SPE

SPE 20776

Comparison and Analysis of Radioactive Tracer Injection Response With

Chemical Water Analysis Into the Ekofisk Formation Pilot Waterflood
O.B. Skilbrei, L.D. Hallenbeck, and J.E. Sylte, Phillips Petroleum Co. Norway
SPE Members

Copyright 1990, Society of Petroleum Engineers Inc.

This paper was prepared for presentation at the 65th Annual Technical Conference and Exhibition of the Society of Petroleum Engineers held in New Orleans, LA, September 23-26, 1990.

This paper was selected for presentation by an SPE Program Committee following review of information contained in an abstract submitted by the author(S). Contents of the paper,
as presented, have not been reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material, as presented, does not necessarily reflect
any position of the Society of Petroleum Engineers, its officers, or members. Papers presented at SPE meetings are subject to publication review by Editorial Committees of the Society
of Petroleum Engineers. Permission to copy is restricted to an abstract of not more than 300 words. Illustrations may not be copied. The abstract should contain conspicuous acknowledgment
of where and by whom the paper is presented. Write Publications Manager, SPE, P.O. Box 833836, Richardson, TX 75083-3836. Telex, 730989 SPEDAl.

ABSTRACT INTROOUCTION

This paper documents the design, operation, analysis and
use of two different radioactive tracers, tritiLITI and
iodide-125, which were injected in the Ekofisk Formation
Pilot Waterflood Project. This paper also describes how
chemical analysis of water produced in the offset producers
during the pilot project was successfully analyzed,
interpreted, and used as a second tracer method.
Both tracer methods were used to evaluate waterflood
mechanisms and were invaluable to the overall interpreta-
tion of the Ekofisk Formation Pilot. A comparison of both
methods is made to demonstrate thei r repeatabi l i ty for
possible future use.
During the pilot water injection program, clear responses
from the injected radioactive tracers were identified from
the three production wells within the pilot. Analysis of
the produced water from these wells was carried out in the
laboratories of IFE ("Institutt for Energiteknikk") in Oslo
with an accuracy within 1-2 percent.
The Ekofisk Field is located in the Norwegian Sector of the
North Sea and is composed of two naturally fractured chalk
formations, the Ekofisk and the Tor. This essentially
volumetric, solution gas drive reservoir was initially
under-saturated, with an initial pressure of 7120 psi and
a bubble point pressure of 5545 psi at 268 degrees Fahren-
heit. Initial solution GOR at producing separator condi-
t ions was 1530 SCF/STB and ini t i a l oi l gravi ty was 33
degrees API.
Production from Ekofisk was commenced in 1971 through four
subsea producers and switched to three permanent production
platforms in 1975. Production peaked in 1976 at over
300,000 BOPD and declined to below 75,000 BOPD by mid 1986.
Through waterflooding, the oil production has increased to
the current level of 130,000 BOPD. Reservoir pressure
declined through the bubble point pressure of 5545 psi in
1978 after which producing GOR from the field increased and
peaked at 9000 SCF/STB in 1986. Through 1989, Ekofisk had
produced 1,362 million barrels of oil equivalents.

Chemical composition of the produced water was monitored
during the same tracer injection period which showed a
direct correlation. The large differences in sulfate
concentration between Ekofisk formation water and the
injected seawater makes this ion a good alterative to the
more ideal radioactive tritium tracer.
Careful analysis of the various tracers in the Ekofisk
Formation Pi lot confirmed that both waterflood recovery
mechanisms capillary imbibition and viscous displacement
were contributing to the overall waterflood recovery. This
was a key element in the interpretation and simulation of
the pilot.
Waterflooding was first initiated in Ekofisk Field in
November 1987 and was restricted to t~e Tor formation in
the northern two thirds of the field. This was preceded
by extensive study of laboratory imbibition experiments and
a pilot waterflood project to confirm laboratory results
and to gain experifnce from water injection into the highly
fractured chalks.
Initial laboratory data had indicated water injection into
the Ekofisk formation was of l~er potential. In June 1986
a Lower Ekofisk Formation Pilot was initiated to determine
waterflooding potential of that f9rmation. The pilot was
configured similar to the Tor Pilot with injection through
Well B-16 situated in the middle of a triangle formed by
the three producers Well B-19, Well B-22 and Well B-24
(Figure 1).

References and illustrations at end of paper

559
 COMPARISON AND ANALYSIS OF RADIOACTIVE TRACER INJECTION RESPONSE ~ITH
2 CHEMICAL ~ATER ANALYSIS INTO THE EKOFISK FORMATION PILOT ~ATERFLOOD
The operational aspects and the main results of the pilot
project have been thoroughly presented in the SPE paper No.
19838 "Ekofisk Formation Pilot ~aterflood", September 1988.
Since that time, the level of information on the tracer and
chemical results has increased to a point where it is
natural to present a final report on also these aspects.
PART ONE: RADIOACTIVE TRACERS
TRACER DESCRIPTION
Tritium, 3H, is an isotope of hydrogen that has three
protons instead of one. It is radioactive, with a half-life
of 12.26 years, emitting low energy Beta particles. ~hen a
tritium atom is substituted for one hydrogen atom in the
water molecule, a tritiated water molecule, HTO is formed.
Small amounts of HTO is then added to ordinary water, and
a tritiated water mix is obtained. The radioactivity of
this tritiated water then depends on the concentration of
HTO in the water.
Tritiated water is nearly identical to ordinary water, with
no significant differences in chemical properties. This
makes it an ideal water tracer, since selective absorption
and adsorption effects within the reservoir can be ex-
cluded, and there are no precipitation or dissipation
problems that other chemical compounds may exhibit.
Tritiated water mixes readily with the formation water, and
the produced fractions from each origin can be determined
from the produced tritiated water concentrations.
125 1 is a radioactive isotope of iodine. It emits weak
Gamma rays, and has a half-life of 60 days. The tracer was
prepared by dissolving sodium iodide, Nal, where the iodide
is the radioactive isotope, into water. The high solubility
of the isotope in water, and the abundance of iodide and
other similar halogens in the formation water serves to
insure that most of the radioactive iodide ion remains in
the water during the flow through the reservoir.
TRACER INJECTION OPERATION
Two different tracer methods were used, a continuous
injection of the long lived tritiated water, and a slug
injection of the short lived iodide tracer. Table 1
summarizes the main events of the pilot, including tracer
injection.
Continuous injection of tritiated water in ~ell B-16
started in August 1986. By that time a total of 527,000
barrels of water had been injected from the start of
injection in JUly. The amount of HTO tracer added gave the
injection water a radioactive concentration of 2235 Bq/l
through the whole tracer injection period. This corre-
sponded to a net radiation level of 309 CPM (counts per
minute) from a 1/2 liter sample of water, measured in a
liquid scinti llator with 28% efficiency. The tracer was
continuously injected with ~ell B-16 injection water until
the tracer injection was permanently terminated on November
10, 1988. A total of 4.5 Terra8q (4.5*10'2) HTO was
injected with 12.8 MMbbl water. An additional 4.3 MMbbl was
injected following termination of the tracer program.
A slug of 24.4 Giga8q (24.4*10 9 ) 125 1, was injected on
March 25, 1987. The tracer was mixed with 40 liters of
water and injected during a 2 hour period. Cumulative water
injected prior to this slug injection was 4,814,000
barrels.
Samples of produced water from the three producers were
obtained on a weekly basis throughout the tracer injection
560
SPE 20776
period. These samples were analyzed for each particular
tracer and results closely monitored. (Figure 4 and 6)
Initial 'sampling was complicated by the low water produc-
tion rates from the wells and the high reservoir tempera-
ture (268 degrees Fahrenheit). Great care was taken during
the sampling process not to introduce the three producing
wells into the test separator at similar times in order to
reduce the risk of contamination.
TRITIUM TRACER RESPONSE
The tritiated water tracer response is directly related to
4 specific pilot operational phases:
Phase 1: Continuous injection of tritiated water from
August 1986 through July 1987.
Phase 2: Shutdown of injection from August 1987 through
December 1987.
Phase 3: Restart of injection and a recompletion and
restim~lation program of the three production
wells. This included perforating additional
layers.
Phase 4: Terminating tracer injection from November
1988, but maintaining water injection.
Phase 1
~ell B-22: Tritiated water was first observed in October
1986, approximately 50 days after it was first injected in
~ell B-16. At this time a total of 1.5 million barrels of
water had been injected, and ~ell 8-22 was producing with
a 3% watercut. The produced tracer concentration increased
steadil y to 1382 Bq/ l by the end of phase 1 and tota l
watercut had increased to 20%. The tracer concentration
translated to a ratio of 62% injected water versus total
produced water.
~ell 8-19: Tritiated water was first observed in March
1987, 200 days after the start of tracer injection. But
compared with ~ell 8-22, both tracer concentrations and
water production was significantly smaller. By the end of
July 1987 the HTO concentration had increased to 467 Bq/l
( 21% of injected concentration) at a watercut of only 1-
2%.
~ell 8-24: No water production or tritiated water observa-
tions.
Phase 2
~ell B-22: During phase 2, the produced watercut decreased
from 20 to 4 percent, and the tracer concentrations
decreased from 62 to 44 percent. This falloff in tracer
concentration was most probably related to a change in
drainage area of the well during the shut-in of injection.
Initial water breakthrough in this well was believed to be
due to a small high permeability layer directly communica-
ting with the injector ~ell 8-16. ~hen injection was shut-
in this layer stopped contributing injected tracer direct-
ly.
~ell B-19: The watercut dropped to zero, while measurable
tracer concentrations dropped to 10% of the injected con-
centration.
~ell B-24: No response was observed.
SPE 20776 O. SKILBREI L.D. HALLENBECK AND J.E. SYLTE
Phase 3
Yell B-22: The watercut increased, and stabilized at 33%
during this period. At the same time the tracer concentra-
tion began to increase again, and levelled off at about
56%. The tracer concentrations never reached the level seen
during phase 1.
Yell B-19: The watercut steadi ly increased during this
period, and reached 6 percent in October 1988. SillUlta-
neously, the tracer concentrations increased from 28% in
March 1988 to a maximum of 41% in October 1988.
Yell B-24: Yhen Well B-24 started to produce water
(watercut exceeded 10% in August 88), the tracer con-
centration rapidly reached 25% of the injected, and
increased to 35% by "the end of the tritiated water injec-
tion. The increased watercut in all three wells seems to be
mostly associated with the February 1988 restillUlations-
/recompletions, in which additional producing layers were
added and existing layers restillUlated.
Phase 4
Yell B-22: After the tritiated water injection was termina-
ted in November 1988, tritiated water concentrations
decreased from 55 to 40 percent by October 1989.
Yell B-19: A substantial increase in watercut occurred in
the beginning of 1989, when the watercut rapidly increased
from 5 to 15 percent. During the period of increasing
watercut, the tracer concentration decreased from 40 to 20
percent. As the tritiated water injection was now termina-
ted, this behavior was considered indicative of additional
breakthrough of injected water from a high conductive layer
or fracture system.
In apparent contradiction to the above observations, there
were also periods of increased tritiated water concentra-
tions in this phase, as seen in the second part of Figure
5. This behavior will be discussed in later sections, but
was directly connected to periods of lower produced
watercuts, caused by periodically decreased average daily
injection rates (Figure 2).
Yell B-24: The produced tracer concentrations has been
remarkably stable following the tracer injection termina-
tion, and by the end of the pilot, only small reductions in
tritiated water concentrations were observed.
TRlTllit TRACER INTERPRETATION
By November 1988, the pilot wells had produced a total of
approximately 560,000 barrels of water. This volume can be
split into tritiated injected water and non-tritiated
formation water by IlUltiplying the tracer concentration
with the water rate. Figure 3 shows a plot of the cumula-
tive volumes of water produced in Well B-22, where the
shaded area defines the part that is injected water. This
determines that the early water production in this well
consisted of mainly formation water, but by the end of the
pilot, Well B-22 has cumulatively produced more injected
water than formation water. Well B-19 and Well B-24 has
produced much less water, and both has produced more
formation water than injected water.
up to November 1988 a total of 12.8 MM barrels of water has
been injected in the pilot area, while only approximately
286,000 barrels of formation water and 273,000 barrels of
the injected water has been produced from the pilot wells
combined. If the pilot area is defined to be confined
561
3
within a 1215' radius around Well B-16, the 120' height and
35% porosity of the ED layer gives a pore volume of 35 MM
barrels, and 12% water saturation a formation water content
of 4.2 MM barrels. Compared to these values, only 7% of the
format i on water and 2% of the i nj ected water has been
produced from all three wells, while the injected volume
corresponds to as IlUch as 37% of the total pore volume, or
42% of the HCPV. From November 1988 to October 1989 an
additional 4.3 MMbbl of water has been injected, increasing
cumulative water injection to 49% of the pore volume.
The low and delayed response of the tritiated water in Yell
B-24 was a result of low fracture connectivity between Yell
B-24 and Well B-16. This was also verified from the slow
production re~ponse to water injection and earlier pulse
test results.
IOOIDE-125 TRACER RESPONSE
Well B-22:
Iodide-125 was first observed on April 15, 1987, 21 days
after the slug injection. The response increased from 1.8
Bqll to an intermediate high of 20.4 Bq/l on June 23, 1987.
The response levelled off at an average of 16.7 Bq/l until
a second increase was seen from March to May 1988. Due to
the fP.f.rt lifetime of the iodide isotope, the last analysis
of I was conducted on a sample collected on May 26,
1988, which showed a decay corrected tracer concentration
of 28.1 Bqll, which is the highest concentration measured.
Yell B-19:
Iodide response was first observed on April 29, 1987, 35
days after injection. The initial concentration was 1.9
Bq/l. The concentration increased steadily, and levelled
off at 5.5-7.5 Bqll during phase 2. A second large increase
was seen after the well recompletion/restillUlation, where
the concentrations increased from 8.4 Bq/l on February 1 to
22.1 Bq/l on May 25 with a maximum value of 25.2 Bq/l on
April 21.
Well B-24:
No iodide tracer was produced until the water production
increased in February 1988. The first concentration on
February 25 was 13.8 Bqll. The concentration increased
slightly to 15.6 Bq/l on May 23, with a high of 21.2 Bq/l
on April 21.
IOOIDE-125 TRACER INTERPRETATION
The plot of the iodide responses (Figure 6) for Well B-19
and B-22 do not demonstrate the typical IIturtle9backll
shaped response curve expected from a slug injection. (The
Well B-24 response had a shape similar to the two other
wells.) With time, the iodide response was expected to
build to a maximum and then decrease to zero again as the
bulk of the tracer slug had pa~d the pilot wells. But due
to the 60 days half-life of I, it was not possible to
folLow the tracer response Long enough to confirm such a
shape. Thus neither the maximum produced concentrations,
nor the cumuLative produced tracer materiaL is known.
The IlUch quicker breakthrough of the iodide slug compared
to the tritiated water, especially in Yell B-19, demonstra-
ted that a different flow path for the injected water had
been establ ished from the first tritiated water tracer
injection to when the slug tracer was injected. This is
consistent with the mechanism of capillary imbibition. The
tritiated water was injected with the first water into a
 COMPARISON AND ANALYSIS OF RADIOACTIVE TRACER INJECTION RESPONSE WITH
4~ --...-.:C~H~E::::M~IC::.::A~L~W=A~T..::E~R~A~N~A~LY.:..:S::..:I..:::S_I:..::N:..:.T..:::O_T:..:.H:::E....:E::,:Kp0:.:.,.F.=.,:IS::::;K:.....:...FO=:R.::.M::::.:A:.:...T:.::IO::;:N:.....:..PI:.:L:.::O:.:...T....:W:::.A:..:.T.::.ER:.:.;F~L:.::OOO.::::.... --..;S~P..:;E_2;;..0...;.7_76...,

system with only initial water saturation. This early Injected water: The composition of the injected water is
injected water was quickly imbibed from the fractures into presented in Table 3. The salinity of the injected water
the matrix. At the time of the iodide-125 slug injection, is fairly equal to the pilot formation water, which
4.8 MMbbl water had been injected. Paths of water through excludes the use of chloride and sodium content to monitor
the fractures where imbibition was completed, had probably the produced water. The biggest differences in composition
been established. Small traces of the iodide could then lies in the sulphate, and consequently the barium and
reach the producers in a much shorter time. strontium content, as these are the components of the very
insoluble barium sulphate scale and the low soluble stron-
The extremely low produced iodide tracer concentrations tium sulphate scale. In addition there is a large contrast
indicates that a high degree of tracer slug dilution has in magnesium and potassium content between injected and
occurred in the reservoir. The average produced iodide con- formation waters.
centration of approximately 15 Bqll would be equivalent to
mixing the injected slug of 24.4 GigaBq with 10.2 MM Produced water: As previously mentioned, all of the water
barrels of water. This is more than twice the entire produced from the pilot wells has consisted of both
formation water content in the pilot area, and from these injected water and formation water. The chemical composi-
figures it seems likely that the greater part of the tracer tion of such a mix is dependant upon a number of factors,
slug has been imbibed in the matrix, rather than recovered where two main mechanisms can be identified; interactions
in the producers. It must however be noted that not all the between the i nj ected water and the format i on water, and
mechanisms of iodide tracer behavior is known. Sorption and interactions between the injected water and the formation
various chemical reactions could possibly affect the low rock. Table 4 lists a typical example of produced water
tracer concentrations. chemical composition from the pilot area.

up to the end of the tracer measurement period, a total of Chemical interactions

0.67 GigaBq ( 2.7% of the injected slug) iodide tracer was
produced. Since the tracer concentrations at this time was Injected water-formation water:
still increasing, it can be assumed that less than half of
the tracer that has passed the producers was seen before The main variables affecting the injected water-formation
the radioactivity of the tracer decayed below measurable water interactions are:
values. Thus it is estimated that 5 to 10% of the injected
slug was produced during the pi lot. The fraction of the - the composition of each fraction
tracer that is retained in the formation can be interpreted - the concentration of each fraction in the mix
as the fraction of injected water imbibed. This indicates - the salinity of the mix
that 90-95% of the water injected at the time of the iodide - the resulting solubility of the compounds of the mix
tracer injection has been imbibed and retained in the
formation. This value is a little lower than the current One of the compounds mo~t affected by these interactions
overall average of 98% of the injected water retained is the sulphate ion, SO, -. The concentration of this ion
(based on cumulative injected water production equals 2% of is approximately 100 times larger in the injected water
cumulative injection). This fact may be explained by the than in the formation water. Two common scales formed by
assumption that the imbibition efficiency will decrease the sulphate ion is barium sulphate and strontium sulphate.
with time as the water saturation of the matrix increases. The solubility of these scales is very low, and dependant
These assumptions are supported by the increasing watercuts upon the concentration of sulphate ions. In the case of
seen during the last year of the pilot, where the produced mixing injected water and formation water, the comparative-
water rate is approaching 10% of the injected rate. ly large amount of sulphate ions in the injected water
causes the solubility product and consequently the barium
PART TYO: CHEMICAL PERFORMANCE and strontium ion concentration to decrease. This occurs as
precipitation of barium sulphate and strontium sulphate in
Parallel to the tracer analyses of the pilot water samples, the formation.
conventional 11-ion chemical analyses has also been
performed. The objective of this was to evaluate the These assumptions are verified by observed performance.
possibility to monitor the mixing of injection water with When Well B-22 had reached an injected water concentration
formation water from the chemical composition of the of 10-15%, the barium and strontium concentrations had
produced water. As the pilot has progressed, easily recog- dropped f rom the format i on water concent rat ions to less
nizable trends in the dissolved sulphate, barium, stron- than 1 mg/l and 50 mg/l respectively. In terms of mol-
tium, potassium and magnesium composition have been equivalents, it is noteworthy that the amount of sulphate
observed. Comparing this with the tritiated water analyses, ion that will precipitate with the barium is small; less
it has been determined that the chemical analyses can be a than the sulphate concentration already present in the
valuable tool in monitoring the performance of the produced formation water. What will cause most of the sulphate
water. precipitation is the strontium and calcium ion content of
the formation water. The significance of this is dependant
Chemical cClq)OSition upon the relative fractions of injected water and formation
water. At low ratios of injected water, comparatively more
Formation water: The formation water composition utilized of . th;1 i~lphate will precipitate than at higher
in this analysis is presented in Table 2. This composition ratIos. '
has been generated from the water production in Well B-22
after breakthrough, and an offset Ekofisk only producer. Injected water-formation rock:
Determining a correct formation water compositional basis
for interpretation of the chemical behavior of the produced The water-rock interactions are related to the chemical
water was complicated by the fact that none of the pilot condition of the rock surfaces. This chemical condition is
producers had produced any free water prior to startup of initially adjusted to an equil ibrium with the formation
the pilot. Also the areal differences in water composition water. Introducing water with different chemical properties
from well to well complicate the analysis. into this system will lead to a change of equilibrium.

562
SPE 20776 O. SKILBREI L.D. HALLENBECK AND J E. SYLTE
Mechanisms that will occur are adsorption or desorption of
the ions in the water on contact with the rock. Adsorption
of sulphate to the rock surface is only one example of
this. The impact of these interactions on the produced
water are then dependant on the rock surface area exposed
to the injected water. Early in the waterflood it is
believed that the rock surfaces will remove relatively
large fractions of the ion. At later times larger rock
surface areas will already have adjusted to a new equili-
brium with the injected water, and less adsorption will
occur. Another determining factor should be the production
mechanism. Viscous displacement through the large surface
area of the rock matrix will lead to more adsorption than
fracture dominated flow.
Appl ications of Chemical Analysis
Crossplots:
To better evaluate the effects of chemical interactions on
the produced water, crossplots between some of the chemical
compounds has been made. The tritiated water concentration
is included in these as a comparison of the chemical
behavior to the ideal tracer. Figure 8 shows such plots
made from the chemical and tracer analysis of Well B-22.
The crossplots reflect the different fractions of injected
water and formation water in the produced water. The 100%
i nj ected water endpoi nt is the measured i nj ected water
value from Table 3, while the formation water endpoint is
taken from Table 2.
If there had been no chemical preclpltation or adsorption
effects, all the points on every crossplot would have
fallen on the straight line between the endpoint values.
The relative proportion of injection water and formation
water in a sample could then be determined directly from
its pos Hi on on thi s strai ght line. But because preci pi-
tation and adsorption does affect the chemical data, the
points will tend to be shifted down and to the left on the
crossplot. They can plot on either side of the ideal
straight line, depending on which of the compounds that has
been affected most. The HTO concentration is not affected
by chemical reactions, therefore nearly all the data falls
below the straight line. Sulphate and potassium are the
ions falling closest to the ideal line. This demonstrates
that of the chemical components available, the sulphate and
potassium ions are probably the best alternatives for
monitoring the concentrations of injected water in the
produced water.
Injected Water Fraction:
By comparing the sulphate, potassium and magnesium con-
centrations with the tritiated water concentrations of the
produced water in terms of fractions or % of the injected
water values, quantitative estimates of the injected water
content of the produced water have been made. Plots of
these comparisons are displayed in Figures 9 to 11.
The fractions of injected water calculated from each of the
chemical compounds are plotted together with the exact
injected water fraction, determined from the tritiated
water concentrations. The differences between these lines
are then interpreted as the effects of precipitation and
adsorption. This applies to all of the compounds, but it
appears that the potassium and sulphate ions are less
affected. It should be noted from Figure 10 that even as
the injected water concentration stabilized at around 50%
in Well B-22, the sulphate content of the water continued
to increase. This is interpreted as a decreased rate of
rock adsorption as the waterflood is maturing, due to less
563
5
water exposure to unflooded rock area. It is expected that
with time, as the injected water fraction of the produced
water is increasing, the fractions calculated from the
sulphate concentrations will approach the exact fraction.
Thus it may be concluded that in a mature waterflood, the
sulphate together with potassium will act as a nearly ideal
tracer for monitoring the fractions of injected water
produced.
Chelllical Response Interpretation
Well B-22:
This well has provided the most information on the chemical
behavior of the waterflood response. With the tritiated
water tracer analyses, it has been possible to follow the
chemi ca l compos it i on of the produced water both as a
function of the mixing of injected water with formation
water, and as a function of elapsed time. In the early
stages of water production there is a fairly large separa-
tion between the injected water fraction calculated from
the chemical analysis, and the fraction calculated from the
tritiated water concentrations. At later times this
separation is decreased, interpreted to be a result of less
adsorption and precipitation as the waterflood matures.
From the start of 1989, when continuous tritiated water
injection was terminated, the chemical analysis indicates
that the produced injected water fraction is continuing on
a slowly increasing trend, though very dependant upon the
pilot area injection rates.
Well B-19 and Well B-24:
With regard to chemical response, these wells have shown
similar characteristics as Well B-22. A large difference
between the chemical and tritiated water fractions is seen
initially, but less of a converging trend is noted. This
may be explained by the substantially smaller watercut and
lower cumulative water production in these wells thus far
as compared to Well B-22.
Most noteworthy from the plot of Well B-24 analyses (Figure
11), is the very slow response to the tritiated water
injection termination in 1989, and the stable chemical
composition. This indicates that no high permeability
layers or direct fracture communication exists between this
well and Well B-16.
As seen from Figure 5 and 9, Well B-19 experienced an
increase in produced watercut from 5% to nearly 20% in the
early part of 1989. This was accompanied by a significant
increase in sulphate, potassium and magnesium concentra-
tions. Parallel to this, a direct response to the tritiated
water injection termination was seen as a sharp decrease in
tritiated water concentrations. This behavior suggests that
this breakthrough of new water into the well consisted of
mainly injected water. A total injected water concentration
above 70% is estimated from the sulphate and potassium
concentrations. These observations indicates that what
appears to be a layer or/with a high conductive system of
fractures has broken through to this producer, causing the
significant increase in water production.
The differences in chemical and tritiated water responses
between Well B-24 and the two other wells may be inter-
preted as a result of different displacement mechanisms.
The slow response in Well B-24 to the changing injection
condi t ions is i ndi cat i ve of a higher degree of vi scous
displacement; while the quicker response of Well B-22 and
Well B-19 would imply capillary imbibition with good
 COMPARISON AND ANALYSIS OF RADIOACTIVE TRACER INJECTION RESPONSE YITH
6 CHEMICAL YATER ANALYSIS INTO THE EKOFISK FORMATION PILOT YATERFLOOD
interwell communication. The fact that Yell B-24 however
appears ~re intensely fractured in absolute terms than
Yell B-22 , would indicate that fracture connectivity is
more important than fracture intensity in determining the
displacement mechanisms between wells.
In view of the discussion on the chemical response of Yell
B-19, some interpretations may be inferred from the wells
response to the tritiated water injection termination in
phase 4. The increased tritiated water and reduced sulphate
concentrations in periods of injection shutdown (reduced
watercut) may be explained by assuming that the area turns
over from a state of pressure maintenance to a state of
pressure depletion. In the first state, the produced water
is mainly excess non-tritiated fracture water with a high
level of sulphate. In the second state, more of the water
is drained from the" matrix which previously had imbibed
tritiated water, and where more of the sulphate had
adsorbed and precipitated. This could occur as a result of
fluid expansion, solution gas drive, or gravity drainage
under the depletion scenario.
SlJolMARY
Interpretation of the Ekofisk Formation Pilot was enhanced
through the use of continuous tritiated water injection,
slug injection of iodide-125, and through chemical analysis
of the produced water. The tritiated water injection
provided continuous information to the split of injection
water versus formation water that was produced from the
three offset producers. The slug injection provided a
comparison of breakthrough times for the two tracers to
evaluate any changes in water velocity from initial
injection to after 4.8 MM bbl had been injected. Chemical
analysis of the produced water provided confirmation of the
tracer behavior and can serve in the future as a natural
tracer method.
The low overall produced water from the pilot as compared
to the large injected volume demonstrated a high degree of
water retention in the pilot area. Yater production from
all three producers was initially zero, and reached a
maximum of 15 percent by the end of the pilot. This
translated to a cumulative water production of only 7% of
the pi lot area formation water and 2% of the injected
water. This behavior was indicative of a highly efficient
imbibition process.
Mechanical mixing of the injected water and formation water
was an important factor when analyzing the results of the
tritiated water tracer injection into the Ekofisk Pilot.
Produced tritiated water concentrations were initially low
and increased slowly to a maximum of 62% of the injected
concentration. Mixing of the injected and formation water
would be representative of a viscous type displacement in
which the flowing waterfront mixes with formation water
until a water bank consisting of only formation water is
formed. A gradual increase in injected water concentrations
would occur after the formation water bank is produced.
This behavior was clearly observed in the Ekofisk Pilot.
The iodide-125 tracer slug displayed a much more rapid
breakthrough time toward the producers, particularly Yell
B-19, than the tritiated water tracer. Yhen two tracers are
injected at different times, one after significant water
injection, the breakthrough times would be similar if only
viscous displacement was occurring. A capillary imbibition
mechanism would result in a shorter breakthrough time for
the second tracer. This behavior was clearly observed
between Yell B-19 and Yell B-16. The low recovery of the
iodide-125 tracer was due to the fact that a large fraction
564
SPE 20776
of the tracer slug was imbibed and consequently never
produced.
The radioactive tracer and chemical analysis has confirmed
that both spontaneous imbibition and viscous displacement
contributed to waterflood recovery in the pilot area. The
imbibition component is bel ieved controlled primari ly by
the fracture intensity, whi le the viscous component is
controlled by fracture connectivi ty. As the degree of
fracture connectivity decreases, the viscous component will
increase, which is believed to be the case with the Yell B-
24 performance. This can occur even if fracture intensities
are increasing, if the producer and injector are located
perpendicular to the fracture trends. On the other hand,
decreasing fracture intensities does not necessarily result
in an increase in the viscous component if direct fracture
connectivity is achieved, as could possibly occur if the
injector and producer are located along the same fracture
trends, which could explain the Yell B-22 behavior. This
only highlights the need to understand as much as possible
about the fracture geometry in the reservoi r pri or to
designing the waterflood pattern.
CONCLUSIONS
1. The responses of both radioactive tracers injected were
clearly identified and analyzable. The analysis of both
the tracer and chemical responses of the produced water
have been a valuable tool in understanding the waterflo-
od mechanisms in the Ekofisk Pilot Yaterflood.
2. Tracer and chemical responses indicate that the water-
flooding mechanisms are a combination of capi llary
imbibition and viscous displacement recovery. The
ana lyses a l so i ndi cates that the Lower Ekof i sk format i on
has a very high degree of water retent i on and thus
efficient oil displacement potential.
3. The breakthrough time of the second tracer slug, iodide,
was significantly smaller than for the first breakth-
rough of the tritiated water tracer. This was attributed
to the 4.8 MM barrels of water injected between the two
tracer injections having establ ished paths of water
where imbibition was completed. This would be a typical
characteristic of capillary imbibition dominated
fracture flow.
4. Only 2.7"k of the injected iodide tracer slug was recove-
red in all three wells combined. The half-life of 60
days limited the effectiveness of this tracer to yield
a total tracer slug recovery.
5. The high concentrations of the dissolved sulphate and
potassium ions in the injected water compared to the
formation water of the Ekofisk Field provides that these
can serve as natura l tracers. The accuracy of th i s
method depends on the maturity of the flood. Early in
the flood the produced chemical concentrations will be
strongly reduced by interactions between the injected
water and the formation rock. At later times these
effects are reduced.
6. A correlation between pi lot operation and production
mechanisms have been identified with the tracer and
chemical responses. During injection, imbibition
dominates, as the produced water is mainly excess
injected water from the fracture network. During periods
of injection shut-down, drainage production of water
from the matrix into the fracture network becomes the
dominant mechanism.
SPE 20776 O. SKILBREI L.D. HALLENBECK AND J.E. SYLTE 7

ACKN~LEDGEMENTS Table 1. Sequence of events, Ekofisk Formation Pilot

~aterflood.
The authors would like to thank Phillips Petroleum Company
Norway and the Ekofisk coventurers, including Fina Ex- * February 1986: ~ell B-19, ~ell B-22 and ~ell B-24 si-
ploration Norway Inc., Norsk Agip A/S, Elf Aquitaine Norge detracked and completed as Ekofisk only
A/S, Norsk Hydro A/S, Total Marine Norsk A/S, and Den producers.
Norske Stats Oljeselskap A/S for permission to publish this * March 1986: ~ell B-16 sidetracked to serve as a
paper. Support from A.A. Hamouda with Phillips Stavanger, Lower Ekofisk injector.
and the Phillips Petroleum Company's office in Bartlesville * July 26, 1986: ~ell B-16 placed on continuous injec-
is also acknowledged. The IFE, Institute for energy tech- tion at 21,600 B~PD injection rate @
nology, in Oslo is commended for their support and work 2,900 psig injection pressure.
with the tracer analysis part of this project. * August 18, 1986: Start-up of continuous tritiated water
injection in ~ell B-16, giving the
REFERENCES: injection water a radioactive concen-
tration of 2235 Bq/l. 9 125
1. Thomas, L.K., Dixon, LN., Evans, C.E., Vienot, * March 25, 1987: A slug of 24.4 GBq (24.4*10) I, a
M.E. "Ekofisk ~aterflood Pi lot". JPT (February radioactive isotope of iodide, was
1987) 221-232. injected in ~ell B-16.
* August 1987: Four week fieldwide shut down due to
2. Sylte, J.E., Hallenbeck, L.D., Thomas, L.K., "Eko- platform Jack-up operations.
fisk Formation Pi lot ~aterflood". SPE Paper 18276 * January 1988: ~ell B-16 restarted injection since the
presented at the 63rd SPE Annual Technical Con- August 1987 shut-down.
ference and Exhibition, Houston, October 2-5. 1988. * February 1988: Reperforations and restimulations to
add the Upper Ekofisk formation in \Jell
B-19, ~ell B-22 and ~ell B-24.
3. Sulak, R.M., Nossa, G.R., Thompson, D.A., "Ekofisk * November 1988: Tritium tracer injection in ~ell B-16
Fi eld Enhanced Recovery", paper presented at the permanently terminated.
2nd International Conference on North Sea Oil and * October 1989: ~ell B-16 taken off injection. End of
Gas Reservoirs, Trondheim, Norway, May 1989. Ekofisk Formation Pilot injection.

4. Sulak, R.M., Thomas, L.K., Boade, R.R., "Rigorous

Treatment of Compaction Drive in a 3-D Reservoir Table 2. Compositional Analysis, formation water
Simulator: A Case Study of Ekofisk Field". SPE
paper 19802 presented at the 64th Annual Technical !!l9L!. me/l
Conference and Exhibition, San Antonio, October 8-
11, 1989. Sodium, Na 12,300 535
Calcium, Ca 850 41.5
5. Jewhurst, J., and ~iborg, R., "Ekofisk-Reservoir Magnesium, Mg 250 44.3
Management in a Compacting Environment", The Barium, Ba 20 0.36
Offshore Digest, (1986). Vol. 2, 54-67 I ron, Fe 5 0.2
Strontium, Sr 170 3.9
6. Dangerfield, J.A., Brown, D.A., "The Ekofisk Fi- Potassium, K 100 2.6
eld". North Sea Oil and Gas Reservoirs, The Nor-
wegian Institute of Technology (Graham and Trotman,
1987) 3-22
Chloride, Cl 22,000 620
7. Snow, S.E., Hough, LV., "Field and Laboratory Sulphate, S04 25 0.52
Experiences in Stimulating Ekofisk Area North Sea Bicarbonate, HC03 500 8
Chalk Reservoirs". SPE Paper 18225 presented at the
63rd SPE Annual Technical Conference and Exhibi-
tion, Houston, October 2-5, 1988. Table 3. Compositional Analysis, injected water

8. Thomas, L.K., Dixon. LN., and Pierson, R.G., !!l9L!. me/l

"Fractured Reservoir Simulation" SPEJ, (February
1983) 42-54. Sodium, Na 10,900 474
Calcium, Ca 404 20.2
9. Brigham, ~.L, and Abbaszadeh-Dehghani, M., "Tracer Magnesium, Mg 1,320 108.2
Testing for Reservoir Description", JPT, May 1987. Barium, Ba 0.1 0.0
Iron, Fe 0.1 0.0
10. van Gol f-Racht, T.0., "Fundamentals of Fractured Strontium, Sr 7.8 0.2
Reservoir Engineering", Elsevier Scientific Publi- Potassium, K 373 9.5
shing Company, 1982.

11. Vetter, O.J., "The prediction of scale problems due

to injection of incompatible waters", SPE paper Chloride, Cl 20,100 566
7794 presented at the Middle East Oi 1 Technical Sulphate, S04 2.630 54.8
Conference of the SPE held in Manama, Bahrain, 25- Bicarbonate, HC03 130 2.1
29 March 1979.

12. Mitchell, R.~., "The Forties Field Sea-~ater Injec-

tion System" JPT (June 1978) 877-884.

565
 SPE 207":' 6

Table 4. Compositional Analysis, produced water

!!!llLl mell

Sodium, Na 11,600 504

Calcium, Ca 818 40.9
Magnesium. Mg 595 48.8
Barium, Ba 0.6 0.0
Iron, Fe o. , 0.0
Strontium, Sr 41.1 0.9
Potassium, K 233 6.0

Chloride, Cl 20,500 577

Sulphate, 504 1380 29
Bicarbonate, HC03 371 6.0

------:~

lj)~~
// /
10250

/ /
PILOT AREA CONFI tJRATION

"{
10150

1198 II

10100

1353 " 10050

~-

Fig.1 - Ekofisk Formation pilot Well configuration map.

566
 [BWPD] (Thousands)
25

20
t ~~d;F~ injection ~ ~ !~
II ~ f\
15

10
I~
r
I
5

,
o
86.5 87 87.5 88 88.5 89 89.5 90
Time
Fig.2 - Well B-16 injection rates.

en
~

[Bq/l] we [Ofo] (Bq/l]

... 200C :-,~_--------------------- 30, 11200
1800 8-19 I I I ~ we

1600
--8- 8-22 25 -+- Tritiated water 1000
-+- 8-24
1400
20 800
1200
1000 15 600
I

':~
800
400
600
400
200
200
0'
86.5
G~
87
'0
87.5 88
I J
88.5 89 89.5
I
90
0I
86.5
I ---~I ~v_,Y

87 87.5
,'-...I
88
I

88.5
I

89
I

89.5
10
90

Time Time
Fig.4 - Tritiated water response, all wells. Fig.5 - Tritiated water response & we, Well B-19.
 SEE 20776

[Bq/l]
30r--

25 r§ -4- 8-22
3000 ,504 :mg/l]
100% INJ .
... ~
20 ..oo~ _~ ~~_._.... WATE!!

..oo~··_···--_ . ·_···· -/

15 zo, ~--·---,-.-7'::·>~'-"-"·--··"'--
....:...
.
'''~,-- ~._--- ._-~--~_._-_."--

'OOHOO%~Mo
10, lodide-125 R ,L'__ __ __
slug
o ,/
, 000 ....
HTO [Bqjl]
,
o
~

600
~

1000

504 [mg/l]
~

1600
~

2000 2500
~

0 1400 i
Mg [mg/l] uoo,Mg [mg/l]
87 87.5
Time
88 88.5 1200 ~_._--~,------_._--_.,-,--_._-_._-_.,_

1000 f-~'----~'"'''''' '-~"--"'---"'-'I

..__.__._." .._-
"
1000
. f---..-----.
Fig.6 - Iodide-125 response, Well B-19 & Well B-22.

........,;"" ".
--;-- ......,. ..
§!

we [%) ,L' __ ,L'__ __ __ __

I
~ ~ ~ ~ ~_

,
~ ~ ~ ~_-'

50 1000 1500 . .00 100 150 200 250 3..

HTO [Bq/l] K [mg/l]

I' ~ 8-19 I
40 ~ I -B- 8-22 tl __ _
-4- 8-24
Mg [mg/l]
K [mg/l] 1400 i
.00,--

"00

30 .000

ooo~· -........
--~~:-.~"!".!; .-~-
, ~.. -·__..·_··O~-;:·'i--:·"··_-~~------·
20
'OO~--'~-~-"--.-_·
__ ··--·_·I

10
e!
, 100'
HTO [Bq/l]
, ,L'__
,
~

... ~

1000
504 [mg/l]
__ __ __
~

.300 ....
~ ~-'

2600

oI I!~O seoo e e 6' / ~,<P' " Fig.a - Crossplots, Well B-22.

86.5 87 87.5 88 88.5 89 89.5 90

Time
Fig.? - Produced water cut, all wells.
 SPE 2077 6

10f] ~ I I 100 [%~ =

-a- we -a- we
80 f-I w
~ 804
80 _w__
~ 804
+~ , +~

60 f-- +K •
f:IA A
0A{> I 60 +K

40 40

20 20

0' !Tr!mlUJ!N"'U~ r=--t:l - !'H

0
86.5 87 87.5 88 88.5 89 89.5 90 86.5 87 87.5 88 88.5 89 89.5 90
Time Time
Fig.9 - Produced injection water, Well B-19.
...
<II
Fig.10 - Produced injection water, Well B-22.
'"
[%]
100 I
I
HTO
80 H --B-- we
4- 804
-+- Mg

6°1~

40

20
I
I

0
86.5 87 87.5 88 88.5 89 89.5 90

Time
Fig.11 - Produced injection water, Well B-24.