Prediction of long and short residue properties of crude oils from their infrared and near-infrared spectra

Research has been carried out to determine the feasibility of chemometric modeling of infrared (IR) and near-infrared (NIR) spectra of crude oils to predict the long residue (LR) and short residue (SR) properties of these samples. A novel method is described to predict short residue properties at different flashing temperatures based on the IR spectrum of a crude oil measured at room temperature. The resulting method is the subject of European patent application number 07251853.3 filed by Shell Internationale Research Maatschappij B.V. The study has been carried out on 47 crude oils and 4 blends, representing a large variety of physical and chemical properties. From this set, 28 representative samples were selected by principle component analysis (PCA) and used for calibration. The remaining 23 samples were used as a test set to validate the obtained partial least squares (PLS) regression models. The results demonstrate that this integrated approach offers a fast and viable alternative for the currently applied elaborate ASTM (American Society for Testing and Materials) and IP (Institute of Petroleum) methods. IR spectra, in particular, were found to be a useful input for the prediction of different LR properties. Root mean square error of prediction values of the same order of magnitude as the reproducibility values of the ASTM methods were obtained for yield long on crude (YLC), density (D(LR)), viscosity (V(LR)), and pour point (PP), while the ability to predict the sulfur contents (S) and the carbon residue (CR) was found to be useful for indicative purposes. The prediction of SR properties is also promising. Modeling of the IR spectra, and to a lesser extent, the NIR spectra as a function of the average flash temperature (AFT) was particularly successful for the prediction of the short residue properties density (D(SR)) and viscosity (V(SR)). Similar results were obtained from the models to predict SR properties as a function of the yield short on crude (YSC) values. Finally, it was concluded that the applied protocol including sample pretreatment and instrumental measurement is highly reproducible and instrument and accessory independent.     

Spectral variable selection for partial least squares calibration applied to authentication and quantification of extra virgin olive oils using Fourier transform Raman spectroscopy

The limits of quantitative multivariate assays for the analysis of extra virgin olive oil samples from various Greek sites adulterated by sunflower oil have been evaluated based on their Fourier transform (FT) Raman spectra. Different strategies for wavelength selection were tested for calculating optimal partial least squares (PLS) models. Compared to the full spectrum methods previously applied, the optimum standard error of prediction (SEP) for the sunflower oil concentrations in spiked olive oil samples could be significantly reduced. One efficient approach (PMMS, pair-wise minima and maxima selection) used a special variable selection strategy based on a pair-wise consideration of significant respective minima and maxima of PLS regression vectors, calculated for broad spectral intervals and a low number of PLS factors. PMMS provided robust calibration models with a small number of variables. On the other hand, the Tabu search strategy recently published (search process guided by restrictions leading to Tabu list) achieved lower SEP values but at the cost of extensive computing time when searching for a global minimum and less robust calibration models. Robustness was tested by using packages of ten and twenty randomly selected samples within cross-validation for calculating independent prediction values. The best SEP values for a one year's harvest with a total number of 66 Cretian samples were obtained by such spectral variable optimized PLS calibration models using leave-20-out cross-validation (values between 0.5 and 0.7% by weight). For the more complex population of olive oil samples from all over Greece (total number of 92 samples), results were between 0.7 and 0.9% by weight with a cross-validation sample package size of 20. Notably, the calibration method with Tabu variable selection has been shown to be a valid chemometric approach by which a single model can be applied with a low SEP of 1.4% for olive oil samples across three different harvest years.     

Determination of acid number and base number in lubricants by Fourier transform infrared spectroscopy

This paper describes the development of practical Fourier transform infrared (FT-IR) methods for the determination of acid number (AN) and base number (BN) in lubricants through the combined use of signal transduction via stoichiometric reactions and differential spectroscopy to circumvent matrix effects. Trifluoroacetic acid and potassium phthalimide were used as stoichiometric reactants to provide infrared (IR) signals proportional to the basic and acidic constituents present in oils. Samples were initially diluted with 1-propanol, then split, with one half treated with the stoichiometric reactant and the other half with a blank reagent, their spectra collected, and a differential spectrum obtained to ratio out the invariant spectral contributions from the sample. Quantitation for AN and BN was based on measurement of the peak height of the v(C = O) or v(COO) absorptions, respectively, of the products of the corresponding stoichiometric reactions, yielding a standard error of calibration of < 0.1 mg KOH/g oil. The AN/BN FT-IR methods were validated by the analysis of a wide range of new and used oils supplied by third parties, which had been analyzed by ASTM methods. Good correlations were obtained between the chemical and FT-IR methods, indicating that the measures are on the whole comparable. From a practical perspective, these new FT-IR methods have significant advantages over ASTM titrimetric methods in terms of environmental considerations, sample size, and speed of analysis, as well as the variety of oil types that can be handled. FT-IR analysis combining stoichiometric signal transduction with differential spectroscopy may be of wider utility as an alternative to titration in the determination of acid or basic constituents in complex nonaqueous systems.     

On-line dissolution determination of Baicalin in solid dispersion based on near infrared spectroscopy and circulation dissolution system

Drug on-line circulation dissolution system with near infrared spectrophotometer for dissolution determination was reported in this paper and subsequently partial least squares (PLS) calibration model was established for concentration prediction of Baicalin in solid dispersion. When the main factor number in PLS calibration model was 6, the correlation coefficients of PLS calibration samples and prediction ones were all 0.9999 and the relative standard deviations were 0.69% and 1.10%, respectively, which showed good robustness and predictability. Combining drug circulation dissolution system with the PLS calibration model, dissolution of Baicalin in raw material drug and solid dispersion were obtained at different times. The results indicated that the dissolution property of Baicalin in solid dispersion (especially at the early time) had been significantly improved. The accumulated dissolution of Baicalin in the solid dispersion at 45 min reached nearly 40%, increasing by 15% compared with raw material drug (about 25%). The aforementioned PLS model associated with drug circulation dissolution system provided a simple, accurate and on-line support for dissolution determination of drug, especially at the early time of rapid dissolution.     

A novel approach for development of a multivariate calibration model using a Doehlert experimental design: Application for prediction of key gasoline properties by Near-infrared Spectroscopy

Alternative methods for quality control in the petroleum industry have been obtained using Near-infrared Spectroscopy (NIRS) combined with multivariate techniques such as PLS (Partial Least-Square). The process of development and refinement of PLS models usually follows a nonsystematic and univariate procedure. The Standard Error of Cross Validation (SECV), the Standard Error of Prediction (SEP) and the determination coefficient (r²_regr.) are usually the only guides used in pursuit of the best model. In the present work, a novel approach was proposed using a Doehlert experimental design with three input variables (wavenumber range, preprocessing technique and regression/validation technique) varied at 5, 7 and 3 levels, respectively. Besides SECV, SEP and r²_regr., some additional response variables, such as the slope, r² and p_value from the external validation, as well as the number of PLS factors, were simultaneously assessed to find the optimum conditions for PLS modeling. The optimum setting for each input variable was simultaneously defined through a multivariate approach using a desirability function. With the proposed approach, the main and interaction effects could also be investigated. The methodology was successfully applied to obtain PLS models to monitor the gasoline quality through the process of product loading in trucks. To prevent product contamination or adulteration, fast prediction of key properties was obtained from FT-NIR spectra within the 7300-3900 cm^− 1 region with SECV in the range 0.04-0.63% w/w for composition (Aromatics, Saturates, Olefins and Benzene) and 0.0008 for Relative Density 20/4 °C. Each optimized PLS model was obtained with less than 40 modeling runs, demonstrating the efficiency of the proposed approach.     

The effect of flow rate, accelerometer location and temperature in acoustic chemometrics on liquid flow: Spectral changes and robustness of the prediction models

Prediction of chemical composition of flowing liquids using passive acoustic measurements and multivariate regression (acoustic chemometrics) has been reported as a promising in-line measurement method. However, the passive acoustic measurement results are also affected directly or indirectly by other factors than composition of the liquid, i.e. physical conditions of the flow and equipment/pipe properties. The present study focuses on the effects of flow rate, accelerometer location and temperature on the acoustic spectra and prediction of composition of liquids. The studied liquids were two-component mixtures of sucrose and water, and three-component mixtures of ethanol, sucrose and water. Multivariate models were estimated using both local and global calibration on full spectra, and augmented frequency and amplitude matrices derived from full spectra. Flow rate and accelerometer location had the most pronounced effect on acoustic spectra and prediction results from recalibrated local models. Temperature had a minor effect on the acoustic spectra and prediction results. The prediction error for determination of ethanol, sucrose and water increased with increasing flow rate. Changes in flow rate resulted in considerable spectral variations, causing the resultant local calibration model to perform poorly predicting the new samples taken at other flow conditions. Global models performed well on prediction of liquid composition at all studied flow and temperature levels. The global models, however, needed higher number of PLS factors and led to higher prediction errors compared to local models. Using the augmented frequency and amplitude matrices in PLS/PPLS global regression models led to higher prediction errors compared to full spectra models. However, the augmented frequency and amplitude models were more parsimonious (4–6 PLS factors) compared to the full spectra models (10–12 PLS factors).     

Kalman filter for updating the coefficients of regression models. A case study from an activated sludge waste-water treatment plant

A Kalman filter was developed to overcome the problems caused by process drifting. Different types of models were used to predict response variables of an activated sludge waste-water treatment plant. These models were constructed using MLR, PCR, and PLS. The MLR-type regression coefficients were calculated for both the PCR and PLS models. After that, the Kalman filter was used to estimate these coefficients, recursively. Both the PCR and PLS `inner relation' coefficient vectors were also estimated in this way and the results were then compared. The effect of the number of variables was also briefly studied. The testing was carried out using sequential process data. The prediction ability was measured by a Q²-value as a function of a lag in the updating of the coefficients.     

Determination of the weight percentage gain and of the acetyl group content of acetylated wood by means of different infrared spectroscopic methods

The weight percentage gain (WPG) and the acetyl group content of wood due to acetylation with acetic anhydride have been analyzed by means of Fourier transform infrared spectroscopy (FTIR) and near-infrared spectroscopy (NIR). Band height ratios (BHR) (1240/1030 (1230/1030) and 1745/1030 (1740/1030)) of the bands at 1745 (1740), 1240 (1230), and 1030 cm-1 were calculated from FTIR-KBr and FTIR-ATR (attenuated total reflection) spectra. The good linear correlation with a coefficient of determination of about 0.94 over a range from 0 to 27% WPG existing between BHRs and WPG and acetyl group content, respectively, requires only a few samples to calibrate FTIR. Partial least-squares regression models based on second derivatives of the NIR spectra in the wavenumber range from 6080 to 5760 cm-1 resulted in a R2 value of 0.99, number of PLS components (rank) between 3 and 5, root-mean-square error of cross-validation between 0.6 and 0.79%, and a residual prediction deviation up to 10. Although a wide range of input parameters (i.e., various wood species and different procedures of acetylation) was used, highly satisfactory results were obtained. Both FTIR and NIR spectroscopic means fulfill the need for determining the WPG and the acetyl content of acetylated wood. By reason of its additional potential for on-line process control, the NIR method may even outperform the FTIR method.     

Quality and statistical classification of Brazilian vegetable oils using mid-infrared and Raman spectroscopy

Palm oil, soy oil, sunflower oil, corn oil, castor oil, and rapeseed oil were analyzed with Fourier transform infrared (FT-IR) and FT-Raman spectroscopy. The quality of different oils was evaluated and statistically classified by principal component analysis (PCA) and a partial least squares (PLS) regression model. First, a calibration set of spectra was selected from one sampling batch. The qualitative variations in spectra are discussed with a prediction of oil composition (saturated, mono- and polyunsaturated fatty acids) from mid-infrared analysis and iodine value from FT-Raman analysis, based on ratioing the intensity of bands at given wavenumbers. A more robust and convincing oil classification is obtained from two-parameter statistical models. The statistical analysis of FT-Raman spectra favorably distinguishes according to the iodine value, while the mid-infrared spectra are most sensitive to hydroxyl moieties. Second, the models are validated with a set of spectra from another sampling batch, including the same oil types as-received and after different aging times together with a hydrogenated castor oil and high-oleic sunflower oil. There is very good agreement between the model predictions and the Raman measurements, but the statistical significance is lower for mid-infrared spectra. In the future, this calibration model will be used to check vegetable oil qualities before using them in polymerization processes.     

Parallel genetic algorithm co-optimization of spectral pre-processing and wavelength selection for PLS regression

Spectral pre-processing and variable selection are often used to produce PLS regression models with better prediction abilities.We proposed here to optimize simultaneously the spectral pre-processing and the variable selection for PLS regression. The method is based on parallel genetic algorithm with a unique chromosome coding both for pre-processing and variable selections. A pool of 31 pre-processing functions with various settings is tested. In the same chromosome several pre-processing steps can be combined.Three near infrared spectroscopic datasets have been used to evaluate the methodology. The efficacy of the co-optimization is evaluated by comparing the prediction ability of the PLS models with those after pre-processing optimization only. The effect of the number of successive pre-processing steps has been also tested.Concerning the different datasets used here, one can observe two different behaviors. In a first case the GA co-optimization procedure is found to perform well, leading to important improvement of the prediction ability especially when three consecutive pre-processing techniques are applied. In a second case, only the preprocessing optimization is enough to obtain an optimal model. All these models are optimal and more accurate compared to the classical models (build with the “trial and error” methods).     

On-line predictions of the aspen fibre and birch bark content in unbleached hardwood pulp, using NIR spectroscopy and multivariate data analysis

An on-line fibre-based near-infrared (NIR) spectrometric analyser was adapted for on-site process analysis at an integrated paperboard mill. The analyser uses multivariate techniques for the quantitative predication of the aspen fibre (aspen) and the birch bark contents of sheets of unbleached hardwood pulp. The NIR analyser is a prototype constructed from standard NIR components. The spectroscopic data was processed by using principal component analysis (PCA) and partial least square (PLS) regression. Three sample sets were collected from three experimental designs, each composed of known pulp contents of birch, aspen and birch bark. Sets 1 and 2 were used for model calibration and set 3 was used to validate the models. The PLS model that produced the best predictions gave an error of prediction (RMSEP) of 13% for aspen and less than 2% for birch bark. Eight components resulted in an R²X of 99.3%, R²Y of 99.6%, and Q² of 95.3%. For additional validation of aspen, three unbleached hardwood samples from the mill's production were calculated to lie between − 7% and + 6%, regarding to the PLS model. When vessel cells were counted under a light microscope a value for the aspen content of 4.7% was obtained. The predictive models evaluated were suitable for quality assessments rather than quantitative determination.     

Determination of the polymorphic forms of bicifadine hydrochloride by differential scanning calorimetry-thermogravimetric analysis, X-ray powder diffraction, attenuated total reflectance-infrared spectroscopy, and attenuated total reflectance-near-infrared spectroscopy

The pharmaceutical compound bicifadine hydrochloride, which has been found to crystallize in two polymorphic forms, has been characterized by thermal analysis, X-ray powder diffraction (XRPD), infrared (IR) spectroscopy, and near-infrared (NIR) spectroscopy. A series of 22 sample mixtures of polymorph 1 and polymorph 2 were prepared and calibration models for the quantitation of these binary mixtures have been developed for each of the XRPD, attenuated total reflectance (ATR)-IR, and ATR-NIR analytical techniques. The quantitative results were obtained using a partial least squares (PLS) algorithm, which predicted the concentration of polymorph 1 from the XRPD spectra with a root mean standard error of prediction (RMSEP) of 4.4%, from the IR spectra with a RMSEP of 3.8%, and from the NIR spectra with a RMSEP of 1.4%. The studies indicate that when analyses are carried out on equivalent sets of spectra, NIR spectroscopy offers significant advantages in quantitative accuracy as a tool for the determination of polymorphs in the solid state and is also more convenient to use than both the ATR-IR and XRPD methods. Density functional theory (DFT) B3LYP calculations and IR spectral simulation have been used to determine the nature of the vibrational modes that are the most sensitive in the analysis.     

Multivariate calibrations in gas chromatographic profiles for prediction of several physicochemical parameters of Brazilian commercial gasoline

Several Brazilian commercial gasoline physicochemical parameters, such as relative density, distillation curve (temperatures related to 10%, 50% and 90% of distilled volume, final boiling point and residue), octane numbers (motor and research octane number and anti-knock index), hydrocarbon compositions (olefins, aromatics and saturates) and anhydrous ethanol and benzene content was predicted from chromatographic profiles obtained by flame ionization detection (GC-FID) and using partial least square regression (PLS). GC-FID is a technique intensively used for fuel quality control due to its convenience, speed, accuracy and simplicity and its profiles are much easier to interpret and understand than results produced by other techniques. Another advantage is that it permits association with multivariate methods of analysis, such as PLS. The chromatogram profiles were recorded and used to deploy PLS models for each property. The standard error of prediction (SEP) has been the main parameter considered to select the “best model”. Most of GC-FID-PLS results, when compared to those obtained by the Brazilian Government Petroleum, Natural Gas and Biofuels Agency — ANP Regulation 309 specification methods, were very good. In general, all PLS models developed in these work provide unbiased predictions with lows standard error of prediction and percentage average relative error (below 11.5 and 5.0, respectively).     

Fullerenes in Chinese Ink Sticks (“Sumi”)

Small amounts (up to 0.1%) of C₆₀ and C₇₀ have been detected by high-pressure liquid chromatographic (HPLC) analysis of toluene extracts from soots used to manufacture Chinese ink sticks, or Sumi. These soots have been prepared by slow burning of pine wood, and later various seed oils. Turpentine oil, the major constituent of pine wood oil, have been found to produce twice as much fullerenes as does toluene when used as the combustion material. The yields of fullerene from unsaturated combustion materials decrease with the iodine number, reaching zero in the saturated compounds. Possible role of fullerenes as the effective gloss enhancer for black color is suggested in reference to the known technique of adding a natural red pigment to Sumi practiced in Japan.     

A kinetic spectrophotometric method for simultaneous determination of phenol and its three derivatives with the aid of artificial neural network

A novel kinetic spectrophotometric method was developed for determination of pyrocatechol, resorcin, hydroquinone and phenol based on their inhibitory effect on the oxidation of Rhodamine B (RhB) in acid medium at pH = 3.0. A linear relationship was observed between the inhibitory effect and the concentrations of the compounds. The absorbance associated with the kinetic reactions was monitored at the maximum wavelength of 557 nm. The effects of different parameters such as pH, concentration of RhB and KBrO₃, and temperature of the reaction were investigated and optimum conditions were established. The linear ranges were 0.22-3.30, 0.108-0.828, 0.36-3.96 and 1.52-19.76 μg mL⁻¹ for pyrocatechol, resorcin, hydroquinone and phenol, respectively, and their corresponding detection limits were 0.15, 0.044, 0.16 and 0.60 μg mL⁻¹. The measured data were processed by several chemometrics methods, such as principal component regression (PCR), partial least squares (PLS) and artificial neural network (ANN), and a set of synthetic mixtures of these compounds was used to verify the established models. It was found that the prediction ability of PLS, PCR and RBF-ANN was similar, however, the RBF-ANN model did perform somewhat better than the other methods. The proposed method was also applied satisfactorily for the simultaneous determination of pyrocatechol, resorcin, hydroquinone and phenol in real water samples.     

The combined effects of boron and oil heat treatment on beech and Scots pine wood properties. Part 1: Boron leaching,thermogravimetric analysis,and chemical composition

The aim of the study was to screen any possible synergistic effect available with combination of plant oils and boron compounds to decrease boron leaching and improve thermal degradation of wood by means of dual treatment. In order to reduce the leachability of boron compounds, sapwood of beech and Scots pine samples were treated with oils to create a hydrophobic layer on the wood cells that may prevent water uptake. Increase (%) of boron retained in double-treated samples after leaching was approximately 20% compared to boron-treated samples without oil. The most promising results on boron leaching obtained in the case of using waste and sunflower oil. Thermogravimetric residues were increased by the higher loading of boron compounds in double-treated samples. Spectra of FTIR-Photoacoustic spectrometer showed some deformation of lignin and cellulose followed by degradation of hemicelluloses of treated wood samples due to heat effect at 160 °C.     

Sustainable production of bioparaffins in a crude oil refinery

Most prevalent fuels of diesel engine vehicles are the gas oils produced on crude oil bases. According to some disadvantages of fatty acid-methyl-esthers (biodiesels), bioparaffins (mainly isoparaffins) are able to substitute them partially or totally. The production of bioparaffins can be carried out on a wide raw material base. This means the natural triglycerides (different origin non-food vegetable oils, 'brown fat' of sewage works, fat of protein processing, algae oils, etc.), lignocellulose etc., In this paper catalysts (NiMo/Al₂O₃; Pt/SAPO-11; Pt/AlSBA-15, Pt/β-zeolite) for the production of bioparaffins from different natural triglycerides (oils and fats) and high-molecular weight Fischer–Tropsch (heavy; >C₂₂) hydrocarbons (produced from biobased synthesis gas) are introduced. After the proper pretreatment of natural triglyceride feedstocks the suggested two-step technologies (deoxygenation and isomerisation). These are able to produce n- and isoparaffins with yield that approaches the 93–99 % of the theoretical value. Gas oil (diesel fuel) and base oil with high isoparaffin content (≥70 %) can be obtained with simultaneous isomerisation and hydrocracking of Fischer–Tropsch heavy paraffins. Properties of the two diesel fuels (bioparaffin fractions) produced with these methods satisfy the requirements of EN 590:2013 standard. Their cetane number is in the range of 65–75, which makes it possible to use itself and to blend lower quality blending components in higher quantity, which is the source of significant profit increase. The integration possibilities of the suggested technologies into crude oil refineries are introduced, too.     

大豆油脂肪酸甲酯化条件的研究

氢氧化钾-甲醇甲酯化是用气相色谱法测定植物油脂肪酸组成的常用衍生方法。本文采用单因素试验及正交试验,以大豆油中亚油酸为主要考核目标,同时,观察其他主要脂肪酸含量变化,研究氢氧化钾-甲醇法甲酯化的优化条件。     

Nanostructured polymers for detecting chemical changes during engine oil degradation

Organic and inorganic polymers nanostructured by molecular imprinting yield functional materials for detecting engine oil degradation. Using both fresh and waste oils as templates, respectively, allows us to selectively tune the polymer to incorporate either fresh or used lubricant. FT-IR studies on polymer pellets prove bulk incorporation of the analyte. By the same strategy sensitive materials can be designed for selectively determining fuel in the engine oil. Waste oil selective sensor layers yield signals that are clearly correlated both to the oil age (known from test stand measurements) and standard oil parameters, such as the total base number (TBN). TBN and sensors additionally show the same time behavior.     

Cost Prediction of a Turbofan Engine Using the Robust Partial Least Squares Method

In recent years,the cost of engines has become increasingly important to engine manufacturers,who are consistently faced with major problems on how to reduce cost to a minimum.Cost has become a decisive factor for aircraft design.To control the continual rapid increased cost,engine cost prediction is indispensable early in the design phase.But the cost data of an aircraft engine is small;we introduce the Robust Partial Least Squares Method in solving this problem,and reducing or removing the effect of outlying data points,which is different from the Classical PLS.We use the MATLAB software doing several simulations;results and analysis of a real turbofan engine data set show the effectiveness and robustness of the Robust PLS method.The Robust PLS method can effectively be used to estimate Turbofan Engine cost with reasonable accuracy.