See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/277973979

The Measurement of Flash Point of Water-Methanol and Water-Ethanol

Systems Using Seta Flash Closed Cup Tester

Article in Fire science and engineering · April 2015

DOI: 10.7731/KIFSE.2015.29.2.039

CITATIONS READS

4 15,780

3 authors, including:

Dong-Myeong Ha Sungjin Lee

Semyung University University of Seoul
129 PUBLICATIONS 339 CITATIONS 35 PUBLICATIONS 247 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Dong-Myeong Ha on 25 August 2015.

The user has requested enhancement of the downloaded file.

Fire Sci. Eng., Vol. 29, No. 2, pp. 1-5, 2015 [Research Paper]

Seta Flash 밀폐식 장치를 이용한 Water-Methanol과

Water-Ethanol계의 인화점 측정
하동명 · 박상훈* · 이성진**†
세명대학교 보건안전공학과, *에스텍 E&C, **세명대학교 임상병리학과

The Measurement of Flash Point of Water-Methanol and Water-Ethanol

Systems Using Seta Flash Closed Cup Tester
Dong-Myeong Ha · Sang Hun Park* · Sungjin Lee**†
Department of Occupational Health and Safety Engineering, Semyung University
*STEC E&C
**Department of Clinical Laboratory Science, Semyung University

(Received , 2015; Revised , 2015; Accepted , 2015)

요 약

인화점은 액체 혼합물의 화재와 폭발의 위험성을 특징짓는 주된 성질이다. 인화점은 가연성 액체에 불꽃이 가해졌을
때 점화가 발생하는 가장 낮은 온도로 정의된다. Seta flash 밀폐식 장치를 이용하여 이성분계 수용성 혼합물인 water-
methanol계와 water-ethanol계의 인화점을 측정하였다. Wilson 식과 UNIQUAC 식과 같은 활동도 모델식을 활용한 방
법을 이용하여 인화점을 계산하였다. 이 계산치와 라울의 법칙에 의한 계산치를 비교하였다. 그 결과, 활동도 계수 모델
식에 의한 계산치가 라울의 법칙에 의한 계산치 보다 측정값에 보다 근접하였다.

ABSTRACT

The flash point is the major property to characterize fire and explosion hazard of liquid mixtures. The flash point is the
lowest temperature at which a liquid gives off enough vapor to form a flammable air-vapor mixture. The flash points of
two aqueous mixtures, water-methanol and water-ethanol, were measured using Seta flash closed cup tester. A prediction
method based on activity coefficient models, Wilson and UNIQUAC equations, was used to calculate the flash point. The
calculated flash points were compared to the results by the calculating method using Raoult’s law. The calculated values
based on activity coefficients models were found to be better than those based on the Raoult’s law.

Keywords : Flash point, Seta flash closed cup, Water+methanol, Water+ethanol, Wilson, UNIQUAC

1. Introduction mits the emission of the low-boiling-point component in

the mixture into the surrounding atmosphere prior to the
(2,3)
The flash point is the important property of the flamma- application of the flame . On the contrary, the mixture in
ble liquid mixtures to analyze the safety in chemical pro- closed cup tester is separated from the surrounding atmo-
(1)
cesses . The flash point is the lowest temperature at which sphere, thereby inhibiting mass transfer between the mix-
a liquid gives off enough vapor to form a flammable air- ture and the surrounding atmosphere. Reliable and consistent
vapor mixture. Generally the lower the flash point tempera- flash point can be obtained using closed cup tester, particu-
(2) (3,4)
ture has the greater the fire hazard . larly for multicomponent mixtures . Therefore the flash
The flash point can be measured using either open cup point by closed cup tester are most widely used to charac-
tester or closed cup tester. Generally the flash point mea- terize flammable substance hazards in practice.
sured by open cup tester results in a higher value than that United Nations (UN) recommended the worldwide
measured by closed cup tester, because open cup tester per- implementation of Globally Harmonized System of Classi-

†
Corresponding Author, E-Mail: [email protected] ISSN: 1738-7167
†
TEL: +82-43-649-1286, FAX: +82-43-649-1537 DOI: http://dx.doi.org/10.7731/KIFSE.2015.29.2.1

1
2 하동명 · 박상훈 · 이성진

fication and Labeling of Chemicals (GHS) in 2008 and the compared to the results by the calculating method using
flash point of liquid mixtures is the crucial reference prop- Raoult’s law.
(5)
erty in the classification of flammable liquids within it .
Unfortunately experimental flash point data of liquid mix- 2. Experimental Section
tures are scarce. Particularly the flash points of the mix-
tures that have non-flammable component, such as water, Two binary aqueous solutions (water-methanol and
have seen little study and the data that did exist was incon- water-ethanol) were selected for the experimental samples.
sistent. Methanol and Ethanol were purchased from CARLO Erba
Deriving the flash point data using closed cup tester is a Reagents (Italy) with a minimum purity of 99.9%. Water
time-consuming work and the cost of obtaining the flash (HPLC grade) was supplied by J.T. Baker, USA. All these
(5)
point data is vary high . Therefore various alternative meth- chemicals were used without any purification.
ods to predict the flash point of liquid mixture have been The Seta flash closed cup tester manufactured by Koe-
proposed. hler Instrument Company (USA) was used to determine the
(6)
Affens and McLaren developed a predictive model for flash point of binary liquid solutions at different composi-
the flash points of alkane mixtures by combined Raoult’s tions.
law, Dalton’s law, and Le Chatelier’s rule. The model bet- The basic system configuration of the Seta flash closed
(11)
ter predict the flash point of the hydrocarbon mixture. cup tester was presented in our previous paper . The tes-
However, its application is limited to ideal solutions. ter consists of a sample cup, time controller, test flame
(7)
White et al. reduced Affens and Mclaren's model to a device, flame controller, thermometer, temperature control-
simpler equation by ignoring any dependence of the lower ler, etc. The Seta flash closed cup tester was operated
(12)
flammable limit (LFL) on temperature. However, the equa- according to the standard test method, ASTM D 3278 .
tion of White et al. is only valid for ideal solutions.
(8)
Liaw et al. developed a mathematical model for the 3. Flash Point Calculation of Binary
flash point of highly non-ideal solutions using activity Aqueous Mixture
coefficient models. Surely, the Liaw’s model needs the
binary interaction parameters of activity coefficient model 3.1 Flash point prediction using activity coeffi-
to calculate the flash points. Without the binary interaction cient model
parameters, the Liaw’s model is not able to predict the The vapor-liquid equilibrium relation of the component i
(10)
flash points for the binary solutions. can be expressed as the following equation .
(9)
Ha et al. predicted the flash point of the binary flam- 0 0
P i = P i αi = P i γ i x i (1)
mable mixtures using the optimizing methods based on the
0
van Laar and NRTL equations. where Pi is the partial vapor pressure of component i, P i is
Methanol is mixed with water and injected into high per- the vapor pressure of component i, αi is the activity of
formance diesel and gasoline engines for an increase of component i, γi is the activity coefficient of component i
power and a decrease in intake air temperature in a process and xi is the mole fraction of component i in the liquid
known as water-methanol injection. Ethanol is miscible phase.
with water and is a good general purpose solvent. It is Assuming vapor-liquid equilibrium in Seta flash closed
found in paints, tinctures, markers, and personal care prod- cup tester, the Clausius-Clapeyron equation can be applied
(10)
ucts such as perfumes and deodorants. to binary systems .
The purpose of this study was to measure and predict the
d ln Pi ΔHi
flash point to aid in evaluating the safety of such binary ------------- = ---------2- (2)
dT RT
liquid mixtures (water-methanol and water-ethanol). The
flash points for two systems, water-methanol and water- where T is the flash point in absolute temperature, ΔHi is
ethanol, were measured using Seta flash closed cup tester. the enthalpy of vaporization of component i, and R is the
The prediction methods based on activity coefficient mod- gas constant.
(10)
els, Wilson and UNIQUAC equations , were used to esti- The enthalpy of vaporization is a function of tempera-
(10)
mate the flash points. The calculated results were compared ture and is calculated using Watson equation in this study.
with the experimental flash points. Also the results were Integrating Equation (2), and Substituting Equation (1)

한국화재소방학회 논문지, 제29권 제2호, 2015년

 Seta Flash 밀폐식 장치를 이용한 Water-Methanol과 Water-Ethanol계의 인화점 측정 3

Table 1. The Binary Parameters of Wilson and UNIQUAC Equations for Each Binary System
Parameters Wilson* UNIQUAC**

Systems A12 A21 A12 A21

Water (HPLC grade)-methanol (99.9%) 433.0681 271.1382 697.5569 −404.8062

Water (HPLC grade)-ethanol (99.9%) 960.9534 307.4162 188.6515 −086.5941
*
Wilson : A12 = λ12 − λ11, A21 = λ21 − λ22.
**
UNIQUAC : A12 = g12 − g11, g21 = g21 − λ22.

results in : qji − qij

τji = ------------------ (10)
RT
ΔH 1
ln Pi = ln ( Pi, 0 γi xi ) + ---------i ⎛ ----- − ---⎞
0 1
(3)
R ⎝ T0 T⎠ z
lj = --- ( rj − qj ) − ( rj − 1 ), z = 10 (11)
2
where P0i, 0
is the saturated vapor pressure of component i,
T is the flash point of binary aqueous mixture, and T0 are The binary parameters of Wilson and UNIQUAC equa-
the flash point of pure flammable component. tions were abstracted from the literature(13) and are listed in
Below we assume that a flash occurs at the lower limit of Table 1.
flammability in the fuel-air system without the additive, The temperature (T) which satisfies Equation (6), was
0 determined to be the flash point of binary aqueous mix-
P2 = P2, 0 = constant (5)
ture. The results are presented in Table 2~3.
where the subscript 2 indicates the flammable component.
Substituting Equation (5) into Equation (3) results in : 3.2 Flash points prediction using Raoult’s law
Under an ideal solution assumption, the liquid phase
1 1
--- = ----- + ( R/ΔH2 ) ⋅ ln ( γ2 ⋅ x2 ) (6) activity coefficient of component 2 is equal to unity. There-
T T0
fore Equation (6) was reduced to Raoult's law, this being
The activity coefficient (γ2) of component 2, which were described as :
presented in Equation (6), can be calculated using two
1 1
activity coefficient models, Wilson and UNIQUAC equa- --- = ----- + ( R/ΔH2 ) ⋅ ln ( x2 ) (12)
T T0
tions. These equations being described as :
The temperature (T) which satisfies Equation (12), was
Wilson equation :
determined to be the flash point of binary aqueous mix-
Λ12 Λ21 ⎞
ln γ1 = − ln ( x1 + Λ12 x 2 ) + x2 ⎛ ------------------------- − ------------------------- ture. The results are presented in Table 2~3.
⎝ x1 + Λ12 x2 Λ21 x1 + x1⎠
Λ12 Λ21 ⎞
ln γ2 = − ln ( x2 + Λ21 x 1 ) + x1 ⎛⎝ ------------------------- − ------------------------- (7) 4. Results
x1 + Λ12 x2 Λ21 x1 + x2⎠

where Λ is as following 4.1 Experimental results

The flash points of binary aqueous mixtures, water (1)-
λ12 −λ11 λ21 −λ22
v − -------------------
RT
- v − -------------------
RT
-
methanol (2) and water (1)-ethanol (2), are listed in Table
Λ12 = ----2- e , Λ21 = ----1- e (8)
v1 v2 2~3 and Figure 1~2.
The concentrations of component i are given in mole
UNIQUAC equation :
fraction, xi. As shown in Figure 1 and Figure 2, the flash
* *
Φ z θ Φ m points of the systems plotted as a function of mole fraction.
ln γi = ln ------i + --- qi ln ------i* + li − ------i ∑ xj lj (9)
xi 2 Φi xi j=1 As can be seen from the figures, the non-flammable
component (water) inhibits the flash. The flash point of
⎛m ⎞ m θ'j τij
− q'i ln ⎜ ∑ θ'j τji⎟ + q'i − q'i ∑ -------------------
m - aqueous mixtures will be vary depending on the composi-
⎝ j=1 ⎠ j=1
∑ θ' τ k kj tion of the non-flammable component. The more non-flam-
k=1
mable component was obtained as having a the higher the
where, flash point temperature.

Fire Sci. Eng., Vol. 29, No. 2, 2015

4 하동명 · 박상훈 · 이성진

Table 2. The Experimental and The Calculated Flash Points

for Water (x1)-Methanol (x2) System
Mole fractions Flash points (oC)
x1 x2 Exp. Raoult’s law Wilson UNIQUAC
0.000 1.000 10.0 - - -
0.102 0.898 11.5 11.85 11.77 11.80
0.315 0.685 16.0 16.69 15.76 16.07
0.500 0.500 21.0 22.63 19.79 20.66
0.700 0.300 26.0 33.15 25.73 27.66
0.900 0.100 46.0 60.99 41.19 44.56
0.950 0.050 58.0 85.12 55.99 59.28
A.A.E. - 08.66 01.47 00.85

Table 3. The Experimental and The Calculated Flash Points

for Water (x1)-Ethanol (x2) System
Figure 1. The comparison of flash points of experimental
Mole fractions Flash points (oC) and calculated for water (x1)-methanol (x2) system.
x1 x2 Exp. Raoult’s law Wilson UNIQUAC
0.000 1.000 12.0 - - -
0.103 0.897 14.0 13.68 13.49 13.57
0.305 0.695 16.0 17.75 16.04 16.56
0.508 0.492 20.0 23.55 18.35 19.33
0.700 0.300 24.0 32.52 20.83 21.99
0.900 0.100 31.0 56.25 28.02 29.72
0.950 0.050 43.0 75.51 36.15 38.70
A.A.E. - 11.98 02.53 01.54

4.2 The comparison of the experimental and cal-

culated flash point
The calculated flash points were presented in Table 2~3.
The average absolute error (AAE) included in the tables is
(14)
defined as :
Figure 2. The comparison of flash points of experimental
exp cat and calculated for water (x1)-ethanol (x2) system.
N T
i − Ti
AAE = ∑ --------------------------
N
- (10)
i=1
o
8.66 C.
exp
where N is the number of experimental data, the is the Ti As shown in Table 3, for water-ethanol system, the
cal
experimental flash point of component i, and Ti is the cal- AAEs between the experimental data and the results calcu-
culated flash point of component i. The AAE is a measure lated by the methods using the Wilson and UNIQUAC
o o
of agreement between the experimental flash points and the equations are 2.53 C and 1.54 C, respectively. The AAE of
calculated flash points. the results calculated by the method based on Raoult’s law
o
A comparison between the experimental data and the cal- is also 11.98 C.
culated results of water-methanol system is presented It is clear that there is a good agreement between the
in Table 2. The AAEs of the results calculated by the measured flash points and those calculated by the activity
methods using the Wilson and UNIQUAC equations are coefficient method performed in this study, as can be seen
o o
1.47 C and 0.85 C, respectively. The AAE of the results from the AAEs. The activity coefficient method is also
calculated by the method based on Raoult’s law is also more accurate than the method based on the Raoult’s law.

한국화재소방학회 논문지, 제29권 제2호, 2015년

 Seta Flash 밀폐식 장치를 이용한 Water-Methanol과 Water-Ethanol계의 인화점 측정 5

5. Conclusion No. 1, pp. 1-23 (1997).

5. H. J. Liaw, V. Gerbaud and Y. H. Li, “Prediction of Mis-
The flash points for two aqueous mixtures, water-metha- cible Mixtures Flash-Point from UNIFAC Group Contri-
nol and water-ethanol, were determined using Seta flash bution Methods”, Fluid Phase Equilibria, Vol. 300, pp.
70-82 (2011).
closed cup tester.
6. W. A. Affens and G. W. Mclaren, “Flammability Proper-
The experimental data were compared with the calcu-
ties of Hydrocarbon Solutions in Air”, J. of Chem. Ind.
lated values based on the Raoult’s law and are generally in
Eng. Chem. & Eng. Data, Vol. 17, No. 4, pp. 482-488
bad agreement with those. (1972).
On the other hand, the flash points calculated by the 7. D. White, C. L. Beyler and C. Fulper, “Flame Spread on
method using the Wilson and UNIQUAC equations are Aviation Fuels”, Fire Safety Journal, Vol. 28, pp. 1-31
very similar to the experimental data. The activity coeffi- (1997).
cient method can thus be useful to calculate the flash point 8. H. J. Liaw, Y. H. Lee, C. L. Tang, H. H. Hsu and J. H.
of binary aqueous mixture and to help analyse the fire and Liu, “A Mathematical Model for Predicting the Flash
explosion hazard. Point of Binary Solutions”, J. of Loss Prevention in the
Process Industries, Vol. 15, pp. 429-438 (2002).
9. D. M. Ha and S. J. Lee, “Measurement of Flash Points
Acknowledgments
for n-hexanol + n-butyric acid and n-butanol + propionic
acid by Seta-flash Closed Cup Method”, Journal of the
This paper was supported by National Emergency Man-
Korean Society of Safety, Vol. 29, No. 6, pp. 76-80
agement (NEMA).
(2014).
10. C. R. Reid, J. M. Prausnitz and B. E. Poling, “The Prop-
References erties of Gases and Liquids”, 4nd ed., McGraw-Hill,
New York (1998).
1. T. Khalili and A. Z. Moghaddam, “Measurement and 11. D. M. Ha and S. J. Lee, “Prediction and Measurement of
Calculation of Flash Point of Binary Aqueous-Organic the Lower Flash Points for Flammable Binary Solutions
and Organic-Organic Solutions”, Fluid Phase Equilibria, by Using a Setaflash Closed Cup Tester”, Korean J.
Vol. 312, pp. 101-105 (2011). Chem. Eng., Vol. 28, No. 5, pp. 1161-1165 (2011).
2. A. Z. Moghaddam, A. Rafiei and T. Khalili, “Assessing 12. R. C. Lance, A. J. Barnard and J. E. Hooyman, “Mea-
Prediction Models on Calculating the Flash Point of surements of Flash Points : Apparatus, Methodology,
Organic Acid, Ketone and Alcohol Mixtures”, Fluid Applications”, J. of Hazardous Materials, Vol. 3, pp. 107-
Phase Equilibria, Vol. 316, pp. 117-121 (2012). 119 (1979).
3. L. Y. Phoon, A. A. Mustaffa, H. Hashim and R. Mat, “A 13. J. Gmehing, U. Onken and W. Arlt, “Vapor-Liquid Equi-
Review of Flash Point Prediction Models for Flammable librium Data collection”, Vol. I, Part 1-7, DECHEMA
Liquid Mixtures”, Ind. Eng. Chem. Res., Vol. 53, pp. (1980).
12523-12565 (2014). 14. D. M. Ha and S. J. Lee, “The Flash Point Measurement
4. D. White, C. L. Beyer, C. Fulper and J. Leonard, “Flame for Binary Flammable Mixture”, KIGAS, Vol. 18, No. 5,
Spread on Aviation Fuels”, Fire Safety Journal, Vol. 28, pp. 60-65 (2014).

Fire Sci. Eng., Vol. 29, No. 2, 2015

View publication stats