Professional Documents
Culture Documents
Complex: Characterization of The Evaporation Rates of Hydrocarbon Mixtures Under Environmental Conditions
Complex: Characterization of The Evaporation Rates of Hydrocarbon Mixtures Under Environmental Conditions
The characterization of the evaporation rate of complex hydro- O n discute la caracthisation du taux d’kvaporation de mk-
carbon mixtures, such as commercial oil fractions, under langes complexes d’hydrocarbures (tels que les fractions d’huiles
environmental conditions is discussed. It is shown that the commerciales) dans les conditions ambiantes. On dCmontre que
relevant thermodynamic quantity is the total volume of vapour la quantitk thermodynamique pertinente est le volume total
(per unit of initial oil mass) removed under isothermal differ- de vapeur (par unit6 de masse initiale d’huile) enlev6 dans
ential distillation conditions (V m8/kg) as a function of the les conditions isothermiques de la distillation diffbrentielle
mass fraction of the oil remaining (F). Plots of F versus V for (V m8/kg.) en fonction de la fraction de masse de I’huile
various isothermal conditions can be used to estimate evaporation restante (F). Ont peut employer des representations graphiques
rates since it is shown that V is equivalent to the group de F, en fonction de V pour diverses conditions isothermiques,
KsAsrs/Ms where Ks is the evaporation mass transfer CD. pour &valuer les taux d’kaporation, vu qu’on constate que V
efficient (m/s), As is the spill area (ma), tg is the evaporation (5quivaut au groupe KsAs7/Ms, oh Ks est le coefficient de
time (s) and Ms is the spill mas (kg). The basis of a method- transfert de masse par Cvaporation (m/s), As est la surface de
ology for estimating environmental evaporation rates is &us la mare (m,), ts est le temps d‘kvaporation (s) et M, est la
outlined. Methods of obtaining V v m s F data are discussed, masse de la mare (k.g.). On e w e ainsi la base d’une m&hodo-
with calibrated wind tunnel evaporations being preferred. The logie pour Bvaluer les taux d‘bvaporation ambiante. O n discute
estimation of evaporation rates under spreading and non-is* les mkhodes pour obtenir des donndes en tunnel akrodynamiqne.
thermal conditions and the analogous dissolution p~ocessare On discute brikement l’haluation des taux d‘kvaporation dans
discussed briefly. des conditions non isothermiques et impliquant l’expansion de
l’huile.
I V G E 1 , V E 2 ,9, ~
E U M
"7 I~
.
OLD 'IJMP
111r
__
TANK
05k
I~0I"IIM.L D l I l l L l r l ~
<,"d.Oil 11.1 0
'J bI
, cv; ":,TI" 0 , , ,,,,, b, , , ~ ~ ,, , , , y
#,A
.", 1 XCA MI db-'
ISOTHERMAL D l S T l L L A T l O l APPARATUS
Figure 2 - Plot of fractioii of oil remaining F against V for
Figure 1 - Schematic diagram of isothermal distillation appa- wind tunnel evaporation at 25 'C and isothermal distillations
ratus. evaporations at 25°C and 45°C.
E = KPamlRT
0
where m is the molecular weight of cumene ( k g l g m o l ) ,
R is the gas constant ( a f m m'lgmol K ) and T the Environmental evaporation is truly differential
ambient temperature ( K ) . whereas the isothermal distillation apparatus uses a
I t should be noted that i t is assumed here t h a t the finite difference approach. However, these become
liquid phase is perfectly mixed at all times and that equivalent for a large number of cycles. Equivalent
there is no liquid phase mass transfer resistance. fractions will thus be evaporated from the environ-
This assumption is valid when the oil viscosity is mental spill and the isothermal distillation apparatus
low and the eddy diffusivity high. Some qualitative when the temperatures are equal and when K.A,T,f M ,
tests in which particles suspended in the oil were equals nvlMn.
observed demonstrated that the oil had a circulating In a wind tunnel the corresponding mass transfer
motion, induced by the air flow. Provided t h a t the coefficient, area, original mass and time may be
oil "mixing time" (observed to be a few seconds f o r K,, mls, A , mz, M , kg and rtLs giving the comparable
low viscosity oils) is negligible compared to the group K,,A,,rW/M u .
evaporation time (typically several days) there should Experimental evaporation data of either type a r e
be no liquid phase resistance effect. As crude oil thus readily converted to environmental conditions
evaporates its viscosity increases and i t may become provided that the environmental mass transfer coef-
necessary to thoroughly stir the oil a t regular in- ficient can be estimated. Mackay and Matsugu"' have
tervals. correlated the mass transfer coefficient against wind
The mass transfer coefficient is undoubtedly slight- speed U m l s , pool size 2 na, and vapour phase Schmidt
ly dependent on the mass transfer rate but the effect number Sc as follows:
will be small a t low transfer rates and the accuracy K , = 0.0048 ( J O . 7 8 Z-0.11 s c - O . G i
required for environmental calculations is not suf-
ficient to justify introducing this complication. In summary, an initial mass of oil M, ky evaporat-
Again i t proves convenient to plot the data in ing under environmental spill conditions with area
modified form, in this case a s t h e fraction of the A , and mass transfer coefficient K , after r. s will
oil mass remaining ( F ) as a function of the group reach the same fraction evaporated as a wind tunnel
K T A I M ~where
~ 7 is the evaporation time (s) and M W evaporation with corresponding values M,,, A,, K ,
the original oil mass (kg). Figure 2 shows such a and rw and as an isothermal distillation in which
plot for Normal Wells crude oil and for diesel oil. n volumes of v m' are evaporated from an initial M,, k g
The advantage of this conversion is demonstrated when
in the next section where i t is shown that the terms K.A,r,/M, = K,A,r,/M, = nv/Mo = V
K r A I M w from the wind tunnel data and nv/MI, a r e
equivalent. It is assumed that equal temperatures apply to all
three processes and that the oil is perfectly mixed.
Conversion of experimental evaporation data The most convenient method of presenting labora-
to enuironmental conditions tory data is thus as mass fraction of oil remaining
( F ) as a function of n v / M D or K,A,T~/M, i.e.
The evaporation loss of a mass of oil spilled under V mS/kg.
environmental conditions can be estimated using a For a given spill, the larger the area the faster
mass transfer coefficient approach. If a mass transfer the rate of evaporation since there is a greater area
coefficient K , mls applied to an area of A. m2 covered exposed per unit mass and the spill is correspondingly
by oil of initial mass M, kg of vapour pressure P a t m thinner. For spreading spills A , continuously in-
then during an element of time dt s the quantity creases and is a function of time. This effect can be
evaporated will be (K,A,P d t / R T ) moles where T is included in the prediction by integrating the area
the absolute temperature and R the gas constant with respect to time if the spreading rate is known
(atm d / g ~ I<).
d or by using a mean area (with respect to time)
As evaporation proceeds, P falls, and the amount in the equation. Strictly, the effect of a change in
evaporated becomes area should be included in estimating K. since K , de-
pends on 2 (or A 8 ) . However, the effect is slight
& ['Pdt = K . A T r , / R T M , molslkg since the power on Z is small.
The general case of a spill which is spreading and
is subject to varying wind conditions can be solved
by a finite difference method
where 7. is the elapsed time and 7 is the mean vapour
pressure. K , and A. are assumed constant. AM = K.(t)A, ( t ) A t P ( F ) / R T = A F M.
In the isothermal distillation apparatus a volume For the element of time At mean values o f K , ( t )
v ms of vapour (the volume of the tank) is removed and A.(t) a r e selected and the quantity K , ( t )
from an initial M D kg of oil each cycle thus leading A. ( t )a t / M . calculated. This is numerically equal
to a quantity evaporated of v P I R T moles. to AV and t h e values of AF corresponding to
For n equilibrations the amount evaporated is thus this A V can be read off the experimental F versus V
graph directly. If the F , V data are curve fitted, the
value of AF can be calculated analytically. The graph
or fitted equation contains the dependence of P on F
-P will be equal for both evaporation methodsif a
and there is no need to calculate a vapour pressure.
By repeating this calculation, the spill mass, volume
sufficient number of evaporation cycles rt is taken and thickness can be expressed as a function of time.
and the same oil is taken to t h e same fraction evap- The mass loss during time at will be M. AF.
The Canadian Iournal of Chemical Engineering, Vol. 54, August, 1976 293
Methodology evaporation conditions of mass transfer coefficient
(which can be estimated from correlations as a func-
The experimental and theoretical work described tion primarily of wind speed), spill area and mass
here provide a rational and rigorous methodology for (which can be readily determined from the spill thick-
characterising the evaporation rates of hydrocarbon ness and density) and the evaporation time.
mixtures. The data contained i n a plot of V versus With some increase in numerical complexity, the
fraction evaporated are truly thermodynamic in na- evaporation rate of a spreading and non-isothermal
ture. Such plots can be prepared for commercial oils spill can be treated.
which are subject to spillage and can be used to es- Methods of obtaining V versus F data have been
timate evaporation rates under environmental condi- discussed and it is suggested that the most convenient
tions. In many respects this curve is the isothermal method is wind tunnel evaporation under known mass
analog of the ASTM 086 isobaric distillation curve transfer coefficient conditions.
commonly used to characterise oil fractions. A mass An analogous characterisation for dissolution has
transfer coefficient can be readily estimated as a been developed but this is of little practical use.
function of wind speed and pool size for any en- It is hoped that this methodology will be of value
vironmental conditions and the corresponding evapora- in characterising oil fractions and in enabling evapora-
tion rate in terms of time to achieve a certain degree tion rates in the environment to be readily and ac-
of evaporation calculated using the plot, even during curately calculated.
spreading. It would be desirable to have a series of
plots covering the likely range of environmental tem- Acknowledgments
peratu res. The authors are grateful to the National Research Council of Canada
This experimental work demonstrated the general and the Inland Waters Directorate of Environment Canada for finan-
cial support.
feasibility of using both methods to obtain such plots
and has elucidated the nature of the experimental Nomenclature
difficulties. For practical purposes, the best approach
is probably to use a wind tunnel with periodic liquid A = oil spill area m2
a = tray area m2
mixing. Although the isothermal distillation proce- E = evaporation rate kg/s
dure could be refined and improved to give greater F = mass fraction of oil remaining
accuracy, this would require extensive experimental K = mass transfer coefficient m / s
m = molecular weight of cumene kglmol
modification to improve the vacuum system and pre- M = original mass of oil kg
ferably weigh the oil sample in situ. It may be most A M = incremental change in mass kg
useful for handling relatively volatile oils such as n = number of cycles
gasoline where a faster evaporation rate can be P = vapour pressure atm
achieved. p = mean vapour pressure alm
R = as constant atm m3/mol K
A similar approach can be used for characterisa-
tion of dissolution rate but the low solubilities and
s,
i
=
=
8time
chmidt number
s
low mass transfer rates result in only slight changes T = temperature K
in the oil composition which a r e unlikely to be of en- IJ = wind speed m / s
vironmental significance.
v = volume of vapour removed per unit initial mass-of
oil m3/kg
In an environmental emergency such as an oil spill u = tank volume m3
on water or a crash of a road tanker carrying gaso- x = pool diameter m
line, it is unlikely that there would be time to under- z = pool width m
A = incremental value
take any detailed calculations of evaporation rates r = time s
such as are described here. Such data would however
be useful in enabling the duration of the f i r e hazard Subscrifits
to be assessed and the contribution of evaporation in D = under isothermal distillation conditions
removing the oil or modifying its properties to be E = under dissolution conditions
estimated. Accordingly, i t is suggested t h a t the data o = original value at zero time
* = under spill conditions
could be stored and processed in nomograph form for = under wind tunnel conditions
use in emergency situations.
References
( 1 ) Mackay, D. and Matsugu R. S. Can. J. Chem. Eng. 61. 434 (1973)
Conclusions (2) Smith, C. L. and MacIn&e W.G.. Proc. Joint Conf. on Prevenl
tion and Control of Oil Shills, API. p. 119, Washington. D.C.
(1973).
It has been shown that in characterising the (3) Harribon, W.. Winnick, M. 0..Kwong, P. T. Y. and Mackay. D..
evaporation rate of hydrocarbon mixtures the ap- Env. Sci. and Tech., 9. 231 (1976).
(4) Sivadier, H.O. and Mikolaj. P. G., Proc. Joint Conf. on Preven-
propriate truly thermodynamic quantity is the total t’p_n__andControl of Oil S p j l l ~ , API, p. 476, Washington, D.C.
(lY“3).
volume V of hydrocarbons vapour ( a t its ambient ( 6 ) Scott, B. F.. “Investigation of the Weathering of a Selected Crude
vapour pressure) evaporated under isothermal dif- Oil in a
Cold Environment” in “Water Quality Parameters”, AS’Rd
STP 673 p. 614 Philadelphia PA (1976).
ferential distillation conditions per unit mass of oil ( 6 ) Matsup;, R. S..’M.A.Sc. Thesis Univ. of Toronto (1973)
and the corresponding mass fraction remaining un- (7) Mackau, D.. Charles, M. E. an’d Phillips C. R. Crude Oil Spills on
Northern Terrain (2nd Report) Infohnatio; Canada Cat. No.
evaporated F . Plots of F versus V for various tem- R67-34/1974.
(8) Regnier, 2. R. and Scott, B. F., Env. Sci. and Tech 9, 469 (1976).
peratures provide a convenient and rigorous method (9) Liss, P. S. and Slater. P. 0..Nature, 247, 181 (1974):’
of storing data on the isothermal evaporation char-
acteristics of an oil mixture. Manuscript received October 22. 1975 ; accepted for publication March
24, 1976.
It has been shown that V is equivalent to the
group K.A.7.1 M. which represents the environmental * * *
P p