[PDF] Vapor-Liquid Equilibrium Data on the System Natural Gas-Water

View - Download Vapor-Liquid Equilibrium Data on the System Natural Gas-Water

Previous PDF

Next PDF

T.P. 2848

VAPOR.LIQUID EQUILIBRIUM DATA ON THE SYSTEM

NATURAL GAS·WATER·TRIETHYLENE GLYCOL AT

VARIOUS TEMPERATURES AND PRESSURES

JOE A. PORTER,* BLACK, SIVALLS AND BRYSON, INC., KANSAS CITY, ·MO., AND LAURANCE S. REID, UNIVERSITY OF OKLAHOMA, NORMAN, OKLA., MEMBER AIME Gas dehydration plays an important part in the production of natural gas. Effective dehydration prevents formation of gas hydrates and the accumulation of water in transmission systems,"·" insuring uninterrupted gas deliveries at maximum efficiency under the most adverse weather conditions. At the present time, most gas companies require a maximum water vapor content of seven Ib per million standard cu ft of gas, so that virtually all gas tendered for sale must be dehydrated to meet this specification. For a number of years it has been common practice to pro duce gas and gather it at a complOn point for dehydration prior to discharge into the transmission system.""U,16,l1 How ever, higher transmission line pressures, long gathering lines and relatively low ground temperatures have made it neces sary to dehydrate gas at, or near, individual wells in order to gather gas from a number of'llewly developed fields without unusual difficulty. Where gas has been dehydrated at pressures ranging from

300 to 800 psi in the past, future trends indi

cate that these processes may be operated at pressures as high as

2,000 psi.

Economics of gas dehydration

are of great importance, par ticularly where facilities must be provided to process rela tively small quantities of gas, such as the production from an individual well. Although the adsorption of water vapor from gas on a granular sorbent material such as activated bauxite, activated alumina, or one of the gels is highly effective and produces virtually "bone dry" gas, the cost of a small unit of this type is substantially greater than that of an absorption process which, through proper selection of the absorbing liquid, will dehydrate the gas sufficiently to meet pipe line specifications.

For this reason, a great deal of empha

sis has been placed on the development of small, inexpensive dehydration units·'" and the search for more effective absorb ent liquids has been intensified. lReferences are given at end of paper. Manuscript received at the office of the Petroleum Branch September

27, 1949. Paper presented at the Branch Fall Meeting in San Antonio,

Texas, October 5-7, 1949.

*American Gas Association Supply Men's Research Fellow in Chemical Engineering at the University of Oklahoma. This paper is a condensa tion of a thesis submitted by J. A. Porter in partial fulfillment of the requirements for the M.Ch.E. degree. conferred by the University of

Oklahoma in August. 194V.

A wide variety of methods for dehydrating gas are known' and many of these have been used in industry. Earlier appli cations of the absorption process employed concentrated solu tions of calcium and lithium chlorides as the absorbent. The severe corrosion problems inherent in handling these solutions and the relatively small dew point depressions obtained caused early abandonment in favor of, or conversion to, diethylene glycol when it was found that aqueous solutions of this organic liquid were more hygroscopic than the brines and were non corrosive. Processes employing diethylene glycol-water solu tions are widely used for gas dehydration at pressures ranging as high ai 1,200 psi."'''''' At nominal pressures a dew point depression of

45° to 50°F may be obtained and the data of

Russell

et al!' indicate that a minimum dew point is obtained from the effluent gas at a pressure of approximately 1,200 psi when the gas is in equilibrium contact with a 95 per cent by weight diethylene glycol solution. In a number of instances the dew point depression obtained with diethylene glycol-water solutions is not sufficient to produce a specification product without cooling the inlet gas.

In a recent search for a better

absorbent, triethylene glycol was used in a small commercial dehydration unit and subjected to rather exhaustive field tests.· The data obtained were encouraging and indicated that, at pressures ranging from 300 to 500 psi, triethylene glycol produced a substantially greater dew point depression than diethylene glycol. These results led to an investigation of the system natural gas-water-triethylene glycol in an effort to obtain vapor-liquid equilibrium data, to determine pressure limitations, and to develop other data pertinent to the design of gas dehydration processes.

A review of the literature

has failed to reveal any data which permit reasonably accurate calculation of the vapor liquid equilibrium conditions for a solution of water and tri ethylene glycol in contact with natural gas at high pressure. Since these constituents form a non-ideal system, the Poynting equation,·,2l or the usual combination of Raoult's and Dalton's laws,·,,·,22 would not be valid. Correction of Raoult's and Dal ton's laws by the use of activity coefficients" is not feasible for available data are insufficient for the prediction of the actual increase in the ratio of the activity of one component in the vapor phase to its activity in the liquid. Therefore, ex peri-

Vol. 189, 1950 PETROLEUM TRANSACTIONS, AIME 235 Downloaded from http://onepetro.org/JPT/article-pdf/2/08/235/2238744/spe-950235-g.pdf by guest on 02 October 2023

T.P. 2848

VAPOR-LIQUID EQUILIBRIUM DATA ON THE SYSTEM NATURAL GAS-WATER- TRIETHYLENE GLYCOL AT VARIOUS TEMPERATURES AND PRESSURES

DEW POINT ,

T"'ESTER ,--lQ1-----

FLOW DIAGRAM OF APPARATUS

FIG. 1

mental equilibrium data were observed in the laboratory, using

95 per cent by weight triethylene glycol, 5 per cent water solu

tion in contact with a relatively nitrogen-free natural gas at pressures ranging from 500 to 2,000 psig and at temperatures ranging from 60° to lOO°F.

APPARATUS

The experimental apparatus used in this investigation is shown in Fig. 1 and is, essentially, "a duplicate of the con tinuous flow system used by Russell" to investigate the system natural-gas-water diethylene glycol.

In this system, the natural

gas was compressed to the desired pressure by a three-stage Rix compressor and passed through a coil in a constant tem perature bath to a saturator where the moisture content of the gas was adjusted to the desired level. Saturated gas from this vessel then passed through a gas scrubber where any entrained water was separated and removed from the gas. Saturation conditions were checked by dew point measurement of the gas at a point immediately downstream from this scrubber."

A measured quantity of

saturated gas was then passed to a second constant temperafure bath containing the gas-liquid contactor and a separator. A measured quantity of 95 per cent triethylene glycol solution was pumped into the contactor coil by means of a gas-powered plunger pump so that the gas and glycol solution flowed concurrently in intimate equilibrium contact to the separator, where the two phases were separated and removed from the system. The effluent gas stream was split, with one portion passing through a

U. S. Bureau of Mines Dew Point Tester maintained

at system pressure, thence to a pressure-reducing regulator prior to measurement at atmospheric pressure in a wet test meter. The second portion was passed to a second pressure reducing regulator, thence to a train of drying tubes filled with anhydrous magnesium perchlorate where water and glycol vapors were adsorbed prior to measuring the gas in a second wet test meter. Gas from both meters was then vented to atmosphere. The glycol solution was withdrawn from the separator at a constant rate in order to maintain a uniform level in that vessel. A portion of this solution was accumulated under pres sure in a separate vessel, equipped with a high-pressure reflex liquid level gauge glass, where the solubility of gas in the glycol solution was determined at the end of each run. Accessory apparatus included a special viscosimeter used to determine the glycol concentration of all solutions used in this investigation.'·

EXPERIMENTAL MATERIALS AND

TECHNIQUES

The natural gas used in this inveitigation was obtained from the mains of the Oklahoma Natural Gas Co. at Norman, Okla. A representative analysis of this gas is shown in Table I. The triethylene glycol used was a technical grade obtained from the

Carbide and Carbon Chemicals Division of Union

Carbide and Carbon Corp. Physical constants and other data pertaining to this glycol are shown in Table II. The glycol was blended with distilled water to the desired concentration of

95 per cent by weight glycol and 5 per cent by weight water.

A specific

operating procedure was used throughout the investigation which was developed from preliminary experi ments. In pressuring the apparatus, the saturator was by passed to prevent condensation of moisture in the system. When system pressure reached 200 psig, glycol injection was started. After the desired operating pressure was attained, the gas was turned through the saturator and the gas flow to the contactor was adjusted to a rate of six standard cu ft of gas per hour. The glycol injection rate was adjusted to 60 gal per MC£ of gas. These rates, and the established temperatures and pressures, were maintained constant throughout the run. After operating conditions were established, the system was op erated for one hour before any samples were drawn for analysis or effluent gas dew points observed.

At the end of this pre

liminary period, an inlet gas dew point was observed to start the run and the drying tube train was connected into the proper effluent gas stream. A total volume of approximately 20 standard Cli ft of gas was passed through the drying tubes in order to minimize error. The weight of the glycol vapor in the effluent gas stream was obtained by difference, calculated from the pounds of water and glycol adsorbed from one cu ft of gas passing through the drying tubes, minus the pounds of water vapor per C4 ft of gas as determined by dew point measurement in conjunction with the water vapor content correlation by

McCarthy et al." Prior to use, the adsorbent

used in the drying tubes was saturated with hydrocarbons by passing dehydrated gas through the train, thus insuring that only glycol and water vapors would be adsorbed during the investigation.

Dew points of

inlet and effluent gas were determined by means of a

U. S. Bureau of Mines Dew Point Tester, using

liquid propane as the refrigerant. Preliminary investigation of the effect of glycol vapors on the dew point of gases treated with triethylene glycol failed to reveal errors of the magnitude predicted by Reisenfeld and Frazier" so that the dew point method was accepted as adequate. It is interesting

Table I

Analysis

of Natural Gas

Component

Carbon Dioxide

Nitrogen

Methane

Ethane

Propane

Butanes

Pentanes

Hexanes

Heptanes and Heavier Gas

Volume %

or Mol % 0.70 1.95 82.52
8.20 3.67 1.86 0.72 0.19 0.19

100.00

Specific Gravity (Re: Air = 1.000): 0.6932

236 PETROLEUM TRANSACTIONS, AIME Vol. 189, 1950 Downloaded from http://onepetro.org/JPT/article-pdf/2/08/235/2238744/spe-950235-g.pdf by guest on 02 October 2023

JOE A. PORTER AND LAURANCE S. REID

T.P. 2848

Table II

Some Physical Constants and Properties of

Triethylene Glyco}23

Structural Formula HO·CH,·CH,·O-CH,-CH,-O-CH 2 -CH,-OH

Molecular Weight 150.17

Specific

Gravity at 20° /20°C 1.1254

Boiling

Point at 760 mm Hg. 287.4°C

Decomposition

Temperature" 206.5°C

Solubility in Water Completely miscible

in all proportions to nute that the dew points obtained were sharp and unmis takable; wholly unlike the results of similar determinations on gases treated with diethylene glycol solutions. This is probably due to the fact that triethylene glycol exhibits approximately 15 per cent of the vapor pressure of diethylene glycol at the temperatures employed. The water concentration of the glycol solution was deter mined using an adaptation'· of the method proposed by Fried, Bigg and Jennings.' This method is based on the fact that the viscosity of a glycol-water solution is directly related to the glycol concentration.

The specially constructed apparatus was

calibrated at a constant temperature of 30°C using solutions of known concentrations prepared from a sample of 100 per cent triethylene glycol supplied by the Carbide and Carbon

Chemicals Division

of Union Carbide and Carbon Corp.

All glycol solution used in

this investigation was collected, analyzed and re-hlended to 9S per cent glycol concentration >-Ix en I z I en z o u 1i m .0 ::i

3·00

o w .000 I .1

II--I---

V I V V 50

EQUILIBRIUM CONSTANTS

FOR WATER IN

95% BY WEIGHT

TRIETHYLENE GLYCOL

SOLUTION

V V V V '?'i>')'t/ ,I» V V V I I V V V o '? 7' V ./ # v V ,<;,oo,?'i>'!>? V <;,00, ,?'i>'!>. Y V ,000

60 70 eo 90 100

TEMP 'F

FIG. 2

.005 Ef

EQUILIBRIUM CON STANT S

>-Ix I Z z o u .000 ,-I I b..", I-- I 500

FOR WATER IN

i-f-----

95'0 BY WEIGHT

TRIETHYLENE GLYCOL

SOLUTION

I-- oo·Jr---V V ---nt)' V

I-----

19.j V V 60'
f V

1000 1500

TOTAL PRESSURE -PSIA.

FIG. 3

--1 -b V V ..____V "7 V V 2000
by adding fresh glycoL Fresh sli!ution requirements were small due to the fact that the glycol-gas circulation rate was so great that, even at 500 psig pressure and 100°F, the solution was diluted to but 94.0 per cent by weight glycoL At higher pressures and lower temperatures, the dilution was of the general order of magnitude of one to two-tenths of one per cent so that the equilibrium data obtained were, substantially, for 95 per cent glycol concentrations throughout the investigation.

Solubility

of gas in glycol solution was determined by lib erating the gas evolved from solution differentially and meas uring it by means of a wet test meter. The gas was liberated at a very low rate so that the process was substantially iso thermal. The volume of the residual glycol solution was meas ured in a standard 1,000 ml graduated cylinder. Accuracy of the data obtained in this investigation is con sidered to be quite satisfactory. Dew point measurements were reproducible to 2°F. Determination of the glycol concentra tion of the various solutions was accurate to 0.1 per cent, and check determinations on gas solubility in glycol solutions varied from 3 to 5 per cent. The method employed for the determination of glycol concentration in the -effluent gas proved unsatisfactory, particularly at pressures exceeding 1,000 psig and at the lower temperatures employed, due to the fact that it depended on a small weight differential between two relatively large weights. As this weight differential became progressively smaller the error became larger.

DISCUSSION OF RESULTS

The data o-btained from this investigation are shown in graphical form for sake of brevity. Fig. 2 is a semi-logarithmic plot of the equilibrium constants for water, calculated from the obEerved data, versus temperature using pressure as the parameter. The curve for atmospheric pressure was calculated from published equilibrium data." The best straight line has been drawn through the experimental data points for each higher pressure, parallel to the atmospheric isobar. From this plot the equilibrium constant for water may be obtained at intermediate temperatures. Fig. 3 is a cross-plot of 2, where the equilibrium constants for water are plotted versus system pressure with temperature the parameter. This plot indicates that, over the temperature range investigated, a mini-

Vol. 189, 1950 PETROLEUM TRANSACTIONS, AIME 237 Downloaded from http://onepetro.org/JPT/article-pdf/2/08/235/2238744/spe-950235-g.pdf by guest on 02 October 2023

T.P. 2848 VAPOR-LIQUID EQUILIBRIUM DATA ON THE SYSTEM NATURAL GAS-WATER TRIETHYlENE GLYCOL AT VARIOUS TEMPERATURES AND PRESSURES 100
1>.10 o I Z W I Z o o Q: o CL Q: 0 Wquotesdbs_dbs31.pdfusesText_37

[PDF] végétaux chlorophylliens exemple

[PDF] les végétaux chlorophylliens

[PDF] groupement de texte français

[PDF] groupement de texte exemple

[PDF] processus production dun produit

[PDF] système de production dune entreprise

[PDF] groupement de textes bac francais

[PDF] procédés de fabrication cours

[PDF] gestion de la production industrielle ppt

[PDF] groupement de texte definition

[PDF] étendue d'une série

[PDF] questionnaire cest pas sorcier

[PDF] diagramme de fabrication du pain industriel

[PDF] fabrication pain boulangerie

[PDF] fabrication pain 8 étapes