RESERYOIR STIMULATION

Bertaux (1986) observedthat in silicoaluminatescon- l This processsupposesthe initial adsorptionof H30t

taining potassium, such as illite, potassium fluoborate hydronium ions on the clay surface,then its exchange
forms after treatmentwith fluoboric acid; potassiumhex- with NH:, so that HF is generatedin situ. Such ex-
afluosilicate forms after treatment with mud acid. This changeand replacementof H30t by NH4’ depends
showsthat nondamagingbyproductsare formedin the case on manyparametersand cannotbe ascertained.There-
of fluoboric acid, whereasformation plugging by alkali fore, ,eventhe generationof HF is dubious.
fluosilicatescan occur with mud acid. This is anotherad-
vantageof using fluoboric acid in some“acid-sensitive”
formations.
During the injection period, fluoboric acid behaveslike
a weak hydrofluoric acid solution, but one where the HF
is constantly replenished.The small amount of fluoride
ions availableat any time limits the dangerof precipitat-
ing aluminum species.Only the first acidity is useddur-
ing this step.Hydroxytrifluoboric acid, HBF30H, buffers
the solutionandpreventsother undesirableprecipitations.
During the shut-in HBF30H continuesto react, but at
a slow pacebecauseits hydrolysis is minimal. The HF,
thus liberated, reactsfurther with mineral species.It also
reacts by topochemical reactions, where the aluminum
from the undissolvedclay structure is put into solution
by forming one of the fluoaluminate complex ions (de-
pendingon [F-l), and the surfaceof the mineral is there- Figure 14-8-SEM picture showing kaolinite and illite clays
fore enrichedin silicon and boron. An amorphouscoat- in a pore, before attack.
ing of silica andborosilicate “glass” is then formed over
the remainingsilicateand fine silica grains, welding them
to the framework and thus preventing their migration.
This effectis clearly seenin Figs. 14-6and 14-7, where
the samepore, containing two different clays (kaolinite
and illite), is shownbefore and after reaction with a fluo-
boric acid solution. The quartz is hardly etched, while
the high surfacearea, fast reactingillite is completelydis-
solved.The kaolinite plateletsare abouthalf dissolvedand
anamorphousmaterialis coatingthe undissolvedkaolin&e,
welding themtogetherandto the underlying quartzgrain.

14-6.2 Sequential Mud Acid

This systeminvolvesthe in-situ generationof hydrofluoric
acid, occurringfrom the alternateinjectionof hydrochloric
acid and of ammonium fluoride (Hall et al., 1981). The
reactionsof hydrofluoric acid are thoughtby someto take
place at the rock surfaceby adsorption followed by ion
exchange,but the yield of this heterogeneousprocess
seemshighly doubtful for several reasons.
l If HF acidwere to be generatedthrough sucha process,
it would be a very small quantity, hardly enough to
etch the surface of the clay material.
l As this process is based on the CEC of the clays,
migrating kaolinite would hardly be touched.
Figure 14-7-SEM picture of the same pore after an 8%
HBFh treatment.
14-6.3 Alcoholic Mud Acid
Alcoholic mud-acid formulations are a mixture of mud
acid and isopropanolor methanol(up to 50%). The main
application is in low-permeability dry-gas zones. Dilu-
tion with alcohollowers the acid/mineralreactionrate and
provides a retarding effect,

14-16

Cleanup is also facilitated; acid surface tension is
decreasedby the alcoholswhile the vapor pressureof the
mixture is increased, which improves gas permeability
by reducing water saturation.
14-6.4 Mud Acid Plus AlC13 For Retardation
An acidizing systemto retard HF-minerals reactionshas
beenproposedin which aluminum chloride (AK&) is ad-
dedto mud-acidformulationsto complexsomeof the fluo-
ride ions in the injected mixture, according to the reac-
tions (Gdanski, 1985):
AlC& + 4HF + HZ0 $ AIFd- + 3HCl + H30+,
AlF‘,- + 3H30+ F+ AIF; + 3HF + 3HZ0.
However, this procedure is tantamountto adding dis-
solutionreatitionproductsto the mixture beforethesereac-
tions occur. In theory this should slow down the rates.
However, the retardationof clay dissolution hasnot been
proved experimentallybecauseof the prime importance
of the high surfaceareason clay reactivity, much more
important than a slight depletion of acid at high tempera-
tures. The risk of earlyprecipitationof damagingproducts,
such as aluminum fluoride (AIFS) or fluoaluminates, is
probably increasedby the use of an acid, already con-
taining aluminum ions, before reaction. Flow testshave
showna smallereffectivelive acid penetrationthan in the
case of mud acid.
14-6.5 Organic Mud Acid
As total acidity speedsup mineral dissolution with mud
acid, organic mud acid involves the replacementof 12%
HCl by organic acids (typically 9 % formic acid, a weak
acid only partially dissociated),mixed with 3 % HF, to
retard hydrofluoric acid spending.This systemis particu-
larly suitedfor high-temperaturewelIs (200O-300oF) , for
which pipe-corrosion rates are diminished accordingly.
This systemalso reducesthe tendency to form sludge.
14-6.6 Self-Generating Mud-Acid Systems
These acidizing systemswere originally developed by
Templeton et al. (1974), and their application has been
widenedrecently by Abrams et al. (1983). They involve
the hydrolysis of organic estersinto the corresponding
carboxyIic acids, followed by the reaction of theseacids
with ammonium fluoride to yield hydrofluoric acid. As
the hydrolysis reaction is activated by temperatureand
the acidity obtained is not as strong as with mud acid,
a low corrosion rate of tubular goodsas well asa delayed
reaction of the progressivelygeneratedhydrofluoric acid
are expected.The latter would allow deeppenetrationof
live HF.
MATRIX ACIDIZING OF SANDSTONES
Dependingon the bottomholetemperature,different or-
ganic esters are used:
methyl formate between 130’ F and 180o F, with the
reactions:
HCOOCH3 + HZ0 * HCOOH + CH30H
slow, rate-controlling reaction
HCOOH + NH4F $ NH: + HCOO- + HF
ammonium salt of monochloroacetic acid between
180’ F and 215” F:
NH; ClCH$OO- + HZ0 e HOCHJJOOH +
NH: + Cl-
methyl acetatebetween I90’ F and 280’ F.
Reagentchoiceis intendedto limit at 30% (maximum)
the generationof hydrofluoric acid during pumping of the
mixture in the tubing; thus, a minimum of 40 minutes of
spendingtime seemsnecessary.However, field tests of
these systemshave not been conclusive. Many precipi-
tatesform in theselow acidic systems,suchas ralstonite
(NHaMgAIFJ , and other fluoaluminates(silicates)upon
spendingof thesemixtures on clays; thus, the useof com-
plexing agentsor acids, such as citric acid, is suggested.
Besides,formation sensitivity after treatni?nthasnot been
tested, and handling problems arise from the high flam-
mability of methyl formate.
Overall, these systemshave many drawbacks. Based
on the hydrolysis of various organic esters,they are tem-
peratureactivated.Unlike fluoboric acid; where new HF
is generated only upon spending, no equilibrium is
reached.This meansthat more HF is geheratedas tem-
perature increasesand the ester can eventually be com-
pletely hydrolyzed long before reaching-the final depth
of damage.The true degreeof retardationdependson tem-
peratureandpumping time. Theseestersare more expen-
sive and more dangerousto handlethan HCl or inorgan-
ic salts (flammability). More precipitatesare formed as
a result of the poor solubility of the organic byproducts.
The orily advantageover the reduced-strength HF is lower
corrosion rates.
14-6.7 Buffer-Regulated Hydrofluoric Acid Systems
Other high-pH acidizing systemsproposedfor use up to
360’ F involve the buffering effect of an organic acid and
its ammonium salt, mixed with ammonium fluoride, as
a hydrofluoric acid precursor (Abrams et al., 1983). To
minimizecorrosion,the useof the sameuninhibitedbuffer,
without ammoniumfluoride as a preflush,hasbeenrecom-
mendedup to 350 ’ F. The ammoniumsalt of the organic
acid is generatedfrom the partial neutralizationof the acid
with ammonium hydroxide. The different proposed
buffered systemsare:
14-17
RESERVOIR STIMULATION
l formic acid/ammoniumformate with pH = 3.5 to 4,
l acetic acid/ammoniumacetateand citric acidlarnmo-
nium citrate with pH = 4.5 to 5.
To extend the application to higher temperatures(up
to 550 ’ F), an excessof ammonium salt is formed by us-
ing a higher ammoniumhydroxide/organicacid ratio. As
the kinetics of clay dissolution increaseswith the fluo-
ride ion concentration,more ammoniumfluoride is added
to compensatefor the pH increase(Scheuerman,1985).
Successfulin-depthstimulationhasbeenobservedwith
this systemonly for bottomholetemperatureslower than
54 ’ C. In mostcases,many damagingprecipitatesareno-
ticed when using this system(fluosilicates,fluoaluminate
often involving ammonium), the formation of which is
relatedto the weakacidity in the near-wellborearea.These
systems suffer from the same drawbacks as the self-
generatedmud-acid system.
14-7 MATRIX ACIDIZING DESIGN
A typical designfor a stimulationjob should involve the
following major steps:
Ensurethat the well is a candidatefor matrix stimu-
lation; i.e., clearly establishthat the well is damaged.
Much of this hasbeencoveredin detail in Chapter1.
21 Establish the cause of damage, its severity, and
location.
3) Selectthe proper formulation of treating fluids, in-
cluding additivesthat will removethe damagewithout
damagingthe rock through formation of secondary
precipitates, sludge, etc. This may require laborato-
ry tests.
4) Determinethe pressureand/or rate for injection with-
in matrix flow regime; i.e., without fracturing the
formation.
Determinethe volume of treating fluid to be injected
per foot of perforated interval.
If the reservoir is multilayered or if the pay zone is
very thick, determine a placementtechniqueto en-
sure each foot of the reservoir receives the proper
amount of fluid.
7) Determine possible shut-in times, plan diverter
cleanup,andmakerecommendations for returningthe
well to production or injection.
Assessprofitability of treatment by estimating in-
creasesin productivity or injectivity vs the cost of
the treatment itself.
14-18
Design engineersmust recognizethat every one of the
abovepoints is equally importantandnonecanbe exclud-
ed. Failurescanoccurby leavingout anysinglestep.Other
factors must be consideredbefore finalizing a proposal
for a pumping schedule:
l circulating out wellbore fluids that could be damaging
to the formation,
* constraints, imposedby completion, that prevent the
use of certain fluids (e.g., corrosion) or techniques
(e.g., mechanicaldiversion), and
l availability of materials.
As for any other design problem, the quality of the
proposal strongly dependson the quality of the informa-
tion collectedon the well condition,laboratorytests,reser-
voir parameters,and completion description.
It is assumedhere that the first three stepshave been
completed; i.e., a comprehensivereservoir justification
of the treatmenthasbeendoneand that stimulation fluids
have been selectedaccording to recommendationssuch
as those described in Sections 14-2 and 14-3.
14-7.1 Pressure For Injection
Matrix acidizing is defined as the injection of fluids at
pressuresbelow fracturing pressure. The injected fluid
flows either through the existing porous medium or
through new passagewayscreatedby the fluid itself.
Any “matrix” treatmentthat exceedsthe fracturepres-
sure is undesirablebecausethe acid may not dissolve or
dispersewellbore damage.While fractures are essential
in low-permeability reservoirs, in higher permeability
reservoirs the newly formed fractures may bypass the
damagedarea and may not clean the formation around
the wellbore. The createdpermeabilitycontrastwould not
be adequate,and the bypasseddamagewould remain.
Fracturesmay also extendthrough zone boundariesand
lead to an increase in water or gas production.
A major step in designing a matrix acidizing job is to
determinethe conditionsfor matrix flow regime.To avoid
fracturing the formation, the bottomhole pressuremust
not exceedthe fracturingpressure.The calculationof frac-
turing pressurehasbeendemonstratedin Chapter2. De-
termining bottomhole pressuresfrom wellhead data is
shown in detail in Chapter 16.
14-7.2 Determination Of Maximum Injection Rate
The maximum possibleinjection rate that doesnot frac-
ture the formation is derived from Darcy’s radial flow
equation:

4.917 x lO-6 kh [(Q x H) - Aps‘+ - p]
q.WIOX=
pB (ln ~JJ-,~+ s)
(14-6)
whereqirnmis the injection rate (BPM), k is the effec-
tive permeability of the undamagedformation (md); 11is
the net thicknessof the formation (ft); gfis the fracture
gradient (psi&); H is the depth (ft); Apsafeis the safety
pressuremargin (200 to 500 psi); p is the reservoir pres-
sure (psi); p is the viscosity (cp); re is the drainageradius
(ft); r,,, is the wellbore radius (ft); and s is the skin fac-
tor (dimensionless).The formation volume factor, B, is
very near unity.
Eq. 14-6 is a simplified inflow performancerelation-
ship; it doesnot accountfor transient effects, multiphase
flow, or reservoir heterogeneities.The injected fluid is
assumedto be incompressible.The effective permeabili-
ty is the permeabilityto the injectedfluid. Theseconcepts
are reviewedin detail in Chapter 16. The value obtained
from Eq. 14-6with the initial skin value can be usedonly
as a guideline for determining the initial rate.
14-7.3 Fluid Volumes: Experience
Acid flows through sandstonesin a uniform and stable
manner.As a consequence,unlike HCI treatmentsof car-
bonates, it is not the purpose of mud-acid treatmentsto
createnew permeability by dissolving part of the forma-
tion. This, in fact, would deconsolidatethe formation. In-
stead, they are intended to dissolve damagingmaterials
plugging the pore channels and thereby increase
permeability.
Acid reactionswith sandstoneformations are complex
sinceHF could reactwith feldspars,clays, and other sili-
cate and carbonateminerals, often of undefined chemi-
cal compositionandwidely variable physicaldistribution.
When acid reaction modeling is not available, volumes
used are the ones experiencehas shown to be most cost
effective,
In general,the mosteffectiveacidizingtreatmentis com-
posedof three parts: namely, a preflush, the acid formu-
lation, and an overflush.
- Prejlush Volumes
The preflush servesseveral important functions. It
l provides a separationbetweenconnatewater and HF
to help prevent the formation of potentially damaging
sodium or potassiumfluosilicates, and
l reactswith carbonateminerals in the formation to pre-
vent their reaction with HF.
MATRIX ACIDIZING OF SAh’DSTONES
Thesefunctions can be satisfiedby hydrochloric acid.
Dependingon the carbonateand clay contentsof the for-
mation and its permeability, 5 % to 15% HCl or 5 to 10%
aceticacid is used. The volume of this acid preflush also
dependson the carbonatecontent, but certain minimum
volumescanbe derivedbasedon the displacementof con-
natewater andinter-mixingof fluids while they arepumped
down the tubing.
The volume required to displaceformation fluids to a
radial distance, rS, can be estimatedby
Vp = 7.48 [t$ (rj? - t-i) x] (14-7)
where r,,, is the wellbore radius (ft); rS is the depth of
damage(ft); and l$ is the pore volume (gal&t).
The volume requiredto dissolveall acid-solublemateri-
als up to a radial distance, rS, is given for HCI by:
where VHcl is the volume of HCI required (gal/ft); XHcl
is the weight fraction of formationmaterialsolublein HCl;
/3is the dissolving power of the acid (gallon of rock dis-
solvedper gallon of acid); rS is the radius of damage(ft);
and r,,, is the wellbore radius (ft). :,
When the carbonatecontentof the sandstonereservoir
is very low, it may seemthat only a small volume of HCl
preflush, or no HCI at all, is necessary.It may appear
that a solventand/orbrine preflushis sufficient.However,
a substantialHCl preflushvolumeis still necessaryto com-
pletely exchangethe cations of the clays by protons. If
this step is not done, clays will be abIe,to exchangeand
extract protons from the main HF acid flush. This will
increasethe pH of this fluid and will certainly induce un-
desirable precipitations (Gdanski and Peavy, 1986).
Acid Volumes
The acid formulation generally usedis a 12%-3% mix-
ture of hydrochloric and hydrofluoric acids plus inhibi-
tor and any other special additives required for the for-
mation to be treated, including diverting agents. Other
strengthsof acidsand other ratios of HCl to HF acidscan
be usedto help prevent secondaryprecipitationproblems.
The HCl acid maintains a low pH to avoid deposition of
CaFz, AlF3, and other complex reaction products. This
issuehasbeenaddressedin Sections14-2, 14-3, and 14-5.
The optimum designfor regular mud-acidvolumes ap-
pears to be 125 to 200 gal per foot of formation. This
value is basedon many treatmentsthroughout the world
and on a statisticalanalysisof 174treatmentsin the Texas
14-19
RESERVOIR STIMULATION

Gulf Coast by Gidley et al. (1976). It is consistentwith spendingand correspondingmineral-concentrationpro-

figures quotedby Lafleur andJohnson(1973)in the North
Seaand by Smith et al. (1965). However, a recent study
by Gidley (1985) showsthat many successfulsandstone
acidizing treatmentshave been performed with smaller
volumes of acid.
Since the purposeof sandstoneacidizing formulations
is damageremoval, the amount and location of which is
poorly known, sometrial-and-error or onsite injectivity
measurementsare used to optimize the volume.
In order to limit such trial-and-error exercise, a real-
time onsite treatment evaluation method (Prouvost and
Economides, 1987and 1988) is usedto determinewhen
the damagehasbeenremovedand decide if enoughfluid
has been used. This techniqueis detailed in Chapter 16.
Oveflush Volumes
The overflush for mud acids canbe either 3 %- 10% HCI,
NH&l, or a light hydrocarbon, such as diesel oil. For
a gas well, nitrogen may be used. Other overflush for-
mulations can be consfderedfor special purposes, such
as to facilitate diverter cleanup.
The volume of overflush used successfullyhas ranged
from 50 gal& to 1.5 times the volume of the acid formu-
lation. Sincethe primary purposeis to displacepotentially
damaging precipitates deep into the reservoir, a 4-ft
penetration radius is usually sufficient. The required
volume may be calculated using Eq. 14-7.
Flowback
Flowbackof spentfluids from conventionalmud-acidtreat-
files after attack. To do that, rock mineralogic composi-
tion is required and minerals are lumped into two
categories:quartzor slow reactingspeciesandall the other
mineralsthat are fast reacting. An exampleof suchsimu-
lation is given in Figs. 14-Sto 14-10, where the attack
of a lOO-md,15% porosity sandstonereservoiris modeled.
In this example,the wellbore radius is 2 in. and the for-
mation is composedof 75% quartz and25 % other miner-
als. A standardmud-acidsolution(3 % HF) is injectedinto
the formation at a rate of 0.01 BPM/ft. The acid-spending
profile on the nondamagedformationis shownin Fig. 14-8
for various volumes of acid. The relative concentrations
of the fast reactingminerals and of the quartz framework
are shownin Figs. 14-9and 14-10for the samedata.These
data demonstratethat large volumes of acid are needed
to achievemoderatepenetrationwith such a homogene-
ous dissolution process.
The objectiveof matrix acidizing is to removedamage,
but the simulation requires damageparametersobtained
from well testing.When parametersareunknown, differ-
ent simulationruns canbe madewith varying parameters.
A = 150 iai/ft
0 =200 gal/ft
+ = 250 gaMt

ments should be accomplished as soon as possible.

Detrimental reaction products will be formed within the
formation if spentmud-acid formulations remain for an
extendedperiod of time. To assistsuchflowback it is be-
coming commonpractice to usenitrogen and coil tubing
for lifting spent fluids.
In productionwells that havebeentreatedfor deepclay
damage,a slow “beaning-up” processis recommended:
the wells are brought back into production very slowly
by graduallyincreasingthe sizeof the chokesover a period
of several weeks and thus keeping fine migration to a
minimum.
Usually, no attemptis made to flow back water injec-
tion wells. Rather, the well is immediately injected and
the spentfluids are pushedout farther into the formation,
away from the critical wellbore area. Radial Distance (in.)

14-7.4 Fluid Volumes: Mathematical Simulation

Only the reactionof plain hydrofluoric acid is considered Figure 14-8-HF acid spending profile in an undamaged
in this section. Using the mineral dissolution models sandstone.
described in Section 14-5.3, one can compute HF acid-

14-20