Microbial CaCO3 mineral formation and 
stability in an experimentally simulated high 
pressure saline aquifer with supercritical 
CO2
Authors: Andrew C. Mitchell, Adrienne Phillips, Logan 
Schultz, Stacy Parks, Lee Spangler, Alfred B. 
Cunningham, & Robin Gerlach
NOTICE: this is the author’s version of a work that was accepted for publication in International 
Journal of Greenhouse Gas Control. Changes resulting from the publishing process, such as peer 
review, editing, corrections, structural formatting, and other quality control mechanisms may not 
be reflected in this document. Changes may have been made to this work since it was submitted 
for publication. A definitive version was subsequently published in International Journal of 
Greenhouse Gas Control, 15, July 2013. DOI#10.1016/j.ijggc.2013.02.001.
Mitchell AC, Phillips A, Schultz L, Parks S, Spangler L, Cunningham AB, Gerlach R, "Microbial 
CaCO3 mineral formation and stability in an experimentally simulated high pressure saline aquifer 
with supercritical CO2," International Journal of Greenhouse Gas Control, July 2013 15: 86–96.
Made available through Montana State University’s ScholarWorks 
scholarworks.montana.edu 
Micro
experi
superc
Andrew
Lee Spa
a Institute of Ge
b Center for Bio
c Department o
d Department o
ABSTRACT
The use of m
sequestered c
precipitation, 
environments
core was inoc
the core. The 
the flow react
increased ove
in the reactor
calcium was 
and atmosphe
3.6% after 48
depressurizati
resulted in a m
scCO2 satura
mineralization
scCO2 injecti
Introduc
Geolog
generated
(scCO2; 
bearing f
Zakkour 
The prim
injection
permeabi
less visco
(Nicot et
One pr
biofilms 
attached 
matrix ofbial CaCO3 mineral formation and stability in an 
mentally simulated high pressure saline aquifer with 
ritical CO2
 C. Mitchell a,b,∗, Adrienne Phillips b,c, Logan Schultz b,c, Stacy Parks b,c, 
ngler d, Alfred B. Cunninghamb, Robin Gerlach b,c
ography and Earth Sciences, Aberystwyth University, SY23 3DB, UK
film Engineering, Montana State University, Bozeman, MT 59717, USA
f Chemical and Biological Engineering, Montana State University, Bozeman, MT 59717, USA
f Chemistry and Biochemistry, Montana State University, Bozeman, MT 59717, USA
icrobiologically induced mineralization to plug pore spaces is a novel biotechnology to miti-gate the potential leakage of geologically 
arbon dioxide from preferential leakage pathways. The bacterial hydrolysis of urea (ureolysis) which can induce calcium carbonate 
via a pH increase and the production of carbonate ions, was investigated under conditions that approximate sub-surface storage 
, using a unique high pressure (∼7.5 MPa) moderate temperature (32 ◦C) flow reactor housing a synthetic porous media core. The synthetic 
ulated with the ureolytic organism Sporosarcina pasteurii and pulse-flow of a urea inclusive saline growth medium was established through 
system was gradually pressurized to 7.5 MPa over the first 29 days. Concentra-tions of NH4+, a by-product of urea hydrolysis, increased in 
or effluent over the first 20 days, and then stabilized at a maximum concentration consistent with the hydrolysis of all the available urea. pH 
r the first 6 days from 7 to 9.1, consistent with buffering by NH4+ ⇔NH3 + H+. Ureolytic colony forming units were consistently detected 
 effluent, indicating a biofilm developed in the high pressure system and maintained viability at pressures up to 7.5 MPa. All available 
precipitated as calcite. Calcite precipitates were exposed to dry supercritical CO2 (scCO2), water-saturated scCO2, scCO2-saturated brine, 
ric pressure brine. Calcite precipitates were resilient to dry scCO2, but suffered some mass loss in water-saturated scCO2 (mass loss 17 ± 
 h, 36 ± 7.5% after 2 h). Observations in the presence of scCO2 saturated brine were ambiguous due to an artifact associated with the 
on of the scCO2 saturated brine before sampling. The degassing of pressurized brine resulted in significant abrasion of calcite crystals and 
ass loss of approximately 92 ± 50% after 48 h. However dissolution of calcite crystals in brine at atmospheric pressure, but at the pH of the 
ted brine, accounted for only approximately 7.8 ± 2.2% of the mass loss over the 48 h period. These data suggest that microbially induced 
, with the purpose of reducing the permeability of preferential leakage pathways during the operation of GCS, can occur under high pressure 
on conditions.
Keywords:
Biofilm
Calcium carbonate
Ureolysis
Supercritical CO2
CO2 leakage
Permeabilitytion
ic carbon sequestration (GCS) is one strategy to reduce the emission of greenhouse gases 
S 
und
nd 
ites through the combus-tion of fossil fuels. GC
critical point = 31.1 ◦C and 7.39 MPa) into 
ormations, deep unmineable coal seams, a
and Haines, 2007). 
ary concern during the operation of GCS s
 well, other monitoring or abandoned wells an
lity cap rock (Pan et al., 2009). Risk of leakag
us than the res-ident pore fluid, which allows i
 al., 2009).
oposed method to mitigate leakage and enhan
to plug porous media (Mitchell et al., 2009). Bio
to surfaces, and embedded in extracellular poly
 mostly polysaccha-rides and proteins (Costertoinvolves the injection of supercritical CO2 
erground for-mations such as depleted oil 
deep saline aquifers (White et al., 2003; 
 is the possibility of leakage through the 
d through fractures or faults in the low-
e arises because scCO2 is less dense and t to migrate upward through the formation 
ce CO2 storage is the use of engineered 
films are assemblages of microorganisms, 
meric substances (EPS). EPS is a hydrated 
n and Stewart, 2001; Lewandowski
and Beye
media an
taminant
Gerlach a
of life for
and Beye
Bacillus m
cores und
32 ◦C), (2
cores by a
(3) biofilm
viability w
2009), an
log reduc
for plankt
cells. The
and scCO
It has
of minera
porous m
2009, 201
enhanced
ancient se
carbonate
groups fo
in solutio
and in th
mediated
the form
enhanced
Mitchell
an impor
through a
the syste
centratio
increase i
solid calc
Urea h
first step
carbonic
The amm
bonate (H
(Eqs. (3)
an increa
carbonate
cipitation
Chen, 200
ing the h
by Eq. (7)
CO(NH2)2
NH2COOH
H2CO3 ↔
2NH3 +2
HCO3− +
CO32− +C
CO(NH2)2
Other
and iron r
(Van Lith
usually d
easily as u
subs
l ure
ate i
by u
g), th
., 20
eral-
plug
ible s
oach
ccess
redu
s far
sity
ugh e
s Sa
e the
nven
ston
onst
s typ
d to
ontro
odat
d for
men
; Va
ation
ren e
and
; Mi
atm
nde
to hy
also
nder
onat
of 3
temp
3 so
3 ac
ere,
on ca
ge e
erate
nthet
3 m
e, in
ss th
xper
High
high
lysis
r hi
press
pres
hetic
ko, Tnal, 2007). Biofilms can reduce the permeability of porous
d have been engineered to control environmental con-
s and flow in oil reservoirs (Cunningham et al., 2003;
ndCunningham, 2010). Biofilms are the predominantway
most microorganisms in natural systems (Lewandowski
nal, 2007). Previous experiments demonstrated that (1)
ojavensis biofilms can be grown in Berea sandstone
er high pressure and scCO2 injection conditions (8.9MPa,
) the biofilm can decrease the permeability of sandstone
pproximately two orders ofmagnitudewithin a fewdays,
s can resist the challengeof scCO2 exposure andmaintain
ith only a slight increase in permeability (Mitchell et al.,
d (4) biofilms formed by B. mojavensis exhibited only a 1
tion in viable cell numbers compared to a 3 log reduction
onic (free floating, not attached to a surface) B.mojavensis
se data demonstrate biofilm viability under high pressure
2 conditions relevant to GCS (Mitchell et al., 2008).
also been proposed that the biofilm-enhanced formation
ls, such as calcium carbonate, can plug free pore space in
edia and result in leakage reduction (Cunningham et al.,
1; Dupraz et al., 2009; Mitchell et al., 2010). Microbially
mineralization has long been recognized in modern and
diments (Van Lith et al., 2003). Biofilms enhance calcium
nucleation by offering negatively charged functional
r cation adsorption and by metabolically driven changes
n composition and pH in the biofilm microenvironment
e bulk solution (Schultze-Lam et al., 1996). Microbially
urea hydrolysis is one mechanism which can enhance
ation of calcium carbonate and is being investigated for
GCS (Cunningham et al., 2009, 2011; Dupraz et al., 2009;
et al., 2010). In this process, bacteria hydrolyze urea,
tant nitrogen compound found in natural environments,
series of reactions which raise the pH and alkalinity of
m through an increase in bicarbonate and carbonate con-
ns. In the presence of divalent cations, such as Ca2+, this
n alkalinity can lead to the saturation and precipitation of
ium carbonate (CaCO3).
ydrolysis proceeds through a number of reactions. In the
, urea (CO(NH2)2) is hydrolyzed to ammonia (NH3) and
acid (H2CO3) (Eqs. (1) and (2)) (Burne and Chen, 2000).
onia and carbonic acid equilibrate in water to form bicar-
CO3−), ammonium (NH4+), and one hydroxide ion (OH−)
and (4)). The increase in hydroxide ions corresponds to
se in pH, which shifts the bicarbonate equilibrium to form
ions (CO32−) (Eq. (5)). In the presence of Ca2+, CaCO3 pre-
occurs (Eq. (6)) once saturation is exceeded (Burne and
0; Castanier et al., 1999). The overall reaction summariz-
ydrolysis of urea and precipitation of CaCO3 is described
.
+H2O → NH2COOH + NH3 (1)
+ H2O → NH3 +H2CO3 (2)
HCO3− +H+ (3)
H2O ↔ 2NH4+ +2OH− (4)
H+ +2OH− ↔ CO32− +2H2O (5)
a2+ ↔ CaCO3 (6)
+2H2O + Ca2+ ↔ 2NH4+ +CaCO3(Overall,process) (7)
microbial metabolic processes, such as sulphate, nitrate
as a
teria
oper
ated
CO2(
et al
(min
will
cess
appr
be a
ing,
zone
visco
thro
Carlo
twic
in co
sand
dem
rock
ulate
by c
T
gate
treat
2010
biliz
War
ery
1996
near
ate u
able
It is
als u
carb
a pH
and
CaCO
H2CO
H
itati
stora
mod
a sy
CaCO
brin
asse
2. E
2.1.
A
ureo
unde
tial
high
synt
(Temeduction could also be used to promote CaCO3 formation
et al., 2003; Van Paassen et al., 2010a), but these processes
epend on microbial growth and cannot be controlled as
reahydrolysis,which simply requires the additionof urea
diameter
internal d
the core
allow thetrate, without the actual requirement of cell growth. Bac-
a hydrolysis also does not require light, and can therefore
n dark subsurface environments. The pH increase gener-
rea hydrolysis also dramatically increases the solubility of
us enhancing the rate of CO2 solubility trapping (Mitchell
10). While there is no net mineral precipitation of CO2
trapping) beyond that from the urea, the mineral formed
subsurface pore spaces. This is particularly useful for inac-
ubsurface environments, where ‘traditional’ engineering
es are not useful. For example, while injection wells may
ible to traditional sealing approaches, such as cement-
cing permeability proximal to the well bore or in fracture
from the well requires novel approaches due to the high
of such traditional sealing materials. Bacteria can move
xtremely small pore throat sizes of 0.4m<(Mitchell and
ntamarina, 2004) and where pore throat size is at least
cell size (Jenneman et al., 1985). Since pore throat sizes
tional reservoir rocks range from about 2 to 0.03m in
es, and from 0.1 to 0.005m in shales (Nelson, 2009) this
rates that bacteria can effectively be transported through
es present in common GCS sites and can likely be manip-
grow and precipitate minerals on space and time scales
lled injection of nutrients (Cunningham et al., 2011).
e, ureolysis-inducedCaCO3 precipitationhasbeen investi-
a number of engineering purposes including wastewater
t (Hammes et al., 2003), soil stabilization (Harkes et al.,
n Paassen et al., 2010b; Whiffin et al., 2007), immo-
of radionuclides (Mitchell and Ferris, 2005, 2006a,b;
t al., 2001), and mineral plugging for enhanced oil recov-
carbon sequestration (Dupraz et al., 2009; Ferris et al.,
tchell et al., 2010). All of these studies have been made at
ospheric pressure conditions. However, in order to oper-
r GCS conditions, ureolytic organisms must be viable and
drolyse urea under high pressure conditions (∼7.5MPa).
important to consider the stability of the CaCO3 miner-
GCS conditions, given the potential for the dissolution of
eminerals in CO2 charged acidic brine,which typically has
.5–4 depending on brine chemistry, formation lithology,
erature (Kaszuba and Janecky, 2009). LowpHcan increase
lubility by shifting thebicarbonate equilibriumtoenhance
tivity (Stumm and Morgan, 1996).
we investigate whether ureolysis-induced CaCO3 precip-
n occur under conditions that approximate subsurface
nvironments, using a unique high pressure (∼7.5MPa)
temperature (32 ◦C) pulse-flow reactor system housing
ic porous media core. We also consider the stability of
inerals in scCO2, and in likely subsurface fluids, such as
the presence of and equilibrated with scCO2, in order to
e stability of biofilm formed minerals in the subsurface.
imental methods
pressure flow reactor setup and synthetic porous media
-pressure flow reactor was built to investigate whether
and CaCO3 precipitation in porous media could occur
gh pressure conditions. The system allowed a differen-
ure to be established between an input and an output
sure accumulator, causing media to pulse-flow through a
porous media core housed in a Hassler-type core holder
ulsa,OK). The synthetic corewasproduced froma2.54 cm
, 5 cm long high density polystyrene rod, with 6×1mm
iameter glass capillary tubes running lengthwise through
(Supplemental Information, Fig. SI1). The capillary tubes
aqueous phase to pass through and microscopy to be
Fig. 1. (a) S react
synthetic po an exp
core holder
performe
high-pres
steel tubi
trol temp
service p
stand pre
Prior t
tor and in
in 6g L−1
sulphate
Richardso
N (TEMC
and core
treatmen
and an ov
hydraulic
connecte
an incuba
2.2. Flow
Sporos
forming,
the prese
and Ferri
was grow
high
lized
. (199
mM
entr
S. pa
25.2
n ex
∼40
injec
g tig
rem
ped
ic pu
ium
nal s
medi
tion
e th
core
syste
ced bchematic diagram of high pressure (∼7.5MPa) moderate temperature (32 ◦C) flow
rousmedia corewithCaCO3 precipitates visible on the inside of the glass tubes after
.
d on any biofilm and mineral precipitates formed. The
sure system was constructed of ¼” (6.35mm) stainless
ng and Swagelok fittings housed in an incubator to con-
erature at 32±0.2 ◦C. The media reservoirs were water
iston-type accumulators (Parker, Inc.) designed to with-
ssures of 21MPa (Fig. 1).
o loading the synthetic core, the input media accumula-
fluent tubing were sterilized by exposure to 10% bleach
TWEEN 80 solution, followed by 2.25g L−1 sodium thio-
solution, followed by a 70% ethanol solution (Barkley and
n, 1994). Next, the synthetic corewas loaded into a Buna-
O) sleeve and the core holder was reassembled. The core
sleeve were not sterilized by autoclaving nor chemical
t. The annulus of the core holder was filled with water
erburden pressure of about 10.5MPa was provided by a
jackpump. Thecoreholder andmedia accumulatorswere
d to the high-pressure flow system which was housed in
tor at 32±0.2 ◦C (Fig. 1).
reactor inoculation and operation
arcina pasteurii (ATCC 11859, gram-positive, spore-
the
steri
et al
187
conc
the
but
tatio
and
was
fittin
lum
pum
stalt
med
inter
the
solu
suriz
the
the
induurease positive), known to induce CaCO3 precipitation in
nce of urea and dissolved Ca (Fujita et al., 2008; Mitchell
s, 2005, 2006b) was used for the experiments. S. pasteurii
n to the exponential growth phase, prior to inoculating
24h, the
mulator.
the input
CMM−wor containing synthetic porous media core. Inset pictures show (b) the
eriment, and (c) the synthetic porousmedia core inside theHassler-type
pressure system containing the synthetic core, in filter
CaCO3 mineralizing medium (CMM) described by Ferris
6) which contained 3g L−1 Nutrient Broth, 333mM Urea,
NH4Cl, 25mM NaHCO3, and was pH adjusted to 6 with
ated HCl. The CMM excluded calcium for preparation of
steurii inoculum and flow reactor inoculation (CMM−)
mM CaCl2•2H2O was included during the main precipi-
periment (CMM+). To inoculate the core, it was isolated
ml of the exponential growth phase culture of S. pasteurii
ted using a 60ml sterile syringe attached to a Swagelok
htenedonto the influent sideof the coreholder. The inocu-
ained in the core for 16h and then sterile CMM− was
through the core at atmospheric pressure using a peri-
mp at 8.86mlh−1 for 16 days. The use of a rich growth
and urea allowed S. pasteurii biofilm to develop on the
urfaces of the core and for ureolysis to occur. On day 17,
a accumulator was filled with 1100ml of sterile CMM−
using a peristaltic pump and N2 gas was used to pres-
e system to 1.43MPa, with a differential pressure across
of 0.1MPa (Fig. 1). After being allowed to acclimatize to
m temperature, sterile CMM− was pulsed into the core,
y the constant differential pressure across the core. Aftermedia in the core was collected in the output media accu-
Sterile CMM− was then pulsed into the core again from
media accumulator. For the next 20 days, fresh sterile
as pulseddaily into the core, andmedium that had resided
Fig. 2. Pictures of the high pressure vessel (HPV) used to challenge CaCO3 precipitates with scCO2 and brine. HPV was also used to incubate Ca free mineralizing medium
(CMM−) de a in t
occurred at e Isco
image of th e HPV
in the cor
period, sy
7.5MPa±
core betw
accumula
procedur
Effluent f
pH, NH4+
2.3. This
S. pasteur
at pressu
iment, CM
inputme
tion, as de
to check
pressuriz
sure vess
2.3. Flow
Efflue
been puls
accumula
polycarbo
entific (F
calibrated
entific). A
Nessler A
results to
determin
Mass Spe
ronmenta
effluent w
forming
a daily b
Saline sol
plemente
30 rando
individua
in a 96 w
vidual we
nfirm
ere
+, in
deta
glass
wer
e.
Calci
ord
s und
ge si
2 sa
rated
value
weig
+ w
thp
ipita
le, un
was
rast
8h
ster
tant
onw
Dust
dica
wer
mat
2, sc
ric p
. Atm
ynth
by Kscribed by Ferris et al. (1996) and Ca inclusive CMM (CMM+) abiotically at 7.5MP
the high pressures used in this study. (a) Supercritical CO2 is pumped by a Teledyn
e HPV shows scCO2 in the headspace and scCO2 saturated brine in the bottom of th
e for 24h was collected in the effluent. During this 20-day
stem pressure was gradually increased from 1.43MPa to
0.1% while maintaining a differential pressure across the
een 0.1 and 0.55MPa±0.1%. On day 39, the input media
tor was filled with CMM+, and the same daily pulsing
e was continued, in order to induce CaCO3 precipitation.
rom the flow reactor was collected daily and analyzed for
, Ca2+ and ureolytic cell counts, as described in Section
procedure was performed in order to elucidate whether
ii viability could be maintained and ureolysis could occur
res at which scCO2 could be present. During the exper-
M− and subsequently CMM+ were sampled from the
dia accumulator in order to check for bacterial contamina-
scribed below. Control experimentswere also performed
ureolysis could not occur abiotically under pressure, by
ing CMM− and CMM+ to 7.5MPa in a separate high pres-
el (HPV) (Fig. 2) in the absence of S. pasteurii.
reactor effluent analysis
nt that had resided in the core for 24h and had then
ed out of the core was collected from the output media
tor and immediately filter sterilized through 0.2m
nate membranes and tested for pH using a Fisher Sci-
air Lawn, NJ) pH/ion/conductivity meter. The meter was
daily using pH 7 and pH 10 buffer solutions (Fisher Sci-
mmonium concentrations were determined using the
ssay (Mitchell and Ferris, 2005) by comparing sample
standards made with NH4Cl. Ca2+ concentrations were
ed using an Agilent 7500ce Inductively Coupled Plasma
ctrometer (ICP–MS) in Montana State University’s Envi-
l and Biofilm Mass Spectrometry Facility. Non-filtered
as used to determine the number of ureolytic colony
units (CFUs) as a measure of ureolytic cell viability on
asis. Effluent was serially diluted in Phosphate Buffered
ution, from 10−1 to 10−5 and plated onto agar plates, sup-
d with 37g L−1 Brain Heart Infusion, and 20g L−1 urea.
to co
pH w
CMM
ther
The
core
scop
2.4.
In
itate
stora
scCO
satu
pH
pre-
CMM
grow
prec
steri
tion
cont
24–4
with
cons
coup
con
perio
they
Infor
scCO
sphe
2.4.1
S
used
mly selected CFUs were individually picked and grown in
l wells containing 250L of BHI +20g L−1 urea solution
ell plate. Plates were incubated at 30 ◦C for 24h, and indi-
lls were tested for ammonium using the Nessler Assay
MgSO4•7
KCl, and 1
filter ster
thetic brihe absence of scCO2 to demonstrate that no abiotic hydrolysis of urea
(Lincoln, Nebraska) pump into the (b) view-ported HP(c) the view port
during an exposure (see also Video 1).
ureolytic activity (Mitchell and Ferris, 2005). CFU’s and
also measured from the influent CMM− and subsequently
order to check for ureolytic bacteria contamination. Fur-
ils are given in the Supplemental Information, Section SI1.
capillary tubes running lengthwise through the synthetic
e removed and imaged on a Nikon SMZ1500 stereo-
um carbonate brine and supercritical CO2 challenges
er to test the stability of ureolysis-induced CaCO3 precip-
er high pressure scCO2 conditions relevant to subsurface
tes, CaCO3 was challengedwith supercritical CO2 (scCO2),
turated brine (brine co-equilibrated with scCO2), brine-
scCO2, and atmospheric pressure brine at two different
s (pH 7.61 and 4, Table 1). Pyrex bottles containing
hed black polycarbonate coupons and containing 250ml
ere inoculated with 1ml of S. pasteurii in the exponential
hase, and incubated at 30 ◦C to induce ureolysis andCaCO3
tion onto the coupons. Coupons were also incubated in
inoculated CMM+ medium to confirm abiotic precipita-
not occurring. Black polycarbonate was used to provide
with the precipitated CaCO3 for light microscopy. After
of incubation and precipitation, coupons were removed
ilized tweezers and dried for more than 24h at 65 ◦C until
weight was achieved. Mineral not firmly attached to the
as gently blownoff the coupon using compressed air (Fal-
Off, Branchburg, NJ). The coupons were weighed again
lly over a period of 24h to confirm constantweight before
e imaged on aNikon SMZ1500 stereoscope (Supplemental
ion, Section SI2). Coupons were then challenged with
CO2 saturated brine, brine saturated scCO2 and atmo-
ressure brine, as described below.
ospheric pressure brine
etic NaCl rich brine of a 162g L−1 TDS (1.5M strength,
aszuba et al., 2003), containing 683mM NaCl, 16.4mM
H2O, 5.5mM CaCl2•2H2O, 480mM MgCl2•6H2O, 264mM
.79mMKBr, with a resulting pH of 7.61was prepared and
ilized. Mineral-covered coupons were placed in the syn-
ne for 4h, 24h, and 168h (Exposures 1–3, Table 1). After
Table 1
Summary of scCO2 and/or brine exposures of CaCO3 undertaken in the high pressure vessel.
Exposure Pressure Temperature Description
1 Ambient Room Coupons in brine for 4h, initial pH 7.61
2 Ambient Room Coupons in brine for 24h, initial pH 7.61
3 Ambient Room Coupons in brine for 168h, initial pH 7.61
4 1300psi 37 ◦C Coupons in scCO2 for 2h
5 1250psi 37 ◦C Coupons in brine-saturated scCO2 for 2h
6 1250psi 37 ◦C Coupons in brine-saturated scCO2 for 48h
7 1250psi 37 ◦C Coupons in scCO2-saturated brine for 2h
8 1250psi 37 ◦C Coupons in scCO2-saturated brine for 48h
9
10
11
removal f
residual b
above.
2.4.2. Sup
Coupo
tem was
flow via
pressure
maintain
4, Table
to accou
gauges.
diately t
imaged w
above.
2.4.3. ScC
350m
into the
coupons
three we
surized a
the brine
at 37±0.
bles and
to have r
urated br
portions
were exp
and supe
tor) at 8.9
6 and 8 r
the coup
remove a
weighed,
Exper
gate the e
coupons
the effec
and CO2
brine occ
and this w
under GC
saturated
scCO2 sa
MSU, equ
probe con
observed
brine (Ha
CO2 in h
., 20
Janec
ged a
in th
der t
e wa
is br
eral r
lutio
rime
of th
mat
esul
Ureo
-flow
he fl
ns w
xper
ually
in th
ation
(Fig.
ys (F
inan
chell
ent g
1.1
e (Fi
+ con
etect
not b
was
ia co
tartin
ric t
3). D
entr
,600
lable
.1×1
∼7.5
3000
g th
lyticAmbient Room
Ambient Room
Ambient Room
rom the brine, couponswere rinsed in DIwater to remove
rine and then dried, weighed, and imaged, as described
ercritical CO2 (scCO2)
ns were loaded into the HPV system (Fig. 2). The sys-
pressurized and scCO2 was pumped under constant
a Teledyne Isco high pressure pump to the HPV until
stabilized at 8.9MPa. Pressure and temperature were
ed at 8.9±0.4MPa and 37±0.2 ◦C, respectively (Exposure
1), well above that required for supercritical conditions
nt for the accuracy of the pressure and temperature
The system was depressurised after 2h, and imme-
he coupons were removed from the reactor, rinsed,
ith the stereoscope, dried and weighed, as described
O2 saturated brine and brine saturated scCO2
l of brine was added under ambient pressure conditions
HPV. Coupons were placed into the reactor so three
were wetted in the brine (bottom portion of reactor) and
re in the headspace of the reactor. The system was pres-
nd scCO2 was pumped into the HPV and bubbled through
until pressure in the reactor stabilized at 8.9±0.4MPa
2 ◦C. The increase in surface area between the CO2 bub-
the brine as well as the resulting mixing are assumed
esulted in sufficient mass transfer to produce scCO2 sat-
ine and brine saturated scCO2 in the lower and upper
of the HPV, respectively. Hence, the coupons in the HPV
osed to brine saturated scCO2 (in the headspace portion)
rcritical CO2 saturated brine (bottom portion of the reac-
MPa and 37 ◦C for 2h and 48h (exposures 5 and 7, and
espectively, Table 1). Immediately after each exposure,
ons were removed from the HPV, rinsed in DI water to
ny residual brine, imaged with the stereoscope, dried and
as described above.
imentswere also performed in order to separately investi-
ffect of pH on mineral mass loss by exposing mineralized
to pH adjusted brines at atmospheric pressure. Isolating
t of pH was necessary because during depressurisation
degassing of the HPV, extreme physical agitation of the
urred which may have accounted for mineral mass loss,
ould not be representative of processes that would occur
S field conditions. In order to estimate the pH of the scCO2
brine used in the current study, pH was measured in
turated brines in a similar high pressure flow system at
ipped with an in-line Barben Analyzer Technology pH
et al
and
char
used
in or
brin
to th
min
disso
expe
ogy
Infor
3. R
3.1.
pulse
T
ditio
the e
grad
tion
liber
(4))
5 da
dom
(Mit
efflu
5.2±
activ
NH4
nod
had
ysis
med
S
sphe
(Fig.
conc
of 14
avai
to 2
tion
of ∼
catin
ureonected to a Campbell Scientific CR1000 data logger. The
pH was 3 in a 0.5 g L−1 TDS brine and 4 in a 5g L−1 TDS
nsen, 2009), consistent with the decreased solubility of
igher salinity solutions (Duan and Sun, 2003; Lagneau
thetic po
the calciu
to 7.5MP
productioCoupons in pH 4 brine for 2h
Coupons in pH 4 brine for 48h
Sterile control
05). These observations agree with reports by (Kaszuba
ky, 2009) who report typical pH values of 3.5–4 for CO2
cidic brines. Therefore the pH of the 162g L−1 TDS brine
e current study was estimated to have a pH of >4. Hence,
o expose the coupons to the most acidic conditions likely,
s adjusted to pH 4 with HNO3, and coupons were exposed
ine for varying lengths of time. Large volume-to-carbonate
atios were used to minimize pH changes due to CaCO3
n and maintain the pH at 4. Minerals from each of the
nts were examined with XRD to determine the mineral-
e precipitates before and after exposure (Supplemental
ion, Section SI2).
ts
lysis and carbonate mineral formation at ∼7.5MPa
conditions
ow reactor was operated at atmospheric pulse-flow con-
ith calcium free medium (CMM−), for the first 17 days of
iment. During this period, NH4+ concentrations increased
from ∼3000mgL−1, which is the free NH4+ concentra-
e CMM− from NH4Cl, to ∼11,000mgL−1, indicating the
of NH4+ from the hydrolysis of urea (Eqs. (1), (2) and
3). The pH also increased from 7 to 9.1 within the first
ig. 3), consistent with buffering by NH4+ =NH3 +H+, the
t buffer in the system, which has a pKa value of 9.3 at 30 ◦C
and Ferris, 2005). Viable cells measured in the reactor
enerally increased during this period, up to amaximumof
×107 CFU, and 100% of the tested CFUs were ureolytically
g. 3). Influent media samples during this period exhibited
centrations of ∼3000mgL−1, a pH of ∼7, and there were
ableCFUs (Fig. 3), demonstrating that the influentmedium
een contaminated by ureolytic organisms, and that ureol-
occurring only in or downstream of the synthetic porous
re.
g onday18, pressurewas gradually increased fromatmo-
o 7.5MPa as the daily pulse-flow of CMM− continued
uring this period, effluent pH remained at ∼9.1. NH4+
ations gradually increased over this period to a maximum
±1229mgL−1, consistent with the hydrolysis of all the
urea, as did ureolytic CFUs, on average which increased
07 ±4.8×106 (Fig. 3). During the period of pressuriza-
MPa, influent media still exhibited NH4+ concentrations
mgL−1, pH was ∼7, and there were no CFUs, again indi-
at the influent medium had not been contaminated by
organisms, and that ureolysis was occurring in the syn-
rous media core (Fig. 3). Pressurization of the CMM− and
m inclusive media, CMM+, in the absence of S. pasteurii
a in the HPV for 24h (Fig. 2) did not result in any NH4+
n, change in pH or, in the case of the CMM+ experiments,
Fig. 3. Pressure, influent and effluent NH4+, pH, Ca2+ and ureolytic colony forming
units from the high pressure (∼7.5MPa) moderate temperature (32 ◦C) flow reac-
tor containing a synthetic porous media core. Ca free CaCO3 mineralizing medium
(CMM−) described by Ferris et al. (1996) was used as the influent media from time
0 to day 38, after which the CMM media included Ca (CMM+).
Table 2
Aqueous chemistry of CaCO3 mineralizing medium (CMM+) described by Ferris
et al. (1996) after exposure to abiotic high pressure conditions (8.9MPa) in the
high pressure vessel. Experiments repeated with CMM excluding calcium (CMM−)
demonstrating that urea hydrolysis is not induced abiotically by high pressure con-
ditions. Concentrations in mgL−1.
Sample pH NH4+ Na+ Mg2+ K+ Ca2+ Ba2+
CMM−+pressure 7.37 3.22 6.76 0.025 1.15 0.200 0.001
CMM− 7.15 3.45 6.60 0.025 1.12 0.18 0.001
CMM++p
CMM+
decrease
pressure
After
influent w
Ca2+. The
12±0.6m
tion for th
there we
tions, or u
and the g
(Fig. 4) an
cated tha
the core,
occurred
no precip
influent C
∼1000m
regionun
not in th
density g
in the up
ImageJ s
images fr
and maxi
tively, w
below th
2011).
3.2. Calci
CO2/brine
In ord
tated CaC
on polyca
tures at h
also atmo
7.61, at a
CaCO3 m
respectiv
eral mas
significan
not sugge
a statisti
timescale
mass cha
not mois
The expo
a signific
variation
this mas
the depre
lence in t
formation
surizationressure 7.31 2.95 6.48 0.025 1.11 9.71 0.001
7.25 2.35 6.49 0.025 1.06 9.20 0.001
in Ca2+ concentrations, which confirmed that increased
alone did not result in the hydrolysis of urea (Table 2).
the flow system had been pressurized to 7.5MPa, the
as switched to CMM+, containing 1000mgL−1 (25mM)
concentration of Ca2+ in the flow reactor effluentwas only
g L−1 after the first 24h, and remained at this concentra-
e remainder of the experiment (Fig. 3). During this period,
re no appreciable changes in effluent pH, NH4+ concentra-
reolytic CFUs. After 45days the systemwasdepressurised
lass capillaries inside the corewere retrieved. Stereoscope
d XRD analysis (Supplemental Information, Fig. SI2) indi-
t calcite had precipitated inside the glass capillaries of
demonstrating that ureolysis induced CaCO3 precipitation
under high pressure conditions ∼7.5MPa. Additionally,
itates were observed in the effluent accumulator, and the
MM+ media exhibited consistent Ca2+ concentrations of
g L−1, indicating that precipitation occurred in the core
der pressure or down theflowpath of the core (Fig. 3), and
e effluent accumulator during depressurisation. A crystal
radient could be observed, with a greater crystal density
flow (influent) end of the core. Crystal size analysis with
oftware (Mitchell and Ferris, 2006a,b) of 5 stereoscope
om throughout the core revealed the average, minimum
mum crystal size was 57m, 10m and 102m respec-
ith a standard deviation of 21m. This ignored crystals
e stereoscope detection limit of ∼5m (Schultz et al.,
um carbonate mineral stability with supercritical
mixtures
er to examine stability of the microbiologically precipi-
O3 during brine and scCO2 exposure, CaCO3 precipitates
rbonate coupons were challenged with scCO2-brine mix-
igh pressure (∼8.9MPa) conditions in the HPV (Fig. 5) and
spheric pressure brines. Incubation of CaCO3 in brine, pH
tmospheric pressure for 4h, 24h and 1 week, lead to a
ass change of −7±0.036%, −10%±4.7% and −23%±2.5%
ely for experiments 1–3. However the difference in min-
s from before and after exposure was not statistically
t for the three experiments (Fig. 5) and therefore did
st that brine exposure at atmospheric conditions led to
cally significant change in the mass of CaCO3 on these
s. Exposure to scCO2 for 48h did not lead to an observable
nge (−0.4±0.01% variation) demonstrating that dry (i.e.
ture containing) scCO2 has no effect on the mineral mass.
sure of CaCO3 to brine saturated with scCO2 resulted in
ant mass change of CaCO3 during both 2h (−84%, ±35%
) and 48h (−92%, ±50% variation) of exposure. However,
s loss is being attributed to an artifact associated with
ssurization process, which resulted in significant turbu-
he brine filled portion of the HPV. Highly intense bubble
was observed during out-gassing of CO2 during depres-
, which occurred at a rate of 1.158MPamin−1 for the 2h
Fig. 4. Images of CaCO3 precipitates formed inside the capillary of the synthetic porous media core in the high pressure (∼7.5MPa) moderate temperature (32 ◦C) flow
reactor. (a) CaCO3 precipitates (in white) inside one of the 5 cm long glass tubes (capillaries) that were inserted in the artificial core. Decreasing crystal density can be seen
in the direc aCO3
images of a coatin
no extensiv
exposure
ing turbu
believed
as eviden
tom sump
above th
with brin
2h of exp
Due to
ing depre
could hav
to brine a
pH of the
saturated
Duan and
2003, 20
dissolutio
a small m
over 48h
re an
ajor
anal
ures
ges
plem
iscu
Ureo
-flow
he u
e-flo
by Stion of flow (top is influent, bottom effluent). (b–e) Stereoscope images: (b) small C
cluster of CaCO3 crystals, (e) view into capillary from influent end showing CaCO3
e EPS rich biofilm was observed.
and 3.08MPamin−1 for the 48h exposure. The result-
lence, clearly visible through the sight glass (Video 1), is
to have mechanically abraded CaCO3 from the coupons,
ced by the discovery of some CaCO3 crystals in the bot-
of the HPV. CaCO3 minerals in the headspace of the HPV,
e scCO2 saturated brine (i.e. exposed to scCO2 saturated
e vapour) exhibited a mass change of −36±7.5% during
osure, and −17±3.6% during 48h of exposure.
the physical disturbance that the coupons undergo dur-
ssurisation and degassing of the brine, and the effect this
e upon the mineral mass loss, coupons were also exposed
t atmospheric pressure, acidified to pH 4, the estimated
1.5M scCO2 saturated brine at 8.9MPa used in the scCO2
brine experiments (Allen et al., 2005; Carey et al., 2010;
Sun, 2003; Kaszuba and Janecky, 2009; Kaszuba et al.,
befo
no m
XRD
mixt
chan
(Sup
4. D
4.1.
pulse
T
puls
ysis05; Lagneau et al., 2005), in order to determine mineral
n associated with low pH conditions alone. This revealed
ass change of only −3.4±0.67% over 2h and −7.8±2.2%
(Fig. 5). Stereoscope images of the CaCO3 precipitates
under pr
age envir
studies o
organismcrystals inside the 1mm internal diameter capillary, (c and d) close-up
g on the inside wall of the capillary. While crystals were clearly visible,
d after exposure to the scCO2-brine mixtures exhibited
morphological changes on the 10–100m scale (Fig. 6).
ysis of the precipitates before and after the scCO2/brine
revealed only calcite, demonstrating that there were no
in mineralogy due to interaction with scCO2 and brine
ental Information, Section SI2, Fig. SI2).
ssion
lysis and carbonate mineral formation at ∼7.5MPa
conditions
se of a unique high pressure, moderate temperature
w system has demonstrated that bacterial urea hydrol-
. pasteurii and associated CaCO3 precipitation can occur
essure conditions that approximate subsurface CO2 stor-
onments. These observations are consistentwith previous
f S. pasteurii at atmospheric pressure, demonstrating the
’s ureolytic capability (Ferris et al., 2003; Fujita et al.,
Fig. 5. Change in mass of CaCO3 on coupons (normalized to initial mineral mass)
before and after challenges in the high pressure vessel (HPV) (see Table 1) with
percentage change of mass listed. Three or more coupons used in each experiment.
2000; W
reported
of mesop
replicatio
1993; Ya
to 5.2±1
pasteurii
7.5MPa,
inhibit ce
Ureoly
during th
when N
14,500±
the hydro
suggests
experime
7.5MPa, u
than ava
the synth
while bio
CFUs in th
olytic bio
days, sinc
by the av
The h
the begin
ureolytic
the expe
associate
system. P
strated th
to porous
Van Paas
2011). In
surface-a
Fig. 6. Stereoscope images of coupons exposedbrine and scCO2 before (a and c) andafter (b andd) cha
No major morphological changes to CaCO3 crystals noticeable on this scale.arren et al., 2001). High pressures have previously been
to significantly affect the metabolic functions and growth
hilic organisms, such as S. pasteurii, by inhibiting DNA
n and protein synthesis (Abe et al., 1999; Gross et al.,
yanos and Pollard, 1969). However, ureolytic CFUs up
.1×107 CFU/mL in the reactor effluent indicate that S.
maintained viability under conditions of approximately
demonstrating that elevated pressures did not fatally
ll function.
tic CFUs in the reactor effluent were, on average, higher
e period of pressurization at 7.5MPa (day 18 onwards),
H4+ concentrations were also at their maximum of
1200mgL−1, a concentration which was consistent with
lysis of all the available urea in the influent media. This
that ureolytic biomass increased during the course of the
ntand indicates thatduring theperiodofpressurizationat
reolysis was limited by themass transport of urea, rather
ilable ureolytic biomass, under pulse-flow conditions in
eticporousmedia core. In contrast, prior topressurization,
masswas increasing in the core, as indicated by increasing
is period, ureolysiswas likely limitedby theavailableure-
mass rather than reactant supply until approximately 17
e NH4+ concentrations were less than could be produced
ailable urea, at only ∼11,000mgL−1.
igh pressure flow system was only inoculated once at
ning of the experiment, yet significant numbers of viable
cells were observed in the effluent over the course of
riment. Therefore, it appears a well established surface-
d community, or biofilm, developed in the high pressure
revious studies at atmospheric pressures have demon-
at S. pasteurii can be effectively distributed and attached
media surfaces, including sand beds (Harkes et al., 2010;
sen et al., 2010b), and glass capillaries (Schultz et al.,
these studies, media flow was induced that allowed a
ssociated community to develop after initial inoculation.
llenge in thehighpressure vessel (HPV). Images are at 40×magnification.
Schultz e
microsco
extensive
surface a
in this st
able und
pH and N
demonstr
in the sy
Biofilms
logical st
organism
2007; Ste
us that pl
at 136 at
viable ce
reduction
suggest th
in the hig
tion unde
form.
The fl
cipitation
be expect
sure by u
carbonate
tates out
is exceed
in the hi
the depre
pH under
CO2 was
present a
effluent w
in solutio
pressure)
of CaCO3
Morgan, 1
surization
gas and o
lack of C
CaCO3 pr
mulator,
carbonate
ubility of
(Pytkowi
the ion a
at atmos
strating t
would be
atmosphe
QCi (Park
standard
tion is rea
urea has
of 7.5MP
ysed befo
be a min
precipita
interfacia
Morgan, 1
2D hetero
at a mini
homogen
at SIs of 1
sure flow
com
roge
bact
hell
e ca
ose c
hanis
xper
lysis
satu
(Mit
ing 2
cell
llary
crys
fast
on ra
er Ca
Ferri
rage
3 pr
arily
ope
433m
s bat
chell
Nev
exp
ted s
ic co
spa
mat
ging
s, as
mosp
; Sch
Calc
brine
he si
e cha
s los
tals f
ation
dega
mal
, the
2h,
onst
. CaC
as be
ns (
., 201
ns u
ly du
ure
ted a
prod
ceb
lutio
ed prt al. (2011) demonstrated using confocal laser scanning
py, that the surface associated S. pasteurii did not form
EPS, but rather a discrete cell coverage on the capillary
nd calcite surfaces, which supports observations made
udy where only CaCO3 crystals but no EPS was observ-
er high pressure pulse-flow conditions (Fig. 4). The low
H4+ concentration in the influent media accumulator
ates the biofilm community must have been present only
nthetic porous media core or downstream of the core.
are often more resilient to physical chemical, and bio-
resses than free floating, or planktonic cells, of the same
(CostertonandStewart, 2001; Lewandowski andBeyenal,
wart, 2003). Indeed, it was previously demonstrated by
anktonic B. mojavensis cultures exposed to flowing scCO2
m and 35 ◦C for 19min lead to a 1000-fold reduction in
ll numbers, while biofilm cultures only showed a 10-fold
in viable cell numbers (Mitchell et al., 2008). These data
at the biofilmstate of S. pasteurii in the porousmedia core
h pressure system may have promoted growth and func-
r high pressure conditions compared to the planktonic
ow system demonstrates ureolysis-induced CaCO3 pre-
can occur under high pressure conditions. This would
ed under the high pH conditions maintained under pres-
reolysis, which shifts the bicarbonate equilibrium to form
ions (CO32−) which, in the presence of Ca2+, precipi-
of solution as CaCO3 if the solubility product for calcite
ed (Burne and Chen, 2000; Castanier et al., 1999). Indeed,
gh pressure flow reactor experiments, pH measured in
ssurized effluent fluids (Fig. 3) will represent the in situ
high pressure conditions because nitrogen gas and not
used for pressurization. If a separate CO2 phase had been
t elevated pressures, degassing of CO2 from the reactor
ould have resulted in a shift of the carbonate speciation
n, leading to an increase in pH relative to the in situ (high
conditions. This could have resulted in the precipitation
during depressurization (Murray et al., 1980; Stumm and
996). However, the precipitation of CaCO3 during depres-
is unlikely to have occurred sinceN2 was the overburden
utgassing of CO2 was likely not a significant factor. The
aCO3 precipitates in the effluent accumulator confirms
ecipitation occurred in the core region before the accu-
and must have been caused by ureolysis induced pH and
ion concentration increases. Pressure increases the sol-
CaCO3, and thus the ion activity product at saturation
cz and Conners, 1964). In artificial seawater at 10MPa,
ctivity product at saturation is 1.093 times greater than
pheric pressure (Pytkowicz and Conners, 1964), demon-
hat under fixed Ca concentrations, slightly more urea
hydrolysed before CaCO3 saturation is reached than at
ric conditions. Saturation index (SI) modeling on PHREE-
hust and Appelo, 1999) with analysis of the media under
conditions (1 atm and 25 ◦C) demonstrates that satura-
ched [SI = log ({Ca2+}{CO3−}/Ksp) = 0]when only 1.05mM
been hydrolysed. This suggests at the maximum pressure
a used the experiments, 1.15mM of urea was hydrol-
re CaCO3 precipitation was induced. However, this will
imum value, as supersaturation is often required before
tion is initiated because the nucleation activation (i.e.,
l) free energy barrier has been surmounted (Stumm and
996). Indeed, in abiotic CaCO3 precipitation experiments,
geneous nucleation on seed particles has been observed
by a
hete
and
Mitc
to th
of lo
mec
E
ureo
thus
cells
gest
near
capi
that
grew
itati
high
and
(ave
CaCO
prim
flow
and
viou
(Mit
SI3).
flow
limi
thet
void
Infor
plug
tant
in at
2011
4.2.
CO2/
T
brin
mas
crys
suris
was
the s
pH 4
over
dem
gible
pH h
actio
et al
ditio
first
mixt
limi
and
surfa
disso
trollmum SI of 0.78 (Teng et al., 2000), and spontaneous 3D
eous nucleation in unseeded solutions has been observed
.83–1.99 (Gómez-Morales et al., 1996). In the high pres-
system, calcite precipitation would be likely be initiated
are so rap
function
meric sub
transportbination of 3D homogenous nucleation in solution and 2D
neous nucleation on the capillary walls, nascent crystals,
erial cell surfaces (Ferris et al., 2003; Fujita et al., 2000;
and Ferris, 2005, 2006b). Crystals were observed attached
pillary walls (Fig. 4b, e and f), as well as agglomerates
rystals forming plugs (Fig. 4c and d) suggesting multiple
ms of nucleation and growth.
imental and modeling studies demonstrate that during
, the highest pH, concentrations of carbonate ions, and
ration, will occur proximal to, and decrease away from
chell and Ferris, 2006b; Zhang and Klapper, 2010), sug-
D nucleation in the flow system will preferentially occur
s. However, the decrease in crystal size densities in the
along the flow direction of the core (Fig. 4a) suggests
tal growth was controlled by reactant supply and nuclei
er in the inlet region, where reactants, and thus precip-
tes are higher due to the increased SI and availability of
2+ and CO32− concentrations (Ferris et al., 2003; Mitchell
s, 2005; Teng et al., 2000). Crystal sizes were much larger
=57m) than previously observed for ureolysis induced
ecipitation (∼1–6m) (Mitchell and Ferris, 2006a,b). This
reflects a higher total reactant supply provided by pulse-
ration of the high pressure experiments (20.16mg Ca2+
g CO32− over the 9 days of CMM+), compared to pre-
ch experiments [5.25–17.5mg Ca and 79–264mg CO32−
and Ferris, 2006a,b)] (Supplemental Information, Section
ertheless at the termination of the high pressure pulse-
eriment, plugging had not occurred due to the relatively
upply of reactants compared to the pore space in the syn-
re, calculated as 0.0185 cm3 CaCO3 relative to 0.196 cm3
ce in the synthetic porous media core (Supplemental
ion, Section SI4). However, permeability reduction and
would have occurred given the supply of sufficient reac-
demonstrated for ureolysis-induced CaCO3 precipitation
heric pressure constant-flowsystems (Cunninghamet al.,
ultz et al., 2011).
ium carbonate mineral stability with supercritical
mixtures
gnificant mass change observed in the scCO2 saturated
llengesover2h (−84±35%mass loss) and48h (−92±50%
s) appeared to reflect physical dislodging of the CaCO3
rom the coupon surface, particularly during the depres-
of the HPV which caused violent bubbling when CO2
ssing from the brine (Video 1). This was supported by
l mass change of CaCO3 in atmospheric pressure brine at
estimated pH of the scCO2 saturated brine (−3.4±0.67%
−7.8±2.2% over 48h). This minor mass change of CaCO3
rated that acidity-driven dissolution of CaCO3 was negli-
O3 dissolution in the scCO2 saturated brine driven by low
en observed in other studies of acidic brine–CaCO3 inter-
Carey et al., 2010; Gledhill and Morse, 2004, 2006; Sanz
1) and would be expected under the experimental con-
sed here. The apparent lack of this phenomenon could be
e to mass transport limitations in the static brine–CaCO3
in the HPV, where reaction kinetics are reduced by the
bility of reactants (e.g. H+) to reach the mineral surface
ucts (Ca2+ and CO32−) to migrate away from the mineral
y advection. Indeed, at lowpHvalues, thekinetics of calcite
n are controlled by mass transport because surface con-
ocesses (addition of ions to or loss from mineral surface)
id (Marini, 2006). Secondly, thismaybedue toaprotective
of the biofilm where the presence of extracellular poly-
stances protects the CaCO3 from dissolving. While mass
limitations might be reduced in flowing systems due to
enhanced
2004), th
reduce th
the disso
sion limit
described
The close
cipitated
(Schultz e
brines,m
stable in
urated br
specifical
also have
CaCO3 m
impact th
pore plug
and pH a
more rea
tic bedro
carbonate
Kaszuba
microbia
5. Concl
Exper
ate temp
porous m
tion to b
subsurfac
pasteurii
media co
the daily
from 7 to
has a pKa
CaCO3 pr
dient wit
limited b
precipita
investiga
ditions th
condition
loss, due
effect on
cant CaCO
brine (pH
ibrated b
scCO2 eq
tation of
at the ter
acidic dis
to the pr
of the br
limited ab
mineral s
induced m
ability of
GCS are a
brine reg
storage s
Newa
leakage, a
environm
be applie
aditi
e w
app
coul
larly
ous
tion
ce w
tion
3 pr
ial di
t of i
tion
proc
thro
caled
ation
san
sing
ns.
owl
undi
) Ze
DE-F
465
Mar
nvir
IP, C
gy In
is a
endi
uppl
d, i
.201
renc
., Kato
ms. Tr
, D.E.,
quest
uel Pro
eld, J.F
inera
eathe
6, 340
ey, W
ood, E
meric
, R.A.
nd Inf
, W., S
ellbor
terna
nier, S
on an
eology
rton, J
lonie
know
4–81.
inghamadvection of reactants and products (Wang and Jaffe,
e possible protective function of the biofilm would still
edegreeofCaCO3 dissolution. Inbiofilm-affectedsystems,
lution might be slowed down significantly due to diffu-
ations (Stewart, 2003) and mineral dissolution has been
to be affected by polysaccharides (Banfield et al., 1999).
association of S. pasteurii biofilms and microbially pre-
CaCO3 minerals has been demonstrated by us previously
t al., 2011). These data demonstrate that in high pressure
icrobially enhanced CaCO3 precipitateswill likely bemost
the long term in the scCO2 region and in the scCO2 sat-
ine region where advection is limited. GCS site geology,
ly the abundance of silisiclastic or carbonate bedrock will
a major control on the in situ pH and the saturation of
inerals both before and after scCO2 injection which may
e effectiveness of microbially induced mineralization and
ging. Specifically, pH and CaCO3 saturation will be higher
nd oversaturation will be maintained in the long-term
dily in pore spaces in carbonate bedrock than silisiclas-
ck sites, from the dissolution and buffering of the native
minerals (Carey et al., 2010; Kaszuba and Janecky, 2009;
et al., 2003, 2005). This will enhance the stability of the
lly produced mineral plugs.
usions
iments using a novel high pressure (∼7.5MPa) moder-
erature (32 ◦C) pulse-flow reactor containing a synthetic
edia core allowed microbiologically induced mineraliza-
e investigated under conditions approximating those in
e carbon storage environments. A surface-associated S.
biofilm became well established in the synthetic porous
re, which was able to hydrolyse all of the urea available in
media pulses at pressures of ∼7.5MPa. This raised the pH
9.1, consistent with buffering by NH4+ =NH3 +H+ which
value of 9.3 at 30 ◦C and induced CaCO3 precipitation.
ecipitated as calcite, and a decreasing crystal density gra-
h distance into the core suggested crystal growth was
y the mass transport of Ca2+ and urea. Challenges of
ted CaCO3 with brine/scCO2 static mixtures for 2–48h
ted the potential stability of the precipitates under con-
at approximated a range of pre- and post-CO2 injection
s. Brine at atmospheric conditions generated a slightmass
to dissolution of the mineral phase. scCO2 had a negligible
mass, but scCO2 equilibrated brine generated a signifi-
3 mass loss. However, control experimentswith acidified
4, the estimated minimum in situ pH of the scCO2 equil-
rine) demonstrated that much of the mass loss in the
uilibrated brine experiments was due to the physical agi-
the CaCO3 crystals during depressurisation and degassing
mination of the experiments. Mass loss was not due to
solution of the CaCO3, which appeared to be limited due
otective effect of the biofilm and the static incubation
ine+CaCO3 where reaction kinetics are reduced by the
ility of reactants and products to migrate to and from the
urface by advection. These data suggest that microbially
ineralization with the purpose of reducing the perme-
preferential leakage pathways during the operation of
pplicable in the scCO2 region andboth the scCO2 saturated
ion and the brine saturated scCO2 region of the subsurface
ite where subsurface fluids are quasi static.
to tr
to th
new
face
simi
tane
injec
surfa
injec
CaCO
spat
poin
injec
tion
pore
up-s
cipit
thick
focu
ditio
Ackn
F
(DOE
No.
08ER
(EU)
the E
DUR
Biolo
8807
App
S
foun
ijggc
Refe
Abe, F
is
Allen
se
F
Banfi
m
w
9
Barkl
W
A
Burne
a
Carey
w
In
Casta
ti
G
Coste
co
is
7
Cunn
nd novel tools to control subsurface permeability and CO2
s presented here, are required in inaccessible subsurface
ents, where ‘traditional’ engineering approaches cannot
d. For example, while injection wells may be accessible
for the
diation
Cunningham
geologi
3245–3onal sealing approaches, reducing permeability proximal
ell bore or in fracture zones far from the well requires
roaches. While the injection of a base into the subsur-
d achieve a pH increase and induce CaCO3 precipitation
to the ureolysis based approach, thiswould lead to instan-
CaCO3 supersaturation and precipitation at the point of
. However, the controlled injection of urea into the sub-
ould allow for transport of urea farther away from the
point before urea hydrolysis induced pH increase and
ecipitation would occur. This is likely to promote a wider
stributionof CaCO3 andavoiduncontrolledplugging at the
njection (Ferris et al., 2003;Mitchell and Ferris, 2005). The
of ureolytic organisms could also enhance the precipita-
ess, particularly since bacterial cells can migrate through
at sizes in common GCS site rocks. Recent work has been
by investigating radial flow and ureolytic CaCO3 pre-
at atmospheric pressures, within 74 cm diameter, 38 cm
dstone cores (Phillips et al., 2012) and ongoing work is
on up-scaling under pressures that approximate GCS con-
edgments
ng was provided from the US Department of Energy
ro Emissions Research Technology Center (ZERT), Award
C26-04NT42262 and DOE EPSCoR Award No. DE-FG02-
27, NSF Award No. 0934696, as well as a European Union
ie Curie Reintegration Grant to ACM (277005). Support for
onmental and BiofilmMass Spectrometry Facility through
ontract Number: W911NF0510255 and the MSU Thermal
stitute from theNASAExobiology ProgramProjectNAG5-
cknowledged.
x A. Supplementary data
ementary data associated with this article can be
n the online version, at http://dx.doi.org/10.1016/j.
3.02.001.
es
, C., Horikoshi, K., 1999. Pressure-regulated metabolism in microorgan-
ends in Microbiology 7, 447–453.
Strazisar, B.R., Soong, Y., Hedges, S.W., 2005. Modeling carbon dioxide
ration in saline aquifers: significance of elevatedpressures and salinities.
cessing Technology 86, 1569–1580.
., Barker, W.W., Welch, S.A., Taunton, A., 1999. Biological impact on
l dissolution: application of the lichen model to understanding mineral
ring in the rhizosphere. Proceedings of theNational Academyof Sciences
4–3411.
., Richardson, J., 1994. Laboratory safety. In: Gerhardt, P., Murray, R.,
.,Willis, A., Krieg, N. (Eds.), Manual ofMethods for General Bacteriology.
an Society of Microbiology, Washington, DC, pp. 715–734.
, Chen, Y.-Y.M., 2000. Bacterial ureases in infectious diseases. Microbes
ection 2, 533–542.
vec, R., Grigg, R., Zhang, J., Crow, W., 2010. Experimental investigation of
e integrity and CO2-brine flow along the casing-cement microannulus.
tional Journal of Greenhouse Gas Control 4, 272–282.
., Le Métayer-Levrel, G., Perthuisot, J.-P., 1999. Ca-carbonates precipita-
d limestone genesis – the microbiogeologist point of view. Sedimentary
126, 9–23.
.W., Stewart, P.S., 2001. Battling biofilms – the war is against bacterial
s that cause some of the most tenacious infections known. The weapon
ledge of the enemy’s communication system. Scientific American 285,
, A., Sharp, R., Hiebert, R., James, G., 2003. Subsurface biofilm barriers
containment and remediation of contaminated groundwater. Bioreme-
Journal 7, 151–164.
, A.B.,Gerlach,R., Spangler, L.,Mitchell, A.C., 2009.Microbially enhanced
c containment of sequestered supercritical CO2. Energy Procedia 1,
252.
Cunningham
Reducin
barriers
Duan, Z., Su
and aqu
Geology
Dupraz, S.,
modelin
aquifer
Ferris, F.G.,
pluggin
Ferris, F.G.,
induced
Geochim
Fujita, Y., Fe
ate prec
305–31
Fujita, Y., Ta
L., Smit
ter to e
3025–3
Gerlach, R.,
dynami
and Bio
Gledhill, D.K
brines a
Gledhill, D.K
Geochim
Gómez-Mo
calcium
169, 33
Gross, M., L
tion of
and ide
Biochem
Hammes, F
remova
Hansen, L.,
ature fl
Thesis,
Harkes, M.P
2010. F
precipit
Jenneman,
nutrien
383–39
Kaszuba, J.
dioxide
Sequest
Geophy
Kaszuba, J.P
a mode
sequest
Kaszuba, J.P
reaction
integrit
Lagneau, V.
du com
aquifèr
Lewandows
London
Marini, L., 2
ments i
Mitchell, J.,
cal engi
1222–1
Mitchell, A.
induced
kinetic
Mitchell, A
and sol
Environ
Mitchell, A.
and gro
ell, A
unnin
ritical
ell, A
009. B
ationa
ell, A
010.
appin
270–5
ay, J.W
f pres
eochim
n, P.H
APG B
, J.-P.,
om ge
akage
., Olde
ioxide
ust, D
rogram
eoche
ons Re
ps, A.J
ingham
iofilm
echno
wicz,
seaw
E.,Ayo
odeli
tz, L.,
iologic
oday 2
tze-La
rial su
art, P.S
m, W.
040 p.
H.H.,
esses a
cta 64
ith, Y
lizatio
dolom
aasse
recht,
cologi
aasse
010b.
iogrou
g 136
, S., Ja
CO2
onver
en, L.A
ium ca
nd for
8, 93–
fin, V
ipitati
17–42
e, C.M
tion a
eologi
aste M
os, A.
olecu, A.B., Gerlach, R., Spangler, L., Mitchell, A.C., Parks, S., Phillips, A., 2011.
g the riskofwell bore leakageofCO2 usingengineeredbiomineralization
. Energy Procedia 4, 5178–5185.
n, R., 2003. An improved model calculating CO2 solubility in pure water
eous NaCl solutions from 273 to 533K and from 0 to 2000bar. Chemical
193, 257–271.
Parmentier, M., Ménez, B., Guyot, F., 2009. Experimental and numerical
g of bacterially induced pH increase and calcite precipitation in saline
s. Chemical Geology 265, 44–53.
Stehmeier, L.G., Kantzas, A., Mourits, F.M., 1996. Bacteriogenic mineral
g. Journal of Canadian Petroleum Technology 35, 56–61.
Phoenix, V., Fujita, Y., Smith, R.W., 2003. Kinetics of calcite precipitation
by ureolytic bacteria at 10 to 20 degrees C in artificial groundwater.
ica Et Cosmochimica Acta 68, 1701–1710.
rris, F.G., Lawson, R.D., Colwell, F.S., Smith, R.W., 2000. Calcium carbon-
ipitation by ureolytic subsurface bacteria. Geomicrobiology Journal 17,
8.
ylor, J.L., Gresham, T.L., Delwiche, M.E., Colwell, F.S., McLing, T., Petzke,
h, R.W., 2008. Stimulation of microbial urea hydrolysis in groundwa-
nhance calcite precipitation. Environmental Science & Technology 42,
032.
Cunningham, A.B., 2010. Influence of biofilms on porous media hydro-
cs. In: Vafai, K. (Ed.), Porous Media: Applications in Biological Systems
technology. Taylor Francis, Boca Raton London New York, pp. 173–230.
., Morse, J.W., 2004. Dissolution kinetics of calcite in NaCl–CaCl2–MgCl2
t 25 degrees C and 1bar pCO2. Aquatic Geochemistry 10, 171–190.
., Morse, J.W., 2006. Calcite dissolution kinetics in Na–Ca–Mg–Cl brines.
ica Et Cosmochimica Acta 70, 5802–5813.
rales, J., Torrent-Burgués, J., Rodríguez-Clemente, R., 1996. Nucleation of
carbonate at different initial pH conditions. Journal of Crystal Growth
1–338.
ehle, K., Jaenicke, R., Nierhaus, K.H., 1993. Pressure-induced dissocia-
ribosomes and elongation cycle intermediates – stabilizing conditions
ntification of the most sensitive functional-state. European Journal of
istry 218, 463–468.
., Seka, A., de Kniff, S., Verstraete, W., 2003. A novel approach to calcium
l from calcium-rich industrial wastewater.Water Research 37, 699–704.
2009. Design and experimental testing of a high pressure, high temper-
ow-through rock core reactor using supercritical carbon dioxide. M.S.
Montana State University.
., van Paassen, L.A., Booster, J.L., Whiffin, V.S., van Loosdrecht, M.C.M.,
ixation and distribution of bacterial activity in sand to induce carbonate
ation for ground reinforcement. Ecological Engineering 36, 112–117.
G.E., McInerney, M.J., Knapp, R.M., 1985. Microbial penetration through
t-saturated berea sandstone. Applied and EnvironmentMicrobiology 50,
1.
P., Janecky, D.R., 2009. Geochemical impacts of sequestering carbon
in brine aquifers. In: McPherson, J.M., Sundquist, E. (Eds.), Carbon
ration and its Role in the Global Carbon Cycle, Volume 183. American
sical Union Monograph, Washington, DC, pp. 239–247.
., Janecky, D.R., Snow, M.G., 2003. Carbon dioxide reaction processes in
l brine aquifer at 200 degrees C and 200 bars: implications for geologic
ration of carbon. Applied Geochemistry 18, 1065–1080.
., Janecky, D.R., Snow,M.G., 2005. Experimental evaluation ofmixedfluid
s between supercritical carbon dioxide and NaCl brine: relevance to the
y of a geologic carbon repository. Chemical Geology 217, 277–293.
, Pipart, A., Catalette, H., 2005. Modélisation couplée chimie-transport
portement à long terme de la séquestration géologique de CO2 dans des
es salins profonds. Oil &Gas Science andTechnology–ReIFP 60, 231–247.
ki, Z., Beyenal, H., 2007. Fundamentals of Biofilm Research. CRC Press,
, p. 452.
006. The kinetics of mineral carbonation. In: Marini, L. (Ed.), Develop-
n Geochemistry, Vol. 11. Elsevier, Chapter 6, pp. 169–317.
Carlos Santamarina, J., 2004. Biological considerations in geotechni-
neering. Journal of Geotechnical & Geoenvironmental Engineering 131,
233.
C., Ferris, F.G., 2005. The co-precipitation of sr into calcite precipitates
by bacterial ureolysis in artificial groundwater – temperature and
dependence. Geochimica Et Cosmochimica Acta 69, 4199–4210.
Mitch
C
c
Mitch
2
n
Mitch
2
tr
5
Murr
o
G
Nelso
A
Nicot
fr
le
Pan, L
d
Parkh
p
g
ti
Philli
n
b
T
Pytko
in
Sanz,
m
Schul
b
T
Schul
te
Stew
Stum
1
Teng,
c
A
Van L
si
in
Van P
d
E
Van P
2
b
in
Wang
to
C
Warr
c
a
1
Whif
c
4
Whit
ra
g
W
Yayan
m.C., Ferris, F.G., 2006a. Effect of strontium contaminants upon the size
ubility of calcite crystals precipitated by the bacterial hydrolysis of urea.
mental Science & Technology 40, 1008–1014.
C., Ferris, F.G., 2006b. The influence of Bacillus pasteurii on the nucleation
wth of calcium carbonate. Geomicrobiology Journal 23, 213–226.
Zakkour, P.
logical
Journal
Zhang, T., K
tation..C., Phillips, A.J., Hamilton, M.A., Gerlach, R., Hollis, W.K., Kaszuba, J.P.,
gham, A.B., 2008. Resilience of planktonic and biofilm cultures to super-
CO2. Journal of Supercritical Fluids 47, 318–325.
.C., Phillips, A., Hiebert, R., Gerlach, R., Spangler, L., Cunningham, A.B.,
iofilmenhanced subsurface sequestration of supercritical CO2. The Inter-
l Journal on Greenhouse Gas Control 3, 90–99.
.C., Dideriksen, K., Spangler, L.H., Cunningham, A.B., Gerlach, R.,
Microbially enhanced carbon capture and storage by mineral-
g and solubility-trapping. Environmental Science & Technology 44,
276.
., Emerson, S., Jahnke, R., 1980. Carbonate saturation and the effect
sure on the alkalinity of interstitial waters from the Guatemala Basin.
ica Et Cosmochimica Acta 44, 963–972.
., 2009. Pore-throat sizes in sandstones, tight sandstones, and shales.
ulletin 93, 329–340.
Oldenburg, C.M., Bryant, S.L., Hovorka, S.D., 2009. Pressure perturbations
ologic carbon sequestration: area-of-review boundaries and borehole
driving forces. Energy Procedia 1, 47–54.
nburg, C.M., Wu, Y.-S., Pruess, K., 2009. Wellbore flow model for carbon
and brine. Energy Procedia 1, 71–78.
., Appelo, C., 1999. User’s guide to PHREEQC (Version 2) – a computer
for speciation, batch-reaction, one-dimensional transport, and inverse
mical calculations. U.S. Geological Survey, Water-Resources Investiga-
port 99-4259, p. 312.
., Lauchnor, E., Eldring, J., Esposito, R., Mitchell, A.C., Gerlach, R., Cun-
, A.B., Spangler, L.H., 2012. Potential CO2 leakage reduction through
-induced calcium carbonate precipitation. Environmental Science &
logy 47, 142–149.
R.M., Conners, D.N., 1964. High pressure solubility of calcium carbonate
ater. Science 144, 840–841.
ra, C., Carrera, J., 2011.Calcitedissolutionbymixingwaters: geochemical
ng and flow-through experiments. Geologica Acta 9, 67–77.
Pitts, B., Mitchell, A.C., Cunningham, A.B., Gerlach, R., 2011. Imaging
ally-induced mineralization in fully hydrated flow systems. Microscopy
011 (September), 12–15.
m, S., Fortin, D., Davis, B.S., Beveridge, T.J., 1996. Mineralization of bac-
rfaces. Chemical Geology 132, 171–181.
., 2003. Diffusion in biofilms. Journal of Bacteriology 185, 1485–1491.
, Morgan, J.J., 1996. Aquatic Chemistry. John Wiley & Sons, New York,
Dove, P.M., De Yoreo, J.J., 2000. Kinetics of calcite growth: surface pro-
nd relationships tomacroscopic rate laws. Geochimica Et Cosmochimica
, 2255–2266.
., Warthmann, R., Vasconcelos, C., McKenzie, J.A., 2003. Microbial fos-
n in carbonate sediments: a result of the bacterial surface involvement
ite precipitation. Sedimentology 50, 237–245.
n, L.A., Daza, C.M., Staal, M., Sorokin, D.Y., van der Zon, W., van Loos-
M.C.M., 2010a. Potential soil reinforcement by biological denitrification.
cal Engineering 36, 168–175.
n, L.A., Ghose, R., Van der Linden, T., Van der Star, W., Van Loosdrecht, M.,
Quantifying biomediated ground improvement by ureolysis: large-scale
t experiment. Journal of Geotechnical and Geoenvironmental Engineer-
, 1721–1728.
ffe, P.R., 2004. Dissolution of a mineral phase in potable aquifers due
releases from deep formations; effect of dissolution kinetics. Energy
sion and Management 45, 2833–2848.
., Maurice, P.A., Parmar, N., Ferris, F.G., 2001. Microbially mediated cal-
rbonate precipitation: implications for interpreting calcite precipitation
solid-phase capture of inorganic contaminants. Geomicrobiology Journal
115.
.S., van Paassen, L.A., Harkes, M.P., 2007. Microbial carbonate pre-
on as a soil improvement technique. Geomicrobiology Journal 24,
3.
., Strazisar, B.R., Granite, E.J., Hoffman, J.S., Pennline, H.W., 2003. Sepa-
nd capture of CO2 from large stationary sources and sequestration in
cal formations – coalbeds and deep saline aquifers. Journal of the Air &
anagement Association 53, 645–715.
A., Pollard, E.C., 1969. A study of effects of hydrostatic pressure onmacro-
lar synthesis in Escherichia coli. Biophysical Journal 9, 1464.
, Haines, M., 2007. Permitting issues for CO2 capture, transport and geo-
storage: a review of Europe USA, Canada and Australia. International
of Greenhouse Gas Control 1, 94–100.
lapper, I., 2010. Mathematical model of biofilm induced calcite precipi-
Water Science & Technology 61, 2957–2964.