Article pubs.acs.org/est
Thermally Released Arsenic in Porewater from Sediments in the Cold Lake Area of Alberta, Canada Muhammad Babar Javed and Tariq Siddique* Department of Renewable Resources, University of Alberta, Edmonton, AB T6G 2G7, Canada S Supporting Information *
ABSTRACT: Elevated arsenic (As) in aquifers in close proximity to in situ oil sands extraction in the Cold Lake area, Alberta, Canada is attributed to high temperature steam (∼200 °C) injected into oil sands deposits to liquefy bitumen. Heat propagated from hot injection wells alters physicochemical properties of the surrounding sediments and associated porewater. Seven sediments from four different cores drilled up to ∼300 m depth collected from different locations in the area were used to study the thermal effect (∼200 °C) on As distribution in the sediments and its release into porewater. Sediments were moistened with synthetic aquifer or deionized water according to the moisture regimes present in aquitard, aquifer and fractured zones. Heat application greatly released As in the porewater (500−5200 and 1200−6600 μg L−1) from aquifer and fractured sediments, respectively. Mass balance of As chemical fractionation showed that ∼89−100% of As in porewater was released from exchangeable and specifically adsorbed As in the sediments. Heat application also altered As distribution in the sediments releasing As from exchange surfaces and amorphous Fe oxides to soluble As fraction. The results provide great insight into As release mechanisms warranting development of strategies to mitigate groundwater As contamination during industrial operation.
■
INTRODUCTION Arsenic (As) is one of the most abundant trace elements in the earth’s crust with some of its chemical forms being extremely toxic to living organisms.1 In recent years, As groundwater contamination and its subsequent serious health effects have received global attention, where ∼100 million people in the world have been at risk of consuming As contaminated water.2−4 In the Cold Lake area of Alberta, Canada, higher concentrations of As (>10 μg L−1; maximum acceptable concentration in drinking water in Canada)5 have been found in groundwater6,7 which is extensively used for agricultural and household purposes.6,8 The studies conducted by the Alberta Health and wellness, Alberta Environment and Alberta geological survey correlated higher concentrations of As in groundwater to the bedrock geological formations containing marine shale in the region.6,9,10 To better understand the natural source of groundwater As, we characterized ∼135 sediment samples collected from 5 cores drilled up to ∼300 m depth in surficial deposits/glacial till developed on marine shale at different locations in the Cold Lake area for As chemical fractionation and mineralogy.11 The total As concentration in the sediments varied from ∼1 to 35 μg g−1. The higher As (∼32 μg g−1; n = 2) was found in sediments from Lea Park formation developed on shale in core 14−01w.11 Most of the As (up to 46% of total As) was reported in the exchangeable and specifically adsorbed forms along with a significant amount (up to 34% of total As) tied up in pyrite and arsenopyrite minerals.7,11 In addition to the presence of naturally higher As in the sediments, the area is also enriched in bituminous sands (oil sands) attracting growing industrial activity. Significant oil © 2016 American Chemical Society
sands deposits (∼97%) are located at greater depths (∼400 m deep) requiring the use of in situ technique such as cyclic steam stimulation (CSS) or steam assisted gravity drainage (SAGD) for bitumen extraction. The bitumen is very viscous at normal temperature; therefore, heat is required to decrease the viscosity so that it can be brought to the surface (http:// www.ramp-alberta.org/ramp.aspx). During the high temperature (∼200 °C) steam injection for bitumen liquefaction, heat also radiates away from the hot wells to the surrounding sediments and associated porewater.12,13 A numerical modeling study to predict the transport of heat plume during CSS suggested that the thermal plume could transport to a distance of over 100 m in a 5 m thick aquifer with a velocity of ∼25 m year−1.14 A possible consequence of heat propagation is mobilization of As from sediments into aquifer.12 A monitoring program conducted by Canadian Natural Resources Ltd. (CNRL) found a plume of dissolved As approximately 360 m down-gradient from CSS injectors. The As concentration was ∼50 μg L−1 (5 times higher than the As drinking water guideline) in the leading edge of plume.15 The extent of As released into porewater may depend on As distribution in different inorganic and organic phases in sediments16 that can be altered by localized heating of sediments.13 The thermally induced changes in As distribution have the potential to deteriorate groundwater quality by mobilizing As from different sediment host phases.13 Received: Revised: Accepted: Published: 2191
September 17, 2015 January 13, 2016 February 3, 2016 February 3, 2016 DOI: 10.1021/acs.est.5b04555 Environ. Sci. Technol. 2016, 50, 2191−2199
Article
Environmental Science & Technology
Table 1. Physicochemical Properties of Sediments (S1−S7) Used to Study Thermally Released Arsenic (As) from Sediments into Porewater sample ID S1 S2 S3 S4 S5 S6 S7 a
As (total) μg g−1
TOC %
sulfide %
pH
lithologya
formationa
core IDa
± ± ± ± ± ± ±
0.6 0.1 0.6 0.3 0.7 2.8 ND
0.1 0.1 0.7 0.3 0.6 1.9 2.2
7.3 7.2 9.1 7.4 6.6 6.2 7.1
Clay and silt Till ND Mudstone Clay and silt Till ND
Empress Marie Creek Shale Bonnyville Ethel Lake Bonnyville Shale
WR991 WEPA991 WR991 WEPA992 WEPA992 WEPA991 14−01W
3.5 4.2 7.0 8.8 10.5 19.9 28.4
0.1 0.3 0.7 0.9 1.1 1.8 1.8
The information about lithology, land formation and core ID was obtained from Andraishek28 and Javed et al., 2014.11
hydrogeochemistry of the region. Seven sediment samples (S1−S7) collected from four different cores acquired from different locations were used in this study. Those sediments contained different As concentrations (∼4−28 μg g−1), varied in their physicochemical properties and belonged to different geological formations such as Marie Creek, Ethel Lake, Bonnyville, Empress and Shale in the region (Table 1). The comprehensive description of the cores and the lithology of the area have been reported previously.11 The sediments were ground with agate mortar and pestle and passed through 125 μm sieve before their physicochemical characterization and exposure to intense heat. The pH of sediment slurries prepared with 0.05 M CaCl2 solution (1:5 ratio)24 was measured using accumet AR20 pH meter. The pH ranged between 6 and 9, but most sediments had pH around 7. Total organic carbon (TOC) contents in sediments determined by loss on ignition (LOI) method25 were 0.05) was found in the As release using two different apparatuses (data not shown). The grinding of sediments before thermal treatment may enhance As release from sediments due to higher surface area of sediments exposed to thermal treatment. The particle size distribution of the ground sediments is provided in SI Table S1. To compare the As release into porewater from ground sediments versus intact core sediments, two intact core sediments (S6 and S7) chosen based on higher total As (20− 28 μg g−1) were moistened with synthetic aquifer water. Before moistening the intact core sediments for thermal treatment, sediment cores were wrapped with a parafilm (1337412, Fisher Scientific) to keep the core intact covering top and bottom of the cores with filter papers (Whatman No. 40, Fisher Scientific) to facilitate saturation of the cores when partially submerged overnight in synthetic aquifer water. After saturation that brought the core at ∼25% moisture content, the sediments were thermally treated. Because liquid phase could not be separated for As analysis after thermal treatment, a sequential extraction was applied on these sediments for As fractionation to compare the effect of thermal treatment on As fractionation and phase transfer in ground versus intact core sediments. Sequential Extraction of Arsenic from Sediments. To determine the effect of heat on As distribution in different inorganic and organic phases such as soluble, exchangeable, bound to carbonates, and associated with amorphous and crystalline Fe and Mn oxyhydroxides, sulfides and organic matter in the sediments, a sequential extraction procedure (SEP; Javed et al.16 SI Table S2) was performed for As fractionation before and after the thermal treatment of the sediments. Briefly, sediments (0.4 g) were sequentially extracted with 40 mL of extractants for appropriate period of time as given in SI Table S2. The sediment-extractant suspensions were shaken for the specified duration, and then centrifuged for 40 min at 6000g to separate the supernatants which were decanted without removing any sediment particles.
■
RESULTS Thermal Effect on Arsenic Release from Sediments into Porewater. Seven sediments (S1−S7) collected from 2193
DOI: 10.1021/acs.est.5b04555 Environ. Sci. Technol. 2016, 50, 2191−2199
Article
Environmental Science & Technology
As while less As was released from Empress (S1) and Marie Creek (S2) sediments (Figure 1A). The sediments moistened with deionized water also exhibited a similar trend of As release after thermal treatment (Figure 1B) where maximum As (∼4500−6600 μg L−1) was released from the Bonnyville and Ethel Lake sediments and less As (∼1100−1800 μg L−1) released from Empress (S1) and Marie Creek (S2) sediments. Deionized water was more aggressive even at lower water to sediment ratio (1.4:1). Overall, the extent of As release into porewater was significantly higher where deionized water was used to moisten the sediments compared to the synthetic aquifer water. Thermal Effect on Arsenic Fractionation in Sediments. To investigate which As host phases (fractions) of the sediments contributed to As release into the porewater, only one sediment (S7) containing high As was moistened at 1.4:1 and 3.7:1 with synthetic aquifer or deionized water to solid ratios and subjected to sequential extraction protocol after thermal treatment. The results showed that exchangeable (F2) and specifically adsorbed (F3) As phases in the sediment significantly (P < 0.05) released ∼3 μg g−1 and ∼4 μg g−1 As into synthetic aquifer and deionized porewater, respectively after the thermal treatment (Figure 2A and B) that constituted 89−100% of the released As into the porewater. The association of As with different solid constituents such as crystalline Fe oxides, silicates, sulfides, and organic matter did not change significantly (P > 0.05) by the thermal treatment. The sediments (S1−S7) moistened at 15 and 50% water contents did not yield any aqueous phase for As analysis, therefore, they were analyzed for As fractionation after thermal treatment. The rationale was to observe changes in the As host phases in sediments after thermal treatment which might potentially contaminate groundwater if thermally affected sediments come in contact with water. The results of As fractionation at 15 and 50% water regimes (SI Tables S3−S6) were averaged due to their similarity (P > 0.05) to simplify the description of the results (Figure 3). The percentage distribution and actual concentration of As in different fractions are shown in Figure 3 and SI Figure S2, respectively. The As fractionation in sediments before thermal treatment (without adding any water) showed that the As was initially partitioned among sulfides (F8, 10−30% of total As), specifically adsorbed on mineral surfaces (F3, 14−29% of total As), crystalline and amorphous Fe oxides (F5 and F6, 10−33% and 6−22% of total As, respectively), silicate minerals (F7, 6−25% of total As) and exchangeable phase (F2, 5−14% of total As) in the sediments (Figure 3). Soluble As (F1) in the sediments along with the As associated with carbonate (F4) and organic matter (F9) were negligible (
