SPE 106573 Application of rock strength in drilling evalution

R. Nygaard University of Calgary, G. Hareland University of Calgary, University of Calgary

Copyright 2007, Society of Petroleum Engineers This paper was prepared for presentation at the 2007 SPE Latin American and Caribbean Petroleum Engineering Conference held in Buenos Aires, Argentina, 1518 April 2007. 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, as presented, have not been reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material, as presented, does not necessarily reflect any position of the Society of Petroleum Engineers, its officers, or members. Papers presented at SPE meetings are subject to publication review by Editorial Committees of the Society of Petroleum Engineers. Electronic reproduction, distribution, or storage of any part of this paper for commercial purposes 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 where and by whom the paper was presented. Write Librarian, SPE, P.O. Box 833836, Richardson, Texas 75083-3836 U.S.A., fax 01-972-952-9435.

Rock strength calculations The rock mechanical parameter that is most important when conducting drilling analysis is unconfined compressive rock strength (UCS)1. The UCS can be determined from Mohr Coulomb failure criteria. The Mohr-Coulomb failure criterion in terms of peak loads is given as:
S ' = UCS + S ' tan v h

(1)

Abstract Different sources can be used to develop rock strength information along the wellbore. Such strength information is important when assessing the stability of the wellbore, selecting mud weights and designing casing programs. However, there are other areas, especially in drilling, where rock strength information is applicable, but still underutilized. A methodology is developed to estimate drilling time and bit wear before drilling if rock strength is known. To estimate drilling time and bit wear, effects of other parameters like drilling parameters, well bore size and drilling bit design has to be normalized. This methodology has been used to estimate drilling time and estimate bit wear and further evaluate drilling performance while drilling. After drilling the additional information has been used to conduct a post analysis and transfer knowledge from well to well. The advantage of this methodology is it eliminates the effect of geological variability when comparing performance between wells and fields. Introduction Different sources can be used to develop rock strength information along the wellbore. Such strength information is important when selecting mudweight and assessing the stability of the wellbore, selecting mud weights and designing casing programs. Strength information is also used for completion and hydraulic stimulation design. However, there are other areas, especially in drilling, where rock strength information is applicable, but still underutilized. To obtain the rock strength along the well bore, logs, rock mechanical tests or even drilling data can be used. In this paper we address how to obtain this rock strength and some areas where rock strength has been provided to give valuable information for drilling purposes.

Where Sv is vertical effective stress, Sh is horizontal effective stress, and is failure angle. Effective stresses are defined the difference between total stresses and pore pressure.
S ' = S pp v v

(2)

Sv is the total stress, pp is the pore pressure. There are several methods to obtain UCS along the well bore. In most cases, the availability of data determines which methods to choose. Different methods for obtaining UCS are described below. Rock Mechanical laboratory testing Rock mechanical laboratory testing on preserved core samples are the most accurate method for calculating rock strength. Rock mechanical tests are conducted on cylindrical rock samples in a triaxial laboratory cell2. The triaxial cell can control and monitor the confining stress, vertical load and pore pressure. The vertical and horizontal strain (i.e. normalized deformation) of the rock sample is also recorded. First are the samples consolidated with isotropic confining stress. It means that the sample is loaded when vertical and horizontal stress is kept equal. After consolidation, in the shear phase, the vertical stress is increased until the sample reach peak strength. To obtain a failure criterion for a specific core depth several triaxial tests are conducted with different consolidation stresses. Figure 1 shows the shear phase results, in the vertical effective stress vs. horizontal effective stress space, of 3 North Sea sandstone triaxial tests3. The tests had confining stresses of 2, 5 and 10 MPa. In the shear phase the horizontal stress is kept constant while the vertical stress is increased. The Mohr Coulomb failure criterion calculates the UCS to be 76 MPa. Figure 2 shows the shear phase of 3 North Sea shale triaxial

[SPE 106573]

tests with confining stresses of 5, 10, 15 MPa. Low permeability materials like shale are tested undrained (i.e. pore pressure valve is closed). The change in pore pressure during the shear phase will also change the effective horizontal stress since the total horizontal stress is kept constant. The Mohr Coulomb failure criterion gives UCS equal to 8 MPa for the shale. The disadvantage with using triaxial tests results for drilling applications is the lack of available core material for testing and that testing is time consuming. Therefore triaxial tests are more often used to calibrate rock strength to other information sources like petrophysical properties or well logs. Rock strength from logs The use of sonic velocity logs to determine elastic properties of rock is well established. There is published several correlations between rock strength and sonic travel time 5-7. Sonic travel time measured on cores can be correlated with unconfined compressive strength derived from failure criteria8. The failure criteria were established from triaxial compressive tests on sandstone and shale cores. The sonic vs UCS were analyzed using Equation 37

example of cuttings derived rock strength log is given on Figure 78-9. Rock strength from ROP models The use of drilling data to predict rock strength has been developed based on Rate of penetration (ROP) models10-12. And specific ROP models have been developed for specific bit types like rollercone, poly crystalline compact bits (PDC) and natural diamond bits (NDB)10,11,13. The ROP models needs to include all the parameters that influence ROP, such as operational drilling parameters (e.g. WOB, RPM, flow rate, mudweight and type), bit types and wear and the rock formation properties. The rock formation properties that are included are lithology, abrasiveness, pore pressure and rock strength. For a previous drilled well all the above mentioned information is recorded except the rock strength. The data from a previous drilled well is used to generate a rock strength by the inversion of a bit specific rate of penetration models. The effects of operating parameters, bit design and wear, drilling hydraulics, mud rheology and pore pressure are normalized model. The inverted rate of penetration provides then a calibrated measure of the rocks strength under actual drilling conditions and simultaneously determines the wear characteristics of the bits used in drilling the relevant sections. This wear character is an evaluation of the bits performance while drilling various formation types and under a variety of operating conditions and includes, detailed bit geometry and its resistance to wear. Figure 9 shows an example of two ROP model based rock strength logs from the North Sea. The two logs correlates well. The ROP model based rock strength is also included in Figure 5 and 8. In Figure 5 it is good mach between the sonic rock strength and the ROP model based rock strength. Also the mach is good between the ROP model based rock strength and cuttings based rock strength in Figure 8. So, several sources can create similar rock strength logs which make it possible to obtain strength information in most cases. However, meter by meter based logs and more than one reference will give the best predictions. Therefore it is preferable to use the either log based or drilling based methods for obtaining rock strength logs. Application of rock strength in drilling simulation When planning a new well section it is valuable to know in advance the total drilling time, to be able to predict bit wear and further recommend optimized drilling parameters. Therefore a detailed analysis was initiated of a 12 inch section in the North Sea. The section was planned through the Hordaland and Rogaland shale formations which caps most reservoirs in the North Sea. Plan First, rock strength from a reference well had to be established to be able to analyze drilling time, bit wear and optimized parameters. The rock strength was based on a previous drilled vertical 12 section in the area. The recorded drilling data (ROP, WOB, RPM, Flow rate) is reported in Figure 10. The section was drilled from 1755m down to 2730m with two IADC 437 Roller cone bits. The total drilling time was 44 hours. The average ROP was 27 m/h. The

UCS =

1.00 +k k1 (t c k 2 ) 3

(3)

Where tc is travel time (s/ft), UCS is unconfined compressive strength (MPa) and k1, k2, k3 are experimental constants. In Figure 5 the shale and sandstone data are plotted with a best curve fit for the sandstones, shales and both lithologies combined. The best fit experimental constants for equation 9 are given in Table 1. When sonic logs are available these data can give a continuous strength profile along the wellbore as shown in Figure 5. However, in situation where very limited information exists the same approach can be applied by using average seismic travel times for each formation (Figure 6). In Figure 6, UCS derived from triaxial tests is also included. The mach between the rock mechanical tests and the velocity trend is good. However this approach will only give indicative results since there are no strength variations within the specific formations. Neutron density logs can also be used for deriving rock strength from logs. Porosity information can be obtained along the wellbore from Neutron density logs. Figure 7 shows that there is a good correlation between UCS and porosity for sandstones3. So if Neutron logs are available they can be used in similar fashion as sonic logs to obtain rock strength Rock strength from cuttings Another approach to overcome the time-consuming triaxial testing is to conduct rock mechanical tests on small cutting samples9. The rock mechanical tests are performed on a few mm-sized rock samples (cuttings) by a small spike, which is indented into the sample. When deformation occurs in the sample without increase in vertical load the critical transition force (CTF) is reached. CTF is correlated to UCS (unconfined rock strength) derived from triaxial laboratory tests. One

[SPE 106573r]

ROP model rock strength was calculated from the drilling data. The reference rock strength log represents the rock strength along the reference well path. The well path for the planned well hit the Rogaland formation at different depth which can be seen by comparing formation tops on Figure 9 and 10. Figure 10 shows the formation tops along the planned well path of the new well. This reference rock strength had to be shifted approximately 50m to mach the new well. The procedure assumes geological continuity and homogeneity i.e. the lithological properties of the individual formations will be broadly similar in the offsets and planned well. For deviated wells this process is more complex since the reference well path (at the inclination of the offset well) has first to be plotted against true vertical depth. Then the formation tops had to be adjusted for the new well site before it can be transposed along the new well path. In most instances due to the complexity of geological structures this may not be an exact representation of the strengths along the new well path but a fair approximation, except for an area with any major structural changes. However, in this area there exists detailed geologic information and the rock strength for the planned well given in Figure 10 based on the reference well should be a close approximation to the actual rock strength log for the new well. When the rock strength log is established for the new well simulations based on ROP models can predict drilling time, bit wear and optimized parameters. The simulations performed on the rock strength log were done in a commercial available drilling simulator14. The effect of any number of drilling parameters and operating conditions, within known limitations of operational parameters and the rig, are then evaluated. The performance of several bits can be evaluated to enable selection of the best bit and the corresponding optimal set of parameters to use while running the bit. Multiple scenarios were evaluated with variations in: WOB/RPM combinations. Changes in operating parameters as a function of rock strength variations. Bottom-hole assembly configurations. Bit hydraulics Bit types, impregnated, NDB, PDC, and roller cone This approach ensures that an optimal solution is obtained. Based on the performed simulation a set of recommended parameters on WOB, RPM and Flow rate was determined for a specific PDC bit. The PDC bit wear was prognosed to be an average of 1.2. The simulated bit wear of 1.2 is equivalent to an IADC reported bit wear on the rig of 1-1 or 1-2 (inner-outer wear on cutters). The total drilling time for the new section was prognosed to be 22 hours and with an average ROP of 40m/h. The prognosed drilling plan for the new well improves the drilling compared to the reference well. This prognosis (Figure 10) with the selected PDC bit and recommended parameters was sent to the rig. Follow up The drilling progress was continually evaluated during drilling. During this phase updates were conducted daily or several times a day to verify the predictions or if needed,

modify the model predictions. The effects the actual drilling progress had on the predicted performance, bit wear condition and drilling time prognosis where evaluated and informed back to the drilling operation. Practically speaking, Figure 10 was updated with the actual ROP, WOB, RPM and Flow rates as the drilling went on. The bit wear and rock strength were re-simulated based on this actual information and included as well. The information on Figure 10 provided to be a good common picture of the situation in discussions between the rig and the office. In Figure 11 the final update plot are shown where the prognosis of the well are overlaid with the actual outcome of the section. The rock strength log from the reference well is slightly higher than the actual well but the correlation is good. Also the drilling time prognosed was same as the actual. Average simulated bit wear was slightly less than reported and prognosed. The reason for less bit wear is the small reduction rock strength and reduced RPM. The overall result shows a very good mach between the prognosis and the actual outcome. This example shows that rock strength can be a very valuable tool for developing drilling prognosis. The quality of the prognosis will be determined on the lithological homogeneity and on the availability of data for constructing the rock strength log. Conclusions Comparable rock strength logs along the wellbore can be created from different sources such as rock mechanical laboratory testing, rock mechanical testing on cuttings, drilling data, sonic or neutron density logs, or even average travel times. The quality of the predictions will be highest for the meter by meter based logs. Strength logs can be a very valuable tool to use in combinations with ROP models for conductiong drilling analysis such as predicting drilling time, bit wear and recommend optimum parameters.

Nomenclature MD =Shale/sand porosity ratio (dimensionless)

t c
Flow k1, k2, k3 MD p pp RPM ROP Sh Sv Sv UCS

= Sonic Compressional Travel Time (sec/ft) = Flow rate (L/minute) = Experimental constants = Measured depth (m) = porosity (%) = Pore pressure (MPa) = Rotations per minute = Rate of penetration (m/h) = Horizontal effective stress (MPa) = Vertical total stress (MPa) = Vertical effective stress (MPa) = Unconfined compressive strength (MPa)

[SPE 106573]

WOB References
1. 2.

= Weight on bit (MPa)

Vertical stress (MPa)

160 140 120 100 80 60 40 20 0 0 2 4 6 8 10 12 UCS=76 MPa

3.

4.

5.

6.

7.

8.

9.

Unconfined rock strenght (MPa)

10. 11.

12.

13.

Goodman, R. 1989. Introduction to Rock Mechanics. John Wiley & Sons, New York. P562. Berre, T., Tunbridge, L. & Heg, K., 1995, The measurement of small strains and K0values in triaxial tests on clay-shales, 8th Int. Congress on Rock Mech., Tokyo, pp 1195-1199. Nygaard, R., Bjorlykke, K., Heg., K. and Hareland, G. The effect of diagenesis on stress-strain behaviour and acoustic velocities in sandstones. 1st Canada-U.S. Rock Mechanics Symposium, May 27-31, 2007, Vancouver, British Columbia, Canada. Nygaard, R., Gutierrez., M. and Heg., K. Shear failure and Shear fracturing in Shales and Mudrocks, Accepted in 1st Canada-U.S. Rock Mechanics Symposium, May 27-31, 2007, Vancouver British Columbia, Canada. Kasi, A. Zekai, S. & Bahsa-Eldin, H., 1983. Relationship between Sonic Puls Velocity and Uniaxial Compressive Strengths of Rocks. Proc. Of the 24th U.S. Symp. On Rock. Mech. Texas A&M University, 20-23 June 1983, TX, US: 409-419. Tokle, K, Horsrud, P. & BratIi, R.K., 1986. Predicting Uniaxial Compressive Strength From Log Parameters: 61st Ann.Tech. Conf. and Exh. of the Soc. of Petr. Eng., 5-8 October 1986. Orleans, LA USA. SPE15645. Onyia, E.C., 1988. Relationships between Formation Strength, Drilling Strength, and Electric Log Properties. 63rd Ann. Tech. Conf. Houston October 2-5 1988, TX, USA. SPE 18166. Hareland, G., Nygrd, R. Calculating unconfined rock strength from drilling data, Accepted in 1st Canada-U.S. Rock Mechanics Symposium, May 27-31, 2007, Vancouver, British Columbia, Canada. Zausa, F., Agip Spa; Civolani, L., Brignoli, M., Santarelli, F.J., 1997. Real-Time Wellbore Stability Analysis at the Rig-Site. SPE/IADC Drilling Conference, 4-6 March 1997. Amsterdam Netherlands. SPE37670. Warren, T.M. 1987. Penetrationrate Performance of Roller Cone Bits. SPE Drilling Engineering: 9-18. Hareland, G. & Hoberock, L.,1993. Use of Drilling Paramters to Predict In-Situ Stress Bounds. SPE/IADC Drilling Conf. 23-25 February 1983, Amsterdam Netherlands. SPE 25727. Rampersad. P.R., Hareland, G., & Boonyapaluk, P. Drilling optimization Using Drilling Data and Available Technology. 3rd Latin American/Caribean Petr. Eng. Conf. 27-29 April 1994 Buenos Aires, Argentina. SPE 27034. DDS user manual. Drops Technology AS. 2006.

Horizontal stress (MPa)

Figure 1. Results from rock mechanical tests on sandstone. Unconfined compressive strength is calculated to be 76 MPa.
Vertical effective stress (MPa) 40 35 30 25 20 15 10 5 0 0 5 10 15 Horizontal effective stress (MPa) 20 5 MPa 10 MPa 15 MPa UCS=8 MPa

Figure 2. Results from rock mechanical tests on shale. Unconfined compressive strength is calculated to be 8 MPa.
80 70 60 50 40 30 20 10 0 50 70 90 110 130 150 170 190 Sonic travel time (microsec/ft) Sandstone Shale Sandstone Shale Oniya

Tables and Figures Table 1. Experimental constants for rock strength correlation based on sonic logs. ______________________________________________ 2 k1 k2 k3 r ______________________________________________ Sandstone 0.0011 50 3.42 0.9 Shale 0.0013 50 -2.66 0.9 Combined 0.0012 50 0.22 0.9 _____________________________________________

Figure 4. Correlation of Unconfined compressive strength vs sonic travel time for shale and sandstone.

[SPE 106573r]

Strength (MPa) 0 1000 1250 1500 Depth (m) 1750 2000 2250 2500 2750 3000 25 50 75 100

Rock Strength (MPa) 0 1350 1375 25 50 75 100

Drilling strength Sonic strength

1400 1425 1450 1475 1500 Depth (m) 1525 1550 1575 1600 1625 1650 1675 1700 1725 1750 Cutting strength ROP

Figure 5. Unconfined compressive strength calculated based on sonic travel times and ROP model strength.
0 500 Average travel times strength 1000 UCS triaxial tests Depth (m) 1500 5 UCS (MPa) 10 15 20 25

Figure 8. Drilling data rock strength compared with rock strength from rock mechanical tests done on drilling cuttings.

0
2000

Rock Strength (MPa) 10 20 30 40 Well 1

50

400 500 Well 2

2500

600 700 Depth (m) 800 900

3000

Figure 6. Rock strength estimated based on seismic travel times.

140 Vertical stress at failure (MPa) 120 100 80 60 40 20 0 0 10 20 Porosity (%) 30 40 UCS = 137-34Ln(p) R = 0.91
2

1000 1100 1200 1300

Figure 9. Rock strength calculated based on ROP models.

Figure 7. Correlation between strength from rock mechanical tests and porosity.

[SPE 106573]

ROP [m/h]
Formation Tops 1550 0 20 40 60 80 100 125

RPM
150 175 200 0 5

WOB [Ton]
10 15 20 2000 2500

Flow [lpm]
3000 3500 4000 0

Avg. Bit Wear

1 2 3 4
0 10

Time (hours)
20 30 40 50

1650

1750

1850

Hordaland

1950

2050

Rogaland

Figure 10. Drilling data for reference section.

Rock Strength [MPa]
Formation Tops 1750 0 20 40 60 80 100

MD

2150

2250

2350

2450

2550

2650

2750

RPM
125 150 175 0 2.5

WOB [Ton]
5 7.5 10 3000 3250

Flow [lpm]
3500 3750 4000 0

Avg. Bit Wear

1 2 3 4
0

Time (hours)
10 20 30

1850

1950

Hordaland

2050

2150

MD

2250

Figure 11.

2350

2450

Rogaland

2550

2650

Figure 11. Drilling prognosis of new section.

[SPE 106573r]

Rock Strength [Mpa]

Formation Tops 1750 0 20 40 60 80 100 100

RPM
125 150 175 0 2.5

WOB [Ton]
5 7.5 10 3000 3250

Flow [lpm]
3500 3750 4000 0

Avg. Bit Wear

1 2 3 4
0

Time (hours)
10 20 30

1850

1950

Hordaland

2050

2150

Rogaland

MD

2250

2350

2450

2550 Reported wear Inner Cutter Wear - 1.0 Outer Cutter Wear - 1.0

2650

Figure 12. Prognosis and actual results for the new well.