The use of multivariate statistics and mathematically modeled IR spectra for determination of HVO content in diesel blends

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F60461373%3A22320%2F25%3A43929841" target="_blank" >RIV/60461373:22320/25:43929841 - isvavai.cz</a>

Výsledek na webu

<a href="https://www.sciencedirect.com/science/article/pii/S0016236124021124?via%3Dihub" target="_blank" >https://www.sciencedirect.com/science/article/pii/S0016236124021124?via%3Dihub</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1016/j.fuel.2024.132963" target="_blank" >10.1016/j.fuel.2024.132963</a>

Jazyk výsledku

angličtina

Název v původním jazyce

The use of multivariate statistics and mathematically modeled IR spectra for determination of HVO content in diesel blends

Popis výsledku v původním jazyce

Diesel fuel containing one or more bio-components is currently increasingly represented on the market. The most common bio-component for diesel fuels is FAME (Fatty Acid Methyl Esters), but the share of diesel with HVO (Hydrotreated Vegetable Oil) content has been increasing in recent years. HVO has many advantages, compared to FAME, but the fact remains that it is difficult to detect and determine HVO in blends with mineral diesel fuel. The reason is the great similarity in the composition of HVO and mineral diesel fuel. A commonly used method for determining HVO content in diesel is radiocarbon C14 analysis, which can be performed by using LSC (Liquid Scintillation Counting) or AMS (Accelerator Mass Spectrometry). However, the important fact is that both methods are significantly expensive. For this reason, infrared spectrometry followed by chemometric data processing was used in this work. Infrared spectroscopy using ATR technique was utilized for primary data collection. Five PLS statistical models were prepared for the determination of HVO in blends with mineral diesel fuel. Analytical standards used in these models consisted of ten different mineral diesel fuels and five different HVO fuels. Spectra of these neat components were measured, spectra of calibration standards of blends were created mathematically by linear combination of the neat fuel components spectra (HVO and diesel). Two sets of validation standards which were applied in the first PLS model were created the same way as the calibration standards. The first model used only mathematically generated spectra for calibration and validation. Another four models were validated with real measured spectra, but they differed in the scope and method of calibration. All these five models were afterward optimized for minimalization of RMSEC (root mean square error of calibration) and RMSEP (root mean square error of prediction). Mean centering and other smoothing techniques were used for optimization. The RMSEP ranged from 1.67 to 2.22 vol% for models validated using the real spectra. For the mathematically validated model, it was less (0.32 vol% after optimization).

Název v anglickém jazyce

The use of multivariate statistics and mathematically modeled IR spectra for determination of HVO content in diesel blends

Popis výsledku anglicky

Diesel fuel containing one or more bio-components is currently increasingly represented on the market. The most common bio-component for diesel fuels is FAME (Fatty Acid Methyl Esters), but the share of diesel with HVO (Hydrotreated Vegetable Oil) content has been increasing in recent years. HVO has many advantages, compared to FAME, but the fact remains that it is difficult to detect and determine HVO in blends with mineral diesel fuel. The reason is the great similarity in the composition of HVO and mineral diesel fuel. A commonly used method for determining HVO content in diesel is radiocarbon C14 analysis, which can be performed by using LSC (Liquid Scintillation Counting) or AMS (Accelerator Mass Spectrometry). However, the important fact is that both methods are significantly expensive. For this reason, infrared spectrometry followed by chemometric data processing was used in this work. Infrared spectroscopy using ATR technique was utilized for primary data collection. Five PLS statistical models were prepared for the determination of HVO in blends with mineral diesel fuel. Analytical standards used in these models consisted of ten different mineral diesel fuels and five different HVO fuels. Spectra of these neat components were measured, spectra of calibration standards of blends were created mathematically by linear combination of the neat fuel components spectra (HVO and diesel). Two sets of validation standards which were applied in the first PLS model were created the same way as the calibration standards. The first model used only mathematically generated spectra for calibration and validation. Another four models were validated with real measured spectra, but they differed in the scope and method of calibration. All these five models were afterward optimized for minimalization of RMSEC (root mean square error of calibration) and RMSEP (root mean square error of prediction). Mean centering and other smoothing techniques were used for optimization. The RMSEP ranged from 1.67 to 2.22 vol% for models validated using the real spectra. For the mathematically validated model, it was less (0.32 vol% after optimization).

Druh

J_imp - Článek v periodiku v databázi Web of Science

CEP obor

—

OECD FORD obor

20704 - Energy and fuels

Projekt

—

Návaznosti

S - Specificky vyzkum na vysokych skolach

Rok uplatnění

2025

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název periodika

Fuel

ISSN

0016-2361

e-ISSN

1873-7153

Svazek periodika

379

Číslo periodika v rámci svazku

January 2025

Stát vydavatele periodika

GB - Spojené království Velké Británie a Severního Irska

Počet stran výsledku

8

Strana od-do

132963

Kód UT WoS článku

001312388300001

EID výsledku v databázi Scopus

2-s2.0-85203403736