Winter 2005 • GasTIPS 5

ROCK MECHANICS

D evelopment of more efficient andlower cost drilling technology willsignificantly increase gas production

by allowing economic exploitation of diffi-cult formations, such as deep, hard rockreservoirs. The estimated yearly cost to drillhard rock in the United States exceeds $1billion. Potential savings of $200 million to$600 million are possible if the penetrationrate in hard rock can be doubled while main-taining bit life, according to Tibbitts et al.

There is evidence the combination of per-cussion and rotary drilling techniques canpotentially provide significant improvementin rate of penetration in hard rock environ-ments (see review by Samuel, 1996). In addi-tion to faster penetration, other benefitsinclude the ability to use lower weight on bit,less contact time with rock and therefore less abrasion and longer bit life, improvedhole deviation control and generation oflarger cuttings allowing improved geologicinterpretation. These potential and theoreti-cal advantages for combined percussion androtary drilling, however, have not been con-sistently demonstrated in the field.

The fundamental rock mechanics proces-ses associated with combined percussion androtary drilling have not been fully definedand adequately modeled, and there are nopractical simulation tools available to helpdesign and optimize drilling operations. Thishas led to cost and reliability concerns, limit-ing the widespread application of percussion

drilling by industry. Terralog Technologies,with partial funding provided by the U.S.Department of Energy, Office of FossilEnergy, National Energy TechnologyLaboratory (NETL), is pursuing a compre-hensive research program to significantlyadvance fundamental understanding of thephysical mechanisms involved in combinedpercussion and rotary drilling. The projectteam, headed by Terralog Technologies andsupported by TerraTek, has extensive andunique experience and capabilities in funda-mental rock mechanics, geomechanical sim-ulation, and full-scale rock mechanics anddrilling experiments. The research programincludes three primary efforts:

• analytical investigations to develop animproved understanding of the funda-mental rock mechanics processes involvedin percussion drilling;

• development of advanced simulationtechnology for the drilling process tak-ing into account coupled structural, par-ticle and fluid flow mechanics; and

• investigation and validation of theseimproved characterization and modelingapproaches with full-scale laboratoryexperiments.

Background In rotary drilling (Figure 1a), the bit rotationproduces impact and shearing forces. Thethrust on drag bits provides a penetratingforce normal to the direction of movementthat breaks the bond holding the rock parti-cles together. The stress (energy) is built upuntil relieved by the formation of tension orshear fractures along the direction of thrust.While the impact creates compression, a cut-ting force perpendicular to the penetrating

Advanced Simulation Technology for Combined Percussion and Rotary Drilling and Cuttings TransportBasic research is required to advance simulation technology and help industry more economicallyrecover vast untapped gas resources contained in deep, hard rock environments.

By Michael Bruno, Gang Han and Claudia Honeger,

Terralog Technologies USA, Inc.

Figure 1. Rock damage process for rotary (a) and percussion drilling (b).

A B

6 GasTIPS • Winter 2005

ROCK MECHANICS

direction may cause tensile fractures thatextend from the bit tip to the rock surface atabout 80˚. Chip formation occurs discontinu-ously ahead of the drag bit, and the penetrat-ing and cutting forces oscillate during cutting.

In percussion drilling (Figure 1b), the bit and cutter oscillate axially to impact the rock, imparting compression loads.Developed by the Chinese more than 4,000years ago, percussion drilling first involvedthe raising and dropping of heavy piercingtools to cut and loosen earth materials. In1859 at Titusville, Pa., Col. F. L. Drake com-pleted the first oil well using a cable tool per-cussion-type machine. One of the earliestreports of percussion drilling technique wasdocumented in 1949. Since then, differentterms have been used, such as downholehammer, percussion hammer, percussive drill

and percussive-rotary drill.Percussion drilling with and with-

out rotation has been shown toimprove rate of penetration in somehard formations, such as siliceousgranite, sandstone, limestone anddolomite. In addition to a faster pen-etration, other benefits include theability to use lower weight on bit, lesscontact time with rock and thereforeless abrasion and longer bit life,

improved hole deviation control and genera-tion of larger cuttings to allow for improvedgeologic interpretation.

But the potential and theoretical improve-ments in drilling efficiency using combinedpercussion and rotary drilling have proved dif-ficult to achieve consistently in the field. Theproject’s objective is to significantly advancethe fundamental understanding of the physi-cal mechanisms involved in combined percus-sion and rotary drilling, and thereby facilitatemore efficient and lower cost drilling andexploitation of hard rock reservoirs.

A conceptual model of the drilling processis illustrated in Figure 2. We attempt to bet-ter characterize and simulate for fundamen-tal processes during drilling:

• drillbit penetration with compression,rotation and vibration;

• stress propagation and damage accumu-lation;

• rock failure and disaggregation; and • cuttings transport away from the bit face

and up the wellbore annulus.These are coupled physical processes, with

different physics related to the tool and bitmechanics, rock mechanics, and fluid andcuttings transport mechanics. A coupledsimulation system is illustrated schematicallyin Figure 3. The tool hits the rock face,imparting an impact and shearing load. Therock provides resistance to the tool motion.As the rock becomes damaged and fails, thesolid material becomes crushed and disag-gregated and adds cuttings to the mudflowstream. The mud system also may influencethe tool and bit movement through dampingand pressure resistance.

Tool and bit mechanics The tool and bit motion can be describedthrough the fundamental structural dynamicsequations relating force (F) to the combinedinfluences of mass (M) times acceleration(d2U/dz), damping (C) times velocity (dU/dz)and stiffness (K) times displacement (U).

Fz = Mz d2Uz/dz + Cz dUz/dz + KzUz (1)

F� = M� d2U�/dz + C� dU�/dz + K�U� (2)

There is one structural dynamics equationto define the axial motion, with the subscript z shown in equation (1), and one structuraldynamics equation to define the rotationalmotion, with the subscript � given in equation(2). In our coupled model simulation process,we solve the equations for a given time increment and impose displacements onto the rock interface.

Rock mechanicsThe rock mechanics system (Figure 4) reactsto that imposed deformation during the giventime increment. The elements in the rock system deform and may become damaged orfail during that time period, while at the sametime, the rock provides resistance to the tool

Figure 2. Conceptual drilling model and physical processes.

Figure 3. Coupled simulation software.

Winter 2005 • GasTIPS 7

ROCK MECHANICS

advance (through a resisting stiffness). In ourgeomechanical simulation, we apply a contin-uum model for the rock system, using theFLAC3D software from Itasca.

Stresses and strains are propagated dynami-cally through the rock based on fundamentalcontinuum mechanics equations. Applicationof the continuum of the momentum principleyields Cauchy’s equation of motion:

�ij, j + �bi = � (3)

where � is the mass per unit volume, b is thebody force per unit mass, v is rock velocityrelated to rock displacement (ui) throughaaaa , and the subscripts i and j are coordi-nate system indices.The usual summation con-vention is implied. The rock total strain incre-ment is a sum of elastic strain increment ,shear plastic strain increment and tensileplastic strain increment :

(4)

Rock is modeled as a Mohr-Coulombtype of elastoplastic material with strainhardening and softening:

(5)

The yield surface (f ) where rock starts tobehave plastically is defined by dynamicstresses (�ij) calculated from equation (3),plastic strain ( ) if the stressesexceed the yield surface and a hardeningparameter (�) that describes rock strengthbehavior with plastic deformation.

Stresses are related to strains through rock

properties, and these change during time asmaterial becomes damaged. When a finiterock element becomes sufficiently damagedso as to lose its inherent strength, it will dis-aggregate into discrete particles. Generally,this occurs across the entire bit face during agiven time interval, and the tool and bit pen-etrates to the next level of elements.

A conceptual one-dimensional model toillustrate this process is shown in Figure 5.The percussion tool oscillates, rotates andimpacts the top of the rock column, whichresists the movement by supplying a returnstiffness. Rock elements may reach their crit-ical strain limit and fail in tension, compres-sion or shear. The element disintegrates intocuttings and induces a small displacementjump in the tool motion as the bit assemblypenetrates deeper into the hole.

Coupled simulationTo simulate cuttings transport mechanics,Terralog has developed a coupled fluiddynamics and particle mechanics model thatcaptures not only macroscopic fluid behavior,but also the effect of solid particles on mud-flow and interactions among these particles.A particle suspended in a fluid is subjected toa number of hydrodynamic forces. Themomentum of a solid particle moving with afluid can be described as:

(6)

where Vp is particle volume, �p is its density,SSSis the particle velocity vector, and T, repre-

senting all forces between fluid and particle, isthe instantaneous stress tensor that must sat-isfy the Navier-Stokes equations. A set of con-stitutive models are developed to calculate var-ious forces from fluid-particle interactions,such as drag forces because of fluid viscosityand pressure difference across a particle, buoy-ancy forces and particle collisions. The influ-ence of pipe rotation on fluid transportation isconsidered through a solution of the circum-ferential velocity for the Taylor Couette exper-iment. Analytical solutions are available forlaminar, Newtonian flow (with patterns shownin Figure 6), while numerical solutions withthe help of computational fluid dynamicssolvers are required for turbulent and non-Newtonian flow conditions.

Particles generated through rock damageare modeled as individual spheres or clumps ofparticles to simulate irregular blocks andplates. They are introduced into the drillbitslots and then flow along the annulus. Figures7 and 8 illustrate two sample simulationsshowing cuttings transport in vertical and hor-izontal wells for varying bit penetration andmudflow conditions.

Figure 4. Continuum model for rock mechanics. Figure 5. Conceptual for bit and rock interaction and penetration.

Figure 6. Radial and axial flow and cut-tings transport in well annulus.

dvi

dt

8 GasTIPS • Winter 2005

ROCK MECHANICS

Cuttings transport mechanicsThe combined models defining the tool andbit mechanics, the rock mechanics and thecuttings transport mechanics are coupled tosimulate the complete drilling process.

First, the tool model generates an impactvelocity applied to the rock model. The rockmodel then calculates loading stresses anddetermines whether rock failure occurs, andif so, how much. The rock properties, includ-ing rock stiffness resisting further bit pene-tration, are then updated based on damageaccumulations. The bit model recalculates bitvelocity, which is in turn used by the rockmodel again. This cycle continues until bitvelocity is reduced to zero, which indicatesthe beginning of bit retreat after impact.

The rock model calculates how much vol-ume of rock will fail based on the implementedfailure criteria. This will trigger the cuttingstransport model to discretize the failed rockinto a number of spheres or clumps, whichinfluences fluid viscosity and therefore itsvelocity, and which also loads the fluid systemwith additional cuttings particles. Simulationsfrom the cuttings model are used to determinewhether failed rock can be efficiently carriedaway from the bit-rock impact surface.

Efforts and discussionThe next phase of this project will include lab-oratory testing and validation studies. We willfirst simulate and test single cutter assemblies,followed by full-scale mud hammer tools for avariety of loading conditions. TerraTek Corp.in Salt Lake City, Utah, will perform full-scalelaboratory testing under simulated downholeconditions at its drilling simulation facility. Atleast six drilling tests would be completedusing two rock types. During each drillingtest, parameters such as borehole pressure,rotation, axial load, weight on bit, rotary speed,hammer frequency and amplitude will be mea-sured. During each test, we will measure therate of penetration for varying rotation, ham-mer frequency and amplitude.

The fundamental rock mechanics processes

associated with percussion drilling have neverbeen fully defined and combined into a com-prehensive treatise. Critical processes includedynamic load and energy transfer from thereciprocating and rotating drillbit to the rock,geomechanical processes of dynamic damageand fracture at the bit face, and coupled fluidand cuttings particle flow around and awayfrom the bit. This project will advance thestate-of-the-art in each of these areas, provid-ing industry with fundamental algorithms tobetter characterize the drilling process, a revo-lutionary new type of simulation technologyand new laboratory experimental data. ✧

References1. Han, G., Bruno, M., Lao, K., PercussionDrilling in Oil Industry: Review and RockFailure Modeling, American Association ofDrilling Engineers Paper AADE-05-NTCE-59,AADE National Technical Conference &Exhibition, Houston, April 2005.2. Harpst, W.E., and Davis, E.E., RotaryPercussion Drilling, Oil and Gas Journal, 182-187, March 1949.3. Itasca Consulting Group, Inc., PFC3D: User’sGuide, Minneapolis, Minn., 1999.4. Itasca Consulting Group, Inc., FLAC3D:Theory and Background, Minneapolis,Minn., 2002.5. Melamed, Y., Kiselev, A., Gelfgat, M., Dreesen,D., and Blacic, J., Hydraulic Hammer DrillingTechnology: Developments and Capabilities,Journal of Energy Resources Technology, 122 (1):1-8, 2000.6. Pratt, C.A., Modifications to and Experiencewith Percussion Air Drilling, Society of PetroleumEngineers/lnternational Association of DrillingContractors Paper No. 16166, the SPE/IADCDrilling Conference, New Orleans, March 1987.7. Samuel, G.R., Percussion Drilling…Is it a LostTechnique? A Review, Society of PetroleumEngineers Paper No. 35240, the Permian Basin Oil& Gas Recovery Conference, Mildland, Texas,March 1996.8. Taylor, G.I., Stability of a Viscous LiquidContained between Two Rotating Cylinders,

Phil. Trans. R. Soc. Lond., 223: 289, 1923.9. Tibbitts, G.A., Long, R.C., Miller, B.E., JudzisA., and Black, A.D., World’s First Benchmarkingof Drilling Mud Hammer Performance atDepth Conditions, Society of PetroleumEngineers/lnternational Association of DrillingContractors Paper No. 74540, the IADC/SPEDrilling Conference, Dallas, Feb 26-28, 2002.10. Whiteley, M.C., England, W.P., Air Drilling Operations Improved by Percussion-bit/Hammer-tool Tandem, Society of PetroleumEngineers Paper No. 13429, SPE DrillingEngineering: 377-382, Oct 1986.

AcknowledgementsThis project is partially funded though U.S.Department of Energy, Office of Fossil Energy,National Energy Technology Laboratory con-tract DE-FC26-03NT41999.We thank Timothy Grant and John Rogers foruseful direction and discussion. Technical contri-butions and discussions with Professor MauriceB. Dusseault at University of Waterloo, Dr.Marian Wiercigroch and Joe Emans at theUniversity of Aberdeen, and Dr. Sidney Greenand Mr. Alan Black at TerraTek also are greatly appreciated.

Figure 7. A sample simulation of cut-tings transport with high viscosity andhigh flowrate in a vertical well.

Figure 8. A sample simulation of cut-tings transport with low viscosity andlow flowrate in a horizontal well.