Formation Damage Tests of High=

Density Brine Completion Fluids

L.N. Mor$enthaler, SPE, Shell Development Co.

5P~ I 5“81/”
Summary. Laboratory core flow tests were cenducted to assess the formation damage pntential of brines with
densities between 13.4 and 19.2 ibm/gal [1606 and 2301 kg/m3]. Unfavorable tluid/rock interactions are not
evident. Recovery of oil pa’rneabilhy ia slow but complete, in agreement with established displacement theories.
Observed damage effects are the result of the instability of the brine or incompatibti~ with water in the coiea.
The. effectiveness of ZnBr2 in preventing impairment suggests acid-snluble impairing materials.

Introdrrotton
Solids-free fluids elimimte a major cause of formation in the 13.4-to- 14.2-Ibnr/gal [1606-to-1702-kg/m3 ] range
darrrage, invasion by fine particulate. The lack of solid by mixinf+ 14.2- and 11.6-ibm/gal [1702- and
filter-caking agenta, however, compounds the problerrr of 1390-kg/rri ] brines. Brines with densities between 14.2
fluid invasion of producing formations. 1 Highdensity and 15.0 ibm/gal [1702 and 1797 kg/m3] were prepared
brines have extremely high levels of dissolved electro- with 14.2-lbm/gal [1702-kg/m3 ] brine and technical
lytes compared with typical formation waters. It is well grade CaC12 or CaBr2, or reagent grade ZnBr2. The
known that radical reduction of electrolyte content can reference brine was a 3% NaCl solution prepared with
ahnck cfays arrd result in swefling, dispersion, and loss reagent-grade cbernicals. SoItroITM, a refined low-
of perrneabilky. Other more subtle effects of unfavora- aromatic-content oif, was usad in the two-phase study.
ble fluidlreck interactions have been reported. 2 Siigle- AU fluids were filtered through a 0.45-ym membrane ilk
phaae flow testa were conducted in part to asaess pnten- before use.. No corrnsion iniibitora were added to any
tial unfavorable fluidhck irrteractiona with typical aand- brines. Potter5 examined formation +rnage effects of
stmre reservoirs that might result from Ore ir3creaae ir3 corrosion itilbltora.
electrolytes aasnciatcd with dense brine completion fluid.
High brine electrolyte concentrations a3so incrcaae the Apparatua . . ...
pmaibility of unfavorable interaction with forrrretion These tests were carried out in a Hassler-type sleeve core
waters. Our resrdta indicate that unfavorable interaction holder. Alf metal surfaces contacting heavy brines were
with formation waters are the major potentiaf causes of Haatalloy CTMtn eliite corrosion and plugging pmb-
formation damage with heavy brines. Iems resulting from corrosion prnducta. Precision dk-
H@ brine viscnsitiea might lead to peer brine displace- placcment pumps were used in the conatsnt-flow-rate testa,
ment by oif during well flowback after formation inva- and a regulated N2 supply waa used in the constant-
aion. Two-pbaae flow tests were conducted to investigate pressure testa. A regulated backpressure of at least 300
the impact of thk effect on oil permeabfity. Our testa psi [2.07 MPa] was applied in all constant-rate tests.
show that recovery of oil permeability after invasion by Cores were confined in elastomer sleeves with a cOn-
heavy brines ia slow but ultimately complete. These re- stant confining pressure of between 800 and 1,200 psi
sults are conaiatent with established theories of fluid flow [5.52 and 8.27 MPa]. Tests were conducted at constant
in pernus media. 34 tempsraturea between room temperature and 150”F
[65.6”C]. In the teata at elevated temperatures, a beat ,ex-
Coraa
changer waa used to bring fluids up to test temperature
Cores tested in thk program (ace Table 1) were selected before they contacted the cores.
to span the range of clay types and crmtents normally en- In the constantdifferential-pressure two-phase tests,
countered in sandstone reservoirs. Core plugs abnut 2 in. flow rates were meaaured by observation of brine and oil
[5 cm] long and 1 in. [2.5 cm] in diameter were drilled levels in a burette. Injection pressure of lCKIpsi [0.69
with either fiquid N2 or 3% NaCl aa the cutting fluid. MPa] waa used with Banders coma and 30 psi [0.21 MPa] .
When flow-tested with fresh water, one core of each type with Berea. The backpressure regulator was eliminated
suffered at leaat 90% reduction in penneabilky, demon- to minimize apparatus volume. Apparatus volume was
strating the presence of chemically active clays. meaaured by a series of displacements witbout the core.
Fluid saturations in cores were determined by material
Fluids
balance.
.Comrnercial 19.2-lbm/gal [2301-kg/m 3] CaBr2 /Zrd3r2,
15.7 -lbrn/gal [188 I-kg/ms ] CaC12/CaBr2 /ZnBr2, [Link] Teata
14.2-lbm/gal [1702-kg/m3] CaBr2, and 11.6-lbnr/gal Prowdrmes. AUmeawere evacuated and then filled with
[1390-kg/m3] Ca~2 brines were used. We made brines 3% NaCl solution (reference brine). They were mounted
OwWhL1934acclew.1 PeIrolwn Er@wm and flowsd with reference brine at a rate of 5 to 10 cm3i13

432 SPE Production Engineering, November 1986

 I TABLE 1–COMPOSITION OF CORE TYPSS USED IN THIS STUOY I

Quartz ‘ Feldspar Dolomite Clay

Care Type (%) (%) (%) (O,@ Type
Bandwra” 65 10 10 15 70% kaolinite,
30% illite
Berea” 91 4 4 100% kaolinite
Gulf coast’ 90 : 5 70% kaolinite,
3Q%illite
Sandpack”” 92.5 7.5 100% motrtmofillonite

[Link] analysb,
..htmlr,ti bymalwh!s.1 Prepamllc.n.

TASLE 2–SINGLE-PHASE FLOW TESTS

.-.
Retum-
, Test Brine Exposure Time Pe&:t:lity
Core TYP8 ($d) (Ibmlgal) (hours)
Bandera 11.5 13.4~ 24 1.0
Sandera 11.5 [Link] 24 1.0
Bandera 21.5 14.2” 24 0.71
Berea 72.0 14.2” 24 0.75
Bandora 11.5 [Link] 24 0.77
Bandera 62.0 14.9d 24 0.90
Bandera 103.0 14.9” 24 1.0
Serea 213.0 14.9f .24 1.0
Sandpack 9.0 15.79 36 1.0
Sandpack 7.4 19.29 50 1.0
Gulf ouast 146 19.29 20 1.0
: Pwnw.blllty103% NaC1.
0aCl,lCaBr2
;CaBr,
~ ~Br2. -% casr2.
:8% 2.0,2,32%cm,.
9TG~zn~r,.~~ ca~,,.
Ca3r2mw2 mnlwc!A N,”*

for a minimum of 12 hours. In all the tests repcn’tedhere, of damage by more thai a factor nf two, and brines for-
permeability was stable for at lesst the last 6 hours. This mulated with 8 nnd 16 wt% Zxd3r2 csuse no dmnage. at
value is the initial undamaged permt?abfity used in these all. High-mnccntradon CaBr2 brines are neutral to s@ht-
tests. Test high-density brine was then flowed at a rate ly alkaline, but zinc ion is acidic. The etktiveness of
of 10 cm3/h for a minimum of 24 hours (34 PV’s). zinc ion iu counteracting the damage suggest3 thst Oreim-
Reference brine was reintroduced into the core and flowed pairing material is acid-soluble.
at between 5 and 10 cm 3h until full permeability was Two calcium salts that might lead tn the obsem?d ef-
recuvered or until permeabdity remsined st2hle for a mini- fects are Ca(OH)2and CaC03. C![Link] were
mum of 24 hours. present in the cores. Precipitation of either sslt could be
induced by a temperature increase. 6 Samples of CaSr2
Results and Discussion. Results of the single-phsse brine brine and mixtures of C6Br2 brine and reference brine
flow tests are summarized in Table 2 and Fig. 1. An ex- core effluent were heated in an attempt tn isolate a precipi-
amination of these results shows that most of the brines tate, but none WBSever observed We can only infer that
tested were not damaging. Damage was observed only a crystal surface may [Link] m indum precipitation.
with brines having densities beeween 14.2 and 15.0 On tie basis of these data, the impairing material cannot
ibm/gal [1702 and 1797 kg/m3], and then only for some be identified positively. These results, which 616highly
brine cornpsitiuns. The dme requked for complete recov- significant despite the kickof pusitive identification, show
ery of perrneabfity increases with brine viscusity. huse that highly concentrated brines, particularly those con-
the damage effects appear to be sssuciated with brine com- -g cdtiuM, have a tendency to cause precipitation
pmition nnd not core type, we wm infer that damage is of insoluble salts.
not caused by unfavorable clay interactions. Field sud laboratory tests7 show that incompadbtity
Examination of the penneabti~-vs.-time curves for a between calcium chloride completion fluids and fornra-
number of 15.O-lbm/gal [1797-kg/m3 ] brines (Ilgs. 2 and tion watm can result in d2m2ging in0rg2nic precipitates.
3) shows that the damage occurs during heavy brine flow Riley et al. 7 indicated that thk problem is more severe
aud that the amount of damage increases with increasing if fornrmion watera contain significa!rt sulfate or bicar-
volume throughput. An oil mmun-permeabi@ test (dk- bonate concentrations.
cussed in detail in the following section) shows the same
magnitude of damage, indicating that the impairing ma- Tw@Phase Tests
terial is not significantly soluble in reference brine. For- When a .&ter-basd completion fluid invades an oil-
mulation of a brine with 4 W% ZuSr2 reduces the rate bearing formation, much of the oil is displaced frnm the
SPE %xiuctkm Engineering, November 1986 433
 ~~-% .
---=--
----- _ _ ‘\
_—— —- ....
....
0.9 — ..- 1 .. .
f -.’
5 =,
s 0.,
— ,x 2.s,2
,---

5 ,
— — 4s .?”s,2 ,
I
I.’ #
-. . . . .. ,“s,2
f .’
.! -
.!
/----—--—
$0.4–

0“:=
—..
—...

i ---- “ .“

., - -—. .

, , , !
‘! ..204! “-X.” w !,.
,.., -, ,,: 1“ “w-r.

Fig. I—Return permeability ratio vs. time for displacement Fig. 2—Effecf of 14.9. to [Link]/gaI brines on perme
of heavy brines. bllity.

nem-wellbore region. Full production potential camtotbe dictions based on immiscible disulacement3,4 theow can
realiied until thii iniccted water is dkdaced by oif flow. account for the observed sahmtion and oil permeably
The two-phase test; were conducted io meas~e the ef- changes. In two tests, oif flow was continued until the
fect of heavy brine completion fluids on these two reference-saturation condhion was obtained. Both return
processes. “ penneabfity and PV of flow required to attain these con-
Two-phase tests arc complicated because permeability ditions were measured.
toaziven fluid deoends on its saturation. The reference A simpler, constant-flow-rate oil/water displacement
(und”magcd) con&ion ii these tests must therefore be test was conducted to determine whether a reduction in
defined in terms of the permeability to a known fluid at oil permeability was observed with 14.2-lbm/gal
a known saturation. In these tests, the reference condl- [1702-kg/m3] CaBr2 brine. In this test, oil permeability
tion was defined as the oil permeability and saturation at was compared at a singfe saturation condition.
an oil fractional flow of 0.99 where the oil phase is Soltrol
and the water phase is 3 % NaC1. Rehtm-pemteabfity tests Procedures. Cores were first saturated with reference
mrist consider not only the permeability recovery attire brine under vacuum and inserted in the core holder. Per-
reference saturation, but afso the rate of recovery of the meab~lfy at 100% reference brine (3’%NaCl) saturation
reference stste. was then measured. Oil was flowed through the core un-
A series of constant-pressure oifhine displacement ex- til the oil frsctionaf flbw was 0.99, the endpoint fraction-
periments was conducted to measure the effect of al flow chosen for these rots. Reference brine was
19.2-lbm/gal [2301-kg/m3] brine on oil permesbii~ and reintroduced into the core and flowed until the endpoint
saturation. Oil saturation and permeability at the refer- fractional flow was obtained. This cycle (oil followed by
ence oil fractional flow of 0.99 after exposure to heavy reference brine) was repeated. The endpoint oil satura-
brine was compared with the reference condition. Pre- tion and permeability usually varied slightly between the

.B.. -
-——
,=--
5-
$U
z

c.. -
f a 9“”
i _ “4 .,.,,..,
[Link]”.
. -
---”:: %-
I
‘! w m
M“”.-, . m.

Fig. &DlspIacement of brine by 011.

Fig. 3-Effect of exposure to 14.2-lbm/gal brine on 011por-
meahilltv.

434 [Link].g,November1986
 I TABLE 3–SATURATION CHANGES I

Core
BD3
BD6
9S6
A79
A66
(#d)
3.2
3.4.
80
103
98

TABLE 4-RETURN-OIL-PERMEABILITV RATfOS ANO DISPLACEMENT

VOLUMEB FOR OILFLOOOS

Return Permeability Displacement FV,S

L&
0.06
0.0s
0.14
0.15
0.15
1“ -.

Ratios Endpoint
Core km .— Equal SW
(brine) (red) Endpoint Equal SW Reference Brine Heavy Brine Heavy Brine
BD3 2.6 0.44 5,1 8.0
(~;:)
1,8 0.41 1.0 3.4 4.1 96
(19.2)
98B 55 0,48 1,0 4.9 7.5 196
O&.:)
66 0.47 3.8 5.3
(&
50 0.32 2.3 3.4
(19,2)
Berea 41.5 0.76 4.0 120
(14.2)

TABLE 5—COMPARISON OF EXPERIMENTAL AND CALCULATE VALUE3

Waterflood
Mobifity”
Core Ratio
BD3 0.15
BD6 0.092
98B 0.11
A79 0.17
A68 0.16

.M -~, 396NaCltisolacingSc-trol,
m

first (displacing 100% brine) aod later cycles (oil displac- Results rmd Discussion. Remvery of oil panncabfity af-
ing brine in the presence of trapped oil). Oil was then ter exposure to heavy brines follows a pattern similar to
flowed to the endpoint fractional flow a third time. III all rrxovery of brine pcrmeabdity. As Fig. 4 shows, more
tests, the oil permeability and saturations measured at the oil flow is reqoircd to recover parrneability after exposura
end of the second and third oilfloods were identical, in- to 19.2-lbm/gsl [2301-kg/m3 ] brines than after exposura \
dicating that a stable distribution of oil in the pora net- to 3 % NaCl, but complate recovery is observed. Core pcr- .
work had been obtsined. The oil saturation and mcabfity decreases during flow of 14.2-lbm/gsl [1702-
parrnesbility so defined is the reference (umfmnrtged)con- kg/m3] brine, snd recovery is incomplete during oil flow
dition fnr these tests. just as it ia during brine flow (see Fig. 3). The implica-
Test brine (19.2 lbndgd [2301 kg/m3]) was flowed tions of these results with 14.2-Ibm/gsl [1702-kg/m3]
into the core at the same injection pressure as the refer- brine have already been discussed. A more detailed ex-
ence brine floods (30 psi [0.2 MPa] for Berea, 100 psi arnhmtion of the 19.2-ibm/gal [2301-kg/m3] displa&mmt
[0:69 MPa] for Bandera cores). Flow was continued un- experiments indicates that the principal cause of the slow
til a brine fractional flow of 0.99 wss obtained. Brine satu- recovery of permeab~@ is the low mobility of the heavy
ration. was measured, and oil was injwted at the same brine.
pressure rmdl the oil fractional flow W6S 0.99. In two tasts, These mobfity et%cts are demnnstratcd by the satura-
oil flow was continued until the reference-smta oil satu- tiorrs observed at the 0.99 fractiorml flnw endpoint (see
rmion was attained. Table 3). The dstB show that little dhTerence in endpoint
SPE Prodwdon “EIIgixmimg, November 1966 435
 sshrmtion is obaarved between reference-brine snd heavy- highdensity brines. Use NaCl, KC1, or NH4CI brines
brine displacements of oil. On the other hand, when oil instead of seawater.
displaces a heavy brine, a higher water saarration is ob- 3. Increaaed viscosities (without viscosifiers) observed
served at endpnirrt thmr when oil &places reference brine. in these brines slow the displacement of brine by oil or
The eonsequericx2 of the incressed water aatumtion is water. At a given water saturation, however, oil perrne-
a reduction of endpeint d perrneabtity by a factnr of mvo abfity is not reducsd hy the presence of heavy brine.
after [Link] to 19.2-lbdgaJ [2301-kg/m3] brint h
Table 4, kro is the reference oil permeability defined aa Nomenclature
the permeability tGSoltml at SDoil fractional flOW of 0.99 k = brine permeability at 100% saturation, md
when the water phaaa ia 3% NaC1. The endpoint retmn- ki = brine permeability at 100% saturation
pmmeabtity ratio ia the perrneabfity to Soltrol at the
before exposnre to heavy brines, md
reference oil fractional flow (0.99) when the water phase
k. = oif perrneabdity at oif fractional flow of
is 19.2-lbm/gal [230 l-kg/m3] ZnBr2 brine divided by
km. In Cores BD6 and 98B, oil flow wss continued un- 0.99, Soltxol displacing heavy brine, md
til the reference water saturation (see Table 3) waa ob- km = reference oil permeability at oil fractional .—
tained. In bGth maes, oil pr33Jeabtity w88 completely flow of 0.99, SOItfOldisplacing 3%
recovered. NaCl, md
More oil flow waa mqukrd to displace hesvy brine from kw = reference brine permeability at brine ,
tbe cores compamd with tbe refenmm state. The diapl8ce- fractimrd flow of 0.99, 3% NaCl
rnent volume required to reach the oil fractional flow when displacing SoltTol, md
hesvy brine wss displaced [Link] 1% times that re- M = mobility ratio
quired in the reference cas~ i.e., 1M times aa much oil SW = average water saturation
flow was reqkxi to achieve lower oil Wmration and per-
PO = oil viscosity, cp [Pa.s]
meabiihy. Displacement of heavy brine to the reference
water aaturstion and a retnrrr-permeability ratio of 1.0 re- P ~ =’ water viscosity, cp [Pa.s]
quifed 30 to 40 dines as much oil flow. These TWUItS show
that recovery of oil permeabfity is slowed when high- Aoknow[edgmenta
density brines invade the formation. I thank G.C. Adarna for conducting the experiment and
Applicable theories of displacement under segregated A. Abmrna and D.J. OMeara Jr. of Shell Development
flow COJJditiOJIS3,4 provide a way to calculate the effect Co. for msny helpfol discussions. I SISOthank Shelf De-
of chmrges in fluid viscosity on saturation and the return- velopment Co. for permission to publish this paper.
penneabtity ratio at a given fractional flow. In these ex-
psrimerm, dmplacementa occur under segregated flow
conditions becanse of gravi~ forces for brine diaplacin 1. Spies,RJ..1 al.: “Field [Link] Hish-Dmsiq. Brines
oil 3 end viscous instability for oil displacing brine. % as Completion Fluids,-, JPT (May 1983) 881-88.
Fractional flow data obtsined in this study are consistent 2. Reed, M.G.: “Fomtion Permeability Damage of Mica AlrerafiorI
with tbia interpretation. Remits of the calctrlatiom, sunr- and Caxbmate Dissolution,,, papx SPE 6CC9presatxd at tie 1976
SPE Annual Technical Confemmceand Exhibition, New Grl&ns.
mdrized in Table 5, indicate that the saturation change Ocr.3-6.
and rctrrrn-permeab~lty decrease are cauaed by the in- 3. D&e, L.P.: Fumdzmemafs of Resermir En@uering, Ekvi.w
creased brine viscosity ad not by any unfavorable 3cietic Publishers, Amsferdam (1978).
fhrid/rock or fluid/fluid interactions. 4. .. .. of
van [Link]. P. and [Link] Pod. C.: ‘-[Link] Dmmimio”
Water-Drive Recesses Involving Viscous Fingering, S, Trims.,
~ (1958) 213, 103-12.
Conclusions 5. Porrer, G. F.: ,&FonnarionDmp,e of Brine Corrosion Jnbibitors,>,
1. Brines in the 13.4-to-19.2-lbrn/gal [1606-to-2301 - papr SPE 12495 presentedat rhe 1984 SPE Symposium o“ For-
kg/m3 ] range formulated with various combhtiona of madcm Damage Co”tml, Bakersfield, Feb. 13-14.
6. Cowan, J.C, and Weintirt, D.J.: Wakr-Fomed scale Deposirs,
CaC12, CaBr2, snd ZnBr2 did not cause formation Gulf Publishing Co., Housfon (1976).
damsge became of unfavorable clay interactions in any 7. Sky, S.M. er al.: “ConuoUing Jncompalibiliriesof Formation ahd
of the sandstone cores tested. Kill Weight Fluids Durin8 Workover,”papmSPE13131 presamd
2. Damage observed with highly czmcentrated (> 14.2 at the 1984 8PE Ammal Technical [Link]”.x and Exiibitkm,
Houston, Sept. 16-19.
lbrn/grd [ >1702 kg/m3]) brines formulated with CaBr2
and/or CaC12 is caused by precipitation of MJacid-soluble
calcium salt. Brines with densities greater thsn 14 Ibm/gal S1 Metric Conversion Factor
[1702 kg/m3] should be formulated with a minimmn of Ibm/gsl X i.198 264 E+02 = kg/m3
8% ZnBr2 to lower pH snd tG prevent precipitation.
Tbcse results suggest thst incompatibfity with formation SPEPE
watsra is a potcntkd cause of formation damage with high- OdQlna! rmnuec,l#-bed 1“the3C+W01PelrohumEnglrwma officeMarch1,,
denaity brine completion fluids. Special care should be [Link].x@cce#tifor~,bllcmllm Od. .4, 1S9S, Rwleed IIMIIUSMF4 mlvti June
11, [Link]+r(SP31SS1 7) ITS Pmmted at lha 1S&3 SPE P,cduc!lmOFSr@Or@
tskan to prevent timg of seawater (ligh in SOT ) with a~~lum heldh OkJnhoma city, Maw lc-12.

436 SPE Reduction %ginecrinz, +%.emkr \986