Published February 9, 2016
Agricultural & Environmental Letters
Approaching Brine Spill Remediation from the Surface: A New In Situ Method
Research Letter Aaron L. M. Daigh* and Aaron W. Klaustermeier Core Ideas • In situ remediation of brine spills is possible via surface extraction of salts. • A crystallization inhibitor allowed for 29 to 57% of salts to be harvested from the soil surface. • Methods for surface extraction of salts can aid in expediting brine spill remediation timelines.
Abstract: Well drilling for energy resources innately produces brine waters. These brines often contain sodium in the 10,000 to >100,000 mg L-1 range. In situ soil remediation of brine spills traditionally consists of diluting salts with organic materials and then infiltrating divalent cations. This leaching technique can require years to centuries, as a function of soil clay content, to remove salts from the root zone. We present a new in situ remediation method that extracts salts from the soil surface and expedites remediation. Surface application of a crystallization inhibitor (ferric hexacyanoferrate) to brinecontaminated soils followed by an evaporative flux resulted in dendritic crystal growth above the soil surface. This process allowed easy harvest of 29 to 57% of salts within 7 d without mechanical disturbance to the soil. Future studies should include loadingrate optimization, in-field testing, evaluation of reaction product fate and transport, and identification of other amendments to rapidly extract salts.
W
A.L. M. Daigh and A.W. Klaustermeier, North Dakota State Univ., Soil Science Dep., Walster Hall, Fargo, ND 58108.
Copyright © American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America. 5585 Guilford Rd., Madison, WI 53711 USA.
This is an open access article distributed under the terms of the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Agric. Environ. Lett. 1:150013 (2016) doi:10.2134/ael2015.12.0013 Received 2 Dec. 2015. Accepted 21 Dec. 2015. *Corresponding author (
[email protected]).
ell drilling for oil, gas, and geothermal infrastructures produce saline-sodic wastewaters (i.e., brines). Since 2001, official brine spill reports filed at the North Dakota Department of Health and the North Dakota Industrial Commission exceed 13,000 cases with an average of 42 reports filed per week during 2014 (North Dakota Department of Health, 2015). These brines contain sodium and chloride at >10,000 and >100,000 mg L-1, respectively, with electrical conductivities (EC) often above 200 dS m-1, values similar to those reported by others across North America (Tomlinson et al., 2015; Dresel and Rose, 2010; Jury and Weeks, 1978). The traditional in situ remediation of brine-contaminated soil is the incorporation of organic materials to dilute salts and subsequent infiltration of divalent cations (Sublette et al., 2007). The infiltration of divalent cations (mainly Ca2+ based) is to displace sodium from soil exchange sites, minimize soil dispersion, and increase effective pore diameters to leach salts from the root zone. Remediation timelines can range from 1 to >1000 yr while deploying this traditional method (Jury and Weeks, 1978). Given these time scales, it is surprising that no efforts attempting to bring salts to the soil surface for harvesting are reported in the literature. Such methods, if feasible, could expedite remediation and reduce the potential for Cl- contamination of groundwaters. The objective of this study was to determine if harvesting brine salts via the soil surface is possible without mechanically disturbing the soil. During a thorough literature search, we found that crystallization inhibitors can disrupt salt precipitation within the porous networks of building walls and bricks by delaying salt precipitation and result in salt formations on the outer surfaces (i.e., salt efflorescence; Gupta et al., 2012; Rodriguez-Navarro et al., 2002; Selwitz and Doehne, 2002). We hypothesize that the infiltration of a dilute crystallization inhibitor suspension followed by an evaporative flux on brine-contaminated soil will allow dendritic crystal growth above the soil surface and subsequently allow for easy removal of salts with minimal to no soil disturbance. Page 1 of 1
Procedures A sandy loam, loam, and silty clay soil with saturated paste extract EC’s < 2 dS m-1 were air-dried, ground, sieved through a 2-mm sieve, and packed to a 6 cm depth in 18 polyvinyl chloride columns (six columns per soil texture; column dimensions: 7.7 cm diameter by 10 cm height). The bottom of each column was sealed with Plexiglas to achieve a zero flux boundary for water and solutes. The top of each column remained open to the atmosphere to allow the infiltration of solutions and evaporative fluxes. Deionized water was saturated with NaCl to create a brine solution (EC of 244 dS m-1), and then 110 mL of the brine was infiltrated into each of the 18 soil columns. The brine-treated soils were allowed to air-dry for 7 d until salt precipitants were visible. For each soil texture, a 110-mL suspension of the crystallization-inhibitor hexacyanoferrate and deionized H2O was then infiltrated into four of the six brine-contaminated soil columns. Of the four treated soil columns, two columns received a 0.00001 M ferric hexacyanoferrate suspension and two columns received a 0.01 M ferric hexacyanoferrate suspension for each soil texture. The remaining two columns not treated with the crystallization inhibitor were used as controls and received no additional fluids following the initial brine contamination. The 18 columns were allowed to evaporate again in the laboratory at 20.7 ± 0.3 оC and 22.5 ± 1.6% relative humidity, and salts were harvested using a scoopula 7 d after initial efflorescence began to occur. During a preliminary test, little to no dendritic crystal formation occurred with the procedure described above. The infiltrated NaCl saturated water would have lowered the soil’s osmotic potential to approximately -5 MPa and limited the evaporative flux. Additionally, the conditions for ferric hexacyanoferrate to solubilize to Fe and the crystallization-inhibitor hexacyanoferrate may have been limited under our laboratory conditions. Therefore, 20 mL of liquid ammonia was added to the ferric hexacyanoferrate suspensions before application to obtain efflorescence of salts at the soil surface. Ammonia enhances the liquid evaporative flux due to its low intermolecular forces (i.e., high vapor pressure and low boiling point [-33 оC]) and its solvent properties for ionic compounds, including metal cyanogens (Fernelius and Johnson 1928). Dendritic crystal formations were obtained with the ferric hexacyanoferrate suspension over the 7-d period when the evaporative flux, and potentially the solubility of the hexacyanoferrate, was enhanced with ammonia (Fig. 1 and 2). The moisture retention in the dendrite formations prevents hardening and crusting and allows for easy harvesting of salts. The dendrite formations were carefully hand harvested as to not remove any soil minerals and initially weighed wet. A salt subsample was weighed, dried at 105 оC for 48 h, and immediately weighed after oven drying to obtain a salt water content. The total amount of salts harvested from each soil column was then corrected for moisture, and the oven-dry salt weight is reported below as a percentage of the applied brine NaCl salts. Page 2 of 4
Fig. 1. Image of dendrite formation following the infiltration of the crystallization-inhibitor ferric hexacyanoferrate in a brinecontaminated soil column.
To determine if soil texture and the crystallization inhibitor’s loading rate affected (i) the amount of salts harvested and (ii) if the quantity of harvested salts was different from zero, a mixed model analysis of variance was used in SAS with means separate using LSD at the 0.05 level (version 9.4, SAS Institute, Inc.). The treated and untreated (control) soil columns could not be directly compared in the statistical analysis because all the untreated soil columns had zero harvested salts and, therefore, zero variance. Because of the zero variance issue, the treated soils were deemed as being significantly different than zero if the LSD did not overlap with zero.
Results and Discussion We successfully extracted and harvested an average of 46, 57, and 29% of the brine salts from the sandy loam, loam, and silty clay soils, respectively, within 7 d after efflorescence began using a single application of the 0.01 M ferric hexacyanoferrate suspension. Statistically, the sandy loam and loam soils produced similar and significantly greater quantities of harvestable salts than the silty clay soil. Additionally, all soil columns treated with the 0.01 M suspension had harvestable salt quantities significantly different than zero. The standard deviation of the two replicates per soil texture was 9, 6, and 1% for the sandy loam, loam, and silty clay soils, respectively. These data also suggest greater variability in harvestable salt as sand content increases. In the 0.00001 M suspension-treated and the control soil columns, a cemented salt crust formed within the top several millimeters of the soil surface as expected. These cemented salt crusts were not harvestable without mechanically disturbing and inadvertently removing soil minerals. Therefore, harvestable salt quantities were zero for all 0.00001 M ferric hexacyanoferrate suspension treated and control soil columns (i.e., zero variance among experimental units). Ammonia itself does not appear to AGRICULTURAL & ENVIRONMENTAL LETTERS
noferrate is applied as suspended colloids and do not decompose rapidly except under low pH (Meeussen et al., 1992). Therefore, the infiltration of the ferric hexacyanoferrate suspension, as done in this study, results in a thin layer of visible colloids at the soil–air interface. Further evaluation is needed to determine the effect of incorporating ferric hexacyanoferrate into the soil surface layers on salt extraction. The ferric hexacyanoferrate is considered as an iron cyanide mineral and is not toxic (Sharpe 1976). According to Meeussen et al. (1992), referencing Sharpe (1976), “these complexes are generally considered to be extraordinarily stable and are even called kinetically inert.” Once the ferric hexacyanoferrate does decompose, the hexacyanoferrate reaction product has a half-life of 10 to 1000 yr between soil pH of 5.5 to 7.5, respectively (see Fig. 8 in Meeussen et al., 1992). Even if a soil’s pH decreases to 4, the half-life of the hexacyanoferrate is 1 yr. The reaction products of the hexacyanoferrate are hydrogen- and free-cyanide. However, we use a dilute suspension of the ferric hexacyanoferrate, and the proposed use of this compound is not in a small, enclosed space where the volatile hydrogen- and free-cyanides will accumulate to toxic levels. Additionally, as the soil is reclaimed and vegetation reestablished, microorganisms can readily decompose and assimilate the cyanides (Yu, 2015; Luque-Almagro et al., 2005; Kjeldsen, 1999; CCME, 1999). Moreover, the ferric hexacyanoferrate is used here as just an example; other crystallization inhibitors with low-toxicity reaction products may produce similar results for remediation purposes. This experiment demonstrates that a crystallization inhibitor can be used to induce harvestable salt growths at the soil surface, and we recommend the development and evaluation of more crystallization inhibitors as tools for rapid brine spill remediation.
Conclusions
Fig. 2. A time series of images showing a brine-contaminated soil immediately after infiltrating the crystallization-inhibitor ferric hexacyanoferrate, the initiation of salt efflorescence during an evaporative flux, the coverage of the soil surface with dendrite formations as salt efflorescence progresses, and the subsequent vertical growth of the dendrites.
have properties to induce dendritic crystal growth since the same quantities of ammonia was applied in both of the 0.01 M and 0.00001 M treatments. The growth of dendritic crystals above the soil surface is expected to be continuous if an evaporative flux persists due to the convective transport of liquids and solutes to the soil surface. Therefore, the 7-d period used here is arbitrary, and a greater proportion of the brine salts may likely be extracted under ample water supply conditions. Additionally, the ferric hexacyaAGRICULTURAL & ENVIRONMENTAL LETTERS
In situ remediation of brine contaminated soils via surface extraction of salts is possible without mechanically disturbing the soil. The method described here resulted in the rapid harvesting of 29 to 57% of salts from NaClcontaminated soils within 7 d using the 0.01 M crystallization inhibitor suspension. New in situ methods of deploying crystallization inhibitors for rapid salt extraction show great potential and could be useful for brinespill first responders coupled with other remediation methods to expedite the full remediation of brine-contaminated soils. Such methods could reduce the risk of chloride contamination to groundwaters. Future studies should include (i) loading rate optimization, (ii) in-field testing, (iii) further investigation of the fate and transport of ferric hexacyanoferrate reaction products during the remediation process, and (iv) the evaluation of other crystallization inhibitors, nanotechnology, and coatings that can induce dendritic crystal growth on soil surfaces for the purpose of rapid salt removal with minimal toxicity to the environment. Page 3 of 4
Acknowledgments This work is supported by the USDA National Institute of Food and Agriculture Hatch project 1005366 and funded by the North Dakota State University Agricultural Experiment Station. The authors thank Radu Carcoana for his technical assistance during laboratory analysis.
References Canadian Council of Ministers of the Environment (CCME). 1999. Canadian soil quality guidelines for the protection of environmental and human health: Cyanide (free) (1997) In: Canadian environmental quality guidelines, 1999. Canadian Council of Ministers of the Environment, Winnipeg. Dresel, P. E. and A. W. Rose. 2010. Chemistry and origin of oil and gas well brines in western Pennsylvania. Open File Rep. OFOG 10–01.0. Pennsylvania Geological Survey, Harrisburg. Fernelius, W.C., and W.C. Johnson. 1928. Liquid ammonia as a solvent and the ammonia system of compounds: I. Liquid ammonia as a solvent. J. Chem. Educ. 5:664. doi:10.1021/ed005p664 Gupta, S., K. Terheiden, L. Pel, and A. Sawdy. 2012. Influence of ferrocyanide inhibitors on the transport and crystallization processes of sodium chloride in porous building materials. Cryst. Growth Des. 12:3888–3898. doi:10.1021/cg3002288 Jury, W.A., and L.V. Weeks. 1978. Solute travel-time estimates for tiledrained fields: III. Removal of a geothermal brine spill from soil by leaching. Soil Sci. Soc. Am. J. 42:679–684. doi:10.2136/ sssaj1978.03615995004200050003x Kjeldsen, P. 1999. Behavior of cyanides in soil and groundwater: A review. Water Air Soil Pollut. 115:279–308. doi:10.1023/A:1005145324157
Page 4 of 4
Luque-Almagro, V.M., M.J. Huertas, M. Martinez-Luque, C. MorenoVivian, M.D. Roldan, L.J. Garcia-Fil, F. Castillo, and R. Blasco. 2005. Bacterial degradation of cyanide and its metal complexes under alkaline conditions. Appl. Environ. Microbiol. 71:940–947. doi:10.1128/ AEM.71.2.940-947.2005 Meeussen, J.C.L., M.G. Keizer, and F.A.M. De Haan. 1992. Chemical stability and decomposition rate of iron cyanide complexes in soil solutions. Environ. Sci. Technol. 26:511–516. doi:10.1021/es00027a010 North Dakota Department of Health. 2015. Environmental incident reports. Oilfield environmental incidents. www.ndhealth.gov/EHS/Spills/ Rodriguez-Navarro, C., L. Linares-Fernandez, E. Doehne, and E. Sebastian. 2002. Effects of ferrocyanide ions on NaCl crystallization in porous stone. J. Cryst. Growth 243:503–516. doi:10.1016/ S0022-0248(02)01499-9 Selwitz, C., and E. Doehne. 2002. The evaluation of crystallization modifiers for controlling salt damage to limestone. J. Cult. Herit. 3:205–216. doi:10.1016/S1296-2074(02)01182-2 Sharpe, A.G. 1976. The chemistry of cyano complexes of the transition metals. Academic Press, New York. Sublette, K.L., J.B. Tapp, J.B. Fisher, E. Jennings, K. Duncan, G. Thoma, J. Brokaw, and T. Todd. 2007. Lessons learned in remediation and restoration in the Oklahoma prairie: A review. Appl. Geochem. 22:2225– 2239. doi:10.1016/j.apgeochem.2007.04.011 Tomlinson, H., A. Klaustermeier, R. Limb, A.L. Daigh, and K. Sedivec. 2015. The impacts of brine spill and vegetation parameters in the Bakken region of western North Dakota. In: Annual Meeting Abstracts, Society for Range Management, Sacramento, CA. 31 Jan.–6 Feb. 2015. Society for Range Management, Wheat Ridge, CO. Yu, X.Z. 2015. Uptake, assimilation and toxicity of cyanogenic compounds in plants: Facts and fiction. Int. J. Environ. Sci. Technol. 12:763–774. doi:10.1007/s13762-014-0571-6
AGRICULTURAL & ENVIRONMENTAL LETTERS
