The Essentials of Fiber-Optic

Distributed Temperature Analysis

The Essentials of
Fiber-Optic Distributed
Temperature Analysis
 Schlumberger 20052009
Schlumberger
225 Schlumberger Drive
Sugar Land, Texas 77478
Produced by Schlumberger Educational Services.
All rights reserved. No part of this book may be reproduced,
stored in a retrieval system, or transcribed in any form or by any
means, electronic or mechanical, including photocopying and
recording, without prior written permission of the publisher.
09-CO-0139

An asterisk (*) is used throughout this document to denote a mark of Schlumberger.
 Contents

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Fiber-Optic Distributed Temperature Measurement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Single-ended versus differential-light-loss compensated (double-ended)
temperature measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
Thermal resolution and statistics reduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
Flow resolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Example: Flow resolution between layers 25 m apart . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Oilwell Thermal Basics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
The geothermal gradient . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
The Joule Thomson effect . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Flow up the well . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Single-Phase Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
Multizone Reservoirs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Example: Quicklook interpretation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
Crossflow between zones . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
Multiphase Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
THERMA* Thermal Modeling Software . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
Thermal workflow analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
Data input . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
Temperature calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
DTS data analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
Flow calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
Example: Flow from a multilayered reservoir . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
Joule Thomson Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
Flow in horizontal wells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
Flow within the reservoir . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
Example: Horizontal-well Joule Thomson flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
Water Injection Wells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
Hot-water-injection technique . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
Hot-slug velocity measurement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
Example: Hot-slug velocity in a horizontal injector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
Flow Between Wells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
Example: Cold-water breakthrough . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49

The Essentials of Fiber-Optic Distributed Temperature Analysis Contents iii

 Low-Rate Gas Wells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
Kuntz and Tixiers subtangent method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
Example: Low-rate multizone gas well . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
High-Rate Gas Wells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
Water Detection Using Temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
Low Joule Thomson effects (oil wells) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
High Joule Thomson effects (gas wells) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

iv
 Introduction

Temperature logs have been used to monitor producing wells since the early 1930s,1 and a con-
siderable number of papers have been dedicated to temperature measurements and their
analysis over the years. The data generated have been used to calculate the flow contribution in
oil and gas wells and to evaluate water injection profiles and the effectiveness of fracture jobs,
as well as to identify cement tops, crossflow between zones, flow outside the casing, and other
flow- and wellbore-related events.
Normally analysis of the temperature log recorded with the production logging tool is consid-
ered secondary to analysis of the spinner log, which measures flow velocity directly; temperature
is conventionally used only as an indicator of gas in flow. The main disadvantage of relying on a
temperature log for well productivity analysis is that only one thermal profile is obtained when
the production logging tool is run.
However, the Sensa* fiber-optic distributed temperature sensor (DTS) system can generate
many temperature logs (as frequently as every 7 s) over the life of the well because it is prein-
stalled with the completion. In addition, the system identifies the source of a change in well
performance when it occurs, rather than later during a production logging run, and unlike pro-
duction logs, DTS measurements do not require interventions that can change the wells status.
In recent years the cost of production logs has increased considerably because many wells are
drilled horizontally through the reservoir and the logging tools must be conveyed on coiled
tubing or by well tractors. In certain cases, such as some subsea wells, production logging tool
conveyance is not possible. Consequently, alternative technologies become viable if they can be
installed in wells at a reasonable cost and their data can be interpreted to give flow and other
production information. The Sensa DTS system is just such a technology.
This manual outlines the current technology and best practices of distributed time- and
depth-based temperature analysis for use with the Sensa DTS system.

George Brown
Manager, Interpretation Development

The Essentials of Fiber-Optic Distributed Temperature Analysis Introduction 1

 Fiber-Optic Distributed
Temperature Measurement

Fiber-optic distributed temperature measurement uses an industrial laser to launch 10-ns bursts
of light down the 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 to measure
the temperature along the fiber. Because the speed of light is constant, a spectrum of the
backscattered light can be generated for each meter of the fiber by using time sampling, and that
allows generation of a continuous log of spectra along the fiber (Fig. 1).
A characteristic of each spectrum of backscattered light is that the ratio of the Stokes Raman
bands to the anti-Stokes Raman bands is directly proportional to the temperature of the length
of fiber from which it is generated. Consequently, a log of temperature can be calculated every
meter along the whole length of the fiber by using only the laser source, analyzer, and a refer-
ence temperature in the surface system. There is no need for calibration points along the fiber
or for calibrating the fiber before installation.

10-ns pulses of laser light Fiber

Laser

Incident
Brillouin Raleigh
Analyzer
light

Stokes
Anti-Stokes Raman
Backscattered Raman
light band
band

Spectra of back- Wavelength

scattered light

Figure 1. Distributed temperature measurement.

The Essentials of Fiber-Optic Distributed Temperature Analysis Fiber-Optic Distributed Temperature Measurement 3
 Temperature from the non-temperature-sensitive Stokes (NTS) and temperature-sensitive
anti-Stokes (TTS) Raman bands is calculated using the following algorithm:

1
=
1

(( x
) (
ln TTS / NTS / TTS / NTS )c )
,
Temp Tref Sens
where
Temp = calculated temperature (K)
Tref = reference-coil temperature (K)
Sens = temperature sensitivity
(TTS/NTS)x = anti-Stokes/Stokes ratio at the point of interest along the fiber
(TTS/NTS)c = anti-Stokes/Stokes ratio in the reference coil.
Spectrum acquisition times can be specified from as little as 7 s to hours, and the length of the
acquisition time defines the accuracy and resolution of the measured temperature log. Typically
a resolution of 0.1C is required for reservoir surveillance.
Figure 2 shows NTS and TTS data acquired from both ends of a 9,000-m fiber installed in a
double-ended configuration in a well with a depth of 4,500 m. The fiber turnaround sub is at the
bottom of the well. The counts decrease exponentially with fiber length and there is a noticeable
effect of light loss at the connections at the wellhead.
Usually a measurement interval of 1.0165 m is used, which allows the measurement of up to
12 km of fiber. Longer fibers, up to 30 km, can be used, but they require longer measurement
intervals of up to 10 m.

30,000

25,000
TTS 1 Anti-Stokes
20,000

TTS 2 Anti-Stokes
Counts 15,000

10,000
NTS 1 Stokes NTS 2 Stokes
5,000

0
0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 9,000
Fiber length (m)

Figure 2. NTS and TTS data along a 9,000-m fiber.

4
 Single-ended versus differential-light-loss-compensated
(double-ended) temperature measurements
A differential-light-loss-compensated temperature measurement is made by sending pulses of
light down the fiber from one end and then from the other. NTS and TTS traces are obtained for
both sets of measurements. Sensa DTS software uses both sets of measurements to produce a
temperature trace that is automatically corrected for the differential light losses that can occur.
In a single-ended temperature measurement, the light loss is corrected by using a differential
light correction (DLC) factor that assumes the loss is constant along the fiber. Variations in the
manufacturing, purity of the fiber, and bends in the control line are factors that can invalidate
this assumption.
There are four primary advantages of differential-light-loss-compensated (double-ended)
measurements over single-ended measurements.
The absolute accuracy of a double-ended measurement can be specified, whereas it cannot be
for a single-ended measurement because of uncertainties in the light-loss correction.
Nonuniform light loss can result in apparent thermal anomalies that could be misinterpreted
from single-ended measurements to be flow or other well events.
Single-ended temperature measurements can change over time as the loss characteristic of the
fiber changes (e.g., new connections, reconnected fibers, different DTS, fiber deterioration).
With a double-ended installation, the fiber can easily be pumped out and replaced if necessary.
Given the same acquisition time, a double-ended measurement has poorer thermal resolution
because measurement is required from both ends of the fiber and computation time is increased.
The same thermal resolution can easily be achieved by increasing acquisition time.
In wells where thermal changes are small and it is important to track them over time (e.g.,
producing oil wells), the double-ended technique is essential. However, in wells where the ther-
mal changes are large (e.g., steamflood wells), the single-ended installation is appropriate and
more cost-effective.

Fiber-Optic Distributed Temperature Measurement 5

 Figures 3a and 3b compare a single-ended measurement (installed in a J configuration with
fiber running partially back up the well beyond the turnaround sub) and a double-ended mea-
surement in a 4,500-m oil well. The single-ended measurement DLC value has been adjusted so
that the temperature from 3,000 m down to the turnaround sub is the same as the temperature
from the turnaround sub back to 3,000 m and both measurements are symmetrical around the
turnaround sub. The absolute difference between the single- and double-ended measurements,
the loss-induced shift at 3,100 m, and the slope change below 3,100 m were caused by inadequate
differential-light-loss correction on the single-ended measurement.

80

75

Temperature
(C) 70

65

Double-ended measurement
60 Single-ended measurement
0 1,000 2,000 3,000 4,000
Depth (m)

80
Loss-induced slope change
Loss-induced shift Absolute
79
error

Temperature 78
(C)

77

Double-ended measurement
76 Single-ended measurement
2,000 3,000 4,000
Depth (m)

Figures 3a and 3b. Comparison of single-ended and double-ended measurements.

6
 Thermal resolution and statistics reduction
The DTS temperature measurement is subject to statistical variation, so the longer the acquisi-
tion time, the better the thermal resolution of the measurement. Figure 4 shows that the
root-mean-square (RMS) thermal resolution is also dependent on the length of the fiber and
whether the installation is single-ended or double-ended (i.e., looped back so the laser can pulse
light alternately from either end).

1,000
4-km DE
8-km DE
10-km DE
100 12-km DE
4-km SE
8-km SE
Acquisition 10-km SE
time (min) 10 4-km DE
8-km DE
10-km DE
12-km DE
1
4-km SE
8-km SE
10-km SE
0.1
0.1 1 10
RMS resolution (C)

Figure 4. Acquisition time versus RMS temperature resolution. (DE = double-ended, SE = single-ended)

The data shown in Figure 4 are for a Sensa DTS 800 system with a 1-m measurement length;
they show that long acquisition times will be needed in deep wells where high thermal resolution
on the order of 0.1C is required. In these cases, it is possible to use shorter acquisition times if
several adjacent depth levels, or time steps at one depth level, are combined with appropriate
statistical variation-reducing algorithms. But shorter acquisition times can result in reduced ver-
tical or temporal resolution. Raw Sensa DTS data can appear to be noisy because the
acquisition time is too short to achieve the required thermal resolution.
However, there are often good reasons for using shorter acquisition times. Deep producing
wells are one example. Acquiring a temperature trace every hour or two throughout the day
rather than a higher-quality curve once a day allows identifying trends that occur during the day.
A noise-reduction algorithm can be applied to the data to reduce either depth or temporal reso-
lution, but achieving the required thermal resolution for flow analysis or for comparison of
temperature trends takes longer time periods. Acquiring data too often (e.g., every quarter hour
in a producing well) increases the likelihood of observing short-term thermal events, but results
in massive datasets.
A second example is tracking a hot slug in an injection well. In this situation it is important
to acquire data rapidly (every minute or at even shorter intervals); thermal resolution is less
important.

Fiber-Optic Distributed Temperature Measurement 7

 Raw DTS data should be acquired at a frequency higher than needed to provide the required
thermal resolution (typically 0.1C in producing wells), and statistics-reduction algorithms
should be applied to achieve the required resolution. Following these recommendations improves
the ability to observe short-term trends. Figures 5 and 6 show the effect of applying a statistics-
reduction filter to data acquired over a 6,000-ft reservoir interval to improve thermal resolution.
In this example a filter was applied in depth and time to enhance the thermal resolution.

150

Temperature
140
(F)
130 02/05 19:52
02/05 00:43
X0,000 01/05 00:46
X2,000 30/04 00:49

X4,000 29/04 00:53

Depth
28/04 00:56
(ft)
Figure 5. Raw data over the reservoir interval.

150

Temperature
(F) 140

02/05 19:52
130 02/05 00:43
01/05 00:46
X0,000 30/04 00:49
X2,000 29/04 00:53

Depth X4,000 28/04 00:56

(ft)

Figure 6. Filtered data over the reservoir interval.

8
 Flow resolution
The accurate resolution of flow contribution from temperature logs in vertical and deviated wells
depends primarily on the resolution of the temperature measurement, the geothermal gradient,
and the separation between the two flowing intervals. In horizontal wells the subsequently dis-
cussed Joule Thomson effect is the significant factor.
Assuming that raw DTS data can be filtered to achieve a thermal resolution of 0.05C (Fig. 7a),
Fig. 7b can be used to estimate the flow resolution as a function of the geothermal gradient and
vertical separation of the zones.

Example: Flow resolution between layers 25 m apart

Figure 7b shows that if the vertical zone separation is 25 m and the geothermal gradient is
2C/100 m, it is possible to resolve flow contribution to an accuracy within 10% of the total
flow rate.

(a)

0.05C
Geothermal gradient

Upper reservoir

Vertical
Lower reservoir separation

(b)
100
0.05C (geothermal = 1C/100 m)
0.05C (geothermal = 2C/100 m)
0.05C (geothermal = 3C/100 m)

Flow-resolution %
(lower flow/total flow) 10

1
1 10 100 1,000
Vertical-zone separation (m)

Figures 7a and 7b. Flow resolution as a function of geothermal gradient and vertical zone separation.

Fiber-Optic Distributed Temperature Measurement 9

 Oilwell Thermal Basics

The key to understanding temperature logs in oil and gas wells is knowing how the fluids flow-
ing in the well gain or lose heat because of the external effect of the geothermal gradient and
the internal Joule Thomson fluid effect. The combination of these effects creates a characteristic
time-dependent thermal profile that can be recorded using the Sensa DTS system and analyzed
to determine the flow in the wellbore.

The geothermal gradient

Because of the tremendous difference in temperature between the core of the Earth and
space beyond our atmosphere, the Earth is constantly losing heat by conduction through its crust
(Fig. 8). The Earths crust consists of layers of rock with different geological, petrophysical, and
thermal properties; therefore the geothermal gradient is not necessarily a straight line.
Reservoir fluids lose heat to the cooler surrounding environment as they are produced up the
wellbore. Quantifying this heat loss requires an accurate measurement of the geothermal gradient.

Very cold
Layers of rock with differing
thermal properties
Space
Geothermal
gradient
Surface

Earth

Thermal
gradient

Very hot Temperature Temperature

Figure 8. The geothermal gradient.

Injection or circulation of fluids in the well changes the temperature of the near-wellbore
region, and time must be allowed for the transient effects to dissipate before the real geother-
mal gradient can be identified. Thermal equilibrium often takes days, or even weeks, so
knowledge of the near-time well history is important when defining the geothermal gradient.
Typical geothermal gradients vary from 0.6 to 1.6F/100 ft [1.0 to 3.0C/100 m], with a typical
average value of 1.0F/100 ft. (Refer to Charts 1 and 2 on pages 63 and 64.)

The Essentials of Fiber-Optic Distributed Temperature Analysis Oilwell Thermal Basics 11

 The Joule Thomson effect
The Joule Thomson effect is a change in the temperature of a fluid during expansion in a steady-
flow process involving no heat transfer or work at constant enthalpy.2 It occurs in throttling-type
processes, such as adiabatic flow through a porous plug or an expansion valve, and as fluids flow
into and up the wellbore, where a drop in pressure occurs.
Joule Thomson heating of injected water is also referred to as friction heating because it
results from the viscous dissipation of pressure in flow through porous media.
The temperature change per unit pressure change at constant enthalpy is defined by the Joule
Thomson coefficient, JT :

T T 1
JT = = t ,
p
H c

where
T = temperature
p = pressure
H = enthalpy
t = coefficient of thermal expansion
= fluid density
c = specific heat capacity.
This effect means that as liquid or gas flows in the reservoir toward the well, depending on its
Joule Thomson coefficient, it heats or cools because of the pressure drop at the wellbore. A liquid
or gas also heats or cools as it flows up the wellbore because of the friction pressure drop there.
In general a pressure drop causes slight heating of flowing oil and water but a large Joule
Thomson cooling for flowing gas.
Figure 9 shows the thermal response of the Joule Thomson effect on gas and oil flowing radi-
ally from a reservoir into a wellbore. The largest cooling (or heating) takes place immediately
around the wellbore, which is where the largest pressure drop occurs.

Pressure and
temperature

Inflow temperature (oil), a small increase

Reservoir temperature

Inflow temperature (gas), a large decrease

Reservoir pressure

Bottomhole
flowing pressure

Radial distance from the wellbore

Figure 9. Joule Thomson effect in the reservoir.

12
 In a horizontal pipe or wellbore, once the system has achieved thermal equilibrium, the pres-
sure drop along the well causes a linear increase (or decrease) in temperature along the well
because of the steady flow of fluid and Joule Thomson effect (Fig. 10).

Temperature

Single-point inflow

Figure 10. Joule Thomson effect along a horizontal well.

Flow up the well

When oil flows up the well, its pressure decreases, and gas comes out of solution once the pres-
sure has dropped below the bubblepoint (Fig. 11). The flowing volume of oil produced up the well
is reduced by the oil formation volume factor, Bo.

Gas
Oil As oil and water flow
Water to the surface, the oil
volume shrinks, and
gas comes out of solution
below the bubblepoint
pressure.

Bubblepoint
This change in volume
pressure
and fluid properties
must be accounted for
in thermal modeling.

Oil
Water

Figure 11. Upward flow of oil and gas.

Oilwell Thermal Basics 13

 The oil formation volume factor is defined as
Volume of oil (including dissolved
gas at downhole conditions)
Bo = .
Volume of oil at 60F and 14.7 psi
Bo has a typical value between 1.2 and 1.4 depending on the gas/oil ratio (GOR).
The amount of gas produced at the surface depends on the gas formation volume factor (Bg),
which is defined as
Volume of gas at downhole
temperature and pressure
Bg = .
Volume of gas at 60F and 14.7 psi
The change in the volumes of fluid and gas as they travel up the well causes the fluid and gas
velocities to change. Thus as oil flows up a well at a constant rate, its properties and velocity are
changing with depth as gas comes out of solution and the oil shrinks. Heat carried up the well by
the fluid is lost to the surrounding formation by conduction through the completion hardware
and cement. The recorded temperature curve is a result of all of these effects, which are also
time dependent. This is a complex process, but the algorithms defining it are well understood
and can be derived from conventional equations of heat transfer.
To solve the equations, a NODAL* analysis solution is required to calculate the pressure drop
up the well tubing on the basis of the fluid properties under flowing conditions and then to cal-
culate the transient-temperature distribution depending on how long the well has been flowing.

14
 Single-Phase Flow

Ramey 3 proposed an approximate solution to the heat-transmission problem caused by unsteady

radial conduction to the surrounding Earth when hot or cold fluid is injected into a steady-state
wellbore. The solution allows the calculation of steady-state temperatures by depth and time.
This solution was adapted for use in producing wells by Schlumberger over 30 years ago,4
although the solution was only for single-phase flow and did not account for changes in fluid ther-
mal properties when fluids or gas flow up the well (Fig. 12).

0
Temperature
profile
1,000

T (z, t)
2,000 Tgz

Geothermal
Depth (ft) gradient
3,000

z
4,000
Reservoir

5,000
Tge

50 100 150
Temperature (F)

Figure 12. Temperature response for upward single-phase flow.

The Essentials of Fiber-Optic Distributed Temperature Analysis Single-Phase Flow 15

 According to Curtis and Witterholt,4 the fluid temperature at any point T(z,t) at a height z
above the producing reservoir (at geothermal temperature Tge) at any time t after the com-
mencement of production is given by the following equations:

rce
()
f t = ln
2 t 1 /2
0.290 ,
( )

[ ( )]
A = 1 .66 Q f Cf f t ,

( ) ( )
T z ,t = Tge gG z + gG A + gG A e z /A,

where
rce = casing radius (ft)
= formation thermal diffusivity (ft2/D)
t = time (D)
A = relaxation distance (ft)
Q = flow rate (B/D)
f = fluid density (g/cm3)
Cf = fluid-specific heat (Btu/lbm in F)
gG = geothermal gradient (F/100 ft).
The equations show that for a given flow rate the observed temperature profile forms an
asymptote to a line parallel to the geothermal gradient that increases exponentially with time.
In temperature log analysis, it is important to know the flowing history of the well and to
understand that any changes in flow rate result in a time-dependent change in the temperature
curve. The well warms up rapidly in the first few days of production and changes more slowly
thereafter. Figure 13 shows the effect of changing flow rate on the steady-state producing tem-
perature log.

0
500 B/D
250 B/D 1,000 B/D
1,000
Geothermal
gradient
2,000

Depth (ft)
3,000

4,000

Reservoir
5,000

50 100 150
Temperature (F)

Figure 13. The effect of changing flow rate on the temperature log.

16
 Figures 14a and 14b show the effect of time on the temperature log as the well heats up and
achieves thermal equilibrium at a constant flow rate of 1,000 BOPD.

0
0.5 D 50 D
5D 500 D
1,000

2,000
Geothermal
Depth (ft) gradient
3,000

4,000

Reservoir
5,000

50 100 150
Temperature (F)

150

2,500 ft

Surface
Temperature (F)
100

50
0 5 10 15 20 25 30
Time (D)

Figures 14a and 14b. The effect of time on the temperature log, given a constant flow rate.

Single-Phase Flow 17
 Figure 15 shows the temperature profiles calculated using the Curtis-Witterholt equations
for oil flowing at 500 B/D from reservoirs at 4,000- and 5,000-ft depths for the first three days
of production.
If temperature data are recorded continuously during the warmup period of a well producing
at a constant flow rate, the thermal parameters of the equations can be accurately derived and
the temperature data used to calculate flow.

Flow zones

150

100 Temperature
(F)

50
29/03 15:12
29/03 05:32
28/03 16:15 4,000
28/03 02:59
27/03 13:42 2,000 Depth
27/03 00:26 (ft)
0

Geothermal Gradient Definition

Lower reservoir depth (ft) 5,000

Depth 1 (TVD ft) 0

Temperature 1 (F) 50

Depth 2 (TVD ft) 5,000

Temperature 2 (F) 150

Geothermal gradient (F/ft) 0.02

Upper flow (B/D) 500

Lower flow (B/D) 500

Thermal diffusivity (ft2/D) 0.98

Casing diameter (in.) 7

Fluid density (gm/cc) 0.8

Fluid specific heat (Btu/lbm F) 1

Figure 15. Calculated temperature profiles for the first three days of production.

18
 Multizone Reservoirs

Where flow from two or more zones is commingled in a single wellbore, flow from the upper zone
enters the well at a temperature lower than that of the deeper zone. The addition of this colder
fluid to the flowing stream decreases the streams temperature at the mixing point, an anomaly
that can be used to clearly identify the point of fluid entry.
The thermal response of the anomaly is a function of the combined flow rate above the upper
zone and flow rate from the lower zone below the anomaly (Fig. 16). Thus, given the geothermal
gradient and the measured temperature profile, the proportional contribution from two or more
commingled zones can be calculated.

Flow rate (B/D)

0 1,000 2,000
0

Total flow
1,000

Upper
flow
2,000
Lower flow Tt Tg Lower flow
=
Depth (ft) Total flow Tl Tg
3,000
Reservoir
Tt Tl
4,000
Reservoir Tg

5,000

50 100 150
Temperature (F)

Figure 16. Thermal response of multizone flow.

The temperature decrease at the confluence of the two fluid streams compared with the differ-
ence between the lower zones flowing temperature and the geothermal temperature at the mixing
point reflects the percentage of flow between the lower reservoir and total fluid production.
This can be expressed by a simple, quicklook equation:
lower flow rate mixed temperature geothermal temperature
= .
total flow rate lower flow temperature geothermal temperature

The Essentials of Fiber-Optic Distributed Temperature Analysis Multizone Reservoirs 19

 Example: Quicklook interpretation
Figure 17 shows the thermal response of flow from three zones at 4,000, 4,500, and 5,000 ft. The
response was calculated using a thermal model, with the flow rates shown in the figure.
The red temperature curve shows 2,000 BOPD production below the 4,000-ft inflow zone and
3,000 BOPD above it.
Temperature of flow immediately below the 4,000-ft zone, Tl = 146F
Temperature of the mixed flow immediately above the 4,000-ft zone, Tt = 141F
Geothermal temperature at the 4,000-ft zone, Tg = 130F
As noted in Fig. 16, lower flow = total flow (Tt Tg)/(Tl Tg).
Lower flow = 3,000 (141 130)/(146 130)
Lower flow = 3,000 0.69
= 2,062 BOPD
Accuracy is 62 BOPD, or 3.0%.
This quicklook approach assumes no Joule Thomson warming or cooling in the reservoir.

Flow rate (B/D)

0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000
3,500

Flow distribution Reservoir

4,000

Reservoir Geothermal
temperature
Depth (ft) 4,500

Reservoir

5,000

Calculated
temperatures
5,500
80 90 100 110 120 130 140 150 160
Temperature (F)

Flow Rates per Zone

Depth (ft) Red Flow Rate Blue Flow Rate Orange Flow Rate
(BOPD) (BOPD) (BOPD)
4,000 1,000 1,000 500
4,500 1,000 500 500
5,000 1,000 1,500 2,000

Figure 17. Thermal model of multizone flow.

20
 Crossflow between zones
Where reservoir intervals have different reservoir pressureseither naturally or from the effects
of productionand are connected through the wellbore, they start flowing from one zone to the
other when the well is shut in at the surface.
This crossflow can also occur between zones supposedly sealed behind cemented casing where
the cement job is less than perfect and causes increased corrosion of the casing and unwanted
depletion of intervals thought sealed.
Whether crossflow results from different reservoir pressures or ineffective sealing, the tem-
perature log responds to the flowing fluid and can be used to identify the flow direction and
magnitude as shown in Fig. 18.

Flow rate (B/D)

0 1,000 2,000
0

1,000

2,000

Depth (ft)
3,000
Reservoir Upward crossflow

4,000
Flow Reservoir
distribution
5,000

50 100 150
Temperature (F)

Figure 18. Upward crossflow between zones.

It makes no difference whether the flow, or the installed fiber, is inside or outside the casing
because the thermal effect generated is conducted through the casing wall. Identification of flow
behind casing is a classic temperature log application.
The spinner log is often difficult to analyze or assessed incorrectly in wells with slotted liner
and external casing packers, or in the case of sand screen completions with gravel packs, because
the flowing cross-sectional area is required to convert the spinner velocity to flow rate. This vari-
able is unknown and can vary with depth. However, the temperature log, which responds to the
mass flow rate, indicates the flow rate whether flow is inside the casing, outside the casing, or both.

Multizone Reservoirs 21
 Multiphase Flow

Where there is multiphase flow, the friction pressure losses caused by flow up the tubing result
in changes in the thermal properties of the liquids and gas. These changes must be taken into
account when calculating the thermal profile.
In Fig. 19, 1,000 B/D of 30 API, 100 GOR oil; 1,000 B/D water; and 1.0 MMcf/D gas are flowed
from a reservoir at 5,000-ft depth and are shown together with the geothermal gradient. It has
been assumed that steady-state flowing conditions prevail, that is, the well has been flowing at
this rate for more than one year, and the permeability is high enough that there are no reservoir
Joule Thomson effects.
The difference in the shape of the thermal profiles results entirely from the differences
between fluid and gas thermal properties, whereby water loses less heat to the surrounding envi-
ronment than an equivalent flow of oil.
As we approach the reservoir, the difference between the oil and water temperature profiles
decreases. At the reservoir level the Joule Thomson thermal effect from the near-wellbore region
becomes significant while the thermal effects of upward flow become negligible unless the reser-
voir intervals have a large vertical separation.

0
1,000 BWPD
1 MMcf/D
1,000
1,000 BOPD

2,000
Geothermal
gradient
Depth (ft)
3,000

4,000
Reservoir

5,000

50 100 150
Temperature (F)

Figure 19. Modeled thermal profiles of oil, water, and gas.

The Essentials of Fiber-Optic Distributed Temperature Analysis Multiphase Flow 23

 THERMA Thermal
Modeling Software

Thermal workflow analysis

To convert measured temperature data to a measurement of flow from different reservoir zones,
a thermal model of the well and near-wellbore region is required. The software must be able to
calculate the flowing well temperature based on reservoir, fluid, well, and completion properties.
Variables that control flow in the reservoir, such as permeability, are adjusted until the measured
and calculated data coincide. Once a match is achieved, the distribution of the variables in the
reservoir zones is uniquely defined and the model is used to determine the flow rate from the
individual zones.
THERMA thermal modeling and analysis software was developed by Schlumberger specifi-
cally for analyzing distributed temperature data in the wellbore and near-wellbore area to define
flow. The workflow of a typical thermal analysis using THERMA software is outlined below.

Data input
Basic information about the well, such as on land or offshore, the height of the wellhead above
the reference level, and the wells trajectory with respect to vertical, must be specified.
The user then inputs the geothermal gradient. It can be a linear interpolation between the sur-
face and bottomhole temperature, adjusted for true vertical depth (TVD) where necessary, or a
TVD-versus-temperature list if the user has more detailed knowledge of the geothermal gradient.
When modeling a well thermally, it is important to enter all aspects of the completion, includ-
ing the casing and tubing steel and surrounding fluids and cements, to ensure the most accurate
simulation.
A comprehensive layered reservoir model is required to define flow for each zone. The near-
wellbore reservoir model should employ permeability (horizontal and vertical), pressure, rock
properties, flowing-fluid properties, reservoir thickness, skin effect, and drainage radius.
Different models are available to model horizontal or vertical well flow.
The flowing fluids are defined by their black-oil properties: API gravity, GOR, water cut,
and gas- and water-specific gravities. If known, the calibration data at bubblepoint can also
be entered.

Temperature calculation
To successfully model the temperature response in a flowing well, the modeling software first
determines the pressure response throughout the system. This is required for calculation of the
fluid properties and flow rates in the formation and at all depths up the well to determine the
thermal properties of the fluids and thus the temperature response of the system. The pressure
response is modeled using NODAL analysis. The user has the option to calculate the flowing well
pressures by specifying any two of these three factors: surface flowing pressure, reservoir pres-
sure, or flow rate.
The program outputs a plot of the flowing wellbore pressures together with the reservoir pres-
sure for each flowing interval.

The Essentials of Fiber-Optic Distributed Temperature Analysis THERMA Thermal Modeling Software 25
 Once the flow rates and properties have been determined, the software calculates the tem-
perature response in the well as a function of the duration of flow (steady-state or user-specified
transient flow), convection, and conduction between the reservoir layers, the well, and sur-
rounding formations. The calculated temperature in the wellbore or at any radius into the
formation can also be determined.

DTS data analysis

To analyze the DTS temperature data using a thermal model, temperature data representing
stable flowing conditions should be chosen. DTS data can be imported into the software and
depth corrected to the wellhead and turnaround-sub depths. Appropriate filtering is then applied
to produce a temperature curve of the required depth and thermal resolution for flow analysis.
The required resolution is typically 0.1C. The temperature curve can be imported into the ther-
mal analysis model for comparison with its output.

Flow calculation
To calculate the flow rate by zone, the near-wellbore model must be adjusted until the computed
temperature profile matches the DTS-measured profile. Parameters that can be varied to control
the model-calculated flow rate are permeability, reservoir pressure, skin effect, flowing fluid
properties, and water cut. Once the fit between measured and calculated temperatures is
achieved, the flow rate calculated by the software reflects the actual well flow rate. This match
can be achieved in several ways.
Initially when the well is flowing a single-phase fluid (oil), the main unknown is reservoir
permeability. Reservoir pressure can be obtained from formation tests, and water cut is assumed
negligible. Thus the fit to the model is achieved by varying the layer permeabilities.
If a good estimate of reservoir permeability is also available, it is possible to vary the near-
wellbore skin effect to achieve a fit between the model and the DTS data for identifying zones
that have yet to clean up.
Over time and assuming the well is still flowing a single-phase fluid, zone contributions may
change because the layer pressures are changing. Using the permeabilities obtained in the early-
time fit, a match to later data can be achieved by varying reservoir pressure, thereby inferring the
effect of well production on the reservoir pressures over time.
If water breakthrough occurs, DTS data can indicate which layers have flooded by correlating
the change in layer flow rate with an independent surface measurement of water cut. The model
can be rerun to achieve best fit with the measured DTS data by changing an individual layers
water cut to an appropriate value and decreasing the layers permeability to be consistent with
the change in the relative permeability as a function of water cut.
Water-breakthrough thermal effects are often too subtle to be observed by the temperature
measurement, but they result in an associated change in layer flows. Consequently, it is not pos-
sible to determine pressure and water cut at the same time because there are too many unknown
variables.
A genetic algorithm allows the user to automatically calculate the optimum values of perme-
ability, pressure, or skin effect, rather than having to repeat the calculation and estimate new
values each time.

26
 Example: Flow from a multilayered reservoir5
Well installation
This deviated well was drilled from an offshore platform, and oil is produced by an electrical sub-
mersible pump (ESP). The DTS was permanently installed across the reservoir by attaching the
1
4-in. control line protecting the optical fiber to a 2 38-in. tubing stinger hung below the ESP. The
reservoir interval was completed with an expandable sand screen.

Well startup
Geothermal data were recorded before the well was brought on production. Figure 20 shows the
reservoir thermal profiles warming up as the well starts to produce. Major inflow intervals are
identified as temperature steps in the thermal profile. They correlate with the major sands shown
on the resistivity log.

X200
ESP
Warm-up
with time
X300

X400
Depth (m)
Significant inflows

X500
Geothermal
Well startup
Startup + 2 D
X600 Startup + 4 D
Startup + 6 D

50 55 60 65 70 1 10 100
Temperature (C) Resistivity (ohm-m)

Figure 20. Correlation of major inflows with the resistivity log.

THERMA Thermal Modeling Software 27

 Flow allocation by permeability variation
THERMA thermal modeling software calculates the inflow contribution by comparing a temper-
ature profile with simulated data. Figure 21 shows that 10 zones are contributing various amounts
of oil. The model does not calculate temperature rise at the ESP, but it does account for pressure
change at the pump.

Flow rate (B/D)

0 500 1,000 1,500 2,000 2,500 3,000 3,500 4,000
X200
DTS
data
X300
Calculated
profile
X400

Depth (m) Flow distribution

X500

X600

X700
50.0 52.5 55.0 57.5 60.0 62.5 65.0 67.5 70.0
Temperature (C)

Figure 21. Inflow distribution profile.

28
 Inflow distribution changes
Zonal contributions vary as the reservoir pressure in each zone changes over time. Figure 22
shows two thermal profiles of the well taken on different dates. The zone at X475 m was initially
a major contributor; however, after 8 months of production, it stopped contributing. The thermal
profile also indicates that flow stopped below X600 m; that change is attributed to sand fill.

X200

X300

X400
Depth (m) Flow increases

Flow
X500 decreases

Geothermal
Intial flow profile
X600 8-month Flow
production stops
50 55 60 65 70 1 10 100
Temperature (C) Resistivity (ohm-m)

Figure 22. Production changes over time.

THERMA Thermal Modeling Software 29

 Flow allocation by pressure variation
By using the reservoir permeabilities calculated during well startup, THERMA modeling software
can calculate the inflow contributions at later times by fitting the calculated temperature to the
measured temperature at varying reservoir pressures. Note that the pressure in the thin zone at
X470 m has decreased significantly whereas the pressure in the interval below decreased by a
smaller amount (Fig. 23).

X200
March reservoir pressure
November reservoir pressure
X300

X400

Depth (m)
X500

X600

X700
1,200 1,300 1,400 1,500 1,600 1,700 1,800
Reservoir pressure (psi)

Figure 23. Reservoir pressure changes over the time.

30
 Shut-in crossflow
On shut-in the temperature profile should cool back to the geothermal gradient; however, this is
not the case in this well. The shut-in temperature profile shows that flow from both above and
below X475 m is crossflowing into a permeable zone (Fig. 24). The crossflow confirms lower reser-
voir pressure in this zone than in the surrounding reservoir layers, as calculated by the THERMA
software, and explains why it had ceased to flow previously.

X200

X300

X400
Depth (m)

Crossflow
X500
Geothermal
Shut-in after
8 months
Intial flow profile
X600
8-month
production
50 55 60 65 70 1 10 100
Temperature (C) Resistivity (ohm-m)

Figure 24. Crossflow during the shut-in.

Summary of results
The well has been continuously monitored with the DTS system from the time of completion, and
oil production from more than 10 zones in this multilayered reservoir can be clearly identified.
The data confirm well cleanup when production commences and show that two zones change
their flow contributions significantly over the first eight months of production.
Also shown is the cessation of flow below X600 m, which is attributed to sand fill.
On shut-in, oil crossflows into the depleted zone, confirming that it has a lower reservoir
pressure than the surrounding intervals.

THERMA Thermal Modeling Software 31

 Joule Thomson Effects

Flow in horizontal wells

In a perfectly horizontal flowing well, the geothermal gradient does not provide any measurable
thermal response; however, the Joule Thomson effect can cause a flow-related change in the
well-fluid temperature. This can be interpreted to indicate flowing intervals and flow contribu-
tions.
In horizontal oil wells, the Joule Thomson effect is relatively small, necessitating the highest
possible resolution in the acquisition of temperature data.
Figure 25 demonstrates the effect of Joule Thomson warming resulting from one- and two-
point inflow and a uniformly distributed pattern of inflow. The well is horizontal with a 412-in.
completion along the reservoir, flowing 20,000 B/D of 30 API oil with a GOR of 100 at a reservoir
depth of 3,000 ft TVD.
A single-point inflow at the toe (a) is characterized by a linear increase in temperature along
the horizontal section.
Two-zone flow (b) is characterized by an increase in slope of the temperature profile at the
point of the second inflow.7
In a homogeneous reservoir interval (c), the inflow profile is linearly distributed over the entire reser-
voir interval. A distributed inflow is characterized by a nonlinear temperature increase that is
smaller than a one- or two-point inflow increase because it is caused by multiple single-point inflows.

Flow rate (B/D)

0 10,000 20,000 30,000 40,000 50,000
4,000
c a
b
5,000

6,000
Flow distribution

Depth (ft) 7,000

8,000
c Calculated
a temperatures
9,000
b

10,000
144 146 148 150 152 154
Temperature (F)

Figure 25. Joule Thomson warming with distributed inflow.

The Essentials of Fiber-Optic Distributed Temperature Analysis Joule Thomson Effects 33

 The amount of Joule Thomson warmingor cooling in the event of gasis dependent on the
fluid properties, flow rate, and casing diameter that define the pressure drop along the casing. In
low-flow-rate wells with large-diameter casing, the effect is negligible and increases as the casing
size decreases and flow rate increases.
Figure 26 shows the magnitude of the Joule Thomson effect on 30 API oil with a GOR of 100
over a range of flow rates and casing sizes for distributed flow and for a single-point entry at the
toe along a 6,500-ft horizontal interval with a geothermal gradient of 1.5F/100 ft.
The effect of decreasing the API gravity of the oil is also shown. Gas has the opposite effect
and causes cooling rather than heating.

8
4.5-in. single entry
5.5-in. single entry Decreasing
7.0-in. single entry API
6 4.5-in. distributed
5.5-in distributed
7.0-in. distributed
4.5-in. SE 22 API
Temperature
4
rise
(F)

0
0 10,000 20,000 30,000 40,000 50,000
Flow rate (BOPD)

Figure 26. Magnitude of the Joule Thomson effect on horizontal flow.

Flow within the reservoir

Horizontal wells normally have a low well drawdown; so in high-permeability reservoirs the near-
wellbore Joule Thomson effects are small. In some low-permeability reservoirs, however, there is
enough drawdown-pressure drop in the near-wellbore region to generate significant cooling at
the sandface in gas or high-GOR oil wells.
The cooling is directly related to the reservoir permeability and the GOR of the oil. As the GOR
increases, so does the cooling effect.
The example shown in Fig. 27 is a horizontal well producing at a rate of 3,000 B/D with five
permeable intervals. The two heel intervals have permeability of 5 mD, the middle interval
0.5 mD, and the two toe intervals 0.1 mD. The model shows that increasing GOR from 2,000 (a)
to 3,000 (b) and 4,000 (c) results in increased cooling in the more-permeable intervals.
The reason the low-permeability interval has a lower Joule Thomson effect than the high-
permeability interval for similar values of reservoir drawdown is that the flow rate in the
low-permeability interval is much lower and the cooled fluid is warmed by the surrounding
geothermal rock faster than the fluid in the higher-flow-rate interval is warmed.

34
 Flow rate (B/D)
0 2,000 4,000 6,000 8,000 10,000
0
a

2,000

Flow distribution c b
4,000

Depth (ft)
6,000

8,000
Calculated temperatures

10,000
60 80 100 120 140 160
Temperature (F)

Figure 27. Reservoir Joule Thomson cooling as a function of reservoir permeabilityfirst example.

Figure 28 shows an example of the effect of the two heel zones preferentially flowing com-
pared with the two toe zones or the whole interval with the same flow rate and GOR. The reduced
drawdown as the whole interval flows results in lower Joule Thomson cooling. Although it is rel-
atively simple to determine the permeability decrease toward the toe of the well as the well
cleans up, it is not possible to determine the permeability decrease toward the heel of the well
because the temperature of the flow from the toe does not change.

Flow rate (B/D)

0 2,000 4,000 6,000 8,000 10,000
0

2,000

4,000
Flow distribution
Depth (ft)
6,000

Calculated temperatures
8,000

10,000
60 80 100 120 140 160
Temperature (F)

Figure 28. Reservoir Joule Thomson cooling as a function of reservoir permeabilitysecond example.

Joule Thomson Effects 35

 Example: Horizontal-well Joule Thomson flow6
Well installation
This horizontal well has an openhole completion and is produced by gas lift. The double-ended
DTS system was permanently installed across the reservoir by attaching the 14-in. control line
protecting the optical fiber to a 2 38-in. tubing stinger hung below the production tubing.

Well cleanup
The well was put on production in July and by August was producing with stabilized flow. The
DTS-measured temperature profile indicates that the well was primarily flowing from X4,000 ft
up to the heel. The increase in Joule Thomson cooling from X4,000 to X6,000 ft over time indi-
cates that the well is cleaning up across this interval and that the flow distribution along the
reservoir has changed (Fig. 29).

112
Geothermal
August 19
October 29
109 No change from
geothermalno flow
Joule Thomson cooling

Temperature 106
(C)

Change
103 AugustOctober

Inflow JulyAugust

100
X1,000 X2,000 X3,000 X4,000 X5,000 X6,000 X7,000
Depth (ft)

Figure 29. Reservoir Joule Thomson cooling as a function of well cleanup.

36
 Figure 30 shows the change in the computed inflow profile over time produced with the
THERMA model. The model was run using given reservoir permeabilities and pressures. The
change in flow was achieved by varying the skin factor; it was initially high toward the toe and
had to be reduced to fit to the October data.
Although the wells total production decreased over the period, production increased from the
X3,500- to X6,000-ft interval as a result of well cleanup.

112 1,600

109 1,200

Temperature 106 800 Flow rate

(C) (BOPD)

103 400

Casing shoe
100 0
X1,000 X2,000 X3,000 X4,000 X5,000 X6,000 X7,000
Depth (ft)

Geothermal THERMA August flow

August 19 THERMA October flow
October 29

Figure 30. Flow calculated by the THERMA model as a function of well cleanup.

Summary of results
The well was monitored in August and October with the DTS system. Well cleanup with depth
over time can be clearly identified.
The data were used to determine the rate of cleanup of this long, horizontal, high-GOR
producing zones and enabled modifying well stimulation procedures to improve the cleanup
response and monitor the effect on future wells.

Joule Thomson Effects 37

 Water Injection Wells

Distributed temperature surveillance has been used conventionally to monitor the performance of
water injectors with the warm-back technique, wherein the well is shut in for a period of time and
the temperature response is recorded while the well warms back toward the geothermal gradient.
Under normal injection conditions, the cold water injected into the well cools all the sur-
rounding rock, including the nonpermeable intervals above the reservoir. The only information
that can be obtained during injection is the lowest extent of fluid injection. If the flow rate is low,
the profile is determined by using the geothermal gradient and inflow into the reservoir could be
analyzed; however, this is normally not the case.
Once injection ceases, the surrounding rock warms back to the geothermal gradient over time
as a function of the formations thermal properties (Fig. 31). If a permeable interval was injected
with cold water, the rock cools to a much greater radius from the wellbore than the intervals that
were not injected with cold water, such as those behind casing. The injected intervals warm back
at a much slower rate than those that were not injected. The magnitude of this effect is a function
of the injection rate, interval permeability, time, and the thermal properties of the fluid and rock.

Injection rate (B/D)

0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000
0

10-D injection
1,000

1-D warm back

2,000
10-D warm back
Depth (ft)
3,000

Injection
distribution
4,000

5,000
100-D warm back
0 20 40 60 80 100 120 140 160
Temperature (F)

Figure 31. Injector warm-back response.

The Essentials of Fiber-Optic Distributed Temperature Analysis Water Injection Wells 39

 Figure 32 shows the modeled response of a 500-ft thick interval of 10-mD rock with a 100-mD
interval in the middle. The injection rate is 3,000 BWPD at 85F for 10 days. Because the higher-
permeability interval takes most of the water, it is preferentially cooled and stays at the injection
temperature, while the lower-permeability intervals start to warm back 24 hr after the well is
shut in. The effect of the permeability contrast increases as the shut-in time increases.

Injection rate (B/D)

0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000
0

10-D injection
1,000

1-D warm back

2,000
10-D warm back

Depth (ft) 100-D

3,000
warm back
Injection
distribution
4,000

5,000

0 20 40 60 80 100 120 140 160

Temperature (F)

Figure 32. Injector warm-back response (10 days of injection).

40
 If injection has taken place for longer than a year, the zones taking water can be identified,
but a long shut-in time is required to see permeability contrasts in the formation (Fig. 33).
The time required to achieve a clear warm-back profile that highlights permeability contrasts
in the formation is proportional to the injection time, but it can be reduced by injecting water of
a different temperature for a short period immediately prior to performing a warm back or by
tracking the hot water produced by the warm back in the tubing as injection is recommenced.

Injection rate (B/D)

0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000
0

1-yr injection
1,000

1-D warm back

2,000
10-D warm back
Depth (ft) 100-D warm back
3,000

Injection
4,000 distribution

5,000

0 20 40 60 80 100 120 140 160

Temperature (F)
Figure 33. Injector warm-back response (steady-state injection longer than 1 year).

Water Injection Wells 41

 Hot-water-injection technique
If injection has been ongoing for an extended period, the shut-in period must be very long, which
is normally impractical. An alternative approach is to inject water of a different temperature for
a short period to highlight permeability contrasts.
Figure 34 shows the response of injecting water at steady-state conditions (1 year) into a
reservoir. A 24-hr shut-in would not produce an observable response over the reservoir interval,
although zones that have not been taking fluid would exhibit warm back.

Injection rate (B/D)

0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000
0

1,000
24-hr injection
of hot water
2,000
1-D warm back
Depth (ft) 5-D warm back
3,000

4,000 Injection
distribution

5,000
Cool back
0 20 40 60 80 100 120 140 160
Temperature (F)

Figure 34. Injector warm-back response (steady-state injection, followed by 24-hr injection of warm water and warm back).

A 24-hr injection of 10F warmer water followed by a shut-in and warm back can be used to
identify the high-permeability interval as the reservoir interval cools down toward the steady-
state injection baseline temperature. The time needed to highlight the permeability contrasts is
thus reduced significantly.
A short period of water injection at a different temperature, followed by a short warm-back
period, enables identification of the zones taking fluid, and the high-permeability layers can be
defined from the warm-back response.

42
 Hot-slug velocity measurement
If the well has been on injection for an extended period, the reservoir will have completely cooled
to the injection temperature and will stay cold long after injection has stopped. An alternative to
the hot-water injection technique is to allow the warm-back period to generate a slug of hot water
in the tubing immediately above the reservoir and to track it once injection is recommenced.
When the well is shut in, the water in the tubing above the reservoir warms up quickly, usually
in a few hours, because of conduction from the uninvaded formation. These conditions produce
a volume of hot water in the tubing just above the reservoir.
Once injection is recommenced, this hot slug of water can be tracked by the DTS fiber-optic
system, recording at a suitably high acquisition rate, as it moves down the reservoir interval (Fig. 35).
The velocity of the slug can be determined and represents the flow profile into the reservoir.
As with a spinner log, the measured velocity is slightly lower than the true liquid velocity. This
effect results from heat transfer at the interface between the hot slug and cooler water below.
Given a particular injection scenario, the correction necessary to obtain the true velocity can be
calculated, but normally only the percentage injection into a particular interval is required and
can be obtained directly from the analysis.
Unfortunately the hot slug cannot be injected from the surface, except in very shallow wells,
because the slug loses its heat by conduction to its surroundings as it travels down the well. The
hot slug has an advantage over the hot-water-injection technique because it requires only a short
shut-in of the well and has much better resolution and velocity accuracy than can be achieved by
warm-back analysis.

Flow velocity
High permeability
Low permeability
Low permeability

120

110
Temperature 27/03 01:06
(F) 27/03 01:02
100 27/03 00:53
27/03 00:44
4,500
4,600 27/03 00:35
4,700
4,800
4,900 27/03 00:26
Depth
(ft)

Figure 35. The principle of hot-slug velocity measurement.

Water Injection Wells 43

 Example: Hot-slug velocity in a horizontal injector6
Well installation
This horizontal water injector has an openhole 6 14-in. diameter completion. The DTS was perma-
nently installed across the reservoir by attaching the 14-in. control line protecting the optical fiber
to a 2 38-in. tubing stinger hung below the production tubing.

Well warm back

Figure 36 shows the injection temperature profile down the well and along the reservoir, along
with the warm-back response during a 24-hr shut-in. The reservoir injection intervals are iden-
tified by the colored backgrounds.

Slow warm back

200

Temperature
(F) 150 14/03 20:14
14/03 16:59
14/03 11:23
14/03 05:48
X,000
14/03 00:13
X5,000

Depth (ft) X0,000 13/03 18:37

Figure 36. Hot-water slug created in the tubing by a 24-hr shut-in.

During injection, water was flooding only down to X0,000 ft, and below this depth the temper-
ature increases to the geothermal temperature. On shut-in, the interval taking flow could easily
be identified because it stayed cold while the water in the tubing rapidly warmed back toward
the geothermal gradient. Also, a large temperature difference developed at the tubing shoe
between the intervals that had been flooded and those that had not.
Figure 37 shows what happens when the injection is restarted. The hot water generated at the
tubing shoe during the warm-back period floods down the reservoir and can be tracked by the
DTS system and converted to the percentage inflow distribution.

44
 Hot slug

200

Temperature
(F) 150 14/03 22:29
14/03 22:04
14/03 21:38
X,000 14/03 21:12

X5,000 14/03 20:46

Depth (ft) 14/03 20:21

X0,000

Figure 37. Monitoring the hot slug after injection is recommenced.

Figure 38 shows the inflow distribution calculated for March and October using the hot-slug-
injection method. The results indicate that the injection profile of the well is changing with time,
and the well is flooding deeper in October. This may be caused by pressure building up at the heel
or by slow cleanup from the heel to the toe over time.

1.2
March distribution
October distribution
1.0

0.8

Inflow 0.6
distribution
0.4

0.2

0
X7,000 X8,000 X9,000 X0,000 X1,000
Depth (ft)

Figure 38. Hot-slug velocity measurement results.

Summary of results
The injection monitoring showed that the water was flooding the reservoir farther along the
horizontal section as time passed.
The data confirm the decision to drill long injectors in this reservoir, show that the injectors
will clean up eventually, and indicate the injection profile at any given time.

Water Injection Wells 45

 Flow Between Wells

Injecting cold water into a reservoir causes the reservoir rock to cool over time, as determined
by the temperature of the injected water, rate of injection, and duration of injection. The water
at the periphery of the flood warms rapidly to the geothermal temperature of the formation
through which it is passing. However, as large volumes of cold water are injected into the reser-
voir, the rock cools radially from the wellbore with time, forming a thermal cold front that follows
the waterflood front into the reservoir at approximately half the radial distance (Fig. 39).
As injected water from a nearby well is produced, the temperature of the injected water arriv-
ing at the producer will eventually drop from the geothermal temperature to a value depending
on the zone thickness, injection rate, and the temperature and distance between the injector
and producer.

Cold front Flood front

Reservoir temperature

Temperature

Figure 39. Radial injection of cold water.

The Essentials of Fiber-Optic Distributed Temperature Analysis Flow Between Wells 47

 Nearly 50 years ago Lauwerier 8 outlined the theory for the movement of a hot or cold interface
in a reservoir caused by the injection of water of a different temperature. A calculation using his
equations is shown in Fig. 40.

80
Reservoir temperature
60
Water thermal front Waterflood
Temperature
40 front
(C)

20
Injection temperature
0
0 100 200 300 400 500
Radius (m)

Figure 40. Thermal front resulting from the radial injection of 10,000 B/D of 20C water into a 30-ft, 25%-porosity zone for 2 years.

Using Lauweriers equations and assuming a radial flood, it is possible to calculate the dis-
tance of the thermal front from a typical water injector injecting at a rate of 10,000 B/D for a
variety of zone depths and times (Fig. 41).

2,000
5-ft zone
20-ft zone
1,500 100-ft zone
300-ft zone
Radial
distance 1,000
(ft)

500

0
0.1 1 10
Time (yr)

Figure 41. Thermal front movement through a 25%-porosity zone with respect to time and zone thickness.

The larger the injection interval, the longer the thermal front takes to reach the producing
well. The thermal front has a radius approximately half that of the waterflood front, so a produc-
ing well that has a decrease in produced water temperature must have been producing water for
a significant period.
Nevertheless, if it is known when injection commencedeither because a new injector was
drilled or because an injector was shut in for a time and then restartedthe arrival of the ther-
mal front at a nearby producer indicates the sweep velocity between the injector and producer.
The thermal front can be readily detected by distributed temperature monitoring.

48
 Example: Cold-water breakthrough7
Well installation
This highly deviated producer has a 7-in. diameter cased hole completion with fluids pumped
to the surface by an ESP. The DTS was permanently installed across the reservoir by attaching
the 14-in. control line protecting the optical fiber to a 2 38-in. tubing stinger hung below the pro-
duction tubing.

Breakthrough scenario
Figure 42 shows a measured DTS temperature profile over the reservoir interval and also the
THERMA model temperature and calculated flow rate. The lowest producing interval is close to a
seawater injector, and relatively cold (133F) injection water is being produced from that interval.
Farther up the well, the cold seawater is mixing with oil flowing in at the geothermal temper-
ature from the other producing intervals. The thermal model has also been run with the lowest
interval producing water at the geothermal temperature for comparison. In this case the flowing
temperature loses its resolution and becomes less than 2.0F from the geothermal gradient.
In this well the sole reason for the large temperature difference is that relatively cold injec-
tion water is being produced from the zone at the toe. If cold-water breakthrough occurred in any
of the higher intervals, the effect would be lessened because the water would be mixing with a
flow of oil at approximately the geothermal temperature.

Flow rate (B/D)

0 2,000 4,000 6,000 8,000 10,000
X8,000
DTS data
Flow ESP
X9,000 distribution

X0,000

X1,000
Depth (ft)

X2,000

Cold-water
X3,000 breakthrough Simulated
geothermal-water
breakthrough
X4,000

120 130 140 150 160

Temperature (F)

Figure 42. The effect of cold-water breakthrough at the bottom producing interval.

Summary of results
This well has been continuously monitored for more than five years, and although the flow
contributions have changed over this time, the well is still producing cold water from the toe and
a significant inflow from the zones close to the heel.

Flow Between Wells 49

 Low-Rate Gas Wells

The difference between oil- and gas-well temperature profiles is that gas wells may exhibit large
Joule Thomson cooling effects as the gas enters the borehole. Consequently, the initial wellbore
flowing temperature is lower than the geothermal gradient; however, the equations defining the
thermal profile from the reservoir up the well are the same for gas as for oil and water (Fig. 43).

Gas rate (MMcf/D)

0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
0

1,000

Calculated
2,000 temperature

Depth (ft) 3,000

4,000 Flow
distribution
Joule Thomson
5,000 cooling on inflow

0 20 40 60 80 100 120 140 160

Temperature (F)

Figure 43. Gas-well thermal profile.

The Essentials of Fiber-Optic Distributed Temperature Analysis Low-Rate Gas Wells 51

 Figure 44 shows the Joule Thomson cooling effect for a 10-MMcf/D radial flow of gas at steady-
state conditions in a 50-ft interval of 10-mD rock. The temperature change is approximately 20F,
and half of that change occurs in the region very close to the wellbore.
The thermal and pressure radial responses are similar in character.

155 3,100
Pressure 3,000
150

2,900
145
Temperature 2,800
Temperature (F) 140 Pressure (psi)
2,700
135
2,600
130 2,500

125 2,400
0 200 400 600 800 1,000
Radial distance (ft)

Joule Thompson Calculator

Steady-State Conditions (Assumes Radial Flow)
Reservoir pressure (psi) 3,000

Reservoir temperature (F) 150

Flow rate (MMcf/D) 10

Formation volume factor, Bg 0.004

Fluid viscosity (cp) 0.03

Joule Thompson coefficient (F/psi) 0.0416

Formation permeability (mD) 10

Formation thickness (ft) 50

Well diameter (in.) 7

Reservoir radius (ft) 1,000

Pressure decrease (psi) 491.5

Temperature decrease (F) 20.4

Figure 44. Joule Thomson cooling in the formation.

52
 Kunz and Tixiers subtangent method
In 1955 Schlumberger published a method by Kunz and Tixier for the interpretation of temperature
logs in gas wells that is particularly applicable to low-rate (less than ~1 to 2 MMcf/D) gas wells.9
Kunz and Tixier calculated that the flow rate is proportional to the subtangent of the temper-
ature curve above the inflow interval (Fig. 45).
Figure 46 shows a thermal model of three gas inflow zones contributing 100 Mcf/D each with
~20F Joule Thomson cooling at each inflow.

C(Tg Tw )
1,000 Q= ,
dT

dH
2,000
Geothermal where
gradient
Depth (ft) dT Q = flow rate (MMcf/D)
3,000
dH C = time-dependent constant
Tw Tg = geothermal temperature (F)
4,000 Tg
Tw = well temperature (F)
dT
= slope of intercept
5,000 dH

50 100 150
Temperature (F)

Figure 45. Kunz and Tixiers subtangent method.

Gas rate (MMcf/D)

0 1 2 3 4

Flow 100 Mcf/D

distribution
4,000

100 Mcf/D
Joule
Depth (ft) Thomson
cooling
100 Mcf/D

5,000

Calculated
temperature

80 100 120 140 160

Temperature (F)

Figure 46. Model of three gas inflow zones in a low-rate gas well.

Low-Rate Gas Wells 53

 In this example the Kunz-Tixier function can be used to estimate flow rate in the lower
interval. The function becomes increasingly less stable as the flow rate increases, and the tem-
perature profile trends toward a straight line of negative slope, as seen in the upper interval.
The constant C is a function of the time of production and the gas and formation thermal proper-
ties, so the subtangent curve can be converted to a flow rate if C is calculated. It is customary to
identify only the inflow zones and the proportion of flow from each zone, and the subtangent approach
does this with a reasonable degree of accuracy. This analysis method is also valid for oilwell flow.

Example: Low-rate multizone gas well

Well installation
This vertical gas-producing well has an openhole 3 12-in. diameter completion over the reservoir
interval and is producing 1.1 MMcf/D gas. The DTS was temporarily installed across the reservoir
by running an optical fiber encased in a 316-in. steel tube into the well while it was flowing.

Flowing temperature profile

Figure 47 compares the temperature profile computed by the THERMA model with the DTS-
measured values and the calculated inflow distribution. This reservoir is partially depleted, so
some intervals have lower drawdown and Joule Thomson cooling than others. The interval at
X3,950 ft is clearly not cooling as much as the interval below; rather it is adding warmer gas to
the flow stream from below because it has a lower reservoir pressure. In this example, reservoir
permeabilities were assumed and the fit of the profile calculated by the THERMA model to the
temperature data was achieved by varying the reservoir-layer pressures.

Gas rate (MMcf/D)

0 1 2 3 4
X1,000
DTS data
Calculated
temperature
X2,000
Flow
distribution
Depth (ft) X3,000
Differing
layer
pressures
X4,000

250 275 300 325 350

Temperature (F)

Figure 47. Low-rate multizone gas well.

Summary of results
Analysis of the DTS temperature data indicates which intervals are flowing and their percentage
contributions.
The data also indicate which intervals have lower pressures because of depletion.

54
 High-Rate Gas Wells

In high-rate gas wells the Kunz-Tixier subtangent becomes negative because of Joule Thomson
cooling of the gas related to the pressure drop in the wellbore and the cooling caused by the pres-
sure drop in the formation. Therefore it is not possible to use the Kunz-Tixier approach to
determine the percentage of flow from each producing interval.
Figure 48 shows the effect of three reservoir intervals producing 10 MMcf/D each, where the
Joule Thomson cooling in the reservoir is the same for each zone. It is reasonable to assume that
this is the case if the reservoirs have similar drawdowns; that is, all have a similar reservoir pres-
sure gradient to that in the flowing wellbore. The temperature profile above each entry is
essentially a straight line.

Gas rate (MMcf/D)

0 30 60 90 120
Calculated
Flow temperatures
distribution 10 MMcf/D
4,000

10 MMcf/D
Depth (ft)
10 MMcf/D

5,000

Joule Thomson cooling

80 100 120 140 160
Temperature (F)

Figure 48. High-rate gas flow (10 MMcf/D per zone).

If all the reservoir intervals have the same reservoir pressure, the well will not crossflow on
shut-in, and a first approximation to the Joule Thomson cooling for each interval is a line parallel
to the geothermal gradient through the inflow temperature of the lowest producing zone (assum-
ing that the producing zones are close to one another). Individual zone values can be further
defined by monitoring the zone warm-back temperatures during a well shut-in and extrapolating
them back to the flowing conditions.
The quicklook method outlined for multizone geothermal reservoirs can be used to estimate
the contribution of each zone using the Joule Thomson cooling line parallel to the geothermal
gradient as the reference line toward which additional inflows trend.
If the reservoir intervals have different pressures and exhibit different Joule Thomson inflow
cooling, it will be difficult to determine the flow contribution of each interval because there are
too many unknowns.

The Essentials of Fiber-Optic Distributed Temperature Analysis High-Rate Gas Wells 55

 Water Detection
Using Temperature

The DTS system measures in situ well temperatures that can readily be related to the mass flow
rate of oil and gas. The system, however, cannot measure water cut in typical oil-producing inter-
vals a few hundred feet thick with a typical geothermal gradient of 1.0F/100 ft.
Although the thermal difference between oil and water at the reservoir level is negligible,
other factors cause the flow rate from the watered-out zone to change. In some cases a change
in flow rate may be correlated with the increase in surface-water production to identify the inter-
vals where water has broken through.
Other factors include the change in the Joule Thomson effect in the near-wellbore region,
fluid viscosity, relative permeability, and a change in reservoir-zone pressure.

Low Joule Thomson effects (oil wells)

In typical stacked multizone oil reservoirs the thermal response to an increase in water cut may
be subtle because the decrease in fluid viscosity is partially offset by a decrease in the relative
permeability in the radial flow equation and because the near-wellbore Joule Thomson warming
for oil and water is small, particularly in high-permeability sands. A change in temperature may
be caused by a change in flow associated with a change in reservoir-zone pressure or draw-
downor bothas the lift characteristic of the well changes, rather than the zone water cut.
Because thermal changes are small, temperature can be used to monitor reservoir pressure
over time, regardless of water-cut changes, as long as the effects of the other contributing para-
meters are known to be small.

The Essentials of Fiber-Optic Distributed Temperature Analysis Water Detection Using Temperature 57
 Figure 49 compares the thermal responses of three producing intervals 100 ft apart, where (a)
all three zones are producing oil, (b) all zones are producing oil with 30% water cut, and (c) the
upper and lower zones are producing oil while the center zone produces 90% water.

Flow rate (B/D)

0 400 800 1,200 1,600 2,000
8,000
(b) Oil with 30%
water cut in
zones 1, 2, and 3
8,500
(a) Oil
all zones (c) Oil with
90% water
cut in zone 2
9,000
Depth (ft)

9,500
Flow distribution

10,000

141 142 143 144 145 146 147 148 149 150 151
Temperature (F)

Figure 49. The effect in a vertical oil well of an increase in water cut in the reservoir.

There is a small thermal difference between the three scenariosless than 0.2Fwhich
would not be detectable with the DTS system. However, the flow rate varies because of the change
in the wells lift characteristic when water flows, which alters the drawdown for each zone, and
because of the change in relative permeability in the middle zone when it waters out.
If the produced water is sourced from a cold-water injector or up a fault from a lower, hotter
interval, the produced water temperature will change over time as the hot or cold water eventu-
ally reaches the producing well. (See Flow between wells, page 47.) The production of
nongeothermal liquid from the reservoir can be identified by the associated change in tempera-
ture response.

58
 High Joule Thomson effects (gas wells)
In typical stacked multizone gas reservoirs, the thermal response to an increase in water cut may
be significant because the water breakthrough decreases the Joule Thomson cooling effect of the
gas in the near-wellbore region. If the well is monitored frequently and water breaks through, the
water-producing interval may be identified by its decrease in Joule Thomson cooling.
When water breaks through, the wells lift characteristic and zone drawdowns also change,
which may change the Joule Thomson cooling effects of all the producing intervals. Changing
reservoir-zone pressures over time can have the same result.
The effects are much larger than those for an oil well and are more easily identifiable. Figure 50
shows the analysis of a gas-producing vertical well. Layer pressures were changed to achieve the
fit required between the measured data and the Joule Thomson cooling at the near-wellbore
region of each zone. Some zones flow with warmer gas than others when their layer pressures are
lower than the original reservoir pressure. A second model with the water cut of the layer second
from top changed to 75% shows a significant change in the Joule Thomson cooling of this layer
with an associated change in the gas flow rate.

Gas rate (MMcf/D)

0 1 2 3 4
X0,000

DTS data

X1,000
Simulated
Flow water
distribution breakthrough
Depth (ft)
X2,000

Differing
layer
pressures
X3,000

250 275 300 325 350

Temperature (F)

Figure 50. The effect in a vertical gas well of an increase in water cut in the reservoir.

Water Detection Using Temperature 59

 References

1. Doll, H.G. and Perebinossoff, A.A.: Temperature Measurements in Oil Wells, presented at the
1936 Meeting of the Institute of Petroleum Technology.
2. Joule, J.P. and Thomson, W.: On the Thermal Effects of Fluids in Motion, Philosophical
Transactions of the Royal Society of London, London, England (1853), 143, 357365.
3. Ramey, H.J.: Well Bore Heat Transmission, paper SPE 96 presented at the SPE Annual
Meeting, Dallas, Texas, USA (October 811, 1961).
4. Curtis, M.R. and Witterholt, E.J.: Use of the Temperature Log for Determining Flow Rates in
Wells, paper SPE 4737 presented at the SPE Annual Technical Conference and Exhibition,
Las Vegas, Nevada, USA (September 30October 3, 1973).
5. Fryer, V., ShuXing, D., Otsubo, Y., Brown, G., and Guilfoyle, P.: Monitoring of Real-time
Temperature Profiles Across Multi-zone Reservoirs during Production and Shut-in Periods
Using Permanent Fiber-Optic Distributed Temperature Systems, paper SPE 92962 pre-
sented at the SPE Asia Pacific Oil & Gas Conference and Exhibition (APOGCE), Jakarta,
Indonesia (April 57, 2005).
6. Brown, G.A., et al.: Monitoring Horizontal Producers and Injectors During Cleanup and
Production Using Fiber-Optic-Distributed Temperature Measurements, paper SPE 84379
presented at the SPE Annual Technical Conference and Exhibition, Denver, Colorado, USA
(October 58, 2003).
7. Brown, G.A., Kennedy, B., and Meling, T.: Using Fibre-Optic Distributed Temperature
Measurements to Provide Real-Time Reservoir Surveillance Data on Wytch Farm Field
Horizontal Extended-Reach Wells, paper SPE 62952 presented at the SPE Annual Technical
Conference and Exhibition, Dallas, Texas, USA (October 14, 2000).
8. Lauwerier, H.A.: The Transport of Heat in an Oil Layer Caused by the Injection of Hot Fluid,
Applied Scientific Research (1955) 5, A, 145.
9. Kunz, K.S. and Tixier, M.P.: Temperature Surveys in Gas Producing Wells, presented at the
AIME Annual Meeting, Chicago, Illinois, USA (February 13, 1955).

The Essentials of Fiber-Optic Distributed Temperature Analysis References 61

 Appendix

350

t
0f
F 10

1.4
t
0f
F 10

1.2
250 0 ft
F 10
1.0
ft
Temperature F 100
0.8
(F)
0 ft
F 10
0.6
150

50
0 5,000 10,000 15,000 20,000
Depth (ft)
Chart 1. Geothermal gradientF/ft.

The Essentials of Fiber-Optic Distributed Temperature Analysis Appendix 63

 220

0m
10
C
3.0
m
170 100
C
2.5

0m
C 10

2.0

Temperature 120
(C) 00 m
C 1
1.5

m
100
1.0C

70

20
0 1,000 2,000 3,000 4,000 5,000 6,000
Depth (m)
Chart 2. Geothermal gradientC/m.

64
 Table 1. Pipe Size Versus Fluid Velocity for a Flow Rate of 1,000 B/D
Pipe OD Weight Pipe ID Fluid Velocity
(in. [mm]) (lbm/ft [kg/m]) (in. [mm]) (ft/min [m/min])
1.9 [48.26] 2.75 [4.09] 1.61 [40.89] 276.0 [84.12]

2.375 [60.33] 4.00 [5.95] 2.041 [51.84] 172.0 [52.43]

4.60 [6.85] 1.995 [50.67] 180.0 [54.86]

5.80 [8.63] 1.867 [47.42] 205.0 [62.48]

4.5 [114.30] 9.50 [14.14] 4.09 [103.89] 42.7 [13.01]

11.60 [17.26] 4 [101.60] 44.7 [13.62]

13.50 [20.09] 3.92 [99.57] 46.6 [14.20]

15.10 [22.47] 3.826 [97.18] 48.8 [14.87]

5.5 [139.70] 13.00 [19.35] 5.044 [128.12] 28.1 [8.56]

14.00 [20.83] 5.012 [127.30] 28.5 [8.69]

15.50 [23.07] 4.95 [125.73] 29.2 [8.90]

17.00 [25.30] 4.892 [124.26] 29.9 [9.11]

20.00 [29.76] 4.778 [121.36] 31.3 [9.54]

23.00 [34.23] 4.67 [118.62] 32.8 [10.00]

7 [177.80] 17.00 [25.30] 6.538 [166.07] 16.7 [5.09]

20.00 [29.76] 6.456 [163.98] 17.2 [5.24]

23.00 [34.23] 6.366 [161.70] 17.6 [5.36]

26.00 [38.69] 6.276 [159.41] 18.2 [5.55]

29.00 [43.16] 6.184 [157.07] 18.7 [5.70]

32.00 [47.62] 6.094 [154.79] 19.3 [5.88]

35.00 [52.09] 6.004 [152.50] 19.8 [6.04]

38.00 [56.55] 5.92 [150.37] 20.4 [6.22]

Appendix 65
www.slb.com/completions
09-CO-0139