Made available through Montana State University’s ScholarWorks Ureolysis-induced calcium carbonate precipitation (UICP) in the presence of CO2- affected brine: a field demonstration Catherine M. Kirkland, Arda Akyel, Randy Hiebert, Jay McCloskey, Jim Kirksey, Alfred B. Cunningham, Robin Gerlach, Lee Spangler, Adrienne J. Phillips © This manuscript version is made available under the CC-BY-NC-ND 4.0 license https:// creativecommons.org/licenses/by-nc-nd/4.0/ 1 Ureolysis-induced calcium carbonate precipitation (UICP) in the presence of 2 CO2-affected brine: a field demonstration 3 4 Catherine M. Kirkland1,2*, Arda Akyel1,3, Randy Hiebert4, Jay McCloskey4, Jim Kirksey5, Alfred B. 5 Cunningham1,2, Robin Gerlach1,3, Lee Spangler6, and Adrienne J. Phillips1,2 6 7 1Center for Biofilm Engineering, Montana State University, 366 Barnard Hall, Bozeman, MT 8 59717 9 2Department of Civil Engineering, Montana State University, 205 Cobleigh Hall, Bozeman, MT 10 59717 11 3Department of Chemical and Biological Engineering, Montana State University, 214 Roberts 12 Hall, Bozeman, MT 59717 13 4Montana Emergent Technologies, 160 W Granite St, Butte, MT 59701 14 5Loudon Technical Services LLC, 1611 Loudon Heights Rd, Charleston, WV, 25314 15 6Energy Research Institute, Montana State University, P.O. Box 172465, Bozeman, MT 59717 16 17 Highlights 18 • UICP in the presence of CO2-affected brine reduced injectivity in a wellbore cement 19 channel by an order of magnitude. 20 • Ultrasonic imaging logs conducted after UICP treatment showed additional solids behind 21 the casing which extended at least 30 m (100 ft) above the injection zone. 22 • Urease enzyme from heat-treated microbial cultures was an effective catalyst for UICP 23 reactions. 24 25 Graphical Abstract 26 27 28 1 29 Abstract 30 Biomineralization is an emerging biotechnology for subsurface engineering applications like 31 remediating leaky wellbores. The process relies on ureolysis to induce precipitation of calcium 32 carbonate in undesired flow paths. In geologic storage of CO2, there is a potential for leakage 33 and low pH conditions, thus, ureolysis-induced calcium carbonate precipitation (UICP) was 34 tested at field scale to seal a channel in the wellbore cement annulus in the presence of CO2- 35 affected brine. Conventional oil field methods were used to deliver UICP-promoting fluids 36 downhole to the treatment zone approximately 1000 feet (305 m) below ground surface (bgs). 37 Over 4 days, 242 L (64 gal) of heat-treated Sporosarcina pasteurii cultures (22 bailers) and 329 38 L (87 gal) of urea – calcium chloride solution (30 bailers) were injected. The UICP treatment 39 resulted in a 94% reduction of injectivity and ultrasonic well logging showed a noticeable 40 increase in the percentage of solids in the channel outside the casing, including more than 30 m 41 (100 ft) above the injection point. Subsequent well logging 11 months after the field 42 demonstration showed that a significant portion of the new solids remained but the seal was 43 compromised following sustained pumping. The results of this experiment suggest that UICP 44 can be promoted in the presence of CO2-affected brine to seal leakage pathways. Additional 45 research is required to optimize long term seal integrity to ensure storage of CO2 in geologic 46 carbon sequestration scenarios. 47 Keywords 48 Ureolysis-induced calcium carbonate precipitation, UICP, wellbore integrity, Sporosarcina 49 pasteurii, CO2 sequestration 50 51 1. Introduction 52 1.1 Background 53 Rising levels of CO2 in the atmosphere must be reduced to achieve climate change 54 mitigation goals and limit the projected average global temperature increase 1. A significant part 2 55 of this effort is carbon capture and sequestration (CCS) wherein CO2 captured from point 56 sources like coal-fired power plants or chemical processing facilities is compressed to a 57 supercritical state and injected into subsurface storage reservoirs for sequestration on geologic 58 time scales 2. Maintaining wellbore integrity is essential to ensuring that sequestered CO2 stays 59 trapped in subsurface reservoirs and is not re-emitted to the atmosphere via leakage pathways 60 in the near wellbore environment. Current state of the art for mitigating leaky wellbores 3 where 61 conventional squeeze cementing is not appropriate includes use of fine cement, resins 4, and 62 nanomaterials 5, 6. Many of these methods are appropriate for leakage pathways on the order of 63 100 µm or larger but cannot penetrate micro-scale cracks due to the high viscosity of the sealing 64 agents. Biocement, produced during the process of ureolysis-induced calcium carbonate 65 precipitation (UICP), is emerging as a contender to aid in wellbore sealing applications where 66 more conventional methods are unsuccessful or inappropriate 7-9. In this study, we report on a 67 field study which applied UICP in the presence of CO2-affected brine to repair compromised 68 cement in a test well. 69 70 1.2 UICP Fundamentals 71 It has long been known that enzyme catalysis can be harnessed to hydrolyze urea and 72 precipitate calcium carbonate 10. The enzyme urease is widespread in nature and can be found 73 in microbes like bacteria and algae, as well as in fungi and plants 11. Ureolysis results in a pH 74 increase and produces bicarbonate (HCO3-) and carbonate (CO32-) ions. When calcium (Ca2+) 75 activity is sufficient to exceed saturation conditions, calcium carbonate (CaCO3) precipitation 76 can occur [Eqn. 1]. 77 78 CO(NH2)2 + 2H2O + Ca2+ → 2NH+4 + CaCO3 (s) [1] 79 3 80 During ureolysis-induced calcium carbonate precipitation (UICP), ureolytic bacterial cultures 81 can serve as both the source of the urease enzyme and can form a biofilm matrix on surfaces. 82 Additionally, the biofilm is believed to provide nucleation sites for precipitation to initiate 12-14. 83 Over multiple injections of ureolytic cultures and UICP-promoting media, the CaCO3 mineral 84 thickens, bridges pore throats and fractures, fills voids, and ultimately seals flow pathways 15. 85 86 1.3 Project overview 87 The primary goal of this project was to assess the ability of UICP to form a bio-cement seal 88 in the near wellbore environment in the presence of CO2-affected brine. Dissolved CO2 reduces 89 the brine pH and the fraction of CO32- ions in solution, thereby potentially making CaCO3 90 precipitation less likely and instead favoring dissolution of carbonate minerals. Since UICP has 91 been successful in sealing leaky wellbores in the oil and gas industry 7, 8, it is important to 92 assess the extent to which the process could be applied in the context of CO2 sequestration 93 where exposure to CO2-affected brine is possible. 94 A secondary goal of this UICP field demonstration was to evaluate the extent to which heat- 95 treated microbial cultures can effectively seal undesired flow pathways. The ureolytic 96 bacterium, Sporosarcina pasteurii, was the source of the urease enzyme in this study but 97 cultures were briefly heat-treated prior to injection to inactivate the microbes while preserving 98 enzyme activity. The use of heat-treated cells with intact enzyme activity may offer a viable 99 alternative in situations where injection of live cells is problematic from either a logistical or 100 regulatory standpoint. The heat treatment also mimics conditions in the deeper subsurface 101 where elevated temperature would impact cell survival. 102 The field demonstration of UICP sealing in the presence of CO2-affected brine was 103 conducted in a 24.4 cm (9.625 inch) outer-diameter cased test well at the William Crawford 104 Gorgas Electric Generating Plant (Alabama Power, Southern Company) near Parrish, Alabama, 105 USA, hereafter referred to as “Gorgas”. The well was drilled to a depth of 1498 m by the U.S. 4 106 Department of Energy to assess the potential for geologic carbon sequestration. The target 107 formation(s), however, proved to be unfavorable for CO2 storage so the well has been used 108 instead for research purposes. Previous field tests, which used live microbial cultures as the 109 source of the urease enzyme, focused on sealing a hydraulic fracture in a sandstone formation 110 340 m (1115 ft) below ground surface (bgs) 16 and sealing a channel in the well cement 310 m 111 (1017 ft) bgs 9. Additional details about the well are available in Phillips et al. (2016)16 and in 112 Section 2.1 below. UICP injections 0 1 2 3 6 327 Project Timeline (Days) 113 Not to Scale 114 Figure 1. The experimental timeline of the field demonstration. Samples were collected from 115 the downhole injection zone four times (red arrows). Injections of 5% HCl and NaHCO3 were 116 used to generate CO2 in the wellbore annular defect. Ultrasonic imaging (USIT) logs were 117 collected twice during the field demonstration (Day 1, 6) to image the materials in the wellbore 118 annulus, before and after injections of UICP-promoting fluids. Biomineralization in the annular 119 cement channel was monitored during UICP injections using pressure and flow rate 120 measurements. A third USIT log was acquired 11 months after the field work ended (Day 327). 121 122 123 Figure 1 summarizes the types of activities performed and order of events during the current 124 field demonstration. Detailed descriptions of each activity follow in later sections. Water 125 samples were collected four times over the course of the study (red arrows) to characterize 5 Establish injectivity USIT log Generate CO2 in wellbore USIT log USIT log 126 downhole geochemical and microbiological conditions. (This data will be reported elsewhere by 127 project collaborators.) Because flow behind the casing at the study depth, 310 m (1017 ft) bgs, 128 had been effectively minimized with UICP during the second field demonstration 9, acid 129 injections were used first to dissolve the calcium carbonate biomineral and establish a flow path 130 in the wellbore cement annulus. Once a channel was formed, CO2 was produced behind the 131 casing and injections of UICP-promoting fluids began. Throughout injection, pressure and flow 132 rate were monitored to assess changes in injectivity, defined here as the flow rate divided by the 133 injection pressure. Injectivity serves as a proxy for permeability or transmissivity since flow path 134 geometry through the compromised cement is unknown. Decreasing injectivity is an indication 135 of mineral formation in the channel15, 17-20. In addition, ultrasonic imaging logs were conducted 136 twice to assess the extent of mineral formation behind the casing due to UICP. The field 137 demonstration was ended when the injection pump’s maximum pressure and minimum flow rate 138 were reached. A subsequent seal integrity check was performed approximately 11 months (327 139 days) after the field demonstration ended and additional ultrasonic well logging was performed. 140 Following the seal integrity check, the well was plugged and abandoned in accordance with 141 regulatory requirements. 142 143 2. Materials and Methods 144 2.1 Well preparation and experimental design 145 The current work is the research team’s third field demonstration of ureolysis-induced 146 calcium carbonate precipitation in the Gorgas well. This field demonstration was conducted at 147 310 m (1017 ft) bgs, the same depth as the previous well cement sealing demonstration where 148 the temperature is approximately 15°C. The static water level in the well was approximately 10 149 m (30 ft) bgs at the beginning of the demonstration. The three sidewall perforations in the steel 150 casing drilled during the previous study at 310.0, 310.3, and 310.9 m (1017, 1018 and 1020 151 feet) bgs were again used to access the cement annulus 9. The 7.3 cm (2-7/8-inch) steel 6 152 injection tubing string extended from the surface to a depth of 312.7 m (1026 ft) bgs, including a 153 1 m (3 ft) perforated pup joint ending in a collar stop at 311.6 m (1022 ft). A bull plug closed off 154 the end of the tubing at 312.8 m (1026 ft) bgs. The target injection zone was isolated by a 155 packer above the casing perforations and a bridge plug below. The packer was placed between 156 the tubing string and casing at a depth of 296.6 m (973 ft) bgs. A 24.4 cm (9-5/8-inch) cast iron 157 bridge plug was set at a depth of 316 m (1037 ft) bgs and was topped with 1.5 m (5 ft) of 158 cement to prevent fluid from migrating down the casing below the injection point. An 11 L (2.9 159 gal) slickline dump bailer was used to deliver fluids to the subsurface. Fluids were pumped into 160 the bailer at the wellhead with transfer pumps and hoses dedicated to each fluid type. The 161 bottom of the bailer was fitted with a glass disk that shattered upon impact with the collar stop at 162 the target injection zone. Fresh water pumped down the tubing string after impact flushed the 163 fluids from the bailer, tubing string, and wellbore through the sidewall perforations and into the 164 cement channel behind the casing. Bailer delivery downhole and return to the surface took 165 approximately 20 minutes. 166 The initial injectivity [L/min*MPa (gal/min*psi)] of the system was obtained by monitoring the 167 wellhead pressure during an initial injection test of water down the tubing string. Injectivity was 168 low initially since the prior field demonstration 9 had sealed the flow channel behind the casing 169 in the same injection zone. 170 After the initial injection test, 3 bailers (33 L) of 5% HCl were delivered downhole to expand 171 the flow path in the cement by dissolving biomineral formed during the previous field 172 demonstration in the well. Following the expansion of a flow path in the cement, three bailers 173 (33L) each of 5% HCl and 100 g/L NaHCO3 were injected in an alternating fashion to produce 174 CO2 in the near wellbore environment according to Eqn. 2. 175 176 𝐻𝐻𝐻𝐻𝐻𝐻 + 𝑁𝑁𝑁𝑁𝐻𝐻𝐻𝐻𝐶𝐶3 → 𝑁𝑁𝑁𝑁+ + 𝐻𝐻𝐻𝐻− + 𝐻𝐻2𝐶𝐶 + 𝐻𝐻𝐶𝐶2(𝑔𝑔) Eqn. 2 177 7 178 Direct CO2 injection would have required a rigorous EPA Class VI injection permitting review 179 process, which was not feasible during the timeframe of this project. The amount of HCl (1.26 180 kg or 34.8 mol) and NaHCO3 (3.3 kg or 39.2 mol) injected for CO2 production have a theoretical 181 CO2 production of approximately 35 mol, assuming that all the reactants were consumed in the 182 intended reaction. The HCl injections to open a flow path in the wellbore annulus could have 183 produced additional CO2, depending on the extent of the reaction with calcium carbonate 184 produced during the previous field demonstration. 185 Following CO2 production, injections of UICP-promoting fluids began. For the first 2 days of 186 injecting UICP-promoting fluids, one bailer of heat-treated cells was followed by two bailers of 187 urea – calcium solution. Starting the end of the second day, the bailer contents alternated 188 between heat-treated cells and urea – calcium solution. 189 190 2.2 Well characterization – USIT logs and sampling 191 An ultrasonic imaging (USIT) log was collected twice during the field demonstration using an 192 Isolation Scanner (Schlumberger) which produces a Solid-Liquid-Gas (SLG) map of the 193 wellbore annulus via measurement of the cement impedance and flexural wave attenuation. The 194 first USIT log was recorded after establishing a flow path with acid injection but before CO2 195 production. The second USIT log was collected at the end of the field demonstration after 4 196 days of injecting UICP-promoting fluids. Another USIT log was acquired approximately 11 197 months (327 days) after the end of the field demonstration prior to plugging and abandoning the 198 Gorgas well. Image analysis using ImageJ software (v1.8.0_172) was performed to quantify the 199 solids detected behind the casing in each USIT log. The area of the image occupied by tan 200 pixels, which represent solids, were measured after applying a color threshold to the image 201 histogram. Solids correspond to hue values between 17 – 50 on the 0 – 255 scale. The full 0 – 202 255 range was applied for both saturation and brightness. 8 203 At the end of the field demonstration, samples of precipitates were scraped from the 204 injection tubing and ‘sludge’ from the well sump and bottom plug was collected for analysis. 205 These samples were subjected to X-ray diffraction (XRD) analysis to characterize the mineral 206 composition. Samples were dried and pulverized then deposited onto a glass slide coated with a 207 thin layer of petroleum jelly and analyzed with a Scintag X1 Powder X-ray diffractometer 208 equipped with a copper (Cu) k-alpha X-ray source housed at the Imaging and Chemical 209 Analysis Laboratory (ICAL) at Montana State University. 210 211 2.3 UICP-promoting fluids 212 Microbial cultures were started by inoculating 1L 18.5 g/L BHI media (Becton Dickinson, 213 Franklin Lakes, NJ) plus 20 g/L urea (Fisher Scientific) with 10 mL (4.4 x 106 CFU/mL) of frozen 214 stock of Sporosarcina pasteurii (ATCC® 11859™). After approximately 16 hours, the 1L culture 215 was transferred to 45.4 L (12 gal) yeast extract medium (15.5 g/L yeast extract, 24 g/L urea 216 (Dyno Nobel, Inc, Deer Island OR), 1 g/L NH4Cl). The 45.4 L cultures were grown at 30°C for 217 approximately 24 hours in 56.8-L (15-gal) conical bottom reactors equipped with aeration, 218 temperature control, ventilation, and recirculation as described in Kirkland et al. (2020) 7 and 219 shown in Figure 2. bio-reactors cooling tank heat-treatment tank bio-reactors 220 9 221 Figure 2. Heat-treatment system developed in the custom MSU mobile laboratory received flow 222 from 24-hr cultures grown in the bioreactors (left). The 60°C heat-treatment tank (left) was 223 wrapped in insulation to reduce heat losses. The cooling tank, located outside of the trailer, 224 contained water below 30°C. Cultures were pumped through stainless steel coils in both the 225 heating and cooling tanks. 226 227 After approximately 24 hours of growth in the conical bottom reactors, the S. pasteurii 228 cultures were combined in a holding tank from which they were pumped with a peristaltic pump 229 into the heat-treatment system (Figure 2). Microbes were heat-treated as they flowed through 230 coiled tubing suspended in reactor heated to 60°C. During the first two days of the field 231 demonstration, a 15.24 m (50-ft), 1.27 cm (1/2-inch) OD 316 stainless steel (SS) coil was used. 232 For the remainder of the experiment, a second coil was added in series (15.24 m (50-ft), 0.95 233 cm (3/8-inch) OD 316 SS) to allow for higher flowrates while maintaining adequate residence 234 time in the 60°C reactor (i.e. 8-13 minutes). Temperature was maintained at 60°C in the hot 235 water bath with immersion heaters and insulation was wrapped around the tank to reduce heat 236 loss. Heat-treated cells were cooled using a 15.24 m (50-ft), 0.95 cm (3/8-inch) OD 316 SS coil 237 submerged in a < 30°C water bath to protect the urease enzyme from inactivation due to 238 prolonged exposure to elevated temperature (Figure 2) 21. Cooled cultures were stored in a 239 holding tank at ambient temperature until they were delivered into the well. 240 Microbial cultures were sampled each day before and after heat treatment. Population 241 analysis was performed using the drop plate method 22 to count colony forming units (CFU). 242 The detection limit was 1 x 103 CFU/mL. Before filling the bailer for an injection, the heat- 243 treated cultures were also sampled for pH, electrical conductivity (EC), urea analysis with the 244 Jung Assay 23, and optical density. Optical density (OD600) was measured in triplicate to monitor 245 growth in the reactors from 200 µL samples in a 96-well plate at 600 nm using an Infinite F50 246 plate reader (Tecan, Switzerland). OD600 of the fresh YE medium blank was 0.05 ± 0.002. 10 247 Urea-calcium solution (U+C) was prepared daily with 72 g/L urea and 124 g/L CaCl2 248 (Peladow, Occidental Chemical Corp., Dallas, TX) in a 132-L (35-gal) tank. The U+C solution 249 was mixed with a drill-powered mixer and sampled for pH, urea, and Ca2+ analyses. 250 251 2.4 Batch studies – Ureolytic activity 252 The ureolytic activity of heat-treated cultures was compared to that of actively growing 253 microbial cultures using EC [mS/cm] measurements. Due to the production of NH4+ ions during 254 ureolysis (Eqn. 1), increases in EC over time can be used as a proxy to compare ureolysis rates 255 between samples 24. Ten mL of 20 g/L urea (Fisher Scientific) solution was added to 10 mL 256 samples of both actively growing S. pasteurii cultures and heat-treated cultures in 50 mL conical 257 tubes. Mixtures were vortexed initially and immediately prior to EC measurements. EC 258 measurements were performed every 30 minutes for 3 hours. 259 260 3. Results 261 3.1 Reduced Injectivity 262 Wellhead pressure and pumping flow rate were monitored and recorded as water was 263 injected to push each bailer of fluids into the wellbore channel. Figure 3 shows the injectivity of 264 each bailer delivery as a function of total injection volume. Approximately 242 L (64 gal) of 265 heat-treated microbial cultures (22 bailers) and 329 L (87 gal) of U+C media (30 bailers) were 266 injected over 4 days. The UICP treatment resulted in an order-of-magnitude reduction of 267 injectivity in the channel (Figure 3) which is indicative of biomineral formation in flow pathways 268 outside the well casing. The preliminary low injectivity of 0.76 L/min*MPa (1.38x10-4 gpm/psi) 269 measured during initial freshwater injection increased to 9.33 L/min*MPa (1.73x10-3 gpm/psi) 270 after the injection of HCl to open a flow path (Figure 3). The highest injectivity during the 271 experiment occurred during the injection of HCl and NaHCO3 to produce CO2 when the flow rate 272 was 8.33 L/min (2.2 gpm) at 0.49 MPa (708 psi) for an injectivity of 17.1 L/min*MPa (3.11x10-3 11 273 gpm/psi). After CO2 production but before UICP treatment, the channel conveyed 4.9 L/min at 274 0.49 MPa (1.3 gpm at 712 psi), for an injectivity of 10.0 L/min*MPa (1.83x10-3 gpm/psi). During 275 UICP treatment, injectivity was stable for two days before decreasing rapidly. On the third day 276 after beginning UICP injections and at a cumulative injection volume of approximately 1740 L 277 (460 gal), injectivity decreased significantly before leveling off, then decreasing again in a ‘stair- 278 step’ pattern (Figure 3). The final flow-pressure relationship at the end of the field 279 demonstration were 0.757 L/min at 0.761 MPa (0.2 gpm at 1104 psi) which equates to an 280 injectivity of 0.99 L/min*MPa (1.81x10-4 gpm/psi). An injection test 11 months later produced an 281 injectivity of 2.56 L/min*MPa (4.67x10-4 gpm/psi) which increased to 8.62 L/min*MPa (1.57x10-3 282 gpm/psi) following sustained pumping. Red arrows in Figure 3 indicate when USIT logs were 283 acquired. 284 285 Project Day 1 2 3 4 5 6 327 18 16 14 12 10 8 6 4 2 0 0 500 1000 1500 2000 2500 286 Cumulative Injection Volume (L) 287 288 Figure 3. Injections of 5% HCl were used to open a flow path behind the well casing and then 289 injections of 5% HCl and NaHCO3 were used to generate CO2 in the annular cement channel. 290 Subsequent injections of UICP-promoting fluids reduced the injectivity by an order-of-magnitude 12 Injectiivty (L/min*MPa) 291 over days 3-6. Fresh water injections (Δ), 5% HCl injections (•), HCl/NaHCO3 injections to 292 produce CO2 (■), UICP-promoting injections (◊), and pre-abandonment injection test (x). Red 293 arrows indicate USIT logs. Vertical lines indicate days along secondary x-axis; time scale is not 294 linear. 295 296 3.2 USIT logs 297 The Gorgas well was logged ultrasonically twice during the field demonstration and again 11 298 months later (Figure 3, red arrows). USIT logs (Figure 4) show a 2D projection of the materials 299 detected immediately behind the casing (right panel of each log) as a function of depth in the 300 borehole (left panel of each log). In the first USIT log acquired during the current field work 301 (left), the acid treatment appears to have expanded a channel outside the casing (black oval) 302 relative to the final USIT log recorded at the end of the prior wellbore sealing demonstration9. 303 The corresponding injectivity was 9.33 L/min*MPa (1.70x10-3 gpm/psi) (Figure 3). After UICP 304 treatment in the presence of CO2-affected brine (middle), there was a noticeable increase in the 305 presence of solids in the channel in the region of the side wall perforations 300 – 310 m (990- 306 1019 ft) bgs and more than 30 m (100 ft) above the injection point (red ovals). Compared to the 307 initial USIT log, the post-UICP log shows a 58% increase in the image area occupied by solids 308 behind the casing for the depth interval shown in Figure 4. The USIT log recorded on Day 327, 309 approximately 11 months after the end of the field work (Figure 4, right) prior to plugging and 310 abandoning the well, shows that significant solids remain despite the loss of solids relative to 311 the Post-UICP log. The image area occupied by solids measured in this last USIT log is 30% 312 greater than in the initial USIT log. The relative abundance of solids in the well-log images is 313 consistent with injectivity values recorded over the course of the fieldwork. Higher proportions 314 of solids correspond to lower injectivity and vice versa. 315 13 INITIAL POST-UICP 11 MO. LATER 316 317 Figure 4. USIT logs showing a 2D projection of the materials immediately outside the casing 318 before and after UICP treatment, where tan=solids, blue=liquid and red=gas detected. Fluids 319 were able to access the cement defect/channel in three potential locations (990, 1004, 1017- 320 1019 feet below ground surface) indicated by red arrows due to previously drilled sidewall 321 perforations. Before production of CO2 and UICP treatment, a channel was observed in the solid 322 material detected behind the casing (black oval on left panel). After UICP treatment a significant 323 increase in the amount of solid was observed (red ovals, middle panel). A follow-up USIT log 11 324 months later (Day 327, right panel) shows that significant solids remained in the location despite 325 some loss. 326 327 3.3 Sample analysis 328 XRD analysis of mineral samples collected from the injection tubing show that the mineral was 329 74.9% calcite (CaCO3), 22.2% vaterite (CaCO3), and 2.9% quartz (SiO2). A sample from the 330 well sump, which is located between the collar stop and the bottom plug, was 67.2% quartz and 14 331 32.8% magnetite (Fe3O4). Two samples of solids collected from the bottom plug were 332 predominantly iron species, magnetite and iron (III) oxide hydroxide (FeO(OH)), with quartz 333 comprising 33% and 25% of each sample, respectively. 334 335 3.4 Heat treatment of microbial cultures 336 Sporosarcina pasteurii cultures were grown to an average concentration of 2.5x108 CFU/mL 337 in the mobile laboratory conical bottom bio-reactors prior to heat-treatment. No colonies of 338 heat-treated cells were observed above the detection limit on the agar plates after 24 hours. 339 Batch studies to compare the ureolytic activity of growing vs. heat-treated cultures showed 340 larger increases in EC in the heat-treated cultures than in the growing cultures. EC increased 341 on average 41% more in heat-treated samples, with a standard deviation of 23 percentage 342 points. Excluding Day 2 data, which was an outlier, EC increased 51% more in heat-treated 343 cultures (St.dev = 7 percentage points). 344 345 4. Discussion 346 Presence of CO2 in the wellbore annulus does not appear to have impeded calcium 347 carbonate mineral formation as evidenced by the order-of-magnitude reduction in injectivity and 348 the increased solids detected in the post-UICP treatment USIT log. These findings corroborate 349 previous studies which have found that UICP can occur in the presence of CO225, 26.The ‘stair 350 step’ behavior of injectivity reduction is characteristic of UICP sealing in channels or fractures, 351 where substantial mineral precipitation is often necessary before measurable flow restriction 352 occurs. In these channel-type systems, mineral formation must reach a critical threshold before 353 the ratio of pressure and flow measurements, or injectivity, registers significant change. Once 354 the flow path is partially restricted, however, conditions are favorable at that location for 355 additional mineral to form and injectivity decreases dramatically over a relatively short time 356 frame 15. In this field demonstration, this pattern appears to repeat three times (Figure 3), with 15 357 the decreases occurring the morning of the third day of injecting UICP-promoting fluids (at a 358 cumulative injection volume of approximately 1741 L (460 gal)), the morning of the fourth day 359 (approximately 2214 L (585 gal)) and at the final bailer injection (2362 L (624 gal)). Similar 360 behavior has been observed in other studies monitoring sealing of channels with UICP 15, 16, 18-20, 361 27, 28 though some report only a single ‘stair step’ before the channel sealed. The results 362 obtained during this field demonstration suggest that when a preferred flow path sealed due to 363 UICP, the reactants were re-routed through another path in the wellbore annulus until that one 364 also sealed, and so on. This sealing behavior is in contrast to that reported where the seal 365 formed within a higher permeability sandstone rock matrix in a waterflood injection well 7, 8 or 366 within a laboratory bio-reactor sand pack 29. In these cases, mineralization and sealing of the 367 pore spaces proceeded gradually over time in a nearly linear fashion. In systems where 368 permeability is more evenly distributed, each pore that fills or is blocked with bio-mineral 369 fractionally reduces the injectivity of the system until few pathways remain open30. 370 The XRD data from the tubing string samples confirms that calcium carbonate precipitation 371 occurred downhole following the injection of UICP-promoting fluids. The fact that calcium 372 carbonate polymorphs were not detected in the sump and bottom plug suggests that the 373 mineralizing fluids were successfully transported out of the wellbore and that the precipitation 374 reactions proceeded within the cement channel as intended. The presence of iron species in 375 the well sump and bottom plug is to be expected. Raising and lowering the bailer repeatedly 376 within the steel tubing string, injection of acid to open a flow path, and injection of acid and 377 NaHCO3 to form CO2 likely contributed to the accumulation of iron species at the bottom of the 378 tubing string. 379 Extended exposure to elevated temperatures (greater than 65°C) is known to inactivate the 380 urease enzyme which could inhibit ureolysis-induced mineral precipitation 21, 31. In this study, 381 however, EC data points to higher rates of ureolysis in the heat-treated cultures. This apparent 382 contradiction is related to the temperature, 60°C, and duration of exposure the microbial 16 383 cultures experienced which was limited to 8 – 13 minutes. The goal was to inactivate the cell 384 without inactivating the enzyme. Thus, one explanation for the observed increase in ureolytic 385 activity in heat-treated cultures relates to the integrity of the cell wall following heat stress. A 386 partial breakdown of bacterial cell walls could enhance transport of urea into or urease enzyme 387 out of the cell, thereby increasing urea hydrolysis rates. In the case of the live cells, transport of 388 urea into the cells (or urease out of the cells) is recognized as a rate-limiting step in urea 389 hydrolysis 32, 33. 390 The USIT logs acquired during this project show both the potential of the UICP 391 biotechnology to seal leakage pathways in wellbores as well as some remaining questions 392 related to seal durability. The Post-UICP USIT log shows extensive solids precipitation not only 393 in the immediate vicinity of the injection, but also far above. Due to the presence of a bridge 394 plug in the casing, it was not possible to observe the extent of mineral formation below the 395 injection point though there is no reason to assume mineral could not have formed there. 396 The available data regarding the durability of UICP mineral seals in this well is mixed and 397 highlights the need for additional research on the topic. First, the preliminary water injection, 398 shown as the first data point in Figure 3, yielded a low initial injectivity of 0.76 L/min*MPa 399 (1.38x10-4 gpm/psi) and indicates that the seal produced during the previous demonstration of 400 UICP in the well 9 remained in place during the 18 months between field demonstrations. 401 Moreover, our small business team members have successfully deployed UICP to seal leakage 402 pathways in 11 commercial oil and gas wells in Colorado and Wyoming 7 with no additional 403 treatment required to satisfy regulatory standards for operation or abandonment. Four of these 404 wells in the Denver-Julesburg Basin of Colorado were horizontal production wells in the 405 Niobrara formation experiencing sustained casing pressure ranging from 80 psi (0.55 MPa) to 406 899 psi (6.2 MPa) prior to UICP treatment. Immediately after treatment, casing pressure in all 407 wells was reduced to 0 psi (0 MPa). After six months in operation following UICP treatment, two 408 of the wells produced casing pressure readings of 5 psi (0.03 MPa) and 2 psi (0.01 MPa) 17 409 (unpublished data), which is well below the regulatory value of 50 psi (0.34 MPa) that triggers 410 remedial action. The remaining two wells, which had initially yielded the highest sustained 411 casing pressures of 395 psi (2.7 MPa) and 899 psi (6.2 MPa), remained at 0 psi (unpublished 412 data). These findings from the Gorgas well and commercial production wells provide evidence 413 that the calcium carbonate biomineral seal can be durable over a timeframe of months to years. 414 At the same time, the USIT log acquired prior to abandoning the well (Figure 4, right) shows 415 a loss of solids relative to the Post-UICP USIT log acquired at the end of the field demonstration 416 (Figure 4, middle). Injectivity data suggests that the seal was further compromised following 417 sustained pumping during the subsequent injection test. There are several possible 418 explanations for this observed behavior. First, residual CO2 in the near wellbore environment 419 may have produced low-pH conditions favorable for calcium carbonate dissolution. Second, 420 heat-treatment of cells may have altered the microbial phenotype such that cells did not firmly 421 attach to surfaces as a biofilm prior to mineralizing. It is not possible, given the available data 422 from this field demonstration, to say with certainty which of these scenarios, if either, caused the 423 loss of solids after the field demonstration ended. Two points, however, are worth noting. 424 First, acidic pH conditions (e.g. < pH 7) are known to promote calcium carbonate dissolution 425 wherein the rate of dissolution depends on complex geochemical and hydrodynamic factors 426 related to fluid saturation states at the solid-liquid interface25. The surface area of the mineral 427 seal exposed to wellbore fluid is of critical importance. We hypothesize that the amount of 428 calcium carbonate that dissolves due to low pH conditions is proportional to the exposed 429 surface area. Therefore, the seal in Figure 4, which has flow pathways around the well 430 circumference and exposed surface area distributed over several meters along the wellbore 431 axis, would be expected to be impacted more by acidic conditions than a similar aperture seal 432 with fewer flow paths and less exposed surface area. The relatively uniform dissolution of 433 exposed seal edges, apparent in the two USIT logs, should be noted. 18 434 Second, despite exhibiting higher ureolysis rates than live cells, the heat-treated S. pasteurii 435 cultures do not appear to have produced as robust a seal as live cells have done in previous 436 demonstrations 9. Ma et al. 34 indicate that intact cell structure promotes nucleation of mineral 437 during UICP. Secchi et al. 35 found preferential attachment of elongated, motile cells (like 438 S.pasteurii) on the downstream side of surface roughness under moderate hydrodynamic 439 conditions. Heat-treatment would be expected to modify both of these phenotypic 440 characteristics – cell shape and motility – via damage to the cell membrane, thereby decreasing 441 the capture efficiency or attachment of cells to the surfaces in the wellbore and by reducing the 442 potential of the cell to act as a nucleation site. Heat-treated cells would be more likely to 443 behave instead as passive particles in the flow where they would be intercepted by surfaces on 444 the upstream side of wellbore annulus surface roughness 35. This ‘passive particle’ behavior 445 may be useful in explaining the formation of mineral so far above the injection zone. Heat- 446 treated cells may have been carried in the flow up the well casing beyond the injection point 447 until they intercepted a surface due, in part, to cell lysis and other phenotypic changes. 448 Further numerical simulations 36, 37 and validation of these hypotheses in the lab could 449 elucidate the mechanisms that promote a durable seal and advance the process to develop 450 deployable microbial cultures for large-scale adoption of the biotechnology in the field 38. 451 452 5. Conclusions 453 The presence of CO2 did not impede mineral formation during the field demonstration as 454 evidenced by the reduction in injectivity observed during UICP treatment as well as the 455 additional solids present in the post-treatment USIT log. Additionally, urease enzyme from heat- 456 treated microbial cultures was an effective catalyst for the UICP reactions and higher ureolysis 457 rates were observed in the heat-treated cultures than in live cultures, perhaps due to cell lysis. 458 At the same time, the seal formed in the wellbore annulus in this field demonstration was less 459 robust than a previous seal produced by UICP in the same well without the presence of CO2- 19 460 affected brine and using live microbes. The results of this experiment suggest that UICP can be 461 promoted in the presence of CO2 impacted brine to seal leakage pathways and may be effective 462 to ensure storage of CO2 in geologic carbon sequestration scenarios. Further research is 463 needed to study the durability of the seal over extended time scales under adverse wellbore 464 conditions. 465 466 Acknowledgements 467 This material is based upon work supported by the Department of Energy under Grant No. DE- 468 FE0024296. Any opinions, findings, conclusions or recommendations expressed herein are those 469 of the authors and do not necessarily reflect the views of the Department of Energy (DOE). This 470 work was performed in part at the Montana Nanotechnology Facility (MONT), a member of the 471 National Nanotechnology Coordinated Infrastructure (NNCI), which is supported by the National 472 Science Foundation (Grant# ECCS-1542210). The authors thank Abby Thane Davis and Laura 473 Dobeck at the Center for Biofilm Engineering, as well as Nathaniel Rieders at ICAL. 474 475 References 476 477 1. Tong, D.; Zhang, Q.; Zheng, Y.; Caldeira, K.; Shearer, C.; Hong, C.; Qin, Y.; Davis, S. J., 478 Committed emissions from existing energy infrastructure jeopardize 1.5 °C climate target. Nature 2019, 479 572 (7769), 373-377. 480 2. Agency, I. E., The Role of CO2 Storage. IEA: Paris, 2019; Vol. Exploring Clean Energy Pathways. 481 3. Jafariesfad, N.; Sangesland, S.; Gawel, K.; Torsæter, M., New Materials and Technologies for 482 Life-Lasting Cement Sheath: A Review of Recent Advances. SPE Drilling & Completion 2020, 35 (02), 262- 483 278. 484 4. Todorovic, J.; Raphaug, M.; Lindeberg, E.; Vralstad, T.; Buddensiek, M. L., Remediation of 485 Leakage through Annular Cement Using a Polymer Resin: a Laboratory Study. 8th Trondheim Conference 486 on Co2 Capture, Transport and Storage 2016, 86, 442-449. 487 5. Genedy, M.; Kandil, U. F.; Matteo, E. N.; Stormont, J.; Taha, M. M. R., A new polymer 488 nanocomposite repair material for restoring wellboreseal integrity. International Journal of Greenhouse 489 Gas Control 2017, 58, 290-298. 490 6. Genedy, M.; Matteo, E. N.; Stenko, M.; Stormont, J. C.; Taha, M. R., Nanomodified Methyl 491 Methacrylate Polymer for Sealing of Microscale Defects in Wellbore Systems. J. Mater. Civ. Eng. 2019, 31 492 (7). 20 493 7. Kirkland, C. M.; Hiebert, R.; Hyatt, R.; McCloskey, J.; Kirksey, J.; Thane, A.; Cunningham, A. B.; 494 Gerlach, R.; Spangler, L.; Phillips, A. J., Direct Injection of Biomineralizing Agents to Restore Injectivity 495 and Wellbore Integrity. SPE Production & Operations 2020, Preprint (Preprint), 8. 496 8. Kirkland, C. M.; Thane, A.; Hiebert, R.; Hyatt, R.; Kirksey, J.; Cunningham, A. B.; Gerlach, R.; 497 Spangler, L.; Phillips, A. J., Addressing wellbore integrity and thief zone permeability using microbially- 498 induced calcium carbonate precipitation (MICP): A field demonstration. Journal of Petroleum Science 499 and Engineering 2020, 190, 107060. 500 9. Phillips, A. J.; Troyer, E.; Hiebert, R.; Kirkland, C. M.; Gerlach, R.; Cunningham, A.; Spangler, L.; 501 Kirksey, J.; Rowe, W.; Esposito, R., Enhancing wellbore cement integrity with microbially induced calcite 502 precipitation (MICP): a field scale demonstration. Journal of Petroleum Science and Engineering 2018, 503 171, 1141-1148. 504 10. Fujita, Y.; Ferris, E. G.; Lawson, R. D.; Colwell, F. S.; Smith, R. W., Calcium carbonate 505 precipitation by ureolytic subsurface bacteria. Geomicrobiol. J. 2000, 17 (4), 305-318. 506 11. Krajewska, B., Ureases I. Functional, catalytic and kinetic properties: A review. Journal of 507 Molecular Catalysis B-Enzymatic 2009, 59 (1-3), 9-21. 508 12. Oppenheimer-Shaanan, Y.; Sibony-Nevo, O.; Bloom-Ackermann, Z.; Suissa, R.; Steinberg, N.; 509 Kartvelishvily, E.; Brumfeld, V.; Kolodkin-Gal, I., Spatio-temporal assembly of functional mineral 510 scaffolds within microbial biofilms. Npj Biofilms and Microbiomes 2016, 2. 511 13. Stocks-Fischer, S.; Galinat, J. K.; Bang, S. S., Microbiological precipitation of CaCO3. Soil Biol. 512 Biochem. 1999, 31 (11), 1563-1571. 513 14. Zambare, N. M.; Naser, N. Y.; Gerlach, R.; Chang, C. B., Mineralogy of microbially induced 514 calcium carbonate precipitates formed using single cell drop-based microfluidics. Scientific Reports 2020, 515 10 (1), 17535. 516 15. Kirkland, C. M.; Norton, D.; Thane, A.; Hiebert, R.; Hommel, J.; Kirksey, J.; Esposito, R.; 517 Cunningham, A.; Gerlach, R.; Spangler, L.; Phillips, A. In Biomineralization and wellbore integrity: a 518 microscopic solution to subsurface fluid migration, 14th Greenhouse Gas Control Technologies 519 Conference, Melbourne, Australia, 21-26 October 2018; Melbourne, Australia, 2019. 520 16. Phillips, A. J.; Cunningham, A. B.; Gerlach, R.; Hiebert, R.; Hwang, C.; Lomans, B. P.; Westrich, 521 J.; Mantilla, C.; Kirksey, J.; Esposito, R.; Spangler, L., Fracture Sealing with Microbially-Induced Calcium 522 Carbonate Precipitation: A Field Study. Environ. Sci. Technol. 2016, 50 (7), 4111-4117. 523 17. Cuthbert, M. O.; McMillan, L. A.; Handley-Sidhu, S.; Riley, M. S.; Tobler, D. J.; Phoenix, V. R., A 524 Field and Modeling Study of Fractured Rock Permeability Reduction Using Microbially Induced Calcite 525 Precipitation. Environ. Sci. Technol. 2013, 47 (23), 13637-13643. 526 18. Minto, J. M.; MacLachlan, E.; El Mountassir, G.; Lunn, R. J., Rock fracture grouting with 527 microbially induced carbonate precipitation. Water Resour. Res. 2016, 52 (11), 8810-8827. 528 19. Tobler, D. J.; Minto, J. M.; El Mountassir, G.; Lunn, R. J.; Phoenix, V. R., Microscale Analysis of 529 Fractured Rock Sealed With Microbially Induced CaCO3 Precipitation: Influence on Hydraulic and 530 Mechanical Performance. Water Resour. Res. 2018, 54 (10), 8295-8308. 531 20. Kirkland, C. M.; Norton, D.; Firth, O.; Eldring, J.; Cunningham, A. B.; Gerlach, R.; Phillips, A. J., 532 Visualizing MICP with X-ray uCT to enhance cement defect sealing. International Journal of Greenhouse 533 Gas Control 2019, 86, 93 - 100 534 21. Feder, M. J.; Akyel, A.; Morasko, V. J.; Gerlach, R.; Phillips, A. J., Temperature-dependent 535 inactivation and catalysis rates of plant-based ureases for engineered biomineralization. Engineering 536 Reports 2020, n/a (n/a), e12299. 537 22. Herigstad, B.; Hamilton, M.; Heersink, J., How to optimize the drop plate method for 538 enumerating bacteria. J. Microbiol. Methods 2001, 44 (2), 121-129. 539 23. Jung, D.; Biggs, H.; Erikson, J.; Ledyard, P. U., New colorimetric reaction for endpoint, 540 continuous-flow and kinetic measurement of urea Clin. Chem. 1975, 21 (8), 1136-1140. 21 541 24. Shashank, B. S.; Minto, J. M.; Singh, D. N.; El Mountassir, G.; Knapp, C. W., Guidance for 542 Investigating Calcite Precipitation by Urea Hydrolysis for Geomaterials. J. Test. Eval. 2018, 46 (4), 1527- 543 1538. 544 25. Mitchell, A. C.; Phillips, A.; Schultz, L.; Parks, S.; Spangler, L.; Cunningham, A. B.; Gerlach, R., 545 Microbial CaCO3 mineral formation and stability in an experimentally simulated high pressure saline 546 aquifer with supercritical CO2. International Journal of Greenhouse Gas Control 2013, 15, 86-96. 547 26. Verba, C.; Thurber, A. R.; Alleau, Y.; Koley, D.; Colwell, F.; Torres, M. E., Mineral changes in 548 cement-sandstone matrices induced by biocementation. International Journal of Greenhouse Gas 549 Control 2016, 49, 312-322. 550 27. Phillips, A. J.; Eldring, J.; Hiebert, R.; Lauchnor, E.; Mitchell, A. C.; Cunningham, A.; Spangler, 551 L.; Gerlach, R., Design of a meso-scale high pressure vessel for the laboratory examination of 552 biogeochemical subsurface processes. Journal of Petroleum Science and Engineering 2015, 126, 55-62. 553 28. Phillips, A. J.; Lauchnor, E.; Eldring, J.; Esposito, R.; Mitchell, A. C.; Gerlach, R.; Cunningham, 554 A. B.; Spangler, L. H., Potential CO2 Leakage Reduction through Biofilm-Induced Calcium Carbonate 555 Precipitation. Environ. Sci. Technol. 2013, 47 (1), 142-149. 556 29. Kirkland, C. M.; Zanetti, S.; Grunewald, E.; Walsh, D. O.; Codd, S. L.; Phillips, A. J., Detecting 557 Microbially Induced Calcite Precipitation in a Model Well-Bore Using Downhole Low-Field NMR. Environ. 558 Sci. Technol. 2017, 51 (3), 1537-1543. 559 30. Ferris, F. G.; Stehmeier, L. G.; Kantzas, A.; Mourits, F. M., Bacteriogenic mineral plugging. J. Can. 560 Pet. Technol. 1996, 35 (8), 56-61. 561 31. Illeova, V.; Polakovic, M.; Stefuca, V.; Acai, P.; Juma, M., Experimental modelling of thermal 562 inactivation of urease. J. Biotechnol. 2003, 105 (3), 235-243. 563 32. Beckers, G.; Bendt, A. K.; Krämer, R.; Burkovski, A., Molecular identification of the urea uptake 564 system and transcriptional analysis of urea transporter- and urease-encoding genes in Corynebacterium 565 glutamicum. J. Bacteriol. 2004, 186 (22), 7645-7652. 566 33. Mobley, H. L.; Island, M. D.; Hausinger, R. P., Molecular biology of microbial ureases. 567 Microbiological Reviews 1995, 59 (3), 451-480. 568 34. Ma, L.; Pang, A.-P.; Luo, Y.; Lu, X.; Lin, F., Beneficial factors for biomineralization by ureolytic 569 bacterium Sporosarcina pasteurii. Microbial Cell Factories 2020, 19 (1), 12. 570 35. Secchi, E.; Vitale, A.; Mino, G. L.; Kantsler, V.; Eberl, L.; Rusconi, R.; Stocker, R., The effect of 571 flow on swimming bacteria controls the initial colonization of curved surfaces. Nature Communications 572 2020, 11 (1). 573 36. Hommel, J.; Akyel, A.; Frieling, Z.; Phillips, A. J.; Gerlach, R.; Cunningham, A. B.; Class, H., A 574 Numerical Model for Enzymatically Induced Calcium Carbonate Precipitation. Applied Sciences 2020, 10 575 (13). 576 37. Cunningham, A. B.; Class, H.; Ebigbo, A.; Gerlach, R.; Phillips, A. J.; Hommel, J., Field-scale 577 modeling of microbially induced calcite precipitation. Computational Geosciences 2019, 23 (2), 399-414. 578 38. Landa Marbán, D.; Tveit, S.; Kumar, K.; Gasda, S., Practical approaches to study microbially 579 induced calcite precipitation at the field scale. 2020. 580 22