1、GPA RR-25 77 3824b99 0004670 807 - Research Report RR-2 5 The Vapor-Liquid Equilibrium of the CH,-C02 System at Low Temperatures Project 739 S. C. Mraw S.-C. Hwang Riki Kobayashi William Marsh Rice University Chemical Engineering Department Houston, Texas April, 1977 Tulsa, Okla. 74103 - Phone: 918/
2、582-5112 Copyright Gas Processors Association Provided by IHS under license with GPANot for ResaleNo reproduction or networking permitted without license from IHS-,-GPA RR-25 77 3824699 0004671 743 = - FOREWORD Methane and carbon dioxide are constituents of natural gas. Carbon dioxide is usually loo
3、ked on as an impurity which can cause corrosion of flowlines and processing equipment, can freeze in a cryogenic plant designed for high ethane recovery, and is expensive to remove from the natural gas. As gas becomes more valuable, low quality natural gases can be processed economically to recover
4、the valuable components. A substantial amount of low quality natural gases contain significant amounts of nitrogen, methane, carbon dioxide, ethane, and hydrogen sul- fide. occurs on a widespread basis, gases with high quantities of carbon dioxide must be processed. To evaluate the possibility of ph
5、ysical separation and recovery of the valuable components, vapor-liquid equi- librium data are required. The methane-carbon dioxide binary system is an important system for processing gases, and this study was undertaken to obtain good data which can be used for the design of equipment for the cruci
6、al separations. In addition, when carbon dioxide injection into oil reservoirs The GPA K Value Steering Committee members and former members who aided in developing this project were Mel Albright, Jack Davis, Doug Elliot, Dan Forbes, Don Granicher, Gene Harlacher, Bob Jacoby, Mike Kesler, Karl Kilgr
7、en, John Sweny, Warren White, and Lyman Yarborough. The committee members express their appreciation to Professor Kobayashi and co-workers for their excellent work and diligence in completing this work. K Value Steering Committee M. A. Albright, Chahn Technical Section F Technical Data Development C
8、opyright Gas Processors Association Provided by IHS under license with GPANot for ResaleNo reproduction or networking permitted without license from IHS-,-GPA RR-25 77 U 3824699 OOOqb72 b8T - . TABLE OF CONTENTS Page rntroduction . 1 Experimental Results . 2 Table i . 4 Table II 5 Table 111 . 7 Tabl
9、e IV . 10 Experimental Derails . 14 Acknowledgement . 16 Reference 17 Appendix . 18 Copyright Gas Processors Association Provided by IHS under license with GPANot for ResaleNo reproduction or networking permitted without license from IHS-,-GPA RR-25 77 3824677 0004673 5Lb - - INTRODUCTION Pikaarl re
10、ported the compositions of the liquid phase in equilibrium with solid CO2 and of the vapor phase in equilibrium with solid COZ in the solid- liquid and solid-vapor regions, respectively, of the phase diagram for the system methane-carbon dioxide. Hwang et reported the compositions of the vapor phase
11、 in equilibrium with liquid in the liquid-vapor region of the dia- gram. In order to complete the phase diagram of this important system, the pre- sent work was undertaken to determine the liquid compositions in the same liquid vapor region as that studied by Hwang et al. For a typical example of th
12、e iso- thermal phase diagrams of the CH4-CO2 system at low temperatures, see Figure 1, which shows those portions of the diagram which have been determined by the various investigators. The liquid compositions in equilibrium with vapor were measured as a function of pressure on isotherms at -65.00,
13、-81.40, -94.00, -112.00, -130.00, -148.00, and -184.00F (-53.89, -63.00, -70.00, -80.00, -90.00, -100.00, and -120.00“C). The smoothed results were combined with the previous dew-point datas to provide tables and plots of isobaric and isothermal K-values. The ex- perimental method used was the vapor
14、-recycle method in a temperature-controlled equilibrium cell with chromatographic analysis of the phases. A calibration curve of detector response position of methane-carbon dioxide mixtures was prepare( w U 3 # # w a or the range .OZ to ,997 mole fraction methane. I . v r CONSTANT s+v . .- BUBBLE-P
15、OINT DATA (THIS WORK) -DEW-POINT DATA (HWANG et al 1 -*- FREEZING-POINT DATA (PIKAAR) / I /* -*o FROST-POINT DATA (PIKAAR) 0= 0. .M. .- -.-a- O MOLE FRACTION METHANE FIG. 1-Phase diagram for the methane-carbon dioxide system at low temperature -1- Copyright Gas Processors Association Provided by IHS
16、 under license with GPANot for ResaleNo reproduction or networking permitted without license from IHS-,-GPA RR-25 77 = 3824b77 O004674 Y52 - EXPERIMENTAL RESULTS As is the tradition in this laboratory when a new apparatus is built, some data points are taken in order to check the performance of the
17、new appara- tus against the previous, well-established equipment. In this case, some va- por compositions were measured at several temperatures since these could be compared to those arrived at by the elution method of Hwang et a1.2 When these vapor compostions were plotted vs. pressure at a given t
18、emperature, the present data a the higher temperatures fell on a curve parallel to the previous work, but displaced in composition. Typical data are shown on Table 1. It was originally though that each apparatus had an approximate error of about 1% in the ability to determine the mole fraction of ca
19、rbon dioxide in a given mixture. approximate error is given as 1% of the ratio mole fraction CHq/mole fraction COZ. However, for concentrations less than 20% C02, this error is of the same order of magnitude as the simpler formula given by Hwang et al., i.e. 1% of the mole fraction COZ) It was troub
20、lesome, therefore, to observe the discrepancy in mole fraction of CO2 shown in Table 1 at the higher concentrations of C02. Four factors were considered in trying to discover the source of the discre- pancy: 1) pressure, 2) temperature, 3) calibration error, and 4) failure of one or the other appara
21、tus to sample properly. (As mentioned in the calibration section of this paper, the 1) pressure could be ruled out since the values of the composition of the vapor phase at -65, -81.4, and -94F depend very little on pressure (see Figure 4) in the regions where the discrepancies were observed. Large
22、inaccuracies in the pressure measurement could be made and still not produce anywhere near the magnitude of the error observed in composition. 2) Temperature was considered in that either of the platinum thermometers used in the two apparatus might be inaccurate and tend to shift the data in just su
23、ch a parallel fashion as was observed. However, the two thermometers were inter-compared (as were their re- spective Mueller Bridges) and found to match easily within .lo“ F (.05“ C) in this temperature region. In fact, a third, independent thermometer was barrowed and found to check the thermometer
24、 used in the present work towithin .04“F (.02“C) This was nowhere near the temperature discrepancy needed to produce the observed discrepancy in concentration. Since the error is systematic, 3) calibration could be the reason for the discrepancy, although both calibration procedures were checked aga
25、inst the two Matheson “primary standard mixtures and found to agree far closer than could explain the discrepancy in Table 1. And, despite much consideration and experimentation to check the possibility that 4) “failure of one or ther other apparatus to sample correctly“ could be the reason, no such
26、 exp lana t ion cou Id be found . It seems probable then that the original estimates of 1% accuracy in Cali- bration combined with low temperature sampling may have been too optimistic for both methods and perhaps a 2 2% total error would be more realistic. alone does not remove the total discrepanc
27、y, one can only assume that the pre- sent case must be of those instances where several errors add in an infortunate manner to give a resultant discrepancy larger than originally expected. Other than this assumption, there is as yet no explanation for the observed discrepancy Since this -2- Copyrigh
28、t Gas Processors Association Provided by IHS under license with GPANot for ResaleNo reproduction or networking permitted without license from IHS-,-GPA RR-25 77 3824699 O004675 399 = - - It should be pointed out, however, that the difference between the pre- sent work and that of bang et al. is stil
29、l not a tremendous amount, and the difference gradually decreases with decreasing CO2 concentration until, at -112”F, there is no real discrepancy, even on a percentage basis, outside of experimental error. No comparisons between the two apparatus were made at temperatures lower than -112F. excellen
30、t extrapolation of the liquid phase data to the accurately known va- por pressure of pure CHq would seem to preclude any chance of a serious error at the temperatures -130 and -148F. However, s can be seen from Figure 5, the All the experimental data are shown in Table II and Figures 2 and 3, Note t
31、hat there is an inflection point in which are largely self-explanatory. and IV and in Figures 4 and 5. each of the liquidus curves for temperatures above the critical point of methane. It should also be pointed out that the data at -184F were taken under rather severe conditions. It can be seen from
32、 the calibration curve of Figure 7 that the calibration only extended to .997 mole fraction methane (.O03 CO2). the range of calibration, and the fourth is on the very end of it. Smoothed data are given in Tables III Therefore three of the compositions reported at -184F are beyond To arrive at these
33、 approximate values for composition, the calibration data were extrapolated. Therefore, the points at -184F must carry more than the usual encertainty in composition. Furthermore, it will be seen that all the pressures reported at -184F are actually above the accepted vapor pressure of pure CHq (171
34、.0 psia). 3 This is especially surprising since the extrapola- tion of the results at -130 and -148F pointed right at the vapor pressure of CH4, Figure 5. The discrepancy in pressure at -184F can only be attributed to the 1000 psia Heise gauge used in the present work, whose stated accuracy is 0.251
35、, 6.34f -76 , 5 0,200 6-00 b 31.4a 1.0 6-80 - 4.0 0,702 6.27 -40 , O 0,418 5-50 -65.0 - 0.263 4.77 -69.88; -79.0 O,23ge O . 197 4.56f 4.11 a, Saturation temperature of carbon dioxide from Ref, - 6 6 c. Triple-point temperature of carbon dioxide from - f. KsLv cH4 d. Triple-point temperature of mixtu
36、re, Ref, ( 7 , 8 ) SLV =, %O2 - 10- Copyright Gas Processors Association Provided by IHS under license with GPANot for ResaleNo reproduction or networking permitted without license from IHS-,- GPA RR-25 77 W 3824697 0004683 465 W MOLE FRACTION CO2 0 Hwang et al, Ref. 2 A Vapor pressure of CO,. Ref.
37、6 - O 0.20 0.40 0.60 080 I .o MOLE FRACTION CH4 54c 520 500 480 O u 4 _- a E 380 360 3 u) Il MOLE FRACTION CO2 0.05 0.0 4 0.03 0.02 0.01 O I I I I -148.00 OF / 0 This work Hwang et al, Ref. 2 n Vapor pressure of methane, Ref. 3 I I I 0.95 0.96 0.97 (171) ( 170) - 184.00 OF (169) I72 I I I I 0.992 0.
38、996 1 0.98 0.99 1.00 FIG. 2 - Pressure-composition diagram for methane-carbon di- oxide system for above critical temperature of methane FIG. 3 - Pressure-composition diagram for methane-carbon di- oxide system for temperature below critical temperature of methane. Dotted lines at -130“ and -148“ F
39、represent vapor lines normalized for slight Heise gauge discrepancies. Pressure in parenthesis at -184“ F represent pressures when the lines are nor- malized to the known vapor pres- sure of methane, 171.0 psia, Ref. 3. MOLE FRACTION CHA Copyright Gas Processors Association Provided by IHS under lic
40、ense with GPANot for ResaleNo reproduction or networking permitted without license from IHS-,-6 51 41 31 2i Il 9.1 8.1 7.1 6.1 5.1 .% 4.1 Y 0 3. s I- = 2 d g 2. Q W 1.1 O. O. O. 0.1 O. O. O. O. O. GPA RR-25 77 3824699 0004b84 3Tl 1 I ,1111 CO, TRIPLE POINT h LEGEN0 THIS WORK KAMINISHI, ET AL (1968)
41、PREDICTED -_ I - - - - - - - - - - - - -_ LOCUS OF K? (NOT AN ISOTHERM) (NOT AN ISOTHERM) - - LOCUS OF KsLv - Tt (CO,) -69 88 F P CO2 CRITICAL CO, TRIPLE POINT -184.F -148.F -130F _ .-. T ru I-IICL*C - I I rn Y-, TO K.0.157 AT P.1.7 PSIA P P, ( CH, 1 IP-65.F 11111 L I l 11 ,111 2000 60 70 80 90 I00
42、200 300 400 500 600 800 io00 PRESSURE PSIA FIG. 4-Isothermal K values for methane-carbon dioxide system -12- Copyright Gas Processors Association Provided by IHS under license with GPANot for ResaleNo reproduction or networking permitted without license from IHS-,-6C IO K I .o O. I GPA-RR-25 77 9 38
43、24b 0004b85 238 9 IL * O I II t- LL o O I II I- - - - t o u- O, E- in order to maintain the temperature constant at the desired value, the cooling was then balanced by heat supplied by a Thermotrol temperature controller which main- tained the temperature constant to generally better than ? .02“ F (
44、2 .OluC). The vapor in the cell was recycled through the liquid by means of a magnetic pump5 until equilibrium was achieved. CO2 and CHq were introduced into After equilibrium was reached, the sample line to the chromatograph was e- vacuated, and a valve mounted on the cell was opened to withdraw sa
45、mple into the line. The parts of the sample lines which pass through the bath fluid are heated to ensure immediate vaporization of the sample. Before entering the chromato- graphic oven, the sample streams entered a stainless steel mixing vessel, consist- ing of a teflon-coated stirring bar which, w
46、hen magnetically driven, agitated a number of glass beads against the walls and lid of the vessel. The emerging gas stream was thus ensured of complete homogeneity. This stream was then taken in- to the chromatographic unit and analyzed as has been described for the calibration mixtures. The methane
47、 used was Matheson Purity CH4 (minimum purity 99.99%) and the car- bon dioxide used from Ma theson Gas Products . was Coleman Instrument Grade CO2 (nimimum purity 99.99%), both - 14- Copyright Gas Processors Association Provided by IHS under license with GPANot for ResaleNo reproduction or networkin
48、g permitted without license from IHS-,-GPA RR-25 77 3824699 0004687 000 = CORRESPONDING PERCENT CH 4 2 IO 50 90 99 99.9 1 I I l I -2 -I O I 2 3 LOG (ACH, 1 Aco, 1 FIG. 7-Calibration points for the methane-carbon dioxide system -15- Copyright Gas Processors Association Provided by IHS under license w
49、ith GPANot for ResaleNo reproduction or networking permitted without license from IHS-,-GPA RR-25 77 3824699 0004b88 T47 H ACKNOWLEDGEMENT We thank Ray Martin for his superior effort in the excellent mechanical design and construction of the apparatus, and Fran Leland for his help with much of the electronics
