1、91 GPA TP-LZ) 85 3824699 OOLL248 963 Technical Publication TP-12 Liquid Densities of Ethane - Propane Mixtures W. R. Parrish Phillips Petroleum Co. Ba rtlesvi I I e, Oklahoma February 1985 ,812 First Place Tulsa. Okia 74103 Phone 918i582.51i2 GPA TP-32 85 II 3824699 0033249 8TT COMPRESSED LIQUID DEN
2、SITIES OF ETHANE-PROPANE AND PROPANE-NORMAL BUTANE MIXTURES FOREWORD Each day, millions of barrels of ethane-propane (EP) mixtures and natural gas liquids (NGL) are bought and sold in the United States. However, for both types of mixtures there are limited acclirace density data in the compressed li
3、quid region where metering occurs. liquid density data for binary ethane-propane mistures and for binary propane- n-butane mixtures in the 50 to 122F (and 140F for the propane-butane system) at pressures up to 1400 psia. In addition, limited ethane-propane-n-butane data also are included. This publi
4、cation presents compressed The data were taken at even increments in temperature, press=re and composi- tion. This permits determining the effect of each variable on ti-. ensity with little or no interpolation. Therefore, the data should be valuab in developing and testing new density correlations.
5、The publication consists of three separate parts. Part I presents experimental results for EP mixtures. Where possible the data are compared with the work of others. Part II compares the experimental data with three different methods for calculating EP densities. Accurately predicting densities of E
6、P mixtures is coqlicated by the fact that the mixture may be in the critical region which is difficult to model-. The COSTALD method gives fairly good results except in the region near the critical temperature of the mixture. A modification of the extended correspondinn states method was found to gi
7、ve the most accurate results over the region of interest. A FORTRAN listing of the program is shown as an appendix. Part III compares the measured densitites of propane-butane and ethane-propane- butane mixtures with the data of other investigators and with two methods Lor predicting densities. For
8、this system, both COSTALD and a modified version of the extended corresponding states method worked well. The actual data are tabulated in an appendix. (j47 :b ar1 Sutton Secre tarp 1 GPA TP-12 85 382Yh77 0011250 511 Part I Wm. R. Parrish Phillips Petroleum Company Bartlesville, Oklahoma 74004 2 GPA
9、 TP-12 85 3824699 OOLL25L 458 M Abstract Densities of compressed liquid ethane-propane mixtures are reported at five temperatures between 10.00 and 48.9OoC. cover eight compositions ranging between 30 and 95 mol percent ethane at pressures up to 9.65 MPa. tive, extremely dependent on temperature and
10、 pressure, and can represent up to a 50 percent correction in mixture volume. For each isotherm the data The excess volumes derived from the data are nega- i 3 Introduction Cryogenic distillation, combined with demand for ethane as a feed- stock, makes ethane-propane (EP) mixtures an important commo
11、dity chemical. Typical custody transfer measurements for EP mixtures involve mass measurement via a densitometer and a turbine flowmeter. also must be analyzed for ethane since sales are based on mass of ethane transferred. pressure and composition and then predict density. accurate predictive techn
12、ique which is based on extensive, accurate ex- perimental data in the temperature region where ethane, and possibly the mix- ture, are close to the critical temperature. However, samples of the mixture It would be extremely Useful to be able to measure temperature, To do this requires an Seven sets
13、of density data for liquid EP mixtures exist in the open literature. temperatures and compare their values with those of Shanaa and Canfield (1968). Recently, Orrit (1983) reported Saturated liquid densities for a nearly equimolar EP mixture at temperatures between 94 and 230 K. (1973 ) presents sat
14、urated liquid densities at 15.6OC compressed liquid data for four binary EP mixtures st emperatures between 7.2 and 84.9OC and at pressures up to 138 bar. mixtures containing small amounts of methane and n-butane. Acosta (1975) measured compressed liquid densities for three EP mixtures at -45.5, -17
15、.8, 10.0 and 37.8OC at pressures up to 13.8 MPa. report densities of EP mixtures at temperatures ranging from O to 6OoC and at pressures up to 138 MPa. between two and nine additional components. peratures greater than -45.5OC were obtained using dilatometry. Hiza, et at. (1977) report Saturated liq
16、uid densities at WC Kare Tomlinson (1971) lists He also obtained densities on p Finally, Provence, et al. (1972) However, all of the mixtures studied contained All of the density data at tem- GPA TP-12 85 U 3824699 0033253 220 = This paper presents compressed liquid densities at 10.00, 15.56, For ea
17、ch isotherm, data were obtained for eight 26.67, 37.80 and ,48.90C. compositions, ranging from 30 to 95 mol percent ethane at pressures up to 9.65 MPa. These data should be useful for systematically evaluating the ef- fect of temperature, pressure and composition on the accuracy of existing density
18、correlations and for providing an aid in developing new correlations. Experimental Apparatus and Procedure Figure 1 shows a -matic of the experimental apparatus. The sys- tem contains a vi0 .sing tt; ,ensitometer (Mettler-Paar Model DMA 512), a 100 an3 positive displacement piston pump and a magneti
19、cally-driven mixing pump. These components are immersed in a constant-temperature glycol bath which is stable to - +O.0loC. A calibrated, platinum resistance thermometer monitors the bath temperature. System pressure is found by measuring the pressure required to null a differential pressure gauge h
20、aving a resolution of 70 Pa. pressure is read using a bourdon gauge with a digital readout resolution of 0.01 psia (70 Pa). gauge. 250 cm3 positive displacement pumps. jacket for temperature stability. (+O.iC) - at 24.5OC which is within 0.5OC of ambient temperature. fluid temperature is monitored u
21、sing a calibrated thermocouple. 0-2000 psia (0-13.8 MPa) pressure transducer measures the pressure in the pumps and manifold to the system. displacing weighed amounts of pure water. The nulling This gauge was calibrated against an oil dead-weight Pure ethane and propane are charged into the system b
22、y using two These motor-driven pumps have a water Water in the jacket is maintained constant The actual A calibrated Pump volume displacement was calibrated by 5 GPA TP-12 5 e 3824699 0011254 167 = Prior to, and following each the densitmeter was calibrated at For the 10 1.4 MPa increments using pur
23、e ProPane and either ethane or argon. and 15.6OC isotherms, pure ethane was Used. pressure at 26.7OC and compressible behavior at the higher temperatures made it difficult to use as a calibration fluid. these temperatures. molecular sieve bed in an ice bath. pane with stated purities of 99.9 and 99.
24、99 mol percent, respectively, were used without further purification. However, ethanes high vapor Therefore, argon was used at Pure grade argon (99.998 percent) was dried over a Phillips Research Grade ethane and pro- After calibrating the densitometer and pressure transducer at a given temperature,
25、 pure ethane was metered into the densitometer - 100cm3 piston pump system. desired composition. A known amount of propane then was added to obtain the The fluids were added slowly so that the liquid-full 250 em3 and 100 cm3 line pressure. accuracy of the piston pumps could be adjusted to This preve
26、nted vaporization which composition measurement. maintain a constant (25OkPa) could adversely affect the Once the materials were add.a the pure fluid densities. analysis was made for the 37.8OoC isotherm which involved using argon as a calibration fluid. Uncertainties in temperature, pressure and eq
27、uation of Uncertainty in the volume of pure fluid added Table 1: lists the estimated total uncertainty (defined as three This error The error analysis shows that the major uncertainty in the experi- ment is due to the uncertainty of the pure fluid densities. tainty in the composition, the error will
28、 be systematic instead of random since the pure fluids were charged at the same temperature and pressure each time. computed composition.) Excluding the pure fluid uncertainty, temperature be- comes the most important source of error primarily because it affects the pure fluid densities more than pr
29、essure. For the uncer- (A gas chromatographic analysis of a mixture sample agreed with the Discussion of Results I Of the seven sets of EP mixture densities in the literature only 9 GPA TP-12 85 3824639 0011258 802 D Acostas (1975) data can be coipared directly with the present work. shows the effec
30、t of composition on the excess volume at 10.0 ad 37.80%. lines were computed from a Redich-Kister expansion (Prausnitz, 1969, pp 195) fitted to the present data. The figure also compares our work with Acostas data for the 6.89 MPa isobar. The excess volumes were computd from his re- ported densities
31、 and the pure fluid densities computed from the modified BHR equation of state given in the appendix. He claims an uncertainty of +0.4%. Based on the uncertainties for the two experiments, the data agree well. Figure 3 The - Figure 4 shows the effect of pressure at 10.0 and 37.8OC on the equimolar e
32、xcess volume. Acostas data at 0.50133 mole fraction ethane are included for comparison. Again, the agreement between investigators is goOb. For the present work, the equimolar value was determined from the fitted Red- lich-Kister expansion. isotherm were omitted to provide resolution to the figures.
33、 Values for pressures below 6.89 MPa on the 37.8OC Finally, Figure 5 shows the dramatic effect of temperature and pressure on the equimolar excess volume. As expected, the pressure effect increases rapidly with increasing temperature since ethane becomes so com- pressible in this temperature range.
34、Conclusions Densities of ethane-propane mixtures have been measured in the 10 to Because the data are near the critical temperature of 49OC temperature range. ethane, the mixtures show large, negative excess volumes which are strongly dependent on temperature and pressure. 50% deviation from ideal m
35、ixing and must be considered if liquid densities are These excess volumes represent up to 10 GPA TP-12 85 E 3824699 001Li59 7q9 _- - to be computed accurately from temperature, pressure and composition measurements. 1 - In a later paper we will evaluate various density computation tech- I niques usi
36、ng the results reported here along with other literature data. Appendix To compute the pure fluid densities needed in this work we usd the 32-term modified equation of state (McCarty, 1980). The mathematical form of the equation is, P = ,AT + p2(C(l)T + G(2)Tli2 + C(3) + C(4)/T + C(5)/T2) + p3(C(6)T
37、 + C(7) + G(8)/T + C(9)/T2) + p4(C(10)T + C(11) + G(12)/T) + p5(C(13) + p6(G(14)/T + G(15)/T2) + p7(C(16)/T) + p8(C(17)/T + C(18)T2) + p9(C(i9)/T2) + p3(2)/2 + (213) exp(yp2) + p5(C(22)/T2 + G(23)/T4) exp(rp2) + p7(C(24)/T2 + G(25)/T3 exp(rp2) + p9(G(26)/T2 + G(27)/T4) exp(rp2) + p1l(G(28)/T2 + G(29
38、)/T3) exp (rp2) + P(G(O)/T + C(3i)/T3 + G(32)/T4) exp(yp2) where P, p, R and T represent pressure, density, gas constant and absolute temperature, respectively. ing the densities of pure argon, ethane and propane. obtained by PVT data sinuitaneously with other properties including heat capacities. U
39、nits for the ethane and propane equation are bar (lo5 Pa), mol/liter and Kelvin. liter and Kelvin. Table A-1 lists the coefficients used for comput- These coefficients were For argon, they are atmospheres (0.01325 MPa), mol/ 11 GPA TP-12 5 3824677 OOLL2bO 460 = Acknowledgments The author thanks R. L
40、. Brandon, J. L. Dum and M. R. Lee for their contributions in constructing the apparatus and taking the experimental data. Uso, we thank D. L. Embry for developing the computer program to calculate the pure fluid densities. Conversion Units Used in this paper S.I. Density Mass Volume 8 cm3 10-3 10-3
41、 106 .3 10-3 m3 Temperature OC OC+273.15 E I 12 References Acosta C., J. R., - 1975. pane, and n-Butane. Ph.D. Thesis, Kansas University, LdWrenct Kansas, 272 pp. Densities and Viscosities in the Sys+ il Ethane, Pro- Ely, J. F., personal communication, 1982. Hiza, M. J., Haynes, W. M. and Parrish, W
42、. R., - 1977. sities and Excess Volumes for Binary Mixtures of Low Molar-Mass Alkanes and Nitrogen between 105 and 140 K. J. Chem. Thermo., 9:873-96. Orthobaric Liquid Den- Kahre, R. C., - 1973. 3. Chem. Eng. Data, 18:267-70. Liquid Density of Light Hydrocarbon Mixtures. McCarty, R. D., - 1980. mody
43、namic and Transport Properties of Selected Cryogens. Tech. No. 1025. National Bureau of Standards, U.S. Interactive FORTRAN IV Computer Program for the Ther- Oritt, J. E., 1983. - Mixtures. Fluid Phase Equilibria, 12:253-281. Orthobaric Liquid Densities of Natural-Gas-Component Prausnitz, J. M., B.
44、Prentice-Hall, Englewood Cliffs, New Jersey, 523 pp. Molecular Thermodynamics of Fluid-Phase Equilibria. Provence, T. K., Wiener, L. D. and Walton, D. K., - 1972. High-Ethane Raw Make Streams. sociation, Tulsa, Oklahoma, 124 pp. Liquid Densities of Technical Publication TP-2. Gas Processors As- Shan
45、aa, M. Y. and Canfield, F. B., 1968. Liquid Density and Excess Volume of Light Hydrocarbon Mixtures at -165%.rans. Faraday Soc. 64: Part 9: 2281-6. Tomlinson, J. R., - 1971. Mixtures. Oklahoma, 21 pp. Liquid Density of Ethane, Propane, and Ethane-Propane Technical Publication TP-1, Gas Processors As
46、sociation, Tulsa, 13 GPA TP-12 85 m 3824697 00112b2 233 Table 1. Typical Densitometer Calibration at 26.7OC using Argon and Propane as the Calibration Fluids mon Propane Calibra tion2 Pressure 2.76 0 * 04474 1.400429 0 49546 1.467503 2.3431 4.14 O 06754 1.403939 0 49933 1 4680% 2 * 3433 5.52 O. 0906
47、0 1.407487 O. 50293 1.468668 2.3432 6.89 0.11392 1.411070 O. 50630 1.469191 2.3439 8.27 O. 13745 1 + 414673 O 50948 1.469489 2.3444 9.65 0.16113 1.418291 O. 51248 1.470165 2 3449 Average 2 3438 1. Pure fluid densities are cmputed from the modified BWR equation of state (see appendix). tency only and
48、 do not represent absolute accuracy. Five significant figures are given for internal consis- 2. Calibration constant calculated using Equation 1. GPA TP-12 85 3824699 OOL12b3 1T - - _- Table 2. Comparison of Measured and Calculated Ethane Densities Tempera ture Pressure (OC) MPa ) 26.7 37.8 48.9 5.5
49、2 6.89 e 89 7 65 5.52 6.89 8.27 9.65 Dgnsity Measured 1 o. 3397 o. 3579 O. 3136 O. 3369 o i9 o. .*39 o. 2299 O. 2918 O. 3182 0.06 0.04 0.03 0.03 0.01 -1.26 0.42 -0.04 -0.03 I k GPA TP-12 85 = 3824679 0011264 O06 Table 3 Densities and excess volumes of Ethane-Propane Mixtures as a function of temperature, composition, and pressure. Exces Volume DensiSy (cm 4 /mol) Temperature Mol Fraction Ethane Pressure (OC 1 (MPa 1 Wcm 1 o. 9511 4.14 O. 3957 -0.8 5.52 O. 4042 -0.6 10.00 6.89 O. 4112 -0.5 8.27 O. 4173 -0.4 9.65 O. 4227 -0.3 O. 9006 O. 8005 O. 7007 2.76 O.
