F.B. THOMAS, N. DEDORA*, X. ZHOU**, D.B. BENNION Optimization/Optimizing... · gas condensate reservoir. Two poblems ... Rich condensate reservoirs often exhibit an inherent...

F.B. THOMAS, N. DEDORA*, X. ZHOU**, D.B. BENNIONHycal Energy Research Laboratories Ltd.

*now with Elan Energy Inc.**now with Green Pipe Industries Inc.

for the two phase Z factor material balance conecUon which isextensively written of in the literature<IA). The hiib-mobility gasmakes it extremely difficult to produce the liquid phase hydrocar-bon once it has CCMIdeIIscd; moreover ~ ~ ~tscan seriously reduce gas productivity.

These factors, although bard to oven:ome, Deed not precludeeconomic operation of. gas condensate reservoir. Two poblemswhich are commonly observed in gas condensate systems are nowdiscussed.

Problem 1: What is the Character of theFluid In Situ?

Rich condensate reservoirs often exhibit an inherent difficul-ty-char8cterization of die fluid. Often one has a rich gu which isin contact widt one or mOle separate and distinct oil phases in situ.In approximately SOCII of die cases which have been anaIysed.when separator gas and liquid samples ~ ~bined in die 1abo-ratory. the saturation ~ure of the ~ombined fluid exceeds themaximum ~ure of the reservoir. Figure 2 describes die situa-tion. For example. the GOR observed at the field separator isshown by dte GOR asterisk. As one contacts dte separator gaswith the separator liquid at reservoir temperature, one mayobserve a saturation pressure much higher than the reservoir JXe5-sure. The choice one has is either to increase CK decrease the GORin order to reduce dte saturation pressure to reservoir JXe5sure. Inmost cases it is preferred to decrease the GOR since, due tomobility effects. gas flow would normally be expected to be high-er dtan that co~sponding to the actuaJ in situ fluid. Using dtistechnique provides at least a necessary condition for the recombi-

Technical Background

The problems associated widl gas condensate systems are as(or more) challenging as those associated widl oil reservoin.Figure I shows a typical JXessure-temperatUIe diagram. To the leftside of the critical point one observes OObblepoint systems. Heavyand conventional oils are examples of these. To the right of diecritical point one observes dewpoint systems. The literature is~plete widl discussions of P- T diagrams whe~in one observesthe maximum p~ssure as die cricondenbar and die maximum~ as the criCOIxIenthemt<I.2,3). If the prasure and tem-perature of the specific composition place die material in thereservoir to the right of the phase loop (at a reservoir temperaturegreater than die criCODdendJenn) then ~ can perfonn any pres-sure manipulation whatsoever and a single phase (in situ) is the~It. These are nonnaIly called dry gases. The region in betweenthe critical point and the cricondentbenn is where gas condensatesappear.

The quality lines describe die amount of liquid and gas whichare at equilibrium at a specified p~ssu~ and temperature. ForreservoiR whose bebavi<JUr is very close to die critical point. veryhigh liquid dropout can be incuned at ~ close to saturationp~ssure. If production by primary depletion is chosen. as die liq-uid condenses from the gas phase and ~umulates in the reser-voir. cash flow decreases. By leaving die ~ condensatebehind in the ~servoir. one has not only decreased the totalhydrocarbon which will be produced. but the components ~main-ing in the reservOO are of the highest value. This forms the basis

()ctd)er 1997. ti.ne 36. No. 9 43

. c -L--:-"

~--~Ii

"'

GaR

FIGURE 2: Typical separator gasltiquid P-GOR diagram.

nation procedure. In no way does dUs mean that the saturationpressure must be equal to die reservoir pressure. In some cases thesaturation pressure of d1e in situ fluid may be lower d1an d1e reser-voir pressure.

In most of the cases depicted by Figure 2 the much higher satu-ration pressure may be due to excessive preferential gas flow orcontamination from a liquid phase in situ wherein a free liquid.such as a heavier oil. is produced at the same time as d1e gas con-densate. In some cases die oil produced is of a much lower APIgravity and very little contamination may result in an extremelyunrealistic dewpoint pressure. In these cases one can only hope toobtain representative samples by adhering more closely to sam-pling techniques described in the literature<S.6). Sometimes bottom-hole samples provide more representative compositions of reser-voir fluids. Systems such as dUs can normally be evaluated andconfinDed by aoalysing d1e heavier ends of the liquid phase pr0-duced and also by aoalysing d1e saturation pressure characteristicsas have been described.

At this point if one detennines that con~~~~tiOl1 exists. thendtere may not be much benefit to performing laboratory studies.This needs to be detailed before lab work begins.

Problem 2: Retrograde Condensate EffectsRetrograde condensation results in a number of problems. The

most obvious and serious of d1ese is lost productive CaplK:ity dueto accumulation of liquid in the reservoir. This bas two facets: diefirst is associated with not being able to produce die higher valueliquid components aDd secondly, die increased liquid saturationresults in reduced gas flow rates. These facton work in conceitaIKf die more serious die liquid dropout, die greater die reductionin gas relative penneability.

For a system which is single phase initially, relative pemleabil-ity effC(:ts are absent. Relative penneability should be viewed as adependent variable detennined by three odleI' parameters or influ-ences. These general influences are associated with:

1. intelfacialleDSion effC(:ts,2. viscosity ratio,3. pore size distribution.

By definition, IFf effC(:ts are only involved when two or morephases are present. The interfacial tension is important because ofcapillary pressure, which can be viewed intuitively as Equation(I), wherein in order to sweep through a pore in a two phase sce-nario, die differential pressure must be equal to or exceed die cap-illary pressure.

decreases, the capillary pressure holding the liquid in the poreincreases. Therefore, to be able to produce retrograde liquid fromsmall pore duoats, one must have either a very high differentialpressure driving force or low interfacial tension.

Extrapolations of this thinking would indicate, that for gas con-densate systems which exhibit high interfacial tensions where thepore throats are very small, which may correspond either to lowpenneability rock or higher permeability rocks but with very largecoordination number, the success of flowing the liquid from therock, once it has condensed. will be limited. In such cases vapour-ization (lean gas cycling) or injection of IFf reducing agents(COz, c, or ~) may be the only option to enhance d1e perfor-mance. On the other hand if the equilibrium gas and liquid exhibitlow interfacial tension, then the liquid may flow freely from mostof the pore duoats in the rock and very little retrograde conden-sate relative penneability reduction will be observed.

Where the pore tlu-oat diameters are much larger, even thoughdIe interfacial tension may be high, it may be easy to overcomethe capillary forces which are keeping the liquid in those porethroats. Therefore. for larger pore throat systems which may cor-respond to higher penneability rocks and/or small coordinationnumbers, the interfacial tension effect may not be very important.

Unfortunately, other factors cloud the issue. One of the factorswhich complicates this development is the mobility effect. H onewere to analyse the viscosity ratio between most equilibrium gasand condensate systems, the viscosity of dIe gas would normallybe at least 10 - 20 times lower than that of the condensate phase.Due to the inherent nature of the less viscous phase to flow morereadily, it will tend to take d1e path of least resistance and willprefereDtlally flow through the larger pore throats. GaIdner<7> hasshowo correlations wherein for gas/liquid flows. expooentlal vis-cous finger growth is often seeD. Therefore in light of the IFf cri-terion [EquatioD (I)J, one may automatically conclude that thelargeI" pore throat sizes will contribute to much easier prodUctiODfrom a reservoir which exhibits liquid condeusate effects. This isusually true. However, eveD dtough the fluids may have thecapacity to flow through most of the rock where the pore throatsare larger, die gas may preferentially "choose" only the largestpore throats and may bypass die rest. Therefore even dtougb dIegas may have a low enough IFf to poteDtially sweep all porethroats, it may, due to mobility effects and inhereot differences inresistance to flow through different parts of the rock only contactthose of larger dimensioD. Therefore a compromise will bereached betWeeD interfacial tension and mobility. H d1e system ismobility or least path of resistaDce-dorn~at~. then the only wayto effectively reduce the liquid saturation may be to vapourizecompooeots from the liquid into die flowing gas phase. Whed1erd1e free liquid will flow, or if improved recovay will have to rely00 extIactioo effects depends 00 the compromise reached betweenIFf and mobility in the presence of the actua1 porous media.

In order to assess the optimal way to produce a gas condeDsate.one must be ~ to perform the testing in die pcsence of theporous media. H the actual reservoir rock is not used, one mayhave die appropriate viscosity ratio and interfacial teDSion, but theconclusions drawo may be inappropriate in light of die fact thatdie compromise between mobility and IFf is Dot d1e one whichwould be consistent with die reservoir rock.

It should be noted at this point that in light of the couplednature of interfacial tension, mobility and die pore size distribu-tion of the porous media laboratory testing must adequately repre-sent each of these three factors. H any portlOO of laboratory analy-sis does not adequately represent these parameters, then the cre-dence which one can lend to the conclusions drawo from a moreroutine laboratory study may be minimal. That is, if viscosity ratiois used as a reference for atmospheric relative permeability test-ing, but the interfacial tension of the fluids is Dot IIIatcl1ed. thenone has a priori biased the conclusions in favour of the systembeing mobility dominated. What then are the laboratory testswhich should be performed to evaluate a gas condensatereservoir?

Interfacial Tension

Radius~.. «

.(1)

One can observe from Equation (I) that as d1e interf~ial ten-sion decreases, ~ capillary pressure decreases. Convenely as theradius of the pore throat which contains the retrograde liquid

The Journal of Canadian Petroleum T ectv1Oiogy44

TABLE 1: Comparison of typical gas condensate characterizations.

Plat c ReservoirSeparator

GOR(ms/m~

ModifiedGOR

(mJ/..~

System #

laitial Plat

(psi) (PSI)(kPag)~

2413219805

i 17237

~~

~':

~,

i

9901100

13994

~

280<2S(K

~

is measured. On this basis one can determine if the liquid is ofhigh enough volumetric proportion to cause a problem.Usually, those fluids which exhibit less than 1% retrogradecondensation show very little tendency to reduce gas relativepenneability. However, in some cases, even at low liquiddropouts, the liquid tends to migrate into the production well-bore and result in reductions in gas penneability. The evalua-tion via CVD, however, introduces one to how serious theproblem may be and initiates the overall evaluation on a labora-tory scale. Figure 3 and Table 2 describe some of the dataavailable from a CVD experiment. In addition to liquiddropout. one also obtains needed information for material bal-ance calculations and preliminary characterization.

If on the basis of the CVD testing one bas determined dIat diedegree of retrograde condensation is minimal. one may proceedto perionn pressure-temperature diagram testing and die scnsi-

Recommended Experimental Proceduresfor Gas Condensate Evaluation

In light of the many factors which are involved in the produc-tion of gas condensate reservoirs, it is the experience of theauthors that a number of tests be perronned in order to optimizethe production strategy. The recommended tests are:

1. Appropriate characterization of the fluids.2. Extent of retrograde condensation.3. The retrograde condensate relative permeability reduction.4. Efficiency of cycled gas in ameliorating gas productivity in

~ ~ of retrograde effects.5. Quantification of the influences of gas cycling, including

composition and pressure effects.6. Simulation.These will now be commented on in order.

Reservoir Fluid CharacterizationAs was mentioned initially, one must know the nature of thefluids which have been sampled. If contaminants are involved,or if non representative gas and oil production rates are mea-sured, then the fluid characterization will not be adequate. Thetechniques described previously should be foUowed and theproper sampling procedures adhered to. Those reservoirs whichexhibit the most challenging problems are those which do haveseparate liquid and gas phases in situ, as wen as those reser-voirs which require greater drawdown pressures to produce thefluids. In some cases bottomhole samples may be required, andeven then such may be difficult to capture. Table 1 shows a setof typical results obtained on systems which showed deviationsfrom reality in terms of saturation pressures at separator GORs.

Evaluation of Retrograde CondensationOne of d)e standard tests which is often applied to detenninethe response of a gas condensate system is the constant volumedepletion test wherein the liquid phase as a function of pressure

45October 1997, Vdume 36, No.9

pressure. As the liquid accumulates, one may see drasticdecreases in gas relative penneability. Figure 5 provides evi-dence of this effect in the presence of the same hydrocarbonfluid but widt changing core permeabilities. The lower perme-ability core exhibited the most severe relative permeabilityimpairment whereas the higher permeability cores exhibitedless severe reduction. This is intuitively consistent with theIFI'/mobility discussion provided previously since dte gas/liq-uid IFI' for this system was between 10 and 20 dyne/cm andthe viscosity ratio was about 30.

Evaluation of Cycled Gas for AmelioratingGas Productivity

In light of the results of d1e previous testing, wherein reductiOI'of gas permeability was observed, gas injection can be evaluat-ed widl respect to its effect on restoring gas productivity. Thus,when retrograde condensate reduces gas relative penneabilities,die injection of potential lean gas or rich gas streams can beconducted to quantify dle gas' ability to remedy die pemleabili-ty impainnenL For example, if pursuant to d1e relative penne-ability reduction caused by retrograde condensation, dlere is noapparent improvement in gas permeability by d1e injection ofmethane, then Cl would be an unlikely candidate for gascycling operations. If, on the other hand, a combination ofethane and metl1ane were to be effective at increasing die rela-tive pemleability of gas subsequent to retrograde condensation,d1en this would qualify for further testing- and evaluation. Tooptimize die gas used for cycling one may dlen need to knowwhich effect is dominant in ameliorating dle gas relative per-meability; is it the change in IFf ,improvement in mobilityratio; or simple reduction in liquid volume due to vapourization

tivity of such on addition of lean gas in order to effectivelyreduce die critical temperature of dte reservoir fluid to a pointwhere the cricondentherm is equal to or less dtan the reservoirtemperature. This is described pictorially in Figure 4. This fIg-ure shows that after three months of lean gas cycling, the effec-tive pressure-temperature diagram of the reservoir bas reduceddie cricondentberm. However, one must cycle for at least twoyears in order to reduce die cricondenthenn of die effectivereservoir fluids to a point where it is less dtan die reservoirtemperature. After two years of cycling, according to Figure 4,one can dien discontinue dte cycling operation and simplydeplete the reservoir. This is advantageous because once thisoccun, then primary depletion win recover virtually all of thehydrocarbon in place since the system is now outside the phaseloop.

Unfortunately die scenario described in Figure 4 is usually tooexpensive to implement. and particularly for diose systemswhich are much higher in retrograde liquid saturations.

Retrograde Condensate Evaluation in theCore

A necessary condition for detennining whether retrograde con-densate effects are going to be a problem is to perf OrIn depres-surization steps in a core and observe die reduction in gas per-meability at different pressures. Figure 3 shows an exantple ofincreasing liquid dropout at pressures below die dewpoiDt pres-sure. Knowing die profile as in Figure 3 one can dJen producedJe desired liquid saturation in d1e ~ simply by reducing die

46 The Journal of Canadian PetmIeJ.n T ~

IL-

~

iI!

Pr-.e

FIGURE 9: Expected gas cycling behaviour.nGun 8: Gas cycling raWts interfadal tension proC'lJe.

of components? The next phase of testing defines theseinfluences.

Cycling TestsOn the basis of the retrograde condensate evaluation in thecore, wherein the influence of retrograde condensation on rela-tive penneability bas been quantified. one can detennine whatwould be the optimal way to produce a gas condensate reser-voir. The focus of these tests is to produce the most pertinentdata which can be connected to field scale modelling efforts.One must evaluate what the influence of cycled gas composi-tions and operating pressures will be. In perfonning thesemulti-cycle tests, the following parameters are quantified:

a. Mass transfer between the retrograde liquid and the

cycled phaseb. a.aoges in viscosity of the upper and lower phases

c. IFf changes

The most conventional application for gas injection bas been toinj~ sufficient lean gas into a reservoir which will result in asuppression in the overall critical temperature of the system.This is shown schematically in Figure 4.

MOle typically a gas cycling scheme is implemented wherebydie deleterious effects of retro~ condensation are reduced.The cycling bas two effects, namely, the extraction of com-pounds from die liquid phase thus resulting in the recovery of~ of die liquid, simply due to phase behaviour. Figure 6shows compositional d1anges OCCUlTing in die liquid phase as afunction of discrete cycling of gas in a laboratory experimentInformation which is gained from this type of test can be veryimportant for a number of reasons.

The fust set of infonnation which is available from Figure 6 isd1at it identifies die components which are going to be avail-able from d1e oil due to gas cycling. In Figure 6 one can seethat components up to CII are extracted from d1e liquid phaseby d1e gas phase. Based upon die concentration of these com-ponents in d1e liquid, one can then get an idea of how muchrevenue this would account for if the only liquid recovered bygas cycling was due to extraction of d1ese components.

One can also contrast d1e effect of odler gases, as in Figure 7wherein a comparison is made in die composition of d1e gasproduced. The injection gas has been equilibrated wid1 die liq-uid phase which has condensed in d1e reservoir and the equilib-rium gas has been flashed to standard conditions. ~ is moreeffective at extracting components from die liquid than is C,and H2S is more effective than CO2- This will have d1e twoeffects as deflDed previously. Infonnation such as this is alsoextremely important in d1e development of d1e d1eoreticalmodel to ~ct this type of mass transfer in subsequent simu-lation work. For example, the standard characterization proce-dure would normally be based upon a Change in K values ond1e basis of correlated critical properties. In that case, onewould probably see d1e ~ fraction divided into five or sixhypothetical components, and typically these componentswould reflect the following character:

i) ~-Ctoii) CII - C'5ill) Ct6 - c,.,iv) Cn-~v) ~-~vi) c".

The infonnation from Figure 6 states that all d1e components

...~(%)

N-.~Row

--0-

~.1 ,/',.

~~

FIGURE 10: Expectedrelatloashlp betweenp~un, ~Yery andcash now. PresstK8

47October 1997. Volume 36. No. 9

into account in the gas design. One would see liquid yield in(Cfms of bbVmillion, for example, ploaed VI. JXeSSure. In diiscase one may see a minimum liquid yield a( a press~ some-where in the middle of die two phase loop (die phase loop asassociated in Figure I). Depending upon the reservoir tempera-ture i( may be technically superior to operate 81 higher pressureand to inject a makeup gas in order to see die effects reponedin Figure 6 and be able to recover many components from theliquid phase. Volatility effects will only be associated widicomponcnts usually less than c. silx:e, for eumpIe. the roilingpoint temperature of ~ is in the I07OC range. However. widigas cycling, due to the extraction, one may see substantiallyincreased liquid yields by maintaining a higher pressure in diereservoir and cycling the gas.

SimulationOnce the evaluation has been perfonned and the data set devel-oped, different production strategies can dten be compared byusing a compositional simulator. As mentioned. the ability toconnect the fluid flow to the phase bebavioor properties is acrucial issue and thus attention must be given to the relativepenneability model. The data used should be generated accord-ing to procedures described previously<') aI¥I these can then beused to model single wellbore systems. On the basis of theweUbore modelling, (MIe can dten prod1x:e a p&c. such as Figure10, wherein superimposed on Figure 10 are the results ofFigure 9, along with the economics. In such a case, that whichwould produce maximum liquid yield may ~ appear to pr0-duce the best net cash flow. The major beacfit of implementingcompositional simulation, which bas been tuned to an adequatedata set, is to be able to examine the potential for revenue gen-eration as opposed to the D'~r1-l!'jz~tion of hydrocarbon recov-ery from the reservoir. Of course, parametas such as gas avail-ability and am as well as compression eater stroogly into thisequation but once the data are available from the experimenta-tion, one can produce the best possible estimates of businessstrategy on these gas condensate systems.

SummaryI Gas condensate reservoirs often exhlmt difficulties associat-

ed with characterizing the fluids in place-This is particularlytrue where simultaneous production from a rich gas zoneand a separate oil phase in am is ,-,bsu-~ aose 81ten1ionto this must be given in order to ~ ..-o:seDtative Iab<xa-tory data sets.

2 The effect of reuograde coodeDSatiOll on gas condensateoptimization is ~y impcxtaDt ~ a ~ in gasproductivity OCCUR at the lame time the more high valueadded components accumulate and remain behind in areservoir.

3 The success of a mixed gas reservoir production strategywill be determined by a compromise reached between:

L Mass transfel' effectsb. Interfacial tensionc. The viscosity ratio between the gas and the liquidd. The pore size distribution in which the retrograde

condensate occun4 Once these ~ have been evaluated, then coupling

this into a reservoir simulator is the best way to obtaininsight into the ecoIK)Inic forecasts.

beyond CI1 are essentially non volatile and are not involved indie mass transfer. In light of this, ~fore. die data would sug-gest that a better characterization scheme for the ~ regionwould be the following:

i) c. as a library componentii) c, - library componentill) c. - library component

iv) c, -library componentv) Cia -library componentvi) C11+ would be the last hypothetical component

In d1is ~ ~ has used the ~ number of compooentsbut has associated the discretization in the ~ region to thosecomponents which are involved in the mass transfer. Thus withthis data one stands a ~ ~ of being able to simulatethe mass transfer effects which are observed in the system.

The second effect is that the gas which is cycled may result inmanipulation of the interfacial tension as weD as the effectiveviSQ)Sjty ratio between the gas and the liquid. Figure 8 showsthe influera of cycle numbec on die IFf between the gas andthe liquid. The teChnique for measuring IFf is a pendant droptechnique similar to Rotenburg<i). As one would expect fromFigure 6 one wiD initially see ~ mass tl'aDSfer between the1jquid and the upper phase which IIKXIId result in a decrease inintafacial tension between that tquilimum gu and the equilib-rium liquid. However, what is plotted in Figure 8 is the interfa-cial tension ~hed between the cycled gas and the equilibrium1jquid and u the cycle number inaeases for medWIe and as thecompooeots which have been ex1IKted from the liquid phaseIre produced, the liquid effectively becomes a heavier phase.Thus, with the methane, which is probably only extracting c.or c, from the liquid phase (depeoding upon the temperanue),~ ~ expect to see inaeasing IFf u a fuJx:tion of cyclenumber u the liquid phase gets heavier. Therefore, withcycling methane, although ~ of the liquid may be vapour-ized and subsequently produced. which could result in a lowersanuation in the region of the weUbore, it may never be able todisplace liquid from the poIa ~...ux of the interfacial ten-sion. H one comparcs chat to what is observed with the CO2,one may see a deaeasing interfacial tension due to solubility of~ in the liquid phase and exttxUon of components from theoil. One ~ initially deaeasing interfacial tCIIsion as the ~IOlubilizcs in the liquid phase and d1is 00minates dte fact thatd1ere is extraction of the in~~ate components into the gasphase which Ire subsequeDdy produced 8M coodeosed at sur-f~ Howe~, ~ the mariml'm solubility of the ~ in theliquid phase is 8Cbieved, Iben the only d1ing left to occur isIUbsequeat mass transfer of the intamediates from die liquidinto die gas phase and d1is would presage the minimum IFfwhid1 is observed widilUbsequeDt iDc:reasing IFf as a functionof cycle number.

This indicates that if die liquid is not produced from the regionof the weUbore and the benefit of gas cycling is totally reliantupon exb'8Cbon of COInpoDCDts from the liquid into the gasphase, ~ the ttmsieot bebaviow' must be aMlSistent widl thisrepleoishing and subsequent extraction of these componentsfrom die weUbore region in order to maintain sufficiently lowliquid satunDon in die production weD region and benefit froma cycling operation.

The most important thing to realize with many of these systemsis that much of the hydrocarbon liquid will be lost once thephase loop has been en~ into and dJe main objective is totry to reach a compromise between the expense for dJe gascycling operation and trying to keep die liquid phase to a mini-mum so that the gas productivity in dte region of the weUboreis not too much impaired. Figure 9 describes the standardvolatility VI. extraction scenario in d1e context of liquid yieldVI. pressure. To dte left of d)e minimum point on the liquidyield one may see a system dominated by volatility, whereas tothe right of that point one may begin to see the extractioneffects. All of these parameters need to be optimized and taken

REFERENCESI. DAKE, L.P., Fundamentals of Reservoir EngiDeering; Elsevier

Pllblishing. Oxford. (1978).2. CRAFT, B.C. 8Id HAWKINS. M.F.. AppIiaI PmoIeum Reservoir

Ealineering; Preftlice-Halllftc., EIIg'-ood Cliff., HJ. 1959.~r2.

3. GILCHRIST. R.E. and ADAMS, J.E.. How to Best Utilize PVTRepOI1S; Petroleum Engineer Intemat;OflGl. (J.I, I99J).

4. MOSES. P.L.. Engineering Applicalions of Phase Behaviour of

The JoumeI of Canadian ~ T ~48

Authors' Biographies

Brent Thomas holds a doctorate in chemi-cal engineering. He has worked onenhanced oil recovery applications for thelast 10 years. including gas injection.chemical flooding. solids precipitation andthennal applications. He is JXesently vicepresident of Hycal Energy ResearchLaboratories Ltd.

Nancy (Holowach) Dedora received aB.Sc in chemical engineering from theUniversity of Calgary and is currentlyenrolled in the master of engineering IXO-gram. After working with PetroCanadaResources in the oil sands department, shejoined Hyca1 Energy Resealdl LabontOOesLtd. as a research engineer where sheworked in the areas of PVT analysis andmiscible and chemical flooding. She is cur-rently a reservoir engineer with ElanEnergy Inc.

Crude Oil and Condensate Systems; Jollrllal of PetrolellmTechnology. JlIly 1986. pp. 715.723.

S. STRONG, J., mOMAS, F.B., BENNION, D.B., Reservoir RuidSamplins and Recombination Techniques For LaboratoryExperiments; Paper pres~ted at the CIM 1993 AMIlaI TechnicalCoII/erence ill Calgary. May 9 - /2. 1993.

6. MCCAIN, W.O. and ALEXANDER, R.A., Sampling GasCondensate Wells; Society of Petrolellm EIIgilleers ReservoirEngineering, AI/glUt 1992. pp. 358-362.

7. GARDNER, J.W., VPMA, J.G.1., An Investigation of PhaseBebaviour'--~ Bypusina lDlenection iD COz Aoodin&;SPE 10686. SPEIDOE SYlIIposi"", on En1I4IIced Oil Reco"~ry,TII/sa. April 4 - 7. 1992. -

8. ROTENBERG, V., BORUVKA, L., and NEUMANN, A. W.,Determination of Surface Tension and Contact Angle from theShapes of Axisymmetric Ruid Interfaces; Jollrnal of Colloid andInterface Science. Vol. 93.. No.1. May 1~.

9. BENNION, 0.8. and mOMAS, F.B., Recent Improvements inExperimental and Analytical Techniques for the Detennination ofRelative Pemleability Data From Unsteady State Row Experiments;pruellt4d at the SPE 1~ Technical ~nnc~ and Expositiolllteldill Port~Spain. TrlIIidIJd. J- 26-28, 1991.

Provenance-Original Petroleum Society manuscript,Optimizing Production from Gas Condensate Reservoirs, (94-04), fiBt presented at the 45dt Annual Technical Meeting, June12-1S, 1994, in Caiguy, Alberta. Absb"Kt submitted for reviewJune 23, 1994; editorial comments sent to die author(s) November12, 1996; revised manuscript lueived December 3, 1996; paper8PiXOved for pre-press December 4, 1996; fina18PiXOval April 1,1997 ..&

Xuelong7Jaou obtained a B.Sc. in petrole-um engineering from East Qlina PetroleumInstitute in Shangdon Province in 1983 anda M.Sc. in chemical petroleum engineeringfrom the University of Calgary in 1995.Since dIat time be has worked in the area ofPvf analysis (both experimenlally and the-oretically). Since 1991 he has beeninvolved in research and developmentrelated to fluid phase flow, multi phase~uilibria, and EOS modelling. Xeulong iscurrently a chemical engineer with ~nPipe Industties Inc.

Brant Bennion is currently president ofHycal Energy Research Laboratories Ltd.of Calgary, Alberta. He is involved inresearch in multi-phase flow in porousmedia. formation damage studies, miscibleand thermal EOR. Brant received a B.Sc.from die University of Calgary in 1984 andis a member of APEGGA and a boardmember of the Calgary section of ThePetroleum Society.