ENHANCED OIL RECOVERY recent developments in slim tube testing

If the slim tube is packed with angular grains, the variability of pore and pore throat sizes and shapes simulates reservoir rock but allows each test to be run at a fraction of core flo

ENHANCED OIL RECOVERY

and D.B BENNION Hycal Energy Research Laboratories Ltd

ABSTRACT

Testing of solvent effectiveness related to a given crude oil under

simulated reservoir conditions began in the early 1950s The

in-itial apparatus used a small-diameter stainless steel tube,

capa-ble of sustaining high internal pressures, to contain a porous

medium similar to reservoir rock This "slim tube" was

mount-ed within an oven maintainmount-ed at reservoir temperature The

desired test result was the minimum pressure required to achieve

solvent/oil miscibility, characterized by high oil recovery from

the slim tube

As interest in enhanced oil recovery by the hydrocarbon

mis-cibleflood (HCMF) process has increased, slim tube test

equip-ment and procedures have become more sophisticated The

current state of the technology is discussed, including

labora-tory equipment, instrumentation, operating procedures and

analysis of results

Procedures for analysis and interpretation of slim tube test

results are discussed, leading to improved solvent design

con-cepts Examples are provided to demonstrate the impact of

sol-vent design on the hydrocarbon miscible flood process

Introduction

The object of solvent design is to find the combination of

avail-able NGL and mixing gas streams to give a hydrocarbon

mix-ture whidt demonstrates interaction with the reservoir oil as

good as or exceeding some reference parameters Design

engi-neers seek to find a solvent which will have a reasonable chance

of attaining the desired increase in oil recovery in the reservoir

Enrichment with intermediates must exceed an economically

ac-ceptable safety margin above the minimum miscibility

concen-tration (MMC), which defines the boundary between

multi-contact miscible (MCM) and immiscible perfonnance This

usually implies selecting the solvent analysis corresponding to

one or more percentiles of slim tube oil recovery above the MMC

recovery, or a critical temperature one or more degrees higher

than for MMC This definition requires results from

success-ful slim tube tests with solvents characterized as both

immisci-ble and multi-contact misciimmisci-ble The plot of recovery against

Keywords: Slimtube Solvent flood Hydrocaft)oQ Labocatory tcst

Sol-vent design, Miscible flood Enhanced oil recovery.

solvent mole average critical temperaturetll can then be used

as the basis for solvent selection

Slim tube tests are performed to observe actual solvent-oil interaction in a physical simulation of reservoir pore space If the slim tube is packed with angular grains, the variability of pore and pore throat sizes and shapes simulates reservoir rock but allows each test to be run at a fraction of core flood cost The slim tube can then be considered to represent a single chain

of connected reservoir pores exhibiting realistic solvent displace-ment efficiency Diffusion and convective dispersion effects con-Current with the fluid flow direction are included in the result, but not the tri-dimensional sweep efficiencies characteristic of

a real reservoir element

Data enhancement procedures, verified by improved agree-ment between measured and calculated initial oil density and effluent reservoir volumes, improve comparisons between slim tube runs Interpretion benchmarks can be established to

clari-fy the characterization of each run as MCM or immiscible A set of four key performance correlations for each run show how solvent performance compares to benchmarks, and aid charac-terization of the dominant solvent process

Historical Equipment

Until recently, the slim tube equipment in use varied little from the original designs of the early 19505 The slim tube was usually

13 metres of nominal 6.2 mrn 0.0 stainless steel tubing packed with glass beads or 100 mesh sand and coiled into a tight helix The pore volume was usually 90 to 140 Rml

After pore volume determination, the procedure was to fill the tube with reservoir oil Solvent was then injected from a high pressure container located outSide the oven The effluent ejected from the tube flowed through a capillary sight glass and the back pressure valve, then proceeded outside the oven to a continuously operating ambient separator Periodic samples of separator gas were analyzed to C6 Separator gas and oil volu~es were accumulated and an over-all analysis of each was obtained at the end of the test The slim tube was usually washed with solvent and the residual solvent and oil delermined from the extracted fluid volume and composition to check the mass balance for the run

The data developed by this equipment comprised the "oil recovery" derived from the combined separator liquid volume~ Paper reviewed and accepted for publication by the Editorial Board of the Journal of Canadian Petroleum Technology

FIGURE 1 Experimental slim tube apparatus

and analysis; and the indication of solvent/oil mixing zone

arrival associated with appearance of two phases in the

capil-lary sight glass The information could not be checked to iden-"

tify errors in reponed volumes or analyses, and the indicated

solvent performance could not be critically appraised

mit photographs of the phase condition for each injection step The equipment gives the advantage of steady state observation and quantitative evaluation of phase volumes accumuJated over

a period of time, rather than observation of unstable bubbles passing through a capillary sight glass

The equipment is enclosed in a 0.7 m by 0.7 m by 1.7 m oven, with 5.08 cm thick walls filled with high temperature insula-tion The oven is split into a lower portion containing the high pressure fluid storage cylinders and an upper portion contain-ing the slim tube, effluent accumulators and density measure-ment cell The two compartmeasure-ments are separated by a perforated baffle (10 holes/cm2), and have separate access hatches to per-mit work in the lower chamber during a slim tube run with minimal impact on temperature stability in the upper chamber The front access hatch is equipped with two, 2 cm thick pieces

of explosion-proof tempered glass to permit observation and photography of the ~ cells and their contents without open-ing the oven door The improved oven design allows the temperature-sensitive effluent densitometer to be used in the upper chamber with only minor fluctuations

Figure 2 desaibes the ambient conditions separation system The heart of this apparatus, a gasometer capable of 0.1 ml accuracy, is initially evacuated and purged with helium to elimi-nate unwanted atmospheric residual gas contamination Slim tube effluent is choked to ambient pressure through a needle valve against a baffle plate Oil and water fall into a graduated centrifuge tube and the separator gas enters the gasometer for volume measurement and correction to standard conditions The centrifuge tube is capped; spun at 210 rad/sec for 15 miD; and read to give the separator oil volume Portions of each separa-tor product are transferred by gas and liquid syringes to a chro-matograph for analysis

Improved Equipment

FIgUre I provides a schematic illustration of improved slim tube

equipment used to evaluate solvent-reservoir oil interactions

The slim tube is 18.29mof9.53 mmO.D 316 SS tubing, packed

with 200 mesh angular crushed quartz by mechanical vibration

and coiled into a 35.6 cm diameter tight helix Each end

con-tains 2 mm of fiberglass packing and a 2 micron in-line filter

The pore s~ce, approximately 225 Rml, assures sufficient

volume errors associated with the system piping This represents

an important improvement over the historically used 6.2 mm

O.D by 13 m slim tube, which was much harder to pack and

contained only half the pore volume

Injected fluids are stored in high pressure cylinders, located

in-the oven with the slim tube to ensure thermal equilibrium

in the system and maintained at the upstream entry pressure

fluids in the oven are handled by mercury displacement to avoid

solubility effects related to CO2 and H~ The mercury

injec-tion rate is controlled by highly accurate positive displacement

pumps at 0.4 to 240 mVh up to 70 MPa with 0.01 ml accuracy

This fluid handling equipment gives much better experimental

stabilitY than previous equipment, which had the fluid cylinders

stored outside the oven

Slim tube effluent first passes through a high pressure

dens-ity cell, then is collected inside the oven by water displacement

from two high pressure visual cells Standard operation keeps

one cell on line to the slim tube, while the other is being

emp-tied through the separator and refilled with distilled water The

duration of this cycle defines each solvent "injection step" for

the slim tube run The water displaced passes through a

back-pressure regulating system rated to 70 MPa, accurate to 0.1'1

of the setpoint value and monitored by a high accuracy

record-ing transducer The water volume is measured in a graduated

cylinder and used to confirm the slim tube output related to

each injection step The visual effluent accumulator cells

per-34

Intrumentation

System pressure and pressure differential across the slim tube during a test are measured using test gauges, accurate to SO kPa The gauges are calibrated using a deadweight tester An injec-tion pressure profde provides another generally useful indica-tion of solvent performance, with a pressure drop usually observed at breakthrough

The Journal of Canadian Petroleum Technology

The separator samples are subjected to compositional analysis

using gas chromatographs, with helium carrier gas The

analy-sis processing deletes oxygen and nitrogen associated as air for

the gas samples, and reports components to C6 for the gas

and alkanes to C.,- for the liquid

Another significant improvement in the slim tube apparatus

is a high pressure remote density meaurement system for

con-tinuous measurement of the effluent density at conditions up

to 35 MPa and 175°C This aids in the accurate determination

of the solvent/resen.oir oil mixing zone configuration after 18 m

of transit through the porous medium The densitometer

con-sists of a small stainless steel U tube mounted against a

counter-weight vibration damper The U tube is ultrasonically excited

on a continuous basis, and the resonant frequency of

oscilla-tion is measured The frequency of oscillaoscilla-tion ("T" factor) is

directly related to the density of the material contained in the

U tube With temperature control under normal operating

con-ditions, accuracy is approximately 0.5 kg/m)

The densitometer must be calibrated at the specific test

tem-perature and pressure with standards of known density that

bracket the range of densities expected Ethane and propane

are generally selected as standards, due to the availability of

density-data for these materials over a wide range of

tempera-tures and pressures

Procedure and Data Assembly

The slim tube is initially cleaned with toluene and tri-chloroethane to remove any residual oil and asphaltic heavy ends from the previous run This is followed by methanol and then

a vigorous nitrogen flush at reservoiuemperature and ambient pressure for 24 hours to vaporize residual methanol The slim tube is then evacuated for a 24-hour period Mercury manom-eters, attached to the injection and production ends of the slim tube, indicate when a satisfactory vacuum has been obtained The slim tube is saturated at reservoir temperature and pres-sure with toluene to determine the pore volume If the pore volume has decreased compared to previous measurements, the tube is either recleaned or replaced

Initial Conditions

Recombined oil is then displaced through the toluene-saturated slim tube at reservoir conditions, with effluent being processed through the separator This is a first-contact miscible displace-ment of the toluene from the slim tube, leaving it saturated with reservoir oil The operation is complete when the effluent gas/oil ratio is equal to the solution gas/oil ratio of the initial recom-bined oil and the liquid phase analysis is comparable to the stock tank oil analysis from the single-stage flash test

The slim tube run commences by the injection of an addi-tional 0.20 pore volumes of pure live oil through the slim tube and densitometer Effluent is separated and collected in 0.10 pore volume increments, with gas and liquid volumes and analyses reported This portion of the test provides an analyti-cal benchmark for the initial oil composition

Slim Tube Operation

Salven injection then begins, continuing until 1.2 PV has en-tered the tube Normal fluid injection rates are 0.1 to 0.15

PV /hr, with effluent accu~ulated in 0.1 PV increments for separation and analysis Effluent density is logged at least In times during each increment The effluent accumulator is pho-tographed at the end of each increment Upstream and differen-tial pressures across the slim tube are recorded throughout each run

The data available for each sample interval of a slim tube run includes separator gas/liquid pair volumes and analyses; the fraction of the upper phase in the effluent accumulator from photographs; and the effluent accumulator water withdrawals

In addition, the raw data set includes the measured slim tube pore volume; the solvent analysis and measured density; and the measured density, sOlution GOR and FVF for the reservoir

Fluid Tested

Reservoir oil is obtained by recombimng separator gas and liquid

samples at reservoir conditions to yield the desired bubble point

pressure and solution gas/oil ratio A pressure/volume test is

conducted on the live oil to check the bubble point A

single-stage flash test is conducted to determine the solution gas/oil

ratio of the oil, and to provide the formation volume factor

for use in later recovery calculations

Figure 3 provides an illustration of the equipment used to

synthesIze hydrocarbon solvents from samples of the NGL mix

and mixing gas available at the reservoir under study Mixing

gases are transferred by cryogenic condensation into a high

pres-sure storage cylinder and heated and prespres-sured to test

condi-tions Heavier components (C5 to C.O> are added as liquids

under pressure- The solvent analysis is then checked and

densi-ty measured with the effluent densitometer A pressure/volume

test is also conducted on the solvent to ensure that it is single

phase at test conditions

TABLE 2 Example mass balance - unadjusted

interval 0.3 to 0.4 PV nominal injection

TABLE 1 Example slim tube raw data set

interval 0.3 to 0.4 PV nominal injection

CUM

rOTAL 1.0000 180.33 21.87 22.14 0.13

-888.8/377.3 IMea.uredl

COMP MOLE FR MOLE FR MOLE FR GRAMS

RECOMB

COMP MOLEFR GRAMS

GRAMS GRAMS

H,s

co

N,

C

c

c

IC,

nC,

'c

nc

C

c

c

c

c

Co

0.0000

0.0131

0.0243

0.5427

0.1832

0.1523

0.0181

0.0578

0.0116

0.0116

0.0024

0.0013

0.0003

0.0002

0.0000

0.0000

0.0000 0.0000 0.0000 0.0232 0.0171 0.01&0 0.0051 0.0171 0.0158 0.0157 0.0701 0.1021 0.012.

0.073&

0.0&05 0.4&87

0.0000 0.0185 0.0328 0.~05 0.0844 0.1483 0.0000 0.0866 0.0000 0.0485 0.0144 0.0172 0.0000 0.0000 0.0000 0.0000

TOTAL 1.0000 1.0000 1.0000

kg/m3

kg/Rm3

nil

Em ACCUMULATOR FRACTION UPPER PHASE

0.00 0.07 0.08 1.01 0.84 0.88 0.17 0.51 0.22 0.22 0.64 1.05 1.07 0.16 0.87 13.20 21.67

0.00 0.17 0.18 1.50 0.48 1.14 0.00 0.88 0.00 0.58 0.21 0.30 0.00 0.00 0.00 0.00

0.00 0.18 0.18

~.44 0.48 1.11 0.00 0.88 0.00 0.58 0.21 0.30 0.00 0.00 0.00 0.00

712.1

888.1'377.3 30.11

oil An example raw data set over the interval 0.3 to 0.4 PV total

injection (0.] to 0.2 PV solvent injection) is shown in Table I

The toluene produced from the cleaning of the slim tube can

also be subjected to composition analysis to determine the

volume of remaining residual oil at the conclusion of a run

This data can be used together with the recovery data from the

test to close the material balance on the system This data was

not available for the example slim tube run

Data Interpretation

The gas and liquid analyses can be used to define the

cor-responding densities at standard conditions using the molal

volume procedurefZ) and GPAO) pure component properties

The analyses may also be converted to mass fraction, and the

molecular mass of each miXture calculated Combination of the

densities, volumes and mass fraction analyses then yields a

com-pilation of the mass in, out and retained in the slim tube for

each component at the end of each 0.] PV increment, as shown

in Table 1 Trial I calallation results for the example data are

summarlzd in Table 2

The initial mass in place for each component can be defmed

by averaging the results for the flfSt 0.2 PV of the displacement,

when only pure reservoir oil was displaced from the slim tube

The reduction in octanes plus content in the slim tube, or «oil

recovery", can then be calculated for each sample interval

dur-ing the solvent displacement This calculation eliminates the need

to correct the oil recovery for "breakthrough solvent" because

the solvent contains no significant octanes plus concentration

Exposure of Inconsistencies

The mass balance results permit critical review of the raw data

to ensure the final interpretation of solvent performance is on

a sound foundation For example, the initial mass in place for

the reservoir oil used in the calculations contains a fIXed

frac-tions of octanes plus The octanes plus mass reduction during

each ]0 PV increment should thus be 10'11 Any discrepancy

indicates inconsistencies in the raw data provided, particularly

36

with respect to the reported injection The total pore volume can then be used to convert the octanes plus mass reduction

to reservoir volume, which must compare favourably with the volume of effluent accumulator water ejected Application of these calculations to the example data yielded 22.49 Rml for 100/0 PV, compared to 30.99 Rml for effluent accumulator water

coUected and 31.07 Rml for resevoir oil withdrawals based on octanes plus depletion This comparison suggests the actual in-jection interval to be great-er than 1007 PV

As an additional check, the initial oil density can be calcu-lated from the produced oil and gas volumes, their respective calculated molal average densities and the measured FVF If this result is significantly higher than the measured effluent den-sity for the first 0.2 PV (oil displacing oil), the separator oil analysis is suggested to be deficient in light components In ad-dition, the measured density for the 0.2 to 0.3 PV injection step should not be significantly lower than measured for the first 0.2 PV, unless breakthrough of the solvent leading edge at the downstream end of the slim tube is clearly indicated from the accilmulator phase behaviour or densitometer data Such defi-ciencies again suggest loss of light components from the sepa-rator liquid during handling

As previously described, the slim tube f1lling process involved displacement of dead oil with live reservoir oil This operation may not have been complete at 0.2 PV total injection To off-set this problem, the fluid analyses for the 0.2 to 0.3 PV inter-val can be used as the original reservoir oil, providing there is

no solvent breakthrough indication by the densitometer data The mass balance calculations can then be rerun to repeat the test described above This redefinition of the initial reservoir oil analysis can be better than the initial trial, but may stil1 be deficient in the same ways For the example data in Table 3, the Trial 2 redeftnition of original reservoir oil for the example run reduced the initial mass in place by 2'1 but did not affect the octanes plus "fraction The octanes plus depletion was still 14'1 for the first 10'" PV increment, indicating more oil production than solvent injection

Component Retention

The material balance results also make possible the assessment

The Journal of Canadian Petroleum Technology

TABLE 3 Example mass balance - unadjusted

interval 0.3 to 0.4 PV nominal, trial 2

TABLE 4 Example adjustment factors - trial 2

RESERVOIR OIL SOLVENT UNITS CUM

CUM XS SOLVENT

RECOMB

COMP MOLE FR GRAMS

NItrogen/Methene Retlo Carbon Dioxide/Ethane Ratio Hydrogen Sulfide/Propane Ratio 'aobutane/Nonnal Butane Ratio laopentene/Normel Pentane Ratio Separator Uquld

0.0702 0.1171 0.0000 0.3243 0.8870 22.7 18.28

2511.3 3.15

0.0474

0.124' 0.3881 0.0000 0.0000 0.0000

gIg gIg gIg gig gIg ml g

ml g

H,S 0.0000 0.00 0.00 0.00 0.00 100.0

CO, 0.0074 0.55 0.07 0.08 0.00 100.0

N 0.0124 0.58 0.08 0.08 0.00 100.0

C 0.3118 8.30 1.01 1.18 0.00 100.0

C, 0.0887 4.82 0.84 0.88 0.00 100.0

C, 0.1018 7.45 0.88 1.04 0.00 100.0

IC 0.0128 1.23 0.17 0.17 0.00 100.0

"C 0.0388 3.12 0.51 0.53 0.00 100.0

IC, 0.0128 1.64 0.22 0.22 0.00 100.0

"C, 0.0133 1.68 0.22 0.22 0.00 100.0

Co 0.0321 4.&8 0.84 0.84 0.00 100.0

C, 0.0453 7.68 1.05 1.05 0.00 100.0

C 0.0382 7.44 1.07 1.04 0.00 100.0

Co 0.0277 8.64 0.17 0.82 0.00 100.0

c 0.2130 84.30 13.20 13.20 0.00 100.0

TOTAL 1.0000 157.18 21.87 21.88 0.00

OCT + 0.3118 115.08 18.10 18.11 0.00

SOL GOR 108.0 (M Ul.dl

FVF 1.347 (M ur.dl

DENSITY kg/Rm3 e88.8 (Calculated}

88e.8/877.8 (Mea.ured)

Exceaa Total Wlthdrewala rru on

Separator ae

Maaa fraction Propane

a non-steady-state compressible displacement Investigation sub-sequently revealed the poSSlole source for this to be initial in-equity between the solvent container pressure and the slim tube entry pressure, which could cause injection volumes different from those indicated by Ruska pump reading These could, in turn, cause increases in produced volume as the back pressure regulator opened to prevent pressure increase in the slim tube

Data Enhancement

The purpose for the following activities is to -establish confi-dence in each set of slim tube data by examining the internal consistency of the various performance parameters recorded; and to ensure that companion slim tube runs can be compared

on a common basis

Comparison of slim tube runs is complicated by the physical constraints within laboratory equipment, and unavoidable var-iances in operating procedure Review suggests the following sources for these problems: sample contamination by air and/or toluene in equipment;

loss of carbon dioxide to distilled water used for fluid handling;

loss of hydrogen sulphide to distilled water and/or by diffu-sion through steel tubing;

non-standard identification of iso- and nonnal- butanes and pentanes during GC analysis integration;

unreported or lost light hydrocarbons resulting from separa-tor liquid handling and/or air extraction while centrifuging

or pipette sampling; and

in-equities at the start of injection

The intended benefits from elimination of the impact of these problems is avoidance of erroneous conclusions about the rela-tive reservoir risk for the solvents being compared

Lost Light Components

The equilibrium constant values<3> for light components at separator conditions suggest their concentrations relative to ead1 other in the separator gas and recombined oil to be essentially the same The adjustment of the reported separator gas and liquid analyses, therefore, begins with establishment of five parameters for the initial separator gas and solvent mass frac-tion analyses: nitrogen/methane ratio, gig

carbon dioxide/ethane ratio, gig hydrogen sulphide/propane ratio, gig iso-butane/normal butane ratio, gig

iso-pentane/normal pentane ratio, gig

Values for the example data set are shown in Table 4 The fraction solvent contained in ead1 produced effluent sample can

be estimated from the mass fraction propane in the solvent and reservoir oil Weighted average factors can then be applied to the combined effluent analysis to give a first approxiniation ad-justment

The adjusted effluent is then flashed to separator conditions using the near-linear correlation of the log of equilibrium

cons-of retention for each solvent component as the flood

progress-es Expected "true oil" component effluent mass produced can

be dermed from the initial oil analysis and the effluent mass

of the heaviest component Production in excess of this "amount

can then be deducted from the injected mass, and the

indicat-ed net accumulation expressindicat-ed as a percent of the mass

inject-ed The cumulative form of this calculation was found to be

most useful in comparing slim tube runs

The example calculations in Table 3 indicate that solvent

breakthrough has not yet occurred at 0.4 PV total injection

This result is in direct disagreement with effluent density, which

abruptly declines from 686.9 kg/RM3 to 377.3 kg/RM3 when

the interval is 70'1 complete The comparison of calculated true

oil to total withdrawals gives general indication that the octanes

plus mole fraction reported in the liquid is too high The most

likely reason is loss of light ends while handling the separator

liquid in equilibrium with air

Volume Balance

A reservoir voidage version of the mass balance can also be

cal-culated for each 10 PV increment Expected "true oil"

production for each component can again be calculated from

the effluent mass for the heaviest component, the initial

sepa-rator gas and liquid analyses and the solution GOR from the

recombination Production by component in excess of the "true

oil" expected can be derived to make up an excess fluid

analy-sis The excess fluid density at reservoir conditions can then be

estimated by calculations<2l, yielding the excess fluid FVF The

excess fluid reservoir volume can be calculated and added to

the "true oil" reservoir volume for direct comparison to the

volume of effluent accumulator water ejected The

compari-son between the calculated excess fluid cut in the effluent

accumulator and the upper phase volume fraction is also

in-teresting, indicating whether the excess fluid is condensed in

the reservoir oil or is an extraneous, vaporizing phase Such

calculations based on the data in Table 3 indicated no excess

produced volume

The divergence of reported PVthroughput (22.49 ml) from

measured (30.99 ml) and calculated (31.48 ml) strongly suggests

November-December 1988, Volume 27, No.6

TABLE 7 Example volume balance - adjusted interval 0.3 to 0.4 PV nominal, trial 2

TABLE 5 Example adjusted data set

-interval 0.3 to 0.4 PV nominal, trial 2

GRAMS GRAMS

COMP MOlE FR MOLE FA MOlE FR GRAMS

Tabla 6 Tabla 6

21.63 0

887.7 ko/Rm3 31.45 Rml

1.347 Rmi/mi 23.35 ml 22.70 mI 0.00 nol 108.0 mUmi 2545.1 mI 2511.3 nol 0.00 mi' 31.45 Rnol 224.8 Rml 0.1388 PY 0.5178 PY 0.1378 PV 0.5080 PY

True Oil Wltlldrewa'8 True Oil Oen81ty True Oil Wltlldrewa'8 FVF

0.00 0.23 0.28 2.04 0.83 1.57 0.00 1.22 0.00 0.80 0.30 0.41 0.00 0.00 0.00 0.00

H.S 0.0000 0.0000 0.0000

co, 0.014 0.0003 0.0185

N 0.0242 0.0001 0.032.

C, 0.1808 0.010 0.1401

Co 0.1703 0.01.1 0.0844

Co 0.111 0.0448 0.1483

IC 0.0112 0.0144 0.0000

ftC 0.0348 0.0445 0.08

IC, 0.0032 0.02$8 0.0000

ftCo 0.00$3 0.024$ 0.0485

Co 0.0038 0.0842 0.0144

C 0.0017 0.0841 0.0172

C 0.0005 0.085$ 0.0000

Co 0.0001 0.0884 0.0000

CM 0.0000 0.05.0 0.0000

C 0.0000 0.4121 0.0000

-TOTAl 1.0000 1.0000 1.0000 21.41

YOl 1 2511.$ 22.1 $1.0$

DENS 1.2 804.$

kg/m3

MOLE 28." 171.14 $1.30

MASS.kG

YEAS 0EN8fTY 241.$

kg/Rm3

EI'R WAnR

nil

EI'R ~TOR FRACT1OM UPPeR PHASE

0.00 0.08 0.0.

1.11 0.87 1.02 0.17 0.62 0.21 0.22 0.8S 1.02 1.05 0.13 0.85 12.8S

0.00 O.2S 0.28 1.08 0.8S 1.58 0.00 1.22 0.00 0.10 O.SO 0.41 0.00 0.00 0.00 0.00 7.48

Tabla I

T.~I 5

Tabla 6

Tabla 5

7.48

True Seperator OU Reported Saparator 011 EKee.e Separator 011 Solution GOR True Solution Ga.

Reported Saparator a

EKea Saparator G

Total Withdrawal.

Slim Tuba PV Tota' Wlthdr.wal.

Cum Wlt"'awal.

Reported Total Wlthdr.wal.

Cum Reportad Wlthdr.wal.

"dlu.ted InJection Int.rval

Tabla 5

887.7

111.1'377.3

30.'1 0.0

TABLE 6 Example mass balance - adjusted

Interval 0.3 to 0.4 PV nominal, trial 2

CUM TOTAL TRUE CUM XS SOLveNT WITHO OIL WITHO RET GRAMS GRAMS GRAMS %

RECOMB

COUP MOLE FA GRAMS

H.8 0.0000 0.00 0.00 100.0

co 0.0078 0.57 0.01 100.0

He 0.0124 0.&1 0.01 100.0

c 0.1074 8." 1.1& 88.3

c 0."14 4.11 0.81 100.0

IC 0.0127 1.21 0.17 100.0

nc 0.0182 S.74 0.52 100.0

IC, 0.0121 1.52 0.21 100.0

nc 0.0132 1.M 0.22 100.0

c 0.0321 4.&3 0.83 100.0

c 0.0471 1.42 1.02 100.0

c 0.0417 1.32 1.05 100.0

c 0.0131 8.11 0.83 100.0

Co 0.0283 8.45 0.15 100.0

c 0.2125 82.82 12.13 100.0

TOTAL 1.0000 154.81 21.41

OCT + 0.3173 113.38 15.88

SOL GOR 108.0 CM.a.Uf.dl

FVF 1.347 CM.a.uredl

DENSITY kO/Rln3 811.1CC8~at.dl

818.8/111.ICM.a.ufedl

True OIC WIthdrawal - 11z.ISI82.12HCompon.nt initial 011 In Plac.1

Exce - Total Wlthdrawa" - True 01

0.00 0.01 0.01 1.12 0.88 1.00 0.17 0.52 0.21 0.22 0.83 1.02 1.01 0.83 0.1' 12.13 21.83 15.88

0.00 0.00 0.00 0.03 0.00 0.02 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.05 0.00

tant (K value) with the fourth power of critical temperature,

as shown in Figure 4 Upward or downward adjustment of the

equilibrium constants caused by temperature or pressure

fluc-tuations can be simulated by multiplying the entire K value set

38

by a factor Iterative calculations find the required multiplying factor to provide a mass balance match within I'" error and

a match on the reported vapour/liquid ratio within S" error The resulting separator liquid can thus be made to contain light component mass not previously reponed, but in better

equilibri-um agreement to the content of the separator gas As a result, the mole fractions of the heavier components in the r~mbined effluent will be reduced

Revised Balances

The adjusted separator gas and liquid analyses and reponed volumes are then used to rerun the mass balance calculations desaibed previously The ratio of the unadjusted to adjusted mass of the heaviest component can be used to revise the

repon-ed separator liquid volumes, so that the reinstatement of lost light components does not affect the mass of the heavier com-ponents produced Adjustment calculations for the example produced effluent (fable 5), show an increase of 0.2 ml in the reponed separator liquid volume for the example increment, a(XX)mpanied by redistribution of the component mole fractions

in the produced fluids

The mass and volume balance calculations can then be re-run, using adjusted separator gas and liquid analyses and ad-justed volumes Two trials are required, the first to confarm the agI ~.ait between the calaIlated reservoir withdrawals and the volume of effluent acxumulator water injected The real in-jection increment can then be deflDed, as shown in Tables 6 and 7, together with "true" incremental solvent injection

Interpretation

Comparison of data enhancement results to the raw data often shows that the adjustments increase the intermediates (ethane-pentane) concentration in the separator liquid This, in turn, redefines the compositions of the initial oil-in-place and the effluent from each injection step Although the total mass produced is usually not affected, the withdrawal of propane and lighter is increased The adjustment usually improves the comparison between "true oil" and effluent withdrawals for the lighter components as shown by comparing Tables 2 and

6 The addition of light ends indicates excess effluent mass/ volume in agreement with the onset of density decline and ap-pearance or an upper phase volume (I") in the effluent ao:umu-lator, as shown in Table 6

Reinstatement of the light components and use of the oil

com-The Journal of Canadian Petroleum Technology

~

(/)

z

0

~

::>

~

CD

:J

5

0

0

5

~

0

Q,

~

FIGURE 4 CorftlatiOD of equDlbrfum COOStaDt to critical temperatures at staDdard coDditioos.

~

100

c

~

~

~

t-:I:

<-'

W

:s::

~

=>

U

LL-0

~

c

w

Z

~

w

~

j

SOLVENT

~ r;

fOOO ~ U)

~

t-O ~ 3 15

L~

~

71 % AT 1.2 PV SOLVENT

60

LEGEND

IDEAL

m

dj :

ACTUAL

DENSITY

40

2

TOTAL HPV INJECTED, FR

position produced during the first solvent injection increment

to deriDe the initial reservoir oil significantly improves the

com-parison between calculated and measured densities in Table 6

compared to Tables 2 and 3 This agreement lends support to

revision of the pore volume scale, improving the comparisons

interpre-tation of the solvent performance For the example the data

enhancement procedures indicated the pore volume inj~OD scale to be in error by 0.125 PV by the end of the test, reducing the fma11y accepted octanes plus recovery by 3 percentiles

~

rd 100

~~, " '

' ,',','

"" : ' '.:'" .

W '

<-' '

'-,WO' ,

'I~~ , ' ,-,u'

:Z::>"

-c( o ',-'

WI}:

" '", " " ,"

f " ", '" "

" " ,

" ", '" ,'.~

" " " ,.

","'.'

:J.' ", '" ,'.'

-,.,.,.", ';~

a

~ 90

I.aJ

,

6

~

<-'

W 80

~

~

=>

0 7°i'" ~

~

a

I.aJ

z

~

~ 60

~o

~z

SOLVENT

1000

0

~~~

68 % AT 1.2 PV SOLVENT

1\

LEGEND IDEAl

ACTUAL

DENSITY

1.8-so'

FIGURE 6 JFCM solveat performaace - cumulative etbaae mentioa

2

100

80 % AT 1.2 PV SOLVENT

~~

tt}

~

d-~

~

1000 t= II) Z I.J 0 I-I.J

0 ~ I.J

73 % AT t2 PV SOLVENT

~~

LEGEND

IDEAl

a

w

~

~

~

t-:I:

c.:>

i;j

~

:i

U

Lo-0

~

a

w

z

~

W

Q:

ACTUAL

DENSITY 1.8

TOTAL HPV INJECTED, FR

u

m

ItJ

b

~<>

2 FIGURE 7 JFCM solvent perfonDance - cumulative propane retention

Process Interpretation

The confid~ resulting from the data enhancement pr~ures

leads to the following definitions intended to establish the

ba-sis for design of acceptable cost-effective solvents from slim

tube data

Key Performance Correlations

The interpreted slim tube results yield a large potential array

40

of perfonnance correlations with pore volume solvent injected: a) separator gas and oil mole fraction of each component or

of component groups; b) recombined effluent mole fraction of each component or of component groups; c) component reten-tion in the slim tube as per cent of cumulative mass injected; d) recovery factor for R~ wmpoocot or for component groups; and e) effluent density measured at the slim tube outlet Because there are 16 components and 9 or more meaningful

The Journal of Canadian Petroleum Technology

I')

~

~

<.:>

~

~

~ 0

I-~ 3 I::

~

ffiJ :0

~ l.a

to,

80

70

60

It- ~1~O % AT ~.2 PV

SOLVENT 98.7 % AT 1.2 PV SOLVENT

~ DEFlECTION AT i

MIXING ZONE ~

BREAKTHROUGH +

::~,~~:

" w ,<-'

"oZ "

~ )WO"

','

~-J

~

~

1000 ~ Iii

~

w

~

~

~

<-'

~

~

J

<

E

~

~

~

f;3

a:

LEGEND

IDEAL

f::;;1

~

b

ACTUAL

1.8

0 11:2" O.~ 0.6 0.8 1 1.2

TOTAL HPV INJECTED, FR

nGURE I JFCM SOlyeDt performance - camuJati.e odaDes plus recG'YKJ

.-,.-.

100

Q

g

~

.-:I:

0

~

~

=>

u

~

0

~

8

z

~

w

~

SOLVENT

65 % AT 1.2 PV SOLVDIT

LEGEND

IDEAL

ACTUAL

r::m Iji:I

!

DENSITY 40

TOTAL HPV INJECTED, FR

COmponent groups, each slim tube run analysis could generate

Over 100 correlation plots Many of the potential alternatives

are redundant

The cumulative solvent retention plots give meaningful,

fOCUssed information about the process Solvent breakthrough

can be easily identified because retention begins to decline below

100'" at that point Arrival of the true mixing zone at the slim

tube outlet is usually associated with the beginning of

straight-line, stable decline of the cumulative solvent retention

Reten-tion at 1.4 PV total injecReten-tion (1.2 PV solvent injecReten-tion) gives

a measure of the relative values of solvent components in dis-placing slim tube oU These properties of the correlation can

be compared directly with the effluent density correlation and the phase behaviour in the effluent accumulators for confIr-mation The most meaningful correlations are those for methane, ethane, carbon dioxide and nitrogen as a group ("ethane and lighter"), ethane and propane

The interpretation of recovery correlations is much more

80

60

40

20

1000 ~

If) Z IoJ 0

£P

~

""'.,

J 0"

0:'

~'~a :'

, ,

~

~~

>=

I-ii\~ z

\AI a

I-z

\AI

1000 3

~

\AI

"

~,.I

~

a

i&J

~

,

~

.-:I:

w

3::

~

=>

U

L

0

~

a

i&J

z

~

i&J

Q:

TOTAL HPV INJECTED, FR cumulative etbane meatio

1

FIGURE 10 JMCM solvent performance

0

fusing, because the recovery is always negative for solvent

com-ponents, becomes increasingly more negative until

breakthrough, changes variably during leading edge production

and increases with the arrival of the real mixing zone For

com-ponents with low concentrations in the solvent, the correlations

can even be horizontal, indicating little or no component

rCt;Ov-ery because production is being replaced by injection The most

meaningful correlation is the one for the octanes plus group,

which contains no solvent components and can indicate oil recovery without the need for breakthrough solvent corrections

Ideal FCM Performance

The characteristics of the key correlations can be used to iden-tify solvent-reservoir oil 5Y5tCln$ which perform as fir5t con-tact miscible (FCM) The effluent density plot should be horizontal up to 0.8 PV solvent injection then decline as an

90

80

70

60