SPE 109544

Applications of Temperature Transient Analysis for Welltest Interpretation and

Well Productivity Evaluation
Nick Bahrami / Sharif University of Technology , Jamal Siavoshi / Schlumberger-Testing

Copyright 2007, Society of Petroleum Engineers

be influenced by skin and reservoir permeability2. Many
This paper was prepared for presentation at the 2007 Asia Pacific Oil & Gas Conference and
Exhibition held in Jakarta, Indonesia, 30 October-1 November, 2007. analysts rely on pressure derivative curve to diagnose wellbore
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
storage period and radial flow regime on pressure transient
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 data. However, there are field examples that flow regimes
publication review by Editorial Committees of the Society of Petroleum Engineers. Electronic
reproduction, distribution, or storage of any part of this paper for commercial purposes without the
written consent of the Society of Petroleum Engineers is prohibited.
can’t be accurately determined.
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 wellbore storage effect delays the formation pressure
the paper was presented. Write Librarian, SPE, P.O. Box 833836, Richardson, Texas 75083-3836
U.S.A., fax 01-972-952-9435.
response and distorts the early portion of pressure transient

Abstract data. Diagnosing of the radial-flow regime3 is crucial to

During transient tests, both pressure and temperature change quantitative interpretation since it provides values for

depending on downhole flow rate. In gas producing wells, permeability and skin. Unit slope and the plateau on the

Joule-Thomson cooling and frictional heating effects are the pressure derivative curve as well as Horner plot are usually
main dynamic factors that cause flowing bottomhole used to identify pure wellbore storage and radial flow regime

temperature to differ from the static formation temperature at as shown on Fig.1. The interpreter’s first task always is to

that depth. When a gas well is shut in, JT cooling effect is identify the unit slope line and derivative plateau to identify

vanished and this causes a sudden sharp increase in sandface flow regimes 4.

temperature. Then as the effect of wellbore storage ends, Since different factors including wellbore storage, skin and

wellbore temperature gradually cools down due to heat reservoir heterogeneity affect pressure response, detecting end
conduction with near wellbore region. of wellbore storage and flow regimes might have uncertainties.
This paper demonstrates the applications of temperature Horizontal wells, for example, pose two special problems for

transient data and proposes a new technique for using the reservoir engineer. The first is the unavoidably large

temperature transient data in gas wells in order to determine wellbore storage effect as horizontal section may extend for
end of wellbore storage. Also, the effects of permeability and thousands of feet and cannot be isolated from the transient.
well productivity on temperature behavior are discussed. Three The second is the more complex nature of transient, which

field examples are shown in which both temperature and makes diagnosis more difficult. So wellbore storage may

pressure transient data were analyzed for more accurate distort the early time flow regimes and cause uncertainty in
welltest interpretation. This paper shows how knowledge of detecting first plateau on derivative curve. Therefore,
Joule-Thomson cooling effect and frictional heating effect can determining accurate time at which wellbore storage stabilizes,

be applied for reservoir characterization. has a significant impact on the analysis and interpretation5 of
pressure data.
Introduction One way to identify wellbore storage duration is to record
In pressure transient tests1, the early portion of the well test downhole flow rate using a flowmeter such as spinner in
data is usually affected by the wellbore storage effect and can which end of wellbore storage effect can be determined, as
2 SPE 109544
downhole flow rate becomes zero. However, downhole flow
rates which are less than the spinner threshold cannot be
detected. Therefore, spinner may underestimate the end of
wellbore storage period. This technique might not always be
applicable since often a combination of pressure and
temperature gauges is run in wells. Also, a flowmeter has to be
located above producing zones during transient tests.
This paper explains how use of temperature transient data
helps measuring the time for end of wellbore storage on
pressure transient data for an accurate analysis and
interpretation.
Temperature Behavior During Transient Tests
6
A difference in temperature results in a heat transfer . Heat
is a form of energy that is transferred between objects by
conduction, convection and radiation.
In dynamic conditions, when a fluid expands at constant
enthalpy because of a pressure drop, the temperature of fluid
changes. This phenomenon is named the Joule-Thomson
effect7 which is one of the main factors controlling inflow
temperature during production. An example of this type of
process is a flow stream passing through a throttle, i.e.
perforation in casing. The change in temperature per unit
pressure is called Joule-Thomson Coefficient, JT which
is derived in most thermodynamics texts as:
 T 
 JT   (1)
 P  h
For an ideal gas, the Joule-Thomson coefficient is zero
meaning that when an ideal gas expands at constant
enthalpy, there is no temperature change while as real fluids
expand, cooling occurs if  JT is positive and warming
occurs if  JT is negative.
Typically, for natural gases up to pressure of around 5000-
7000 psia,  JT is positive during production and it ranges
from 0.01 to 0.06 °F per psia. At higher pressures, above
8000 psia,  JT might be negative for gas and raising gas
temperature during flow. The magnitude of the temperature
change is complex and depends on the fluid composition,
pressure, temperature and drawdown.
The fluid temperature is also controlled by frictional
heating9 effect. During production, as fluid flows from
reservoir towards wellbore, reservoir fluid temperature
increases due to friction between fluid molecules and porous
medium. Therefore, during production, both JT cooling (or
JT heating) and frictional heating effects are the main
factors causing sandface temperature to differ from the
original reservoir temperature at that depth. For oil
producing wells, sandface temperature is always higher than
original reservoir temperature, while for gas producing
wells, sandface temperature can be either higher or lower or
equal to original reservoir temperature due to simultaneous
effect of frictional heating and JT cooling effect.
The temperature change is essentially proportional to
pressure drop. From reservoir towards wellbore, since
reservoir fluid travels a long distance and pressure decreases
gradually, frictional heating has the dominant effect
compared to JT effect. Effect of permeability on temperature
behavior is also shown in Fig. 2. The less the reservoir
permeability, the more will be the friction effect.
In near wellbore region, since any throttling caused by
perforation or damage skin causes a significant extra
pressure drop, JT effect can be dominant. The JT cooling
effect is very localized and depends directly on near
wellbore pressure drop. The larger localized pressure drop
(high total skin), the higher JT cooling effect. Rate
dependent skin caused by high rate gas wells increases JT
cooling effect.
Analysis of Temperature Transient For Gas Wells
For gas producing wells, both the Joule-Thomson cooling
effect and frictional heating effect are main factors causing
the inflow temperature of the fluid to be different from the
geothermal temperature 10
at that depth. This implies that
flowing wellbore temperature can be either higher or lower
than static reservoir temperature depending on which
phenomenon is more dominant. The temperature log data
are useful for evaluation and diagnosis of well production
problems. A set of typical temperature profile for different
well conditions and applications are presented in Appendix
section, Figures 10-27.
SPE 109544
In pressure buildup tests, the early time pressure and
temperature data are affected by wellbore storage in which
there is a downhole flow rate from reservoir towards
wellbore. During wellbore storage, flow rate decreases
within a short period of time approaching a zero value;
causing JT cooling effect to be vanished. This will cause a
sharp increase in temperature in the early time pressure
buildup due to inexistence of JT cooling effect. A typical
wellbore temperature behavior during flow and shut-in is
depicted in Fig. 3. When wellbore storage effect ends and
there is no downhole flow rate, wellbore temperature
decreases gradually due to radial thermal heat transfer
between wellbore and reservoir. As a result, bottom hole
temperature starts cooling down over shut-in time and
showing a decreasing trend. The amount of jump in
temperature in early time build-up can provide qualitative
information on reservoir permeability, near wellbore
pressure drops and well Productivity Index. The amount of
jump in temperature can be very small in case of a high
permeability and negative skin, and can be significantly
large when there is a low permeability and/or a large
positive skin.
The application of temperature data in welltest analysis has
been shown in the field examples given below. Note that
temperature transient data are depicted on a Semi-Log scale
for shut-in time in order to be compared readily with pressure
derivative plot.
Field Examples
Multi-rate test followed by pressure buildup test was
conducted in an oil well and two gas wells. The oil well
testing was performed using Drill Stem Testing. The gas
11
wells were tested by Production Logging tool, in order to
record downhole rates, pressure and temperature. The
acquired data for oil well and gas wells are shown in Figs. 5,
6 and 7 respectively. Analyzing pressure buildup data for
the two gas wells is also illustrated in Fig. 9.
In order to illustrate the behavior of the JT effect
quantitatively for the gas wells, a plot of JT coefficient vs.
pressure at different temperatures was prepared using a
commercial PVT simulator for given gas compositions. As
shown in Fig. 4, at given pressure/temperature
3
corresponding to gas wells condition, JT coefficient has a
positive value implying that JT generates a cooling effect.
Therefore, for the gas well in the study, inversion Joule-
Thompson (warming effect) cannot occur at the given
wellbore condition.
During flow after flow, stabilized wellbore temperature is a
combination of friction and JT effects. As pressure
drawdown is increased, either a decreasing or an increasing
trend in wellbore temperature can be observed depending on
JT and friction factors. For gas producing wells, any
increase in choke size at wellhead causes a sudden pressure
drop at bottomhole. This causes a large expansion
accordingly and results in a sharp reduction in stabilized
wellbore temperature, over a short period of time, due to JT
cooling effect. This phenomenon is seen on recorded
temperature for gas wells demonstrating JT cooling effect.
Given values for skin and permeability have been estimated
from the plateau on pressure derivative curve. The
productivity indices for wells have been calculated from
flow after flow tests.
Case 1: The well is a horizontal oil producing well
completed as open hole. Figs. 5a and 5b show plots of
pressure and temperature respectively during flow and shut-
in, and Fig. 5c shows pressure and pressure derivative vs.
shut in time on log-log scale for the build-up test.
As shown in Fig. 5b, there is significant increase in
temperature during flowing periods caused by a combination
of frictional heating and JT heating effects. During shut-in
periods, since the dynamic factors are vanished, temperature
cools down immediately and shows a decreasing trend to
reach original reservoir temperature.
Case 2: This well is a vertical gas well with large openhole
producing interval. The acquired data are shown in Fig. 6.
During first flow period (18 MMSCFD) where pressure
drawdown is 22 psia, since skin effect is small, frictional
heating effect is the dominant factor and causes an increase
of 1.0 °F in temperature. When pressure drawdown
increases to 60 psia and rate is 50 MMSCFD, flowing
temperature continues to increase indicating a dominant
frictional heating effect associated with a small JT cooling
effect implying that there is no significant extra pressure
4 SPE 109544
drop around wellbore.
Temperature transient data during build up can also be used
to detect end of wellbore storage region. Figs. 6d, 6e and 6f
show spinner, temperature and pressure derivative vs. shut
in time. When the well is shut-in, JT cooling is vanished
and temperature starts increasing sharply. When WBS ends
and downhole flow approaches a zero value, frictional
heating effect is also diminished. At this time, wellbore
temperature starts declining due to heat conduction with
near wellbore region. Therefore, for gas wells, during build
up test, the wellbore storage starts when temperature starts
increasing sharply and it ends when temperature starts
decreasing. The relatively small jump in temperature during
early time pressure build-up, 0.7 °F, indicates a small total
skin, which is coherent with the results given by well test
interpretation. Temperature transient data shows that
wellbore storage ends after 0.02 hrs while the spinner under-
estimates end of WBS, 0.012 hrs, because of the spinner
threshold. Note that, the time corresponding to end of
wellbore storage on derivate curve was determined from
both pressure derivative and temperature data which was in
agreement with spinner rotation data as well. The welltest
interpretation showed an average permeability of 49 md,
average skin of –0.5 and well Productivity Index of 719
MSCFD/Psia.
Case 3: This well is a vertical perforated gas well. The
acquired data are shown in Fig. 7.
During first flow period (22 MMSCFD) where pressure
drawdown is about 80 psia, since skin effect is very small,
temperature has a significant increase, almost 2.7 °F,
indicating a high frictional heating effect generated by a low
permeable zone. When pressure drawdown increases to 600
psia and flow rate is 50 MMSCFD, JT cooling effect
increases accordingly, and becomes significant. As shown in
Fig. 5c, this causes a reduction in wellbore temperature in
higher rates. The dominant JT cooling effect implies that
there is a high pressure drop around wellbore. This is
usually the case for perforated wells in which area open to
flow is much smaller (throttling) compared to openhole
wells causing a high pressure drop. The significant extra
near wellbore pressure drop is due to rate dependent skin
caused by high flow rates and partial perforation.
Fig. 7d shows that flowmeter records zero flow rate after
0.04 hrs while wellbore temperature continues to increase.
The flowmeter has under-estimated wellbore storage
duration due to spinner threshold. Temperature behavior
during build up, as shown in Fig. 7c, depicts a 2.7 °F jump
in temperature indicating that wellbore storage effect ends
after 0.06 hrs. Welltest interpretation estimated an average
permeability of 6 md, skin of –3.3, and well Productivity
Index of 31 MSCFD/Psia. The given low permeability, 6
md, and the small skin effect (-3.3), validates temperature
analysis and the high frictional heating observed in low flow
rates. The low well productivity (31 MSCFD/Psia) also
confirms the dominant effect of rate dependent skin at high
rates.
Having accurate wellbore storage duration and knowing
temperature behavior can provide more reliable
interpretation of pressure derivate.
Discussion
Figs. 8a, 8b and 8c show a comparison between the
measurements for the two gas wells during flow after flow,
and Figs. 8d, 8e and 8f show the comparison during pressure
build-up.
Fig. 8c shows two different temperature behaviors for the
two gas wells. One well shows a frictional heating effect
over flowing periods while the other well shows a cooling
effect. The amount of pressure drawdown for case 2
indicates that JT cooling effect should be significantly
smaller compared to case 3. This implies that less cooling
effect is observed in case 2.
For low flow rates where rate dependent skin is small, JT
cooling effect will be small accordingly. This results in an
increase in temperature due to dominant frictional heating.
The low permeability zone shows more heating in low rates.
The wellbore temperature is a combination JT cooling and
frictional heating effects. There is always heating generated
by friction no matter what is the permeability. Low or high
permeability will only decrease/increase magnitude of
heating effect.
In high production rates, although permeability in case 2 is
SPE 109544 5

higher than case 3, smallness of JT cooling effect due to 8. For gas producing wells, the amount of jump in
high PI has caused frictional heating effect to be more temperature in early time of build-up provides
dominant. In contrast, for case 3, although permeability is qualitative information on well productivity index and
low, largeness of JT cooling effect due to low PI counteract near wellbore pressure drop.
the effect of frictional heating causing decreasing trend for
Acknowledgement
temperature. For case 3, throttling caused by perforation and
The authors would like to appreciate Adil Gurkan Ceyhan
rate dependent skin (low well productivity) plays a major
(Schlumberger), Thormod Ekely Johansen (Memorial
role in increasing JT cooling effect.
University of Newfoundland), and Mohammad Abshenas
Moreover, temperature jump in early time build-up in the
(Schlumberger) for many helpful discussions on this work.
low PI gas well is larger than that of in the high PI gas well.
The welltest interpretation results for the gas wells are Nomenclature
shown in Figs. 9a and 9b. The temperature transient analysis P Pressure
T Temperature
is in good agreement with the pressure data interpretation
 JT Joule-Thompson coefficient
confirming validity of applied techniques.
References
Conclusions
1. Earlougher, R.C., « Advances in Well Test Analysis »
These case studies show that temperature transient data can be
Monograph Series, SPE Richardson TX 1977.
a useful tool to analyze pressure transient data and reduce
2. Ramey, H. J. Jr., « Short-Time Well test data
uncertainties in well test interpretation. From these field
Interpretation in the Presence for Skin Effect and
examples and showing application of the temperature-transient
Wellbore Storage », JPT, Jan. 1970.
data in the context of well test interpretation, we can draw the
3. Bahrami H. , Siavoshi J., « Second Derivative Yields
following conclusions:
New Insights to Well Test Analyses », Oil&Gas Journal
1. Use of both pressure and temperature in analysis of
(Dec. 5, 2005) 46.
transient tests help reducing uncertainty in welltest
4. Economides C. E., Use of the Pressure Derivative for
interpretation.
Diagnosing Pressure-Transient behavior, SPE 18594.
2. Temperature transient analysis is a fast and reliable tool
5. Bourdet D., « Well Test Analysis: The Use of
for determining end of WBS in well test interpretation.
Advanced Interpretation Models » ELSEVIER 2002.
3. Frictional heating and Joule-Thomson cooling effects
6. Roberts, J. K.; « Heat and Thermodynamics » Blackie
control temperature during transient tests.
and Son, limited, London (1942).
4. Bottom hole flowing temperature can be either higher
7. Roger J. Steffensen and Robert C. Smith, «The
or lower than static reservoir temperature depending on
Importance of Joule-Thomson Heating or Cooling in
which phenomenon is dominant.
Temperature Log Interpretation », SPE 4636, 1973.
5. The less the reservoir permeability, the more the
8. Yoshioka K., Zue D., 2005. A comprehensive model of
frictional heating effect.
temperature behavior in a horizontal well , SPE 95656.
6. Magnitude of JT cooling effect is largely dependent on
9. Katz, Donald L.; « Thermodynamics Analysis of
total skin. The more the total skin, the more will be JT
Frictional Heat Effects in Pipeline Flow » Oil and Gas
cooling effect.
Journal (March 6, 1972).
7. In pressure buildup tests for gas wells, an increase in
10. Kabir, C., Establishing geothermal gradient using a new
temperature represents effect of wellbore storage
static temperature analysis method, SPE 38667, 1997.
period. Wellbore storage ends when wellbore
11. Schlumberger series, « Production Logging », Schlumberger
temperature starts decreasing.
Wireline and testing, Jan 1997.
6 SPE 109544

Fig. 1: Diagnosis of wellbore storage and radial flow on a typical pressure derivative curve and Horner plot.

Fig. 2: Effect of reservoir permeability on frictional heating effect during flow assuming no JT cooling effect.
The less the permeability, the more the frictional heating effect (temperature raise)
SPE 109544 7

Fig. 3: Evaluation of JT cooling and frictional heating effects on wellbore temperature during flow
and shut in for gas wells.

Fig.4: PVT properties and JT coefficient for reservoir fluid of the gas field
8 SPE 109544

(a)

(b)

K=51 md
S=-1
PI= 17 STBD/Psia

(c)

Fig 5: Flow after Flow and Pressure Build up test for case 1
(a) Pressure during flow after flow and build up
(b) Wellbore temperature during flow after flow and build up
(c) Log-Log plot of pressure and pressure derivative during build-up
SPE 109544 9

(a) (d)

(b) (e)

(c) (f)

Fig 6: Flow after Flow and Pressure Build up test for case 2
(a) Downhole flow rate during flow after flow and build up
(b) Pressure during flow after flow and build up
(c) Wellbore temperature during flow after flow and build up
(d) Spinner vs. shut-in time during build-up
(e) Semi-log plot of wellbore temperature during build-up
(f) Log-Log plot of pressure and pressure derivative during build-up
10 SPE 109544

(a) (d)

(b) (e)

(c) (f)

Fig 7: Flow after Flow and Pressure Build up test for case 3
(a) Downhole flow rate during flow after flow and build up
(b) Pressure during flow after flow and build up
(a) Wellbore temperature during flow after flow and build up
(b) Spinner vs. shut-in time during build-up
(c) Semi-log plot of wellbore temperature during build-up
(d) Log-Log plot of pressure and pressure derivative during build-up
SPE 109544 11

Flow after Flow Build-up

(a) (d)
(b) (e)

K=6 md
S=-3.3

PI=719 PI=719

K=49 md
S=-0.5 PI=31
PI=31

(c) (f)

Fig 8: Effect of dynamic reservoir properties on sandface temperature behavior for cases 2 and 3;

(a) Downhole flow rates during flow after flow test

(b) Pressure drawdown during flow after flow test
(c) Temperature change during flow after flow test
(d) Downhole flow rates during build up test
(e) Pressure change during build up test
(f) Temperature change during build up test
12 SPE 109544

(a)

(b)

Fig 9: Welltest analysis results for Pressure Build up data of cases 2 and 3

(a) Diagnosis of pressure build-up data, case 2.

(b) Diagnosis of pressure build-up data, case 3

SPE 109544 13

APPENDIX

Figure 13: Normal temperature profile (a) vs

Figure 10: Fluid Entry and temperature profile temperature profile with cross-flow (b)

Figure 14: Temperature profile in horizontal wells –

low flow rate

Figure 11: Effect of flow rate on temperature profile

Figure 15: Temperature profile in horizontal wells –

high flow rate

Figure 12: Effect of gas entry on temperature profile

Figure 16: Multiple entry of reservoir fluid in horizontal

wells
14 SPE 109544

Figure 17: Liquid Entry and temperature profile

Figure 21: Water injection wells (or hydraulically

fractured well): temperature log run after water
injection at different times, can identify the fractured
zone

Figure 18: Gas Entry and temperature profile

Figure 19: Water channeling behind casing

Figure 22: Water channeling in injection wells in

multi reservoir zones (or multi-stages fractured
wells): running temperature log at different times
after water injection (or after fracturing) is completed,
can identify the target reservoir zone as well as water
channeling behind casing into the other reservoir
zones
Figure 20: Gas channeling behind casing
SPE 109544 15

Figure 23: Effect of liquid level in wellbore on

temperature log response

Figure 26: Temperature log in a multi-layered oil

well: Flowing conditions with no significant cross
flow, versus, shut-in conditions with significant
cross-flow (into the thief zone)

Figure 24: Effect of water re-circulation in wellbore (a)

on velocity profile (b) and temperature log response (c)

Figure 27: Temperature log in gas wells: a low

productivity index well (a), versus, a high
productivity index well (b)

Figure 25: Temperature log in a gas well with liquid re-

circulation downhole
(flowing vs shut-in)