Rapid prediction of total petroleum hydrocarbons in soil using a hand-held mid-infrared field instrument

Rapid Determination of the Total Petroleum Hydrocarbon Content of Soils by Handheld Fourier Transform Near-Infrared Spectroscopy

For successful soil remediation and hydrocarbon exploration operations, determining the total petroleum hydrocarbon (TPH) content of soils is an indispensable process step. This paper reports on the performance of a handheld Fourier transform near-infrared (FT-NIR) spectrometer for rapid and quantitative determination of TPH content of soils from two different sites by diffuse reflection measurements. For rapid decisions for exploration work or environmental site assessment projects, a quick─preferably on-site─determination of TPH content is valuable. Diffuse reflection NIR spectra were recorded from soil samples of two different sites with TPH reference values ranging from 350 to 30,000 ppm, as determined by capillary gas chromatography and flame ionization detection with hydrocarbon fingerprinting C1-C44. However, this paper not only addresses the development of site-specific partial-least squares (PLS) calibrations but also demonstrates the locally-weighted PLS (LW-PLS) technique, which can be used to develop global, site-independent PLS calibrations without significant penalty in calibration performance. As a first step, the diffuse reflection spectra were used to develop conservative, site-specific PLS calibration models with root-mean-square calibration/cross-validation errors (RMSEC/RMSECV) of 1043/1106 and 741/785 ppm TPH, respectively, and the average absolute prediction errors for samples not contained in the calibration set were 451 and 293 ppm for the two sites, respectively. In a further step, significant degradation of the RMSE values of a conservative PLS model based on the NIR spectra of both sites was then compared to the application of the LW-PLS method, with only a slight loss of the prediction accuracy relative to the site-independent models. This study confirms the ability of next-generation portable FT-NIR spectrometers to predict low TPH levels in various soil types through both─soil-specific and site-independent─calibrations, giving these spectrometers the potential to become rapid screening tools in the field.

Comparing the effect of different sample conditions and spectral libraries on the prediction accuracy of soil properties from near- and mid-infrared spectra at the field-scale

The prediction accuracy of soil properties by proximal soil sensing has made their application more practical. However, in order to gain sufficient accuracy, samples are typically air-dried and milled before spectral measurements are made. Calibration of the spectra is usually achieved by making wet chemistry measurements on a subset of the field samples and local regression models fitted to aid subsequent prediction. Both sample handling and wet chemistry can be labour and resource intensive. This study aims to quantify the uncertainty associated with soil property estimates from different methods to reduce effort of field-scale calibrations of soil spectra. We consider two approaches to reduce these expenses for predictions made from visible-near-infrared ((V)NIR), mid-infrared (MIR) spectra and their combination. First, we considered reducing the level of processing of the samples by comparing the effect of different sample conditions (in-situ, unprocessed, air-dried and milled). Second, we explored the use of existing spectral libraries to inform calibrations (based on milled samples from the UK National Soil Inventory) with and without 'spiking' the spectral libraries with a small subset of samples from the study fields. Prediction accuracy of soil organic carbon, pH, clay, available P and K for each of these approaches was evaluated on samples from agricultural fields in the UK. Available P and K could only be moderately predicted with the field-scale dataset where samples were milled. Therefore this study found no evidence to suggest that there is scope to reduce costs associated with sample processing or field-scale calibration for available P and K. However, the results showed that there is potential to reduce time and cost implications of using (V)NIR and MIR spectra to predict soil organic carbon, clay and pH. Compared to field-scale calibrations from milled samples, we found that reduced sample processing lowered the ratio of performance to inter-quartile range (RPIQ) between 0% and 76%. The use of spectral libraries reduced the RPIQ of predictions relative to field-scale calibrations from milled samples between 54% and 82% and the RPIQ was reduced between 29% and 70% for predictions when spectral libraries were spiked. The increase in uncertainty was specific to the combination of soil property and sensor analysed. We conclude that there is always a trade-off between prediction accuracy and the costs associated with soil sampling, sample processing and wet chemical analysis. Therefore the relative merits of each approach will depend on the specific case in question.

VIRS based detection in combination with machine learning for mapping soil pollution

Widespread soil contamination threatens living standards and weakens global efforts towards the Sustainable Development Goals (SDGs). Detailed soil mapping is needed to guide effective countermeasures and sustainable remediation operations. Here, we review visible and infrared reflectance spectroscopy (VIRS) based detection methods in combination with machine learning. To date, proximal, airborne and spaceborne carrier devices have been employed for soil contamination detection, allowing large areas to be covered at low cost and with minimal secondary environmental impact. In this way, soil contaminants can be monitored remotely, either directly or through correlation with soil components (e.g. Fe-oxides, soil organic matter, clay minerals). Observed vegetation reflectance spectra has also been proven an effective indicator for mapping soil pollution. Calibration models based on machine learning are used to interpret spectral data and predict soil contamination levels. The algorithms used for this include partial least squares regression, neural networks, and random forest. The processes underlying each of these approaches are outlined in this review. Finally, current challenges and future research directions are explored and discussed.

Factors affecting in situ analysis of total petroleum hydrocarbons in contaminated soils by using a mid-infrared diffuse reflectance spectroscopy

The hand-held mid-infrared diffuse reflectance infrared Fourier transform (MIR-DRIFT) spectrometer was used to assess the applicability of on-site and real time monitoring of total petroleum hydrocarbons (TPH) in contaminated soils during site characterization and remediation. Field measurement devices (MIR-DRIFT and turbidimetric screening test kits) were used to analyze reference soils with concentration ranging from 713 to 54790 mg/kg and compared with the results by a gas chromatography/mass spectrometry method (GC/MS). In situ field measurement of 147 petroleum-contaminated soil samples from 11 contaminated sites was correlated with laboratory-determined soil TPH levels by GC/MS. The concentrations of TPH by MIR-DRIFT were significantly correlated to the concentrations of TPH by GC/MS. Detection of TPH by the MIR spectrometer was not affected by the weathering effects of diesel-contaminated soils. Soils contaminated by mixed fuels with high content of gasoline constituents may cause the potential interference in MIR measurement. In field practice, interference may be attributed to soil moisture, soil organic matter, and soil texture. Soil moisture below 5% is required to reduce variation of infrared beam reflected from high level of surface liquid. When measuring the contaminated soil with a high organic matter content, the results may be overestimated due to the possible effects of surface reflection and interference. Clay and partial silty clay soils were not suitable for MIR spectrometer detection due to a potential shielding effect to reduce the infrared radiation absorbed by TPH. Future research is warranted to reduce the variation caused by soil texture and heterogeneity in TPH prediction.

Determination of Total Hydrocarbons in Contaminated Soil with "Thin Layer Sorptive Extraction Coupled with Attenuated Total Reflectance-Fourier Transform Infrared Spectroscopy"

Soil analysis using infrared spectroscopy has been proposed as an alternative to conventional soil analysis to detect soil contamination. This study therefore aims to develop an innovative, in situ, rapid, precise, and inexpensive method that is easy to implement in order to assess soil contamination with hydrocarbons. This work describes the development and validation of a new extraction method by thin-layer sorptive extraction and attenuated total reflectance-Fourier transform infrared spectroscopy (TLSE-ATR-FTIR). First, this method allows the preconcentration of thermodesorbed pollutants on a polymer thin film and then, their quantification by ATR-FTIR using a standard addition method. A five factor fractional factorial design was used to identify the most significant factors impacting the analysis. These factors include soil texture, total organic carbon (TOC), humidity, and concentrations of contaminants. The results showed that TOC, nature (clay, sandy, and loamy) of the soil, and the concentration of pollutants can affect the infrared absorbance. The analytical method has been validated by verifying the different performance criteria such as linearity, accuracy, precision, and quantitation limit. The comparison of the results obtained by TLSE-ATR-FTIR to the results of conventional analyses carried out by accredited laboratories confirms that the use of the proposed method can become an effective alternative to the current methods for the determination of the total hydrocarbons in soils.

Quantification of Hydrocarbon Abundance in Soils Using Deep Learning with Dropout and Hyperspectral Data

Terrestrial hydrocarbon spills have the potential to cause significant soil degradation across large areas. Identification and remedial measures taken at an early stage are therefore important. Reflectance spectroscopy is a rapid remote sensing method that has proven capable of characterizing hydrocarbon-contaminated soils. In this paper, we develop a deep learning approach to estimate the amount of Hydrocarbon (HC) mixed with different soil samples using a three-term backpropagation algorithm with dropout. The dropout was used to avoid overfitting and reduce computational complexity. A Hyspex SWIR 384 m camera measured the reflectance of the samples obtained by mixing and homogenizing four different soil types with four different HC substances, respectively. The datasets were fed into the proposed deep learning neural network to quantify the amount of HCs in each dataset. Individual validation of all the dataset shows excellent prediction estimation of the HC content with an average mean square error of ~ 2 . 2 × 10 - 4 . The results with remote sensed data captured by an airborne system validate the approach. This demonstrates that a deep learning approach coupled with hyperspectral imaging techniques can be used for rapid identification and estimation of HCs in soils, which could be useful in estimating the quantity of HC spills at an early stage.

The application of a handheld mid-infrared spectrometry for rapid measurement of oil contamination in agricultural sites

Rapid analysis of oil-contaminated soils is important to facilitate risk assessment and remediation decision-making process. This study reports on the potential of a handheld mid-infrared (MIR) spectrometer for the prediction of total petroleum hydrocarbons (TPH), including aliphatic (alkanes) and polycyclic aromatic hydrocarbons (PAH) in limited number of fresh soil samples. Partial least squares regression (PLSR) and random forest (RF) modelling techniques were compared for the prediction of alkanes, PAH, and TPH concentrations in soil samples (n = 85) collected from three contaminated sites located in the Niger Delta, Southern Nigeria. Results revealed that prediction of RF models outperformed the PLSR with coefficient of determination (R²) values of 0.80, 0.79 and 0.72, residual prediction deviation (RPD) values of 2.35, 1.96, and 2.72, and root mean square error of prediction (RMSEP) values of 63.80, 83.0 and 65.88 mg kg^-1 for TPH, alkanes, and PAH, respectively. Considering the limited dataset used in the independent validation (18 samples), accurate predictions were achieved with RF for PAH and TPH, while the prediction for alkanes was less accurate. Therefore, results suggest that RF calibration models can be used successfully to predict TPH and PAH using handheld MIR spectrophotometer under field measurement conditions.

RemScan: A tool for monitoring the bioremediation of Total Petroleum Hydrocarbons in contaminated soil

Total Petroleum Hydrocarbons (TPH) represent major environmental contaminants which pose a significant risk to ecosystems and humans heath if left untreated. Bioremediation represents a simple, cheap and environmentally-safe approach to clean up TPH-contaminated sites. Traditional TPH analysis is expensive and time-consuming. Here we assess, for the first time, the potential of RemScan as a fast, accurate and cost-effective portable device to be used as a tool to monitor the bioremediation process. A variety of TPH-contaminated soils were subject to TPH quantitative analysis using RemScan. The TPH values obtained were validated and compared against the results obtained from an accredited external laboratory, which uses Gas Chromatography / Mass Spectrometry (GC/MS) for TPH analysis. •RemScan showed a correlation coefficient (R²) of 0.998 in comparison with the traditional methods, but importantly with a significant reduction in both time and cost.•RemScan was successfully used to measure TPH concentrations in bioremediated, weathered-contaminated and highly contaminated soil samples with TPH concentrations varying from 100 to 100,000 mg kg^-1.•The RemScan Laboratory Station was used to minimize the source of errors associated with human manual handling.

Infrared spectroscopy and multivariate methods as a tool for identification and quantification of fuels and lubricant oils in soil

Environmental contamination caused by leakage of fuels and lubricant oils at gas stations is of great concern due to the presence of carcinogenic compounds in the composition of gasoline, diesel, and mineral lubricant oils. Chromatographic methods or non-selective infrared methods are usually used to assess soil contamination, which makes environmental monitoring costly or not appropriate. In this perspective, the present work proposes a methodology to identify the type of contaminant (gasoline, diesel, or lubricant oil) and, subsequently, to quantify the contaminant concentration using attenuated total reflection Fourier transform infrared (ATR-FTIR) spectroscopy and multivariate methods. Firstly, gasoline, diesel, and lubricating oil samples were acquired from gas stations and analyzed by gas chromatography to determine the total petroleum hydrocarbon (TPH) fractions (gasoline range organics, diesel range organics, and oil range organics). Then, solutions of these contaminants in hexane were prepared in the concentration range of about 5-10,000 mg kg^-1. The infrared spectra of the solutions were obtained and used for the development of the pattern recognition model and the calibration models. The partial least square discriminant analysis (PLS-DA) model could correctly classify 100% of the samples of each type of contaminant and presented selectivity equal to 1.00, which provides a suitable method for the identification of the source of contamination. The PLS regression models were developed using multivariate filters, such as orthogonal signal correction (OSC) and general least square weighting (GLSW), and selection variable by genetic algorithm (GA). The validation of the models resulted in correlation coefficients above 0.96 and root-mean-square error of prediction values below the maximum permissible contamination limit (1000 mg kg^-1). The methodology was validated through the addition of fuels and lubricating oil in soil samples and quantification of the TPH fractions through the developed models after the extraction of the analytes by the EPA 3550 method adapted by the authors. The recovery percentage of the analytes was within the acceptance limits of ASTM D7678 (70-130%), except for one sample (69% of recovery). Therefore, the methodology proposed here provides faster and less costly analyses than the chromatographic methods and it is adequate for the environmental monitoring of soil contamination by gas stations.