Keywords

11.1 Introduction

Over the years, diverse remediation technologies and approaches have been developed to remediate petroleum hydrocarbon (PHC)-contaminated sites, often targeting benzene, toluene, ethylbenzene, and xylene (BTEX) compounds due to their toxicity and physico-chemical properties. The knowledge and the experience gained were applied to remediation technologies increasingly more effective and for which diverse practical guidance documents were issued to assist field application (USACE 2014; NAVFAC 2013; EPA 2017). With respect to known BTEX chemical and physical properties and site-specific conditions, these remediation technologies are implemented, individually or combined, to stimulate either destructive mass removal processes (e.g., biodegradation or chemical oxidation) or non-destructive mass removal processes (e.g., volatilization). Regardless of the selected remediation technology, the fate of BTEX compounds and treatment cost-effectiveness are nonetheless closely related to successful field implementation of the selected remedial technology. Proper establishment of the intended mass removal process thus requires specific attention to ensure optimal performance of the remediation technology. However, although a remediation technology can be designed to stimulate a specific mass removal process, additional mass removal process(es) may co-occur, often inevitably. To demonstrate that the intended mass removal process is occurring as initially designed, advanced characterization tools capable of discriminating co-occurring processes are required.

CSIA is a process-based advanced characterization tool that tracks the stable isotope composition of selected volatile organic compounds (VOCs) present in groundwater. In essence, the change in stable isotope composition of selected VOCs is monitored during the treatment period to reveal valuable information about the attenuation process that concentration analysis cannot provide. Application of CSIA as a characterization tool has been studied for more than two decades, the subject of numerous scientific review publications (Kuntze et al. 2019; Thullner et al. 2012; Vogt et al. 2016; Elsner 2010), described in textbooks (Aelion et al. 2010; Jochmann and Schmidt 2012), and summarized by environmental agencies (EPA 2008) and coalition groups (ITRC 2011) for field practitioners. While most of the work cited above concerns CSIA application in a context of natural attenuation and biodegradation of contaminants, it is only recently that exhaustive development efforts have been directed toward engineered in situ remediation (Bouchard et al. 2018a). With CSIA being able to discriminate between co-occurring mass removal processes, the application of CSIA in engineered in situ remediation thus represents a great asset in assessing treatment performance. By demonstrating the occurrence of the intended mass removal process for newly implemented treatment or following modification actions, CSIA supports decision-making for timely system optimization, resulting in remediation cost efficiency. In this chapter, a methodology to deploy CSIA specifically to assess BTEX remediation treatment performance is presented with focus on groundwater assessments. In the attempt to remain concise and field application oriented, this chapter does not aim to provide a complete review of theoretical knowledge behind CSIA application or a thorough discussion of all parameters commonly included in exhaustive performance monitoring [as for instance suggested in technical guidance documents such as EPA (2017)], but rather focuses on key concepts allowing field practitioners to confidently utilize CSIA at their field sites.

11.2 CSIA Principles

11.2.1 Background and Concepts

Stable carbon (C) and hydrogen (H) atoms present in nature are commonly considered to have an atomic weight of 12 g and 1 g per mole and are respectively denoted as 12C and 1H. However, some naturally occurring stable C and H atoms have an additional neutron in their nucleus. The additional neutron increases the molecular weight to 13 g and 2 g per mole, respectively denoted as 13C and 2H. These 13C and 2H atoms are called heavy isotopes and represent a proportion of 1.07% and 0.0115% of each of the C and H atom pool, respectively (IUPAC 1991). Since PHCs are organic molecules uniquely composed of C and H atoms, it is expected to observe inclusion of a heavy atom in some molecules. Isotopically different benzene molecules, i.e., lightmolecules and heavymolecules, are depicted in Fig. 11.1. The 13C and 2H atom can be found at various positions within the heavymolecules. Due to the naturally low abundance of 13C and 2H atoms, it is generally assumed that PHCs such as BTEX compounds will include only one heavy isotope per molecule.

Fig. 11.1
Three benzene molecules, with the second and third having isotopes 13 and 2, respectively. A table has 4 columns and 3 rows. The rows list the chemical formula, molecular weight, and degradation rate for 1 light and 2 heavy molecules.

Illustration of isotopically different benzene molecules with selected characteristics. The heavy isotope in molecules 2 and 3 can occupy other positions than those indicated. See text for additional specificities regarding degradation rate and heavy isotope positioning in the molecule

Full size image

The presence of a heavy isotope confers to molecules slightly different physical properties. One key characteristic is that chemical bonds between two light isotopes are more rapidly broken compared to bonds between a light and a heavy isotope. More specifically, molecules with a heavy isotope present at the reactive position react more slowly than molecules containing only light isotopes (Elsner et al. 2005). This difference in isotopic composition hence causes slower degradation rates for heavymolecules compared to lightmolecules (Fig. 11.1). As a consequence, accumulation of heavymolecules in the remaining contaminant pool is expected over time, whether the reaction taking place is biotic (biodegradation) or abiotic (chemical oxidation or reduction) (Aelion et al. 2010; Elsner et al. 2005). For simplicity, degradation will further be used throughout the chapter as a general term referring to both biotic and abiotic reactions. This gradual accumulation of heavymolecules in the remaining fraction over time, reflected by isotope ratio changes with time, becomes unequivocal evidence of compound destruction. It is specifically the tracking of these isotope ratio changes that defines the essence of the CSIA method, which has proved to be a reliable means for assessing the fate of organic contaminants released in the environment.

The progressive change in isotope composition observed in the remaining contaminant pool is related to the change in contaminant concentration by the Rayleigh equation (Eq. 11.1) (Mariotti et al. 1981):

$$R_{t} = R_{0} f^{\alpha - 1}$$
(11.1)

where Rt and R0 are respectively the stable isotope ratio (13C/12C or 2H/1H) of the compound at time t or at initial time 0, f is the remaining contaminant fraction expressed by the ratio of concentrations measured at time t and at initial time (i.e., Ct/C0), and α is the isotope fractionation factor. The isotope ratios (Rt and R0) and VOC concentrations are obtained via analytical measurements carried out in the laboratory on groundwater samples collected on site (see Sect. 11.2.2). In contrast, α is a pre-determined parameter normally determined by experiments conducted in the laboratory. By definition, the parameter α quantifies the difference between the heavymolecules and lightmolecules degradation rates, and can be expressed by Eq. 11.2:

$$\alpha = \frac{{{}^{{{\text{heavy}}}}k}}{{{}^{{{\text{light}}}}k}}$$
(11.2)

where heavyk and lightk are the degradation rates of heavymolecules and lightmolecules, respectively (Fig. 11.1). As k is slower for heavymolecules compared to lightmolecules, the α coefficient is typically smaller than 1. The greater the difference between heavyk and lightk, the faster the isotope fraction occurs during the reaction and the greater the heavymolecule accumulates in the remaining contaminant pool. For convenience, α is often transformed into an enrichment factor (ε) using Eq. 11.3:

$$\varepsilon = \left( {\alpha - 1} \right)$$
(11.3)

The ε is commonly multiplied by 1000 to avoid working with small fraction numbers, and consequently be expressed in unit of per mil (‰). The ε transformation and terminology will be used in this chapter.

11.2.2 Isotope Analysis and Delta Notation

Analytical measurements are conducted to determine the 13C/12C and 2H/1H isotope ratios for each compound of interest included in a mixture. The initial analytical procedure step consists of compound separation via a gas chromatography unit. The compounds of interest are then conveyed to a combustion oven (for 13C/12C isotope ratio) or to a pyrolysis oven (for 2H/1H isotope ratio) prior reaching the isotope ratio mass spectroscopy unit (IRMS). To insure accurate and precise measurement of the isotope ratio, a nominal mass of 1 nmol of C or 8 nmol of H for each compound of interest needs to reach the IRMS source detector (EPA 2008). To meet this mass requirement for environmental samples with low VOC concentrations, a common strategy is to combine a preconcentration device integrated in the analytical setup and processing a large sample volume.

By convention, measurements of isotope ratios are reported using the δ notation, and expressed relatively to the Vienna Pee Dee Belemnite (VPDB) or the Vienna Standard Mean Ocean Water (VSMOW) international reference standards for carbon and hydrogen, respectively (Eq. 11.4) (Clark and Fritz 1997):

$$\delta = \left( {\frac{R}{{R_{{{\text{std}}}} }} - 1} \right)$$
(11.4)

where R and Rstd are the isotope ratio of the sample and the international reference standard, respectively. The δ value obtained is then multiplied by 1000 to conveniently express the results in units of per mil (‰) or in milli-Urey (mUr) as used in some recent publications. The former unit of ‰ is used in this chapter.

11.2.3 Isotope Fractionation Processes and Quantification

As introduced earlier, bond-breaking reactions involving a heavy isotope located at the reactive position of the primary enzymatic attach will proceed at a slower rate, which leads to an observable isotope fractionation in the remaining substrate pool. However, bond-breaking reactions can be induced by a variety of biotic and abiotic degradation processes. As each degradation process can be carried out by different bond-breaking mechanisms, the extent of isotope fractionation will vary. In addition, the type of chemical bond uniting the atoms, dilution of the heavy atom inside increasing molecular structure, the commitment to catalysis (i.e., masking isotope effect due to several sequential reaction steps inside the cell) (Gandour and Schowen 1978; Huskey 1991; Northrop 1981), and mass transfer limitation (Thullner et al. 2013) are also key factors influencing the extent of isotope fractionation. For these reasons, the ε values are compound-specific and process-specific (i.e., reaction mechanism-specific together with bacteria and enzyme-specific in the case of biodegradation). Isotope fractionation and corresponding ε values related to biological and chemical processes are briefly overviewed below. Detailed explanations on isotope fractionation variations are beyond the scope of this chapter and can be found in previous publications (Elsner et al. 2005; Aelion et al. 2010).

Isotope fractionation has also been assessed for non-bond-breaking processes (physical mass removal processes) such as volatilization from non-aqueous phase liquid (NAPL) or water (Julien et al. 2015; Kuder et al. 2009), sorption to organic matter (Imfeld et al. 2014), or aqueous and gaseous diffusion (Wanner and Hunkeler 2019; Bouchard et al. 2008a). However, when assessing the performance of in situ chemical oxidation (ISCO) or enhanced biodegradation treatment in saturated zones, physical mass removal processes are less likely to have a significant contribution and thus not considered in this chapter. For a field study assessing the performance of an air sparging system using CSIA (hence including physical mass removal processes), the reader is referred to Bouchard et al. (2018b). For further details on the fundamentals, design, and implementation of bioremediation and ISCO technologies, the reader is referred to Chaps. 14 and 15, respectively.

Laboratory-based determination of enrichment factors is commonly carried out using closed experimental setups, which ensure that solely the destructive process (biodegradation or chemical oxidation) is causing the concentration decrease. While single-isotope experiments used to be carried out, insight gained when combining measurements of two isotopes (for instance δ13C and δ2H) during the same experiment is substantial. Dual-isotope assessments allow coupling of δ13C changes with δ2H changes, hence revealing a specific isotope enrichment pattern associated to the destructive mechanism (Zwank et al. 2005; Kuder et al. 2005; Fischer et al. 2007). The combined δ13C and δ2H change can be quantified by the Λ value. The Λ value is equivalent to the slope of the linear regression obtained when plotting δ2H changes as function of δ13C changes, and approximated by Eq. 11.5 (Hӧhener and Imfeld 2021):

$$\Lambda = \frac{{\ln \left[ {\left( {{\updelta }^{2} {\text{H}}/1000 + 1} \right)/\left( {{\updelta }^{2} {\text{H}}_{0} /1000 + 1} \right)} \right]}}{{\ln \left[ {\left( {{\updelta }^{13} {\text{C}}/1000 + 1} \right)/\left( {{\updelta }^{13} {\text{C}}_{0} /1000 + 1} \right)} \right]}} \approx \frac{{\Delta {\updelta }^{2} {\text{H}}}}{{\Delta {\updelta }^{13} {\text{C}}}} \approx \frac{{\varepsilon_{\text{H}} }}{{\varepsilon_{\text{C}} }}$$
(11.5)

where Δ (in ‰) is the difference between two δ measurements. The two approximate transformations diverge from the exact definition with increasing \(\varepsilon_{{\text{H}}} /\varepsilon_{{\text{C}}}\), causing for instance an overestimated Λ by 5% when \(\varepsilon_{\text{H}} /\varepsilon_{\text{C}}\) = 22 (Hӧhener and Imfeld 2021).

11.2.3.1 Biological Process

As discussed in Chaps. 5, 9, and 14, biodegradation of PHCs and BTEX in particular may occur through a wide variety of bond-breaking mechanisms, which differ whether the process is taking place under oxic or anoxic conditions, and also accordingly to which enzymatic pathway is being used. As a consequence, these specificities will generate different magnitudes of carbon and hydrogen fractionation, and when both isotopes are combined, will exhibit specific carbon and hydrogen enrichment patterns. Numerous laboratory studies were conducted to determine ε-C and ε-H values for BTEX biodegrading under oxic or various anoxic conditions and revealed these differences in enrichment patterns. A selection of literature ε-C and ε-H values is provided in Table 11.1 for aerobic and anaerobic BTEX biodegradation. For each BTEX compound, the largest and the smallest ε-C values were selected and are reported in Table 11.1 whether or not δ2H changes were considered during the experiment and regardless of the prevailing electron acceptor during the anaerobic biodegradation. The latter ε sub-set will be used to interpret single-isotope assessments further described in Sect. 11.4.2.1. In addition, experiments showing the largest and the smallest Λ value (reported by the respective study or calculated using Eq. 11.5 with ε-C and ε-H issued from the same experimental series) and regardless of the electron acceptor for anoxic conditions are also reported in Table 11.1. This second ε-C and ε-H value sub-set (or Λ) will be used to interpret dual-isotope assessments further described in Sect. 11.4.2.2. Note that among the published studies, the isotopic evaluations were carried out either using a pure microbial strain, a microbial consortium, or the native microbial population. The data selection presented in Table 11.1 was made regardless of these specificities to identify and use the widest possible ranges. Finally, because only few Λ values are reported in the literature for xylene, Table 11.1 reports the Λ values related to biodegradation regardless of the isomers (o, m, or p). A recent compilation of ε-C and ε-H with corresponding Λ values can be found in Vogt et al. (2016) reporting Λ values by enzymatic reaction.

Table 11.1 Carbon (ε-C) and hydrogen (ε-H) enrichment factors and related Λ values measured for BTEX compounds related to biodegradation (for aerobic and anaerobic) and chemical oxidation (for various oxidizing agents and activation method) processes
Full size table

11.2.3.2 Chemical Process

Chemical oxidation of VOCs initiated by oxidizing agents involves bond-breaking reactions. An oxidation mechanism pathway is likely specific to each oxidizing agent, which can furthermore depend on the activation method producing diverse radicals (Zhang et al. 2016; Matzek and Carter 2016). In addition, the nature of radicals formed can change due to contact with other species naturally present in groundwater such as background organic matter, halide, phosphate, and carbonate (Lee et al. 2020; Li et al. 2017). Therefore, the chemical oxidation process will lead to 13C and 2H enrichment in the remaining VOC pool, and the extent of isotope fractionation will depend on the selected oxidizing agent, activation method used, and prevailing radicals. Since the ε value database is yet restrained for chemical oxidation reactions, Table 11.1 lists the different ε-C and ε-H values available for BTEX when chemically oxidized by persulfate, hydrogen peroxide, or permanganate under various activation methods. When possible, Λ values (reported by the respective study or calculated using Eq. 11.5 with ε-C and ε-H values issued from the same experimental series) are also reported in Table 11.1. Nevertheless, given the limited knowledge regarding isotope fractionation generated by chemical oxidation, it is highly recommended to carry out laboratory experiments with aquifer material at the prevailing redox conditions to derive site-specific ε-C and ε-H values.

11.3 CSIA Implementation for Field Site Evaluation

With the understanding of the processes leading to isotope fractionation for BTEX compounds, the next step is to take advantage of this knowledge to evaluate the performance of a remediation treatment at the field-scale level. Field implementation and sampling strategy to evaluate the performance of a remediation treatment using CSIA should be built given the objective of the remediation treatment, site-specific conditions, but also to CSIA specificities. This section aims to provide general and specific considerations to bear in mind when establishing a CSIA-based remediation performance monitoring program, with the focus being given to groundwater assessments.

11.3.1 Approach and Sampling Strategy Considerations

The general application goal for CSIA is to document an isotopic shift for BTEX compounds caused by the treatment being applied. For this purpose, a baseline characterization is essential to document pre-treatment isotopic compositions, to which post-treatment results will be compared to. In addition, to ensure capturing the most valuable and timely information leading to a successful application, the following specific considerations should be addressed:

  1. (i)

    Regardless of the remediation treatment to be deployed, a CSIA baseline characterization should be considered for each sampling location to be included in the post-treatment sampling program.

  2. (ii)

    The number of sampling locations included in the sampling program should be sufficient to meet the objectives and be appropriate for the size of the treated area.

  3. (iii)

    The use of injection wells as monitoring wells is discouraged by technical performance monitoring guidance documents as injection of large fluid volumes may create water displacement leading to momentarily VOC concentration decrease (Payne et al. 2008). Since dilution does not create isotope fractionation, CSIA is actually the right tool to assess such potential VOC displacement. Nevertheless, field practitioners should keep in mind that unrepresentative excessive and prolonged reactions occurring inside (or near) the injection well might not be representative of distant areas (Huling and Pivetz 2006).

  4. (iv)

    Following the CSIA baseline characterization, at least two or three post-treatment sampling events should be considered. Since it can be challenging to predict exactly when maximum degradation of VOC mass would be observed given potential presence of VOC concentration rebounds, planning several post-treatment events is a judicious strategy.

  5. (v)

    Post-sampling event periodicity can be estimated using the expected lifetime of the emplaced amendment and estimated groundwater flow velocity. Chemical oxidizing agents with short half-lives should have shorter intervals between sampling events (few weeks), whereas oxygen-releasing products designed to last longer could have extended intervals between sampling events (few months). The groundwater flow velocity is also to be considered for long lasting amendments, or when forced by an engineered system.

  6. (vi)

    BTEX concentration should be considered when pre-selecting the sampling locations. Following the treatment, the remaining BTEX concentration should be above the isotope analytical detection limits (see Sect. 11.2.2). In contrast, if BTEX concentration is initially too high (strongly suggesting presence of NAPL), the BTEX mass decrease due to treatment might not be sufficiently strong to counteract BTEX mass input by NAPL dissolution, which would impede observation of an isotope enrichment.

11.3.2 CSIA Sampling Requirements and Procedures

Groundwater sampling for CSIA is normally performed using the same sampling procedure, vial sizes (typically 40 mL VOA glass vials), and preservative agents (to inhibit microbial activity) as for the common VOC concentration analysis. In cases where BTEX concentrations are expected to be low, larger glass containers can be considered to allow the laboratory reaching better analytical detection limits. However, when an ISCO treatment is performed, a quenching agent should be added to the samples to consume the remaining oxidant, hence avoiding post-sampling oxidizing reaction with BTEX (EPA 2012). This specific quenching procedure or vial size selection should be discussed with the selected CSIA laboratory prior to the sampling event. Upon sampling completion, the vials can be stored, handled, and shipped to the CSIA laboratory in a standard cooler using the same procedures for samples dedicated for VOC analysis. Finally, BTEX concentrations are required to perform the CSIA measurement in the laboratory, and thus samples for BTEX concentration analysis should always be collected and analyzed during the same sampling event.

11.4 CSIA Field Data

This section introduces assessment approaches and interpretation of CSIA field data. First, general and specific considerations for sound CSIA interpretation are discussed, either introduced by recalling the state of knowledge, factors affecting the isotope fractionation, or by identifying potential pitfalls that should be avoided. Then, interpretation schemes making use of only one element (single-element isotope assessment) or two elements (dual-element isotope assessment) are introduced.

11.4.1 Interpretation Considerations and Pitfalls

The Rayleigh equation (Eq. 11.1) is the key mathematical feature to interpret isotope fractionation. This equation relates the isotope change to the decreasing VOC concentration. It is worth mentioning that the equation was originally developed to describe the isotope fractionation occurring in a system where a single process controls the change in concentration. However, for complex systems like contaminated aquifers, one can foresee potential for VOC concentration decrease due to several co-occurring processes (for instance biodegradation, dispersion, dilution, diffusion, and sorption). While some processes such as dilution and dispersion do not cause fractionation, other processes may result in some isotope fractionation, hence calling for cautions when evaluating field site data. As biodegradation was more often shown to be the dominant controlling isotope fractionation process in natural systems, application of Rayleigh equation to assess contaminant fate in groundwater is generally accepted (EPA 2008). Nonetheless, some studies are underlying the need for specific application limitations for complex and heterogeneous aquifer systems, especially when performing quantitative assessments (Thullner et al. 2012). In the present work, degradation is assumed to be the controlling isotope fractionation process in groundwater, and that sorption and aqueous diffusion are processes not likely to cause significant isotope fractionation. In cases where remediation systems are emphasizing physical processes, such as air sparging or soil vapor extraction systems, or being conducted in an organic rich aquifer, considering isotope fractionation caused by physical processes may be indicated.

Notwithstanding the Raleigh equation application concept and restrictions, there are notable additional interpretation considerations and pitfalls that deserve to be considered, and should be kept in mind when interpreting CSIA field data set:

  1. (i)

    Due to uncertainty errors related to field and analytical procedures, an isotope shift of at least 1‰ or 10‰, respectively for δ13C and for δ2H, respectively, should be observed to suggest presence of compound degradation, whereas an isotope shift of 2‰ or 20‰ can be considered as strong evidence of degradation activity (EPA 2008). It should nevertheless be kept in mind that, especially for aerobic biodegradation of BTEX, large extent of biodegradation is required in some cases before observing carbon isotope shift larger than 2‰. Accordingly, observed field shifts smaller than 2‰ are not necessarily indicating an absence of biodegradation activity (see Sect. 11.4.2.2 and explanation of Fig. 11.2), but perhaps a dominant biodegradation mechanism generating insufficient isotope fractionation. This limitation hence underlines the importance of a multi-line of evidence approach.

    Fig. 11.2
    A line graph highlights anaerobic and aerobic benzene degradation. Two lines with lambda values 8.7 and 8 divide the increasing extent of biodegradation.

    Dual-carbon and hydrogen isotope plot illustrating expected isotope trends (delineated areas) for aerobic and anaerobic benzene biodegradation. Note that both areas are slightly overlapping. The Λ values were selected from Table 11.1. The grey dashed box represents no significant isotope change between two measurements whereas the grey dotted box is only suggestive of biodegradation effect. For each biodegradation delineated area, the magnitude of isotope fractionation expected for increasing extent of benzene biodegradation (from 50 to 99%) is indicated

    Full size image
  2. (ii)

    Interpretation of CSIA data should be carried out taking into account VOC concentration data and conventional field parameters (geochemical parameters and terminal electron acceptors) to validate the information supported by CSIA. As discussed in Chap. 10, additional specialized molecular biological tools (such as qPCR analyses or 16S rRNA gene amplicon sequencing to determine the structure of the microbial community) can also be included as an additional line of evidence complementary to CSIA.

  3. (iii)

    The initial isotope composition (or source signature) of each contaminant before the occurrence of the spill is rarely known. To overcome this data gap, it is a common strategy to use the most negative δ13C and δ2H value measured at the site as the source value, for each targeted VOC.

  4. (iv)

    All suspected source zones on the site must be sampled to determine the likelihood of multiple sources (i.e., two distinct spills). Different spills would most likely have distinct BTEX isotope compositions. While in this case CSIA indicates the presence of multiple sources, the assessment of BTEX biodegradation will become challenging if both plume zones are overlapping.

  5. (v)

    A survey of carbon isotope composition of BTEX compounds collected from different gasoline sources showed variations ranging from  −23.5 to −31.5‰ for benzene, from −22.9 to −30.4‰ for toluene, from −22.9 to −31.1‰ for ethylbenzene, and from −22.4 to −30.6‰ for xylene (regardless of the isomer) (O'Sullivan and Kalin 2008). Accordingly, field measurements providing more positive δ13C values than the higher-end value range would suggest that a degradation process has already affected the compounds.

  6. (vi)

    Applying CSIA inside a source zone where NAPL is present may fail to demonstrate isotope fractionation although degradation may be occurring in the aquifer. Due to the limited biodegradation inside NAPL (because of toxicity) and negligible isotope fractionation related to dissolution process, the dissolution of VOCs from NAPL will consistently bring in groundwater VOCs with the original isotope signature. This process will thus influence the isotope ratio of dissolved VOC in groundwater. In some cases, the mass of freshly dissolved VOC can dominate the overall δ13C (and δ2H) signature in groundwater, hence masking the isotope fractionation caused by the degradation process intended by the treatment. While this dissolution process impedes degradation demonstration, it however validates the presence of NAPL in the vicinity of the sampling location.

  7. (vii)

    Given the limitation discussed in point vi above, application of the Rayleigh equation to quantify biodegradation extent or rate in the source zone should be done with precautions. While Rayleigh’s law assumes a finite dissolved VOC mass, introduction of additional mass of VOC into groundwater (caused by NAPL dissolution and VOC desorption) will cause an underestimation of the degraded mass (see further discussion in Sect. 11.4.2.1).

  8. (viii)

    Sampling a monitoring well with the screen positioned across the water table may result in groundwater samples mixing oxic water (upper part of the aquifer with BTEX undergoing aerobic biodegradation) with anoxic water (lower part of the aquifer with BTEX undergoing anaerobic biodegradation). Such mixing of reduction–oxidation conditions may affect evaluation of the dominant biodegradation mechanism controlling the isotope fractionation.

11.4.2 Assessment Approach

Two CSIA assessment approaches are discussed below, using either one or two isotopes. While a single-isotope assessment approach is meant to discriminate VOC concentration decrease due to degradation from dilution, the dual-isotope assessment approach has the advantage to provide more specific information by identifying the dominant VOC degradation mechanism. Both assessment approaches are described below by proposing CSIA data sorting and interpretation schemes to gain the most insight from the available information.

11.4.2.1 Single-Isotope Assessment Approach

The use of a single isotope (either δ13C or δ2H) will provide a direct line of evidence for BTEX mass destruction when dilution is potentially co-occurring during the treatment. Depending on site-specific assessment objectives, this can turn out to be sufficient to demonstrate treatment success. For such single-isotope assessment, comparison of isotopic shift between baseline and collected samples post-treatment for each BTEX compound is the basic interpretation scheme. The significance of the isotope shift should be interpreted based on threshold values as introduced above for δ13C or δ2H measurements. The larger the isotope shift, the more likely the treatment was effective for BTEX compounds. Nevertheless, the isotopic shift to be observed yet depends on the ε value, the degradation rate constant and, if the assessment compares the injection well to downgradient wells, groundwater velocity.

The interpretation can further be developed to evaluate the VOC mass proportion affected by the treatment and the degradation rate. The decrease in VOC concentration observed between two distal monitoring wells is usually due to a combination of destruction and dilution processes, which is expressed by Eq. 11.6 (Van Breukelen 2007b; Aelion et al. 2010):

$$f_{{{\text{overall}}}} = f_{{{\text{deg}}}} *f_{{{\text{dil}}}}$$
(11.6)

where the overall remaining mass fractionation (foverall) is the product of remaining mass fractionation related to degradation (fdeg) and dilution (fdil). The fdil considers all physical processes, such as dispersion and sorption. It is worth recalling that fdeg, calculated based on the Rayleigh equation, only involves the mass of contaminant dissolved in groundwater. Since both fdeg and fdil are reducing the VOC concentration, coefficients will be < 1.

When considering a monitoring well located in the source zone with presence of NAPL, the process of VOC dissolution will counteract both degradation and dilution processes. Accordingly, to evaluate the change in VOC concentration in groundwater at a monitoring well periodically sampled over time, Eq. 11.6 is amended to become Eq. 11.7:

$$f_{{{\text{overall}}}} = f_{{{\text{deg}}}} *f_{{{\text{dil}}}} *f_{{{\text{diss}}}}$$
(11.7)

where fdiss is the mass fraction input caused by VOC dissolution in groundwater. Since fdiss is a mass input process, the coefficient will be > 1. When groundwater flow velocity is negligible within the sampling periodicity, one can expect to have fdiss ≫ fdil, hence making fdil negligible.

For both Eqs. 11.6 and 11.7, the foverall can be calculated using concentration data measured either at the same well but at different sampling time (C0/Ct) of for two distal wells aligned with respect to groundwater flow (C0/Cx). As shown in Eq. 11.8, for fdeg, the latter can be determined using the isotope measurements and a rearrangement of Eq. 11.1:

$$f_{{{\text{deg}}}} = {\text{exp}}^{{\left( {\updelta ^{13} {{C}}_{{{t}}} -\updelta ^{13} {{C}}_{0} } \right) /\varepsilon }}$$
(11.8)

where δ13Ct and δ13C0 are isotopic field measurements obtained at time = t and time = 0, respectively. Ideally, microcosm experiments are carried out with aquifer material at the prevailing redox conditions to determine site-specific ε values. The alternative option is to select an appropriate ε value from the literature. Since there are many different ε values related to BTEX biodegradation (as discussed in Sect. 11.2.3) the selection must be carefully made considering the specific conditions prevailing at the site. Conventional field parameters or specialized molecular biological tools (such as microbial genetics) carried out in parallel can assist the field practitioner in the selection. Nevertheless, it is good practice to calculate a range of estimated fdeg by using the largest and the smallest ε values available for each VOC (Table 11.1) to support interpretation or decision-making. The largest ε value will provide a more conservative estimate, hence reducing the risk of overestimating fdeg. If Eq. 11.7 applies to the field site conditions, the VOC mass input process (fdiss) will additionally bias the calculations, also leading to an underestimated degraded mass estimate.

Assuming a constant groundwater flow velocity and a constant first-order degradation rate in the aquifer, Eq. 11.1 can also be transformed to obtain an in situ degradation rate (k) for the targeted contaminant. Assuming a first-order degradation rate controlling the evolution of the remaining mass fraction as expressed by Eq. 11.9 (Aelion et al. 2010):

$$f_{{{\text{deg}}}} = {\text{exp}}^{{\left( { - k*t} \right)}}$$
(11.9)

Insertion of Eq. 11.9 into Eq. 11.8 provides Eq. 11.10:

$$k = - 1*\left( {\frac{{\updelta ^{13} {{C}}_{{{t}}} -\updelta ^{13} {{C}}_{0} }}{\varepsilon *t}} \right)$$
(11.10)

where k is the in situ degradation rate (day−1) and t is the time period (in days) between two sampling events within the same monitoring well. Equation 11.10 becomes applicable for evaluation using a single well over time as the injected product degrades VOCs faster than source input, which momentarily interrupts the steady state conditions. If Eq. 11.7 applies due to presence of NAPL in the vicinity of the monitoring well, the k value will be underestimated due to fdiss. To calculate k based on measurement from two distal monitoring wells (and measured during the same sampling event), the parameter t can be replaced by the associated site-specific distance (x) and groundwater velocity (v) to provide Eq. 11.11 (Aelion et al. 2010):

$$k = - 1*\left( {\frac{{\updelta ^{13} {{C}}_{{{x}}} -\updelta ^{13} {{C}}_{0} }}{{\varepsilon *\left( {x/v} \right)}}} \right)$$
(11.11)

Equation 11.11 assumes the injected product has reached the distal monitoring well due to groundwater flow (i.e., not due to large radius of influence related to injection activity) and that steady state conditions are re-established. The calculated k value can then be used to estimate the length of the treatment area required to meet the groundwater criteria for BTEX concentrations. Since fdil is not considered in Eq. 11.11, the estimated length of the treatment area will likely be conservative. To increase their reliability, these equations should only be applied when δ13Ct or x−δ13C0 > 2‰ (or 20‰ for δ2H).

11.4.2.2 Dual-Isotope Approach

As described earlier, biological and chemical oxidation have specific δ13C and δ2H isotope fractionation processes which lead to different isotope enrichment patterns during the reaction. The use of dual isotopes (δ13C and δ2H) takes advantage of these specific isotope enrichment patterns associated with diverse degradation mechanisms. Therefore, a dual-isotope assessment not only demonstrates BTEX mass destruction, but also indicates the dominant degradation mechanism affecting BTEX mass while additional co-occurring process(es) may be present during the treatment. Compared to single-isotope assessments, dual-isotope assessments are more rigorous in demonstrating establishment of the intended mass removal process by the treatment.

For dual-isotope assessment, interpretation of CSIA field data is performed by comparing the trend observed with field measurements to the theoretical trends expected for biodegradation (aerobic and anaerobic) or chemical oxidation process (depending on the remediation treatment). The first step in the interpretation sequence is to establish a supportive δ13C versus δ2H reference plot illustrating theoretical enrichment trends for potentially occurring degradation processes. This is achieved by plotting calculated δ2H data in function of δ13C, using respective Λ value for each degradation process listed in Table 11.1. The resulting timeless isotopic enrichment trend represents the expected δ13C and δ2H evolution as the degradation process is progressively occurring. To account for the isotope fractionation variability caused by diverse biodegradation mechanism pathways (due to site-specific microbial population composition), the minimum and the maximum Λ values reported in Table 11.1 for each selected process should be used to calculate extremity enrichment trends. Both extremity enrichment trends form a process-specific delineated area. An illustrative example is provided in Fig. 11.2, where process-specific delineated areas were calculated for aerobic and anaerobic biodegradation of benzene using the approximate transformation provided in Eq. 11.5. Note the simplification made for anaerobic biodegradation, where the selection of extremity Λ values was made regardless of the terminal electron acceptor. Figure 11.2 also indicates, for each biodegradation delineated area, the isotope shift expected for increasing extent of benzene biodegradation. For the selected Λ values, one can observe that at least 50% of anaerobic benzene biodegradation is needed to observe significant isotope shifts (> 2‰ for δ13C or > 20‰ for δ2H), whereas larger extents are needed for aerobic biodegradation. The proportion of biodegraded benzene (b) was calculated by a simple rearrangement of Eq. 11.8, leading to Eq. 11.12:

$$\text{b} \left( \% \right) = \left( {1 - f_{{{\text{deg}}}} } \right)*100$$
(11.12)

As a second step, field measurements are introduced in the reference isotopic plot. For each selected compound from a specific sampling location, the carbon and hydrogen isotope shift (or ∆δ13C and ∆δ2H) between baseline and a post-injection event is calculated and the resulting data point (field measurement) is plotted in the reference isotopic plot. Additional points can be plotted for every post-treatment measurement conducted over time at the same monitoring well. The positioning of field measurements relative to process-specific delineated areas can then be evaluated. In addition, the significance of isotope shift should be interpreted based on threshold values (2‰ and 20‰) as introduced above for δ13C and δ2H measurements. The field measurements lying inside a process-specific delineated area suggest the related process as the dominant mass removing process. In some cases, field measurements may lie between two process-specific delineated areas, which would suggest a combination of the two processes. The evaluation can be deepened by calculating the contribution of each two different on-going processes using extended Rayleigh equations. These extended equations were previously developed to consider two concurrent degradation pathways with different ε values (van Breukelen 2007a).

11.5 Examples of Field Case Applications

Application of the CSIA method to evaluate the performance of different remediation approaches for BTEX-contaminated aquifers is presented below for two pilot scale tests. For each pilot test, CSIA was included to the remediation performance monitoring plan and used as a diagnostic tool to demonstrate that the BTEX mass degradation processes were established as initially intended.

11.5.1 In situ Chemical Oxidation Application

A pilot scale test was conducted at Site 1 to evaluate the potential of NaOH-activated sodium persulfate in reducing dissolved BTEX mass in a PHC source zone. CSIA was used as a performance metric to validate BTEX mass destruction although co-occurrence of dilution related to large volume fluid injections contributing to decrease BTEX concentration was expected. The discussion of the results will focus on δ13C values measured for toluene, ethylbenzene, m,p-xylene, and o-xylene (TEXX) results. Benzene concentrations were below the detection limit for isotope analysis.

11.5.1.1 Field Approach

The injection of persulfate was performed through an injection–extraction system using two wells. The injection and extraction well were located 12 m apart and aligned with groundwater flow. The groundwater recirculation system was maintained in operation for the duration of persulfate injection and was then halted to leave the persulfate lingering in the aquifer. To demonstrate the destructive effect of ISCO treatment on dissolved TEXX, groundwater was collected before and during the treatment from monitoring well MW-S1 located midway between the two recirculation wells. The sampling event timing and the parameters analyzed are summarized in Table 11.2.

Table 11.2 Technical description of the pilot scale test conducted at Site 1
Full size table

11.5.1.2 Results and Interpretation

For the four TEXX compounds in groundwater at MW-S1, a concentration decrease was observed during the post-injection event 1 (Post 1) compared to baseline level and was followed by an increase observed during the post-injection sampling event 2 (Post 2) (Fig. 11.3). In addition, baseline sulfate (SO4) concentration increased from 32 mg/L to 1100 mg/L (Post 1) to then decrease to 810 mg/L (Post 2) (Fig. 11.3). The presence of SO4 in groundwater is related to persulfate reaction as the former is a by-product of decaying persulfate. Similarly, the strong pH increase (from 7.2 to 12.5) also confirms the arrival of the NaOH-activated persulfate slug in the vicinity of the monitoring well. The sampling timing for Post 1 was established based on the persulfate decay rate previously determined during a bench test. The decay rate was used to ensure a sampling interval sufficiently long for persulfate to degrade a significant TEXX mass, but still with sufficient remaining persulfate restraining TEXX concentration to rebound. Based on TEXX and SO4 concentration patterns observed (Fig. 11.3), the timing of Post 1 was adequate to capture persulfate effects on TEXX mass.

Fig. 11.3
Three line graphs for changes in p H, S O 4, concentration, and delta super 13 C versus days. The lines for p H and S O 4 form an inverted v-shape. The lines for concentration form a v-shape; for delta 13 C, the lines depict an increasing trend.

Changes in pH, SO4, concentration and δ13C for toluene, ethylbenzene, m,p-xylene, and o-xylene observed in groundwater at MW-S1 for baseline (day 0) and after 26 days (Post 1) and 38 days (Post 2) following NaOH-activated persulfate injection at Site 1. The dashed lines in panel C represent the original isotopic signature determined for respective compound at the site

Full size image

Single-isotope assessment: the δ13C values measured for TEXX in groundwater at MW-S1 are presented in Fig. 11.3. The δ13C values measured for TEXX at Post 1 were all more positive compared to their respective baseline value. The calculated Δ δ13C for TEXX, respectively 2.2, 2.4, 3.2, and 2.1‰, were systematically larger than the prerequisite of 2‰ for strong evidence of degradation activity. These significant trends toward more positive values (i.e., enrichment of heavymolecules among the remaining TEXX pool), coinciding with the decline in TEXX concentrations, can be attributed to persulfate oxidation reactions. Accordingly, the δ13C values measured strongly support that the temporal decrease in concentration was not solely caused by dilution, but by substantial degradation reactions affecting the TEXX mass. For Post 2, TEXX concentrations were observed to significantly increase. Although no δ13C measurements were carried out, it is strongly expected that δ13C values for TEXX would return to their original values (i.e., dashed lines in Fig. 11.3). In this case, freshly dissolved VOC mass coming from the NAPL (hence with the original isotope composition), coinciding with persulfate mass exhaustion, is expected to overwhelm the overall δ13C signature in groundwater.

Remaining mass fraction: The isotopic evaluation can be further developed to evaluate the effect of persulfate treatment on TEXX mass. The remaining mass fraction foverall and fdeg were determined for toluene, ethylbenzene, and o-xylene using respective concentration and δ13C (Eq. 11.8) values measured at baseline and Post 1 (Fig. 11.3). For fdeg, the respective ε-C value for toluene, ethylbenzene, and o-xylene related to NaOH-activated persulfate oxidation was used (Table 11.1). The resulting calculated fdeg values for toluene, ethylbenzene, and o-xylene are smaller than those respectively obtained for foverall (Table 11.3). We recall here that calculation of foverall indistinctively considers the processes of destruction (fdeg), dilution (fdil), and dissolution (fdiss) (Eq. 11.7). However, in a source zone, VOC dissolution from NAPL can significantly contribute to increased dissolved concentrations, hence counteracting the former two processes. In addition, contribution of fdil may also be reduced using temporal data measured from the same monitoring well. Accordingly, the fdil * fdiss values > 1 obtained for toluene, ethylbenzene, and o-xylene (Table 11.3) strongly suggest the occurrence of NAPL dissolution during the treatment period counteracting the apparent VOC destruction process. Accordingly, the δ13C dataset reveal greater TEXX destruction by persulfate than concentration data alone would suggest. Due to the evidence of NAPL dissolution, calculation of degradation rate k (using Eq. 11.10) was not attempted as very likely to be underestimated.

Table 11.3 Remaining mass fraction foverall, fdeg, and fdil calculated for toluene, ethylbenzene, and o-xylene using concentration or carbon isotope data
Full size table

11.5.1.3 Benefits of Using CSIA

The pilot scale test demonstrated that base-activated persulfate was effectively oxidized TEXX compounds. Although such remediation approach required injection of large amount of fluid, dilution was apparently not the only process controlling the VOC concentration variations observed. The significant changes in δ13C to more positive values strongly support the occurrence of a TEXX mass destruction process. Furthermore, the remaining mass fraction calculations suggested that dissolution process (from residual NAPL) and oxidation process outweighed the dilution process as the calculated fdeg values were systematically smaller than foverall values. These findings contribute in supporting the decision to move forward with the full-scale application of the treatment inside this source zone area. Nevertheless, field practitioners should bear in mind that these isotopic calculations remain semi-quantitative, especially when treating a source zone. Since the process of VOC dissolution from NAPL represents a mass input to groundwater, the calculated mass proportion degraded is likely under predicted. In addition, the enrichment factor value used here was derived from a NaOH-activated laboratory experiment assuming BTEX degradation by SO4 radicals. Under field conditions, degradation via unactivated persulfate and/or via OH radicals may also have occurred. Since different enrichment factors (Table 11.1) are observed due to different degradation mechanisms, the calculated fdeg values would then change accordingly. It is re-emphasized here that determination of site-specific ε value for NaOH-activated persulfate (through a laboratory bench test with site soil and groundwater) will allow more accurate isotope assessments, especially if a full-scale treatment is considered.

11.5.2 Bioremediation Application

A pilot scale test was conducted at Site 2 to evaluate the potential of stabilized hydrogen peroxide (H2O2) to enhance and sustain aerobic biodegradation of dissolved BTEX mass present in an anoxic plume zone. Hydrogen peroxide was used as source of dissolved oxygen (DO), and was combined with nutrients, a petroleum-specific facultative microbial inoculant (bioaugmentation), and a stabilizer reducing the decomposition rate of H2O2 reactions (hence maintaining an extended DO source). CSIA was used in combination with microbial gene assays as performance metrics to (i) assess the aquifer turnaround from well-established anoxic biodegradation conditions to oxic conditions, (ii) assess aerobic BTEX biodegradation, and (iii) assess the co-occurrence of dilution related to large volume fluid injections for the H2O2 delivery. The discussion of the results is focusing on toluene (δ13C), ethylbenzene (δ13C and δ2H), and m,p-xylene (δ13C and δ2H). Benzene concentrations were below the detection limit for isotope analysis.

11.5.2.1 Field Approach

The injection of stabilized H2O2 solution was conducted through an injection–extraction system as described above for Site 1. To assess toluene destruction and to assess aerobic biodegradation of ethylbenzene and m,p-xylene, groundwater was collected before and during the treatment from monitoring well MW-S2 located midway between the two recirculation wells. The sampling event timing and the sampling parameters analyzed are summarized in Table 11.4.

Table 11.4 Technical description of pilot scale tests carried out at Site 2
Full size table

11.5.2.2 Results and Interpretation

Geochemical parameter results are shown in Fig. 11.4 for baseline, Post 1 and Post 2 sampling events at MW-S2. An aquifer under anoxic conditions was prevailing before the treatments, as suggested by a depleted DO level and strong negative ORP value (and by other geochemical parameter results not shown). Following the H2O2 solution injection, DO levels significantly increased up to 15 mg/L, and lasted above 2 mg/L for at least 40 days. Although high DO levels able to support oxic conditions were present, concentrations of Fe2+ and Mn2+ increased and concentration of SO4 decreased following the injections, suggesting that anaerobic biodegradation was likely co-occurring in the aquifer (data not shown).

Fig. 11.4
Four line graphs for changes in D O, O R P, concentrations, and delta super 13 C versus days. The lines for D O, O R P, and concentrations have a downward and upward decreasing trend whereas for delta 13 C, the lines are upward increasing.

Changes in dissolved oxygen (DO) and oxidation–reduction potential (ORP) (panel a), toluene, ethylbenzene and m, p-xylene concentration (panel b), gene copies of toluene dioxygenase (TDO) and Benzylsuccinate synthase (BssA) (panel c), and δ13C values (panel d) observed in groundwater at MW-S2 during the treatment. The dashed lines in panel d represent the original isotopic signature determined for respective compound at the site

Full size image

To assess the effect of oxygen addition on the microbial population dynamics, gene assays targeting toluene dioxygenase (TOD—indicative of aerobic BTE biodegradation) and benzylsuccinate synthase (BssA—indicative of anaerobic TEX biodegradation) were assessed (Fig. 11.4). The baseline conditions indicate low concentrations (gene copies/L) of both genes. After injection, a significant increase in TOD aerobic genes was observed only for the Post 2 sampling event, although high DO levels were already observed for the Post 1 sampling event. Also at Post 2 sampling event, the TOD gene concentration became larger than the BssA concentration by almost two orders of magnitude. These results suggest a slow transition to an aerobic population that has benefited from high DO concentration (Post 1 event) followed by depleted DO concentration (Post 2 event) in the aquifer. The BssA anaerobic population increased for the Post 1 event, benefiting first from the arrival of nutrients, but then decreased likely due to increasing oxic conditions.

Single-isotope assessment: Concentration and δ13C values measured for toluene, ethylbenzene, and m,p-xylene in groundwater at MW-S2 before and after the injection are presented in Fig. 11.4. The ethylbenzene concentration for the Post 1 sampling event was similar to the baseline condition, but then decreased at Post 2. The δ13C value followed the same trend; remaining unchanged between baseline and Post 1 (equivalent to the original signature determined for ethylbenzene at this site), to then becoming enriched (more positive value) by 2.0‰ at Post 2. The enrichment observed at Post 2 is equal to the prerequisite of 2‰ for significant evidence of (bio)degradation process. In contrast, the m,p-xylene concentration at Post 1 significantly increased compared to baseline condition and coincided with δ13C value changing to depleted value (more negative at Post 1). These changes for both parameters are indicative of m,p-xylene being released into solution (either due to desorption from the soil matrix and/or dissolution from some residual NAPL). The δ13C value measured at Post 1 was the most negative value measured, hence suggestive of the original carbon isotopic signature for m,p-xylene at this site. At Post 2, the m,p-xylene concentration decreased while the δ13C value changed to significantly enriched value (by 5.6‰). Accordingly, the isotopic shifts measured for both ethylbenzene and m,p-xylene between Post 1 and Post 2 (≥ 2‰) strongly support concentration decrease due to a destructive process. Lastly for toluene, while concentrations remained stable (and low) between baseline and Post 1, an increase was observed at Post 2. This increase in concentration coincided with a decreasing δ13C value (between Post 1 and Post 2), suggesting that the toluene mass input into groundwater was momentarily greater than the mass biodegraded. Nevertheless, δ13C values measured at Post 1 and Post 2 are more positive compared to the original signature (−29.1‰), hence indicating that toluene is nonetheless being biodegraded at the site. Note that toluene concentration at baseline was too low to conduct δ13C analysis.

Dual-isotope assessment: In Fig. 11.5, the Δδ2H calculated for ethylbenzene and m,p-xylene between Post 1 and Post 2 sampling events were plotted as a function of Δδ13C. The field measurements are compared to expected enrichment trends for aerobic and anaerobic biodegradation for the respective compounds. The delineated areas for aerobic and anaerobic biodegradation and the positioning of the field measurements were established as described in Sect. 11.4.2.2, using respective Λ values listed in Table 11.1. For ethylbenzene, the Λ value for Δδ2H > 100‰ was selected to represent the lower part of the anaerobic delineated area. The field measurements for both ethylbenzene and m,p-xylene positioned inside their respective delineated area expected for anaerobic biodegradation (Fig. 11.5), suggesting that these two compounds were mainly anaerobically biodegraded.

Fig. 11.5
A and b are line graphs that highlight anaerobic and aerobic biodegradation. Two lines with lambda values 40 and 35 divide the anaerobic and aerobic biodegradation. I b, two lines with lambda values 29 and 15 mark the angle of anaerobic biodegradation in between.

Expected trends for δ13C and δ2H (reported as shift (Δ) relative to baseline value) during progressive aerobic and anaerobic biodegradation of ethylbenzene (a) and m,p-xylene (b), and the measured field value obtained for monitoring well MW-S2 during the post-injection sampling event 2 (reported as shift (Δ) relative to baseline value). The Λ values are taken from Table 11.1. Due to limited Λ values for m,p-xylene, the selection for anaerobic biodegradation was made regardless of the isomers (o, m, or p), whereas no Λ values are available for aerobic biodegradation

Full size image

Biodegradation rates: To evaluate the first-order biodegradation rate (k) for ethylbenzene and m,p-xylene using carbon isotope data, Eq. 11.10 was used. For this purpose, the largest and the smallest ε-C values for anaerobic biodegradation of ethylbenzene and m,p-xylene (Table 11.1) were used to provide a range of estimated k values for each compound for the time period between Post 1 and Post 2 (33 days). The calculated k values for ethylbenzene and m,p-xylene are listed in Table 11.5. The latter in situ rates are further compared to reported in situ rates measured under natural attenuation conditions (unspecified reducing biodegradation conditions) and estimated by the conventional concentration-based approach (Aronson and Howard 1997). The in situ k values derived at Site 2 are in the high end, or larger, than the reported values for natural attenuation. These higher rates support the presence of enhanced biodegradation at the site due to the treatment, more likely due to injection of nutrients and microbial inoculants than dissolved oxygen.

Table 11.5 In situ biodegradation rates (k) calculated for ethylbenzene and m,p-xylene using site-specific δ13C data
Full size table

11.5.2.3 Benefits of Using CSIA

Inclusion of the CSIA method (single- and dual-isotope assessments) in this pilot scale test evaluating the performance of biostimulation has provided valuable information that BTEX concentration data and geochemical parameters cannot reveal or easily address. Certainly, the most valuable information is that CSIA provided evidence of ethylbenzene and m,p-xylene degradation during the treatment. The geochemical parameters and microbial genetic evaluation suggested evolving presence of both aerobic and anaerobic biodegradation. However, while the intended contaminant mass removal process by the treatment was aerobic biodegradation, CSIA suggested that ethylbenzene and m,p-xylene were still mainly anaerobically biodegraded during the course of the pilot test. The period with high DO concentration and evolving aerobic gene copies observed between Post 1 and Post 2 sampling events (for at least 33 days) suggest a transition from a prevailing BTEX-related anoxic microbial population to an oxic population. Nonetheless, the delivery of DO was not sufficiently long to sustain oxic condition and observing significant aerobic ethylbenzene and m,p-xylene biodegradation. Accordingly, numerous injection events (or a prolonged continuous injection event) will likely be required to sustain aerobic biodegradation, which will need to be considered in the treatment cost of the full-scale design. On the other hand, the in situ k rates derived using carbon isotope results for ethylbenzene and m,p-xylene were in the high end or larger than values reported in the literature for natural attenuation under reducing conditions. These higher degradation rates suggest the benefits of nutrients addition and microbial bioaugmentation. Finally, note that all the observations mentioned above are limited to groundwater samples repeatedly collected from a single monitoring well. Although this highlights the importance of sampling timing following the injection event to capture timely information, additional sampling points in space and time would have further supported the conclusions and strengthened the information gained from this pilot test.

11.6 Summary and Future Development

This chapter described a field procedure to support field practitioners when applying CSIA as a tool to assess the performance of in situ remediation treatments of BTEX-contaminated groundwater. The foremost benefit in applying this process-specific isotopic tool is the gain of direct destructive evidence for each BTEX compound assessed without relying on conventional VOC concentration analysis. When conducting dual-isotopes assessment, CSIA furthermore improves the understanding of the dominant mass degradation process(es) that are occurring during a treatment. Such valuable assessment tool confirming establishment of the intended mass removal process undeniably leads to cost-effective in situ treatment operations. Nevertheless, CSIA should be perceived as an assessment tool that needs to be used as part of a multi-line of evidence approach. Although the use of CSIA application is expected to increase in a near future, one current main application limitation is the limited availability of enrichment factor coefficients. As the enrichment factor is a key element to interpret the change in δ13C and δ2H observed during the treatment, the accuracy of the CSIA approach will undoubtedly benefit from enlarging the enrichment factor database (for both ε-C and ε-H) relatively to BTEX biodegradation, to chemical oxidation initiated by various oxidizing agents, and to physical mass removal processes enhanced by an engineered treatment (such as air sparging or soil vapor extraction). Finally, additional field scale demonstrations of CSIA application on treatments applied in saturated and unsaturated zones will also contribute to better disseminate CSIA use among the field practitioner community.