18 GCPII Glycol Dehydration

8/4/2019 18 GCPII Glycol Dehydration

1/36

18GLYCOL DEHYDRATION

A number of liquids possess the ability to absorb water from gas. Yet, there are very few which meet thecriteria for a suitable commercial process: are highly hygroscopic, do no solidify in a concentrated solution, arenoncorrosive, do not form precipitates with gas constituents, are easily regenerated to a high concentration, can

be separated easily, are essentially nonsoluble in liquid hydrocarbons, and are relatively stable in the presence ofsulfur compounds and carbon dioxide under normal operating conditions.Several of the glycols come the closest to meeting all of these criteria. Diethylene (DEG), triethylene(TEG) and tetraethyiene (TREG) glycols all possess suitable traits. However, almost 100% ot the glycol dehy-drators use TEG.DEG is somewhat cheaper to buy and sometimes is used for this reason. But, by the time it is handledand added to the units there is no real saving. Compared to TEG, DEG has a larger carry-over loss, offers lessdew point depression and regeneration to high concentrations is more difficult. For these reasons, it is difficultto justify a DEG unit, although a few are built each year.TREG is more viscous and more expensive than the other processes. The only real advantage is its lowervapor pressure which reduces absorber carry-over loss. It may be used in those relatively rare cases where gly-col dehydration will be employed on a gas whose temperature exceeds about 50(.This ch ..pter will concentrate on TEG, even though property data are shown in Appendix l8A for theseveral glycols. Some of the system characteristics apply also for all glycols.

THE BASIC GLYCOL DEHYDRATION UNITFig. 18.1 shows the basic glycol unit, regardless of the glycol used. Not shown is a full size separatorahead of the absorber, an essential piece of equipment. Also not shown is any cooling equipment that may be apart of the dehydration processes. When it is possible to cool the entering wet gas with air or suitable waterahead of the absorber, do so. Such cooling is the least expensive form of dehydration.The entering wet (rich) gas, free of liquid water, enters the bottom of the absorber and flows counter-current to the glycol. The dried (lean) gas leaves the top of the absorber. The unit shown has internal coils for

precooling the lean glycol entering. This feature is used primarily on small, portable units.The lean glycol enters onto the top tray and flows downward, absorbing water as it goes. It leaves rich in

water.It is convenient to use the word "rich" to describe the bottom of the absorber and the word "lean" for thetop. At the bottom, both the entering gas and glycol leaving are rich in water; at the top end they both are lean

in water.In Fig. 18.1(A) the rich glycol is filtered before entering the coil in the combination surge tank/heatexchanger. There it IS preheated before entering the stripping column. A portion of the water and any othervolatile components in the glycol are vaporized in this column. The partially regenerated glycol then goes to t~ereboiler where the desired concentration of lean glycol is produced. Said lean glycol exchanges heat and IS

pumped to absorber pressure, to complete the cycle.


2/36

\Jas I..OUOH'OIlIlI& dUO t"OLe~~lIlg - "ie tqlJ:,'menl rvrocutes- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

MAIN llAiWETG

.~~1'''1I~I STRIPPING~1:1COLUMN~.~~~-----...,' ~ \ . ' li : . ' . i 1r~I

ABSORBER

STEAM OR _DIRECT FIRED

WATER RICHGLYCOL FILTER

LEAN GL r c ot,

(A )

"IIII

t : ; , ! l1. /;:'I~;:ISTRIPPING~I~COLUMN~ . . . . . .-----..,.,r-I;'!~t ' A o 1t~,

"III ~ __'REBOILER

S T R I P P I N GG A S SURGE TANK

(B ) (C )

Figure 18.1 Basic Glycol Dehydration Unit with Two Stripping Gas Alternatives.

306

W A T E RV A P O R

I S O - O C T A N E


3/36

G ly co l D e hy d ra ti on

The regeneration unit shown is designed to operate at prevailing atmospheric pressure. The initial ther-mal decomposition temperatures of the glycols areGlycol TemperatureEG 165C, 329FDEG 164C, 328FTEG 206C, 404FTREG 238C, 460F

These are the temperatures at which measurable decomposition begins to occur in the presence of aiDEG is no more stable than EG because it pyrolyzes in contact with carbon steel.In the normal unit containing no air (oxygen), it has been found that one can operate the reboiler verclose to the above temperatures without noticeable decomposition. So, they fix the composition of the leaglycol which leaves at its bubble point. .At the pressures involved, Raoult's Law applies:. .

(18.1)where:

x =P " =P -v -yw =mol fr water in lean glycolsystem pressurewater vapor pressure at reboiler temperaturemol fr water in the reboiler vapor(in equilibrium with xw)

Once one has used the highest allowable temperature (to maximize vapor 'pressure], "P" and/or "y" must blowered in value if lower water concentrations are needed. A vacuum pump or ejector may be used to lowepressure.The addition of some other vapor to the reboiler will decrease the "yll for water. As shown in Fig18.1(8), this may be accomplished with stripping gas. Any inert gas is suitable. A part of the gas being dehy-drated, or exhaust from a gas-powered glycol pump (if used), is suitable. The quantity required is small. Itheory, adding gas to a packed unit between the reboiler and surge tank is superior. In commercial units

makes little difference how you get the gas into the reboiler so long as the quantity is right. It is common to usa distributor pipe along the bottom of the reboiler.A second stripping gas alternative is to add a material like iso-octane to the unit, as shown in Fig18.1 (C ) . It vaporizes at reboiler temperature but can be separated from water vapor in a partial condenser.Extra equipment is required but little or no hydrocarbons are discharged to the atmosphere. Venting the smalamount of stripping gas in water vapor offers no real environmental hazard but can become an issue.The unit shown in Fig. 18.1 is not an optimum one from the viewpoint of thermal efficiency, operatingconvenience and safety. It should be used only when simplicity and compactness override these considerationsin small, portable units. Fig. 18.2 shows examples of two systems using TEG that incorporate additionalfeatures.The upper flow sheet is for an offshore unit. The inlet scrubber is in the bottom of the absorber. Three-phase separation is required. The gas rises through a "chimney tray" to the absorber. The hydrocarbon and

water are separated as shown. Three-phase separation saves on deck space and is less expensive, but many othe existing units are unsatisfactory because they provide inadequate separation.The rich TEG from the chimney tray goes to a degassing pot (flash tank) which is operated at a highenough pressure to send the gas to fuel. In some systems the pressure is sufficient merely to enter the mainflare system. The purpose of this is several fold: (1) use or dispose safely any volatile components picked up bthe TEG in the absorber, and (2) keep corrosive sulfur compounds and carbon dioxide out of the high tempera-

ture reboiler.The true solub ility of paraffin hydrocarbons is very low in the g lycols. But, separator carryover andentrainment does introduce hydrocarbons into the rich glycol. Many of these are "heavier" than air and can besafety problem unless disposed of properly. In addition, aromatic components are very soluble in TEG. Thesecan also be a safety concern when discharged to atmosphere at the top of the packed section.

307 ____- _-----_---~


4/36

Gas C onditioning and Processing - T he Equipment Modu les

Mai n d oc k _

G a s

l e a n glycolSleam 10v en r s l a ck

Y i ~ : : : : = - a 3 ~ e - J _ . - G _ Q _ S ( ~ : ~ I : _ ~ ~ : ~ : : ~ : _ 0 1 _ i n _ e ~.J ~~~;~)V ~ ----~~==~-------aler 10 see16 ,Iycal pump. 3 glycol, I . , . , . " ., . . . . . . . , . S u m p t a n k r e g e n e r a t o r s8 s ep a ra to r. c on la n on

Sphm launch.rP i p e l i n eto shoref lo e . .. . ullill

8 w el l s( S CI .U h c 't l M , II ,

DRIED GAS

ENTRAINMENTSEPARATOR

r

I1.._ .. -C"

WET GAS

BLANKETGAS

WATr:R

cro~


5/36

G ly co l D eh yd ra ti on

Both sulfur compounds and carbon dioxide are very soluble in water and react to some degree with thglycols. The degassing in the flash tank prior to the stripping column reduces their concentration and minimizeshigh temperature corrosion. This degassing is more efficient if the rich glycol is preheated first, as shown in thlower flow sheet in Fig. 18.2.External gas- TEG exchange is shown in the upper flow sheet, as opposed to external cooling in the loweone. Offshore it is suitable to use sea water cooling in plate exchangers, provided treated sea water is availablefor other purposes. In temperate latitudes aerial cooling also is a viable alternative.An external glycol-glycol exchanger of the type shown in the lower flow sheet would be preferred if fuel iexpensive. The closest economic temperature approach possible reduces reboiler heat load. Plate exchangersare widely used in this service.Both flow sheets show a fired reboiler. The use of hot oil, steam, waste heat or electrical resistance coilsare all suitable if they are readily available at the site. Oftentimes the use of electrical resistance offshore is verycost-effective and safe.No filter is shown on the upper flow sheet. It is on the low pressure side of the flash tank in the loweone. Location, pressure-wise, obviously affects cost. I prefer locating the filter at some point ahead of thereboiler to r1Iinimize the "gunk" accumulating therein. For effective operation it is imperative that full-flow,glycol filters be installed in the system .

BASIC PROCESS DESIGN FACTORS

All factors controlling the behavior of absorption systems also apply for TEG dehydration. In fact, fromprocess viewpoint, TEG is one of the simpler absorption processes being employed in the petroleum industry.In order to properly design a unit one needs to know maximum gas flow rate, maximum temperature andpressure, gas composition and required water dew point or content of the outlet gas. From these one cancalculate:

1. The minimum concentration of TEG in the lean solution entering the top of the absorber requiredto meet outlet gas water specification.2. The lean TEG circulation rate required to pick up from the gas needed amount of water necessary

to meet the outlet gas water content specification,3. The amount of absorber contact required to produce the necessary approach to equilibriumrequired in (1) above at the chosen circulation rate.

To obtain these answers it is necessary to have a vapor-liquid equilibrium correlation for a TEG-watersystem. From this basic input, one can size equipment and develop mechanical specifications.The procedure that follows is very straight-forward and can be performed manually. In all but a fewexceptional applications, it will give results as reliable as more complex (appearing) methods. Following an out-line of the basic calculation procedure, each major equipment component will be reviewed.

MINIMUM LEAN TEG CONCENTRATION

If water-saturated gas is placed in a static cell with a given concentration of TEG-water solution atfixed P and T, equilibrium would be attained in time. Assuming the liquid had a sufficiently lower waterconcentration, water would transfer to this liquid from the gas. At equilibrium, the mol fraction water in the gasdivided by its mol fraction in the liquid equals the K value for this system.Fig. 18.3 is a convenient, reliable correlation based on K value data to determine the minimum lean TEGconcentration required. {18.1} The water dew point is the result expected if equilibrium was established i':l ~ testcell w ith the glycol concentration shown, at a g iven temperature. Data show that pressure effects are triv ial up

to about 17 MPa [2500 psia].

309


6/36

Gas Conditioning and Processing - The Equipment Modules

45

40

35

30

25".~ 20

15

10to)0- 5z0A. 0~I . & J0 -5a ::I . & J~ -100


7/36

Glycol Dehydration

-80 ~V-90~ 40 eo 60 70 80 90 100 110 120

I N L ET G A S T E N PE R A T U R E , of130 140 leo 160 170

F igure 18 .3 a Equ ilib rium (M in im um ) W ater D ew Poin t O b tainab le w ith a G ivenL ean T EG C oncen tration for D ifferent E ffectiv e C ontactorTempera tures .

311--- ------------- -----------


8/36

Cas Conditioning and Processing - The Equipment Modules

A TEG absorber is essentially isothermal. The heat of solution is about 21 kJ /kg of water absorbed [91Btu/Ibm] in addition to the latent heat. But, the mass of water absorbed plus the mass of TEG circulated istrivial to the mass of gas. So, the inlet gas temperature controls. The temperature rise due to heat of absorp-tion seldom exceeds 2( except when dehydrating at pressures below about 1.0 MPa. In low pressure servicesome temperature adjustment may be desirable.The diagonal lines represent weight % TEG in a TEG-water mixture entering the top of the absorber.What is the lowest water dew point one could attain with a given concentration at a given temperature?

Example: What equilibrium water dew point could be obtained at 40C with a lean glycol solutioncontaining 99 wt % TEG?In Fig. lS.3 locate 40C on the abscissa, go vertically to the 99 wt % line and thenhorizontally to the ordinate. Read -15C.

This water dew point could be attained in a test cell but not in a real absorber. The gas and TEG are notin contact for a long enough time to reach equilibrium. Numerous tests show that a well designed, properlyoperated unit will have an actual water dew point S.S-S.SoC [10-lS0F] higher than the equilibrium dew point.This "approach" to eouilibriurn can be used to specify minimum lean glycol concentration. The procedure is asfollows.1. Establish the desired outlet water dew point needed from sales contract specifications or fromminimum system temperature.2. Subtract the approach from (1) to find the corresponding equilibrium water dew point.3. Enter the value in (2) on the ordinate of Fig. lS.3 and draw a horizontal line.4. Draw a vertical line from the inlet gas temperature on the abscissa.5. The intersection of the lines in Steps (3) and (4) establishes minimum lean TEG concentrationrequired to obtain the water dew point in Step (1).

If water content is specified or calculated in mass per unit gas volume, a water content, pressure, dewpoint temperature correlation is required. Fig. 6.1 from CHAPTER 6 is reproduced in Appendix lSA at the endof this chapter for use in calculations.Example: The gas sales contract specifies an outlet water content of 100 kg/lOG std m3 at a pressureof 6.9 MPa. The inlet gas temperature is 40C. What minimum lean TEG concentration isrequired?

Metric: For 100 kg/lOG std m3 and 6.9 MPa, the equivalent dew point from acorrelation is -2C. If we use an SoC approach the equilibrium dew point is-10C. From Fig. lS.3 at -loC and 40C contact temperature, wt % TEG= 9S.S.English: At 104F [400C) and 1000 psia [6.9 MPa1 the dew point is 2SoF for a watercontent of 6 Ib MMscf. An approach o f 14.4F [SOC1 give~ an equilibriumtemperature of about 14F. From Fig. lS.3a, lean TEG concentration equals9S.S wt % .

The dashed line in Fig. lS.3 at about 9S.S wt % represents the concentration of lean TEG that can beproduced routinely in a regenerator operating at standard atmospheric pressure and 204C [400F]. This is asafe value for design and specification purposes. Concentrations of 98.7 -9S.S wt % are common; some to 99.1wt % have been reported but represent a special case where incoming hydrocarbons provided natural strippingand/or the pressure was lower than standard atmospheric.

Since the capital cost of ordinary gas stripping accessories is trivial, they always should be included.Conditions can change to where they may be required.It is necessary to fix a lean TEG concentration for subsequent calculations. For the first consideration,use the results from Fig. 18.3. If the concentration obtained is less than 98.5 wt % , use 98.5 wt % for thecalculation unless you plan to reduce the reboiler temperature below 204C.The minimum lean TEG concentration may not be the one used. A higher concentration than this maybe specified to minimize circulation rate and optimize cost.

312-_.._ ... ---------_ ..._-__------


9/36

G ly co l D eh yd ra ti on

TEG REGENERATION

A given lean TEG concentration is produced in the reboiler and still column (regenerator) section by cotrol of reboiler temperature, pressure and the possible use of a stripping gas. So long as no stripping gasused, the concentration of the lean TEG leaving the reboiler is independent of the rich TEG entering.When stripping gas is used, the concentration of rich TEG leaving the absorber is found by a wa

material balance around that absorber. By definitionwt % rich TEG = wt lean TEG (100) (18wt lean TEG + wt water absorbed + wt water in lean TEG

The weight quantities in this equation may be found per unit of time or per unit of gas flow. In any case,values used depend on circulation rate. This rate depends on dew point requirements, lean TEG concentrationamount of absorber contact and economics. The latter dictates a rather low circulation rate. This rate usuawill be 25-60 li,ters of lean TEG solution per kg of water absorbed from the gas [3-7 U.S. galjlb water].minimum'rate is governed by the rate required for effective gas-liquid contact in the absorber; the maximumlimited by economics.Because this regeneration takes place at low pressure, under ideal gas conditions, the calculation is v

routine. Fig. 18.4 has been calculated to predict regenerator performance. {18.2}The minimum wt % lean TEG on the top abscissa is found from Fig. 18.3. The wt % of rich TEG onbottom abscissa is found from Eq. 18.2. If one neglects the small amount of water in the lean TEG, theTEG concentration can be approximated from Eq. 18.3.

Rich TEG = (p)(Lean TEG)p + ( l jm) (18

Metric9.3 IbjU.S. galU.S. galjlb

Englishwhere: p = liquid density _ _,..m = lean TEG ratfO "Jlhl vn>.je,-Rich TEG = wt % TEG in rich TEG solutionLean TEG = wt % TEG in lean TEG solution

1.12 kgjLLjkg

The diagonal lines in the lower left portion of Fig. 18.4 represent various amounts of stripping gas. Iseldom that one would use more than 0.08 m3 gas per liter of TEG [10 scfjU.S. gal]. At some point the practcal impact of stripping gas diminishes with rate.Three temperature lines are shown. Where high concentrations are desired, the specification of 204Cnormal unless the gas being dehydrated contains oxygen. This is close to the thermal decomposition temperature (in air). In the usual case where the natural gas is oxygen free, the use of 204C has proven very satisfac

tory.The diagonal lines at upper right in Fig. 18.4 represent the effect of regeneration pressure in mm Hg,

760 mm Hg = 14.7 psia = 101.325 kPa100 kPa = 1 bar = 750 mm Hg

Unless a vacuum is being used, it is customary to use the 760 mm Hg line for design calculations.Notice that at 760 mm Hg pressure and a reboiler temperature of 204C Fig. 18.4 shows a lean Tconcentration of 98.7 wt % . If in using Fig. 18.3 you obtain a concentration less than this, use 98.7 wt % as

desired concentration when utilizing Fig. 18.4.

313


10/36

Vl0-0-

Z0. . . . .,_:::J...:l 00til 0-...:l 0-0U>-...:lt,;)~:u...:lZ Vl

co0'>L:.:i-( > . 0i-:::0CO0-

T I L- , ./ :/, ~ I i- 100 1--,,- ./Vw ~ IV /;W ~I VV .: ~ \~ c . . - :- :>V /[7 ;W 1\\f \ " > l >-~~ ./ Y ' C .L '_ j I/h \1 \ 1 ' - % . . ,. . . < . - VI r V f 7 ' \ '< '/ 1 \ 1 \ Y '~'V VI V I \ \ ~,.o I0 ;)--- -yV V-j;J- --- -- -t-- - r\-r\ f\.'~I- . !/V 7 I \ "\~I l\_ ~ I~/I/ A /v '\ 1 \ 1 ~ + -/I~~7 V V/1 o vI q p 0.5~ / / I . I I ~~O> O~-~ v~~i I I I I I , r\'j,; '\;~~V/ I I I I ~~~~ / l / I I I I I ' " r \ o ; . . \I M ~ ....\0" ..jMN'.....'\~t,_ , - > : ' 00000 ~I~ 0... . .e .... 00000 0/1 I' r\A - ~ - I l I r I ,I-< ~ I ,I I \ ~I \ I \- 5 ; : ;

U) - I ~ 1~0I-< U) I, .0 , I \ ~ ,I--I-< , '\ I \I 0.... t>O I II-f-....=0 I ~ \ \ I I~ II0I- ~~~ It \ 1 \ T, ,~ 1 \ \ ~~ \ \0 0 I "2s: '" r:: .., s, I - ~ I " \~ \ 1E o - r:: Q) r:: Q)It) 0 0":: "- o VI 0 \IIs . . . . So-,_ . . o _ . _ , _ , . . _ o _ ::: ~ 1 \ . I 1 \ i \>"- 0 .., .., (ij tU It) E ._ E o6-lE"O>,::IO VI ::I r:: I .~ \1 \ \ \ \r::~O-l- UO::lQ)O I I::::I0>Q)4- 0 0. u..::~0 ~E~"_''_ I g : 1 1 I Il4-.r::VI"O 00> .~.~ u,_ QJ NU E VI . . . .. r : : : Vl_..c: > 000 I-< I I. - s . . . . . . . . . , 0 IJ)'_C:S-o.. . . . IE .~ 0 4- E tilQ)~4-r::Q) u"- 0 \ I \" 0 o~ l- Q) Q) >,- >, u::I U . . . . , l- ~ ~>,r::Vll-1t) ::I 0> ~ ~ I I '\I -~- ~'-h-~OEQ)l- . . . . , Q) It) 0 --. .. , t :7 '0 '_ So - . . . . . . , It).r:: u u u""'Q)It)r:: s, U r::.~ e-, \ I I \ \ \r::1t)""'~0 Q)._ to .., ....-o It) s; .~ o.l-El-o>. . c : : Q . J " " " Q J _ " ' " E l- Q) l\ \ \ \r::.r:: It) It) Q) ~ 0 > r:: I'" QJ"" _....., ...- ....,"0 "- It)~ 4- U 0::1 Q) l-"O Q) 1 \ \ \ ,0 r:: ""'u s....V'lQ)QJ~ I.:: 0 s, Q)::I0.Q)VI r:: u N >,.~ ~ U Q) , i ' I T \ I \'.D U .,_ V1 c: o.cQ).~ Q)00 o-os-.._,c : :. .. .. ., .r :; . .. .. .. " " ' O r - .D 0 I 1 \ 1~ ~ \_ rc .,_ Q) 0 Q)"''''' 0.... l- r:: u s, It) U ~ . . . ,. . . . , o-.~ >, O>Q)"O l,~ \ , \ , I"0 r:: .LLJE~ ~ "-Q)Q) I) Q)M l-C'> It) 0>"- >, r::.r:: U '>,Q) E r:: Q) o~ I I IVlr::OO.D...,"O l-'~ ~~roo ..... Q) r:: 00.00. ..j N . . . . . I-"OU "0"0 It) r::o.r::EQ) .....0 co \0 Q) .~ Q) s; ." . . . . .Q)Q)O>XQ)VI Q) s; VI ::I . . . . . bO. . ::: _c; o _ o _ s . . . . ee .r::""'It)Q)"'" (Jf-f-w..4-1t)0> ..... VI..::.D1t) III tI):::J314

o0-C>

or::COL?O-Z> -...JVlL?~o

c: :z

o

.v'


11/36

G ly co l D e hy d ra ti on

The general procedure for using Fig. 18.4 is as follows:

1. Atmospheric Pressure, No Stripping Gas - B

Wt % rich glycol is not a variable. Proceedvertically from 0 stripping gas and temperatureline intersection. You will read 98.7 wt % TEGat 204(; 98.4 wt % at 193(.

III T~

2. Atmospheric Pressure, Stripping Gas -

a. Proceed vertically from B to temperatureline and then horizontally.B

b. Proceed vertically from A.III

~T

c. Intersection of two lines from A and Bfixes amount of stripping gas.

3. Vacuum, No Stripping Gas -Ba. Proceed vertically from intersection of 0gas line and temperature line to atmo-spheric line (760 mm Hg).

b. Proceed horizontally from point in (a) topressure line necessary to fix value ofpoint B.

atm.

o

In that rare case where both stripping gas and vacuum are used, procedures (2) and (3) are combined.

Example: An example is shown on Fig. 18.4 for use of stripping gas and vacuum. A 96.84 wt % ricglycol enters a regenerator using 0.03 m3 of stripping gas per liter of glycol solutionscfjU.S. gal]. Proceeding to 204( and then vertically, one reads 99.16 wt % if atmo-spheric pressure is used. If a vacuum is employed and the absolute pressure is 500 mmHg, the lean glycol concentration is 99.41 wt%.

As a general rule, vacuum is avoided unless necessary to simplify unit operation. Vacuum pumps can ba nuisance. An ejector can be used to produce necessary vacuum in the right circumstances.The amount of gas shown in cubic meters is the actual volume at system pressure. In general, most wuse a slightly higher amount than shown. This is the theoretical amount of gas and its effect depends somewhaton distribution. The amount of gas should be measured. An approximate measure, such as that provided byrotameter, is satisfactory.

CIRCULATION RATE - ABSORBER CONTACTS

The above calculation is dependent on glycol circulation rate if stripping gas or vacuum are usedAlthough a number of circulation rates are possible, the minimum feasible one should be used. As circulationrate increases, so does operating cost. The minimum feasible rate is fixed by absorber characteristics and cost.

11"


12/36


1 The figure at left shows theeffect of TEG concentration and cir-culation rate on dew point depressionfor a fixed amount of absorber con-tact of absorber contact. Notice thatthe curves become relatively flat athigh circulation rates.For ordinary pipeline waterdew point control, the glycol circula-tion rate usually will be 25-60 litersper kg water absorbed from the gas[3-7 U.S. gal/lb water]. A circula-tion rate above this range can bejustified economically only in specialapplications for it results in excessutility consumption.

For a given circulation rate a given number of absorber contacts are needed. The relationship betweenrate and amount of contact is very adequately described by the Kremser-Brown method described in CHAPTER13. In the glycol application, mol fractions may be used in instead of regular absorption parameters because ofthe concentrations involved. The basic absorption equation thus may be rewritten as:

TEG Circulation Rate ----~

(18.4)where:

Y N + l = mol fr water in entering wet gasY l = actual mol fr water in dried gas leavingY o = water content of dried gas if it is in equilibrium with theentering lean glycol (value is less than Y l)A = absorption factor, A = L/~KV)L = glycol circulation rate, mo es /unit timeV = gas flow rate, moles/unit timeK = equilibrium constant for water between water in gas and waterin a TEG-water solution, y = KxN = no. of theoretical plates in the absorber

s; = 1.33(E-06) Wv; = 2.11 (E-OS) W

Thus,

=

The mol fraction water, yw' is related to W, the mass of water per -standard volume of gas by a fixedconversion factor. As noted in Chapter 6,

=

The subscripts on "W " have the same significance as on "y."

where: W = kg/106 std m3where: W = Ibm/~Mscf

(18.5)

For a given calculation the values of inlet water content, outlet water content, gas flow rate, and absorberpressure and temperature are fixed. Using a correlation for determining the equilibrium K value of water in aTEG-water system, values for y (or W ) and K are available. The only variables left in Eq. 18.4 are L (theTEG circulation rate) and N (the ~umber ~f theoretical trays). In theory, there are an infinite number of combi-nations of Land N that satisfy Eq. 18.4. In practice, the choices are limited by economics and absorber perfor-mance.The cost of purchasing and operating any absorption unit is a function of circulation rate. It is thus goodpractice to operate at, or near, the minimum rate necessary to meet absorption specifications.

316


13/36

G ly co l D e hy dr at io n

Fig. 18.5 is a plot of Eq. 18.4 that is convenient for manual calculations. This uses what could be calledan overall absorption factor. The ratio L/V varies slightly throughout the absorber. Lo is the rate of lean TEGentering the top tray and Y N+l is the gas rate entering the bottom tray. As a later numerical example will illus-trate, the variation of L/V will have a calculable, but usually nonsignificant, effect on unit design.1. 00

.995.99

. . . . . 309853. . . . . . . . . . .9 8+z: z:::;: ::;:

.9 6. . . . 0 .9 4-, c-,I. . . . . . . . . . .+ + .9 2: >,Z:>,

.9 0

.8 8

.8 6

.8 4

i H i l t W I ! H i l H . l l i'n t - t r . - ~ 2.0

':",,..." : ; r u t H ::11U ~ C I!m l ~:::: l!!~ "j i;H : : : : . : . :

. 8 I; I I ! ! ii i In ~ il l j i! ~m~ I H ii l l W ... , W ; ! ~ u rz: it Ii i ! 1 ~ \ a n i i m t i l;1 I f I : : I:it l il ~! H i i ! ; : , i ifif:

I ! i f l t iU ls~I f ::t: ;~:~t r,::: ::::jl , u !nl!:::t i l = - " ' : ,=1 \:;; it .i i iJ 1i . l i t ! tE in!ffi 1 h ;:. ]i t l f : : i 1 3 g

a 4 8 1 2 1 6 20 24 28A b s o r p t i o n F a c t o r , A = L o / ( K V N +1 )

Figure 18.5 Plot of Absorption Factor Equation.

The left-hand ordinate of Fig. 18.5 could be called absorption efficiency. It is the actual amount of waterremoved, divided by the maximum amount removable. The values of N encompass the range of theoretical traysusually employed. The line for N equals infinity and represents the minimum absorption factor, e.g., minimumcirculation rate. All curves for finite values of N become asymptotic to this line. Note that the scale changes onthe ordinate.Calculation of Lean TEG Rate for a Given Absorption Efficiency and N -

1. Calculate Yo (or W 0)2. Determine absorption efficiency3. Use Eq. 18.4 or Fig, 18.5 to find absorption factor A for a given value of N4. Knowing VN+l and K, solve A for Lo' the lean TEG circulation rate

Calculation of N for a Given Lean TEG Rate and Absorption Efficiency -1. Calculate Yo (or W 0)2. Determine absorption efficiency3. Calculate absorption factor A4. Determine N from Eq. 18.4 or Fig. 18.5

317--- ---_---------


14/36


It is usual to repeat the calculation to obtain three lean oil rate/absorber contact values that satisfy therequired absorption efficiency. The final choice is economic. This usually involves selection of standard modulesto make up the unit.This calculation should be made at the lowest pressure and highest temperature anticipated for theentering wet gas, to obtain the maximum water loading. Historically, the tendency has been to choose a designtemperature lower than that actually obtained.The overall tray efficiency in a well-designed TEG unit will vary from 25-40%. It is recommended that25% be used for most applications. This provides an affordable safety factor to help compensate for theinherent errors in the design specifications.

Equm~rium RelationshipsVarious studies have been made of the equilibrium behavior of water in the TEG-water system.(18.3-7)All provide rather consistent data. The use of an activity coefficient b) is a convenient and reliable method forcalculating K. Using this relationship .

K = ( Yw ) ( r r ) = (B)(W)(-y) (18.6)where: K =Y w =

-y =W =B ==

equilibrium constant for water in a TEG-water systemmol fr water in the gas at saturation over 100% liquidwater (from regular water content correlation)activity coefficient for water in the TEG-water system asfound from Fig. 18.6water content on a mass per volume basis, at saturation, asfound from a regular water content correlation1.33(E-06) when W = kg/l06 std m3 (measured at 15C and 100 kPa2.11(E-05) when W = Ibm/MMscf (measured at 60F and 14.7 psia)

Notice that " Y and thus K vary with TEG concentration, which in turn varies throughout the absorber. Anaverage K at average concentration cannot be found until the circulation rate is fixed. So, a simple trial-and-error calculation is involved. One can assume the inlet lean TEG concentration as a first try. Most often theaverage TEG concentration will turn out to about 98.3 wt%. For routine dehydration ilp:,lir.


15/36


Example:Calculate the circulation rate of 98.7 wt % lean TEG needed to dry 106 std m3/ d of gas7.0 MPa and 40C in a four tray absorber (1 theor. tray) to achieve an exit gas water content of 117 kg/l06 std m3. The inlet water content is 1100 kg/l06 std m3 (saturated gas)1. From Eq. 18.8, Xo = 0.0992. From Fig. 18.6, 'Y = 0.4753. W is the water content of saturated gas at 7.0 MPa and 40C or 1100 kg/l06 st

ml in this case.4. From Eq. 18.7, Wo = (1100)(0.475)(0.099) = 51.73 kg/l06 std m35. The left-hand side of Fig. 18.5 is: (1100-117)/(1100-51.73) = 983/1048 = 0.9386. From Fig. 18.5, for N = 1, A = 15.4

(In order to solve for L one must recognize that it is different at each point in the absorption tower. The conservative approach is to assume the gas volume is constant and solv

for Lo = from A, using VN+l = 1.0 mol. This will yield a circulation rate slightly highthan the more rigorous calculation.)7. Lo: ;= (A)(K)(VN+1)From Eq. 18.6, K = (1.33 x 10-6)(1100)(0.475) = 0.000 695

So, Lo = (15.4)(0.000695)(1.0) = 0.0107 kmol TEG/kmol gas8. Mol gas/h = 1739 (106 std ml/d)

So, kmol TEG/h = (0.0107)(1739)(1.0) = 18.619. MW lean glycol = (0.099)(18) + (0.901)(150) = 13710. kg TEG/h = (18.61)(137) = 255011. Density of TEG is 1.12 g/cml = 1.12 kg/liter

Circulation rate is 2550/1.12 = 2277 liter /hIn one hour (1100 - 117)/24 or 41.0 kg H20 is absorbed.Circulation rate is 2277/41 = 55.5 liter/kg H20 absorbed.

In traditional English units the calculation follows the same format.1. 2., The same3. W = 67 Ib/MMscf4. W0 = (0.475) (67) (0.099) = 3.15 Ib/MMscf5. (67 - 7)/(67 - 3.15) = 0.9406. A = 15.97. K = (2.11 x 10-5)(67)(0.475) = 0.00067

Lo = AKVN+1 = (15.9)(0.00067)(1.0) = 0.0107 mol TEG/mol gal8. The flow rate is 34.92 MMscf (1 x 106 std m3). There are 2636 Ib mol/MMscf.So, Lo in mol/hr = (0.0107)(34.92)(2636)/24 = 411b mol/h

9. MW = 13710. Ib/TEG/hr = (41)(137) = 561711. Density of TEG is about 9.3 Ib/U.S. gal

Circulation rate is 5617/9.3 = 604 U.S. gal/hrIn one hour a total of 92 Ib of water is absorbed.Circulation rate is 604/92 = 6.56 U.S. gal/lb water absorbed

319


16/36


.45

.70

.6 5~. . . . .c:(lJ.~ .6 0u. . . . .. . . . .(lJ0ue-,. . . . . 55>. . . . .u


17/36


Remember that the minimum TEG concentration found from Fig. 18.3 assumes that the tower providethe contact required, e.g., the proper number and efficiency of trays. The result desired is achievable with thconcentration only if both the mechanical design of the absorber and the circulation rate are proper. On tother hand, if you don't have at least that concentration, the result is unattainable, regardless of the amountcontact or circulation rate.Random packing has been used successfully for gas-liquid contact in a glycol absorber but is not generally recommended. One can encounter liquid distribution problems. If undue foaming occurs, the tower c

flood at lower than normal gas rates and cause excess glycol losses. Solids in the gas tend to plug the packing.Structured packings are becoming increasingly popular in glycol dehydration service. Some of the mocommon types include Sulzer Mellapak, Montzpak and Koch Flexipak. Structured packings offer the advantagof considerably smaller diameter towers as compared to trayed contactors. This feature can significantly reducapital costs, especially offshore. In addition, glycol losses may be reduced due to less agitation of the glycsolution at the top of the absorber. The primary disadvantage relates to poor liquid distribution design and tinability to wet the packing.

fFor sizing packed glycol contactors using structural packing, the following parameters are useful.

Metric EnglishKs 0.1 mls 0.33 ftlsHETP 1.5-2.0 m 60-80 in.

Structured packing consists ofprefabricated elements approximately0.2 m [8 in.} thick. Each section is ori-ented 900 relative to the adjacent unit toinsure good liquid distribution. Sealing isprovided at the vessel wall with wallwipers which come with the packing.Two sections of structured packing areshown to the right.One of the primary applications of

structured packing has been to increasethe gas handling capacity of an existingtrayed contactor. Retrofits increasingabsorber capacity by over 30% have beenreported. When retrofitting a contactoradequate pump and regenerator capacitymust also be checked.C ou rtesy N utter Eng ineerin

Some vendors offer an absorber other than a conventional vertical unit. One cannot judge their worthexamining a drawing. However, some significant measure of countercurrent flow is required. The gas leavinmust last contact the lean glycol entering. There must be enough contact ahead of this point to remove the buof the water in the gas. Be cautions with "gimmick" units. The proven standard absorber seldom offers weighor space problems.

REGENERATION AND HEAT EXCHANGE

The glycol circulated must be heated to reboiler temperature and then cooled before re-entering the absorber. Efficient rich glycol-lean glycol heat exchange will minimize fuel andlor utilities cost.The detailed calculation involves a heat and material balance around the regeneration system for thmaximum anticipated glycol circulation rate. The reboiler heat load depends on the efficiency of exchange.

321


18/36

e ReboilerCommonly a direct-fired reboiler is used. The outside temperature of the fire tube surface, covered bycol, is surprisingly cool, if the unit is properly fired. Fig. 18.7 shows one temperature profile obtained on test.

: . : :u:! 2 0 6 0 e 20 8 0 e nooev' /

48 20 e 59 3C 760C )LYCOL BULK TEHP. 204C~42~e 1 0 9 30e

/ /22 10e 2 18C 2 1 6 0 e 21 30eFigure 18.7 Typical Temperature Profile in Gas Fired Glycol Reboiler.

Outside surface. temper~ture may be lo~e.r than steam or hot oil units because .of ,high ~Im resistance onlow pressure gas side. This temperature IS important LV ~,,:;,;ar.t c r; :; ;: .i ,; ; ,b : ; ':::~'~)l'0n;;, _ r _: : . .;gradation of the glycol.To maintain proper skin temperature, prevent "hot spots" and obtain satisfactory fire tube life, it mustve sufficient area. The following are recommended:

Max. heat flux across fire tube wall -Recommended heat flux for max. life -Burner capacity -

25 kW /m 220 k W /m 230 kW /m2

18000 Btu I h r ft2]16000 Btu/hr ftL]110 000 Btu/hr ft2 ]

tice that the burner possesses extra capacity to obtain firing flexibility.The typical heat balance will indicate a reboiler heat load of 390-450 kJ /Iiter of TEG circulated 11400-00 Btu/U.S. gal]. Experience has shown that extra heat capacity is desirable. Based on the maximum pumppacity, I recommend that the minimum reboiler rating be

560 kJ/liter TEG 12000 Btu/U.S. gal TEG] The direct fired reboiler is very efficient but there can be safety problems, particularly offshore. Hot liq-, steam, electrical resistance coils and waste heat can be used. The choice depends on what else is on thetform. The heat load of the glycol unit is rather small so it seldom justifies a separate source.

e Still ColumnThis is a packed column except in large sizes. The upper part of this is really a rectification section tovent glycol loss, No formal amount of reflux is necessary with TEG. Just enough is needed to keep thecking wet. Since the vapor load is low, the sizing of this column is not critical. It should, however, conformerally to the diameter approximated in Eq. 18.10.

d = (A)(m)O'S (18.10)Metric Englishwhere: d = diameter of packed tower cm in.m = glycol circulation rate L{min U.S. gal/minA = empirical constant 1.9 9.1

322


19/36

Glyc o l D e hyd ra ti on

Glycol-Glycol Heat ExchangerThis is the basic heat exchanger. Its efficiency has a direct effect on reboiler heat load. The rich glycolfrom the absorber enters at about the temperature of the inlet gas (Pt. 1). The lean glycol from the regenerator(Pt. 3) enters usually at about 204C. The temperature of the lean glycol at Pt 4 should not be greater than60-65C and should be slightly warmer than the gas leaving the tower.

4In most cases a 15-20oC approach in the heat ex-

changer is desirable. If it is too hot, glycol loss may be en-hanced.In this exchanger

Rich 1 (18.11)Glycol

The easiest way to make this enthalpy calculation is tolook up the average specific heat of the glycol in Appendix 18Aand multiply it by AT across the heat exchanger to find Ah forthe lean mixture. The temperature of the rich glycol enteringthe heat exchanger will be about the same as that of the en-tering wet gas.If an efficient glycol-glycol exchanger is used there may be no need for an additional cooler on the lean glycolstream. The cooling coil in the top of the absorber is not recommended for any application other than smallwellhead units. The most common types of glycol-glycol exchangers are pipe-in-pipe and plate. The first areused on small units, while the latter find extensive USe in larger units and offshore. Plate exchangers provide ef-fective heat exchange but are very susceptible to fouling in dirty service. Clean filtered glycol is imperative iplate exchangers are used.

FILTERSGood filtration is critical. The full-flow type is preferred. I recommend two filters in parallel, with noby-pass lines, so that full filtration is assured.A cloth fabric element that is capable of reducing solids to about 100 ppm by weight is preferred. Paperand fiberglass elements generally have proven unsatisfactory. Filter size in a properly operated glycol systemshould be 5-10 urn. Larger sizes 25-50 urn may be required during start-up and in dirty service.It may be impossible to judge the effectiveness of filtration by color alone. Even well filtered glycolprobably will be black. But, removal of most of the solids will reduce corrosion, plugging and deposits in thereboiler, and may reduce foaming losses. Good filtration is critical for satisfactory performance. It is desirable tomeasure the pressure differential across the filter and change the element when it reaches about 170 kPa [25psi].The use of a carbon purifier downstream from the filter often is recommended. This will produce essen-tially water-white glycol. Maintenance of this color has proven desirable because it tends to increase dehydra-tion efficiency and minimum foaming, a major source of glycol loss.Coal-based activated carbon should be used because wood-based charcoal tends to break up in use.This carbon can be placed in a metal canister or fill a vessel. In either case, good screens are needed to preventcarbon loss into the system. Said carbon particles, much like iron sulfide, tend to promote a stable foam.Glycol filters are only effective when used. In especially dirty glycol systems, filters are often bypassed to

avoid frequent filter change-out. The problems with this should be obvious. If filter plugging is excessive, trylarger filter size and look for source of problem (e.g., poor inlet separation, degradation, corrosion products, etc.).

PUMPSA number of types of positive displacement pumps have been used in glycol service. In small units, vari-ations of chemical feed pumps have been used. A Kimray pump of the type shown in Fig. 18.8 is used com-monly when a gas-powered pump is desired. They have a unique power recovery feat.ure. that minimizes theamount of power gas required. In many cases, the exhaust power gas can be used as stnpprng gas or burned as

fuel.

323


20/36

Gas Conditioning and Precessing - The Equipment Modules

Spud [enI",.' lV/WI

1 > ' ' ' ; ; ; ; ' 2 Pvmp P,S/"" Au""oly1:::::1Yd 6Iyc/J/ from Ahcreer r"'f"tp..,. ....-TJOJ 11,/ 6f.yt.1 n . .R,bc;/~.~(i.-P


21/36


Electric powered plunger type, triplex pumps also are used commonly. These provide a steady flow rto the absorber. Hardened or chrome plated plungers usually are recommended. Scoring and pitting has beenproblem. Piston speeds should be kept below 0.6 m] [120 ft/min].

INLET SEPARATION

You cannot afford the dehydrator if you cannot afford to place an effective separator on the gas inlSalt water will enter the reboiler, evaporate, and coat the walls (causing fire-tube failure) or fill up the unThis happens frequently on wells that supposedly do not produce salt water "officially."A good separator also should remove the bulk of the compressor oil, drilling mud, corrosion inhibitorpipeline dirt, formation solids, and the like, which are somewhat incompatible with glycol unit operation. A fusized unit meeting the criteria in Chapter 10 should be used.Placing the separator in the bottom of the glycol absorber may be satisfactory but seldom is there enouheight to provide performance equivalent to a regular separator. Consider this carefully and do not be too ideaistic. In some circumstances this approach represents poor economy.

OPERATING PROBLEMS

The glycol unit should be essentially trouble-free. It seldom is. Many of the problems stem from inadequate design and/or operational faults. The basic simplicity of the unit and the availability of "standard" untends to obscure the need for attention to mechanical design details. The glycols are very reactive chemicaland need to be protected from contamination.One common symptom of many problems is excess glycol loss. This loss is due to one, or a combina-tion, of the following:1. Foaming2. Degradation3. Salt plugging the regenerator still column4. Inadequate mist extraction5. Inadequate absorber design for flow conditions6. Loss of glycol from pinholes in a gas-glycol coil in the top of the absorber or in a chimney trabove a separator section in the bottom of the absorber7. Spillage of glycol or pump leakage8. Lean glycol to absorber is too hot

Glycol likes to foam. It will foam whenever allowed to. Ordinary foaming may not be critical if the unitcarefully designed. Any foam tends to be more stable when aromatics and/or sulfur compounds are presenMetallic sulfides and sulfites, and degradation products, all contribute to the problem.Foams are only broken using surface and time, or chemicals. Tray spacing must be large enough so th

foam cannot fill up the space between trays and form a continuous liquid phase. A mist extractor does nbreak foam effectively. Once foam fills the absorber, there is a continuous liquid phase for glycol to go ooverhead.

Use of an antifoam agent can reduce the problem. Fig. 18.9 shows the. effect of adding triocty!phospha~eto maintain a concentration pf 500 ppm. There are many antifoam agents available. One that works In on~ Umay not work equally well in another. Some trial-and-error testing of anantifoam agent, and concentrationthat agent, is often necessary.

Avoid adding too much antifoam agent. If too much is added it may accelerate foaming. Set up a carefcontrol policy so operators keep unit concentration within the limits specified.


22/36


10

19~4 II .. '" " t9~, 10 '1 I1951Figure 18.9 Possible Effect of Foam Inhibitors in TEG 5ystems.(18.8}

Degradation is a natural occurrence and is accelerated in the presence of sulfur compounds. The answereffective filtration. Degradation products contribute to foaming but they also are major sources of corrosionblems. The best filtration system uses a regular filter to remove the large "chunks" and then an activatedbon filter to remove hydrocarbons as well as fine contaminants that pass through the first filter. The initialst is higher but the carbon filter may offer a favorable benefit cost/ratio.Salt is a continuing problem. Good separation ahead of the absorber is mandatory. Any salt arriving atregenerator deposits either in the still column or in the reboiler. It is common for packed still columns tog up to the point glycol is lost overhead. If this does not occur 1 salt can plug the reboiler and cause failure.t providing good separation is inexcusable.The water vapor in gas is relatively fresh but is slightly saline. NaCI is soluble in TEG to some degree.50C about 3.3 kg will dissolve in 100 kg of TEG. So, some salt is always present. The soluble salt hy-lyzes to HCI and lowers the pH of the glycol.

Glycols are very reactive with sulfur compounds. The resultant materials tend to polymerize and formunk" which is very corrosive. Also, the glycol pH becomes lower. Corrosion inhibitors alone cannot solve theblem satisfactorily. The real solution is good mechanical design and good filtration supplemented by a cor-on inhibitor. Good design involves factors like control of fluid velocities, long radius ells and a host of little details thatseldom are done properly. In many cases, good mechanical design will eliminate the need for expensive alloy

If feasible to do so, the glycol pH should be maintained above 6.0. Some become so pre-occupied keep-it at 7.0 or above that they add copious amounts of caustic, sodium carbonate and the like to the unit. Theult is seldom satisfactory. Adjustment of pH is proper but the cure can be worse than the disease if it is

Corrosion inhibitors which plate out on metal surfaces and form a film can be effective in minimizingrosion. Many materials are available. A product called Nacap has been used often. Some have used thenes effectively.Fig. 18.10 shows one result of using a corrosion inhibitor in a glycol system.Notice that corrosion was not eliminated; it merely was reduced to a satisfactory level. In a corrosiveironment, the total elimination of corrosion is an unrealistic goal. The proper goal is reducing it to economi-ly tolerable levels.

326


23/36

Glyco l D e hydr at io n

I ~O PPM "INf-t'BITOR ...... ST......4Te:O I I II~H .........JU~TItO TO a \lVITM C......STIC ! : I""'! I ! II ~ INL7: s~ :~ L...~ g L ..~A ' II ~ I I II ~ -. tNHSTOA ..........or:.oJ I I~ " " ' \ f I I II // \ I I CONT ....CTOR PRI!:SSUA~. 000 ,asia.I 0 ....S COM!DOSI"TION I.1 H2$~aG~ /100 C : . , . .M~RC""'' 'T'' 'NS. ~4 OA. I 100 c.",., C02. 1.30 %. Tr:::t...c : 1 ! ! : . 0 '" ....,~I/o-- I I , I I I-+ j II i 1I \ I ! I -I \ i i I -_JI I \ , I I I , I , - ,.,.-l--'",

I \i7~ ! ' , // ,iI 1 _ I : II/~!-=- II, I .'-J --- I I

'00

eo':>0"030

20

'0

.0.:>. . .

IQ!!I;O Ig~1 , , , , , . , .

Figure 18.10 Possible Effect of Corrosion Inhibitors in TEG Systems. (18.8)

Figure 18.11 shows the solubility of H2S in TEG.(18.9) This is true absorption that takes place inabsorber. It lowers pH and provides a mechanism for reactions. Figure 18.12 shows solubility of CO2 in a 9wt% TEG solution. (18.10) Solubility of C~2 in pure TEG is approximately 20% higher. Reference 18provides additional data on the solubilities of H2S and CO2, as well as C1 C2 and C3 in TEG.

S olub ility of hyd rogen sulf id e in trie thyle ne glyco l a tvarious pa rtia l p re ssures in the range o f 30 to 2 300 F .F i 110 0 '''I.

-PartIal premn

o f I rydrog.a '>llfh".:;: ~...::psIg .I S O -

-10 0

-S O

~0 - -

-

10

20

2 0 60 a o 100 120 140 160 110 200T t f II I* ' I IIUn F .Figure 18.11 Solubility of Hydrogen Sulfide in TEG (18.8)

327

22 0


24/36


LL.0 1500to'0g'"~-I 100o!0uE


25/36

Gly col Deh yd ra ti on

I . , . , . . . . . . . . .1 11 11M "1. r... ,...., ...., , , . . . . . . "J. W " . . . . ,.... .,1 "'_I "I.""' ,......., " II, ,._ .. ...,

1 . " " " _ _11.~w....,,,,,,,,,,,".~""'-1 , ,....,...,14 .11- ,., .,. .. "" . ..IS. g". I I....16. l i t o . . . . r.-..&w .........

Figure 18.14 layout of One Offshore Platform for Dehydration (18.13)

REFERENCES18.1 Worley, M.S., Gas Condo Conf., Univ. of Oklahoma, (April 1967).18.2 Perry, C. R., personal communication.18.3 Townsend, F. M., Ph.D. Thesis, Univ. of Oklahoma, Norman, Oklahoma, (1955).18.4 ScauziJIo, F. R., Jour. Petro Tech., (July 1961), p. 697.18.5 Rosman, A., lbld., (Oct. 1973), p. 297.18.6 Worley, M.S., Cdn. Petr., (June 1967), p. 34.18.7 Parrish, W.B., et al., Proceedings GPA 65th Annual Meeting, San Antonio, TX, 1986.18.8 Swerdloff, W., Ibid., (April 29, 1957), p. 122.18.9 Blake, R. J., Oil Gas J., (Jan. 9, 1967), p. 105.18.10 Takahashi,S., et sl., GPA Technical Publication TP-9 (1982).18.11 Jou, F.Y., et al., Fluid Phase Equilibrium, 36 (1982), p. 121.18.12 Carmichael, C. J., lbld., (Nov. 2, 1964), p. 72.18.13 Pierce, D. W. and C. l.Finch, Ibid., (April 29, 1968), p. 70.

329....---~.~~~~-


26/36


u0II) 10000"0 8000roro 6000c..~00 4000. . . . .. . . . ,ro 3000-E"0 2000 2000i . . . . ,'lD III0. . . . . .-Ol~~ 1000 1000(/)< 800 800. ! : I_.J2 600 600=:lI-


27/36

Glycol Dehydration

~' ",Q,, . . ...;-ee. ." "0'". ...!. . . .U


28/36


Phy6 ~cal Proper~ies of GlycolsEG DEG TEG

Formula C,H.D: C.H,.D, C.H I .D.MOlecular Weight 62.1 106.1 150.2BOiling Point at

760 mm Hg of 387.1 472.6 545.9Boiling Point at

760 IIVI\ Hg C 197.3 2H.S 2B SVapor Pressure at 2SoC,mm Hg 0.12 0.01 0.01Density at 2SoC , g/=) 1.11(I 1.113 1.119

at 60C , s/em3 1.OE: ' 1.OBS 1.09:Pounds p er " .: Il lo nat 2SoC 9. ~6 9.~9 ~. 34Freezing Point, =c -1 3 -8 -7Pour Point, c - -54 -58Viscosity inCentipoises ilt2SOC 16.5 28.2 37.3

At 60C 4.6[; 6.99 S.i7S u rf ac e T en si on a t 25CDyne!;/cm 4i 44 45Refractive Index at 2SoC 1.4 r o 1. 446 1.454S pe cif ic H ea t at2 S C 0.5& I 0.55 0.53Flash Point, c (eOC) 116 138 160Fire Point, c (COC) 119 I 143 166

HEAT TRANSFER FACTORS FOR GLYCOLLiquid Heating or Cooling Vapor Hea ting Condensing BoilingMaterial Re < 2 100 Re > 210 0 or Cooling Vapor Liquids

E thy le ne Gl yco l 0 .49 0.1 2 0.60 0.18 0.30Die th len e G ly co l 0 .50 0 .1 1 0.60 0.17 0.30Tr ie thy le ne G ly col 0.49 0 .09 0 .6 0 0 .1 5 0.30The water factor = 1.0. Multiply heat transfer coefficients for water by thefactor shown to obtain glycol coefficient for the service involved.


29/36

Gly co l Dehydr at io n

o 4.50.C~ 4.0


30/36


TEMPERATURE. C.

w(/)o 0..

50~'----~I----~L----~I----~~----~---I~--~--~~--~--~~40 I I -.::::::.] _.-P' "'t---_ I i30 TRIETHYLENE GLYC0,b _...---r::>-.._ ~I IPER CENT BY WEI GH~__""'--- ~ --....._ I - -............._20~ ' : g _ _ " . - ~ r-.; ~r--- - - l10- 40 ----.. :E~~~~s~~~~!~~t--..gl3~'~~---t-I~'~~= = 20 '\' I __ --, I

:- O~~ ~ ~~ ~. ~ _~ :_:'

3 t = = - ~ ~ ~ ~ : - - ~ ! ~ ~ = - ~ - ~ ~ - - - ~-~I--r-~~__-f-ZWU.> -f-(/)oo(/)> 21----L i"\ I I - - - r - - - I r---I I ~L r - - r - - - + - - I--I I

O . s - r--.~ -.6 I -=, _ _ II ----+--=I -0.4 0 10 20 30 40 50 60TEMPERATURE. C.

334

- -


31/36


10050

105

CDIEE .5wa ::::J(J)(J) .1wa::0.. .05a::00..-c> .01.005

.001.0005.0001

.00005.00001-20

-41000500

EXTENDED VAPO R PRES SURE CURVESF O R G LY CO LSTEMPERATURE , O F . .

68 104 140 176 212 284 392 5722! '-f--1--!--!--I-i-f--1--....1---1-- V /1V IfJ, / J Ir----

ETH YLEN E G LY CO L41 / VI_.. - v IIV1 I V V- - t v v LV I TR IETH YLE NE G LYC O L/ V VI J1 . - V V Jf- ~ ~/ K . .I I/' I ~r--:--: DIETH YLENE GL vcoi ,Iv / VA / I' //

V r/ IV i 7// v I1/v y

o 20 40 60 80 100 140 200,TEMPERATURE, "C .

932 1292

soc 100


32/36


1000800600500400

0 300:I:E 200EU Ja :::;, 100VIVI 80Ja ::c, 60a :: 500c, 40 3020

1000700400200

'" 100:I: 70 EE 40- 20a :::;,VI 10VIU J 7a ::a, 4a ::0 2Cl .< I> 0.70.40.2

,, / //' V V ' /1 Y '/ 11 / // / /. '/ ' /' / ,1 1 , '/ .// 'L IL , ./ Y / , ,1 Y V/I ./ /"1 V'/i , , / , I ,

~~ATERtv / ~ V Vy /1 /i' 1 1 LI I I ,V./ V /'VILf_Xl VI I 1 I II V/t~ ViLlA' I I V I I I I I ,./ I I~31 kYI IMJ" " 11-:W/ I ETHYLENE GLYCOL II i ! 1 1I I ,

/ /' /, / ./" "" "-


33/36

Gly col Deh yd ra ti on

1.116

o 1.1000N. . . . . .. . .>- 1.084. . . .>a:< . : : l~ 1.068LL.uWQ. 1.052I

1.036

1.020

1.132A __ r--- --..:::: : : : : : : : - - - - - I--_ ' ::-- f : : : : : - I--- - . : : : . _ _ I--- - I ' : : : : : : : : : t::-- -I--. . . . . . A- - . : : : : r : : : : : : 8C

A TR IETHYLENE GLYCOL8 D IETHYLENE GLYCOLC ETHYLENE GLYCOL0 PROPYLENE GLYCOL

0 I- - I--~ II"--'---LI I---. . . _ _I -- - - 0o 4 8 12 16 20 24 28 32 36 40TEMPERATU RE, C .

I. I 4 r-,----r----:---r---r---r----:----:----:---r-~__

A #~V " ,~lP " I.0 0 L-J.._..I.-______:,__.___._____~ ___,__.._____

1.12o 1.10oN. . . . . .oN.. 1.0 a>-. . . . .>a:

< . : : l 1.06~LL.UWe ; 1.04

1.02

A TRIETHYLENE GLYCOL8 DIETHYLENE GLYCOL

____ A~~--r--r~~~~B'~/CPROPYLEN E GLYCOL~I--~~~'~I /~~

C ETHYLENE GLYCOL

I )/V

I . . . . 0d ~ ' - - . . . . . . . . .Wyo 20 80 1000 60

GLYCOL,PER CENT BY W EIGHT


34/36

Gas C onditioning and Processing - T he Equ ipment M odu les

uu :u~ 0.7~~~~---+--~~~~-r--~--~--+-------~~(/)

0.5L-~~~ __~ __~ __~ __~ __~ __~ __~ __~ __~~-20 o 40 60 800

TEMPERATURE. C.

TRIETHYlENE10 I GLYCOL'. PERI I CENT BY WEI'GHTI I

I I I I0.9

U0. . . . .0-. . . . .

.. :


35/36

Glyc ol D e hydr at io n

.65

6000~ .55E.. . . . .45-. . . .c.J::J~ .40c0C)

" ' .35Ec . . .cD..c~ .30

.25

C)o~ .70E

>. . . .c.J::J

~ .40 ~oC)

" ' .30C-cD..c~

F . d Eth> lene '" , " ++r+r-GI1.col Percent by W ei ht ~tt0 ~ .10

20

30+.o..~ ~0 ~:lf. . . .~ . , . . f+H-

50 ttlF~1~60 ~~. ' '_"H"'~70 , _ .. . . . . . ' i:--:::

80 . .. .. ; !. ..~.::. ~ . . . . .. 90 f- .....+-~-j-":iI~~--+ 100 -'+ +rr 1 - < . : . . ~ t n1:.Jt::r.fTl ' -r+ ' :ttH" 0+' ~'t,+.:.r.L;r .::~_;::_

e Diethylene Glycol. Percent by W eight

o

.60

10 40 50 60

.50

.20 o

20 30Tempera ture C

o/02030405060

1\1\708090......100. . . .

10 50 600 30 40Temperature, C

70 80 90 iOO

70 80 90 100


36/36

Gas C onditioning and Processing - T he Equ ipment Modu les

.0018E. . .U .0016E. . .,;. .0014.U. ... . . . ..: .00120. . .0->-I- .0010>I-U~0z.oooe0u...Jcr: : e .0006c ::w:I:I-

.0004

7

70U0I()C\J'0 65Eu ;;Q. 60'".c:>.-e~ 55z0VIZw 50I-wUcr1 . 1 . .c :: 45~VI

40

35

I I I I I I ITRIETHYLENE GLYCOL, PER CENT BY W EIGHT_ _ - - ~~ ~0 ~ - - - -:----- 10 ~ ~_ - ~_ - - 2030

40 I I II5060708090

100 I

10 20 80 90 1000 40 60 700TEMPERATU RE . C .

5 A I I I IoD~ A DIETHYLENE GLYCOL r--'\ \ B ETHYLENE GLYCOLC TRIETHYLENE GLYCOL~ D PROPYLEN E GLYCOL\:~ -~ I I I\

Date post:	07-Apr-2018
Category:	Documents
Upload:	harisshajar
View:	216 times
Download:	0 times

18 GCPII Glycol Dehydration

Documents