Fluids Handling

Working with
Non-Ideal Gases
Here are two proven methods for
Jimmy Peress, P.E.,
Tritech Consulting Engineers predicting gas compressibility factors.

N UMEROUS METHODS HAVE BEEN

proposed to model pressure-volume-temperature
(PVT) relationships of non-ideal gases (1). The
approaches described in literature are generally classified into
three groups. The first includes virial equations that are de-
available in standard handbooks (2). Thermodynamic ta-
bles for many common compounds often include PVT data
in addition to enthalpy data. When such data exist, Z can
be determined by direct substitution of the tabulated values
into Eq. 2. Product brochures may also provide a good
rived from statistical mechanics. The methods in the second source of PVT data for specific gases or vapors. The fol-
group are represented by analytical equations, such as the lowing example illustrates how Z can be determined from
Redlich-Kwong equation. The third group includes corre- tabulated thermodynamic data.
sponding-state correlations that rely on a dimensional analy- Example 1. A relief valve must be sized to relieve satu-
sis of the gas system to identify the key modeling parameters. rated ammonia at 317.2 psia and 127.4°F. Thermodynamic
This article focuses on two proven and relatively simple data for ammonia indicate that the specific volume, v, of
methods to predict the compressibility factor, Z. One method ammonia vapors at 307.8 psia (20.94 atm) and 125°F
is based on generalized compressibility charts. It is easy to (324.8 K) is 0.973 ft3/lb. The molecular weight, MW, of am-
use, but requires a graphical interpolation that introduces un- monia is 17.03 lb/lbmol. Calculate Z at relieving conditions.
certainty in the estimated value. The second method is de- Solution. The tabulated values are close to the relieving
rived from the Redlich-Kwong equation. It is somewhat te- conditions, so the calculated value of Z will be applicable
dious, but Z can be obtained by mathematical manipulation. for the relieving conditions. Z is obtained by substituting
the PVT data into Eq. 2 and using R = 1.314 atm-ft3/lbmol-
Compressibility factor defined K and MW = 17.03 lb/lbmol. Z = (20.94 × 0.973 ×
An ideal gas is represented by this equation of state: 17.03)/(1 × 1.314 × 324.8) = 0.813.

PV = nRT (1) Estimated compressibility factor

Z factor from generalized compressibility charts: Nel-
For a non-ideal gas, Z is defined as: son and Obert (3–4) developed a set of three generalized
charts representing, respectively, the low-pressure region
Z = PV/nRT (2) (0 atm < Pr < 1 atm), the medium-pressure region (1 atm <
Pr < 10 atm) and the high-pressure region (10 atm < Pr <
The compressibility factor can be viewed as that which cor- 40 atm). In these charts, Z is plotted as a function of Pr and
rects for the non-ideality of the gas. Thus, a gas for which Z Tr. The charts are also available in standard references
= 0.90 will occupy only 90% of the volume occupied by an (1–2). Estimation of Z involves the following steps:
ideal gas at the same temperature and pressure. The values 1. Set the operating temperature and pressure, T and P
of Z range from about 0.2 to a little over 1.0 for pressures 2. Obtain critical temperature and pressure, Tc and Pc
and temperatures of up to 10 times the critical values. Z it- 3. Calculate reduced temperature and pressure, Tr and Pr
self is a complex function of the reduced temperature Tr, re- 4. From the appropriate chart, locate Pr and Tr and read
duced pressure Pr and one or more other parameters, such the value of Z.
as the critical compressibility factor, Zc, or acentric factor, Example 2. Determine the Z factor for saturated ammo-
ω. No simple analytical equations have yet been offered to nia vapors at P = 307.8 psia (20.94 atm) and T = 125°F
express this complex relationship. (324.8 K) from the generalized compressibility charts. Tc is
405.7 K and Pc is 111.3 atm.
Experimental compressibility factor Solution. From the data above, Tr = 0.801 and Pr =
Where available, experimental Z values should be used 0.188. From the generalized compressibility charts in Ref.
for PVT calculations. Compilations of Z factors for some 2, Z = 0.84. The deviation from the experimental value de-
common gases and compounds as a function of T and P are termined in Example 1 is +3.3%.

39 www.cepmagazine.org March 2003 CEP

Fluids Handling

Finding Z from Redlich-Kwong equation of state: The that since the equation is cubic, three roots are possible,
Redlich-Kwong equation of state (5) has constants, a and b: but only one of the solutions will be physically meaningful
and will represent the correct value.
P = RT/(V – b) – a/( T1/2V(V + b)) (3) Example 3. Determine the Z factor for boron trifluoride
(BF3) at 68°F (293.15 K) and 1,500 psig (103.04 atm)
Eq. 3 can be combined with Eq. 2 to give: using the Redlich-Kwong equation of state. For BF3, Tc =
260.9 K and Pc = 49.20 atm. The Z factor reported by the
Z = 1/(1 – b/V) – a/(RT3/2V(1 + b/V)) (4) manufacturer at these conditions is 0.468.
Solution. From the data given above and using Eqs.
Eq. 4 can be rearranged to define Z as: 6–8: a*2 = 0.4278 × 260.92.5/49.20 × 293.152.5 atm–1 =
6.497 × 10–3 atm–1; b* = 0.0867 × 260.9/49.2 × 293.15
Z = 1/(1 – h) – (a*2/b*) h/(1 + h) (5) atm–1 = 1.568 × 10–3 atm–1; h = 1.568 × 10–3 × 103.04/Z =
0.1616/Z. Substituting the values of a*2, b* and h into Eq. 5
where a*2 = a/R2T2.5, b* = b/RT and h = b/V. gives Z = 1/(1 – 0.1616/Z) – [(6.497 × 10–3)/(1.568 × 10–3)]
From thermodynamic considerations, a and b in Eq. 4 × [(0.1616/Z)/(1 + 0.1616/Z)].
can be expressed in terms of the critical properties. A de- Rearranging and simplifying gives: Z = Z/(Z – 0.1616) –
tailed derivation of these terms is beyond the scope of this (0.6695)/(Z + 0.1616). Then, by trial and error, Z = 0.464.
article. Redlich and Kwong have shown that substitution of The deviation from the reported value is –0.9%.
the critical properties Tc and Pc, and rearrangement, lead to Example 4. BF3 is shipped in cylinders that have a net
the following equalities: volume of 43.6 L (1.54 ft3). The temperature of the gas in
the cylinders is 293.15 K and the pressure in the cylinders
a*2 = a/R2T2.5 = 0.4278Tc2.5/PcT2.5 atm–1 (6) is 103.04 atm. The Z factor reported by the manufacturer at
these conditions is 0.468. Estimate the weight of BF3 con-
b* = b/RT = 0.0867Tc/PcT atm–1 (7) tained in the cylinder.
Solution. Rearranging Eq. 2 and substituting the appro-
h = b/V = b*P/Z (8) priate values gives: w = MW × n = (67.80 × 103.04 ×
1.540)/(0.468 × 1.314 × 293.15) = 59.7 lb.
Z can be determined by first substituting the values for
T, P, Tc and Pc and solving the resulting expression by trial Compressibility factor of gas mixtures
and error. The built-in “goal-seek” capability of spread- Gas mixtures that display significant deviations from
sheets such as Microsoft Excel can be used to find the cor- ideal gas behavior can be modeled by using an average Z
rect Z value that satisfies the equation. It should be noted factor that is defined by the following relationship:

Nomenclature Zav = Z1x1 + Z2x2 + Znxn (9)

a = constant in Redlich-Kwong equation of state Z1, Z2, and Zn represent the Z factors for each component
a*2 = term defined in Eq. 5
contained in the mixture, and x1, x2, and xn represent the re-
b = constant in Redlich-Kwong equation of state
b* = term defined in Eq. 5 spective mole fraction of each gas in the mixture.
h = term defined in Eq. 5 Example 5. Compressed air is stored at 200 K under a
MW = molecular weight of gas, lb/lbmol pressure of 98.7 atm. At these conditions, Z = 0.6997 for
n = moles, lbmol O2 and Z = 0.8455 for N2. Determine the Z factor for air
P = pressure, atm containing 20.9% (by volume) O2. The tabulated Z factor
Pc = critical pressure, atm for air at these conditions is 0.8142.
Pr = reduced pressure, P/Pc, dimensionless Solution. The estimated Z factor is obtained by substi-
R = universal gas constant = 1.314 atm-ft3/lbmol-K
tuting the values into Eq. 9: Zav = 0.6977 × 0.209 + 0.8455
T = operating temperature, K
Tc = critical temperature, K × 0.791 = 0.8146.
Tr = reduced temperature, T/Tc, dimensionless
V = volume, ft3 Comparing methods
w = weight, lb The table lists Z factors estimated by the two methods re-
x = mole fraction of gas component in mixture viewed in this article and compares these values with experi-
Z = compressibility factor, dimensionless mental Z factors (where available) at the same T and P. The
Zav = average compressibility factor as defined in Eq. 9, dimensionless conditions selected are somewhat arbitrary, but generally re-
Zc = critical compressibility factor, dimensionless
flect conditions encountered in the chemical industries.
v = specific volume, ft3/lb
ω = acentric factor The analysis reveals that the two methods predict Z fac-
tors that are in good agreement with each other and with the

40 www.cepmagazine.org March 2003 CEP

 Table. Comparison between predicted and actual compressibility factors.*

Actual Conditions Critical Data Reduced Values Compressibility Factors

Case Compound Compressibility Redlich-Kwong
Molecular Actual Charts Equation
Weight T, K P, atm V, ft3/lb Tc, K Pc, atm Tr Pr Z Factors
Z % Error Z % Error
1 Acetic acid 60.05 533.15 21.27 NA 594.80 57.10 0.896 0.373 NA 0.80 NA 0.795 NA
2 Acetone 58.08 463.15 22.47 NA 509.10 47.00 0.910 0.478 NA 0.75 NA 0.734 NA
3 Ammonia 17.03 324.80 20.94 0.9730 405.70 111.32 0.801 0.188 0.8130 0.84 3.3 0.866 6.5
4 Boron trifluoride 67.80 293.00 103.04 0.0258 260.90 49.20 1.123 2.094 0.4682 0.45 –3.9 0.464 –0.9
5 Bromine 159.82 380.00 3.96 0.7640 584.00 101.65 0.651 0.039 0.9684 0.94 –2.9 0.955 –1.4
6 n-Butane 58.12 388.80 20.41 0.2761 425.20 37.50 0.914 0.544 0.6411 0.66 3.0 0.687 7.2
7 1-Butene 56.10 522.00 40.82 0.2460 419.60 39.70 1.244 1.028 0.8213 0.82 –0.2 0.808 –1.6
8 CFC-114 170.92 344.10 7.46 0.2934 418.90 32.20 0.821 0.232 0.8274 0.82 –0.9 0.843 1.9
9 Carbon dioxide 44.01 323.00 59.21 0.1183 304.20 72.90 1.062 0.812 0.7263 0.72 –0.9 0.722 –0.6
10 Carbon monoxide 28.01 300.00 100.00 0.1398 133.00 34.50 2.256 2.899 0.9934 0.99 –0.3 0.980 –1.3
11 Chlorine 70.91 299.80 7.93 0.6357 417.00 76.10 0.719 0.104 0.9074 0.88 –3.0 0.903 –0.5
12 Ethanol 46.07 427.15 6.80 1.6630 516.35 63.00 0.827 0.108 0.9282 0.92 –0.9 0.934 0.6
13 Hydrogen 2.01 300.00 100.00 2.0802 33.20 12.96 9.036 7.716 1.0607 1.05 –1.0 1.063 0.2
14 Methane 16.04 300.00 100.00 0.2101 190.70 45.50 1.573 2.198 0.8549 0.86 0.6 0.855 0.0
15 Methanol 32.04 487.15 39.25 0.3691 513.20 78.50 0.949 0.500 0.7251 0.73 0.7 0.762 5.1
16 Methyl chloride 50.49 349.80 19.31 0.3613 416.30 65.90 0.840 0.293 0.7664 0.78 1.8 0.808 5.4
17 Nitrogen 28.01 300.00 98.70 0.1433 126.20 33.50 2.377 2.946 1.0050 0.99 –1.5 0.990 –1.5
18 Oxygen 32.00 300.00 98.70 0.1191 154.80 50.10 1.938 1.970 0.9543 0.95 –0.4 0.947 –0.8
19 n-Octane 114.23 563.15 21.34 NA 569.00 24.50 0.990 0.871 NA 0.55 NA 0.551 NA
20 Sulfur dioxide 64.06 366.30 23.61 0.2460 430.70 77.70 0.850 0.304 0.7730 0.80 3.5 0.807 4.4
21 Toluene 92.13 553.15 24.07 NA 592.00 41.60 0.934 0.579 NA 0.68 NA 0.690 NA
22 Water 18.01 502.40 27.21 11610 647.00 218.30 0.777 0.125 0.8618 0.87 0.9 0.907 5.2

* Compounds for which the specific volume was not readily available in standard references are marked with “NA.” For these compounds, the actual Z value could not be calculated.

Z factors reported in literature. For the cases evaluated, the The generalized compressibility charts are based on ex-
Z factors interpolated from the generalized compressibility perimental data for 30 gases. Typical deviations are reported-
chart show an absolute average deviation of about 1.5% and ly 1–2%, but some gases such as methane, water and fluo-
a maximum deviation of about 4% for BF3. It should be romethane display deviations of up to 4% in different regions
noted that some regions of the charts are quite crowded and of the charts. The Redlich-Kwong equation is a well-estab-
the interpolation may be somewhat subjective. The lished method to model PVT relationships, but information
Redlich-Kwong method yields Z factors that show an aver- on its accuracy was not readily available to the author. It re-
age absolute deviation of about 2.5%, with a maximum de- portedly yields Z factors that are in good agreement with ex-
viation of about 7% for ammonia and n-butane. Its main ad- perimental values over a wide range of conditions, and is par-
vantage is that it allows the user to obtain Z factors by ticularly recommended for high-pressure conditions. Howev-
mathematical computation without the need for graphical er, this equation is not accurate at or near critical conditions.
interpolation. The use of a spreadsheet program with a
built-in “goal seek” capability can prevent computational Program available
errors and overcome the tedium of repetitive calculations. A Microsoft Excel-based template is available to calcu-
late the compressibility factor using the Redlich-Kwong
Literature Cited equation of state. Readers who are interested in obtaining
the template should contact the author via e-mail. CEP
1. Reid, R. and T. Sherwood, “The Properties of Gases and Liquids
— Their Estimation and Correlation,” 2nd Ed., McGraw-Hill Co.,
New York (1966). JIMMY PERESS is director of Tritech Consulting Engineers, Inc. (Jamaica, NY;
Phone: (718) 454-3920; E-mail: [email protected]). He has over 30 years
2. Perry, R. (Editor), “Chemical Engineers Handbook,” 5th Ed., Mc-
of experience in process engineering, process troubleshooting, software
Graw-Hill Book Co., New York. development and regulatory compliance (emission inventories and air
3. Nelson, L. and E. Obert, Trans. ASME, 76, p. 1057 (1954) as cited permitting). Prior to starting his consulting practice in 1983, Peress was a
in Ref. 1. process development manager and a senior consultant for Chem Systems,
4. Nelson, L. and E. Obert, “How to Use the New Generalized Com- Inc. (Tarrytown, NY). He earned a BS in chemical engineering from the Israel
pressibility Charts,” Chem. Eng., 61, (7) pp. 203–208 (July 1954). Institute of Technology and an MS in chemical engineering from the City
5. Redlich, O. and J. Kwong, “On the Thermodynamics of Solu- Univ. of New York. He is a professional engineer registered in New York, New
tions,” Chem. Rev., 44, pp. 233–244 (1949). Jersey and Delaware, holds seven U.S. patents and is a member of AIChE.

CEP March 2003 www.cepmagazine.org 41