There is a continued interest in the possibility of using brownfield and contaminated sites for amenity and non-food crop such as biomass. However, establishment of vegetation on these sites is poor due to nutrient deficiency, phytotoxicity of contaminants and poor soil physical conditions. Addition of mature composts, which is beneficial to soils in terms of physical properties, nutrient availability and microbial activity, has the potential to promote plant development and restore degraded land into productive use. Yet, the influence of adding large amount of compost to contaminated soils especially on the fate of polycyclic aromatic hydrocarbons (PAHs) and the hydraulic properties of soils remain poorly studied. In this study, the temporal bioavailability changes of 16 PAHs in three contaminated soils amended with two contrasting composts at two different rates was investigated over a period of 8 months. Further to this, the water-release characteristics were determined and the physical quality soil indicators derived. Total and bioavailable fractions were obtained by sequential ultrasonic solvent extraction and hydroxypropy-β-cyclodextrin (HPCD) extraction, respectively after 0, 3, 6 and 8 months. PAHs were identified and quantified by GC-MS.
Water release characteristics were measured using sand table apparatus at water potentials, expressed in water height equivalents, of 0, 30, 50 and 100 cm and pressure cell apparatus at pressure heads of 200, 500, 1000, 1500, 2000, 5000, 10000 and 15000 cm. The Van Genuchten equation was fitted to the water release curve data using the RETC code.
Conjoint analysis (CJ), five-way analysis of variance and machine learning models including multilayer perceptrons (MLP), radial basis function (RBF), support vector regression (SVR), M5 model tree (M5P), M5 rule (M5R) and linear regression (LR) were used to (1) assess the relative importance and interactions of soil and compost type, compost ratio, and incubation time and (2) predict temporal PAH bioavailability changes.
Desorption and degradation contributed to 30% and 70%, respectively, of the PAH loss in the spiked soil, while PAH loss in the coal tar and coal ash contaminated soils resulted from 40% enhanced desorption and 60% enhanced degradation. Compost type (green and catering meat wastes) and application rates (250 and 750 t/ha) had little influence on PAH bioavailability. In contrast compost addition generally increased water retention and improved the physical quality indicators of the contaminated soils while they remained suboptimum to sustain yields of bioenergy crops. The ML models successfully identified the relative importance of each variable including incubation time, organic carbon content, soil moisture content, and nutrient levels on the temporal bioavailability change of individual PAH. For instance, time and moisture were identified as the most important variables influencing naphthalene bioavailability while soil organic carbon content was the most important factor for benzo[a]pyrene bioavailability. Further to ML models, conjoint analysis and five-way analysis of variance highlighted that soil type and contact time were the two most significant factors influencing the PAH bioavailability in compost amended soils. The other two factors, compost type and ratio of compost addition, were less important but their interactions with the other factors were significant. Specifically the 4-factor interactions showed that compost addition stimulated the degradation of high molecular at the initial stage (3 month) by enhancing the competitive sorption within PAH groups. To the best of our knowledge, this is the first study to investigate the influence of multiple interactions in soil-compost-oil matrices on PAH bioavailability. The overall results clearly illustrated that incorporation of multi-factors interactions into risk assessment and site remediation can help to refine decision-making for remediation end points and risk management. Our study further demonstrates that ML can help to predict concentration of a wide range of pollutants in soils which could reduce chemical monitoring at site and further assist decision-making.
Engineered microbial decontamination of soils and subsurface materials is an efficient measure to remediate PAH-contaminated sites. Extended contact time of hydrophobic pollutants and soil constituents, however, may render a fraction of contaminants inaccessible for microbial metabolisation, thus diminishing the potential efficacy of biological treatment. Numerous studies were directed towards the elucidation of sequestration processes and the quantification of the share of pollutants that is susceptible to biological breakdown. A fundamental question is whether spiked aged soils can resemble the sorption and desorption of native PAHs in historically contaminated soils. The recently developed ‘contaminant trap’ [1] was used as an experimental platform for addressing this question.
In the present study, 25 Austrian soils were collected and spiked with four selected polycyclic aromatic compounds. Using the contaminant trap, PAH desorption behaviour from freshly contaminated and aged soils was monitored and then compared with field soils from industrial sites that experienced historical PAH contamination. The aim was to determine fundamental differences in desorption behaviour between spiked and native PAHs with the working hypothesis of a higher retention of native PAHs in historically contaminated soils that cannot be resembled by spiked and aged soils.
Desorption experiments were conducted with freshly contaminated soils (Fluoranthene, Phenanthrene, Benzo(a)pyrene and Benzo(g,h,i)perylene), their aged counterparts and with three soils collected from historically PAH-contaminated sites in Austria. Desorption curves and the desorption resistant PAH fraction were determined for all soils using the contaminant trap [1]. Briefly, soil is added to a solution containing a diffusive carrier (cyclodextrin) and a microbial inhibitor. Desorbed PAHs are captured by an infinite polymer sink (silicon plus activated carbon). Desorption of PAHs was determined for ground and non-ground samples of historically contaminated soils since increased desorption from ground samples would indicate physical entrapment of PAHs by the soil matrix. Desorption experiments were for repeated some soils at high additions of toluene since increased desorption in the presence of high toluene concentrations would indicate competitive binding, which is consistent with adsorption to high affinity sides being the governing sorption process.
PAH concentrations in spiked soils decreased typically by two orders of magnitude during 56 days of incubation in contaminant traps, with the main release occurring within the first two weeks. Desorption of PAHs from historically contaminated soils was much slower during incubation in contaminant traps, and desorption resistant PAH fractions ranging between 25 and 71% were significantly higher than for spiked soils. Aging of spiked soils was not able to reduce the substantial differences between PAH desorption curves for historically polluted soils and spiked soils [2]. The bioaccessible PAH fraction was at least one order of magnitude larger in spiked soils compared to real world samples from historically contaminated sites. The observed differences could not be explained by physical entrapment of PAHs in historically contaminated soils since grinding of these soils did not enhance PAH desorption from the soils. PAHs rather appeared to be bound to high affinity sorption sites, which was indicated by additional PAH release upon addition of toluene. A much lower retention in spiked soils is consistent with absorption by amorphous organic matter or the adsorption to a much larger population of low affinity sorption sites. Irrespective of underlying mechanisms responsible for the dissimilarity in desorption between historically contaminated and spiked soils, this difference has very important implications for real world situations. First of all, these differences challenge the significance of extrapolations of desorption and bioavailability results that were obtained with spiked PAHs. Further, a much higher PAH retention in historically contaminated soils is good news, it suggests limited mobility and exposure of native PAHs. However, the addition of co-solutes can reduce this retention and as a consequence, lead to a re-mobilisation of PAHs which deserves additional attention and research.
References
[1] Mayer, P., et al. 2011. A Contaminant Trap as a Tool for Isolating and Measuring the Desorption Resistant Fraction of Soil Pollutants. Environmental Science and Technology. 45(7): p. 2932-2937.
[2] Scherr, K., et al. 2012. Desorption-resistant fraction in PAH-contaminated soils: spiked and aged soils can not resemble historically contaminated soils. 6th SETAC World Congress, Berlin.
Acknowledgement
This work was financially supported by the Austrian Research Promotion Agency (FFG), and the European Regional Development Fund (EFRE) together with the Government of Lower Austria (project MACATA, WST3-T-95/017-2012).
Many sites in Europe such as former gasworks, wood preservation facilities and coke oven plants may lead to Polycyclic Aromatic Compounds (PAC) pollution in soils and groundwater. In risk assessment, PAH are generally target compounds. But other PAC classes (oxygenated or nitrogenated PAC) are ignored in risk assessments even if they are recognized as toxic, thus leading to an underestimation of the risk. Some of these compounds occur in the original contamination but may be also generated during PAHs degradation (during natural attenuation or remediation operations). However, the evaluation of the treatments efficiency doesn’t take into account the PAC formation. Moreover, these newly generated compounds more water soluble than classical PAH can be mobilized during or after treatments and transferred to surface waters or groundwater.
In a first step, this study focused on the evolution of PACs concentrations during two oxidative chemical remediation treatments: Fenton like oxidation (activated with magnetite), permanganate (MnO4-), and bioremediation experiments (batch incubation) on soils from three types of site: coke oven and gas plants and wood preservation. Each remediation treatment has been carried out in batch during 1h, 24h and one week. 16 PAHs, 11 Oxy-PACs and 4 Nitro-PACs were identified and quantified by using solvent extraction followed by GC-MS.
In a second step, the effects of such chemical oxidations or bioremediation on the water soluble PACs (quality and quantity) was evaluated in order to improve risk assessment. A Solid Phase Extraction method to extract PAC (including PAH and Oxy-PAC) from aqueous samples was first developed. This method was validated with spiked waters but also with waters collected from real coke oven plant sites. The same procedure was also applied on two soil categories (coke oven and gas plants) previously treated or not by chemical oxidation (permanganate and Fenton like oxidation (activated with magnetite)) using different amounts of oxidant or by bioremediation experiments (batch incubation). The PAC repartition between the aqueous and solid phases was studied so as to (i) understand the behavior of each class of contaminants during the treatment and (ii) better evaluate consequences of chemical and biological remediation treatments on water quality.
Results from this project reveals (i) that soils form sites contaminated by PAHs contained polar PAC in important amount, (ii) that chemical and/or biological treatments can lead to an enrichment (relative or absolute) in polar PACs (especially aromatic ketones) and (iii) that polar PAC are preferentially mobilized by water compared to traditional PAH, especially oxygenated PAC initially present but also formed during remediation treatments or during natural attenuation processes.
Background/Objectives ─ Perfluorinated compounds (PFCs), such as perfluorooctane sulfonic acid (PFOS) and perfluorooctanoic acid (PFOA), are a class of compounds widely used in diverse applications, such as carpet protection, surfactants, and shampoos. In particular, PFC-based surfactants have been used in aqueous film-forming foams (AFFF) that have been routinely used in both civilian and US military fire-fighting. Historically, effluents from AFFF fire-fighting activities were neither impounded nor pre-treated prior to discharge to water treatment systems or to the environment. Widespread environmental presence of PFCs has been identified. PFCs are persistent, bioaccumulative, toxic, and are not readily degradable by conventional biological and abiotic treatment technologies. Thus PFCs have drawn increasing public and regulatory concerns including being listed on Annex B of the Stockholm Convention on Persistent Organic Pollutants (POPs) and US EPA’s Unregulated Contaminant Monitoring Rule-3 (UCMR-3) list. The UCMR-3 listing requires that large Public Water Systems sample and analyze six PFCs and has already revealed PFC impacts to some of these systems. These impacts and other PFC related regulatory initiatives around the globe have resulted in a dramatic increase in the number of sites characterized for PFCs.
This paper presents common encounters and lessons learned from characterizing several dozen PFC impacted sites and highlights a number of unique concerns and protocols that must be followed due to the characteristics of PFCs and significant potential for sample contamination. The high solubility of PFCs and associated resulting large dilute plumes, low laboratory detection limits and Health Advisory Levels (HALs), presence of PFCs in many of the products routinely used during groundwater sampling, and potential for non-point sources of PFCs (e.g. dust) all create the need for unprecedented care to ensure an accurate Conceptual Site Model on PFC impacted sites.
Approach/Activities ─ Data and experiences from several dozen PFC sites was gathered via AECOMs PFC Working Group, which collaborates on technical issues to improve our overall understanding of PFCs. The data was then evaluated to identify both common and unique results. Special sampling protocols were also captured from the experience of the PFC Working Group, available literature on the topic, and recommendations provided by analytical laboratories. The experience was then combined to develop a detailed list of sampling protocols and procedures. The site data was also evaluated to identify any trends and outliers or unique situations.
Results/Lesson Learned ─ The results of this evaluation and associated lessons learned provide valuable insight into: sources of background and non-point sources of PFCs, common fate and transport characteristics of PFC soil, groundwater, surface water, and sediment impacts, and reinforce that PFC characterization activities must be very rigorous and involve new protocols and procedures.
Background
Perfluoroalkyl and polyfluoroalkyl substances (PFASs) have been in focus the last 10 – 15 years as persistent and undesirable contaminants. Concern about health and environmental risks led to the EC regulation in 2006 of a specific compound, PFOS (perfluorooctane sulfonic acid) and its derivatives (salts, amides, halides etc.). These are only allowed in very low concentrations or in certain closed loop production systems. This meant that the import or production of PFOS in fire-fighting foams was banned in the EU in 2006, but stockpiles could be used until 27 June 2011. PFOS and its derivatives have been listed since 2009 as a POP (Persistent Organic Pollutant) under the Stockholm Convention. Certain other specific PFAS substances and their derivatives including PFOA (perfluorooctanoic acid) are included on the EU “Candidate List” as Substances of Very High Concern (SVHC).
Because of this concern, many reports (Danish, Swedish, and other EU lands) have been produced with surveys identifying products or waste which might contain PFAS as well as environmental reports mapping the occurrence of PFAS in soil, groundwater and biota.
Objective
Due to general awareness of the risk of soil and groundwater contamination with PFAS especially PFOS and PFOA, the Estates and Infrastructure Organisation (EIO) have implemented screening for PFAS while monitoring groundwater quality at some active and former air bases. The initial focus has been at air bases where fire-fighting exercises have been carried out. The results of the initial screening have been disseminated along with results from other point sources in a publication by the Danish Environmental Protection Agency.
Apart from the further investigations around fire-fighting areas and former waste dumps, the EIO have now implemented an extended project plan to identify and document other activities with products containing PFAS used by the Danish Defence.
The main focus for these investigations is to identify the threat to groundwater reservoirs.
Method
The first phase of the project involves interviews with personnel working for the Danish Defence and Danish Emergency Management Agency. The objective is to attempt to identify products that might have contained PFAS such as cleaning agents, impregnation agents (water repellent agents), paints, hydraulic oil as well as fire-fighting foams. It is the content of PFAS in the final product, the potential flux to soil and groundwater and the risk for migration in the environment which is of interest.
Furthermore, activities whereby these products or activities might cause soil or groundwater contamination i.e. spill, waste water, storage, general handling, decanting etc., are identified and located at Defence Establishments.
Based on this information up to 15 sites are selected for investigation either by sampling and analysis of water samples from existing wells or new wells. Soil samples are also collected and both water and soil samples are analysed for a selection of at least 15 PFAS components.
The second phase of the project will be implemented in the beginning of 2015 and involves supplementary investigations to delineate soil and groundwater contamination with PFAS. The range of PFAS to be analysed will be expanded in response to the newest research concerning PFAS in the environment.
The third phase is planned for summer 2015 and is expected to involve evaluation of suitable remediation techniques.
Results
The initial groundwater screening at Danish military air bases has demonstrated that PFAS substances are present in the groundwater at some, but not all sites, ranging from <10 ng/l to 7500 ng/l for sum of 9 PFAS. In contrast to investigations at some civil air ports or air bases in other countries PFOS and PFOA have not been the dominating components indicating that the fire-fighting foams have been produced by a manufacturer with a different formulation. The C-6 and C-7 perfluorinated carboxylic acids (perfluorohexanoic acid - PFHxA and perfluoroheptanoic acid - PFHpA) have dominated in the initial screening and for this reason EIO has decided to analyse samples for a broader range of PFAS.
The newest results indicate that additionally a C-4 and C-5 perfluorinated carboxylic acids as well as 6:2 FTS (6:2 fluorotelomer sulfonate) are present in groundwater samples at Danish Defence Air bases.
More results from on-going investigations will be presented at AquaConSoil.
Conclusion and relevance
At present, Denmark has no quality criterion for PFAS in groundwater, drinking water or soil. Results are often compared to the German criterion for drinking water of 100 ng/l for the sum of PFOS and PFOA.
The results provide an estimation of groundwater loads around Danish Ministry of Defence properties and an indication of the PFAS composition in groundwater and soil samples. Determination of the PFAS composition is an important contribution with respect to the future definition of a comprehensive Danish quality criterion which should include both individual and sum values for relevant PFAS.
Furthermore, it would be reasonable to expect that the PFAS profile (composition) is dependent on the source and therefore of great interest for future identification of point sources of pollution.