 1: Data Input Range for WOB, String Speed, and ROP.
WOB, kip String Speed, rpm ROP, ft/hr
Minimum 8.2 71 13
Maximum 86.5 251 555
Average 36.9 182 178

One of the most important outputs for the data analysis is observed vs. predicted rate of penetration comparison chart, Figure
6. The chart includes three different data sets for the same database. The available data section is composed of about 900 data
points. The first set of data is the one without any correction. The second data set is the one only with weight on bit
normalization. There is a two fold improvement in the coefficient of determination magnitude (R2) between the first and the
second. The third data set is the interpolated form of data. The number of dataset has been reduced to about 100, by means of
interpolating the data points at a regular depth interval. All parameters have been sampled accordingly to have the most
representative magnitude in reference to what the actual data readings were. Also the bit rotation correction due to the mud
motor has been performed in the third data group. The coefficient of determination is with a magnitude of two fold in
reference to latter.

All_No_Correction (R2 = 0.1243)

WOBvert_Correction (R2 = 0.2788)

Interpolated_WOBvert_RPM-Corrected (R2 = 0.5740)

500

450
Predicted Rate of Penetration, ft/hr

400

350

300

250

200

150

100

50

0
0 100 200 300 400 500
Observed Rate of Penetration, ft/hr

Figure 6: Observed vs. predicted rate of penetration.


SPE 129126 9

The rate of penetration predictions are plotted in Figure 7. The scales are plotted; increasing from top to bottom and left to
right. The additional data set given in this graph in addition to Figure 6 is the interpolated actual rate of penetration. The best
match is observed in the data set in which the data has been interpolated, vertical weight on bit normalization and additional
rotation on the bit due to motor has been taken into consideration.

The optimized weight on bit and bit rotation are given in Figure 8. The optimized magnitudes are calculated based on
interpolated and corrected data and plotted in red color in the chart. They are the recommended magnitudes for two drilling of
the parameters. The optimized weight on bit (Figure 8a) calculation for the given dataset is the vertical component output,
given together with actual occurrences. So, whenever the determined optimum weight on bit is going to be applied the vertical
component is required to be transferred into axial form using the well inclination. It is observed that applying a lower weight
on bit would have resulted in optimization of drilling. The bit rotation (Figure 8b) optimization trend on the other hand was
calculated to have higher magnitudes in reference to what was actually applied. However with the introduction of ample
amount of data points, the trend is getting lower in magnitude and getting closer to actual magnitudes. The most recent
optimum weight on bit and bit rotation are the recommended values for the field in subject. It is observed both weight on bit
and bit rotary speed optimization the initial sections of the charts are with varying trends, however with inclusion of more data
the trends are getting stable.

The way forward


With the increasing number of development wells being drilled all around the world, a reliable optimization methodology
could be adapted initially as a pilot study at the Operation Support Centers, and then starting to the applications of the
introduced project. Each development well could be drilled at relatively lower costs as compared to the ones drilled earlier.

In the scope of a similar study new constants for tooth geometry could be worked on and new parameters for recent bits in use
of the industry could be introduced.
10 SPE 129126

Rate of Penetration vs Depth


Actual ROP Interpolated (Actual ROP)

All_No_Correction (Predicted ROP) WOBvert_Correction (Predicted ROP)

Interpolated_WOBvert_RPM-Corrected (Predicted ROP)

Rate of Penetration, ft/hr


0 50 100 150 200 250 300 350 400 450 500
2900

3100

3300

3500
Depth, ft TVD

3700

3900

4100

4300

4500

Figure 7: Rate of penetration comparison vs. depth for Well-4 12 1/4" hole-Fm3.
SPE 129126 11

WOB Optim ization Chart RPM Optim ization Chart

Bit Weightvert (kip/in.) (W/db)Opt, kip/in Rotary Speed, rpm (N)Opt, rpm

WOB, kip/in Rotary Speed, rpm


1.50 3.50 5.50 7.50 50 100 150 200 250 300 350
2900 2900

3100 3100

3300 3300

3500 3500

Depth, ft TVD
Depth, ft TVD

3700 3700

3900 3900

4100 4100

4300 4300

(a) (b)
Figure 8: Optimized Weight (a) on bit and rotation per minute (b).

Conclusions
Data quality is found to be very important for real-time drilling optimization. Data should be representative and accurate when
used as an input for multiple regression analysis. Findings indicate that instead of including all the available data points a
reduced number of data points when sampled, e.g. being the representative of the existing data trend could give much accurate
results. This is attributed to the spiky trend of the raw data. During real-time operations the data could be sampled at certain
intervals likewise the reporting methodology of directional drilling information, just before being piped from the rig site.

Another important point introduced is the consideration of the wellbore inclination in the analysis. The weight on bit was
converted into the vertical component of the weight on bit, to use the normalized magnitude of the applied weight on the bit. It
was observed that use of the normalized weight on bit resulted in higher accuracy of rate of penetration prediction and
consequently more accurate results for the optimum drilling parameters.

The study demonstrated that drilling rate of penetration could be predicted at relatively accurate levels, based on past drilling
trend. The optimum weight on bit and bit rotation speed could be determined in order to achieve minimum cost drilling. It is
believed that by means of effective communication infrastructures and thorough team efforts having efficient real-time drilling
optimizations based on statistical syntheses are not too distant.

Acknowledgements
Authors are thankful to Eni E&P for providing data, giving support and permission for the publication of this paper.

Nomenclature

Roman
a1 = formation strength parameter
a2 = exponent for the normal compaction trend
a3 = under-compaction exponent
a4 = pressure differential exponent
a5 = bit weight exponent
a6 = rotary speed exponent
a7 = tooth wear exponent
a8 = hydraulic exponent
Cb = bit cost, $
Cf = cost per drilled interval, $
Cr = daily rig cost, $
12 SPE 129126

D = borehole depth, L, ft
db = bit diameter, L, in
dn = equivalent bit nozzle diameter, L, in
f1 = formation strength function
f2 = formation normal compaction function
f3 = formation compaction function
f4 = pressure differential of hole bottom function
f5 = bit diameter and weight function
f6 = rotary speed function
f7 = tooth wear function
f8 = hydraulic function
F = distance drilled by bit, L, ft
gp = pore pressure gradient of the formation, M/L3, ppg
hf = bit tooth dullness, fractional tooth height worn away
H1, H2 = constants for tooth geometry of bit types
J2 = tooth wear composite function used to calculate fractional tooth wear
N = rotary speed, 1/t, rev/min
R2 = coefficient of determination
q = volumetric flow rate, L3/t, gpm
t = time, t, hour
tb =bit drilling time, t, hour
tc =drill pipe connection time, t, hour
tt =round trip time, t, hour
x = drilling rate of penetration independent parameter
x2 = normal compaction drilling parameter
x3 = under-compaction drilling parameter
x4 = pressure differential drilling parameter
x5 = bit weight drilling parameter
x6 = rotary speed drilling parameter
x7 = tooth wear drilling parameter
x8 = bit hydraulics drilling parameter
W = weight on bit, M, 1000 lbs
(W/d)max = bit weight per diameter where teeth fails instantaneously, M, 1000 lbs
(W/d)t = threshold bit weight at which the bit starts to drill, M, 1000 lbs
Wvert = vertical weight on bit component, M, 1000 lbs

Greek
= hole inclination, degrees
= drilling fluid density, M/L3, ppg
c = equivalent circulating mud density at the hole bottom, M/L3, ppg
= apparent viscosity at 10,000 sec-1, m/Lt, cp
H = formation abrasiveness constant, t, hours

Abbreviations
BHA = Bottom Hole Assembly
IP = Internet Protocol
MSE = Mechanical Specific Energy
RTOC = Real Time Operations Center
OSC = Operation Support Centre
WOB = Weight on Bit

References
Bourgoyne, A.T. Jr. and Young, F.S. 1974. A Multiple Regression Approach to Optimal Drilling and Abnormal Pressure
Detection. Society of Petroleum Engineers Journal: 371-384; Trans., AIME, 257. SPE-4238.
Bourgoyne, A.T., Chenevert, M.E., and Millheim, K.K. 1986. Applied Drilling Engineering. Textbook Series, SPE,
Richardson, Texas 2: 232 -240.
Bond D.F., Scott P.W., Page P.E., and Windham T.M.. 1998. Applying Technical Limit Methodology for Step Change in
Understanding and Performance. SPE Drill Eng 13 (3): 197-203. SPE-51181-PA. DOI: 10.2118/51181-PA.
SPE 129126 13

Bradley H.B.. 1987. Mud Logging. In Petroleum Engineering Handbook, Chap. 52, 52-1. Richardson, Texas. Society of
Petroleum Engineers.
Davis J.C. 2002. Statistics and Data Analysis in Geology, third edition. USA: John Wiley & Sons, Inc: 461.
Devereux S. 1998. Practical Well Planning and Drilling Manual. Tulsa, OK: PennWell Publishing Company.
Dupriest F.E., and Koederitz W. 2005. Maximizing Drill Rates with Real-Time Surveillance of Mechanical Specific Energy.
Paper IADC/SPE 92914-MS presented at the SPE/IADC Drilling Conference, 9-12 February 1986, Dallas, Texas. DOI:
10.2118/92194-MS.
Eckel J.R. 1968. Microbit Studies of the Effect of Fluid Properties and Hydraulics on Drilling Rate. Paper SPE 2244 presented
at the Fall Meeting of the Society of Petroleum Engineers, AIME, Houston, Texas, 29 September-2 October 1968.
Galle E.M and Woods A.B. 1963. Best Constant Weight and Rotary Speed for Rotary Rock Bits. Drill. & Prod. Prac., API:
48-73.
Graham J.W. and Muench N.L. 1959. Microbit Analytical Determination of Optimum Bit Weight and Rotary Speed
Combinations. Paper SPE 1349-G presented at the Fall Meeting of the Society of Petroleum Engineers, 4-7 October 1959,
Dallas, Texas. DOI: 10.2118/1349-G.
Iversen F.P., Cayeux E., Dvergsnes E.W., Ervik R., Byrkjeland M., Welmer M., Torsvoll A., Balov M.K., Haugstad E., and
Merlo A. 2008. Offshore Field Test of a New Integrated System for Real-Time Optimization of the Drilling Process. Paper
SPE 112744-MS presented at the IADC/SPE Drilling Conference, 4-6 March 2008, Orlando, Florida, USA. DOI:
10.2118/112744-MS.
Maidla E.E., and S. Ohara. 1991. Field Verification of Drilling Models and Computerized Selection of Drill Bit, WOB, and
Drillstring Rotation. SPE Drill Eng 6 (3): 189-195. SPE-19130-PA. DOI: 10.2118/19130-PA.
Maurer W.C. 1962. The "Perfect - Cleaning" Theory of Rotary Drilling. J. Pet Tech 14 (11): 1270-1274. SPE-408-PA. DOI:
10.2118/408-PA.
Milter J., Bergjord O.G., Hoyland K., and Rugland B. 2006. Maximizing Drill Rates with Real-Time Surveillance of
Mechanical Specific Energy. Paper SPE 92914-MS presented at the SPE/IADC Drilling Conference, 9-12 February 1986,
Dallas, Texas. DOI: 10.2118/92194-MS.
Miska S. 1988. Variable Weight and Rotary Speed for Lowest Drilling Cost. In Developments in Petroleum Engineering, Vol-
2, Chap. 9, 266-275. Houston, Texas. Gulf Publishing Company.
Miska S. ed. 1988. Developments in Petroleum Engineering, (Vol. 2), 266 - 275. Houston, TX: Gulf Publishing Company.
Reed R.L. 1972. A Monte Carlo Approach to Optimal Drilling. SPE J. 12 (5): 423-438. SPE-3513-PA. DOI: 10.2118/3513-
PA.
Mitchell, B. 1995. Advanced Oilwell Drilling Engineering Handbook & Computer Programs. SPE, Richardson, Texas 10: 156
-159.
Monden T., and Chia C.R. 2007. Field Verification of Drilling Models and Computerized Selection of Drill Bit, WOB, and
Drillstring Rotation. Paper SPE 14784-MS presented at the SPE/IADC Drilling Conference, 9-12 February 1986, Dallas,
Texas. DOI: 10.2118/110950-MS.
Montgomery D.C., and Runger G.C. 2003. Applied Statistics and Probability for Engineers, third edition. USA: John Wiley &
Sons, Inc: 482.
Nagel J. 2009. Real-Time communications for remote rig sites. World Oil 230 (9): 91-92.
Reinhold W.B., Close D.A. 1997. Drilling Optimization-The Drillers Role. SPE Drill& Compl 12 (1): 5-12. SPE-29365-PA.
DOI: 10.2118/29365-PA.
Rommetveit R., Bjorkevoll K.S., Halsey G.W., Larsen H.F., Merlo A., Nossaman L.N., Sweep M.N., Knut M.S., and Inge S.,.
2004. Drilltronics: An Integrated System for Real-Time Optimization of the Drilling Process. Paper IADC/SPE 87124-MS
presented at the IADC/SPE Drilling Conference, 2-4 March 2004, Dallas, Texas. DOI: 10.2118/87124-MS.
Schreuder J.C. and Sharpe P.J. 1999. Drilling The Limit A Key to Reduce Well Costs. Paper SPE 57258-MS presented at the
SPE Asia Pacific Improved Oil Recovery Conference, 25-26 October 1999, Kuala Lumpur, Malaysia. DOI:
10.2118/57258-MS.
Simmons E.L. 1986. A Technique for Accurate Bit Programming and Drilling Performance Optimization. Paper IADC/SPE
14784-MS presented at the SPE/IADC Drilling Conference, 9-12 February 1986, Dallas, Texas. DOI: 10.2118/14784-MS.
Ursem L-J., Williams J.H., Pellerin N.M., and Kaminski D.H. 2003. Real Time Operations Centers; The people aspects of
Drilling Decision Making. Paper IADC/SPE 79893-MS presented at the SPE/IADC Drilling Conference, 19-21 February
2003, Amsterdam, Netherlands. DOI: 10.2118/79893-MS.
Warren T,M. 1987. Penetration Rate Performance of Roller Cone Bits. SPE Drill Eng 2 (1): 9-18. SPE-13259-PA. DOI:
10.2118/13259-PA.
14 SPE 129126

APPENDIX-A Abrasiveness Constant

Instantaneous tooth wear could be calculated by means of finding the abrasiveness constant for a known bit record specifically
used in a formation. Formation abrasiveness constant is a parameter used to make necessary calculations when the bit in use in
practical terms, will become inefficient to drill ahead. The instantaneous tooth wear equation is given in combination of tooth
geometry, bit weight and rotary speed for roller cone bits, Eq. (A-1) (Bourgoyne et al. 1986).

W
H1
4 1 + H 2
dh 1 N d b m 2
=
dt H 60 W W 1 + H h (A-1)
2

d b m d b

The references by Bourgoyne (1974, 1986) give the tables for tooth geometry constants and (W/db)max. It is considered that a
specific study could be performed in order to update the respective values provided in the subject tables including values for
bit types introduced in recent years.

It is important to note that the given instantaneous tooth wear equation is going to be normalized at 60 rpm of bit rotation
speed and a constant bit weight of 4,000 lbf/in. The normalization magnitudes are selected accordingly for the specific
conditions in the scope of this study.

In order to be able to calculate the formation abrasiveness constant a tooth wear parameter is required to be introduced, which
is basically the reciprocal of a portion of the terms in composite tooth wear Eq. (A-1). The tooth wear parameter is symbolized
as, J2, as given in Eq. (A-2).

W W

H 1


db m db 60 1
J2 = N (A-2)
W
1 + 2
H
4
db m 2

If both Eq. (A-1) and Eq. (A-2) are combined, one gets Eq. (A-3):

tb hf

dt = J (1 + H
0
2 H
0
2 h)dh (A-3)

Integrating the achieved relation above yields the time the bit would have drilled with the fraction worn away,

h 2f

t b = J 2 H h f + H 2 (A-4)
2

The formation abrasiveness constant could then be written as in Eq. (A-5).

tb
H = (A-5)
h 2f
J2 hf + H2
2

Once formation abrasiveness constant is known, a tb, time of bit rotation as a function of predefined constants and tooth wear
as a fraction could be calculated solving Eq. (A-4). An arbitrary hf value could first be selected and until the calculated
working time of the bit is equal to the actual working time of the bit, a tooth wear fraction could be iterated.

Вам также может понравиться