You are on page 1of 9

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/311547878

Experimental Study of Drilling Fluid's Filtration and Mud Cake Evolution in


Sandstone Formations

Article  in  Journal of Energy Resources Technology, Transactions of the ASME · December 2016


DOI: 10.1115/1.4035425

CITATIONS READS

63 7,851

3 authors:

Chinedum Ezeakacha Saeed Salehi


University of Oklahoma University of Oklahoma
27 PUBLICATIONS   497 CITATIONS    188 PUBLICATIONS   2,593 CITATIONS   

SEE PROFILE SEE PROFILE

Asadollah Hayatdavoudi
University of Louisiana at Lafayette
135 PUBLICATIONS   502 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Polymeric LCM for Wellbore Strengthening View project

WASP - Wabamun Area CO2 Sequestration Project View project

All content following this page was uploaded by Chinedum Ezeakacha on 18 January 2018.

The user has requested enhancement of the downloaded file.


Experimental Study of Drilling
Fluid’s Filtration and Mud
Cake Evolution in Sandstone
C. P. Ezeakacha Formations
Petroleum and Geological
Engineering Department, In real time drilling, the complexity of drilling fluid filtration is majorly attributed to
University of Oklahoma, changing mud rheology, formation permeability, mud particle size distribution (PSD), fil-
Norman, OK 73069 ter cake plastering effects, and geochemical reaction of particles at geothermal condi-
tions. This paper focuses on quantifying the major effects as well as revealing their
S. Salehi contribution toward effective wellbore stabilization in sandstone formations. We con-
Petroleum and Geological ducted an extensive experimental and analytical study on this subject at different levels.
Engineering Department, First, we used field application and the results as guides for our experiments. We have
University of Oklahoma, considered both oil-based mud and water-based mud. Next, we optimized the mud parti-
Norman, OK 73069 cle size distribution (PSD) by carefully varying the type, size, and concentration of well-
bore strengthening material (WSM). Laboratory high pressure high temperature fluid
A. Hayatdavoudi loss tests were carried out on Michigan and Bandera Brown sandstones. The results from
Petroleum Engineering Department, these tests identify the formation heterogeneity and permeability in successful wellbore
University of Louisiana at Lafayette, stabilization. Filter cake permeability calculations, using the analytical model for linear
Lafayette, LA 70503 systems, were consistent with filtration rates, and the expected trend of permeability
declines with time. Finally, we investigated the evolution of internal filter cake and plas-
tering mechanism, using scanning electron microscopic (SEM) analysis. The test results
revealed a significant difference in the formation permeability impairment for the optimal
mud PSD and WSM blend. [DOI: 10.1115/1.4035425]

Keywords: wellbore stabilization, filtration, internal filter cake, external filter cake,
permeability, Michigan Sandston, Bandera Brown Sandstone, particle size distribution,
wellbore strengthening material

1 Introduction as shown in Fig. 1. An internal filter cake is formed when well-


sized mud particles deposit in and around the wellbore, in high
We have organized this paper as follows: Sec. 1 introduces
concentration. This builds rigid aggregates within the available
wellbore stabilization and its mechanism. Section 2 gives a
target pore network [4]. The plasticity of the low permeability
detailed account of the experimental procedure and design consid-
internal filter cake formed in this region assists in reducing the
erations, using different mud blends with wellbore strengthening
near wellbore permeability, lowering subsequent filtrate invasion,
material (WSM) and two different sandstone formations. Section 3
lowering the pore pressure increase behind the filter cake, and iso-
compares the effects of mud particle size distribution (PSD) and
lating formation fluid in a reservoir or nonreservoir section of the
formation permeability on fluid loss and filtrate reduction. In
Sec. 4, we use the microscopic analysis for assessing the mecha-
nism of internal cake evolution from the optimum WSM
blend, while Sec. 5 describes the external filter cake permeability
calculation using a linear system model.
In Sec. 1, we review the wellbore stabilization, also referred to
as mud cake wellbore strengthening. The process occurs when
drilling fluid particles, driven by a potential, penetrate and deposit
in and around the wellbore wall. This process forms a low perme-
ability filter cake [1–4]. The wellbore schematic in Fig. 1 shows
how fluid and particle transport in the axial and radial directions
build filter cake. Considering the variation in mud particles trans-
port rate and wellbore stabilization can be regarded as a time-
dependent process which occurs in different stages.
Spurt invasion, which is the first phase in the early stage of fil-
tration, occurs when drilling mud invades freshly exposed well-
bore wall in the shortest possible time [5]. Fractions of a second
after the spurt invasion, drilling overbalance pressure forces more
solid particles to invade the wellbore pores over a period of time.
This second phase appears to be the most critical in wellbore sta-
bilization because of the evolution of an internal filter cake region,

Manuscript received September 19, 2016; final manuscript received December 1,


2016; published online January 16, 2017. Editor: Hameed Metghalchi. Fig. 1 Wellbore stabilization schematic

Journal of Energy Resources Technology Copyright V


C 2017 by ASME MARCH 2017, Vol. 139 / 022912-1

Downloaded From: http://energyresources.asmedigitalcollection.asme.org/pdfaccess.ashx?url=/data/journals/jertd2/935440/ on 03/03/2017 Terms of Use: http://www.asme.org/ab


well. In addition, when a permeable wellbore is exposed to low Table 1 WSM combination
permeability internal filter cake, a positive change in the effective
stress occurs, which can render the wellbore strengthening by Blend Concentration
mud cake a successful process [6–8].
Mud filtration stage begins after the formation of a stable inter- 1 (OBM) 10% (86 g) fine CaCO3 þ 1% (7 g) graphite
2 (OBM) 10% (86 g) medium CaCO3 þ 1% (7 g) graphite
nal filter cake, which only allows the mud filtrate to invade the 3 (OBM) 10% (86 g) coarse CaCO3 þ 1% (7 g) graphite
formation, while setting the stage for further evolution of the 4 (WBM) 10% (86 g) fine CaCO3 þ 1% (7 g) graphite
external filter cake. The fluid drag effect causes the external filter
cake to undergo compaction by allowing the smaller particles to
penetrate deeper through it [9]. Although the larger sized particles
first form the base of the external filter cake, the high drag force which generates fluid loss response. With the forgoing, we
driving all particles to the cake surface deposits smaller particles employed an optimized drilling fluid particle size distribution
faster. In addition, increasing the overbalance pressure across a approach by carefully varying the type, size, and concentration of
very thin, oil-based-filter cake causes deformation of water drop- wellbore strengthening materials (WSM). The combinations of
let in the mud water phase. This reduces the external filter cake calcium carbonate and graphite shown in Table 1 can capture the
permeability [10]. At some point within the mud filtration stage, effects of both the concentration and size of the wellbore strength-
the growth rate of the external filter cake region decreases until ening materials (WSM) in all blends. Oil-based mud (OBM) was
the filtration rate reaches pressure equilibrium with the formation. used to formulate blends 1, 2, and 3, while blend 4 was for water-
This gives rise to the formation of a heterogeneous filter cake, based mud (WBM).
comprising various particle sizes, at the formation face, while per- Microscopic evaluation of the cake building mechanism of the
mitting only small particles to be exposed to the wellbore flow drilling fluid and its relationship with particle size was very
path [10]. important to our study. For this purpose, we also measured the
Mud filtration can occur in two conditions that vary more in the particle size distributions of various wellbore strengthening mate-
rate of filtrate loss over time. The first condition, which is static rial (WSM) in dry state (Fig. 2) and mud samples, using a laser
filtration, occurs when the drill string is out of the hole. There is diffraction PSD analyzer. For microscopic evaluation, we used
no mud circulation in the well, and the mud particle transport SEM analysis and watershed algorithm to obtain the average pore
toward the external filter cake is not interrupted. As the external throat diameter of the rock specimens. This method is further
filter cake grows with increased compaction and thickness, the fil- described by Salehi et al. [26]. We obtained an average pore throat
trate invasion rate decreases with time, which, in turn, prevents diameter of 200 lm for Michigan sandstone and pore throat diam-
further formation damage. The second condition is dynamic filtra- eter of 61.6 lm for Bandera Brown sandstone. Prior to selecting
tion, which occurs during drilling or mud circulation. The differ- several dry WSM combination, we combined these results with
ence between the two is that the hydrodynamic condition of mud that of Fig. 2. However, we report only the significant blends. In
particles, as observed in static filtration, is superseded by the mud addition to the mud particle size distribution and formation pore
shearing action at the wellbore wall, hence, particle accumulation, throat diameter analyses, we conducted further SEM imaging of
especially of larger sized particles, at the external filter cake sur- the WSM (Fig. 3). Our goal was to have a better understanding of
face is impeded. The outcome is a variation in the external filter the contribution of the WSM particle morphology to wellbore sta-
cake thickness. Eventually, a time-dependent equilibrium thick- bilization. The effects of the particle morphology of WSM, as
ness between particle deposition rate and particle erosion rate will well as the pore and grain morphology of the sandstone core have
be reached, which changes constantly, until a critical invasion rate been detailed in Sec. 4 (Table 2).
appears [10]. In drilling a well, we know that drill cuttings contaminate the
The use of lost circulation materials (LCM) for “proactive” drilling mud. Regardless of the constantly changing downhole
fluid loss control can alter the particle size distribution (PSD) of mud PSD due to cuttings contamination in real time drilling, the
drilling mud. This process significantly impacts the quality of the outcome from a PSD analysis cannot be overlooked, as it provides
filter cake regions, as discussed in the previous paragraphs. A an insight into the effectiveness of any WSM combination and in
carefully engineered mud particle size distribution (PSD) using what size ratio. The final mud PSD results as well as their effects
wellbore strengthening materials (WSM) and lost circulation in lowering the mud fluid loss will be discussed in Sec. 3 (Tables 3
materials (LCM) can bridge off target pore throats and tip of the and 4).
microfractures, thus, preventing further fracture growth [10,11]. We carried out filtration experiments on a laboratory scale
Considerable experimental studies conducted to simulate wellbore using the permeability plugging apparatus (PPA). We considered
filtration conditions and its effect on reducing fluid loss, empha- static filtration condition in a linear system where mud loss and
size the quality of filter cake as a function of cake thickness, cake particle transport in PPA are only in the axial flow direction. The
cohesion, and cake tightness; which quality being, again being a operating principle for this apparatus has been presented by Salehi
function of mud solids type and particle size distribution et al. [26]. They show that PPA can handle very high temperatures
[1,6,12–20]. Most of these studies account for only water-based and pressures up to 400  F and 5000 psi. In our DoE, we consider
mud with bentonite and lost circulation materials (LCM) with
standard, generic additives, a homogenous porous media, and
impracticably low wellbore pressure and temperature conditions.
In reality, wellbore stabilization is a function of time, geothermal
and geo-pressure conditions, formation heterogeneity, permeabil-
ity and porosity, mud type, particle size and concentration, as well
as the mechanism of bridging solids cluster, and other aforemen-
tioned variables [21–25].

2 Experimental Procedure and Design Considerations


The primary objective of the experiments performed in this
study is to characterize drilling fluid filtration which supplies the
driving potential and filter cake evolution in sandstone formations,
using different mud blends. During the design of experiment
(DoE) mud formulation was considered a key control factor Fig. 2 PSD of wellbore strengthening materials

022912-2 / Vol. 139, MARCH 2017 Transactions of the ASME

Downloaded From: http://energyresources.asmedigitalcollection.asme.org/pdfaccess.ashx?url=/data/journals/jertd2/935440/ on 03/03/2017 Terms of Use: http://www.asme.org/ab


Fig. 3 SEM analysis of the wellbore strengthening materials

Table 2 Percentage distribution of WSM

WSM d10 (lm) d50 (lm) d90 (lm)

Fine calcium carbonate 1.032 7.809 27.215


Medium calcium carbonate 1.916 13.954 62.828
Coarse calcium carbonate 2.43 24.37 118.872
Graphite 6.034 28.867 72.664

Table 3 Water-based mud recipe

Ingredients Composition

Water 280.58 Fig. 4 Permeability plugging apparatus (PPA) set up


Gel 25 g
Caustic 0.5 g
Lignite 4.0 g
Barite 80.41 these as noise factors which are environmental variables that
WSM concentration 1% graphite (7 g) þ 10% CaCO3 (86 g) affect fluid loss response. Based on field recommendation, we
Mud weight 11.5 ppg selected 250  F and a differential working pressure of 2000 psi as
operating limits.
The PPA setup, as shown in Fig. 4, can be used for running
multiple fluid loss experiments. To measure the effect of “spurt”
loss on the quality and purpose of WSM, we cut off the flow at
Table 4 Oil-based mud recipe various times between spurt invasion and 30-min time required
for filtration. This procedure enabled us to characterize time-
Ingredients Composition dependent pore plugging of various core specimen based on the
volume of filtrate, mud cake quality, and scanning election micro-
Diesel oil 158 ml scopic (SEM) analysis of the sandstone core specimens.
Primary emulsifier 5g Formation heterogeneity, porosity, and permeability are also
Secondary emulsifier 3g key performance design factors, as they have appreciable influ-
Surfactant 1g ence on wellbore stabilization. To account for these factors, which
Lime 10 g
were often neglected in the previous studies, we used Michigan
28% CaCl2 brine 98 ml
Viscosifier 5g and Bandera Brown sandstone cores, as shown in Fig. 5. For our
Barite 127 g PPA experiments, we cut quarter inch thick disk from heterogene-
WSM concentration 1% graphite (7 g) þ 10% CaCO3 (86 g) ous core, as presented Fig. 5, instead of using homogenous aloxite
Mud weight 11.5 ppg disk as previously were used by others. The choices of sandstones
will capture a close-to-real field plastering scenario and bridging

Journal of Energy Resources Technology MARCH 2017, Vol. 139 / 022912-3

Downloaded From: http://energyresources.asmedigitalcollection.asme.org/pdfaccess.ashx?url=/data/journals/jertd2/935440/ on 03/03/2017 Terms of Use: http://www.asme.org/ab


61.6 lm and 200 lm, respectively. We would analyze the mud
PSD results, in relation to the sandstone pore throat diameters.
With respect to the 61.6 lm Bandera Brown sandstone, the mud
PSD result for blend 1 shows that 91.46% of mud particles are
below 63.25 lm, while 4.03% of mud particles are between
56.4 lm and 63.25 lm. The mud PSD result for blend 2 shows
that 83.22% of mud particles are below 63.25 lm, while 5.26% of
mud particles are between 56.4 lm and 63.25 lm. Blend 2 shows
a slightly better potential for bridging the target pore throat diame-
ter of 61.6 lm, as compared to blend 1. But we can only corrobo-
rate this observation from the filtration experiments as well. The
mud PSD result for blend 3 revealed that 71.24% of mud particles
are below 63.25 lm, while 2.02% of mud particles are between
56.4 lm and 63.25 lm. The mud PSD result for blend 4 shows
that 95.55% of mud particles are below 60.97 lm, while 0.81% of
mud particles are between 56.7 lm and 60.97 lm. The percentage
compositions of mud particles between 56.4 lm and 63.25 lm in
blend 3, and between 56.7 lm and 60.97 lm in blend 4, are con-
siderably smaller than the percentage compositions of mud par-
ticles in blends 1 and 2. Hence, higher fluid loss is expected for
blends 3 and 4.
In relation to the 200 lm pore throat for Michigan sandstone,
the mud PSD result for blend 1 shows that 100% of mud particles
are below 200 lm, while 0% of mud particles are between
178.25 lm and 200 lm. The mud PSD result for blend 2 shows
Fig. 5 Michigan and Bandera Brown sandstone cores and disk that 99.86% of mud particles are below 200 lm, while 0.04% of
specimen mud particles are between 178.25 lm and 200 lm. Blend 3, with
coarse calcium carbonate, has large particles in excess of 200 lm
up to 1000 lm, constituting approximately 21.36% of mud par-
ticles in this blend. The mud PSD result for this blend revealed
that 78.64% of mud particles are below 200 lm, while 0.6% of
mud particles are between 178.25 lm and 200 lm. The mud PSD
result for blend 4 shows that 99.53% of mud particles are below
210 lm, while 0.13% of mud particles are between 195.30 lm and
210 lm.
These results indicate a promising pore throat plugging in Ban-
dera Brown sandstone, than in Michigan sandstone, since there is
little or no percentage composition of mud particles available to
effectively bridge off the Michigan sandstone pores. These obser-
vations will be substantiated from the results of the filtration
experiments. In general, mud particles outside a target pore throat
diameter will not partake in forming an internal filter cake, but
Fig. 6 PSD of different mud blends may act as external filter material [27–29].
We verified the performance of these blends from the experi-
effects of mud particles, which covers a wide range of permeabil- mental filtration results, which we present in Fig. 7. We make sev-
ity in different formations. To prepare the core sample disks, we eral observations regarding the above mentioned results. First, as
used a rigid 2.5 in. coring bit and core cutting machine to cut two expected, the higher the permeability of formation, the higher will
core specimens from each slab, as shown on the top of Fig. 5. be the total amount of filtrate invasion. The results for all the
Using a concrete saw, we sliced the cores into disks to a thickness WSM blends confirmed lower filtration in the Bandera Brown
of 0.25 in. The disk diameter and thickness are boundary con- sandstone. This is an outcome of an early but stable internal filter
straints set by the PPA chamber size. cake formed by mud particles relative to Bandera Brown pore
throat diameter and its low permeability. Similar observation has

3 Effects of Mud PSD and Formation Permeability on


Fluid Loss
As mentioned earlier, mud formulation is a key design factor
affecting fluid loss response and the importance of its PSD to fil-
tration process cannot be emphasized more. Figure 6 shows the
mud particle size distribution (PSD) of four different mud blends.
Blend 1 is composed of 1% graphite mixed with 10% of fine cal-
cium carbonate in OBM. Blend 2 is composed of 1% graphite
mixed with 10% of medium calcium carbonate in OBM. Blend 3
is composed of 1% graphite mixed with 10% of coarse calcium
carbonate in OBM. Blend 4 is composed of 1% graphite mixed
with 10% fine calcium carbonate in WBM. Figure 6 shows a mini-
mum and maximum particle size of approximately 0.4 lm and
100 lm, respectively, for blend 1, 2, and 4. Recall that the pore Fig. 7 Cumulative fluid loss comparison over a period of
throat diameters for Bandera Brown and Michigan sandstones are 30 min

022912-4 / Vol. 139, MARCH 2017 Transactions of the ASME

Downloaded From: http://energyresources.asmedigitalcollection.asme.org/pdfaccess.ashx?url=/data/journals/jertd2/935440/ on 03/03/2017 Terms of Use: http://www.asme.org/ab


been reported in the literature [30]. Their study revealed that in a
plot of fluid loss rate versus formation permeability, formation of
an internal filter cake begins during spurt invasion, followed by
decreasing fluid loss, with an increase in formation permeability.
Additionally, it is evident from Fig. 7 that WSM size affected
the fluid loss in an orderly manner for OBM blends 1, 2, and 3.
The PSD analysis for blend 3 predicted higher filtrate invasion
due to larger size particles in excess of 200 lm and lower percent-
age composition of mud particles required to plug 61.6 lm and
200 lm pore throats. Interestingly, the higher filtrate value in
blend 4, with the same WSM size and concentration, as compared
to blend 1, is clearly due to the fluid type; where a water-based
mud will certainly experience higher filtrate loss than a properly
formulated oil-based mud.
Based on this important finding, we report that blend 1 is the
optimal mix for this study, as it reveals the lowest values of filtrate
invasion for both Michigan and Bandera Brown sandstones.
Although the mud PSD, in Fig. 6, suggested blend 2 as having a
slightly better chance of plugging Bandera Brown sandstone pores
than blend 1, the filtration results reveal only 0.6 ml difference
between the two blends. The small volumetric reading could be
attributed to natural experimental error. Compared to other blends,
the high percentage composition of mud particles in blend 1,
between 56.4 lm and 63.25 lm, is largely influenced by the fine
calcium carbonate and graphite. In this case, WSM packing and
agglomeration within the pore throats is highly probable due to
the increased surface area of the fine calcium carbonate interact-
ing among themselves and the walls of pore throat. In addition,
the irregularity and roughness of graphite surfaces causes a highly Fig. 8 Experimental rock disks with their corresponding SEM
probable reason for well bore strengthening through a newly dis- images before filtration experiment
covered mechanism, which we refer to as particle-pore interlock-
ing and particle–particle (CaCO3 and Graphite) interlocking. This
combined surface interaction increases the cohesive strength of before filtration, as shown in Fig. 8 (left), reveals that a good
the internal filter cake, within the sandstone. Moreover, these particle–pore interlocking can be achieved during wellbore stabili-
newly found properties serve to increase the cohesive strength of zation. This also explains the lower fluid loss result in this rock
the external filter cake for resisting erosive shearing stresses specimen. The SEM image of Michigan sandstone, as shown in
caused by the circulating fluid on the interface of external filter Fig. 8 (right), revealed a smoother morphology, which in addition
cake and the fluid. to its higher permeability, may not promote particle–pore inter-
locking, hence, it yields higher fluid loss when compared to Ban-
dera Brown sandstone.
4 Microscopic Analysis Using SEM For better understanding and delineating the depth particle
Characterizing filter cake is very critical to understanding dril- invasion (in this case disk thickness), we cut three cross sections
ling and well completion operations. The bridging efficiency of of each core for analysis. Figure 9 depicts SEM images of top,
filter cake has a huge impact on wellbore stabilization and forma- middle, and bottom cross sections of the cores, all related to blend
tion damage control or damage mitigation in the pay zone [31]. 1. The study of all images, shown in Fig. 9, first revealed that in
We used SEM imaging technique in this study for qualitative esti- low permeability formation like Bandera Brown sandstone
mation and description of the two filter cake regions formed from (images on the left), a compacted-plastered, formation bridging
the PPA experiments using the optimal blend 1. Prior to conduct- mechanism has been achieved whereas, in Michigan sandstone
ing the SEM imaging, we vacuum dried the experimental core case, unbridged pores can be observed easily. The thickness and
specimens and further preserved them in desiccators to avoid tightness of the filter cake appear to impede further movement of
atmospheric interference and possible chemical or physical altera- filtrate through the formation. This is consistent with the filtration
tion of composite particles within the cores. results where the filtrate loss was lower in the low permeability
Figure 8 shows the rock specimens with their corresponding Bandera Brown than Michigan sandstone. The second observation
SEM images before filtration. The pore morphology of the distinct specifically relates to the top cross section of Bandera Brown
Bandera Brown sandstone pore, under a higher magnification than sandstone, which was fully bridged and plastered with mud par-
the Michigan sandstone image, reveals a circular-like shape. ticles. This observation indicates the early time evolution of the
Although this may not be completely true for all the pore spaces, internal filter cake in a region largely influenced by the fine sized
some of the visible, small pores appear to have irregular shapes. calcium carbonates and graphite. The middle cross section of this
The grains of Bandera Brown sandstone, some approximately low permeability core specimen also corroborates the evolution of
measuring up to 20 lm, appear to be small, discrete, irregular, and internal filter cake plastering effects. The features of the sandstone
have accumulated in large, fragmented quantities. For this reason, morphologies, which have been discussed previously, have also
we increased the magnification during the imaging process. The contributed to the pore bridging observed in Fig. 9.
distinct Michigan pores in Fig. 8 (right), has a rectangular-like To substantiate the heterogeneity of the external filter cake
shape. The edges of the pores appear to be smooth, when com- region, we conducted further SEM imaging of the external filter
pared to the somewhat rough edges of the Bandera Brown pores. cake top and bottom surfaces. We present the external filter cake
In addition, the morphology of the Michigan sandstone, as seen in surface images in Figs. 10 and 11. The morphology of Fig. 10
SEM image, also shows large grains, measuring up to 150 lm and shows that a broad combination of large and small aggregates of
appears to have accumulated in small quantities. mud particles have been distributed within an amorphous phase,
The particle morphology of graphite, as seen in Fig. 3, when forming the base of external filter cake, while Fig. 11 reveals that
compared with the morphology of Bandera Brown sandstone only small particles, at the top of external filter cake, are exposed

Journal of Energy Resources Technology MARCH 2017, Vol. 139 / 022912-5

Downloaded From: http://energyresources.asmedigitalcollection.asme.org/pdfaccess.ashx?url=/data/journals/jertd2/935440/ on 03/03/2017 Terms of Use: http://www.asme.org/ab


Fig. 9 SEM analysis of cross-sections of Michigan (right) and Bandera Brown (left) specimens after filtration.
The arrows indicate the direction of mud filtrate flow from top of the disk in PPA to bottom of the disk; showing
the sequence of particle invasion at various depths (thickness) of the disk.

to the circulating drilling mud. This observation brings to mind a particles, approximately between 0.1 lm and 10 lm, exposed to
sedimentary geological process referred to as fining upward: a wellbore flow.
depositional material flow process whereby, grain size decreases
upward, causing the coarse grains to be deposited at a base, fol-
lowed by finer grains at the top of the accumulated sediments. 5 Determination of External Filter Cake Permeability
This similarity of depositional environment indicate that higher Using some of the analytical models presented in the literature
flow energy per unit volume first carried the larger particles to the [4] for linear systems, we calculated the permeability of the exter-
base of external filter cake then the lower energy per unit volume nal filter cake (Kc). The model used here calculates filter cake per-
deposited the smaller particles at the surface of external filter meability from the slope of the linear trend line of filtrate loss
cake. This is consistent with the process of filtration; that is, “high versus the square root of time. The model requires input variables
energy spurt loss” comes first followed by “lower energy, steady such as porosity changes in the cake, differential pressure across
state particle transport” next. Thus, this study reveals that a heter- rock disk, and cross-sectional area, all to calculate filter cake per-
ogeneous filter cake was formed with particles of various sizes meability. Based on Barkman and Davidson’s work (1985), this
consistent with the transport pressure gradient. The particles of model assumes that the porous media is impaired by pore plug-
approximate sizes between 0.4 lm and 100 lm, are found at the ging followed by the formation of external filter cake [4]. Equa-
base of external filter cake, at the formation face, and only small tions (1)–(3) present details of the analytical model

022912-6 / Vol. 139, MARCH 2017 Transactions of the ASME

Downloaded From: http://energyresources.asmedigitalcollection.asme.org/pdfaccess.ashx?url=/data/journals/jertd2/935440/ on 03/03/2017 Terms of Use: http://www.asme.org/ab


linear trend line, and the square of the slopes of trend line. The
x-intercept in this plot relates to the spurt loss, which could not be
measured accurately after 10 s. This is partly due to the insuffi-
cient volume of the filtrate, that is, already present in the internal
filter cake, in the interstitial pore spaces of the sandstone, and
over the surface of the external filter cake. We calculated the filter
cake bulk density using Eq. (2). In using Eq. (3), we assumed a
constant differential pressure across the filter cake and across the
sandstone disk. Other parameters were input into Eq. (1) to arrive
at the permeability of the filter cake for all specimens. However,
we present the permeability results for the optimum mud blend as
0.11 mD for Michigan sandstone and 0.02 mD for Bandera Brown
sandstone. The filter cake permeability of Michigan sandstone is
approximately five times more than the permeability of Bandera
Brown. This is explained due to a better plastering and formation
of early internal filter cake in Bandera Brown sandstone. One
must also consider effective bridging by fine-sized calcium carbo-
Fig. 10 Base surface of external filter cake in contact with the nates and interlocking mechanism of graphite in the mud system
formation. The arrows point to a combination of large and small and the pore spaces. These wellbore strengthening materials sig-
particles, of various sizes that have accumulated at the base of nificantly impact the quality and permeability of the filter cake
external filter cake bridging the surface of the formation. developed during fluid loss. As mentioned earlier, a successful
well bore strengthening by mud cake can occur when, a permeable
formation, merges with low permeability filter cake to cause a sig-
nificant increase in the effective stress of the formation.

6 Conclusion
In this paper, we present a comprehensive experimental study
on wellbore stabilization in sandstone formations. Our key find-
ings are: the response of fluid loss and filter cake evolution due to
the formation permeability and mud formulation design factors.
The effects of calcium carbonate and graphite in an optimized
drilling mud PSD approach revealed that filtrate reduction is a
function of the type of WSM, size, and concentration. The results
of filtration experiments showed the formation permeability plays
an important role in cumulative filtrate loss and evolution of inter-
nal and external filter cakes. The filter cake build up as well as its
plastering effects for Michigan and Bandera Brown sandstones
were observed from SEM results. The outcome of SEM imaging
of top and middle cross section of Bandera Brown sandstone
Fig. 11 Top of the external filter cake in contact with moving depicted early particle packing, which leads to forming a low per-
drilling mud. The arrows point to small particles that have accu- meability internal filter cake. The heterogeneity of the external fil-
mulated at the surface of the external filter cake. ter cake, having various size particles at its surface and bottom,
has been verified in this study from SEM analysis. Filter cake per-
meability calculations for both high and low permeability cores
reveals low permeability values in Bandera Brown sandstone. As
a future recommendation, we suggest an increase in the concentra-
tion graphite because of its surface irregularity, roughness,
particle–particle interlocking, and particle–pore interlocking
mechanism. The expected outcome is improved filter cake, and
increased cohesive strength and compactness, thus, improving
wellbore strengthening by mud cake.

Acknowledgment
The authors would like to thank Mr. Sloan Spears of the
National Oilwell Varco Fluid Control Laboratory in Lafayette
Louisiana. We also thank Mr. Aderemi Agboola for his help in
Fig. 12 Cumulative filtrate loss versus square root of time conducting laboratory experiments.

S2 wlqw Nomenclature
Kc ¼ (1)
2a2 qc Dp a¼ cross-sectional area of the core, cm2
Kc ¼ cake permeability, mD
qc ¼ ð1  /c Þqp (2) pc ¼ pressure exerted on the slurry, atm
pe ¼ formation external pressure, atm
Dp ¼ Pc  Pe (3) DP ¼ pressure differential, atm
S¼ slope of the curve from cumulative mud filtrate invasion
The first step in this calculation, as shown in Fig. 12, is to plot the w¼ mass fraction of suspended particles, percent by weight
cumulative filtrate loss against the square root of time, obtain the l¼ filtrate viscosity, cp (assume Water)

Journal of Energy Resources Technology MARCH 2017, Vol. 139 / 022912-7

Downloaded From: http://energyresources.asmedigitalcollection.asme.org/pdfaccess.ashx?url=/data/journals/jertd2/935440/ on 03/03/2017 Terms of Use: http://www.asme.org/ab


qc ¼ bulk density of the filter cake, gm/cm3 [16] Cook, J., Guo, Q., Way, P., Bailey, L., and Friedheim, J., 2016, “The Role of
Filtercake in Wellbore Strengthening,” IADC/SPE Drilling Conference and
qp ¼ specific gravity of particle (assume Barite) Exhibition, Fort Worth, TX, Mar. 1–3, Paper No. IADC/SPE 178799.
qw ¼ water density, gm/cm3 [17] Breuer, J., Mammadov, E., Al-Hashmy, W., and Sadat, A. Y., 2016, “Case Study
qu ¼ particle density, gm/cm3 of Enhancing Wellbore Strengthening While Drilling Subnormal and Over Pres-
Uc ¼ cake porosity, fraction sured Formation Types,” IADC/SPE Asia Pacific Drilling Technology Confer-
ence, Singapore, Sept. 22–34, Paper No. IADC/SPE 180544.
[18] Ewy, R. T., and Morton, E. K., 2009, “Wellbore-Stability Performance of Water-
Based Mud Additives,” SPE Drilling and Completions, 24(03) pp. 390–393.
[19] Kang, Y., You, L., Xu, X., and Liao, Z., 2012, “Prevention of Formation Dam-
References age Induced by Mud Lost in Deep Fractured Tight Gas Reservoir in Western
[1] Kiran, R., and Salehi, S., 2016, “Thermoporoelastic Modeling of Time- Sichuan Basin,” J. Can. Pet. Technol., 51(01) pp. 46–48.
Dependent Wellbore Strengthening and Casing Smear,” ASME J. Energy [20] Chellappah, K., Kumar, A., and Aston, M., 2015, “Drilling Depleted Sands: Chal-
Resour. Technol, 139(2), p. 022903. lenges Associated With Wellbore Strengthening Fluids,” SPE/IADC Drilling Con-
[2] Farahani, M., Soleimani, R., Jamshidi, S., and Salehi, S., 2014, “Development ference and Exhibition, London, UK, Mar. 17–19, Paper No. SPE/IADC 173073.
of a Dynamic Model for Drilling Fluid’s Filtration: Implication to Prevent For- [21] Byrne, M., Zubizarreta, I., and Sorrentino, Y., 2015, “The Impact of Formation
mation Damage,” SPE International Symposium and Exhibition on Formation Damage on Core Quality,” SPE European Formation Damage Conference and
Damage Control, Lafayette, LA, Feb. 26–28, Paper No. SPE-168151. Exhibition, Budapest, Hungary, June 3–5, Paper No. SPE 174189.
[3] Salehi, S., Hareland, G., and Nygaard, R., 2010, “Numerical Simulations of [22] Dorman, J., Lakatos, I. J., Szentes, G., and Meidl, A., 2015, “Mitigation of For-
Wellbore Stability in Under-Balanced-Drilling Wells,” J. Pet. Sci. Eng., mation Damage and Wellbore Instability in Unconventional Reservoirs Using
72(3–4), pp. 229–235. Improved Particle Size Analysis and Design of Drilling Fluids,” SPE European
[4] Civan, F., 2007, Reservoir Formation Damage, 2nd ed., Gulf Publishing Com- Formation Damage Conference and Exhibition, Budapest, Hungary, June 3–5,
pany, Waltham, MA, pp. 780–782. Paper No. SPE 174260.
[5] Jiao, D., and Sharma, M. M., 1994, “Mechanism of Cake Buildup in Crossflow [23] Khan, H., Mirabolghasemi, M., Yang, H., Prodanovic, M., DiCarlo, D., Balhoff,
Filtration of Colloidal Suspensions,” J. Colloid Interface Sci., 162(2), M., and Gray, K., 2016, “Comparative Study of Formation Damage due to
pp. 454–462. Straining and Surface Deposition in Porous Media,” SPE International Sympo-
[6] Salehi, S., and Kiran, R., 2016, “Integrated Experimental and Analytical Well- sium and Exhibition on Formation Damage Control, Lafayette, LA, Feb. 26–28,
bore Strengthening Solutions by Mud Plastering Effects,” ASME. J. Energy Paper No. SPE-178930.
Resour. Technol., 138(3), p. 032904. [24] Razavi, O., Vajargah, A. K., van Oort, E., Aldin, M., and Govindarajan, S., 2016,
[7] Tran, M. H., Abousleiman, Y. N., and Nguyen, V. X., 2010, “The Effects of “Optimization of Wellbore Strengthening Treatment in Permeable Formations,” SPE
Low-Permeability Mud Cake on Time-Dependent Wellbore Failure Analyses,” Western Regional Meeting, Anchorage, AK, May 23–26, Paper No. SPE-180467.
IADC/SPE Asia Pacific Drilling Technology Conference and Exhibition, Ho [25] Ezeakacha, C. P., Salehi, S., and Galambor, A., 2016, “An Integrated Study of
Chi Minh City, Vietnam, Nov. 1–3, Paper No. SPE 135893. Mud Plastering Effect for Reducing Filtrate’s Invasion,” SPE International
[8] Nygaard, R., and Salehi, S. A., 2011, “Critical Review of Wellbore Symposium and Exhibition on Formation Damage Control, Lafayette, LA, Feb.
Strengthening: Physical Model and Field Deployment,” AADE National Techni- 26–28, Paper No. SPE-168151.
cal Conference and Exhibition, Houston, TX, Apr. 12–14, Paper No. AADE-11- [26] Salehi, S., Hussmann, S., Karimi, M., Ezeakacha, C. P., and Tavanaei, A.,
NTCE-24. 2014, “Profiling Drilling Fluid’s Invasion Using Scanning Electron Micros-
[9] Tien, C., Bai, R., and Ramarao, B. V., 1997, “Analysis of Cake Growth in Cake copy: Implications for Bridging and Wellbore Strengthening Effects,” SPE
Filtration: Effect of Fine Particle Retention,” AIChE J., 43(1), pp. 33–44. Deepwater Drilling and Completions Conference, Galveston, TX, Sept. 10–11,
[10] Elkatatny, S., Mahmoud, M., and Nasr-El-Din, H. A., 2013, “Filter Cake Prop- Paper No. SPE 170315.
erties of Water-Based Drilling Fluids Under Static and Dynamic Conditions [27] Wang, H. M., Savari, S., Whitfill, D. L., and Yao, Z., 2016, “Forming a Seal
Using Computed Tomography Scan,” ASME. J. Energy Resour. Technol., Independent of Formation Permeability to Prevent Mud Losses—Theory, Lab
135(4), p. 042201. Tests, and Case Histories,” IADC/SPE Drilling Conference and Exhibition,
[11] Salehi, S., and Nygaard, R., 2014, “Full Fluid–Solid Cohesive Finite-Element Fort Worth, TX, Mar. 1–3, Paper No. IADC/SPE 178790.
Model to Simulate Near Wellbore Fractures,” ASME. J. Energy Resour. Tech- [28] Kalantariasl, A., Bedrikovetsky, P., and Vaz, A. S., 2015, “Formation Damage
nol., 137(1), p. 012903. due to Drilling and Well Completion: External Cake Formation and
[12] Bezemar, C., and Havenaar, I., 1966, “Filtration Behavior of Circulating Dril- Stabilization,” SPE Russian Petroleum Technology Conference, Moscow, Rus-
ling Fluids,” SPE J., 6(4), pp. 291–298. sia, Oct. 26–28, Paper No. SPE 176527.
[13] Ba Geri, B. S., Mahmoud, M., Al-Mutairi, S. H., and Abdulraheem, A., 2015, [29] Loeppke, G. E., Glowka, D. A., and Wright, E. K., 1990, “Design and Evalua-
“Effect of Sand Content on the Filter Cake Properties and Removal During tion of Lost-Circulation Materials for Severe Environments,” SPE J. Pet. Tech-
Drilling Maximum Reservoir Contact Wells in Sandstone Reservoir,” ASME. J. nol., 42(03), pp. 328–337.
Energy Resour. Technol., 138(3), p. 032901. [30] Allen, D., Auzerais, F., Dussan, E., Goode, P., Ramakrishnan, T. S., Schwartz,
[14] Nandurdikar, N. S., Takach, N. E., and Miska, S. Z., 2002, “Chemically L., Wilkinson, D., Fordham, E., Hammond, P., and Williams, R., 1991,
Improved Filter Cakes for Drilling Wells,” ASME. J. Energy Resour. Technol., “Invasion Revisited,” Schlumberger Oilfield Review, 3(3), pp. 10–23.
124(4), pp. 223–230. [31] Al-Yami, A. S., Nasr-El-Din, H. A., Al-Shafei, M. A., and Bataweel, M. A.,
[15] Nasr-El-Din, H. A., 2005, “Formation Damage Induced by Chemical Treatments: 2010, “Impact of Water-Based Drilling-In Fluids on Solids Invasion and Dam-
Case Histories,” ASME J. Energy Resour. Technol., 127(3), pp. 214–224. age Characteristics,” SPE Production and Operations, 25(01), pp. 40–49.

022912-8 / Vol. 139, MARCH 2017 Transactions of the ASME

DownloadedViewFrom:
publicationhttp://energyresources.asmedigitalcollection.asme.org/pdfaccess.ashx?url=/data/journals/jertd2/935440/
stats on 03/03/2017 Terms of Use: http://www.asme.org/ab

You might also like