Professional Documents
Culture Documents
Fiber Optic Distributed Temperature Analysis Book
Fiber Optic Distributed Temperature Analysis Book
Fiber Optic Distributed Temperature Analysis Book
Distributed Temperature Analysis
Links are provided on each page to ease
navigation. This PDF may also be browsed
normally.
From the Contents pages, any of the listed
items may be accessed by clicking either
the entry or page number. This eBook is
also bookmarked and thumbnailed.
Click here for the table of contents.
For help using Adobe Acrobat Reader,
press the F1 key or click here to access
Adobe Acrobat online help.
Second Edition
George Brown
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v
Fiber-Optic Distributed Temperature Measurement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Single-ended versus double-ended temperature measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
Thermal resolution and statistics reduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Fiber-Optic Deployment Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
Permanently installed with the completion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
Slickline deployment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Coiled tubing deployment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Oil Well Thermal Basics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Geothermal gradient . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Joule-Thomson effect . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
Flow up the well . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Single-Phase Geothermal Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Multizone geothermal reservoirs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
Example: Quicklook interpretation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
THERMA* Thermal Modeling and Analysis DTS Software . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Thermal transient model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
PIPESIM* steady-state multiphase flow simulator model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
Coupling the thermal transient and wellbore flow models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
Discretization and solution of the finite-difference equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Automatic optimization of parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
DTS data loading and manipulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Multizone Analysis Using Thermal Modeling Software . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
Example: Multizone flow analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
Reservoir depletion in a multizone reservoir . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
Example: Reservoir depletion in a multizone reservoir . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
Effect of Fiber Location . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
Example: Reservoir depletion identified by a fiber located on the outside
of sand screens . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
Multiphase Flow in Multizone Reservoirs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
Water breakthrough . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
Increasing oil GOR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
Example: Increasing GOR on a horizontal ICD and sand screen completion . . . . . . . . . . . . . . . . 32
Gas Wells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
Low-rate gas wells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
Example: Low-rate multizone gas well . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
High-rate gas wells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
iv
Temperature logs have been used to monitor producing wells since the early 1930s (Doll and
Perebinossoff, 1936) and a considerable number of papers have been dedicated to the measurement
and its analysis over the years. The data they generate has been used to calculate the flow
contribution in both oil and gas wells and also to evaluate water injection profiles, determine the
effectiveness of fracture jobs, and identify cement tops, crossflow between zones, flow outside the
casing, and other flow and wellbore-related events.
Normally analysis of the temperature log is viewed as secondary to that of the spinner flowmeter,
which gives flow velocity directly, and temperature is conventionally used only as an indicator of
gas influx with the production logging tool (PLT). The main disadvantage of the PLT temperature
log is that only one thermal profile is obtained at the time the PLT is run, and this makes analysis
difficult on a stand-alone basis.
However, the Schlumberger fiber-optic distributed temperature sensing (DTS) system can
generate many temperature logs (from every two seconds upward) over the life of the well either
preinstalled with the completion or temporarily deployed on slickline or coiled tubing. This
enables well productivity to be evaluated without the need for other log data. In addition, the
system identifies the source of a change in well performance when it occurs, rather than having
to wait for a production log to be run.
In recent years, the cost of PLT logs has increased considerably because many wells are now
drilled horizontally through the reservoir and the PLT tools must be conveyed on coiled tubing or
by well tractors, and in some cases (subsea wells) even this may not be possible. Consequently,
alternative technologies become viable if they can be installed in wells at reasonable cost
and the data generated can be interpreted to give flow and other production information. The
Schlumberger DTS system is just such a technology.
This book, now in its second edition, outlines the current state of the art of distributed time- and
depth-based temperature analysis for use with the Schlumberger DTS system.
George Brown
Schlumberger Advisor
2016
Distributed temperature sensing (DTS) is a fiber-optic The algorithm employed to calculate temperature
distributed temperature measurement that uses an from the Stokes (NTS) and anti-Stokes (TTS) Raman
industrial laser to launch 10-ns bursts of light down bands is
an optical fiber. During the passage of each packet of
light, a small amount is backscattered from molecules
in the fiber. This backscattered light can be analyzed
1 1 ln [(TTS/NTS)x /( TTS/NTS)c] ,
= − (1)
to measure the temperature along the fiber (Al-Asimi Tcal Tref SLE
et al., 2002). Because the speed of light is constant, a
spectrum of the backscattered light can be generated where
for each meter of the fiber using time sampling, thus Tcal = calculated temperature (K)
allowing a continuous log of spectra along the fiber to be Tref = reference temperature (K)
generated (Fig. 1). SLE = temperature sensitivity
A physical property of each spectrum of the back-
scattered light is that the ratio of the Stokes Raman to (TTS/NTS)x = anti-Stokes/Stokes ratio at the point of
the anti-Stokes Raman bands is directly proportional interest along the fiber
to the temperature of the length of fiber from which it (TTS/NTS)c = anti-Stokes/Stokes ratio in the reference
is generated. Consequently, temperature can be calcu- coil.
lated every meter along the whole length of the fiber Temperature trace acquisition times can be varied from
using only the laser source, analyzer, and a reference as few as two seconds to several hours, and this frequently
temperature in the surface system; there is no need for determines the accuracy and resolution of the measured
any calibration points along the fiber or to calibrate the temperature log. Typically, a resolution of 0.05 degC is
fiber before installation. required for reservoir surveillance.
Anti-Stokes Stokes
Raman Raman
band band
Analyzer
Wavelength
30,000
Wellhead Wellhead
25,000
TTS 1
20,000
TTS 2
Counts 15,000
Bottom of well
10,000
NTS 1 NTS 2
5,000
0
0 5,000 10,000 15,000 20,000 25,000 30,000
Fiber length, ft
180
Double-ended (loss corrected)
Single-ended (constant DLC)
170
178
Temperature, 160 Loss-induced Loss-induced
degF 176
shift slope change
174
150
172 Absolute
error
170
8,000 10,000 12,000 14,000
140
0 2,000 4,000 6,000 8,000 10,000 12,000 14,000
Depth, ft
Figure 3. Comparison of single-ended and differential-light-loss-corrected (double-ended) measurements’ thermal resolution and statistic reduction.
100
4-km DTS ASE
4-km DTS Ultra SE
8-km DTS Ultra DE
10
Acquisition time,
min
1
0.1
0.01 0.1 1
Root-mean-square resolution, degC
Figure 4. DTS calibration acquisition time versus rms thermal resolution (DE = double-ended, SE = single-ended).
Table 1. WellWatcher Ultra, WellWatcher Ultra ASE, and WellWatcher Hyperion System Specifications
DTS System Application Range, Spatial Accuracy, Rms Thermal Resolution
mi [km] Resolution, degF [degC]
ft [m]
WellWatcher Ultra General oilfield 7.5 [12] 3.28 [1] ±1.8 [±1] 6.2-mi fiber = 0.72 degF with 12-s acquisition time
[10-km fiber = 0.4 degC with 12-s acquisition time]
WellWatcher Ultra ASE SAGD 3.7 [6] 3.28 [1] ±7.2 [±4] 3.7-mi fiber = 0.54 degF with 10-min acquisition time
[6-km fiber = 0.3 degC with 10-min acquisition time]
WellWatcher Hyperion Remote wells 2.5 [4] 3.94 [1.2] ±5.4 [±3] 2.5-mi fiber = 0.18 degF with 10-min acquisition time
[4-km fiber = 0.1 degC with 10-min acquisition time]
Temperature,
degF
155.8–157.7
154.0–155.8
150 152.1–154.0
Temperature, 150.2–152.1
148.4–150.2
degF 140 146.5–148.4
144.7–146.5
130 May 3 142.8–144.7
140.9–142.8
May 2 139.1–140.9
X 137.2–139.1
May 1
X+1,000 135.4–137.2
X+2,000 April 30 133.5–135.4
131.7–133.5
X+3,000 Time 129.8–131.7
Depth, ft X+4,000 April 29
X+5,000
April 28
Figure 6. Depth-averaged and toe-normalized dataset.
Permanently installed with the completion reduce the possibility of hydrogen in the water causing
the fiber to deteriorate.
Figure 7 shows three options for installing optical fiber
Optical fiber can also be installed in an electrical
permanently with the completion in an oil well. With the
pressure gauge cable using the WellWatcher Neon* DTS,
single- and double-ended pumped system, a ¼-in-diam-
distributed acoustic sensing (DAS), and pressure and
eter control line is first installed with the completion
temperature (PT) gauge system (Fig. 8). The fiber does
and then the fiber is pumped into the control line using
not have to stop at the gauge; an electrical connection
water. A hydraulic control line wet-mate connection can
to the gauge can be split off the cable and the fiber
be used to enable running the completion in two stages.
continued down over the reservoir, if required.
Once in place, the water is backfilled with silicone to
Electric
cable
Hydraulic
control line Gauge
wet-mate
Check valve
Turnaround
sub
Single-ended pumped Double-ended pumped Electric cable
Figure 7. Completion permanently installed fiber options.
Polymeric jacket
Filler rod
Conductor insulation
Optical fiber
18 AWG stranded conductor
Filler rod
OD, in 0.125
INCOLOY® 825-clad Carbon fiber composite
strength member strength member
Max. temperature, degF 240
Weight in air, lbm 32 (per 1,000 ft) SS316 optical-fiber
Weight in water, lbm 27 (per 1,000 ft) tube protection
The key to understanding temperature logs in oil and gas While a well is being drilled, injection or circulation
wells is knowing how the fluids flowing in the well gain of fluids changes the temperature of the near-wellbore
or lose heat due to the external effect of the geothermal region, and sufficient time must be allowed for these
gradient and the internal Joule-Thomson fluid effect. transient effects to dissipate before the real geothermal
The combination of these effects creates a characteristic gradient can be identified. This may take days or even
time-dependent thermal profile that can be recorded weeks to fully dissipate, so knowledge of the near-time
using the Schlumberger DTS system and analyzed to give well history is important for defining the geothermal
the flow in the wellbore. gradient.
If the well has been producing or injecting for some
time (i.e., weeks or months), it can take a similar time
Geothermal gradient to return to the true geothermal gradient when it is shut
in. If there is no preproduction geothermal gradient
The center of the Earth is very hot whereas out in space
defined, this can be extrapolated in the nonreservoir
it is very cold. Through conduction, the Earth is con-
intervals by using a Horner time function at selected
stantly losing heat through its crust (Fig. 11). Because
depths during a short shut-in, in the same fashion that
the Earth’s crust is made up of layers of rock with slightly
a pressure gauge can be used to determine the reservoir
different geological, petrophysical, and thermal proper-
shut-in pressure.
ties, the actual geothermal gradient is not necessarily a
Typical geothermal gradients vary from 0.6 to
straight line.
1.6 degF/100 ft [1.0 to 3.0 degC/100 m] with a typical
Accurate definition of the geothermal gradient is
average value of 1.0 degF/100 ft [2.0 degC/100 m] (Figs.
essential for calculation of the heat loss as fluids travel
A-1 and A-2 in the Appendix).
up the wellbore from the reservoir and heat is lost to
cooler surroundings.
Very cold
0 degC
Layers of rock with different thermal properties
Space
Geothermal
gradient
Surface
Earth
Geothermal
gradient
10
Gas
Water As oil and water flow
to the surface, the oil
Oil
volume shrinks (Bo)
and gas comes out
of the solution below
the bubblepoint
Bubblepoint pressure Pb.
pressure Pb
This change in volume
and fluid properties
must be accounted for
in thermal modeling.
Water
Oil
The Essentials of Fiber-Optic Distributed Temperature Analysis ■ Oil Well Thermal Basics 11
Geothermal gradient
1,000
Depth,
ft 3,000
z
4,000
5,000 Reservoir
Tge
50 75 100 125 150
Temperature, degF
Figure 14. Temperature response for upward single-phase flow. Tgz = geothermal temperature at height z above the producing reservoir.
Ge 1,000 bbl/d
ot
2,000 he
rm
al
gr ad
Depth, ie n
ft 3,000 t
4,000
5,000
Reservoir
50 75 100 125 150
Temperature, degF
Figure 15. The effect of changing flow rate on the temperature log.
0
0.5 day 500 days
1,000
Ge 5 days 50 days
ot
2,000 he
rm
al
gr ad
Depth, ie n
ft 3,000 t
4,000
5,000
Reservoir
50 75 100 125 150
Temperature, degF
Figure 16. The effect of time at constant flow rate on the temperature log.
14
1,000
Total flow
Ge
oth
2,000 er m
al g
r adi
Depth, en t
ft 3,000
Flow rate
Tt Tl
Upper reservoir
4,000
Tg
1,000
Ge
2,000 oth
er m
al g
r adi
Depth, en t
ft 3,000
Tt
Upper reservoir Tl
4,000
Tg
Lower reservoir
5,000
16
A limitation of the Ramey (Curtis and Whitterholt, 1973) relaxation times. Although, in some cases, such as gas- or
equations is that they are valid only for incompressible liquid-level equilibration upon shut-in, this assumption
fluids such as water or diesel, whereas in oil wells, the may be invalid, in general it is adequate for the vast
oil and gas are compressible and this must be taken into majority of cases involving production or injection flows
account to achieve accurate temperature calculations in in the reservoir zone.
the wellbore. In addition, the Ramey equations do not
allow for the effect of Joule-Thomson warming or cooling
of oil or gas flowing from the reservoir, which must also Thermal transient model
be accounted for in any solution.
A 2D radial grid is created to represent the completion
Ideally, any thermal model should account for the
and formation elements. All elements in the model must
transient nature of the temperature response while
satisfy a general energy balance. For all static (i.e., not
also accounting for the transient pressure response at
containing flowing fluids) elements, the basic energy
the same time. The OLGA* dynamic multiphase flow
balance equation relates the rate of change of tempera-
simulator can model fully both transient temperature
ture to the Laplacian of the temperature and thermal
and pressure flow from the reservoir. However, it is not
energy source terms. The application to fluid elements
suited for DTS analysis, which normally requires turn-
relates the thermal properties of a volume element to
around times similar to those of production log analysis.
the density and the thermal heat capacity at constant
To meet this need, Schlumberger developed THERMA*
pressure. In an isotropic medium, the conductivity is
thermal modeling and analysis DTS software, a dedi-
defined for a general element in a completion, and there
cated analysis engine that couples a finite-difference
can be different thermal transfer coefficients across
transient thermal model with the PIPESIM* steady-state
each element boundary.
multiphase flow simulator (Fig. 19). The model assumes
that the time taken to achieve steady-state pressure
flow can be neglected compared with typical thermal
Flow
Reservoir pressure
Reservoir layer
Reservoir layer
18
The Essentials of Fiber-Optic Distributed Temperature Analysis ■ THERMA Thermal Modeling Software 19
Limitations of the quicklook approach are that it does Table 3. Parameters to Create the THERMA Software Model in Fig. 20
not take into account the Joule-Thomson heating or Zone 1 Zone 2 Zone 3
cooling of the oil or gas exiting the formation or the MD top, ft 9,100 9,400 9,700
change in fluid properties due to pressure drop as the MD bottom, ft 9,200 9,500 9,800
oil or gas flows up the well and mixes with flow from the Color
higher zone. To achieve this requires a thermal model of Horizontal permeability, mD 200 200 200
the reservoir and wellbore that solves for convection and Vertical permeability, mD 20 20 20
conduction in the near-well reservoir under radial Darcy Reservoir pressure, psi 4,075 4,165 4,255
flow conditions (or, in a horizontal well, Joshi flow) and Formation type Sand Sand Sand
conduction between the wellbore and surrounding rock Skin 0 0 0
in the nonreservoir interval. Drainage radius, ft 1,000 1,000 1,000
Figure 20 shows the THERMA software results of a Reservoir thickness, ft 100 100 100
thermal model that simulates three producing zones, Model type Vertical Vertical Vertical
each with the same permeability and on the same Geothermal temperature, degF 199.75 204.08 208.39
oil-pressure gradient. The parameters used are listed in Oil type 30 API 30 API 30 API
Table 3. The Joule-Thomson inflow temperature (at the GOR, ft3/bbl 300 300 300
sandface) is shown by the thin, dark green trace, and the
Water cut, % 0 0 0
axial mixture temperature (at the center of the well) is
shown by the red trace.
The model highlights the effect of the Joule-Thomson
inflow temperature on the overall axial flowing tempera-
ture, measured with a production log or fiber located at
the center of the wellbore.
9,000
ie n
t
Depth, ft
9,500
Joule-Thomson
inflow temperature
10,000
180.0 190.0 200.0 210.0 220.0
Temperature, degF
Figure 20. Thermal model of multizone flow.
X+2,000
Joule-Thomson inflow temperature
X+2,400
148 152 156 160 164 168
Temperature, degF
X+2,000
Joule-Thomson inflow temperature
X+2,400
148 152 156 160 164 168
Temperature, degF
Figure 22. Thermal model of multizone well compared with PLT spinner data.
22
Base model
er m
temperature
al
Depleted
g
Flow rate
r ad
flow rate
ie n
9,000
t
9,500
Reduced Joule-Thomson
inflow temperature
10,000
180 190 200 210 220
Temperature, degF
Figure 23. Thermal model of multizone well with reservoir depletion.
The Essentials of Fiber-Optic Distributed Temperature Analysis ■ Multizone Analysis Using Thermal Modeling Software 23
Well flowing
Inflow zones
64.9–65.6
X+200 64.2–64.9
63.6–64.2
62.9–63.6
X+400 62.2–62.9
61.5–62.2
60.8–61.5
X+600 60.1–60.8
59.4–60.1
Depth, ft 58.7–59.4
X+800 58.0–58.7
57.3–58.0
56.6–57.3
X+1,000 56.0–56.6
55.3–56.0
March 9
March 8
X+1,200 March 7
March 6
Geothermal gradient
60 March 4
Temperature, degC March 3
65
Figure 24. Correlation of major inflows with the resistivity log.
24
Geothermal
gradient March flow profile
X+200
Inflow zones
X+800
Flow stops
X+1,000
X+1,200
125 130 135 140 145 150 155
Temperature, degF
Figure 25. Production changes over time.
X+200
X+400
March zone pressures
X+600
X+1,000
X+1,200
X+1,400
The Essentials of Fiber-Optic Distributed Temperature Analysis ■ Multizone Analysis Using Thermal Modeling Software 25
Resistivity, ohm.m
Crossflow on shut-in
X+800
Flow stops
X+1,000
X+1,200
125 130 135 140 145 150 155
Temperature, degF
Figure 27. Crossflow during shut-in.
26
Previous examples in this book are for fiber located in The purple dashed trace in Fig. 29 shows the tempera-
the center of the well either on a stinger in a permanent ture response expected with this installation scenario.
completion or on slickline or CT. However, it is possible In the reservoir sections, the Joule-Thomson inflow
to install optical fiber on the outside of the completion temperature is measured directly by the fiber. Between
as shown on the outside of gravel-packed sand screens the reservoir sections, the axial flowing temperature is
in Fig. 28. measured. The Joule-Thomson inflow temperature ΔT
In this type of installation, the temperature at which is a function of the drawdown Δ P and Joule-Thomson
fiber responds to the reservoir intervals is the tem- coefficient μJT only:
perature of the oil exiting the formation (i.e., the
T = μ JT P. (9)
Joule-Thomson inflow temperature).
Sand screen
Gravel pack
Screen shroud
Sand screen
oth
DTS temperature
er m
al
Base model
g
r ad
9,000 temperature
ie n
t
Depth, ft
Model flow rate
9,500
Joule-Thomson
inflow temperature
10,000
180 190 200 210 220
Temperature, degF
Figure 29. Expected temperature response with fiber located on the outside of sand screens.
Flow profile
X+400
X+500 Joule-Thomson
inflow temperature
X+600
65.0 67.5 70.0 72.5 75.0
Temperature, degC
0 25 50 75 100
Gamma ray, gAPI
Figure 30. Depletion from fiber located on the outside of sand screens.
28
In addition to reservoir depletion, two other production causing a higher zone drawdown. To maintain the upper
scenarios that need consideration are the breakthrough and lower zones with the same drawdown for comparison
of water and that of gas. PLTs have two measurements purposes, reduction of the total flow rate is necessary.
that are used to solve for water cut and GOR change—the The net effect is to produce less water and oil from the
spinner log and Gradiomanometer* specific gravity profile zone than the base dry oil model. Also, the Joule-Thomson
tool and the Digital Entry and Fluid Imaging Tool (DEFT) inflow temperature of the produced fluid decreases
or GHOST* gas holdup optical sensor tool—whereas because the Joule-Thomson coefficient of water is less
with DTS there is only one measurement available: than that of oil. Consequently, the mixture temperature
temperature. This makes identification difficult where the does not decrease as much as it does in dry oil.
changes are small. Also, a thermal model of the formation However, if the water cut of the middle zone is
must take into account the change in relative permeabil- increased further to give a 20% surface water cut, then
ity when the water saturation increases, so the effective with more water being produced and an even lower Joule-
permeability is no longer a constant. Thomson temperature, the temperature effect is greater
than that of the dry oil model and the axial temperature is
now to the left to the dry oil model (Fig. 33).
Water breakthrough Obviously, there is a water cut where there is no
temperature change between the breakthrough and dry
In a typical set of relative permeability curves (Fig. 31),
oil model, and it varies depending on the relative perme-
as the water saturation increases, the permeability to
ability characteristics of the reservoir rock. Consequently,
oil decreases and the permeability to water increases.
it is difficult (and occasionally impossible) to determine
However, the effective permeability of the mixture also
the zone of water breakthrough in a producing well from
varies.
temperature alone, even where there is good before- and
The theoretical model of a three-zone formation in
after-breakthrough DTS data and no great change in the
Fig. 32 shows the effect of increasing the middle-zone
reservoir pressures over the breakthrough period.
water cut to give 5% surface water cut on the temperature
profile. As the water cut increases, the fluid viscosity can
decrease, but the relative permeability also decreases,
1.2
1.0
Oil relative permeability kro
0.8
Water relative permeability krw
Relative 0.6
permeability
0.4
0.2
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Water saturation
Figure 31. Water and oil relative permeability data.
Ge
9,000
ot
he
rm
al
g
ra
Depth, ft Breakthrough model
die
Base flow rate
n
temperature
t
9,500
Breakthrough
Joule-Thomson
inflow temperature
10,000
180 190 200 210 220
Temperature, degF
Figure 32. Baseline versus 5% water-cut thermal models.
Base model
temperature
Ge
9,000
ot
he
rm
al
g
ra
temperature
t
9,500
Breakthrough
Joule-Thomson
inflow temperature
10,000
180 190 200 210 220
Temperature, degF
Figure 33. Baseline versus 20% water-cut thermal models.
30
Ge
9,000
ot
he
rm
al
g
ra
Depth, ft Base High-GOR model
flow rate die
n
temperature
t
9,500
High-GOR High-GOR
gas rate Joule-Thomson
inflow temperature
10,000
180 190 200 210 220
Temperature, degF
Figure 34. Baseline versus gas breakthrough thermal models.
The Essentials of Fiber-Optic Distributed Temperature Analysis ■ Multiphase Flow in Multizone Reservoirs 31
Joule-Thomson inflow
temperature (outside screens) ΔT = μJTΔP
175.5–175.6
174.6 175.4–175.5
175.4–175.4
174.8 175.3–175.4
Temperature, 175.0 175.2–175.3
degF 175.2 175.1–175.2
175.1–175.1
175.4 175.0–175.0
March 30 175.9–175.9
175.6
February 27 175.8–175.8
175.7–175.7
X+2,000 January 21 175.7–175.7
175.6–175.7
December 15 175.5–175.6
X+1,000
175.4–175.5
Depth, ft November 08
X
October 02
Figure 36. Increasing GOR on a sand screen completion (displayed upside down to show the temperature decreases).
32
The difference between oil well temperature profiles and is 13 degF, of which half occurs in the region close to the
those of gas wells is that gas wells may exhibit large Joule- wellbore. The effect, however, varies depending on the
Thomson cooling effects as the gas enters the borehole, and gas properties and reservoir pressure because the Joule-
gas does not carry as much heat up the well as oil or water. Thomson coefficient of gas is pressure dependent. Below
However, although the initial wellbore flowing temperature about 5,000 psi (depending on the gas properties) the ther-
is lower than the geothermal gradient, the equations mal and pressure radial responses have a similar character,
defining the thermal profile from the reservoir up the well at about 5,000 psi the Joule-Thomson coefficient becomes
are the same as for oil and water (Fig. 37). zero, and above 5,000 psi the Joule-Thomson coefficient
Figure 38 shows the Joule-Thomson cooling effect for changes sign and increases so gas heats rather than cools
radial flow of gas at steady-state conditions for a flow rate when subject to a pressure drop.
of 10 MMcf/d in a 50-ft-thick interval of 10-mD rock. This
Gas rate, MMcf/d
0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
0
Calculated temperature
1,000
2,000
Geothermal gradient
Depth, ft 3,000
4,000
Flow profile Joule-Thomson
inflow temperature
5,000
1,000
Q=
(
C Tg Tw ), (10)
dT /dH
2,000
Geothermal dT/dH where
Depth, ft gradient
3,000 Q = flow rate
C = time-dependent constant
Tw Tg Tg = geothermal temperature
4,000 Tw = well temperature
dT/dH = slope of intercept
5,000
50 100 150
Temperature, degF
Figure 39. Kunz and Tixier’s 1995 subtangent method.
3,500
Calculated Geothermal gradient
temperature
4,000
Depth, ft Flow
profile
4,500
5,000
Joule-Thomson
inflow temperature
5,500
80 100 120 140 160
Temperature, degF
Figure 40. Low-rate gas well.
34
Depth, ft X+2,000
Joule-Thomson
inflow temperature
X+3,000
3,500 0 30 60 90 120
Calculated temperature
Flow
distribution
4,000
Ge
oth
er m
Depth, ft 4,500
al
g
r ad
Joule-Thomson
ie n
t
inflow temperature
5,000
5,500
80 100 120 140 160
Temperature, degF
Figure 42. High-rate (10 MMcf/d per zone) multizone gas well.
through the inflow temperature of the lowest producing Figure 43 compares a conventional completion and
zone (assuming that the producing zones are close to one a velocity string completion. Another feature of velocity
another). Individual zone values can be further defined string wells is that there is no need for a packer, so they
by monitoring the zone warmback temperatures during a can be flowed up the annulus if required. Flowing the
well shut-in and extrapolating these temperatures back well up the annulus for a short time (e.g., a day) is suf-
to the flowing conditions. ficient to generate a stable temperature profile that can
Depletion of a zone over time reduces its flow rate be analyzed to give the flow profile without the effects of
relative to the other zones. If the initial reservoir counterflow degrading the temperature response. This
pressures are known and a thermal model is fitted to annular flow period is normally short enough that the
this data, later models will require only a change in the well does not fill up with water. With this scenario, it
surface flow rate and adjustment of the individual zone is possible to run an optical fiber in a slickline system
pressures to achieve a flow solution. down inside the tubing and record both flowing and
shut-in temperature profiles.
Typical velocity string gas wells in the Deep basin
Velocity string gas wells in Western Canada are produced from multiple zones
in low-permeability rock that has been hydraulically
Some low-rate gas wells produce a small amount of
fractured. It is important for the operator to know which
water over time, which eventually fills up the bottom
zones are flowing and by how much, and the only way
of the borehole and prevents the gas from flowing. A
to do this with conventional PLTs is to pull the tubing
solution to this problem is to install a “velocity” string.
above the reservoir and flow it conventionally, which is
This is a small-diameter tubing set down to the bottom
cost prohibitive. The cost-efficient solution is to convert
of the reservoir that acts like a vacuum cleaner, sucking
the well to annular flow for 24 hours before the logging
up the water with the gas and thus allowing the well to
job and then record DTS for a few hours of steady flow
flow continually. However, because the tubing extends
and 6 to 12 hours of shut-in using fiber deployed inside
below most of the reservoir zones, it is not possible to
slickline, with a pressure gauge attached to the end of
run a conventional production log to evaluate the zone
the slickline.
contributions. Also, because the flow from the reservoir
Analysis is then performed using a conventional gas
is down the annulus and then up inside the tubing, the
thermal model. Because these wells have been producing
recorded temperature profile is difficult to interpret
for some time and different zones have depleted at
because flow down the annulus exchanges heat with the
different rates, a uniform gas pressure gradient cannot
flow going up inside the tubing. The result is a tempera-
be normally used for the reservoir pressures. However,
ture profile with greatly reduced temperature character.
36
Packer
X Geothermal gradient
Depth, m
X+400
Calculated
gas rate
Flowing temperature
X+600
X+800
70 80 90 100 110 120
Temperature, degC
5 10 15 20 25 30 35 40 45 50 55
Spinner, rpm
Figure 44. Velocity string analysis compared with a PLT spinner log.
50,000 204
30,000 200
Flow rate, Geothermal gradient Temperature,
bbl/d degF
20,000 198
0 194
4,000 5,000 6,000 7,000 8,000 9,000 10,000
Depth, ft
4½-in completion
16,000 210
4,000 180
Flow distribution
0 170
4,000 5,000 6,000 7,000 8,000 9,000 10,000
Depth, ft
Openhole completion
40
Openhole completion
235
230
Geothermal gradient
225
220
Temperature,
Change over three months degF
215
Joule-Thomson cooling
210
205
200
X X+1,000 X+2,000 X+3,000 X+4,000 X+5,000 X+6,000
Depth, ft
Figure 47. Reservoir Joule-Thomson cooling as a function of well cleanup.
Openhole completion
3,500 235
3,000 230
Geothermal gradient
2,500 225
2,000 220
Flow rate, Temperature,
bbl/d 1,500 215 degF
0 200
X X+1,000 X+2,000 X+3,000 X+4,000 X+5,000 X+6,000
Depth, ft
Figure 48. THERMA software calculated flow as a function of well cleanup.
Hydraulically stage-fractured
horizontal shale wells
In 1991, Mitchell Energy completed its first horizontal Well shut-in Well flowing
well in the Barnett Shale in North Texas and, in 1998,
economically fractured a shale gas well. What followed Figure 49. Flowing and shut-in temperature response.
this success is well known. Shale gas activity has grown
to levels unimaginable 10 years ago, and now oil is also
being produced from similar formations. This growth of the spinners; often in horizontal wells that are cased
and the need for reliable methods to improve efficiency and perforated or with slotted liners or even tubing with
and well life have opened the door for the application of packers, the range of spinner logs is limited. Gas can flow
new technology, including the use of CT DTS surveys for along outside the casing if it is poorly cemented or above
production logging. the slotted liner or tubing in these types of completions.
Not all shale gas wells can be production logged to Temperature analysis reflects the true reservoir flow,
determine where the fractures are located and their flow irrespective of the effects of fluid stratification, whereas
contribution. Of course, if the completion has tubing and spinner analysis may not in some cases.
packers along the reservoir interval, the fractures may More recently, oil has been produced from
occur away from the valves, and spinner logs will show hydraulically stage-fractured horizontal shale wells. In
only the flow through the valves. this case, theory predicts that there should be no large
The advantage of distributed temperature is that Joule-Thomson cooling events at the fractures as for gas
when the well is shut in, the locations of the fractures wells.
can be determined because they have been cooled by However, experience shows that there are cooling
the flow of Joule-Thomson cooled gas during the flow events occurring at the fractures, and these are attributed
period. This cold event remains for a long time once the to the fact that when the fracture is created, a significant
well has been shut in, irrespective of the completion in amount of cold fluid and proppant is injected into the
the borehole. wing of the fracture. This causes the fracture to be
When the well is flowing, the cold gas produced from several degrees below the geothermal temperature for
the fractures flows along the wellbore, warming toward an extended period of time (i.e., months), so when oil
the geothermal temperature as it goes until it mixes with is flowed through the fracture, it is cooled and exits the
gas emanating from the next fracture (Fig. 49). fracture below the geothermal temperature. Clearly, this
The analysis procedure is exactly the same as for phenomena also occurs in gas wells, but the effect is
hydraulically fractured velocity string gas wells, which masked by the Joule-Thomson cooled gas.
are generally vertical whereas the fractured shale gas Thus, it is possible for a few months after a horizontal
wells are horizontal. oil well has been hydraulically fractured to identify the
Figures 50 and 51 show a large difference in the fractures by their cold signature during a shut-in. By
predicted temperature profiles for a hydraulically using the fracture shut-in temperature to define the
stage-fractured horizontal well producing at a rate of magnitude of the cooling occurring inside the fracture,
1.0 MMcf/d with either heel- or toe-biased flow after the it is possible to calculate the flow profile along the
fracture job. wellbore in a similar fashion to gas well analysis.
Where both DTS and multispinner logs have been
run together, the comparison between the results is
generally good. However, it must be remembered that the
multispinner logs reflect only the flow within the diameter
42
2.5
0.5
Flow profile
0 235
10,000 11,000 12,000 13,000 14,000 15,000
Depth, ft
Figure 50. Hydraulically stage-fractured horizontal well with heel-biased gas flow.
4.0 255
Staged hydraulic fractures
3.5
Geothermal
3.0 gradient 250
0.5
Flow profile
0 235
10,000 11,000 12,000 13,000 14,000 15,000
Depth, ft
Figure 51. Hydraulically stage-fractured horizontal well with toe-biased flow.
Distributed temperature surveillance has been tra- Once injection ceases, the rock surrounding the well
ditionally used to monitor the performance of water that has been cooled only by conduction through the
injectors by employing a technique called warmback. casing warms back to the geothermal gradient over a
This involves shutting the well in for a period of time and short period of time as a function of the formation’s
recording the temperature response while the reservoir thermal properties (Fig. 52). However, if a permeable
warms back toward the geothermal gradient. interval has been taking cold injected water, this will
have cooled the rock to a much greater radius from
the wellbore than in the impermeable zones. These
Warmback analysis intervals warm back at a much slower rate than those
that have not been taking injection fluid. The magnitude
Under normal injection conditions, the cold water injected
of this effect is a function of the injection rate, interval
into the well cools all the surrounding rock, including
permeability, time, and the thermal properties of the
the nonpermeable intervals above the reservoir. So, the
fluid and rock.
only information that can be obtained when injection is
Figure 53 shows the modeled response of a 500-ft-thick
occurring is the lowest extent of fluid injection. If the
interval of 10-mD rock that includes two higher-permea-
injection rate is low, the character of the profile should
bility intervals of 20 mD and 30 mD; the water injection
be determined by the geothermal gradient and inflow into
rate is 5,000 bbl/d at 80 degF for 10 days. Because the
the reservoir and could be analyzed as such. However, this
higher-permeability intervals have been taking more
is not normally the case.
1,000
100-day warmback
Completion
2,000
10-day warmback
Depth, ft
3,000
1-day warmback
Injection
Ge
distribution
oth
4,000
er m
al
g
r ad
ie n
5,000
t
10-day injection
0 20 40 60 80 100 120 140
Temperature, degF
Figure 52. Injector warmback response (10-day injection).
1,000
100-day warmback
Completion
2,000
10-day warmback
Depth, ft
3,000
1-day warmback
Injection
Ge
distribution
ot
4,000
he
rm
al
gr
a
die
n
5,000
t
10-day injection
water, they are preferentially cooled and warm back at scale that, even after long injection times, the warmback
a slower rate than the low-permeability zones, although analysis technique can be employed with only a short
there is little contrast after only one day’s shut-in. The warmback period of a day or so (Fig. 55).
effect of the permeability contrast increases as the Typically, long horizontal injection wells have low
shut-in time gets longer. injectivity per foot of rock so these also can be analyzed
However, if injection has taken place for a long dura- using the warmback technique with shut-in times of a
tion (i.e., months to years), the permeability contrasts in day or two.
the formation taking water can be identified only after a One characteristic of all these warmback scenarios
long shut-in time (Fig. 54). is that the nonpermeable rock at the heel of the well
Thus, the optimal time to achieve a clear warmback warms back much more quickly than the reservoir
profile to highlight permeability contrasts in the forma- rock. This means that there can be a large temperature
tion is roughly proportional to the injection time. contrast inside the wellbore at the heel of the well after
If the injection rate is low enough, there is a relation- a short shut-in period. This hot and cold water interface
ship between the injection profile and zone permeability in the wellbore can be used to determine the injection
that can be used to determine the injection profile. Also, profile by tracking it with fast-acquisition DTS measure-
at low injectivity per foot of rock, the warmback response ments (Brown et al., 2004).
to permeability changes occurs at a short-enough time
46
1,000
100-day warmback
Completion
2,000
10-day warmback
Depth, ft
3,000
1-day warmback
Injection
Ge
distribution
ot
4,000
he
rm
al
gr
a
die
n
5,000
t
100-day injection
1,000
100-day warmback
Completion
2,000
10-day warmback
Depth, ft
3,000
1-day warmback
Injection
Ge
4,000
he
rm
al
gra
die
n
5,000
t
3,200
Model injection
response DTS injection
2,400 response 160
Model DTS
Injection rate, 1,600 shut-in shut-in Temperature,
120
bbl/d degF
800 80
Geothermal gradient
Model
0 injection rate 40
X X+400 X+800 X+1,200 X+1,600 X+2,000 X+2,400
Depth, ft
48
Reservoir
Warmback
profile
Inflection points
Temperature
Injection
profile ient
Geothermal grad ½
½
Depth
Figure 57. Hot water velocity measurement.
Openhole completion
Nonreservoir warmup
175–180
170–175
160 165–170
160–165
Temperature, 155–160
degF 140
150–155
o ld
145–150
ys c
125–130
Re s
Depth, ft X+6,000
8:01:34 p.m.
50
176–180
172–175
168–172
165–168
160 161–165
Temperature, 157–161
10:29:09 p.m.
153–157
degF 140 10:05:13 p.m. 149–153
145–149
9:38:31 p.m. 141–145
138–141
134–138
9:11:50 p.m.
130–134
X
126–130
8:45:08 p.m. 122–126
X+1,000
Depth, ft
X+2,000
8:18:27 p.m.
Figure 59. Hot and cold interface velocity measurement with injection resumed.
120
March injection distribution
100 October injection distribution
80
Inflow 60
distribution, %
40
20
0
X X+1,000 X+2,000 X+3,000 X+4,000
Depth, ft
Figure 60. Inflow distribution from hot and cold interface velocity measurement.
Geothermal gradient
Injection + x min
Injection + 2x min
Injection + 2x min
Injection + x min
2D plot
Time
Depth
Figure 61. Hot and cold velocity measurement of three thin permeable zones.
ft X+200
Velocity 0.025 ft/s
X+300
52
Injection of cold water into a reservoir causes the reservoir the producer drops to a value dependent on the zone
rock to cool with time, depending on the temperature of thickness, injection rate and temperature, and distance
the injected water, the rate injected, and the time of injec- between the injector and producer.
tion. The water at the periphery of the flood warms rapidly The theory for the movement of a hot or cold interface
to the geothermal temperature of the formation through in a reservoir due to the injection of water of a different
which it is passing. However, as large volumes of cold water temperature was outlined by Lauwerier (1955) and a
are injected into the reservoir, the rock cools radially from calculation using his equations is shown in Fig. 64. After
the wellbore with time, forming a thermal cold front that two years of injection, the oil/water interface will be at a
follows the waterflood front into the reservoir at approxi- 1,000-ft radius from the injection well while the hot and
mately half the radial distance (Fig. 63). cold interface will be at only a 600-ft radius. This can, of
It follows that after a period of production of injected course, also be modeled using a thermal reservoir simu-
water at geothermal temperature from a nearby injection lator such as the ECLIPSE* Compositional module of the
well, the temperature of the injected water arriving at ECLIPSE industry-reference reservoir simulator.
Cold Flood
Hot Cold front front
Cold Flood
Hot Cold front front
Temperature Reservoir
temperature
180
Reservoir temperature
160
140
Water Waterflood
120 thermal front
Temperature,
degF 100 front
80
60
Injection temperature
40
0 200 400 600 800 1,000 1,200
Radial distance, ft
Figure 64. Thermal front resulting from the radial injection of 60-degF water into a 30-ft 25%-porosity zone for two years at a rate of 10,000 bbl/d.
16,000 180
Gas/liquid
interface
in the annulus ESP
12,000 160
Water flow
0 100
X X+2,000 X+4,000 X+6,000 X+8,000 X+10,000 X+12,000 X+14,000
Measured depth, ft
Cold water
breakthrough
Figure 65. Effect of cold water breakthrough at the bottom producing interval, assuming that all produced fluids are at geothermal temperature.
54
4,000 120
Water flow
0 100
X X+2,000 X+4,000 X+6,000 X+8,000 X+10,000 X+12,000 X+14,000
Measured depth, ft
Cold water
breakthrough
Figure 66. Revised thermal model for the effect of cold water breakthrough at the bottom producing interval.
The Essentials of Fiber-Optic Distributed Temperature Analysis ■ Water Flow Between Wells 55
In heavy oil regions, it is common to reduce the viscos- Monitoring producing and monitor wells in a field
ity of the low-API oil or bitumen by heating it, using under steam injection was the first application of fiber-
steam to make it flow to the surface. In one approach, optic temperature systems in the oil field. This occurred
steam is pumped into injection wells, causing the res- in the San Joaquin Valley heavy oil fields in California
ervoir to become hot between the steam injector and in 1995 and is still being performed in 2016. The high
producer, thereby allowing the lowered-viscosity bitumen temperatures encountered in the initial monitoring
to be pumped from producing wells. A different method efforts resulted in the optical fibers degrading rapidly
employs steam injected directly into the producer well due to hydrogen ingress, which caused poor operating
for a period of time, then the heavy oil and condensed life. However, a significant effort on the development
water are produced from the well in a cycle, which often of optical fiber coatings has resulted in the carbon
requires the use of downhole pumps. This is called the polyimide coating of WellWatcher BriteBlue HT* high-
huff ‘n’ puff process. In a third method, the steam is temperature multimode DTS fiber, which can operate
injected into a separate upper well in a horizontal pair successfully at 570 degF for many years.
of wells (typically 16 ft [5 m] above the producer well).
A steam chamber is created above both wells from which
the lowered viscosity bitumen and condensed water drain Steam monitor wells
down into the producer well by gravity. This last method is
Where steam is injected into one well and oil and bitu-
called steam-assisted gravity drainage (SAGD).
men produced from another, the development of the hot
In all these methods the steam injected ranges from
front between the wells is often observed using monitor
480 to 570 degF, which is outside the range of most
wells situated between the injector and producer so
electrical measurements, and because the produced flow
the reservoir model of the steam front can be refined
is still highly viscous, it is not possible to monitor the
over time. Typically, these monitor wells require only a
flow profile using conventional spinner measurements.
slickline fiber to be run into them on a regular basis and
Measurements in these types of wells are limited to either
a single DTS temperature profile recorded.
thermocouple or fiber-optic temperature measurements.
Previously
steamed zone
458–485
431–458
400 404–431
377–404
350–377
300 323–350
Temperature, 296–323
degF 200 269–296
11/25/2002
242–269
100 3/23/2002 215–242
188–215
X 3/5/2001 161–188
Stea 134–161
X+500 thro m fron 107–134
ugh ts 12/16/2000
the movin 80–107
rese g
X+1,000 r v oi
r 1/29/1999
Depth, ft
X+1,500
1/12/1998
Figure 67. Steamflood tracking using monitor wells.
X+200
Depth, ft
DTS shut-in temperature
X+400
58
X+50
Thermocouple temperature
3 ft from the SAGD wells
Depth, X+100
ft
Steam injector
X+150 Producer
X+200
0 100 200 300 400 500
Temperature, degF
Figure 70. A monitor well showing a high temperature gradient at the producing well.
The Essentials of Fiber-Optic Distributed Temperature Analysis ■ Steamflood and SAGD Well Analysis 59
60
Geothermal gradient
12,000 600
Steam temperature
10,000 500
Calculated
temperature Zone temperature
8,000 400
4,000 200
Cold spots
Flow profile
2,000 100
Geothermal gradient
0 0
X X+500 X+1,000 X+1,500 X+2,000 X+2,500 X+3,000 X+3,500 X+4,000
Depth, ft
Figure 72. SAGD thermal model flow analysis.
The Essentials of Fiber-Optic Distributed Temperature Analysis ■ Steamflood and SAGD Well Analysis 61
Acid reacting with carbonate rock causes an exothermic acid in the rest of the wormhole also reacts with the rock.
reaction, giving off heat. In a process called stimulation, This causes a zone around the wellbore to heat up faster
acid is used to improve near-well permeability in oil and than it would under normal injection conditions. However,
gas wells. During stimulation, acid is bullheaded down the the exothermal heating is a transient event lasting only a
well into the reservoir or placed at selected spots along the short time, although the heated rock remains hot even after
reservoir interval using CT. the exothermal reaction has ceased.
The process of acid dissolution of the carbonate rock Figure 74 shows a well with several reservoir zones
causes “wormholes” (Fig. 73) around the wellbore where (shaded gray) where acid has been bullheaded into the
the reaction and heating take place. This occurs at the tip formation and DTS data acquired during the shut-in warm-
of the wormhole as it expands outward into the formation back period. The “normal warmback” response above the
(McDuff et al., 2010). When injection ceases, the unspent reservoir is of nonpermeable rock that has been cooled
Wormholes
Wormholes
Figure 73. The creation of wormholes around the well by acid reaction.
96.4–98.0
95
94.9–96.4
93.3–94.9
90 91.7–93.3
Temperature, 85
90.2–91.7
degF 88.6–90.2
80 87.1–88.6
85.5–87.1
2:01 p.m. 75 83.9–85.5
82.4–83.9
1:56 p.m. X+600 80.8–82.4
79.2–80.8
1:26 p.m. X+400 77.7–79.2
76.1–77.7
12:55 p.m. 74.5–76.1
Depth, ft
X+200
12:25 p.m.
11:54 a.m. X
Figure 74. Acid heating during warmback.
42.5–43.2
95 41.7–42.5
41.0–41.7
40.2–41.0
Temperature, 39.5–40.2
degF 38.7–39.5
90
38.0–38.7
2:01 p.m. 37.2–38.0
36.5–37.2
35.7–36.5
1:56 p.m.
X+600 35.0–35.7
34.2–35.0
1:26 p.m. 33.5–34.2
32.7–33.5
X+400
12:55 p.m. 32.0–32.7
Depth, ft
12:25 p.m. X+200
X
11:54 a.m.
Figure 75. Acid heating during warmback with the normal nonpermeable warmback response removed.
64
Horizontal section
nt Cg Xg
radie
g
al Nwbk
erm
oth
Ge Cw
e
ur
at
-in er Xw
h ut mp
S k te
ac
r mb Xi
wa X
Injection Ci
temperature x
10.0
Zones in gray correlate
with low log permeability
7.5
Acid response in
low-permeability zones
5.0
VDA acid exothermic warming
in permeable zones only
2.5
Normalization VDA
zone
Temperature, 0
degC
–2.5
–5.0
Preflush
–7.5
Colder than the normalization zone
due to cold water invasion
–10
X X+1,000 X+2,000 X+3,000
Depth, ft
Figure 77. Normalized preflush, VDA acid, and acid responses.
66
Because it is recording temperature along the whole exhibits a large cooling event because the oil flow rate
length of the fiber at the same time, the DTS measure- is low and the Joule-Thomson cooling is large. As the
ment is ideally suited to monitor whether gas lift valves well starts to flow, the cooling event at lower valves is
are working or not by detecting the cooling event that reduced as both the oil flow rate increases and the Joule-
occurs at a valve when gas is passed through it. Often Thomson effect is lessened with increasing annulus
operators run a pressure gauge survey on slickline and pressure (Costello et al., 2012).
plot the pressure gradient inside the tubing to identify Often the bottom valve in a string is an orifice valve,
the operating gas lift valve by the change of gradient. which, as its name implies, has little pressure drop
However, this only identifies the current operating valve across it, so the Joule-Thomson cooling may even be
and does not work if the flowing pressures are not stable. very small. However, the temperature response of gassy
DTS measurements can check that all the valves oil above the valve is different from the response of
operate normally during well startup (they should open nongassy oil, so even where there is no significant Joule-
and pass gas one after the other in sequence) and can Thomson cooling through the valve, it is still possible
identify the operating gas lift valve, even when the flow to identify whether it is flowing gas by comparing the
through the valve is irregular due to a slugging well. temperature response above the valve to that predicted
The cooling event occurring at the gas lift valve is with no gas lift operating.
a result of the pressure drop of gas across the valve, Figure 78 shows a thermal model of gas cooling
causing the gas to cool due to the Joule-Thomson effect. through the top valve in a gas lift string. The cooling is
The magnitude of the cooling is a function of the gas compared with the expected temperature response for
injection rate, pressure drop across the valve, and oil no gas lift cooling, with the well flowing at 1,000 bbl/d
rate up inside the tubing. When the well is started to flow and a 2.0-MMcf/d gas injection rate.
by injecting gas through the valve nearest the surface, it
Depth, ft Cooling
he
rm
valve
gr
a
Reservoir zones
die
n
5,000
t
6,000
Oil rate
7,000
20 40 60 80 100 120 140 160 180
Temperature, degF
Figure 78. Thermal model of gas lift startup with gas flowing through an upper valve.
No gas lift
1,000
Ge
ot
2,000
he
rm
Gas lift through valve
al
g
ra
3,000
die
n
Gas lift valve
t
Depth, ft Gas rate at 4,500 ft
4,000
Cooling at gas
lift valve Reservoir zones
5,000
6,000
Oil rate
7,000
20 40 60 80 100 120 140 160 180
Temperature, degF
Figure 79. Thermal model of gas lift startup with gas flowing through a lower valve.
No gas lift
1,000
Gas lift
Ge
through orifice
ot
2,000
he
rm
al
g
ra
3,000
die
n
Orifice valve
t
6,000
Oil rate
7,000
20 40 60 80 100 120 140 160 180
Temperature, degF
Figure 80. Thermal model of gas lift startup with gas flowing through a lower orifice valve.
68
Geothermal
gradient
200–210
X+2,000
190–200
180–190
170–180
160–170
150–160
Depth, X+4,000 141–150
131–141
ft 121–131
111–121
Gas lift valves 101–111
91–101
X+6,000 81–91
71–81
1:17 p.m. 61–71
12:51 p.m.
12:25 a.m.
100
11:59 a.m.
150
Temperature,
degF 200 11:33 a.m.
Figure 81. Unstable flow causing irregular cooling at a gas lift valve.
Leaks can occur at several places in a well (e.g., in tubing Example: Leaks in a casing identified
or casing, past packers, or even outside the well between
different permeable zones where the cement in the
by injection
outside casing is insufficient to provide a pressure seal). Figure 82 shows leaks through a casing identified by
Although there are several types of casing measure- injecting water down the inside of the casing. There
ment tools, including calipers and even spinner logs, are clearly two leaks occurring at X+5,900 ft and
most of these are limited to measuring the inside the X+7,200 ft with no flow down the casing below the
tubing string. The only measurements that can pene- lower leak.
trate multiple strings and respond to flow outside the Although the upper leak must be through the casing,
casing are acoustic and temperature measurements. there are two options for the lower leak: either there is
The best way to identify a leak with temperature also a hole in the casing at this point or the flow leaking
measurements is to cause the maximum amount of through the upper hole is flowing down the outside of the
flow possible through the leak by injecting into the casing and bleeding off into the formation at this point.
tubing or annulus or bleeding off the annulus pressure. It is not possible with temperature alone to identify
Temperature response to this transient event is recorded which is the correct option.
(Julian et al., 2007).
X+2,000
Geothermal gradient
X+4,000 Completion
End of injection
Depth,
ft X+6,000
Leakoff points
X+8,000
X+10,000
X
Well flowing
X+1,000
Well startup
X+2,000
Valve operating
Flow through holes
X+3,000
Depth,
Completion
Gas lift valve Gas lift
ft through holes
X+4,000
X+5,000
X+6,000
X+7,000
40 60 80 100 120 140 160
Temperature, degF
Figure 83. Tubing leaks identified during gas lift injection with gas lift valve.
X+500
X+1,000
X+1,500
Completion
X+3,000
X+3,500
80 90 100 110 120
Temperature, degF
Figure 84. Flow behind casing observed during shut-in.
72
DTS measurements in oil wells can also be used for a fractured using proppant, which is cold when pumped
variety of other applications. Three are mentioned here. from the surface, the fractured interval is usually easy to
identify. If the fracture is not propped open and has been
created only for a short time (e.g., a “data frac” during
Vacuum-insulated tubing monitoring a hydraulic fracture job), then it may be more difficult
Vacuum-insulated tubing (VIT) is sometimes required to to see the fracture because the fracture opening closes
prevent the region surrounding the well near the surface when injection stops.
from getting too hot during production, which may cause
the collapse of casings with no means of annulus pres-
sure relief. DTS monitoring has been used to monitor the Cement-top detection
effectiveness of VIT and check to make sure the vacuum When a casing has been cemented, the cement sets up
in the tubing remains. with an exothermal chemical reaction that causes it
to heat up for several hours after the cement has been
pumped. The increase in temperature above the geo-
Fracture detection thermal gradient is an indication of where the cement
There are many papers on the use of temperature logs is behind the casing. This effect may also vary with hole
for detecting the height of a fracture in a well. Ideally, size: More cement produces a larger exothermal heating.
the fracture is in line with the wellbore, but it may not Wireline temperature logs are usually run 12 hours after
be, either due to the well trajectory or rock stresses. In a cement job to identify the cement top but, because the
these latter cases the fracture response is more difficult exothermal heating is a function of several parameters,
to interpret. In horizontal wells, the fracture most likely this may not be the optimal time. However, DTS mea-
intersects the wellbore at one point only. surements recorded continuously for 24 hours after the
The method of identifying fractures is to shut in the cement job has occurred are sufficient to identify all the
well after the hydraulic fracture job has been completed intervals in the well where the cement-setup exothermal
and the fractured interval will warmback toward the reaction has occurred.
geothermal gradient slower than the response of a nor-
mally invaded interval. If the well has been hydraulically
Al-Asimi, M., Butler, G., Brown, G., Hartog, A., Clancy, changing injection support around the well,” paper
T., Cosad, C., Fitzgerald, J., Navarros, J., Gabb, A., SPE 135130 presented at the SPE Annual Technical
Ingham, J., Kimminau, S., Smith, J., and Stephenson, K.: Conference and Exhibition, Florence Italy (September
“Advances in well and reservoir surveillance,” Oilfield 19–22, 2010).
Review (Winter 2002), 14–35.
Chertenkov, M., Deliya, S., Semkin, D., Brown, G.A.,
Brown, G.A.: “Monitoring multilayered reservoir Bayanova, A., Kanevsky, E., Nukhaev, M., Shapovalov, A.,
pressures and gas/oil ratio changes over time using and Pormeyster, Y.: “Gas breakthrough detection and
permanently installed distributed temperature production monitoring from ICD screen completion on
measurements,” paper SPE 101886 presented at the Lukoil’s Korchagina field using permanently installed
SPE Annual Technical Conference and Exhibition, San distributed temperature measurements,” paper
Antonio, Texas, USA (September 24–27, 2006). SPE 159581 presented at the SPE Annual Technical
Conference and Exhibition, San Antonio, Texas, USA
Brown, G.A., Carvalho, V., Wray, A., Smith, D., Toombs, M.,
(October 8–10, 2012).
and Pennel, S.: “Monitoring alternating CO2 and water
injection and its effect on production in a carbonate Costello, C., Sordyl, P., Hughes, C., Figueroa, M., Balster,
reservoir using permanent fiber-optic distributed E., and Brown, G.A.: “Permanent distributed sensing
temperature systems,” paper SPE 90248 presented at (DTS) technology applied in mature fields: a Forties
the SPE Annual Technical Conference and Exhibition, field case study,” paper SPE 150197 presented at the
Houston, Texas, USA (September 26–29, 2004). SPE Intelligent Energy International, Utrecht, The
Netherlands (March 27–29, 2012).
Brown, G.A., Field, D., Davies, J., Collins, P., and
Garayeva, N.: “Production monitoring through open hole Curtis, M.R., and Witterholt, E.J.: “Use of the tempera-
gravel-pack completions using permanently installed ture log for determining flow rates in wells,” paper SPE
fiber-optic distributed temperature systems in the 4737 presented at the SPE Annual Meeting , Las Vegas,
BP-operated Azeri field in Azerbijan,” paper SPE 95419 Nevada, USA (September 30–October 3, 1973).
presented at the SPE Annual Technical Conference and
Doll, H.G., and Perebinossoff, A.A.: “Temperature mea-
Exhibition, Dallas, Texas, USA (October 9–12, 2005).
surements in oil wells,” presented at the 1936 Meeting
Brown, G.A., Kennedy, B., and Meling, T.: “Using fiber- of the Institute of Petroleum Technology.
optic distributed temperature measurements to provide
Fahim, M., Keshka, A., Al Marzooqi, A., Alvi, A., Brown,
real- time reservoir surveillance data on Wytch Farm
G.A., and Al Neyaei, F.: “Distributed temperature sensing
field horizontal extended-reach wells,” paper SPE 62952
(DT) enables injectivity visualization to enhance
presented at the SPE Annual Technical Conference and
stimulation efficiency,” paper SPE 141238 presented at
Exhibition, Dallas, Texas, USA (October 1–4, 2000).
the SPE Middle East Oil and Gas Conference, Manama,
Brown, G.A., Storer, D., McAllister, M., al-Asimi, A., and Bahrain (March 6–9, 2011).
Raghavan, K.: “Monitoring horizontal producers and
Fryer, V., Shuxing, D., Otsubo, Y., Brown, G., and
injectors during cleanup and production using fiber-
Guilfoyle, P.: “Monitoring of real-time temperature
optic-distributed temperature measurements,” paper
profiles across multizone reservoirs during production
SPE 84379 presented at the SPE Annual Technical
and shut in periods using permanent fiber-optic
Conference and Exhibition, Denver, Colorado, USA
distributed temperature systems,” paper SPE 92962
(October 5–8, 2003).
presented at the SPE Asia Pacific Oil & Gas Conference
Brown, G.A., and Tiwari, P.: “Using DTS flow and Exhibition, Jakarta, Indonesia (April 5–7, 2005).
measurements below electrical submersible pumps doi:10.2118/92962-MS
to optimize production from depleted reservoirs by
76
350
t t
0f 0f
F⁄10 F⁄10
g g ft
de de
00
1.6 1.4 F⁄1
g
de ft
1.2
F⁄100
250
deg
1.0 ft
g F⁄100
Temperature, de
0.8 ft
degF 100
6 d egF⁄
0.
150
50
0 5,000 10,000 15,000 20,000
Depth, ft
Figure A-1. Geothermal gradient, degF/100 ft.
220
0 m
10
g C⁄
de
3.0 00
m
170
g C ⁄1 m
de 100
2 .5 e g C ⁄
d 0m
2.2 C ⁄ 10
de g
Temperature, 120 2 .0
00 m
degC
e g C ⁄1
d
1.5
00 m
0 d e gC ⁄ 1
1.
70
20
0 1,000 2,000 3,000 4,000 5,000 6,000
Depth, m
Figure A-2. Geothermal gradient, degC/100 m.
78