IADC/SPE 133724

Effect of Permeability Impairment by Suspended Particles on Invasion of Drilling Fluids Tung V. Tran, Faruk Civan, SPE, University of Oklahoma, and Ian Robb, SPE, Halliburton Services

Copyright 2010, IADC/SPE Asia Pacific Drilling Technology Conference and Exhibition This paper was prepared for presentation at the IADC/SPE Asia Pacific Drilling Technology Conference and Exhibition held in Ho Chi Minh City, Vietnam, 1–3 November 2010. This paper was selected for presentation by an IADC/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 International Association of Drilling Contractors or the Society of Petroleum Engineers and are subject to correction by the author(s). The material does not necessarily reflect any position of the International Association of Drilling Contractors or 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 International Association of Drilling Contractors or 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 IADC/SPE copyright.

Abstract This study experimentally and numerically investigates formation damage induced by suspended particles in the drilling fluid

and its effect on limiting their near-wellbore invasion. The study applies the NMR and X-Ray digital radiography tprovide

valuable insights into damage mechanisms along the formation and depth of invasion. Formation damage caused by drilling

fluids is one of the key factors for economic success in oil and gas field developments. The measured permeability reduction

obtained from laboratory test by injecting particulate drilling fluid in a representative core sample is used to determine

empirical parameters used to model the particle migration and deposition in porous media by means of a robust simulation of

the relevant processes. The study provides a concept to develop the ability to evaluate drilling fluids in term of their

formation damage potential.

Introduction

Argiller et al. (1999) analyzed the formation damage potential of water-based drilling fluids. They conducted

experiments with three water-based formulations. It was found that static filtration was mainly governed by external mud

cakes. They also stated that formation damage caused by water-based fluids can be avoided by optimizing the fluid loss

reducer and particle size distribution

Bailey et al. (1999) studied the particle invasion from drilling fluids. The particulate invasion was found to be one of

the primary mechanisms of formation damage caused by drilling fluids. Particles are forced into the formation generally

during the earlier stages of the filter cake growth. The experiments were done with KCl polymer fluid and different grades of

barite and carbonate weighing agents. They found that fine particles penetrated deeply into the formation and could not be

easily removed by back-flushing. Larger particles were observed to deposit near the surface of the injection point. The

permeability reduction was greatest in single-phase brine conditions.

Ding et al. (2004) studied near wellbore damage and natural clean-up of horizontal wells. Near wellbore properties

were altered by drilling fluid, fluid-fluid interaction, and fluid-filtrate invasion during overbalanced drilling operations. The

degree of formation damage was affected by many properties, and operating conditions. The permeability reduction factor

was correlated with flow rate. A rapid fluid loss, indicating as a spurt loss, when the drilling bit contacts the reservoir, was

observed. A deeper particle invasion of the internal filter cake decreased the flow efficiency while preventing the filtrate

2 IADC/SPE 133724

invasion. Serious loss of production occurred with damaging and non-optimized drilling fluids. The formation damage was

much severe with a water-based mud than an oil-based mud.

One key factor in the economics of drilling and completion projects is the performance of wells during their lifetime.

Productivity maintenance requires minimum formation damage. The basic feature that causes formation damage is particle

capture by the porous medium and the consequent permeability reduction (Nabzar et al. 1997; Roque et al. 1995). The trend

toward single well completion places additional emphasis on avoiding formation damage. During drilling and completion,

near-well bore permeability impairment can cause a significant reduction of well productivity; however, the damage may be

avoided if drilling/completion fluid conditions are properly selected.

Formation evaluation has been improved using well logging to provide accurate information on porosity and

formation fluid saturation. However, well logging can not provide a systematic estimation of permeability. Nuclear Magnetic

Resonance (NMR) can be used to determine porosity and permeability impairment during the core flood tests (Coates et al,

1999). In the past decate, X-ray computed tomography (CT) has provided a valuable core analysis tool (Withjack et al.,

2003).The philosophy behind the approach of this paper was to investigate various formation damage mechanisms developed

for a single-phase fluid system based on laboratory core tests using NMR and CT image technologies.

Experimental System and Procedure

The coreflood experimental apparatus included five main components (shown in Fig. 1): (1) an electrical pump

(Ruska), (2) an accumulator which contained fluid with suspended particles, (3) a Hassler-sleeve core holder, (4) back

pressure regulator, and (5) a computer based data acquisition system. The effluents were collected in a graduate cylinder. A

laser diffraction particle size distribution analysis system (LDPSDA) was used to measure the particle size distribution of the

particles in the core flood experiments. The LDPSDA is a system of multifunctional particle characterization tools. Its laser-

based technology permits analysis of particles without risk of missing either the largest or the smallest particles in a sample.

The laser diffraction technology is based on both Fraunhofer and Mie theories of light scattering. All the samples had to be

dried carefully prior to analysis since the system works only with dry samples.

Prior to a core flood test, the equipment was completely cleaned. Water-based drilling muds were prepared by

mixing water and bentonite with additives at desired concentrations. The fluid samples were aged for 48 hours. Fluid

viscosities were measured with a Chann 35 viscometer. Vacuum was used on the test fluid to make sure there were no

trapped air bubbles.

The Berea core samples used were 6.0 inch long and 1.0 inch in diameter. Prior to injection, the cores were

saturated with 3% KCl brine. Different barite particle sizes were added into drilling muds to study the effect of particle size

and concentration on permeability impairment. The flow rate was held constant at 90 cm3/hour.

Data Analysis

NMR and LDPSDA techniques were used to provide particle and pore size distribution of the formation before and

after being damaged.

Particle and Pore Size Distributions:

The volume-average mean particle diameter of barite suspensions used in this study ranged from 2 to 6 μm, while

the minimum particle diameter was 0.5 μm and maximum was 15 μm. Fig. 2 shows particle size distributions expressed in

volume frequency using LDPSDA system.

IADC/SPE 133724 3

When a water-wet rock is fully saturated with water, the T2 value of a single pore is proportional to the surface-to-

volume ratio of the pore, which is measured of the size of the pore. Therefore, the observed T2 distribution for all the pores in

the rock represents the pore size distribution of the rock.

The T2 NMR relaxation parameter is related to the pore surface, S, to volume, V, ratio by the following equation:

br TV

ST 22

11+= ρ ------------------------------------------------------------------------------------------------------------------------------(1)

Assuming a cylindrical geometry and ignoring the bulk term in the equation (1), the relationship reduces to:

rT r21

2

ρ= --------------------------------------------------------------------------------------------------------------------------------------(2)

where ρr is the rock surface relaxivity, r is the diameter of the pore body of the formation.

The T2 relaxation spectrum is a reflection of the variation of pore sizes. Hence, T2 NMR data was scaled to provide

pore size distributions using a relaxivity value of 25.5 μm/s and 16.9 μm/s for sample#1 and sample#2, respectively.

Fig. 3 shows the pore size distribution for the two samples#1(Ko =1,240 mD) and #2 (Ko = 265 mD) using the T2

relaxation data. The mean pore diameter of the 1,240 mD core was 26.1 μm while that of the 265 mD core was 15.7 μm. Fig.

4 is the SEM pictures of the two core samples at magnification of 100 times. The T2 distribution from NMR data offered a

reasonable estimate pore size distribution when the zone was 100% water-saturated. This information is very helpful for

reservoir quality and depositional environment evaluation.

Porosity Variations:

Fig. 5 represents the T2 distribution for the sample#1 at time t = 0, 5, and 18 hours of the core flood experiment. The

injection rate was a constant 36 cm3/hour. Barite (5.4 μm mean diameter) concentration was 5% by weight. Porosity was

estimated using equation (2). It was shown that the overall porosity reduced from 20.6% to 17.6% after 5 hours of flooding.

The porosity did not reduce further as fluid injection was continued. The porosity dropped only 0.1% after another 13 hours.

The cumulative porosity corresponding to relaxation time T2 are also shown as Fig. 6 to illustrate the total porosity of the core

sample at different times during the core flood test.

Several observations can be made from Fig. 5 and 6. The porosity was reduced due to particle capture and

deposition over five hours and then remained constant. The probability of particle being captured inside the core sample

increased as the pore diameter-to-particle size diameter decreased. In another word, the smaller pores would be blocked and

filled with particles before the larger pores. As particles were injected into the core sample, more small pores were blocked.

The flow was diverted to larger size pores and the capture probability rate decreased.

When all the capture sites were filled, no-more particles could be captured. They would be flushed through the large

pore paths. This explains the steady state in porosity established after 18 hours. Additional proof of establishing a steady state

during the particle injection test is examined in Fig. 7.

The effluent particle size distribution was analyzed using the LDPSDA method. The particle size of 5.4 μm mean

diameter was injected at the inlet. The particle size distribution in the effluent was smaller than in the inlet. This shows that

larger particles were captured, while smaller particles were flushed out. As injection proceeded, the outlet particle size

distribution shifted to larger diameter until the inlet particle distribution was almost matched. Hence, the steady state flow

was attained as the effluent particle size distribution and concentration become constant and identical to the inlet suspension.

4 IADC/SPE 133724

Because the pore size-to-particle size ratio is one of the key factors for the capture probability, it is essential to

identify different capture mechanisms during suspension injection. Fig. 8 shows the T2 distribution for the core flooding

experiment in core sample#2 (K0 = 265 mD). The same barium sulfite suspension was used (mean diameter Dp = 5.4 μm). It

was observed that porosity change was small after 5 hours of injection.

From the NMR T2 distribution data, it is also noticed that the peak at longer T2 is an indication of big pore

dominance. As the core was flooded with suspension for 5 hours, the particles filled in the larger pores to create many other

smaller pores inside. Hence, the T2 distribution response gave another peak at a shorter T2. The overall porosity drop was

from 17.5% to 16.9%. As time proceeded, no further change in porosity was observed. This indicates that no more particles

were deposited or captured inside the core. An external filter cake was observed after several hours of injection which

implied that the pore-throat blocking/bridging mechanism was the dominance.

Fig. 9 shows the T2 distribution for the sample#2 experiment at 0, 5, and 18 hours injection of barite suspension (2.2

μm-mean diameter). The peak T2 shifted to the left indicates that the pore sizes got smaller. The larger pores were

substantially filled as time proceeded and the porosity decreased from 17.6% to 14.2% and 13.1% after 5 hours and 18 hours

of the experiment, respectively. The distribution changes shown on Fig. 9 suggest that pore deposition and pore filling

mechanisms were the dominant processes.

The effects of particle size distribution on plugging mechanisms were also examined by injecting a wide range of

particle sizes (barite with mean-diameter of 3.6 μm) into a carbonate core sample#3 (high porosity but low permeability). The

results show that larger particles blocked pore-throats near the injection point and an external filter cake was built to prevent

further particle invasion. The porosity change was insignificant after 18 hours of injection. Fig. 10 represents the T2

distribution of the carbonate sample at 0, 5 and 18 hours of injection. The short and long traverse relaxation times, T2, of the

sample before flooding indicates a wide pore size distribution for this rock (range from 1.2 μm to 12 μm). The small

reduction in porosity is also explained by the presence of a thick external filter cake after the run. The filter cake acted to

prevent further invasion of larger particles into the core.

Fine invasion and simulation resuls:

X-ray computed tomography (CT) was used to estimate the axial porosities along the core samples. The CT-scanned

images were in gray-scale format. After thresholding, the digital images were converterd to binary mode. These images were

processed and analyzed using digital image subtraction technique. The fraction of depositing particles was then calculated.

Fig. 11 shows the processed CT images of the sectional core sample after 2.5 hours of coreflood experiment. The darker area

represented lower porosity. It was observed that porosity was lower at the injection point (Fig. 11-a) resulted from high

amount of particle captured. The porosity was highest at the end section of the core (Fig. 11-f).

The fine migration model developed by Civan and Nguyen (2005) was used to correlate the depositing particle

fraction along the core to estimate the total permeability reduction during the coreflood test. Table 2 shows the input and

adjustable parameter values that yielded a successful match to the experimental data. The volume fraction of particles

retained along the length of the core during the coreflood test is shown in Fig. 12. Overall permeability impairment was

simulated using Civan and Nguyen (2005) model matched with the measured values (Fig. 13).

IADC/SPE 133724 5

Conclusions

1. The experimental core plugging approach taken to investigate the governing damage mechanisms was straight-

forward and led to a fairly good prediction of permeability loss. This approach can help avoid overdesign in

drilling and completion operations.

2. When fines are injected into porous media, the particles are retained in the pores and permeability decreases.

As particle concentration increases, the capture probability increases and leads to external filter cake formation

and restriction of particle invasion into the core sample. Depth of particle invasion also strongly depends on

mud properties such as particle size and concentration.

3. Intergration of X-ray CT scans and Nuclear Magnetic Resonance ability to measure of porosity, pore size

distribution, and permeability of the formation can provide determination of the formation properties during

drilling/completion operations. Procedures have been developed whereby porosity distribution in porous media

can be determined accurately from this technique.

6 IADC/SPE 133724

Nomenclature

A, B, C Empirical parameters in Civan and Nguyen (2005) model

Dp Particle diamter,μm

cp Particle concentration, g/cm3

q Flow rate, cm3/hr

r Pore radius, μm

S Pore surface, μm2

T2 Relaxation time, μs

V Pore volume, μm3

Greek Symbols

μ viscosity of the fluid, cP

ρ average fluid density, g/cm3

ρr relaxivity, μm/s

ρp particle mass density, g/cm3

σi particle volume fraction

References

Argiller, J.F., Audibert, A., Longeron, D.G. 1999. Performance evaluation and formation damage potential of new water-based drilling formulas. SPE Drilling and Completion, 14: 266-273.

Bailey, L., Boek E., Boassen, T., Selle, O., Longeron, D. 1999. Particulate invasion from drilling fluids. Journal of Petroleum

Engineers, SPE 54762.

Coates, G.R, Xiao, L., Prammer, M.G. 1999. NMR Logging Principles and Applications, Gulf Publishing Co., Houston, TX, 234p.

IADC/SPE 133724 7

Ding, Y., Longeron, D.G., Audibert, A. 2004. Modeling of near-wellbore damage and natural clean up of horizontal wells drilled with water-based drilling fluids. Journal of Society of Petroleum Engineers: 252-264.

Nabzar, L., Coste, J.P., and Chauveteau, G. 1997. Water Quality and Well Injectivity. Paper 044 presented at the 9th European Symposium on Improved Oil Recovery, The Hague, 20-22 October.

Nguyen, V. and Civan, F. 2005. Modeling particle migration and deposition in porous media by parallel pathways with exchange. Chapter 11, Handbook of Porous Media, Second Edition, Vafai, K. (Ed), CRC Press, Taylor and Francis Group, Boca Raton, FL: 457-484.

Roque, C., Chauveteau, G., Renard, M. Thibault, G., Bouteca, M., and Rochon, J. 1995. Mechanisms of Formation Damage by Retention of Particles Suspended in Injection Water. Paper SPE 30110, presented at SPE European Formation Damage Conference, The Hague, 15-16 May.

Withjack, E. M., Devier, C., and Michael, G. 2003. The role of x-ray computed tomography in core analysis. Paper SPE

83467, presented at SPE Western Regional/AAPG Pacific Section Joint Meeting, Long Beach, California, May 19–24, pp. 1–12.

8 IADC/SPE 133724

Table 1: Summary of core samples properties

Sample No Permeability Porosity Lithology 1 1,240 mD 20.6% Sandstone 2 265 mD 17.5% Sandstone 3 28 mD 18.9% Carbonate

Table 2: Input and adjustable parameter values for numerical simulation

Parameter Value Parameter Value Lcore, cm 15.24 α 10 dcore, cm 2.54 n1 0.31 Dp, μm 5.6 n2 0.58 cp, g/cm3 0.2 δ, cm-1 0.0016 φ0 0.206 kp, cm-1 2.32 τ 1.41 kd, cm-1 1.54 εp,i 0 ke, Pa-1 0.0008 εnp,i 0 τcr, hr-1 0.000015 ρp, g/cm3 2.72 A 2.16 q, cm3/hr 90 B 0.67 σin, g/cm3 0.2 C 1.12 μ, cP 24 Kp,i, Darcy 0.00124 Knpi, Darcy 1.24

IADC/SPE 133724 9

Fig. 1: Core plugging experimental apparatus

0

2

4

6

8

10

12

14

2 4 6 8 10 12 14 16 18

Volu

me

dist

ribut

ion,

%

Particle size, μm

Dp = 2.2 μm

Dp = 3.6 μm

Dp = 5.4 μm

Fig. 2: Particle size distribution of three different barite samples using LDPSDA system

10 IADC/SPE 133724

0

0.01

0.02

0.03

0.04

0 10 20 30 40 50

Pore

-vol

ume

disi

trib

utio

n, μ

m-1

Pore size, μm

60

Dt = 15.7 μm

K0 = 265 mD

Dt = 26.1 μm

K0 = 1,240 mD

Fig. 3: Pore size distribution of the 265 mD and 1,240 mD core samples

Fig. 4: SEM of the 265 mD and 1,240 mD rock samples (Mag 100x)

IADC/SPE 133724 11

0

500

1000

1500

2000

2500

10-2 10-1 100 101 102 103 104

t = 0 hrt = 5 hrt = 18 hr

Am

plitu

de

Transverse relaxation Time (T2), ms

Sample#1

NMR φ = 20.6%

NMR φ = 17.5%

NMR φ = 17.6%

Fig. 5: T2 distribution plot for sample#1 at time t = 0, 5, and 18 hours of the injection of 5.4 μm barite suspensions

0

5

10

15

20

25

10-2 10-1 100 101 102 103 104

t = 0 hrt = 5 hrt = 18 hr

Cum

ulat

ive

poro

sity

, %

Transverse relaxation Time (T2), ms

Sample#1

NMR φ = 20.6%

NMR φ = 17.6%

NMR φ = 17.5%

Fig. 6: Cumulative porosity plot for sample#1 at time t = 0, 5, and 18 hours of injection of 5.4 μm barite suspensions

12 IADC/SPE 133724

0

2

4

6

8

10

12

14

2 4 6 8 10 12 14

0 hr5 hr18 hr

Volu

me

dist

ribut

ion,

%

Particle size, μm

Sample#1K

0 = 1240 mDt = 5 hours

Dp = 3.3 μm

t = 18 hoursD

p = 4.8 μm

INLETD

p = 5.4 μm

Fig. 7: Particle size distribution of the effluent at time t = 0, 5 and 18 hours of injection of 5.4 μm barite suspensions

0

500

1000

1500

2000

2500

10-2 10-1 100 101 102 103 104

0 hr5 hr

Am

plitu

de

Transverse relaxation Time (T2), ms

Sample#2

NMR φ = 17.5%

NMR φ = 16.9%

Fig. 8: T2 distribution for sample#2 at time t = 0 and 5 hours of injection of 5.4 μm barite suspensions

IADC/SPE 133724 13

0

500

1000

1500

2000

10-2 10-1 100 101 102 103 104

0 hr5 hr18 hr

Am

plitu

de

Transverse relaxation Time (T2), ms

Sample#2 NMR φ = 17.6%

NMR φ = 13.1%

NMR φ = 14.2%

Fig. 9: T2 distribution of sample#2 at time t = 0, 5, and 18 hours of injection of 2.2 μm barite suspensions

0

500

1000

1500

2000

10-2 10-1 100 101 102 103 104

0 hr5 hr18 hr

Am

plitu

de

Transverse relaxation Time (T2), ms

Sample#3NMR φ = 18.9%

NMR φ = 18.1%

NMR φ = 17.4%

Fig. 10: T2 distribution of sample#3 at time t = 0, 5, and 18 hours of injection of 3.6 μm barite suspensions

14 IADC/SPE 133724

a) b)

c) d)

e) f)

Fig. 11: CT images at different sections along the core at injection time t = 2.5 hours. Distance measuring from the injection point: a) 0.5’’, b) 1.5’’, c) 2.5’’, d) 3.5’’, e) 4’’, f) 5.5’’

IADC/SPE 133724 15

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0 0.2 0.4 0.6 0.8 1

CT scan at t = 2.5 hoursCT scan at t = 5.0 hoursSimulation at t = 2.5 hoursSimulation at t = 5.0 hours

Vol.

Frac

tion

of D

epos

ited

Part

icle

s, (ε

/φ)

Dimensionless Length, x/L

Fig. 12: Volume fraction of deposited particles along the core sample at time t = 2.5 and 5.0 hours

0

0.2

0.4

0.6

0.8

1

0 1 2 3 4 5

Experimental dataSimulation

K/K

o

Time, hours

Fig. 13: Permeabilily impairment of the core sample during the coreflood test