IPTC 13440

Water Injection Fall-Off Tests in Deepwater Reservoir:

What Do We Actually See Into Formation?
Sergey Abramochkin, Andrew Carnegie, Saifon Daungkaew, Gordon Goh, Schlumberger, Phil Bee,
Murphy Oil Corporation

Copyright 2009, International Petroleum Technology Conference

This paper was prepared for presentation at the International Petroleum Technology Conference held in Doha, Qatar, 79 December 2009.

This paper was selected for presentation by an IPTC Programme 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 International Petroleum Technology Conference and are subject to correction by the author(s). The material, as presented, does not necessarily
reflect any position of the International Petroleum Technology Conference, its officers, or members. Papers presented at IPTC are subject to publication review by Sponsor Society
Committees of IPTC. Electronic reproduction, distribution, or storage of any part of this paper for commercial purposes without the written consent of the International Petroleum Technology
Conference is prohibited. Permission to reproduce in print is restricted to an abstract of not more than 300 words; illustrations may not be copied. The abstract must contain conspicuous
acknowledgment of where and by whom the paper was presented. Write Librarian, IPTC, P.O. Box 833836, Richardson, TX 75083-3836, U.S.A., fax +1-972-952-9435.

Abstract valuable well test outputs, specifically for short-lived fall-

Water injection is an important recovery method in
deepwater reservoirs. A certain water injection volume is
required to maintain target oil production. A short
injection / fall-off test is therefore often conducted to
evaluate well injectivity performance prior to oil
production. Following that, permanent downhole gauges
are set to monitor full field development scale. Pressure
transient data from the permanent gauges are then also
analyzed as long-term injection and fall-off tests.
Reservoir parameters, such as permeability and skin
factors can be obtained from these short- and long-term
tests. However, several questions must be asked about the
interpretations, such as 1) What do we actually see into
formation, i.e. injected water or reservoir oil? 2) Do the
permeability and skin values change with time? 3) How
can we capture and understand reservoir changes versus
time?
Determination of in-situ reservoir permeability from
pressure transient tests has always been of primary target
in well test analysis. In a single-phase system, an answer
is straight forward. However, in the two-phase injection
system, the single phase assumptions cannot be applied
any longer. When interpreting water injection fall-off
tests, it is important to consider the volume and
consequent injection radius occupied by water pumped
into formation. These factors are significant in log-log
plot interpretation because the first, early radial flow
stabilization plateau reflects the relative permeability to
the water injected [1]. However, the reservoir performance
and production forecasts need to use the effective
permeability to oil. This is usually inferred from long-
term fall-off tests that investigate beyond the injection
water invasion zone. Therefore, an appropriate
methodology should be devised and applied in test design
and interpretation to deliver accurate commercially
offs.
A three-dimensional dynamic simulation has been
selected as a means to show the importance of the above
factors, understand fall-off peculiarities, and discuss the
way the outputs can be achieved when testing injection
wells during continuous pressure maintenance schedule.
The paper demonstrates the application of the simulation
to fall-off tests performed on the two-phase reservoir
model. Responses of pressure fall-off recorded at various
stages of long-term injection sequence on the log-log plot
are produced and compared with actual tests from
deepwater reservoirs. Reflection of fall-off test duration
on phase mobility dynamics is presented. Importance of
rock wettability and relative permeability-saturation
relation is discussed. Contribution of phase-by-phase
saturation related peak flow rates radially propagated in
time from the wellbore away into a reservoir during a fall-
off and their association with the pressure derivative
curve to determine effective phase permeability values are
discussed in this paper. Awareness of changing of the
effective phase permeability values on pressure derivative
that reflects either effective permeability of a phase while
recorded at various moments of injection schedule allows
one to provide a comprehensive well test analysis. To
understand this effect, the correct value should be
reported instead of an overall absolute reservoir
permeability value that could mislead model calibration
issues in field development study or other reservoir
engineering campaigns. Such an approach can be a useful
tool in analyzing both simulated and field fall-off data to
benefit dynamic simulation in the testing business.
Test design and interpretation considerations and
challenges of their use in water injection well tests are
also discussed. In conclusion, the paper is aimed to close
2 IPTC 13440
the gap between technical understanding and the actual
field data during short- and long-term injection fall-off
tests as they become more prevalent nowadays in the
petroleum market place.
Problem Setting
The evolution of testing tools has provided solutions of
new interpretation algorithms and methods. The
downhole pressure transient testing remains an important
phase in field exploration and production to deliver
reliable information about in-situ formation conditions.
One particular output from the well test analysis is the
formation permeability, i.e. a parameter critical to the
determination of field production capacity and planning
of downhole operations for maximizing hydrocarbon
recovery.
When performing pressure transient tests involving a well
shut-in operation in a single-phase system, an answer to
the permeability is straight forward, i.e. the value is
determined from the stabilized section of the pressure
derivative line from the pressure derivative log-log
plot [2], and its magnitude is dictated by how high or low
the stabilization section is with respect to the
dimensionless pressure in a general type-curve analysis
approach. This type of analysis has widely become
computer-aided nowadays. However, in a two-phase
system, particlualrly for water injecting wells in a
waterflooding development plan, the conventional
interpretation approach does not consider the mobility
change between the injecting phase and the formation
fluid. It becomes an important issue to consider the
volume and the associated injection radius occupied by
water pumped into formation. The on-going injection
creates an individual pressure response with additional
stabilization plateau during fall-off and commonly it is
quite apart from the one created by the formation fluid
due to difference in effective permeability, relative
permeability, phase mobility and rock wettability issues.
The response becomes mored dramatic in the case when
long-term injection operation is followed by a short fall-
off that does not have enough duration to see beyond the
waterflooding front, i.e. the value of the investigation
radius is small. In such a case, the permeability
determined will represent only the water phase. Should
the value be taken for the permeability due to oil, it will
result in overestimating the reservoir performance and
production forecast. On the opposite side, when an
injection fall-off is analyzed at the beginning of a well life
after short post-perf injectivity, the radial flow
stabilization plateau due to water can not be identified as
it is masked by wellbore storage and only the one due to
oil is visible in the plot. At this early injection, the
waterflooding front has not propagated deep into the
reservoir yet. Therefore, an appropriate methodology
should be devised and applied in test design and
interpretation to deliver accurate commercially valuable
well test outputs, specifically for short-lived fall-offs.
Summarizing the relevance of the above, there are three
questions that must be asked about the interpretations,
such as 1) What do we actually see into formation, i.e.
injected water or reservoir oil? 2) Do the permeability and
skin values change with time? 3) How can we capture and
understand reservoir changes versus time?
Simulation Model and Results
A three-dimensional dynamic simulation has been
selected as a means to show the importance of the factors
mentioned in the previous section, understand fall-off
peculiarities, and discuss the way the outputs can be
achieved when testing injection wells during continuous
pressure maintenance. The simulation has been set up and
run using a dynamic simulator. The model was designed
to be a radial-grid with the grid size logarithmically
refined from the wellbore diameter of 12 inches and
outwards up to an external diameter of 85000 ft, for
graphical illustration, see Figure 1. Such configuration is
due to a typical radial propagation of the injection flow in
the near-wellbore of a vertical well as well as to ensure
the wellbore pressure transient data recorded at a high
sample rate is accurate and the boundaries are not felt
during a fall-off period. Another reason of using a
logarithmic refining approach is to exclude large number
of cells in the model that would severely slow down the
calculation time. The input data used in the simulation are
stated in Table 1. For reservoir pressure maintenance
issues, a production well was considered in the radial
model located at 6600 ft away from the injecting well
with the production flow rate equal to the injected rates
under complete voidage replacement consideration.
To understand the effect of wettability system on the log-
log response, oil-water fluid model with three (3)
scenarios has been considered corresponding to different
wettability systems: water-wet, intermediate-wet, and oil-
wet, respectively. The examples of the wettability systems
used in the model are shown in Figure 2 (a), (b), and (c).
The relative permeability curves have been generated
artificially using a dedicated software suite.
Each wettability system scenario has featured a post-perf
(short-term) injection period followed by a basic fall-off
and then with four additional cases corresponding to
different durations of a main injection period as can be
seen from Figure 3. The main injection period is then
followed by a final (long-term) fall-off. The durations of
the main injection period are equal to seven (7), thirty
(30), ninety (90), and one hundred and eighty (180) days,
respectively, and for observing the dynamics of a
transition stage between water-related and oil-related
stabilization plateaus (phase mobility change). For the
sake of simplicity, the injection rate was made constant
thoughout the simulation and equal to 5000 bbl/d. The
total volume of the injected water its corresponding radius
of injection has been calculated for each main injection
period case and is stated in Tables 2 and 3. During short-
term and long-term fall-offs simulation runs have been set
to record bottomhole flowing pressure at every 5 secs.
IPTC 13440
The duration of both basic and final fall-offs were 24 hrs.
After a run is completed, fall-off pressure readings were
extracted and represented on a log-log comparison
pressure derivative plot.
The results of simulation are shown in Figure 4. Cases (a),
(b), and (c) are corresponded to the above-mentioned
wettability systems, respectively. Since the recording of
fall-off pressure was simulated to be at the wellbore face,
the wellbore storage is not considered in the plots. In
reality, the first 1.5 to 2 logarithmic cycles should be
attributed to the wellbore storage domination period that
distorts the reservoir response. From the plots mentioned,
two radial flow stabilization plateaus are clearly seen. The
upper one corresponds to the radial flow due to water
(marked as effective water permeability @ S orw ) and the
lower due to oil (marked as effective oil permeability
@ S wi ), respectively. A step-down phase change (or
mobility change) as indicated by the inflection point
between two plateaus in the derivative plot, is also seen.
This corresponds to the phase change between the water
and the formation fluid. In the subsequent fall-off events,
it propagates further as observed along the shut-in time
against the corresponding durations of the main injection
period. The greater the volume of water injected during
the main injection period, the further in time the phase
change response in pressure is observed on the log-log
plot.
Aslo of note, is that the position of radial flow plateaus
with respect to each other and the shape of the phase
change section in the log-log plot are governed by the
type and signature of wettability system, i.e. relative
permeability curves, as previously shown Figure 2 (a),
(b), and (c). The greater the difference between oil and
water end-point permeability values, the more their radial
flow plateaus are apart from each other. The contrast
intensity of the phase transition zone depends on how
expressive is the rock wettability to water phase. On the
contrary, for oil-wet rocks, the transition section is
stretched along the logarithmic shut-in time range. For
intermediate-wet rocks having the characteristics of an
inflected oil-relative permeability curve (i.e., the end-
point relative permeability to oil shown in Figure 2 (b) is
deliberately overstated for emphasis), the transition
change in the log-log pressure derivative creates the hump
in signature before it steps down to the radial flow due to
oil.
For each case, the radius of injection has been calculated
using the formula shown in the below and the results are
stated in Table 3:
Vinj
rinj = 2.37
hnet (1 S wi S orw )
3
where r inj is the radius of injection in ft; V inj is the
injected volume of water, bbl; h net is the net formation
thickness, ft; is the formation porosity; S wi is the
initial water saturation, and S orw is the residual oil
saturation, respectively. The last three values mentioned
are dimensionless in units.
To understand the dynamics of saturation propagation
within the formation, a plot of water saturation
distribution, S w , at various distances from injector well
versus injection program duration has been created from
simulation results. A case for water-wet wettability
system is shown in Figure 5 (a) and (b). For distribution
plot covering the first 20 days of injection, see
Figure 5 (a), note the stabilized straight lines of water
saturation between Day 1 and Day 2. These are due to the
well being shut-in for a basic fall-off that was considered
in the simulation model after the short-term injectivity
test. Here is an example to explain the meaning of this
distribution. For an example to examine the water
saturation behaviour at 300 ft away from the injector, an
observation is made at line 12, Figure 5 (b) that
corresponds to a distance of 275 ft. This figure of 275 ft is
taken as the closest arbitrary distance to the interest of this
investigation because the non-rounded distance values
presented in Figure 5 (a) and (b) are the result of
summation of radial block sizes used in the simulation
model in a logarithmically refined manner. During the
first fifteen (15) days the water saturation in the formation
at the given distance remains equal to the initial water
saturation or water connate saturation since the volume of
injected water has yet to arrive. For our water-wet
scenario, the initial water saturation, S wi , is equal 0.26
and corresponded to zero relative water permeability in
Figure 2 (a), mentioned previously. As the water front
arrives at the given distance, the water saturation starts to
increase and the formation fluid gets desplaced by
injected water until it is completely flushed. At this point
the saturation reaches the residual oil to water, S orw ,
equal to 0.25 in our case and the value corresponds to the
end-point water permeability in Figure 2 (a), respectively.
The water saturation recorded during radial phase change,
i.e. within the range of S wi to S orw will correspond to a
value of relative permeability function, k rel , at the given
water saturation value S w . To get the absolute
permeability, of formation, k abs , the product of the
following equation can be used:
k eff = k rel k abs ,
, where k eff is the effective permeability value that is
derived during downhole pressure testing analysis.
4 IPTC 13440
Another way to represent the saturation propagation in the
formation is given in Figure 6, where the water saturation
is plotted against the distance from the injecting well at
various times. The distribution is less accurate to the one
plotted previously in Figure 5 since with the increased
distance the size of grid blocks in the model gets larger
due to logarithmic refining and, hence, it decreases the
resolution accuracy. Like in Figure 5, the water saturation
distribution vs. distance is provided for the water-wet
scenario.
Thus, discrepancy in phase permeability values at water
injection fall-off tests are mainly determined by relative
permeability curves. The first radial flow stabilization
plateau on a pressure derivative log-log plot determines
the effective permeability to water while the later one to
oil, respectively. However, a question can be raised here
whether the radial flow stabilization plateau due to water
could contain information about the oil phase. To provide
an answer and understand the effect of oil phase on the
pressure derivative curve vs. elapsed time, the residual
phase-by-phase flowrate propagation within formation at
different times, after the well was shut-in for fall-off, was
generated. For the sake of simplicity, only the basic fall-
off and the one following the one-week main injection
period were used in the analysis. For their responses on a
pressure derivative log-log plot, a series of three nodal
points were selected: at the beginning, in the middle, and
at the end of a phase transition, see water and oil phase
series 1-a, 1-b, 1-c and 2-a, 2-b, 2-c in Figure 7. Then,
contribution of phase-by-phase saturation related peak
flow rates radially propagated at the above mentioned
points from the wellbore away into a reservoir during the
abovementioned fall-offs were derived. The association
with the pressure derivative curve to determine effective
phase permeability values is shown in the sequence of
plots from Figure 7 to Figure 9. In Figure 8, the
propagation is plotted against the model cell number and
the injection well is at zero cell number (for our case with
logarithmically refined model the cells get greater in the
radial size away from the injector summing up to 85000 ft
as recorded in Table 1, so the refining effect explains the
distorted response of a phase flowrate at larger cell
numbers in Figure 8).
It is seen from the figure there is not much of water phase
front propagation into the reservoir within a fall-off (no
injection during the well shut-in), as observed in water
series 1 for the first fall-off and that of series 2 for the
final fall-off, respectively. As the water stops propagating,
its flowrate magnitude attenuates with the elapsed fall-off
time, see sub-indices a, b, and c within each oil zone, interpretation of the initial, early injection and
water series 1 or 2, see their corresponding locations in
the plots, i.e. from Figures 8 and 9. On the contrary, the
oil phase continues propagating away from the injector
with less attenuation compared to that for the water and
there is domination in ratio of the oil flowrate over the
water one, respectively. Besides that, the phases are
overlapping with respect to each other. For the nodal
point at the beginning of the phase transition; see series 1,
the oil flowrate line overlaps with the region, where the
water flowrate line reaches the highest magnitude. The
latter confirms the fact the data about the oil phase is
contained in the radial flow stabilization plateau due to
water. In Figure 9, where the same propagation is
represented against distance from the injector, one might
see the wide extension of the oil flowrate. This explains
the reason of using the infinity-sized model, i.e. to
avoid interaction from the model impermeable boundaries
that will reveal itself from the type-curve pressure
transient analysis in the log-log plot.
Field Background and Reservoir Challenges
To test if the hypothesis can be observed in the field,
injection well pressure data was obtained from a
deepwater oilfield in the Malay Basin, offshore Sabah
(East Malaysia). Water injection is a key component of
the project, and required for pressure maintenance, sweep,
and to attain a high recovery factor. Permanent downhole
gauges (PDG) were installed in all water injection wells
on the field to allow for continuous pressure surveillance
without intervention. Both planned and un-planned
pressure fall-off data is recorded at high frequency, and is
used for pressure transient analysis. Planned fall-off
periods include injectivity tests (usually initially on a
well) and as part of regular field surveillance activities.
Un-planned fall-offs occur occasionally when there are
shut-downs in the injection system, and the wells shut
automatically.
The water injection well used as an example, is of subsea
wet tree design, on a manifold with several other injection
wells in water depths of around 1300m. The well was
perforated in a single, homogenous and high-quality
sandstone reservoir of 12m net thickness. The well angle
through the reservoir is 66 degrees, and the well is located
within the oil zone. Immediately post-perforation an
initial injectivity test and fall-off was performed on the
well. Subsequently, a visco-elastic gel (VES) frac
operation was carried out to improve the injectivity of the
well. Continuous injection commenced 12 months later
with no changes to well injectivity after the shut-in
period. Typical injection rates are around 24,000 bwpd.
Injectivity Analysis Challenges
One of the prime aims of injectivity tests is to verify
dynamic (pressure transient analysis-derived)
permeability-thickness, kh , with that derived from
petrophysical logs. In the case of water injection into an
fall-off periods can be problematic unless the presence of
the water flood-front is recognized and accounted for in
the pressure transient analysis. In addition to this, the
effect of temperature on injected fluid viscosity is
important as this will vary the mobility ratio between the
water and oil. Over time and continued injection, cold
injected water will warm as the flood front progresses into
IPTC 13440
the reservoir, and the mobility ratio between the fluids
will change.
During several well injectivity conformance and fall-off
test technical reviews, it is found that there can be a
significant discrepancy between the initial well-test
derived permeability-thickness, kh resulted from post-
perf short-term injectivity and that from the petrophysical
log-derived. This discrepancy can be as high as multiple
of five (5). At the same time, the long-term planned
injectivity which was carried out after one year of
continuous water injection, shows relatively good
coherence of derived kh result compared to that from the
log-derived petrophysical analysis. This has created a
sense that the interpreted perm-thickness result from the
post-perf short-term well-test analysis is dubious until
proven otherwise.
From the actual field data, as shown in Figure 10 [3] shows
viscosity versus temperature for this particular formation.
The viscosity of water at the reservoir temperature is the
same as the oil viscosity (0.3 cp). If water is injected at
cooler temperatures than the ambient reservoir
temperature, it may in turn cool the reservoir fluids near
the well bore, thereby increasing the viscosities of these
fluids. The viscosity difference is less than three times of
the oil viscosity at the near wellbore region and then
temperature increases again away from the wellbore. This
thermal effect is insufficient to explain why the
permeability derived from short injectivity analysis was
observed to be significantly higher than the long term
injection data. To investigate the causes for the
discrepancies, a 3-D dynamic simulation to generate a
representative reservoir injection pressure response while
the injection front propagates was built. The transient
analysis result from this model-derived pressure
derivative is then verified and compared with analysis
from actual field data. With an encouraging match
between these two transient analyses the study has been
carried forward, with some recommendation of further
studies in the future.
In the following discussion, field examples of the
previously discussed pressure transient analysis impact
from early-time, middle time until the late-time injection
events are highlighted. From the post-perforation (i.e.
early-time) injectivity test, to the near wellbore treatment
(i.e. middle-time) injectivity events until the long-term
(i.e. late-time) planned injectivity test performed, the
observation of the injectivity results is made with respect
to the topic of this paper.
(1) Post-perforation injectivity test
This injectivity test of Well A was conducted right after
the formation was perforated and lasted around 2 hours.
The test objective was to establish injection assurance by
pumping treated seawater at maximum stabilized rate of
25000 bbl/d or reaching the wellhead pressure of
3,000 psi, whichever occurs first. The injectivity test was
5
designed to include a series of step rates to evaluate the
injectivity capability and to feature a long stabilized
injection period to ensure the injectivity is stabilized.
However, after few step rate series, the injection pressure
increased above the estimated fracture initiation pressure
and no significant sign of fracturing was observed. The
injection rate was reduced, and it was expected that the
pressure would decrease as well. However, the pressure
kept increasing, so the injectivity test was halted, with a
total of around 1300 bbl of water injected. Following this,
the well was shut in for pressure fall-off for less than
2 minutes. The pressure transient data have shown to be
dominated by wellbore storage period and no radial flow
was observed, as shown in Figures 11 and 12.
(2) Short-term injectivity test during VES fracturing
campaign
With the objectives of the post-perforation injectivity test
not achieved, one month later, Well A was brought back
from suspension for the second injectivity test as a part of
the VES fracturing operation [3]. This time, the injection
and fall-off tests consisted of a series of five injection and
fall-off sequences. All fall-off tests lasting from one to
three hours proved to have the same radial flow
stabilization plateau, even though the skin value changed
at every fall-off due to the VES fracturing events. The
radial flow stabilization plateau delivered the permeability
value of 350 mD and was analysed as oil effective
permeability since the total injected volume for this
operation was less than 3500 bbl. This left the radial flow
due to water unidentified again due to the wellbore
storage period domination in the pressure derivative log-
log analysis plot (as shown in Figures 13 and 14).
(3) Main injection program with pressure recorded
with PDG
The continuous injection program commenced twelve
months later after the VES fracturing operation was
completed. Data analyzed for this paper was recorded
over eight months of continuous injection with an average
water injection rate of 24 000 bbl/d. The well was
equipped with PDGs and the bottomhole pressure data
were captured in real time for reservoir monitoring
purposes. Fall-off data were obtained when the well was
shut-in for operational purposes. The injection pressure-
rate history from PDG is shown in Figure 15. For selected
fall-off periods marked with light blue color in the figure,
a comparison pressure derivative log-log plot was created
and shown in Figure 16. The plot also captures the fall-off
from the post-perforation injectivity test, Fall-off #1, and
the three most representative fall-off data taken from the
VES fracturing period, Fall-offs #2 to #4, respectively.
The rest of the data (i.e. from Fall-off # 5 to Fall-off #10)
correspond to the main injection period. On the whole, the
pressure fall-off data of the Well A follows our
hypothesis: two stabilization plateaus to water and oil are
seen with steady propagation of the waterflooding front
with distance and in time. This radial flow stabilization
plateau corresponding to the effective water permeability
at the residual oil saturation is almost five times less than
6 IPTC 13440
the one corresponding to the effective oil permeability at
the initial water saturation. In the figure, where the fall-
off derivative is updated with the input of correct rate
history, fall-offs #1 to #5 show that only the radial flow
stabilization plateau due to oil is observed. Fall-off
periods #11 and #12 are too short, such that the injection
front transition is way beyond the total investigation
radius. Fall-off #14 delivers a good radial flow
stabilization plateau due to water as its radius of
investigation is greater than the radius of injection (i.e. the
injected volume radius). In general, the comparison
pressure derivative log-log plot shows the significance of
integrated illustration of all pressure fall-off data.
Table 4 lists the results obtained from fall-offs including
the effective permeability to water, the skin, and the
mobility ratio. It can be seen that the average permeability
remains around 113 md. The high skin value at the
beginning corresponded to the fracturing period with
visco-elastic gel and breaker injections. As the main
injection program commenced, skin became negative and
steady at values of -2.5 to -3.0 as a result of maintaining
high injection rates at 24 000 bbl/d. Therefore, it can be
said that skin can change with time at the beginning as the
result of post-perforation operations, cleaning up,
formation treatment and fracturing. However, in
consideration of selecting the radial flow and keeping
flow rates unchanged, the final variations of skin are
reasonable. The last column, the mobility ratio, shows the
ratio of mobility in the inner zone (water) to mobility in
the outer zone (oil). For fall-offs #2, #4, and especially
#5, where two radial flows are identified, it almost
corresponds to the ratio of end-point phase relative
permeability values. As the injection program continues,
the ratio approaches unity. This could be due to the
following reasons: (a) phase transition getting smeared
due to geological heterogeneities, (b) actual reservoir
thinning away from the wellbore, and (c) change of water
viscosity due to thermal effects.
The importance of selecting the proper radial flow
stabilization plateau and the effect of the application of
the output permeability-thickness for specific reservoir
engineering purposes, can be illustrated by examining the
fall-off period #5, which is at an early stage of the water
injection program, as shown in Figure 17. The radial flow
due to water is not seen here as it is masked under the
wellbore storage effect. Selecting the visible radial flow
due stabilization plateau to oil, Figure 17 (a), and
interpreting it using water parameters delivers the
overstated results for permeability and skin, i.e.
k = 520 md and S = +15.7, respectively. However,
putting the radial flow up to a level, where the
stabilization plateau due to water is, Figure 17 (b), as
observed from later fall-offs, where an inflection point
exists in between two stabilization plateaus, the results
become k = 114 md and S = -1.0, respectively. The
decrease between the permeability results is more than
4.5 times. There are obvious consequences that the
incorrect data might have impact on decisions about well
design, completion, production planning and forecasting,
especially if the data was acquired during the appraisal
stage of a field. Consequently, it is important to design an
appropriate pressure transient testing program in order to
allow identification of both of the radial flows in one fall-
off period. This idea of a proper injectivity testing
program in order to fulfill the testing objectives
highlighted in this paper will be discussed with
recommendations in the next section.
The study demonstrates the needs to understand the
limitation of pressure transient analysis, particularly with
computation of the radius of investigation that is based on
the assumption of perfect cylindrical radial displacement
due to homogeneous formation parameters. In reality,
there will be forces acting to distort the flood front away
from cylindrical, i.e. due to gravity, geological
heterogeneity, dip of formation, and lateral geological
changes in all directions away from the well. The real
data will thus have a smeared transition between oil and
water phases, with the shape of the transition roughly
corresponding to a fractional water curve in a vertical
sense. Areally, with the dip and thickness changes with
distance away from the well, and in the example case a
contact downdip, the true transition shape is likely to
remain unknown and it will not be a perfect cylindrical
shape. In some sequences of later fall-offs (#7, #9, and
#13 as examples in Figure 16), one cannot observe two
stabilization plateaus but the pressure responded to a
flattening or thinning of the water displacement front that
could mean a higher actual value of the investigation
radius.
The injection radius data obtained from analysis should be
compared with that from the 3-D dynamic field model,
which incorporates the actual well injection history. The
modelled radial displacement can show sagging effects
like coning of the pumped water near the wellbore and
when the waterflooding front propagates deeper into
reservoir due to gravity effects and formation thickness
changes away from the wellbore.
Test Design and Interpretation Considerations
Problems stated in the previous sections and observations
from field data have raised a reasonable issue on the
needs for fall-off test design to recognize both radial flow
regimes and the phase transition, respectively. It also
becomes clear that observing the waterflooding front
away from a water injection well is generally only
possible early in the field life. Therefore, to gather the
required information, a series of fall-offs should be
incorporated at earlier stage of the water injection plan.
How often these fall-offs should be run and how long they
should be is much in relation with the formation
permeability, which again is the main uncertainty.
Therefore, it is advisable to run petrophysical log to get
the near-wellbore permeability value and use it as the
initial parameter for subsequent fall-off designs.
IPTC 13440
We have observed that those fall-offs, where the
waterflooding fronts and flow periods were clearly seen,
had the investigation radius, r inv , from three up to five
times greater than the calculated radius of injected water
away from the wellbore, r inj . It is evident that at
r inv / r inj 1, the phase transition will not be observed.
Therefore, using the simple approach on homogeneous
formation parameters and cylindrical radial displacement
(the injection radius formula stated previously), for given
r inj , as a rule of thumb, one will require the investigation
radius to be at least five times greater, i.e. r inv =5 r inj .
Then, the r inv value can be recalculated to know how
long a fall-off should be in duration to cover the
abovementioned events.
Alternatively, when analyzing acquired data, if the r inj
can be estimated from actual pressure data plotted on a
log-log pressure derivative graph, where the inflection
point in the derivative line exists and shows two
stabilization plateaus, then rough comparison with the
r inj value based on perfect cylindrical displacement
equation can provide some qualitative insight into flood
front heterogeneity.
If the expected inflection point is not observed in the log-
log pressure derivative line, instead only one stabilization
plateau to water is seen, it could due to causes, such as (a)
water injection-front flattening effect, (b) overall reservoir
thinning away from the wellbore, or (c) change of water
viscosity due to thermal effect. As for the flattening of the
water front due to gravity sagging, geologic
heterogeneities, as well as lateral formation thickness
changes, the investigation and comparison of the injection
radius values is advisable to carry out against 3-D
dynamic field modeling that would incorporate fall-off
scenarios identical to as-run ones. As for the water
viscosity issue, it is related to mobility and diffusivity
components attributing to composite zone behaviour. In
theory, the mobility ratio is dictated by a ratio of relative
end-point phase permeability values. However, in reality,
when the cold water is pumpued down into formation, it
changes its viscosity as the result of the water warming up
while propagating away from the wellbore. The change
could be such, so both the mobility and diffusivity ratios
will remain close or equal unity. This effect can also be
seen in our case, see later fall-offs #9 and #13 in
Figure 16. For these fall-offs, the transition period is no
longer observed in full (Fall-offs #10, 11, 12, and 14 are
too short to make a proper conclusion). In such a case,
additional considerations should be taken to see the water
thermal effect on the success of pressure transient test.
With both the radial flow plateaus clearly identified from
log-log plot, the value of performing diagnostics comes
7
from the comparison of their ratio with the relevant ratio
for end-point phase relative permeability values, provided
relative permeability functions are known. If not, the log-
log plot could give some idea on the wettability system.
To minimize the wellbore storage effect and detect radial
flow stabilization plateau due to water at earlier stages of
the injection program, one should consider the following
actions, such as taking into account tubing internal
diameter when designing completion, setting the
downhole closing valve much near to perforation depths,
reducing the rathole volume, and use of viscous gel for
pulling back total depth.
At all stages of planning, a combination of computer-
aided well test design, single-well and full 3-D dynamic
reservoir models should be used including sensitivity
analysis for a reasonable range of parameters.
Collaboration with geologic and petrophysical disciplines
is a mandatory to identify and verify those scenarios that
might look as controversial at the beginning.
Conclusion and Recommendations
1. This paper presents examples of the challenges in
analyzing fall-off tests in the two phase injection
system. Information about wettability and relative
permeability is required in order to obtain the correct
interpretation. However, this data is not always
available. As a result, an integrated study of all
transient data with corrected rate history is required
for pressure transient analysis.
2. A three-dimensional dynamic model was built to
simulate the effect of injecting water into oil
reservoir for oil-wet, intermediate wet and water-wet
systems.
3. Radially propagated waterflooding fronts as observed
from the model were used to understand the changes
in log-log pressure derivative stabilization plateaus,
and lead to correct interpretation.
4. Field data were then analysed to corroborate the
discussed hypothesis, in order to determine whether
real-world uncertainties hinder the observation of the
waterflooding front in the pressure transient analysis.
5. Care must be taken when analyzing fall-off tests from
injectivity tests into the oil zone in the early life of an
injection well, or during field appraisal. Incorrect
calculation of absolute permeability and skin could
lead to poor development and design decisions.
References
1. Bourdet, D.: Well Test Analysis: The Use of
Advanced Interpretation Models, Handbook of
Petroleum Exploration and Production 3, Elsevier,
Netherlands (2002).
2. Bourdet, D., Whittle, T. M., Douglas, A. A., and
Pirard, Y. M.: A New Set of Type Curves
Simplified Well Test Analysis, World Oil (May
1983), pp 187-195.
8 IPTC 13440

3. Webb, T., Omar, J., Daungkaew, S., Lim, L. M., presented at the 2009 SPE European Formation
Tibbles, R., Munoz, I., Morales, H.: Overcoming Damage Conference held in Scheveningen, The
Poroelastic Behaviour in the Water Injector Wells in Netherlands, 2729 May 2009.
the Kikeh Deepwater Development, SPE 122018

List of Figures and Tables

Table 1 Input model data for simulation

Model Type: vector oriented in radial direction
Model Dimensions: 44 x 18 x 20 cells
Internal Radius: 12 in.
External Radius: 85000 ft
Total Thickness: 40 ft
Horizontal Permeability: 100 md
Vertical Permeability: 10 md
Reservoir Pressure: 4750 psia
Bubble-Point Pressure: 1750 psia
Oil Viscosity: 0.3 cP
Water Viscosity: 0.3 cP
Oil Volume Formation Factor: 1.25 rb/stb
Figure 1 Reservoir model with logarithmically refined radial grid Water Volume Formation Factor: 1.02 rb/stb

Water-Wet Relative Permeability Function Intermediate-Wet Relative Oil-Wet Relative Permeability Function
Permeability Function
1.0 1.0 1.0
0.9 0.9 0.9
Relative Phase Permeability

Relative Phase Permeability

Relative Phase Permeability

0.8 0.8 0.8
0.7 0.7 0.7

0.6 0.6 0.6

0.5 0.5 0.5
0.4 0.4 0.4
0.3 0.3 0.3
0.2 0.2 0.2
0.1 0.1 0.1
0.0 0.0 0.0
0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0
Water Saturation Water Saturation Water Saturation

krow krw krow krw krow krw

(a) (b) (c)

Figure 2 Rock wettability systems considered in the simulation: (a) water-wet; (b) intermediate-wet; (c) oil-wet

Post-Perf Injectivity Test with Fall-Off Main Injection Period Final Fall-Off

24 hours 24 hours 7 / 30 / 90 / 180 days 24 hours

t
0
Fall-Off #1 (Basic) Fall-Off #2
5000

10000
Qinj, bbl/day

Figure 3 Injectivity program considered in the simulation

IPTC 13440 9

Log-Log Comparison Plot for Oil-Wet Case, dp and dp' normalized [psi] vs. dt
Table 2 Total volume of water injected
10000

Post-perf Injectivity Test

After 24 hours 5 000 bbl
Main Injection Period (including post-perf period) 1000

After 7 days 40 000 bbl

Pressure [psia]
After 30 days 155 000 bbl
After 90 days 455 000 bbl 100
Total mobility Effective water permeability @ Sorw

After 180 days 905 000 bbl

Total mobility Effective oil permeability @ Swi
Table 3 Calculated radius of injection
10
Rock Wettability System 0.001 0.01 0.1
Time [hr]
1 10 100

Water Intermediate Oil After 1 day After 1 week After 1 month After 3 months After 6 months

After 24 hrs 47.7 ft 44.3 ft 45.1 ft (c)

After 7 days 135.0 ft 125.2 ft 127.5 ft Figure 4 Log-log comparison pressure derivative plots
After 30 days 265.8 ft 246.5 ft 250.9 ff corresponding to different wettability systems: (a) water-wet; (b)
intermediate-wet; (c) oil-wet
After 90 days 455.5 ft 422.3 ft 429.9 ft
After 180 days 642.3 ft 595.6 ft 606.3 ft
Water Saturation Distribution in Time
Log-Log Comparison Plot for Water-Wet Case, dp and dp' normalized [psi] vs. dt
1
10000 Legend:
1 - @ 17 ft; 2 - @ 22 ft; 3 - @ 29 ft; 4 - @ 37 ft; 5 - @ 48 ft; 6 - @ 62 ft;
0.9
7 - @ 80 ft; 8 - @ 102 ft; 9 - @ 131 ft; 10 - @ 168 ft; 11 - @ 215 ft

0.8
Total mobility Effective water permeability @ Sorw
1
2
0.7
3
1000 4
Water Saturation

0.6
Pressure [psia]

0.5 5

Total mobility Effective water permeability @ Sorw 0.4

100 6 7 8 9 10 11
0.3

Total mobility Effective oil permeability @ Swi

0.2

Total mobility Effective oil permeability @ Swi

0.1

10 0
0.001 0.01 0.1 1 10 100 0 2 4 6 8 10 12 14 16 18 20
Time [hr] Time [days]

After 1 day After 1 week After 1 month After 3 months After 6 months

(a)
(a)
Water Saturation Distribution in Time
Log-Log Comparison Plot for Intermediate-Wet Case, dp and dp' normalized [psi] vs. dt
1
10000 Legend:
10 - @ 168 ft; 11 - @ 215 ft; 23 - @ 275 ft; 13 - @ 351 ft; 14 - @ 449 ft;
0.9
15 - @ 574 ft; 16 - @ 733 ft; 17 - @ 935 ft; 18 - @ 1195 ft

0.8
Total mobility Effective water permeability @ Sorw

0.7

1000
Water Saturation

0.6
Pressure [psia]

0.5

10 11 12 13 14 15
0.4
16
100
0.3
17
18
Total mobility Effective oil permeability @ Swi
0.2
Total mobility Effective water permeability @ Sorw
Total mobility Effective oil permeability @ Swi 0.1

10 0
0.001 0.01 0.1 1 10 100 0 20 40 60 80 100 120 140 160 180 200
Time [hr] Time [days]

After 1 day After 1 week After 1 month After 3 months After 6 months

(b)
(b)
Figure 5 Distribution of water saturation in formation at various
distances in time, water-wet scenario: (a) within first 20 days; (b)
along the whole water injection program of 180 days
10 IPTC 13440

Water Saturation Distribution vs. Distance

1.0

0.9

0.8
Total mobility Effective water permeability @ Sorw

0.7
Water Saturation

0.6

0.5

0.4

0.3

Total mobility Effective oil permeability @ Swi

0.2

0.1

0.0
0 100 200 300 400 500 600 700 800 900 1000
Distance [ft]

1 Day 1 Week 1 Month 3 Months 6 Months

Figure 9 Rate front propagation same as in Figure 8, however,
Figure 6 Distribution of water saturation along the distance away plotted versus distance away from the injecting wellbore
from injecting well at various times, water-wet scenario
Seawater Viscosity vs. Temperature

1.4
Based on 35,000 ppm seawater @ 4800 psi
1.3
Log-Log Comparison Plot for Water-Wet Case, dp and dp' normalized [psi] vs. dt -1.031
y = 81.711x
1.2
10000
1.1

1.0 w = 0.65 cP @ modeled

injection temperature of
0.9 112 F
w = 0.36 cP @

Viscosity, cP
0.8
reservoir
0.7
temperature
1000 of 188 F
0.6 w = 0.80 1.16 cP
@ modeled injection
Pressure [psia]

0.5 temperature of 60 90 F
0.4
w = 0.68 cP @ test o = 0.32 cp @ reservoir
1-a 2-a 0.3 injection temperature of
temperature of 188 F
106 F (Well A)
100 0.2

1-b 2-b 0.1

0.0
50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 210

1-c 2-c Temperature, degree F

10
0.001 0.01 0.1 1 10 100
Time [hr]
Figure 10 Comparison of reservoir oil viscosity, and injected water
After 1 day After 1 week
viscosity.

Figure 7 Nodal points selected at the beginning (a- series), in the

middle (b- series), and at the end (c- series) of phase transition
sections on a pressure derivative log-log plot for pressure fall-offs
after 1 day (1- series) and 1 week (2- series) of injection,
respectively. For these points, rate front propagation curves were
generated, as shown in Figures 8 and 9

Figure 11 First post-perforation injectivity test

Figure 8 Rate front propagation simulated at fall-off times displayed

in Figure 7 and plotted versus model cell number. It is clearly seen
that the maximum change of rate occurs in the oil zone

Figure 12 Log-log plot for the unplanned fall off test

IPTC 13440 11

Figure 16 Log-log plot of all fall-off periods

Figure 13 Field data for Well 1. Pressure Transient recorded during
the injectivity test during and after VES frac operation Table 4 Pressure Transient Analysis Results using two radial
composite model from main fall-off tests

Fall-off name Effective permeability Skin Mobility

to water, md ratio
Fall-off #2 107 +3.8 0.34
Fall-off #4 113 +3.2 0.34
Fall-off #5 114 -1.0 0.22
Fall-off #7 112 -2.0 0.5
Fall-off #9 113 -2.4 0.5
Fall-off #11 98 -3.2 0.6
Fall-off #12 113 -2.7 0.8
Fall-off #13 114 -2.9 1.0
Fall-off #14 128 -2.6 ---

Figure 14 Log-log pressure and derivative plot of all fall-off periods kh = 20800 md-ft
k = 520 md
during and after VES frac operation Skin = 15.7

Radial flow assumed as primary

(a)

kh = 4560 md-ft
Figure 15 Injection pressure-rate history obtained from downhole k = 114 md
Skin = -1.0
permanent gauges

Radial flow assumed as primary

(b)
Figure 17 Fall-off #5 showing importance of selecting the proper
radial flow: (a) to oil; (b) to water