close

Вход

Забыли?

вход по аккаунту

?

The Measurement of High Pressure Vapour-Liquid-Equilibria Part II Static Methods.

код для вставкиСкачать
The Measurement of High Pressure
Vapour-Liquid-EquiIibria
Part II: Static Methods
J.D. Raal* and A.L. Muhlbauer'
Department of Chemical Engineering, University of Natal,
King George V Avenue, Durban 4007, SOUTHAFRICA
Static equilibrium cells in a variety of configurations have been widely used for high
pressure VLE measurement. The method can be divided into analytical procedures
where both vapour and liquid compositions are sampled and measured, and the synthetic
method for which no sampling of the phases is required. Where the isobars and
isotherms have large gradients the synthetic method is the least accurate of the two
methods. For mixtures with more than two components the information obtainable by
the non-analytical method is limited. Attempts to combine features of the analytical and
non-analyticalmethod in a single cell are also described.
The dificulty of sampling from a high pressure space without serious disturbance of
the equilibrium, and the problem of homogenization of a vapourised liquid sample are
reviewed. A static jet mixer for vapourized liquid-sample homogenization is described.
A new static equilibrium cell design is proposed in which the liquid level is made visible
by internal mechanical adjustment, and multiple liquid-phases can be sampled using a
single sampling valve without disturbing equilibrium. A mechanism for producing a
small continuousflow of liquid through a sample valve by use of the internal stirrer is a
novel idea, and is also described.
1. Static Vapour-Liquid-EquilibriumMethods
Description of the static analytical method
The features of a typical static apparatus are shown in Figure 1. The components
under investigation are charged into the equilibrium cell. The liquid components
may be flushed into the cell by the volatile component or pumped in. The volatile
component is usually supplied directly from its storage cylinder. High-boiling
volatile components such as propane and butane may have to be heated and pumped
in by a compressor-type device (Miihlbauer [l]). The contents of the cell are
agitated to promote contact between the phases, thereby shortening the time taken to
reach equilibrium. After equilibrium has been reached the temperature and pressure
are noted, samples of the liquid or vapour or both are withdrawn and their
compositions analyzed. The temperature and pressure of the mixture are controlled
m generate the required isothermal vapour-liquid equilibria.
* Author for correspondence.
I
Sentrachem, Sanaton, South Africa.
88
The Measurement of High Pressure Vapour-Liquid-Equilibria
Part 11: Static Methods
A summary of the important features of a selection of static apparatus reported in
the literature is given in Table 1. In particular, the following may be noted. The
apparatus of Rogers and husnitz [4], Ng and Robinson [91, Figuiere et al. [lo], Bae
et al. 1121, Legret et al. [ll], Guillevic et al. [la], and Nakayama et al. [18] employed
unique methods to sample the liquid and vapour phases. Sampling the liquid phase
without disturbing equilibrium presents the most challenging and difficult problem to
be surmounted for a successful cell design. This problem is exacerbated when two
liquid phases are present. The equipment of Swaid (as described in Konrad et al.
[141) is included as an example of the spectrographic method of phase analysis. The
cell designs of Kalra et al. [8] and Aschcroft et al. [15] are described below. The
methods used by Kalra et al. [8], Ng and Robinson [9], Wagner and Wichterle [19],
and Miihlbauer and Raal[21] to homogenise a vapourized volatile/non-volatile liquid
sample are described. Also discussed are the equilibrium-cell agitation methods of
PRESSURE AND TEMPERATURE
MEASURING DEVICE
'
3
LIQUID SAMPLING
SAMPLING SYSTEM
/ AGITATION
SYSTEM
DEVICE
CONTROLLED ENVIRONMENT
Figure I . Features of the static analytical method.
Kalra and Robinson [22], Bae et al. [12], Figuiere et al. [lo], and the air bath
designs of Rogers and Prausnitz [4], and Miihlbauer and Raal [21].Air bath design is
important as the equilibrium cell must be devoid of all thermal gradients.
Liquid and Vapour Sampling Methods
In the ingenious equilibrium cell of Rogers and Prausnitz [4], vapour and liquid
samples were removed from the equilibrium cell via two sets of moving pistons.
Between each set of two pistons was a small variable volume. During sampling this
volume was extruded from the equilibrium cell into a cylinder and moved down the
cylinder until the sample ports were reached. The sample then expanded through
capillary tubes into a low pressure zone. This sampling technique had the main
advantage that the equilibrium cell pressure was not disturbed during withdrawal of
the samples since the piston movement did not alter the internal cell volume. In the
equilibrium cell described by Kalra and Robinson [22], the vapour and liquid phases
were sampled though specially designed needle valves which were mounted directly
into the wall of the cell. The main features of the valve design were that a sample
89
Rigas
T&lheide
Rogers
Beserer
Sagwa
Klink
Kalra
MI
Figuiere
Legret
Bae
King
Konrad
Kmad
Ashcmft
Guillevic
HWnO
Nakayama
Wagner
Reid
MJhlbauer
Author
(1)
Cell
Volume
1958
w
I21
(21
Ref. Date
(3)
(4)
270
65
2060
360
(51
(6)
(7)
(81
19)
(10)
(11)
(12)
(13)
(14)
(14)
(15)
(16)
(17)
(18)
119)
120)
(21)
100
883
50
w
150
150
101115
500
800
250
150
50
100
300
500
1963
1970
1971
1972
1976
1975
1978
1980
1981
1981
1983
1983
1983
1983
1983
1985
1987
1987
1987
1991
I
3500
1 000
80
100
400
-
Cell
316 SS
PT
B
PTmw
B
B
Press
(61
Measurement Device
TC
TC
TC
TC
TC
B
PT
PT
DWP
PT
TC
TC
TC
TC
B
ss
ss
DWP
Equilibration
time
min
Sample Size
w
0.75
200-300 mg
1OJ gmoles
1
002
200-300mg
6/10 cydes
60
30011 600
1
12
16
8
B
PT
PT
PT
PT
7 20
720
180
900
100
. M l
0.06
818
260
0,06
10' omoies
180
15
8
N.A.
50
60
w
10
120
310
N:A.
QT
BlPT
Ninomic 90 SS
RT
RT
B
RT
N90
N90
MS
316
Sapphire
316SS
ss
-
TC
TC
RT
TC
TC
TC
1s
-
304
316 SS
316SS
316SS
ss
ss
Table 1.
Statlc EquMbdum Apparatus R e p o ~ Whth.Utuature
Operatino Range
423
623
423
310
103
1 000
100
345
175
338
316
673
673
323
2000
2000
690
70
186
200
90
300
200
I
3001450
2931473
333
658
3 101477
450
323
473
433
(1) Due to space limitations only first author mentioned.
-
(2) VV = Variable volume.
131 Material of construction of equilibrium cell. SS = Stainless steel; MS Manganese steel, N 90
(4) TC = thennocouple; RT = Aatinum bsistance Thermometer (pt 100 n).
TS = thermistor; QT = Quartz thermometer.
15) B = Bowdon pressure gauge; PT Pressure transducer; DWP = Dead weight piston gauge.
The Measurement of High Pressure Vapow-Liquid-Equilibria
Part 11: S t a u Mcthodr
could be completely removed into a stream of helium (from an external supply) with
a separate sealing system for the needle and the seat. The sealing arrangement made
it possible to provide the relatively high packing load required to seal off the seat,
without interfering with the lower load required to seal the delicate sampling needle.
The unique feature of the experimental apparatus of Figuiere et al. [lo] was the
vapour and liquid sampling valves. The valves were opened by a hammer activated
by an electromagnet and transmitted to the valves by pushers. The valves were
returned onto their seats by powerful spring washers. The sample volume withdrawn
depended on the opening time of the valve. Opening times of a few hundredths of a
second allowed approximately 1 pl to be withdrawn. The withdrawn sample flowed
through slits machined along the stems where it mixed with the chromatographic
carrier gas. The carrier gas carried the sample to a gas-liquid chromatograph for
composition analysis. A possible disadvantage of this sampling method is that the
sample size will depend on the cell interior pressure and also on the viscosity and
density of the fluid.
Accurate analysis of systems containing high-boiling liquids requires heating of
the withdrawn sample to a higher temperature than the equilibrium temperature to
avoid condensation of the heavier component. Heating the sample in a manifold
attached to the equilibrium cell without disturbing the equilibrium between the
phases becomes a very difficult proposition (Miihlbauer and Raal [21]). Legret et al.
[ 1 11 developed a sampling system which consisted of detachable sampling
microcells. The sampling microcells were transferable to specially designed
chromatographic injection ports. According to the authors, this system permitted
accurate chromatographic analysis of mixtures with high-boiling components.
In the apparatus developed by Guillevic et al. [16] the vapour and liquid phases
were sampled with similar capillary sampling valves. The authors note that the
capillary length, internal section, and the shape of the stem were the consequence of
an experimental study undertaken to obtain representative samples. A sample of
either the liquid or vapour phase was obtained by opening the sampling valve for a
short time. The authors claim that the withdrawn samples were large enough to allow
for chromatographic analysis, yet small enough to cause negligible change to the
equilibrium condition inside the cell.
The liquid phase in the equilibrium cells of Ng and Robinson [9],Bae et al. [12]
and Miihlbauer and Raal [21] was sampled through specially designed devices,
mounted directly into the wall of the cell. The design of the devices was based on
the liquid sampling method of Fredenslund et al. [23]. Bae et al. [ 121 initially used
the same device for vapour-phase sampling. However, it was found to be unsuitable
for sampling the vapour phase as a thin film of liquid was found to have adhered to
the equilibrium cell wall in the vapour space and the vapour sampling technique was
changed. A small portion of the vapour was released into an evacuated collection
device and flushed via a carrier gas stream to the gas chromatograph. A similar
vapour sampling method was adopted by Miihlbauer and Raal [21]. The vapour
sampling system had been developed and described earlier in the literature by Kalra
and Robinson [22].
The two sampling devices used by Nakayama et al. [lS] were a capillary and a
sliding rod sampler. The latter appears to be a combination of the ideas of Rogers
91
J.D. Raal and A.L. Miihlbauer
and Prausnitz [4] and Fredenslund et al. [23]. The sliding rod was brought
completely through the equilibrium cell so the only change in cell volume was that
of the sample volume of approximately 0.1 cm3. The capillary tube sampler was
simpler to design, operate and maintain. However, the cell did experience a larger
volume change during sampling, i.e. 0.25 cm3. To ensure analysis of the
equilibrium phase instead of stagnant material (which could be trapped in the thin
capillary lines) the fmt few withdrawals were discarded.
Nakayama et al. [18] compensated for the pressure change that occurs during
sampling by connecting the equilibrium cell via a stainless steel diaphragm to a
buffer tank (kept at the equilibrium pressure). The volume of the buffer tank was
approximately 5 times greater than that of the cell. The volume and pressure changes
that occurred during sampling were automatically compensated for by the action of
the diaphragm. Great care had to be taken when filling the cell to avoid rupture of
the diaphragm that could occur with large pressure differences.
Konrad et al. [14] used two static equilibrium cells to measure equilibrium data
for non-volatile systems. The cell they developed used a sampling technique, the
other was developed by Swaid [24] and used a spectrographic technique to analyze
phase composition data. Konrad et al. [14] used optical windows for observation of
cell contents. Sampling of the phases was achieved through the use of two capillary
tubes located at the bottom of the cell and in the upper third of the cell. The capillary
tubes were closed by high pressure valves with small dead-volumes. The samples
were removed from the liquid and vapour phases through the valves and separated
into two phases. The sample was decompressed into an evacuated vessel instead of
vapourizing and homogenizing by heating, and the amount of gaseous component
was analyzed using PVT data. The non-volatile components were dissolved in a
solvent and analyzed in a gas chromatograph. This rather cumbersome sampling
procedure and analysis method was necessitated by the thermal sensitivity and high
boiling points of the non-volatile components under investigation.
The equilibrium cell developed by Swaid [24] used near-infrared spectroscopy to
determine composition data in density units according to Beer's law. Other
analytical and synthetic methods obtain cornposition data in mole or mass fractions
that can only be converted to density units with the knowledge of PVT data for the
coexisting phases. This information is normally not available. Synthetic sapphire
windows were situated at the bottom and middle of the cell. Teflon foil was used as a
sealing material between the steel window plugs and the window. The optical axis of
the windows was parallel to the cylinder axis to prevent double refraction and
interference of the absorption bands. Near-infrared spectroscopy was used in
preference to infrared as the absorption bands are usually well separated in the
former. Absorption data were measured with a spectrometer.The electrical signals of
the reference and sample beams were branched out directly behind the detector,
digitized, and evaluated by computer. The molar absorptivity values were
determined in the homogeneous region as a function of pressure, temperature and
composition.
Several different methods have been reported to pressurize the equilibrium cell
contents. Injecting mercury into the equilibrium cell was reported by Kobayashi and
Katz [25]. Huang et al. [17] used a rotary drive mechanism and piston. Konrad et
al. [14] and Swaid [24] used a Bridgman piston. Pressure created by the
92
The Measurement of High Pressure Vapour-Liquid-Equilibria
Par! II: Static Methodr
compression of silicon oil in a screw press was transmitted to the cell contents by a
separator system consisting of either a bellows or piston. If the piston was used, the
cell volumes could be determined by knowing the piston position measured by an
inductive coil.
Ashcroft et al. [ 151 have described a large variable-volume equilibrium cell
design with provision for sampling of multicomponent, multiphase systems, and for
visual phase observations and volume measurements. This was achieved by means of
a unique sampling valve, an optical system, and an accurate method of determining
the cell volume. The equilibrium cell consisted of upper and lower cylinders
connected to each other by a glass capillary tube which formed part of the window
assembly used for optical observations. The lower cylinder was pressurized with
mercury and the upper hydraulically. The use of mercury in these designs is
unfortunate due to the toxicity of its vapour and should be discontinued. The lower
cell contents could be raised or lowered to allow visual observations of the liquid
phase, phase boundary and vapour phase by dual action pumping of the mercury and
hydraulic oil. Once it was established that the liquid or the vapour phase occupied
the space surrounding the sampling valve, a sample of the phase could be taken by a
valve located between the two chambers. Agitation of the cell contents by rocking
the whole assembly is unnecessarily cumbersome.
Sampling components that show great differences in relative volatility:
Methods of obtaining sample homogeneity
In the early 1970's interest developed in the field of supercritical extraction.
Equilibrium data for volatile/non-volatile systems, especially those containing
carbon dioxide, were in great demand. Kalra et al. [8] used an equilibrium cell
similar to that described by Besserer and Robinson [5] to study the carbon dioxiddnheptane system. However, certain major modifications were made to the equipment
to overcome the difficulties associated with the analysis of a volatilehon-volatile
system. In a volatile/non-volatile system, liquid-phase sampling produces flashing.
A liquid mixture containing relatively light and heavy components undergoes a
distillation process when it is throttled across a valve (from a high-pressure region to
a low-pressure region, or to an evacuated space) and the sample obtained for analysis
is not representative of the equilibrium liquid phase. Previous methods of sampling
the liquid phase, namely those based on opening a valve and releasing the liquid
phase into an evacuated space (Kalra and Robinson [22]) were not suitable nor
applicable to volatile/non-volatile systems. A sampling method similar to that of
Rogers and Prausnitz [4] was required, namely the removal of a representative liquid
sample out of the equilibrium cell.
To overcome this fishing difficulty, Kalra et al. [8] modified the equilibrium cell
of Besserer and Robinson [5] by developing a complex new liquid-phase sampling
method. The modification consisted of inserting a four-port ball valve in the line
connecting the vapour to the liquid space. The liquid level in the cell at equilibrium
conditions was raised using a double-actingRush pump until the liquid flooded one
of the through-ports in the ball valve. The valve was rotated 90°, whereupon the
filled port came in-line with hot circulating helium. The helium and sample were
circulated until the liquid sample was completely vapourised. The flow was
93
19.Roal and A L . Miihlbouer
I
-
G
Figure 2. Jet mixer for liquid sample vapourization and homogenization.
A- Pressure transducer, B-cartridge heaters, C-nozzle support, D-Venturi throat,
E-rounded exterior chamber, F-outlet valve, G-sample inlet.
switched to the chromatographic sample valve for analysis. The Ruska pump moved
the upper and lower pistons simultaneously, thus maintaining a constant cell pressure
and temperature so that the equilibrium state was not disturbed. An additional
problem related to the low volatility of the heavy component, n-heptane. In
previous experimental studies [8] systems with higher relative volatilities had been
investigated. To prevent partial condensation of the n-heptane during sampling and
sample transportation, an electrically heated and thermostatically controlled
manifold enclosed in an insulated housing was used. The manifold was part of a
circulation line equipped with a stainless steel bellows-type diaphragm pump. The
pump circulated the vapourized liquid sample and helium carrier gas. The circulation
process ensured homogenization of the liquid sample if any flashing occurred during
the sampling process, and prevented parrial condensation of the heavier component
prior to injection into the gas chromatograph.
Sample analysis problems similar to those described by Kalra et al. [8] were
reported by Ng and Robinson [9]and Miihlbauer and Raal [XI. Wagner and
Wichterle [19] used a capillary sampling method to measure binary and ternary data
for carbon dioxide, 1-hexene and n-hexane. They reported separation of the
withdrawn samples for the volatilehon-volatile systems due to the pressure drop
along the capillary. A 0.8 ml intensively stirred, glass homogenizing vessel was
inserted in the sampling line to rehomogenize the withdrawn samples. The vessel
was kept at approximately 100 K higher than the equilibrium temperature. In the
equipment of Nakayama et a1.[18] the heavy components that could not be
vapourised immediately upon sampling were collected in a cold trap. These
components were returned to the air bath to be gasified. The light and heavy
components were then homogenized with a magnetic pump in a circulation loop
before being injected into the gas chromatograph for composition analysis.
94
The Measurement of High Pressure Vopour-Liquid-Equilibria
Port 11: S t o h Methodr
In the equipment of Miihlbauer and Raal [21], a jet mixer performs the function
of vapourising and homogenizing the liquid sample in preparation for gas
chromatograph analysis. The liquid sample was flashed into the initially evacuated
jet mixer through the nozzle at high velocity, producing a swirling, recirculating flow
until the pressure became uniform. Further mixing of the sample was ensured by
subsequently flushing helium carrier gas through the passage in the cell wall into the
mixer at a controlled pressure. The essential features of the jet mixer are shown in
Figure 2. Jet mixer pressure is a useful indication of liquid sample size and was
monitored with a temperature-compensatedflush-mountedpressure transducer.
Equilibrium Cell Agitation Methods
Mixing of the phases in the equilibrium cell of Kalra and Robinson [22] was
achieved with a Teflon-coated magnetic stirrer. The driving force for the stirrer was
provided by a magnetic pile mounted externally to the cell. The pile consisted of
three magnets mounted in mild steel shoes, encased in an aluminium housing, and
driven by a variable speed d.c. motor. The magnetic stirrer remained coupled to the
rotating pile at speeds in excess of 500 rpm, through 19 mm of 316 stainless steel
and a 15 mm air gap. If the cell was half full of liquid, the stiner generated a deep
vortex in the liquid phase causing vapour bubbles to be continuously drawn into the
underlying liquid phase. It was claimed that equilibrium was rapidly attained
between the phases (30 - 120 minutes) depending on the conditions and the mixture
composition being investigated. The cell agitation methods of Ng and Robinson [9]
and Muhlbauer [ 11 were similar. Equilibration of the phases in the equilibrium cells
of Figuiere et al. [lo], Legret et al. [ll], and Guillevic et al. [16] was achieved by a
magnetic stirrer rotating in an orientable magnetic field induced by coils located
outside the cell. This arrangement is much simpler mechanically and is preferred.
The windowed equilibrium cell of Bae et al. [12] featured a unique magnetically
driven impeller. Mixing in the cell was achieved by an internally vaned metallic
impeller mounted on a hollow shaft. Vapour entered through small holes on the
upper part of the shaft, descended down the hollow shaft, and dispersed into the
underlying liquid phase. Besserer and Robinson [5] have described a variablevolume equilibrium cell capable of measuring refractive indices and vapour-liquid
equilibrium data. The cell consisted of three sections bolted together, two cylinderpiston end sections, and a central windowed section. The windowed section
contained the temperature and pressure-measuring devices and the sampling and
optical systems. Equilibration and mixing of the cell contents was achieved by
transferring the contents between the upper and lower cylinders. A thermocouple
proportional-band temperature-controller maintained the heaters in the surrounding
aluminium shrouds placed over the ends of the cell to within 0.5K of the setpoint.
Thermal environment of equilibrium cell
The equilibrium cell of Ng and Robinson [9] was maintained at the desired
temperature by a 25mm thick aluminium shroud containing eight vertically mounted,
uniformly spaced, pencil-type 250-W electric heaters. The equilibrium cell
temperatures of Konrad et al. [14] and Swaid [24] were also maintained by
electrically heated thermostatting jackets. However, they used additional head and
95
J.D. Raal and AL. Miihlbauer
bottom heaters to reduce the axial temperature gradients to below 0.2K. In the
equipment of Rogers and Prausnitz [4] and Miihlbauer and Raal [211 air baths were
used. Great care was taken in the design of the heating and cooling system by Rogers
and Prausnitz [4] in order to achieve isothermal bath conditions. The inside of the
bath was constructed of copper plates to promote temperature uniformity and
stability. The equilibrium cell was mounted on steel supports with additional copper
plates between them to minimize heat conduction out of the nitrogen bath. Heat
transfer lines were hard soldered to the copper walls of the interior to speed up the
transfer of energy between the bath and the heatinghooling system. A similar
approach (copper lining) for promoting bath temperature uniformity was used by the
present authors (loc. cit).
Description of the Static Non-Analytic Method
A mixture of known composition is prepared and placed into the equilibrium cell.
The temperature or pressure of the mixture inside the equilibrium cell is adjusted
until phase separation of the homogeneous phase occurs. As soon as homogeneous
phase separation starts the pressure and temperature are noted. The mole fractions of
substances making up the mixture can be precisely calculated, as the quantities of
each substance initially loaded into the equilibrium cell are known. After the
appearance of the second phase, the temperature and pressure should be readjusted
into the homogeneous region in order to avoid de-mixing and layering of the phases.
The temperature or pressure is again adjusted until the formation of a new phase is
observed. The pressure, temperature and mole fractions at which these phase
separations start will define points on the phase envelope. The initial results of these
experiments are sets of isopleths (phase boundaries at constant composition). The
method whereby the temperature is varied at constant pressure until a second phase
appears is known as the method of temperature variation. Repeating these
experiments under different constant pressures defines more points on the p(T)isopleth. Conversely, the method whereby the pressure is varied at constant
temperature is known as the method of pressure variation and a T@)-isoplethis
generated. The final results of the synthetic experiments, pressuremole fraction and
temperature-mole fraction phase diagrams, must be obtained by cross plotting from
the two sets of isopleths. de Loos et al. [331 report the use of this type of apparatus
to measure phase equilibria and critical phenomena for the propane/water and
ethane/2-methylpropanesystems respectively.
An example of an application where volumetric data was generated is the
experimentalequipment described by Meskel-Lesavre et al. [26]. The main feature
of the equipment was a very light and small titanium equilibrium cell whose volume
could be altered by a pressurising device, and whose mass could be accurately
determined by a balance (due to its light weight). The liquid component was
introduced into the cell and, after degassing, accurately weighed. The second
component was added and weighed, and the cell inserted into the pressurising
device. The pressure of the cell was increased while the vapour-liquid equilibrium
was maintained by vigorous magnetic stimng. The pressure of the cell was known as
a function of the total cell volume; a correction for thermal expansion and
compressibility of the hydraulic pressurising fluid was taken into account. Accurate
96
The Measurement of High Pressure Vapour-Liquid-Equilibria
Part 11: Slotic Methods
values of the bubble pressure and saturated liquid-phase molar volume were
simultaneously obtained from the pressure-volume plot, where the discontinuity
corresponds to the vapour phase disappearance (bubble point). The mixture pressure
was released and when the vapour phase reappeared the process was repeated at
another temperature. The method is rapid, the authors claiming it took 1 hour to
reach equilibrium and another hour to construct a pressm/cell volume diagram with
10 points. No visual observations were necessary, thus avoiding potentially complex
sealing problems. A further development of this equilibrium cell capable of
withstanding higher temperatures and pressures was described by Rousseaux et al.
[27]. Di Andreth and Paulaitis [28] used a synthetic- type apparatus in which the
volumes of the individual phases of a three or four-phase three-component
equilibrium system could be measured.
Advantages and Disadvantages of the Synthetic Method
No sampling is necessary and therefore no complicated and expensive analytical
devices are required. The experimental method is simple and since there is no need
to wait for equilibration between the phases equilibrium data can be generated
quickly and efficiently. P, V and T and even orthobaric densities may be obtained if
the motion of the pressure transferring element can be recorded with sufficient
precision (Meskel-Lesavre et al. [26]). Critical-state measurements can be carried
out on this equipment. A p(T)-isopleth can be obtained from one filling of the
equilibrium cell.
Visual observation of iso-optic systems, i.e. where the coexisting phases have
approximately the same refractive index, is extremely difficult if not impossible. For
mixtures with more than two components the information obtained from
experimentation is limited, a considerable shortcoming. The method is not suitable
for measurements away from the critical states, i.e. where (aT/X), and (aP/aC), is
large or infinite. Great care must be taken not to overlook a dew point, which can
easily happen if the liquid phase condenses not as a mist but as a thin film on the
wall of the pressure cell. Cooling or heating of a gaseous phase (in particular) must
necessarily introduce a temperature gradient, even when the phase is well stirred.
The Static-Combined Method
In analyzing the isobars and the isotherms of phase diagrams, they are seen to have
large gradients far away from the critical states and small gradients near the critical
states (Deiters and Schneider [29]). Where the isobars and isotherms have small
gradients, slight disturbances in the temperature and pressure of the mixture in the
equilibrium cell can produce large fluctuations in the phase compositions. Therefore,
application of the analytic method for the study of phase behaviour near the critical
state should not be considered. As the non-analytical method does not require
sampling, the above fluctuations are not experienced. Consequently this method is
more accurate near the critical state. Where the isobars and isotherms have large
gradients, the non-analytical method is the least accurate of the two methods. An
error in the overall composition leads to a large temperature deviation in the
generated data.
97
J.D. Raal and A.L. Miihlbauer
Attempts have been made at combining the features of the analytical and nonanalytical static methods into a single equilibrium cell. Provision is made for
viewing of the contents, to allow for sampling of the vapour and liquid phase, and in
recent years some method of accurately determining the volume of the equilibrium
cell. The equilibrium cell of Wisotski is a typical example of this type of apparatus
and is shown in the publication by Deiters and Schneider [29]. The equilibrium cell
was used for cryogenic investigations of volatile substances. For mole fractions far
away from the critical point, the cell was operated in the analytical mode. Samples of
the liquid and vapour phases were transferred from the equilibrium cell into sample
loops connected to six-way valves. The samples were then flushed to a gas
chromatograph for compositional analysis by flowing helium gas. For mole
fractions near the critical point, the cell was operated in the non-analytical mode.
The cell contents were pressurized by a Bridgman piston, and separated from the
pressurizing fluid by a bellows. The position of the bellows was accurately
STEPPER MOTOR 1
PIECE
PISTON ONE .
BALL BEARING
SPACERS
1I
a
.----
MAGNETICALLY:-COUPLED
-.-BALL
BEARING
END PIECE 2 -'
CELL BLOCK END PIECE
STEPPERMOTOR 2
Figure 3. Proposed new static equilibrium cell.
(P,T = pressure, temperature sensing elements).
98
The Measurement of High Pressure Vaput-Liquid-Equilibria
Part 11: Stalk Methods
determined by a magnetic wire, interacting with a high frequency transformer, which
enabled information on phase densities to be collected. Japas and Frank [30] and
Hung et al. [ 171 also describe combined-methodequilibrium cells.
2. Proposed New Cell Design
A shortcoming of the VLE equipment described by Muhlbauer and Raal [21] is its
inability to detect and sample multiple liquid phases, one of the most difficult
problems in high pressure VLE measurement. Applying some novel ideas and from
experience gained during experimentation, a new equilibrium cell was designed, the
principal features of which are shown in Figure 3. The advantages of the proposed
cell lie in the ability to detect multiple phase formation through windows, and to
subsequently sample each liquid phase with a single sampling valve without
disturbing equilibrium.
Features of the equilibrium cell (Figure 3)
The equilibrium cell incorporates see-through windows and can be housed in an air,
oil or water bath. The equilibrium cell cavity is contained between two cylindrical
end-pieces (1 and 2) which are joined together by spacers. The two end pieces and
spacers make up piston 2, which can be moved up and down in the cell block by
stepper motor 2. The ability to move piston 2 allows the cell contents to be viewed at
any level. The formation of multiple liquid phases can therefore be detected and
subsequently sampled with a single Rheodyne 6-portvalve. Piston 1 is used to adjust
the volume of the equilibrium cell, allowing the cell pressure to be set at any desired
value. The seal between the circular end pieces 1 and 2 and the cell block is
provided by a pair of Viton "0"-rings housed in grooves machined to Dowty
specifications, for example. Viton is desirable due to its chemical inertness and its
relative high temperature capabilities (160°C). Piston 2 is aligned via the cell block
end pieces (see Figure 3). Alignment is necessary to facilitate operation and to
prevent stress on the "0"-rings. The piston is prevented from turning with the stepper
motor by a ball bearing. The ball bearing allows the piston rod to rotate freely in the
cylindrical end piece. A similar "0"-ring and ball bearing system is employed on
piston 1 which is used for equilibrium cell volume adjustment.
The volume of the cell is varied by piston 1, which is advantageous for three
reasons. Firstly, the cell contents can be pressurized. This is a desirable capability for
systems containing propane (or components of similar vapour pressure), where
heating the equilibrium cell is not sufficient to pressurize the cell contents to the
critical point and additional forms of pressurization are required. Secondly, the
ability to vary cell volume allows the cell to be used in the synthetic mode of
operation. Thirdly, if the piston position is known by graduation, the cell can supply
volumetric data. To obtain accurate volumetric data, the second piston must be
moved either manually or via a second stepper motor suitably geared to enable it to
move against cell pressure. Hydraulic propulsion was discounted due to the
phenomenon of hydraulic slip. These three considerations suggest the cell to be
"long and thin" rather than "short and wide". The former configuration allows very
sensitive volume changes. A higher liquid level is also produced which allows for
more efficient purging of the sample loop.
99
GAS CHROMATOGRAPH CARRIER GAS
m
s2
B
-I
’
I
I
b)
W
<
z
m
0
2i
CELL SAMPLE LOOP
I
SAMPLE LOOP
Q)
SAMPLE LOOP
The Meosvrement of High Presswe Vapow-Liquid-Equilibria
Part 11: S t a k Methods
The liquid phase to be sampled is brought into position by piston 2. Liquid
sampling is via a single capillary leading to an externally mounted Rheodyne 6-port
valve. A sampling rod extending the full diameter of the cell (Rogers and Prausnitz
[4] and Nakayama et al. [18]) could also be used. The feature of using a singlecapillary sampling port and valve (Laugier and Richon [31] and Renon et al. [32]) as
opposed to sampling by multiple capillaries (Todheide and Frank [3]) is desirable for
a number of reasons, namely:
1. A large number of fittings are required to seal the capillaries and continual
sealing problems can therefore be expected.
2. A large number of capillariescreate stagnant spaces.
3. The Capillary positions are fixed, and the capillary inlets are not necessarily
always at the required level to sample the desired liquid phase.
There are problems associated with capillary sampling into an evacuated space,
due to the volatile component's tendency to flash and consequently create stagnant
regions in the thin capillary. Venting, the standard technique employed to remove the
stagnant region, creates an undesirable pressure gradient from the cell pressure to the
exit pressure of the valve. The stagnant region and venting problems can be
overcome by fitting a funnel to the end of the capillary line. As the piston moves up,
it channels liquid into the funnel, up into the capillary lines, and back out into the
cell, purging the sample line and sample loop (see Figure 4). The sample loop is
thereby filled with a representative liquid sample. A simple method in which the
stirrer is used to produce a return flow through the liquid sampling valve has recently
been devised by Raal [34].
Vapour sampling (not shown in Figure 3) is achieved by displacing a vapour
sample in one of two sampling loops of an 8-port Valco sampling valve. Using a
Rheodyne or Valco valve has many advantages over the conventional sampling
methods discussed under static methods. Primarily, the sample size can easily be
manipulated by changing the valve sample-loop volume. Changing the sample size
with the sampling rods of Nakayama et al. [18] and Muhlbauer and Raal [21] would
require extensive re-machining.
The proposed sampling and analysis configmtions of the Rheodyne valve are
shown in Figures 4 and 5. The liquid sample can be pumped through the jet mixer
by a suitable pump as shown in Figure 5, and the sample can be removed from the
loop and sent to the gas chromatograph. The homogenization scheme shown in
Figure 5 holds certain advantages over use of a jet mixer (outlined above) as
homogeneity of the samples could be achieved more rapidly. However, the risk of
stagnant spaces is greater and the equipment more complex.
3. Conclusions
Several procedures and many different types of equipment used for the measurement
of high pressure VLE measurement have been surveyed. Each equipment or
procedure has advantages and drawbacks. The choice of equipment will be
influenced by the systems to be investigated, e.g. expected relative volatilities or
101
JD. RMI and A L . Miihlbauer
possible multiple liquid-phase formation. Some types of equipment (e.g. for the
synthetic methods) may produce data more rapidly, but also have several drawbacks
as discussed above.
A new design for a static-type cell is proposed which should have minimal
disturbances of the equilibrium state when sampling the liquid phase, and it can
sample multiple liquid phases through a single flow-through type liquid sampling
valve. It can also furnish P-V-Tdata,but the construction is admittedly complex and
will require precise alignment. Another mechanism for producing a small flow
through a liquid sampling valve by use of the internal stirrer is also proposed.
Calibration of gas chromatograph detectors for gas mixtures or gas-liquid mixtures
will be facilitated with a new high-precision volumetric device capable of producing
concentrations from ppm to 50 mole%, as described in Part I. This survey of
equipment and procedures should prove useful to those entering the field or to those
seeking to improve their present equipment or procedures.
References
1. MUhlbauer, A. L. 1990. Measurement and Interpretation of High Pressure Vapour-Liquid
Equilibrium Data. PhD Thesis, University of Natal.
2. Rigas. T.J., Mason, D.F. and Thodos, G. 1958. Vapour-Liquid Equilibria Microsampling
Technique Applied to a New Variable Volume Cell. Ind. Eng. Chem.. 50(9). 1297-1300.
3. Todheide, K. and Franck, E.U. 1963. Das Zweiphasengebiet und die Kritische Kurve im
System Kohlendioxid-Wasser bis zu Druken von 3500 Bar. 2. Phys. Chem..37,38740 1.
4. Rogers, B.L. and Prausnitz. J.M. 1970. Sample-Exmaction Apparatus for High Pressure
Vapour-Liquid Equilibria. Ind. Eng. Chem. Fundam.. 9(1). 174-177.
5. Besserer. G.J. and Robinson, D.B. 1971. A High Pressure Autocollimating Refractometer
for Determining Coexisting Liquid and Vapour Phase Densities. Can. J. Chem. Eng., 49.
651-656.
6. Sagara, H., Arai. Y. and Saito. S. 1972. Vapour-Liquid Equilibria of Binary and Ternary
System Containing Hydrogen and Light Hydrocarbons. J. Chem. Eng. (Japan), 54,339348.
7. Klink. A.E., Cheh, H.Y. and Amich (Jr), E.H. 1975. The Vapour-Liquid Equilibrium of
Hydrogen-n-ButaneSystem at Elevated Pressures. AlChE J.. 21(6). 1142-1148.
8. Kalra, H., Kubota, H.. Robinson, D.B. and Ng. H.J. 1978. Equilibrium Phase Properties of
the Carbon Dioxide-n-HeptaneSystem. J. Chem. Eng. Data, 23(4). 317-321.
9. Ng. H.J. & Robinson, D.B. 1978. Equilibrium Phase Properties of the TolueneCarbon
Dioxide System. J. Chem. Eng. Data, 23(4). 325-327.
10. Figuiere, P.. Hom, J.F.. Laugier. S.. Renon, H., Richon. D. & Szwarc. H. 1980. Vapourliquid Equilibria up to 40 OOO Wa and 400OC: A New Static Method. AlChE J.. 26(5).
872-875.
11. Legret. D., Richon. D. and Renon, H. 1980. Static Still for Measuring Vapour-Liquid
Equilibria up to 50 bar. Ind. Eng. Chem. Fundam., 19,122-126.
12. Bae, H.K.. Nagahama, K. and Hirata, M. 1981. Measurement and Correlation of High
Pressure Vapour-Liquid Equilibria for the Systems Ethylene-1-Butene and EthylenePropylene. J. Chem. Eng. (Japan), 14(1), 1-6.
13. King, M.B., Alderson. D.A., Fallah. F.H.. Kassim, K U . . Sheldon, J.R. and Mahmud. R.S.
1983. Some Vapour-Liquid and Vapour-Solid Equilibrium Measurements of Relevance
I02
The Measurement of High Pressure Vapow-Ligukl-EquiIibria
Part 11: S t a k Methods
for Supercritical Extraction Operations and their Correlation. In Paulaitis, M.E..
Penninger, J.M.L., Gray, R.D. and Davidson. P. (eds), Chemical Engineering at
Supercritical Fluid Conditions. Ann Arbor: The Butterworth Group; pp.31-80.
14. Konrad, R., Swaid, I. & Schneider, G.M.. 1983. High-pressure Phase Studies on Fluid
Mixtures of Low-Volatile Organic Substances with Supercritical Carbon Dioxide. Fluid
Phase Equilibria, 10,307-314.
15. Ashcroft, S.J., Shean, R.B. & Williams C.J.J., 1983. A Visual Equilibrium Cell for
Multiphase Systems at Pressures up to 690 Bar. Chem. Eng. Res. Des.. 61, 51-55.
16. Guillevic, J.L., Richon, D. and Renon, H. 1983. Vapour-Liquid Equilibrium
Measurements up to 558 K and 7MPa: A New Apparatus. Ind. Eng. Chem. Fundam.,22,
495499.
17. Huang, S.S.S., Leu, A.D., Ng. H.J.& Robinson, D.B. 1985. The Phase Behaviour of Two
Mixtures of Methane, Carbon Dioxide, Hydrogen Sulphide and Water. Fluid Phase
Equilibria, 19.21-32.
18. Nakayama, T.. Sagara, H., Arai, K. and Saito, S. 1987. High Pressure Liquid-Liquid
Equilibria for the System of Water, Ethanol and 1,l-Difluoroethane at 323.2 K. Fluid
Phase Equilibria, 38, 109-127.
19. Wagner, 2. and Wichterle. I. 1987. High-Pressure Vapour-Liquid Equilibrium in Systems
Containing Carbon Dioxide, 1-hexeneand n-hexane. Fluid Phase Equilibria, 33.109-123.
20. Reiff, W.E.,Peters-Gerth. P. and Lucas, K. 1987. A Static Equilibrium Apparatus for
(Vapour + Liquid) Equilibrium Measurements at High Temperatures and Pressures.
Results for (methane + n-pentane). J. Chem. Thermodynamics, 19,467-477.
21. Milhlbauer. A.L. and Raal, J.D. 1991. Measurement and Thermodynamic Interpretation of
Vapour-Liquid Equilibria in the Toluene-C02 System. Fluid Phase Equilibria, 64,213236.
22. Kaka, H. and Robinson D.B. 1975. An Apparatus for simultanbeous measurement of
Equilibrium Phase Compositions and refractive index data at low temperatures and high
pressures. Cryogenics, 15,409-412.
23. Fredenslund, Aa., Mollerup, J. and Christensen, L.J. 1973. An Apparatus for Accurate
Determination of Vapour-Liquid Equilibrium Properties and Gas PVT Properties.
Cryogenics, 13,414419.
24. Swaid, I. 1983. Dissertation, University of Bochum.
25. Kobayashi, R.S. and Katz. D.L. 1953. Vapour-Liquid Equilibria for Binary
HydrocarbonWater Systems. Ind.Eng. Chem., 45(2), 440457.
26. Meskel-Lesavre, M.. Richon, D. and Renon. H. 1981. New Variable Volume Cell for
Determining Vapour-Liquid Equilibria and Saturated Liquid Molar Volumes by the Static
Method. Ind. Eng. Chem. Fundam., 20.284-289.
27. Rousseaux, P., Richon, D. and Renon, H. 1983. A Static Method for Determination of
Vapour-Liquid Equilibria and Saturated Liquid Molar Volumes at High Pressures and
Temperatures Using a New Variable-Volume Cell. Fluid Phase Equilibria, 11, 153-168.
28. Di Andreth. J.R. and Paulaitis, M.E. 1987. An Experimental Study of Three - and Four
Phase Equilibrium for Isopropanol-Water-CarbonDioxide Mixtures at Elevated Pressures.
Fluid Phase Equilibria, 32,261-271.
29. Deiters, U.K. & Schneider, G.M. 1986. High Pressure Phase Equilibria: Experimental
Methods. Fluid Phase Equilibria, 29,145-160.
103
J.D. Raal and AL.Miihlbauer
30. Japas, M.L. and Frank, E.V. 1985. High Pressure Phase Equilibria and PVT - Data of the
Water-Nitrogen System 673K and 250 MPa. Ber.Bunsenges. Phys. Chem., 89.793-800.
31. Laugier, S. and Richon, D. 1986. New Apparatus to Perform Fast Determinations of
Mixture Vapour-Liquid Equilibria up to 10 MPa and 423 K. Rev. Sci. Instrum.. 57.
469472.
32. Renon, H., Laugier. S.. Schwartzentruber, S. and Richon. D. 1989. New Determinations
of High Pressure Vapour-Liquid Equilibria in Binary Systems Containing nPropylbenzene with Nitrogen or Carbon Dioxide consistent with the Rausnitz-Keeler Test.
Fluid Phase Equilibria, 51,285-298.
33. de Loos, T.W..POOLW. and Lichtenthaler. R.N. 1984. Fluid Phase Equilibria in Binary
Ethylene + n-Alkane Systems. Ber. Bunsenges. Phys. Chem., 88,855-859.
34. Raal, J.D.1993. Unpublished work, contact author.
Received 4 M a y 1993;Accepted afer revision: 31 January 1994.
104
Документ
Категория
Без категории
Просмотров
1
Размер файла
1 032 Кб
Теги
measurements, part, high, method, pressure, vapour, equilibrium, liquid, statia
1/--страниц
Пожаловаться на содержимое документа