Fuel 291 (2021) 120160

Contents lists available at ScienceDirect

Fuel
journal homepage: www.elsevier.com/locate/fuel

Review article

Rancimat vs. rapid small scale oxidation test (RSSOT) correlation analysis,
based on a comprehensive study of literature
Ferdinand Bär a, *, Markus Knorr a, Olaf Schröder a, Henning Hopf b, Thomas Garbe a,
Jürgen Krahl c, d
a
Coburg University of Applied Sciences and Arts, Coburg, Germany
b
Technical University of Braunschweig, Germany
c
OWL University of Applied Sciences and Arts, Lemgo, Germany
d
Fuels Joint Research Group, www.fuels-jrg.de, Germany

A R T I C L E I N F O A B S T R A C T

Keywords: Due to the fact that modern diesel engines require very well-defined fuel properties to ensure high efficiency, low
Rancimat emissions and a long durability, the fuel properties and their determination methods have to be standardized.
PetroOxy Most diesel fuel standards allow a certain volume fraction of biodiesel that can be blended to fossil diesel fuel.
RSSOT
Biodiesel consists of saturated and unsaturated fatty acid methyl esters (FAME) which are more susceptible to
Biodiesel
Diesel
oxidation compared to fossil diesel fuel. Oxidation products formed by autoxidative processes can lead to higher
Fuel emissions, a higher corrosiveness, fuel filter blocking, clogging of fuel injectors, the formation of deposits in the
Oxidation stability entire fuel system and a decreasing lubricity [1]. Therefore, oxidation needs to be prevented by the addition of
oxidation inhibitors. These antioxidants are consumed during storage, resulting in partial loss of their efficiency
against oxidative stress. After this induction period all antioxidants have been consumed and harmful oxidation
products can be formed. For the determination of the long-term storage stability of diesel fuels rapid oxidation
stability test methods were developed and defined by standards. In Europe, the most important determination
methods are the Rancimat method and the rapid small scale oxidation test (RSSOT). Both methods differ in the
construction of the measurement device and various parameters. However, it is not clear whether the results of
both methods can be correlated in general. Due to the importance of the oxidation stability, we here describe the
results of a literature research that was carried out using 17 literature sources. Possible correlation factors be­
tween the Rancimat and the RSSOT method were analyzed, showing that a universal correlation cannot be found.
In contrast, a comparison of individual series of measurement, e.g. with the same antioxidant at various con­
centrations, can show a good correlation between the methods.

1. Introduction caused by unsaturated and polyunsaturated compounds mainly oleic

acid methyl ester (C18:1), linoleic acid methyl ester (C18:2) and lino­
Biodiesel is a renewable, biodegradable and nontoxic alternative lenic acid methyl ester (C18:3) [1,5,6].
diesel fuel. It consists of fatty acid methyl esters (FAME) of vegetable or The tendency to low oxidation stabilities of these compounds is
animal origin. The use of biodiesel has many advantages. Compared to caused by the low bond dissociation energies of allylic and bis-allylic
fossil diesel fuel biodiesel tends to lower the emissions of carbon mon­ C–H bonds [7,8]. The abstraction of hydrogen atoms can initialize a
oxide (CO), unburned hydrocarbons (HC) and particulate matter (PM) classical free radical chain reaction mechanism (autoxidation) [1,9]:
[2,3]. In addition biodiesel has a high lubricity [4], exhibits very low RH → R● + H●
sulfur and aromatic hydrocarbon contents and makes the energy supply R● + O2 → ROO●
independent from fossil fuel. ROO● + RH → ROOH + R●
Despite its advantages biodiesel is more susceptible to oxidation R● + R● → RR
compared to fossil diesel fuel. The poor oxidation stability of biodiesel is ROO● + ROO● → stable products

* Corresponding author.
E-mail address: [email protected] (F. Bär).

https://doi.org/10.1016/j.fuel.2021.120160
Received 15 April 2020; Received in revised form 1 September 2020; Accepted 5 January 2021
Available online 5 February 2021
0016-2361/© 2021 Elsevier Ltd. All rights reserved.
F. Bär et al. Fuel 291 (2021) 120160

Autoxidation is a multi-step process where primary oxidation prod­ is evaporated to isolate the quantity of filterable and adherent in­
ucts such as hydroperoxides and epoxides decompose, react with each solubles. The sum of filterable and adherent insoluble can be related to
other and form stable secondary oxidation products. Those products can the oxidation stability, expressed as milligrams per 100 ml.
be aldehydes, aliphatic alcohols, formic acid, formate esters, shorter
chain fatty acids and oligomers with molecular weights up to 1600 g/
mol [9–11]. 2.2. DIN EN 15751/DIN EN 14112 (Rancimat method)
These oxidation products can lead to higher emissions, a higher
corrosiveness, fuel filter blocking, clogging of fuel injectors, the for­ The Rancimat method was first proposed by Hadorn and Zurcher
mation of deposits in the entire fuel system and a decreasing lubricity (1974) [34]. The accelerated oxidation test is carried out by heating a
[1]. sample to 110 ◦ C in a sealed test tube while passing a stream of air
To prevent the formation of harmful oxidation products, oxidation through it. The air flow transfers volatile secondary oxidation products
inhibitors (antioxidants) need to be added to diesel fuel. Beside natural such as acetic acid and formic acid into a vessel filled with distilled
antioxidants a number of highly effective synthetic antioxidants have water. This process changes the conductivity of this aqueous phase,
been developed. However, antioxidants are consumed during storage, which can be be used to quantify oxidation products once the aging
resulting in partial loss of their efficiency against oxidative stress. After process has started. The oxidation stability can be correlated to the so-
this induction period all antioxidants have been consumed and harmful called induction time that elapsed from the start of the test until sec­
oxidation products can be formed. For the determination of the long- ondary oxidation products increase the conductivity in the water cell. In
term storage stability of fuels oxidation stability tests are needed. contrast to DIN EN 15751, which defines a sample mass of 7.5 g, DIN EN
For practical reasons oxidation stability tests have to be fast. For the 14112 defines a sample mass of 3.0 g and a smaller reaction vessel.
rapid oxidation stability determination of fuels accelerated aging has to
be used. It is common to accelerate the oxidation process by elevated 2.3. DIN EN 16568 (rapid Rancimat method)
temperature and the addition of oxygen or air.
The detection of the oxidation process can be performed in various The rapid Rancimat method is nearly identical to DIN EN 15751. In
ways. In addition to a set of proposed non-standardized oxidation sta­ contrast to DIN EN 15751 the specified aging temperature in DIN EN
bility determination tests based on detection methods like fourier 16568 is 120 ◦ C. Due to the higher temperature the oxidation rate is
transformed infrared spectroscopy (FTIR) and near infrared spectros­ increased and the formation of secondary oxidation products begins
copy (NIR) [12–19], pressurized differential scanning calorimetry (P- earlier. This results in a faster determination of the induction time.
DSC) [20–22], differential thermal analysis [17,18,23,24], thermo
gravimetric analysis [12,17,18,25], chemiluminescence [26,27], fluo­
rimetry [13,27,28], dielectric spectroscopy [10,12], UV–Vis spectros­ 2.4. DIN EN 16091 (RSSOT method)
copy [12,13,27,29], mass spectroscopy [17,30,31] and nuclear
magnetic resonance spectroscopy [18,32], a number of standardized The RSSOT method was originally presented by Neumann et al.
oxidation stability test methods for diesel fuel have been developed for (2009) [33]. In a hermetically sealed test chamber 5 ml of a sample are
the European market. pressurized with oxygen at 700 kPa and then heated to 140 ◦ C. During
The standardized accelerated oxidation stability tests accepted in the heating the pressure increases to a maximum, the oxygen is
DIN EN 590 are defined in DIN ISO 12205 and DIN EN 15751. consumed and the pressure drop can be recorded every second. The test
DIN EN 14112 is the accepted oxidation stability test in DIN EN ends when a pressure drop of 10% from maximum pressure is reached.
14214 for neat biodiesel. It is a modified version of DIN EN 15751. This The time that passed is called induction time which can be correlated to
test method is usually called the Rancimat test. the oxidation stability. The standard ASTM D7545 specifies the same
In addition, Neumann et al. [33] introduced a more rapid test method which is mainly applied in the United States.
method defined in DIN EN 16091 called the PetroOxy test.
The significant importance of oxidation stability measurements 3. Methodology
warrants a detailed analysis of a possible correlation between the two
methodologies. The literature research carried out in this study investigates whether
While many studies deal with the measurement of oxidation stability a universal correlation between Rancimat and RSSOT results can be
using Rancimat or RSSOT method, only a few reports can be found that found. For this purpose 17 literature sources were analyzed. In accor­
used both methods with the same samples. An even smaller number of dance to Dodos et al. (2014) [35] the comparison was performed by a
studies reported on a comparison of both methods showing only partial linear regression between Rancimat and RSSOT test results. The quality
agreement of the results. The present literature study investigates a of the regressions was quantified by the coefficient of determination R2
possible correlation for Rancimat and RSSOT results by analyzing and which is a statistical measure of how well the regression line approxi­
comparing 17 literature sources. mates the real data points. An R2 of 1 indicates that the regression
perfectly fits the data and that a perfect correlation is found.
2. Oxidation stability test methods The evaluated data points were taken directly from the manuscripts.
In case only diagrams were available the authors were kindly asked to
The following test methods were standardized in Europe for the provide the exact values. In case data was not provided by the authors
measurement of oxidation stability of diesel fuels. the values were taken from diagrams using the Plot Digitizer software
[36].
2.1. DIN ISO 12205 The studies analyzed can be divided into two groups regarding the
RSSOT measurements. One group of authors used the RSSOT device and
DIN ISO 12205 describes a gravimetric method for oxidation stability specification defined in DIN EN 16091 and ASTM D7545. The second
determination of petrodiesel and petrodiesel/biodiesel blends. The group of authors modified the standardized RSSOT method and used a
aging process is accelerated by adjusting a temperature of 95 ◦ C while temperature of 110 ◦ C instead of 140 ◦ C. Both groups were analyzed
oxygen is bubbled through the sample for 16 h (flow rate: 3 l/h). The individually (Figs. 4 and 5).
sample is cooled to room temperature before filtering to obtain the Following the common abbreviation, the amount of biodiesel added
filterable insolubles. The aging vessel is cleaned by trisolvent to collect to conventional fossil diesel fuel determines the blend name (e.g. 7%
all adherent residues. Both samples are then combined and the trisolvent biodiesel = B7).

2
F. Bär et al. Fuel 291 (2021) 120160

300 introduced the RSSOT method. The paper compares the Rancimat
method DIN EN 14112 and the RSSOT method analyzing a set of samples
250 with biodiesel contents of 2% (V/V), 5% (V/V), 7% (V/V), 8% (V/V),
12% (V/V), 15% (V/V), 20% (V/V), 30% (V/V), 100% (V/V) in a round
PetroOxy [min]

200 robin test. The experiments were performed with nationally and inter­
nationally sourced biodiesel samples and blends and it was found that
150 both methods correlate (Table 2).
The data analysis shows a coefficient of determination R2 of 0.9840,
100 which is a good indicator for a linear relationship (Fig. 6).
The study of Araújo et al. (2009) [37] deals with the evaluation of the
50 y = 2.2115x + 8.2646 oxidation stability of castor oil methyl ester and the influence of anti­
R² = 0.8379 oxidants. Araújo et al. (2009) [37] determined a linear correlation be­
0
tween Rancimat (DIN EN 14112) and RSSOT results for castor oil methyl
0 20 40 60 80 100 120
ester with different concentrations of 2,6-di-tert-butyl-4-methylphenol.
Rancimat [h]
The detailed analysis of the data reveals a coefficient of determina­
Fig. 1. Regression for the summarized data evaluation RSSOT (DIN EN 16091/ tion R2 of 0.9907 indicating a good fit of the linear regression curve
ASTM D7545) vs. Rancimat (DIN EN 14112/DIN EN 15751). (Fig. 7).
Hartikka et al. (2013) [38] investigated and compared DIN ISO
12205, the Rancimat method (EN15751) and the RSSOT method to
250 discuss whether these accelerated tests represent real-life conditions.
y = 2.3171x + 6.7283
Various neat fuels and blends of DIN EN 590 diesel fuel (winter and
R² = 0.7691
200 summer grade), hydrotreated vegetable oil (HVO), swedish class 1 diesel
fuel and rapeseed oil methyl ester (RME) were tested before and after
PetroOxy [min]

engine tests. Due to the fact that the Rancimat method is only applicable
150
for fuels containing more than 2% (V/V) of FAME, only EN590 B7
(before and after the engine test) and B30 were tested using this method.
100 The authors conclude that the correlation between EN ISO 12205 and
RSSOT is very weak. Since the authors tested only three samples by
50 Rancimat and RSSOT a reliable statistical analysis can’t be performed
(Fig. 8).
The effectiveness of various phenolic compounds on soybean oil
0 methyl ester and sunflower oil methyl ester oxidation stability was
0 10 20 30 40 50
compared by Deligiannis et al. (2013) [39] by a modified version of DIN
Rancimat [h]
EN 14112 (sample mass 5 g) and by RSSOT method. The phenolic
Fig. 2. Regression for Rancimat induction times < 48 h vs. RSSOT (DIN EN compounds were added at concentrations of 0.02% (m/m), 0.04% (m/
16091/ASTM D7545). m), 0.06% (m/m), 0.08% (m/m), 0.10% (m/m) and 0.12% (m/m) and
the oxidation stability was determined. A comparison or correlation of
both methods was not carried out by the authors.
700 An analysis of the data confirmed good correlations of both methods
for each individual series of phenolic compound at different concen­
600 y = 22.725x + 3.9034
R² = 0.8565 trations with a maximum coefficient of determination R2 of 0.9964 for
500 2,5bis(piperidino-methyl)hydroquinone in sunflower oil methyl ester
PetroOxy [min]

and a minimum coefficient of determination R2 of 0.5813 in soybean oil

400
methyl ester (Figs. 9 and 10). Notably, only three of fourteen data sets
300 y = 11.354x + 66.086 with varying concentrations of an additive give a coefficient of deter­
R² = 0.4273 mination R2 below 0.9100, which indicates a good correlation of Ran­
200
cimat and RSSOT method in test series investigating different
100 concentrations of phenolic compounds in sunflower oil methyl ester or
soybean oil methyl ester. The overall analysis of the results of Deli­
0 giannis et al. (2013) [39] reveals a R2 of 0.6924.
0 10 20 30 40 50
Dodos et al. (2013) [40] investigated the impact of oxidation on
Rancimat [h] lubricating properties of cottonseed oil methyl ester, sunflower oil
summarized evaluated data excluding TBHQ measurments methyl ester, sesame oil methyl ester, palm oil methyl ester, pomace
of Zhou et al. (2016) olive oil methyl ester and high erucic acid rapeseed oil methyl ester and
TBHQ measurements of Zhou et al. (2016) corresponding ultra low sulfur diesel fuel blends with 7% (V/V) of
FAME. The samples were tested by Rancimat and RSSOT method for
Fig. 3. Regression for the summarized data evaluation RSSOT (110 ◦ C) vs. their oxidation stability. It was observed that Rancimat and RSSOT re­
Rancimat (DIN EN 14112/DIN EN 15751).
sults are following the same trend concerning the induction period of the
samples. Further correlation analysis was not carried out by the authors.
4. Literature research and data evaluation The analysis of the data provided in this study results in a R2 of 0.9880
(Fig. 11).
A summary of the studies evaluated can be found in Table 1. In addition to their study of 2013, Dodos et al. presented a supple­
mentary study in 2014 [35] investigating the oxidation stability of
4.1. DIN EN 16091/ASTM D7545 vs. Rancimat castor oil methyl ester, cottonseed oil methyl ester, sunflower oil methyl
ester, sesame oil methyl ester, palm oil methyl ester, pomace olive oil
The oldest data analyzed are from Neumann et al. (2009) [33] who methyl ester and high erucic acid rapeseed oil methyl ester and blends

3
F. Bär et al. Fuel 291 (2021) 120160

250
Neumann et al. (2009)
Araújo et al. (2009)
Hartikka et al. (2013)
Deligiannis et al. (2013)
200
Dodos et al. (2013)
Dodos et al. (2014)
Botella et al. (2014)
Mutra Valle et al. (2014)
PetroOxy [min]

150 de Goede et al. (2015)

Urzędowska and Stępień (2016)
Bär et al. (2018)

100

50

0
0 20 40 60 80 100
Rancimat [h]

Fig. 4. Summary of the data evaluation RSSOT (DIN EN 16091/ASTM D7545) vs. Rancimat (DIN EN 14112/DIN EN 15751) [20,33,35,37–44].

700
Freire et al. (2012)
Medeiros et al. (2013)
600 Damasceno et al. (2013)
Sousa et al. (2014)
Melo et al. (2014)
500
Zhou et al. (2016)
PetroOxy [min]

400

300

200

100

0
0 5 10 15 20 25
Rancimat [h]

Fig. 5. Summary of the data evaluation Modified RSSOT (110 ◦ C) vs. Rancimat (DIN EN 14112/DIN EN 15751) [45–50].

with three different types of ultra low sulfur diesel fuel containing 7% model of the type y = a*xb. Individual investigations for the neat FAME
(V/V) of the above mentioned FAME. The authors presented an exten­ oxidation stability tests and the blend tests were shown.
sive and detailed correlation analysis for Rancimat and RSSOT method It was concluded that a linear regression appears to approximate the
using a linear regression model and a nonlinear power law regression real oxidation stability data slightly better in case of pure FAME. It was

4
F. Bär et al. Fuel 291 (2021) 120160

Table 1
Summary of the studies analyzed RSSOT (DIN EN 16091/ASTM D7545) vs. Rancimat (DIN EN 14112/DIN EN 15751).
Source Biofuel Biofuel Diesel Fuel Additive Additive Concentration Rancimat-
Concentration Methode
Neumann et al. (2009) not mentioned 2 %5% not mentioned – – DIN EN 14112
[33] 7%
8%
12%
15%
20%
30%
100%
Araújo et al. (2009) Castor oil methyl 100% – BHT unknown DIN EN 14112
[37] ester
Deligiannis et al. Soybean oil methyl 100% – BHT 0% (m/m) DIN EN 14112
(2013) [39] ester
Sunflower oil methyl DTBHQ 0.02% (m/m)
ester
Pyrogallol 0.04% (m/m)
4-tert-butylcatechol 0.06% (m/m)
2,5-bis(morpholinomethyl) 0.08% (m/m)
hydroquinone
2,5-bis(piperidinomethyl) 0.10% (m/m)
hydroquinone
2,5-bis(dimethyl- 0.12% (m/m)
aminomethyl)-hydroquinone
Dodos et al. (2013) Cotton seed oil 7% ULSD – – DIN EN 15751
[40] methyl ester
Sunflower oil methyl
ester
Sesame oil methyl
ester
Palm oil methyl ester
Olive oil methyl ester
Rapeseed oil methyl
ester
Hartikka et al. (2013) Rapeseed oil methyl 7% (RME) EN 590 – – DIN EN 15751
[38] ester
Rapeseed oil methyl 7% (RME aged)
ester (aged)
30% (RME)
Botella et al. (2014) Used domestic frying 100% – 4-Allyl-2,6-dimethoxyphenol 0.05% (m/m) DIN EN 14112
[41] oil methyl ester
Porkfat methyl ester Catechol 0.10% (m/m)
Soybean oil methyl 0.30% (m/m)
ester
Rapeseed oil methyl
ester
Sunflower oil methyl
ester
Coconut oil methyl
ester
(neat and 70:30
mixtures)
Dodos et al. (2014) Cotton seed oil 7% ULSD – – neat FAME: DIN
[35] methyl ester (hydrocracking) EN 14112
Sunflower oil methyl ULSD (medium Blends: DIN EN
ester hydrocracking) 15751
Sesame oil methyl ULSD (straight run)
ester
Palm oil methyl ester
Olive oil methyl ester
Rapeseed oil methyl
ester
Castor oil methyl
ester
Valle et al. (2014) Soybean oil methyl 100% – BHT 0.25% (m/m) DIN EN 14112
[20] ester
Soybean oil ethyl Baynox Plus 0.50% (m/m)
ester
0.10% (m/m)
De Goede et al. (2015) Rapeseed oil methyl 2% Summer Type GTL – – neat FAME: DIN
[42] ester EN 14112
Rapeseed oil methyl 5% Winter Type GTL Blends: DIN EN
ester (aged) 15751
10% EN 590
15% US 2D diesel
20%
(continued on next page)

5
F. Bär et al. Fuel 291 (2021) 120160

Table 1 (continued )
Source Biofuel Biofuel Diesel Fuel Additive Additive Concentration Rancimat-
Concentration Methode
100%
Urzędowska and Rapeseed oil methyl 5% DIN EN 590 B5 – – DIN EN 15751
Stępień (2016) [43] ester 30% (winter and summer
grade)
Bär et al. (2018) [44] Rapeseed oil methyl 0% ULSD various hydrazides various molar- and mass- DIN EN 15751
ester concentrations
Distilled rapeseed oil 2% BHT
methyl ester
7% 2,2′ -Methylenebis(6-tert-butyl-
4-methylphenol)
10% Pyrogallol
20% Propyl gallate
50%
100%

Table 2
Summary of the studies analyzed RSSOT (110 ◦ C) vs. Rancimat (DIN EN 14112/DIN EN 15751).
Source Biofuel Biofuel Diesel Additive Additive Concentration Rancimat-
Fuel Methode
Concentration
Freire et al. (2012) [45] neat oil and quaternary 100% – – – DIN EN 14112
mixtures:
Babassu oil
Cotton seed oil
Soybean oil
Damasceno et al. (2013) Soybean oil ethyl ester 100% – Caffeic acid 0.10% (m/m) DIN EN 14112
[47] Ferulic acid
TBHQ
Medeiros et al. (2013) Soybean oil methyl ester 100% – four Rosemary up to 2500 ppm in various DIN EN 14112
[46] Extracts combinations
Citric acid
TBHQ
Melo et al. (2014) [49] Grape seed oil 100% – – – DIN EN 14112
Sesame oil
Oiticica oil
Jatropha oil
Sousa et al. (2014) [48] Soybean oil methyl ester 100% – Curcumin 0.05% (m/m) DIN EN 14112
β-Carotene 0.10% (m/m)
0.15% (m/m)
0.20% (m/m) in combination
Zhou et al. (2016) [50] UCOME 100% – BHA 0.02% (m/m) DIN EN 14112
BHT 0.05% (m/m)
TBHQ 0.10% (m/m)
Pyrogallol
Propyl gallate

70 35
60 30
PetroOxy [min]
PetroOxy [min]

50 25
40 20
30 15
20 10
y = 1.6x + 5.2 y = 3.2604x + 9.0744
10 5
R² = 1.0 R² = 0.9907
0 0
0 10 20 30 40 0 2 4 6 8
Rancimat [h] Rancimat [h]

Fig. 6. Data analysis for results of Neumann et al. (2009) [33]. Fig. 7 Data analysis for results of Araújo et al. (2009) [37].

suggested that the eight hour oxidation stability limit given in DIN EN In case of the blends almost similar observations were reported,
14214 for Rancimat method corresponds to approximately 25 min although a higher variation of data was mentioned. Nevertheless, a good
considering the linear regression and 27 min considering the power law correlation between both methods was found, expressed in an R2 value
regression in the RSSOT test. The linear correlation was demonstrated to of 0.9028. The Rancimat oxidation stability limit of 20 h defined in DIN
be nearly perfect (R2 = 0.9991). EN 590 corresponds to RSSOT induction periods of 52 min (linear

6
F. Bär et al. Fuel 291 (2021) 120160

120 120
100 100
PetroOxy [min]

PetroOxy [min]
80
80
60
60
40
40
20 y = 2.267x + 9.3459
R² = 0.9911 y = 1.7633x + 10.1090
20 R² = 0.9880
0
0 10 20 30 40 50
0
Rancimat [h] 0 20 40 60
Fig. 8. Data analysis for results of Hartikka et al. (2013) [38]. Rancimat [h]

Fig. 11. Data analysis for results of Dodos et al. (2013) [40].

Sunflower oil methyl ester

80 2,5-bis(dimethylaminomethyl)hydroquinone
70 2,5-bis(piperidinomethyl)hydroquinone 250
2,5-bis(morpholinomethyl)hydroquinone
4-tert-butylcatechol
60
PetroOxy [min]

Pyrogallol 200

PetroOxy [min]
DTBHQ
50 BHT
40 150
30
100
20
10 50 y = 2.3552x + 5.0209
0 R² = 0.9249
0 5 10 15 20 25 0
0 20 40 60 80 100
Rancimat [h] Rancimat [h]
Fig. 9. Data analysis for results of Deligiannis et al. (2013) [39].
Fig. 12. Data analysis for results of Dodos et al. (2014) [35].

Soybean oil methyl ester blends with rapeseed oil methyl ester and soybean oil methyl ester was
90 2,5-bis(dimethylaminomethyl)hydroquinone evaluated. Amongst other results, the authors introduced a correlation
2,5-bis(piperidinomethyl)hydroquinone
80 2,5-bis(morpholinomethyl)hydroquinone study of Rancimat and RSSOT method that accounts for factors such as
4-tert-butylcatechol
70 Pyrogallol biodiesel composition, the additives that were used and their concen­
PetroOxy [min]

60 DTBHQ
BHT
trations. The analysis was performed using an ANOVA (analysis of
50 variance) test. It was concluded that the composition of the biodiesel
used does not significantly affect the correlation between both methods.
40
Furthermore, no significant differences can be found in correlation
30 regarding the additives and their concentrations used in the study. Most
20 importantly the authors concluded that there is a significant difference
10 in the linear regression models regarding the intercept and the slope of
0 the regression depending on whether an additive was used or not. Based
0 10 20 30 on these findings, a regression model was developed (Fig. 13):

Rancimat [h]

Fig. 10. Data analysis for results of Deligiannis et al. (2013) [39].
60
regression with additives regression for all
regression) and 55 min (power law regression).
50 data points
Also for the combined test results of the neat FAME and the blends a
satisfactory correlation of the data was displayed by a coefficient of
PetroOxy [min]

40
determination R2 of 0.9245 for the linear model.
Dodos et al. (2014) [35] concluded that the data show strong evi­ 30
dence of correlation for the methods. However, they note that additional
studies seem to be necessary to verify whether a linear relationship of 20 regression
the data can be confirmed (Fig. 12). without additives
The oxidation stability of biodiesel samples and of several blends has 10
been measured by Rancimat and RSSOT method by Botella et al. (2014)
0
[41] with the aim of investigating a possible correlation of both 0 5 10 15 20
methods. Feedstock for the biodiesel was used frying olive oil, pork fat,
Rancimat [h]
soybean oil, rapeseed oil, sunflower oil and coconut oil. Additionally the
influence of 4-allyl-2,6-dimethoxyphenol and catechol on biodiesel Fig. 13. Data analysis for results of Botella et al. (2014) [41].

7
F. Bär et al. Fuel 291 (2021) 120160

Rancimat [min] = (31.89 − 20.63* f)*RSSOT [min] + ( − 214.65 dependent on the base fuel and cannot be applied generally.
It has to be noted that neat and blended RME oxidation test results
+ 319.68 * f)
were included in the analysis although two different Rancimat methods
In this equation f equals 0 if no additive is used and 1 if an additive were used for induction time determination – DIN EN 14112 for the neat
was mixed with biodiesel. It can thus be concluded that the correlation RME samples and DIN EN 15751 for the blends. In addition it was not
cannot be drawn by a single linear relationship. Additionally the authors exactly mentioned which data points were used for the correlation
analyzed the data of Araújo et al. (2009) [37] in terms of the regression analysis.
model and found that all values but one are within a confidence interval An analysis of the data sets obtained from the measurement series
of 95%. It was concluded that the RSSOT method seems to be an alter­ performed with fresh RME and preaged RME in different concentrations
native method for measuring the oxidation stability of biodiesel samples and without consideration of the neat biodiesel samples shows co­
and the linear correlation found between the Rancimat and RSSOT could efficients of determination R2 between 0.4363 and 0.9967 (Figs. 15 and
be used for the prediction of biodiesel oxidation stability in future 16). This indicates that a correlation is not applicable in all cases, which
standards and norms. However, the authors point out that more data are is in agreement with the results of the author’s analysis.
needed. The prediction of threats caused by the low stability of diesel fuel
Murta Valle et al. (2014) [20] performed a comparative study of with a high concentration of biodiesel for engine injector operation was
biodiesel oxidation stability using Rancimat, RSSOT and low pressurized investigated by Urzędowska and Stępień (2016) [43]. For this purpose
differential scanning calorimetry (low P-DSC). The authors used soy­ diesel fuel blends with 5% (V/V) unspecified biodiesel (winter and
bean oil methyl ester and soybean oil ethyl ester with the commonly summer grade) and blends with further 25% (V/V) rapeseed oil methyl
added antioxidants 2,6-di-tert-butyl-4-methylphenol (BHT) and [2,2′ - ester – resulting in B30 – were stored for 16 weeks before testing the fuel
methylenebis (4-methyl-6-tert-butylphenol)] (Baynox Plus) at concen­ in an engine. After 0, 8, 12 and 16 weeks, samples of the ageing summer
trations of 0 ppm, 250 ppm, 500 ppm and 1000 ppm, respectively. Due grade blends were subjected to oxidation stability analysis to gain a
to the fact that the three methods are based on the measurement of better understanding of the long-term storage stability. The oxidation
different parameters the resulting induction times (IT) were investigated stability was tested by DIN EN 12205, Rancimat method (DIN EN 15751
regarding their change when antioxidants were added. The changes and DIN EN 16568) and RSSOT method (DIN EN 16091). The authors
were expressed as the percentage difference compared to the induction found that the Rancimat (DIN EN 15751) and RSSOT results correlated
times obtained without any antioxidants (IT0) calculated by: ΔIT = 100 well, without providing a detailed quantification.
(IT - IT0)/IT. The detailed analysis of the data results in a coefficient of determi­
It was found that the results of Rancimat and RSSOT are quite nation R2 of 0.9813 for a linear regression (Fig. 17).
similar, except for the soybean oil ethyl ester in the presence of Baynox The most recent data analyzed comes from Bär et al. (2018) [44]
Plus. A reason for the different behavior of the data set was not presenting a study dealing with Rancimat (DIN EN 15751) and RSSOT
proposed. measurements of rapeseed oil methyl ester and distilled rapeseed oil
A direct correlation of the Rancimat and RSSOT data sets were not methyl ester stabilized with various additives such as 2,6-di-tert-butyl-4-
offered by Murta Valle et al. (2014) [20] but the analysis results in a methylphenol (BHT), pyrogallol, propylgallate and various hydrazides
coefficient of determination R2 of 0.8352 (Fig. 14). at different concentrations.
Various blends of Gas-to-Liquid (GTL) fuel (summer and winter In addition a series of blends with fossil diesel fuel and rapeseed oil
quality), special blended commercial reference diesel fuel (US 2D) and methyl ester at concentrations of 2% (V/V), 7% (V/V), 10% (V/V), 20%
diesel fuel according to DIN EN 590 mixed with rapeseed oil methyl (V/V), 50% (V/V) and 100% (V/V) were prepared.
ester (RME) and pre-aged rapeseed oil methyl ester (RME), respectively, The authors found a number of differences between Rancimat and
were analyzed regarding their oxidation stability by de Goede et al. RSSOT results. In RSSOT tests isonicotinic hydrazide was the most
(2015) [42]. The study aimed at providing an overview about the sta­ effective hydrazide, whereas the Rancimat tests stated out that manisic
bility performance of different paraffinic and petroleum-derived diesel hydrazide performed best at the same molar concentration. The overall
fuels in combination with biodiesel. Therefore, oxidation stability analysis of the results showed significant differences in the oxidation
measurements such as Rancimat and RSSOT tests were carried out. A stabilizing potential of the additives dependent on the test method.
correlation analysis between the Rancimat and RSSOT results for binary A detailed analysis of a correlation between both methods was not
blends with fresh RME was also part of the study. The biodiesel content performed by the authors. However, a correlation analysis of the indi­
was varied between 2% (V/V) and 20% (V/V) and the data sets were vidual measurement series reveals coefficients of determination R2 from
correlated, resulting in coefficients of determination R2 between 0.6380 0.6405 for the tests performed with various hydrazides to 0.9950 for
and 0.9879. The authors concluded that in any specific series a linear tests carried out with different concentrations of BHT (Fig. 18). How­
correlation can be found. However, the relationship seemed to be highly ever, Rancimat measurements of distilled RME were not considered due

35
400
Summer Type GTL + RME high stability
30 350 Winter type GTL diesel + RME high stability
US 2D diesel + RME high stability
25 300
PetroOxy [min]

EN 590 + RME high stability

PetroOxy [min]

20 250
200
15
150
10
y = 2.5362x + 7.7799 100
5 R² = 0.8352 50
0 0
0 2 4 6 8 10 0 20 40 60 80 100 120
Rancimat [h] Rancimat [h]

Fig. 14. Data analysis for results of Murta Valle et al. (2014) [20]. Fig. 15. Data analysis for results of de Goede et al. (2015) [42].

8
F. Bär et al. Fuel 291 (2021) 120160

400 0.4822 for the set of data points (Fig. 19).

Summer Type GTL + RME preaged A study investigating the efficient formulation of antioxidants in
350 Winter type GTL diesel + RME preaged
biodiesel was reported by Medeiros et al. (2013) [46]. The authors
300 US 2D diesel + RME preaged
tested different concentrations and combinations of antioxidants in
PetroOxy [min]

EN 590 + RME preaged

250 soybean oil ethyl ester and cottonseed oil methyl ester. However, only
200 for soybean oil ethyl ester Rancimat (DIN EN 14112) and RSSOT data
was presented simultaneously. The authors compared both methods and
150 showed a highly diverging behavior for the different antioxidant com­
100 binations. Tests with different concentrations of rosemary ethanol
50 extract and combinations of rosemary ethanol extract and citric acid
exhibited relatively high R2 values between 0.8213 and 0.8806. These
0
results are in contrast to tests performed with different concentrations of
0 10 20 30 40 50
citric acid, tert-butylhydroquinone and combinations of both antioxi­
Rancimat [h] dants, which determines R2 values between − 0.4842 and − 0.1908. An
Fig. 16. Data analysis for results of de Goede et al. (2015) [42]. explanation for the diverging results was not suggested by the authors.
An overall evaluation of the data points available results in a coef­
ficient of determination R2 of 0.9330 (Fig. 20).
70 Damasceno et al. (2013) [47] presented an investigation of the effect
of caffeic acid (CA), ferulic acid (FA) and tert-butylhydroquinone
60 (TBHQ) on the stability of soybean ethyl ester (SEE) during storage. For
this purpose samples of soybean ethyl ester without any additive and
PetroOxy [min]

50
with caffeic acid, ferulic acid and tert-butylhydroquinone at a concen­
40 tration of 1000 ppm were stored for three month, with samples taken
each month. The authors discussed the ratio between the induction
30 periods obtained at the beginning of storage and after 30 days of the
20 samples with caffeic acid and found a good correlation between the
y = 1.5885x + 5.1588 change of induction times in Rancimat and RSSOT values. Further cor­
10 relations between both methods were not carried out.
R² = 0.9813
A detailed data correlation analysis results in a coefficient of deter­
0
0 10 20 30 40 mination R2 of 0.9327 for all data points. The analysis of each individual
Rancimat [h] storage series gives coefficients of determination of R2SEE = 0.9978,
R2SEE CA = 0.9210, R2SEE FA = 0.9496 and R2SEE TBHQ = 0.9691 (Fig. 21).
Fig. 17. Data analysis for results of Urzędowska and Stępień (2016) [43]. The use of the natural antioxidants β-carotene and curcumin in
soybean oil methyl ester was evaluated during a 180 days storage time
by de Sousa et al. (2014) [48]. The tests were performed at concentra­
120 tions of 500 ppm, 1000 ppm, 1500 ppm and with combinations of both
RME with various additives (7,26 mmol/kg)
B2 - B100 antioxidants at concentrations of 1000 ppm and 1500 ppm for each of
100 the additives. The authors point out that, in contrast to β-carotene,
dist. RME with Propylgallate
dist. RME with Pyrogallol curcumin showed a good antioxidant activity determined by 2,2-
PetroOxy [min]

80
RME with BHT diphenyl-1-picrylhydrazyl (DPPH) method, which gives an indication
RME with BH about the radical nature of a reaction. In addition to this evaluation
60
Rancimat and RSSOT tests were carried out. The average induction
times were correlated by a linear regression model, revealing an R2 of
40
0.94954. Based on these results, Sousa et al. (2014) [48] drew the
20 conclusion that the RSSOT method can be used as an alternative tech­
nique for determining the induction time of biodiesel samples. Moreover
0 it was suggested that the former minimum limit of oxidation stability for
0 10 20 30 FAME of six hours (DIN EN 14214:2010–04) measured by the Rancimat
Rancimat [h] corresponds to 133 min measured by the RSSOT method.

Fig. 18. Data analysis for results of Bär et al. (2018).

350
to an induction time of 0 h which is not a reliable value for Rancimat 300
measurements.
PetroOxy [min]

250
4.2. Modified RSSOT method (110 C) vs. Rancimat ◦
200

Freire et al. (2012) [45] presented a study about the oxidative sta­ 150
bility and flow properties of quaternary mixtures of vegetable oils such 100
as soybean oil, cotton oil, jatropha curcas oil and babassu oil. Beside
50 y = 12.346x + 113.96
Rancimat (DIN EN 14112) and RSSOT measurements, tests were per­ R² = 0.4822
formed by pressure differential scanning calorimetry (PDSC). A dedi­ 0
cated comparison between the methods was not done by Freire et al. 0 5 10 15
(2012) [45]. Rancimat [h]
However, the data shows a weak correlation between Rancimat and
RSSOT results expressed by a low coefficient of determination R2 of Fig. 19. Data analysis for results of Freire et al. (2012) [45].

9
F. Bär et al. Fuel 291 (2021) 120160

400 However, the analysis of the sesame oil, oiticica oil, jatropha oil and
grape oil induction times, results in an R2 of 0.3557 which indicates that
350
the methods are not to be correlated for the vegetable oils tested by Melo
300 et al (2014) [49] (Fig. 23). Nevertheless, the authors found consistency
PetroOxy [min]

250 in the data.

Zhou et al. (2016) [50] did an investigation of the oxidation stability
200
of used cooking oil methyl ester (UCOME) stabilized with pyrogallol
150 (PY), propylgallate (PG), butylated hydroxyanisol (BHA), tert-butylhy­
100 droquinone (TBHQ) and 2,6-di-tert-butyl-4-methylphenol (BHT) at
y = 13.5921x + 49.9157 concentrations of 200 ppm, 500 ppm and 1000 ppm. Rancimat tests
50
R² = 0.9330 were carried out according to DIN EN 14112 and RSSOT tests were
0 carried out according to ASTM 7545 (identical to DIN EN 16091) but at
0 5 10 15 20 25 various temperatures (90 ◦ C, 100 ◦ C, 110 ◦ C, 120 ◦ C, 130 ◦ C and 140 ◦ C).
Rancimat [h] The authors found a significant difference between Rancimat and RSSOT
(110 ◦ C) measurements performed with UCOME stabilized with TBHQ.
Fig. 20. Data analysis for results of Medeiros et al. (2013) [46].
The additive showed the highest antioxidant activity in Rancimat tests,
whereas TBHQ shows the lowest effect in RSSOT tests. All other mea­
450 surements are showing consistent results. For the comparison of both
SEE methods the authors determined a stabilization factor (SF) to express the
400
SEE CA antioxidant efficiency of the antioxidants: SF = IPA/IP0, where IPA is the
350 induction period in the presence of the antioxidant, and IP0 is the in­
PetroOxy [min]

SEE FA
300
R² = 0.9327 duction period of UCOME without any antioxidants. It was found that
SEE TBHQ the antioxidants performed much better in the RSSOT method than in
250
200 the Rancimat tests. It was concluded that the Rancimat method may
underestimate the antioxidant capacity of the phenolic compounds
150
investigated. Zhou et al. (2016) [50] could not rule out the possibility
100 that the contradictory antioxidant effects of TBHQ in RSSOT compared
50 to Rancimat tests is mainly because of a possible deactivation or
0 decomposition by the higher pressure of the RSSOT method. However, a
0 5 10 15 comparison with the data observed by Deligiannis et al. (2014) for
Rancimat [h] DTBHQ [39] (Figs. 9 and 10) shows a similar effect of low influence on
RSSOT results.
Fig. 21. Data analysis for results of Damasceno et al. (2013) [47]. An analysis of the data, excluding the TBHQ results, shows a coef­
ficient of determination R2 of 0.9460. The analysis of the results ob­
In contrast to the author’s procedure the data analysis shown in tained from the stability tests performed with TBHQ shows a coefficient
Fig. 22 correlates with the individual induction times of the storage of determination R2 of 0.9677. A combined analysis of the data does not
experiments, but not the average induction times (R2 = 0.9218) appear to be useful (R2 = 0.0696) (Fig. 24).
(Fig. 22).
Melo et al. (2014) [49] presented a study of thermal and oxidative 5. Data interpretation
stability of nonconventional oils for biodiesel production such as
andiroba oil, babassu oil, sesame oil, oiticica oil, jatropha oil and grape The evaluated data shows that good correlations between Rancimat
oil. The tests were carried out using the pure oils without trans­ and RSSOT results can be found in in most of the individual series of
esterification to the corresponding FAME. The oxidative stability was measurement. However, a number of studies presented showed low
tested by Rancimat (DIN EN 14112 with 3 g sample mass) and by the coefficients of determination [20,42,44–46,49]. The reason for the low
RSSOT method to propose alternative sources of oils suitable for bio­ coefficients of determination in those studies might be found in the very
diesel production. The authors note that the determination of the in­ different fuels or additives that were used. The results presented by Zhou
duction times of andiroba oil and babassu oil was not detectable due to et al. (2016) [50] demonstrate that very good correlations can be ob­
an induction time longer than 30 h, although DIN EN 14112:2014-07 tained for UCOME mixed with a number of phenolic antioxidants.
defines a maximum applicability of the Rancimat of 48 h. However, the mixture of UCOME with TBHQ results in a completely

200 300

250 y = 10.894x + 33.641

150 R² = 0.3557
PetroOxy [min]
PetroOxy [min]

200

100 150

100
50
y = 19.424x + 7.0671 50
R² = 0.9218
0 0
0 2 4 6 8 10 0 5 10 15
Rancimat [h] Rancimat [h]

Fig. 22. Data analysis for results of de Sousa et al. (2014) [48]. Fig. 23. Data analysis for results of Melo et al. (2014) [49].

10
F. Bär et al. Fuel 291 (2021) 120160

700 antioxidants and ageing products cannot be lost by evaporation. The

y = 24.524x + 40.419 Rancimat test extracts short chain oxidation products for the determi­
PetroOxy (110 °C) [min] 600 nation of the induction period. However, also antioxidants might vola­
R² = 0.946
500 tilize during the early hours of analysis in accelerated tests such as the
Rancimat method.
400 Santos et al. (2012) [52] investigated the thermal stability of anti­
300 y = 2.2384x + 53.789 oxidants by isostatic and dynamic thermogravimetric measurements
R² = 0.9677 and found that a mass loss of common antioxidants like BHA, BHT and
200 TBHQ is initiated at temperatures lower than 100 ◦ C. In addition the
100 isostatic thermogravimetric analysis showed a major difference between
BHA, BHT and TBHQ. The time required for the complete vaporization
0 of TBHQ is more than ten times larger than the required vaporization
0 20 40 60
time of BHA and BHT. In addition Phang and Dollimore (2001) showed,
Rancimat [h]
that the vapor pressure of BHT and propenyl guaethol can be related to
Fig. 24. Data analysis for results of Zhou et al. (2016) [50]. thermogravimetric data.
Therefore, the vapor pressure of the antioxidants and the formation
different linear regression model (Fig. 24). Bär et al. (2018) [44] showed of azeotropes might influence the correlation between Rancimat and
that RME mixed with different additives such as phenolic antioxidants RSSOT measurements which could explain different correlation results
and hydrazides at the same molar concentration results in poor linear between various antioxidants. A possible relation of the slope of the
regression models. correlation regression and the vapor pressure or thermogravimetric
Fig. 1 shows the overall regression model for the comparison of the properties of antioxidants investigated in the Rancimat test should be
Rancimat results and the RSSOT (DIN EN 16091/ASTM D7545) mea­ done in further studies or by literature research.
surements. The analysis of those data points reveals a coefficient of In addition, no universal correlation can be found whether DIN EN
determination R2 of 0.8379. According to the minimum induction 15751 and DIN EN 14112 were analyzed separated or together.
period of eight hours for neat FAME, defined in DIN EN 14214 for the All data analyzed imply that the measurements were performed with
Rancimat test, the minimum RSSOT induction period corresponds to a proper calibration of each device, suitable purity of utilized oxygen, air
approximately 26 min. This is in good agreement with the results of or water and no chemical variation of the sample between the Rancimat
Dodos et al. (2014) [35]. and RSSOT measurements.
Taking into account that the Rancimat method defined in DIN EN In summary, the data interpretation shows, that a universal corre­
15751 and DIN EN 14112 can only be applied reliably if the induction lation between Rancimat and RSSOT measurements cannot be found.
period is below 48 h a coefficient of determination R2 of 0.7691 can be Individual series of measurement with the same antioxidant at various
obtained (Fig. 2) and the eight hour limit for the Rancimat method concentrations can give a good correlation between the methods.
corresponds to 25.3 min in the RSSOT test. However, the prediction of Rancimat induction times by RSSOT mea­
Fig. 3 shows the regression model for the summarized data evalua­ surements can hardly be performed if no correlation information is
tion of the Rancimat measurements and the RSSOT tests performed at available for the combination of a specific antioxidant and a specific
110 ◦ C. Excluding the experiments of Zhou et al. (2016) [50] performed fuel.
with TBHQ the determination of R2 results in a value of 0.8565. The
eight hour oxidation stability limit given in DIN EN 14214 corresponds 6. Conclusions and summary
to approximately 186 min in the RSSOT test at 110 ◦ C. Considering the
tests of Zhou et al. (2016) [50] performed with TBHQ a much lower The presented analysis of 17 literature sources clearly shows that a
coefficient of determination R2 of 0.4273 can be found. universal correlation between Rancimat (DIN EN 15751/DIN EN14112)
The data presented by Deligiannis et al (2013) [39] and Zhou et al. and RSSOT induction times cannot be found.
(2016) [50] reveal another interesting effect. It can be found that tert- However, it seems to be possible to find correlations in individual
butylhydroquinone and 2,5-di-tert-butylhydroquinone have much lower test series with the same antioxidant at different concentrations. It might
impact on RSSOT test results than on Rancimat test results compared to be possible to predict Rancimat induction times by RSSOT measure­
other phenolic antioxidants such as 4-tert-butylcatechol. At a concen­ ments if correlation information is available for the combination of a
tration of 0.12% (m/m) 2,5-di-tert-butylhydroquinone increases the specific antioxidant and a specific fuel.
RSSOT induction time from 10.0 min to 12.4 min (24% increase) in Further studies examining or comparing Rancimat and RSSOT results
soybean oil methyl ester, whereas it increases the Rancimat induction should consider that antioxidants and oxidation products can be dragged
time from 2.5 h to 11.6 h (364% increase). out of the reaction vessel of the Rancimat device at 110 ◦ C [51] and
In contrast, 4-tert-butylcatechol results in a RSSOT induction time therefore cannot act as an oxidation inhibitor anymore. It seems to be
increase from 10.0 min to 41.4 min (314% increase) and a Rancimat useful to define a limit for RSSOT induction times in DIN EN 590 and
induction time increase from 2.5 h to 11.6 h (364% increase). DIN EN 14214 because the necessary concentrations of antioxidants for
The reason for this behavior might be found in different mechanisms long term storage of diesel fuel might be overestimated by the Rancimat
of action of the antioxidants or interactions between the antioxidants test for more volatile antioxidants. Recommendations for universal
and other substances present in the fuel depending on the method of converting factors between induction times generated by Rancimat and
accelerated oxidation. It is known that antioxidative action and syner­ RSSOT method are not expedient.
gistic or antagonistic effects strongly depend on the temperature and the
partial oxygen pressure [23]. Declaration of Competing Interest
In addition, the volatility of antioxidants and oxidation products
should be taken into account. Studies investigating antioxidant activity The authors declare that they have no known competing financial
in frying oil showed that antioxidants are lost not only by reaction with interests or personal relationships that could have appeared to influence
radicals formed by oil oxidation but also by decomposition and evapo­ the work reported in this paper.
ration before they are able to exert antioxidant activity [51].
Since the RSSOT method uses a hermetically sealed test chamber,

11
F. Bär et al. Fuel 291 (2021) 120160

References [27] Krishnakumar U, Sivasubramanian V. Spectroscopic techniques for stability and

quality assessment of biodiesel. Asian J Chem 2017;29(1):115–8.
[28] Michels FS, Trindade MAG, Falcão EA, Guimarães RCA, Oliveira SL, Caires ARL.
[1] Kumar N. Oxidative stability of biodiesel: causes, effects and prevention. Fuel
The effect of the excitation light intensity during on-line monitoring of biodiesel by
2017;190:328–50.
fluorescence spectroscopy. Fuel 2017;193:395–400.
[2] EPA. A comprehensive analysis of biodiesel impacts on exhaust emissions. EPA420-
[29] Souza FHN, de Almeida LR, Batista FSCL, Rios MAS. UV-visible spectroscopy study
P-02-001. 2002.
of oxidative degradation of sunflower biodiesel. Energy Sci Technol 2011;2(2):
[3] Bär F, Eskiner M, Fan Z, Rossner M, Munack A, Krahl J. In-situ measurement of fuel
56–61.
properties by means of dielectric- and fluorescence spectroscopy. Fachvortrag 12.
[30] Pereira GG, Alberici RM, Ferreira LL, Santos JM, Nascimento HL, Eberlin MN et al.
FAD-Konferenz 5. und 6. November 2014. 2014.
A screening method to evaluate soybean oil-based biodiesel oxidative quality
[4] Knothe G, Steidley KR. Lubricity of Components of Biodiesel and Petrodiesel. The
during its shelf life. J Am Oil Chem Soc 2015;92(7):967–74.
Origin of Biodiesel Lubricity †. Energy Fuels 2005;19(3):1192–200.
[31] Pereira GG, Alberici RM, Fernandes GD, Cunha IBS, Eberlin MN, Dobarganes MC
[5] McCormick RL, Ratcliff M, Moens L, Lawrence R. Several factors affecting the
et al. Ambient sonic-spray ionization mass spectrometry for rapid monitoring of
stability of biodiesel in standard accelerated tests. Fuel Process Technol 2007;88
secondary oxidation products in biodiesel. Eur J Lipid Sci Technol 2014;116(8):
(7):651–7.
952–60.
[6] Karavalakis G, Stournas S, Karonis D. Evaluation of the oxidation stability of
[32] Knothe G. Analysis of oxidized biodiesel by1H-NMR and effect of contact area with
diesel/biodiesel blends. Fuel 2010;89(9):2483–9.
air. Eur J Lipid Sci Technol 2006;108(6):493–500.
[7] Blanksby SJ, Ellison GB. Bond dissociation energies of organic molecules. Acc
[33] Neumann A, Jebsen T, Wierzbicki V. A method for determining oxidation stability
Chem Res 2003;36(4):255–63.
of petrodiesel, biodiesel, and blended fuels. Am Lab 2008(40):20–6.
[8] Zellner R. S.W. Benson: thermochemical kinetics, 2nd Ed. John Wiley & Sons, New
[34] Hadorn H, Zürcher K. Bestimmung der Oxidationsstabilität von Ölen und Fetten.
York-London-Sydney-Toronto 1976. 320 Seiten, Berichte der Bunsengesellschaft
Deutsche Lebensmittel Rundschau 1974;70:57–65.
für physikalische Chemie 1977;81(9):877–8. https://doi.org/10.1002/
[35] Dodos GS, Karonis D, Zannikos F, Lois E. Assessment of the oxidation stability of
bbpc.19770810919.
biodiesel fuel using the Rancimat and the RSSOT methods. SAE Int 2014. 2014-01-
[9] Flitsch S, Neu PM, Schober S, Kienzl N, Ullmann J, Mittelbach M. Quantitation of
2758. doi.org/10.4271/2014-01-2758.
Aging Products Formed in Biodiesel during the Rancimat Accelerated Oxidation
[36] sourceforge. Plot Digitizer 2017;https://sourceforge.net/projects/plotdigitizer/.
Test. Energy Fuels 2014;28(9):5849–56.
[37] Araújo SV, Luna FMT, Rola EM, Azevedo DCS, Cavalcante CL. A rapid method for
[10] Eskiner M, Bär F, Rossner M, Munack A, Krahl J. Determining the aging degree of
evaluation of the oxidation stability of castor oil FAME: influence of antioxidant
domestic heating oil blended with biodiesel by means of dielectric spectroscopy.
type and concentration. Fuel Process Technol 2009;90(10):1272–7.
Fuel 2015;143:327–33. https://doi.org/10.1016/j.fuel.2014.10.080.
[38] Hartikka T, Kiiski U, Kuronen M, Mikkonen S. Diesel fuel oxidation stability: a
[11] Carvalho AL de, Cardoso EA, da Rocha GO, Teixeira LSG, Pepe IM, Grosjean DM.
comparative study. SAE International 2013; 2013-01-2678.
Carboxylic acid emissions from soybean biodiesel oxidation in the EN14112
[39] Deligiannis A, Anastasakos A, Anastopoulos G, Dodos GS, Zannikos F. Effectiveness
(Rancimat) stability test. Fuel 2016;173:29–36.
of Various Phenolic Compounds (Commercial and Non-Commercial) on Biodiesel
[12] Zhou J, Xiong Y, Gong Y, Liu X. Analysis of the oxidative degradation of biodiesel
Oxidation Stability 2013.
blends using FTIR, UV–Vis, TGA and TD-DES methods. Fuel 2017;202:23–8.
[40] Dodos GS, Deligiannis A, Karonis D, Zannikos F. Impact of oxidation on lubricating
[13] Svilans M, Blūms A, Kampare R. Spectroscopic monitoring of biodiesel aging.
properties of biodiesel blends. SAE Int J Fuels Lubr 2013;6(3):657–67. 2013-01-
Mater Sci Appl Chem 2013;28(28):85.
2596. doi.org/10.4271/2013-01-2596.
[14] Peer MS, Kasimani R, Rajamohan S, Ramakrishnan P. Experimental evaluation on
[41] Botella L, Bimbela F, Martín L, Arauzo J, Sánchez JL. Oxidation stability of
oxidation stability of biodiesel/diesel blends with alcohol addition by rancimat
biodiesel fuels and blends using the Rancimat and PetroOXY methods. Effect of 4-
instrument and FTIR spectroscopy. J Mech Sci Technol 2017;31(1):455–63.
allyl-2,6-dimethoxyphenol and catechol as biodiesel additives on oxidation
[15] Pinho DMM, Santos VO, dos Santos VML, Oliveira MCS, da Silva MT, Piza PGT
stability. Front Chem 2014;2:43.
et al. Evaluating the use of EN 14078 for determination of biodiesel in diesel blends
[42] Goede S de, Wilken C, Ajam M, Roets P, Engelbrecht P, Woolard C. A comparison of
sold in the Brazilian market. Fuel 2014;136:136–42.
the stability performance of blends of paraffinic diesel and petroleum-derived
[16] Araújo SV, Rocha BS, Luna FMT, Rola EM, Azevedo DCS, Cavalcante CL. FTIR
diesel, with RME biodiesel using laboratory stability measurement techniques. J
assessment of the oxidation process of castor oil FAME submitted to PetroOXY and
Fuels 2015;2015:1–15.
Rancimat methods. Fuel Process Technol 2011;92(5):1152–5.
[43] Urzędowska W, Stępień Z. Prediction of threats caused by high FAME diesel fuel
[17] Bacha K, Ben-Amara A, Vannier A, Alves-Fortunato M, Nardin M. Oxidation
blend stability for engine injector operation. Fuel Process Technol 2016;142:
stability of diesel/biodiesel fuels measured by a PetroOxy device and
403–10.
characterization of oxidation products. Energy Fuels 2015;29(7):4345–55.
[44] Bär F, Hopf H, Knorr M, Krahl J. Rancimat and PetroOxy oxidation stability
[18] Conceição MM, Fernandes VJ, Araújo AS, Farias MF, Santos IMG, Souza AG.
measurements of rapeseed oil methyl ester stabilized with hydrazides and
Thermal and oxidative degradation of castor oil biodiesel. Energy Fuels 2007;21
antioxidants. Fuel 2018;232:108–13. https://doi.org/10.1016/j.fuel.2018.05.095.
(3):1522–7.
[45] Freire LMS, Filho JRC, Moura CVR, Soledade LEB, Stragevitch L, Cordeiro ÂMTM
[19] Lira LFB de, Albuquerque MS de, Pacheco JGA, Fonseca TM, Cavalcanti EHdS,
et al. Evaluation of the oxidative stability and flow properties of quaternary
Stragevitch L et al. Infrared spectroscopy and multivariate calibration to monitor
mixtures of vegetable oils for biodiesel production. Fuel 2012;95:126–30.
stability quality parameters of biodiesel. Microchem J 2010;96(1):126–31.
[46] Medeiros ML, Cordeiro, Angela M. M. T., Queiroz N, Soledade, Luiz E. B., Souza AL,
[20] Murta Valle ML, Leonardo RS, Dweck J. Comparative study of biodiesel oxidation
Souza AG. Efficient antioxidant formulations for use in biodiesel. Energy Fuels
stability using Rancimat, PetroOXY, and low P-DSC. J Therm Anal Calorimetry
2014;28(2):1074–80.
2014;116(1):113–8.
[47] Damasceno SS, Santos NA, Santos IMG, Souza AL, Souza AG, Queiroz N. Caffeic
[21] Dunn RO. Thermal oxidation of biodiesel by pressurized differential scanning
and ferulic acids: An investigation of the effect of antioxidants on the stability of
calorimetry: effects of heating ramp rate. Energy Fuels 2012;26(9):6015–24.
soybean biodiesel during storage. Fuel 2013;107:641–6.
https://doi.org/10.1021/ef3010543.
[48] de Sousa LS, de Moura CVR, de Oliveira JE, de Moura EM. Use of natural
[22] Zhou J, Xiong Y, Liu X. Evaluation of the oxidation stability of biodiesel stabilized
antioxidants in soybean biodiesel. Fuel 2014;134:420–8.
with antioxidants using the Rancimat and PDSC methods. Fuel 2017;188:61–8.
[49] Melo MAMF, de Melo MAR, Pontes ASGC, Farias AFF, Dantas MB, Calixto CD et al.
[23] Polavka J, Paligová J, Cvengroš J, Simon P. Oxidation stability of methyl esters
Non-conventional oils for biodiesel production: a study of thermal and oxidative
studied by differential thermal analysis and rancimat. J Am Oil Chem Soc 2005;82
stability. J Therm Anal Calor 2014;117(2):845–9.
(7):519–24.
[50] Zhou J, Xiong Y, Xu S. Evaluation of the oxidation stability of biodiesel stabilized
[24] Misutsu MY, Cavalheiro LF, Ricci TG, Viana LH, Oliveira SC de, Machulek A et al.
with antioxidants using the PetroOXY method. Fuel 2016;184:808–14.
Biofuels - status and perspective: Chapter 12; thermoanalytical methods in
[51] Hwang H-S, Winkler-Moser JK. Food additives reducing volatility of antioxidants
verifying the quality of biodiesel. InTech; 2015.
at frying temperature. J Am Oil Chem Soc 2014;91(10):1745–61.
[25] Vega-Lizama T, Díaz-Ballote L, Hernández-Mézquita E, May-Crespo F, Castro-
[52] Santos NA, Cordeiro AMTM, Damasceno SS, Aguiar RT, Rosenhaim R, Carvalho
Borges P, Castillo-Atoche A et al. Thermogravimetric analysis as a rapid and simple
Filho JR et al. Commercial antioxidants and thermal stability evaluations. Fuel
method to determine the degradation degree of soy biodiesel. Fuel 2015;156:
2012;97:638–43.
158–62.
[26] Pei P, Hsu S, Weeks S, Lawson R. Chemiluminescence instrumentation for fuel and
lubricant oxidation studies. Lubric Eng 1989;9(15).

12