Smart wells

Contribution to the “Jaarboek” of the “Mijnbouwkundige Vereeniging”, February 2001,

by J.D. Jansen,
Delft University of Technology, Department of Applied Earth Sciences,
Section Petroleum Engineering, e-mail: j.d.jansen@ta.tudelft.nl
&
Shell E&P Technology Applications and Research, Rijswijk,
The Netherlands: j.d.jansen@siep.shell.com.

Abstract
Smart well technology involves down-hole measurement and control of well bore and
reservoir flow. Drilling and completion techniques have advanced significantly over the last
years and allow for the drilling of complex multi-lateral and extended reach wells, and the
installation of down-hole inflow control valves, measurement devices for pressure,
temperature and flow rate, and processing facilities such as hydro-cyclones in the well bore.
Smart wells may allow us to go from passive/reactive production scenarios to
active/proactive production control. This could e.g. be achieved through influencing the flow
behaviour in the reservoir by imposing a pressure profile along the well bore based on results
of down-hole measurements, and if necessary, continuously updated dynamic reservoir
models. Full development of this potential requires a more systematic analysis of reservoir
and well bore flow in terms of modern measurement and control theory. In parallel this
should be combined with a revision of conventional production scenarios, and the
development of computational tools to rapidly design and assess the value of smart well
solutions.

Figure 1: Artist impression of a smart well with multiple branches, inflow control devices
and reservoir imaging functionality.

Smart Wells, February 2001 1/11

What are smart wells?

What is smart anyway?

Wells equipped with permanent down hole measurement equipment or control valves, and
especially those with both, are nowadays known as smart or intelligent wells; see Figure 1. In
the oilfield, just as in the real world, intelligence is not always a guarantee for success, and
the key question in the development of smart well technology is when the added functionality
also adds value. Dumb wells are sometimes the smartest solution.

Processes
Figure 2 shows a representation of oil and gas production as a feedback control process,
involving measurement, modelling and control. The picture of course not only applies to the
use of smart wells, but also to conventional production activities. However, it forms a good
starting point for the analysis of the added value of smart wells. Two major feedback cycles
occur, each on it’s own time scale:
• Daily production: On a scale of days to weeks, typical input variables are well head choke
settings, water injection pressures, or lift gas rates. Measured output from the process
includes production variables such as pressures, and oil, gas and water rates. Control will
often be driven by short time optimisation objectives, for example production targets or
utilisation rates. Down hole measurement has the scope to improve routine process
control (well surveillance) and production measurement (flow allocation), and down hole
control will allow for rapid reaction e.g. in case of gas or water breakthrough. Extensive
modelling will usually not be required, although some well bore flow and surface network
simulation may be necessary for accurate flow allocation.
• Reservoir management: On a time scale of months to years, the production process
essentially consists of draining the reservoir. In addition to the variables that control daily
production, input includes production engineering activities such as water or gas shut off,
re-completion, stimulation or even side-tracking or in-fill drilling. Measured output
involves production histories, well tests and reservoir images obtained from time lapse
seismic or other sources. Control is usually focused on maximising the asset revenues,
which often translates into maximising ultimate recovery and minimising operating
expenditure (OPEX). It is in this feedback process that the major value of smart well
technology can be expected, through reduced well intervention costs, a reduced number
of wells, accelerated production and, in particular, increased ultimate recovery. System
modelling will often involve extensive reservoir simulation, in addition to well bore and
surface flow modelling.

Smart Wells, February 2001 2/11

 Input System Output
(reservoir & wells)

CONTROL MEASURE

Control
Sensors
algorithms

System model
Measured
output

MODEL
Figure 2: Oil and gas production represented as a feedback control process, involving
measurement, modelling and control.

Integration
Smart well technology can be seen as a next step in the development from vertical wells, via
horizontal wells to multi-lateral wells. As with all these developments, the value of the
technology is not so much in the capability to drill and complete the wells, impressive as
these achievements may be. As was shown above, the value is in improved asset management
through reduced well intervention costs, accelerated production and, in particular, increased
ultimate recovery. Although the majority of the value of smart wells can expected to be
realised during the production phase of the petroleum life cycle, the decisions about the use
of smart well technology have to be made during the development stage, in particular during
field development planning (FDP).
The key objective during FDP is maximisation of the net present value (NPV) within the
constraints of the project. This involves comparison of a large number of development
concepts, usually in combination with a number of subsurface models to reflect geological
uncertainties. Early co-operation of reservoir engineers, production engineers and well
engineers, supported by the appropriate integrated software, is essential to achieve the
objective.
Another integration aspect concerns routing of real-time data to modelling software. This
involves data acquisition, transmission and storage in a data base, data brokering, quality
control, filtering and transfer to modelling tools. Expertise in automated production
operations has until recently mainly been gained for surface production equipment and needs
to be extended to down hole tools and data transmissions systems.

Hardware
Before discussing some examples of the use of smart well technology, it is useful to review
the present state of smart well hardware. The recent rapid increase in smart well applications
have to a large extent been driven by the rapid development of down hole measurement and
control equipment. Although costs are generally still high, the reliability of the equipment has
improved dramatically over the recent years, thus bringing more and more economic
applications within reach.

Smart Wells, February 2001 3/11

Measurement
• Single point measurement of pressure and temperature. Also known as “permanent down
hole gauges” which were already in use long before the term “smart well” became
fashionable. Most used are sensors using resonating quartz crystals: the resonance
frequency of the electrically excited crystals is a function of pressure and temperature.
Recent advances include the development of electric resonating diaphragms which have
the advantage of having no electronics down hole, and fibre brag grating technology
which does away with electronics altogether and uses fibre optics for measurement and
data transmission to surface.
• Distributed measurement of pressure and temperature. A recent development is
distributed temperature sensing (DTS). DTS employs a thin glass fibre optical cable
running along the entire length of the well. Using the effect that light sent through the
cable scatters with characteristics depending on the local temperature, it is possible to
obtain a very accurate (0.1 degree) temperature profile along the entire well. An effective
way of installation of DTS is through pumping it down through a U-tubed ¼ inch control
line that was run with the completion. The value of DTS measurements to interpret
reservoir inflow has yet to be confirmed. A next step in distributed sensing is likely to be
distributed pressure sensing (DPS).
• Flow rate and composition meters for down hole use are still very much in a development
stage. Given the difficulties to obtain accurate three-phase measurement at surface, the
down hole developments make take a while before they reach the stage of routine
application. An exception is the use of venturi meters, which can be used for liquid rate
determination in wells with inclinations up to 30 degrees. Other flow metering concepts
under development include fibre brag grating technology. Compositional meters under
development make use of gamma ray absorption, capacitance or conductance
measurements and electromagnetic helical resonators.
• Information on down hole flow rate and composition can, to a limited extent, be inferred
from down hole temperature and pressure measurements in combination with surface
measurements. Also, the use of down hole inflow control valves allows for the
determination of flow rates from individual well intervals through closing all intervals but
one and using surface measurements only (“well testing by exception”).

Reservoir imaging
In addition to direct or indirect down hole measurement of primary production variables
(pressure and flow rates), there are several developments to obtain reservoir information from
other sources during the producing life of a field. Most notably is the use of “4-dimensional”
(4D) seismic, also known as time lapse seismic, to achieve a picture of fluid front movements
in the reservoir through observation of the differences in seismic images over time.
Other developments, although much more in their infancy, are reservoir drainage imaging
with the aid of continuous resistivity measurements in a well bore or between well bores, or
through listening to micro-seismicity (cracking) around the well bore with down hole geo-
phones.
Yet another possibility is the use of down hole control valves to perform “on-line well tests”,
i.e. to infer information from the reservoir response to deliberately disturbed inflow into the
well bore. This is subject of a recently started PhD project in the section Petroleum
Engineering (See below).

Smart Wells, February 2001 4/11

Control
Down hole flow and pressure control can be achieved through the use of interval control
valves (ICVs). Figure 3 displays the basic concept: A well is completed with a perforated
casing and equipped with a tubing extending below the production packer (also referred to as
an extended stinger). The well is divided in intervals with the aid of packers between the
tubing and the casing, and each interval is equipped with a remotely controllable ICV. All the
major service companies can provide this functionality, and various levels of sophistication –
and costs - can be achieved. At the high end of the scale are electrically controlled
continuously variable ICVs with pressure and temperature measurements and valve position
feedback at each valve. The typical cost of such a valve is in the order of 0.5 million $.
Cheaper solutions employ valves that have a limited number of discrete valve opening
settings, or can just switch between open and closed (on/off valves). In addition to electrically
powered system, hydraulic systems are available.

Figure 3: Well with three perforated intervals completed with interval control valves (ICVs).
ICVs can also be applied to wells completed with a slotted liner or a sand screen instead of a
cemented casing, although this will usually lead to communication behind the casing between
the intervals. Furthermore, ICVs installed in the main well bore of a multi-lateral well can be
used to control inflow from branches. Obviously, the concept can also be used to control
outflow from injections wells, or even cross flow between different zones in a single well
bore.

Down hole processing

The biggest development effort in down hole processing is currently in down hole water
separation with the aid of cyclones. In combination with an inverted electric submersible
pump (ESP) this allows for down hole re-injection, with the potential for a dramatic reduction
in water production to surface. No full-field implementations have been performed to date,
but several pilot tests are ongoing around the world.
Another potential future application of down hole processing is down hole gas compression.

Communications and power supply

Further smart well hardware developments are in the field of power and data transmission.
Signals from down hole measurement devices to surface and vice versa are, at present, sent
mainly electrically or optically (via glass fibre). Recently, several experimental systems for
cable-less communications have emerged.
Power to down hole control equipment is currently provided electrically or hydraulically.
Experimental developments are in the area of batteries which can sustain down hole
temperatures, and down hole power generation (e.g. micro turbines driven by the production
flow).

Smart Wells, February 2001 5/11

Current applications

Water or gas shut-off

A first example of the use of smart well technology is depicted in Figure 4. A reservoir with
water drive and strong horizontal barriers is drained with a single well with perforated
intervals in each separate reservoir layer. Water breakthrough in the layers does not occur
simultaneously because of permeability differences. Using a completion with an on-off ICV
in each interval, well segments can be shut off when water breaks through, thus reducing the
amount of water to be processed at surface and preventing early lift-die out of the well.
Detection of the water could be done, in theory, by using the results from pressure and
temperature sensors at the ICVs. In practice, it will probably be sufficient to assess the effect
of closure of each ICV on the water production of the well at surface. A similar solution
could of course be used to shut off early gas influx. In terms of measurement and control, this
example relates to daily production optimisation, as well as to asset management.

Figure 4: Control of water break-through in a layered reservoir.

Commingled production
A second example is the use of ICVs to allow commingled production from zones with
different pressures, through choking the inflow from the highest pressured zone with a
continuously variable ICV, to avoid cross-flow to the lowered pressured zone; see Figure 5.
The alternative, conventional, scenario would be to sequentially produce the two zones,
through shifting of a sleeve on wire line or coiled tubing, or through work over and re-
perforation of the well. The major value of the smart well solution is in this case the
accelerated production, or, if production is restricted at surface, the maintaining of a constant
production plateau. Additional benefits are the absence of a work over, which is particularly
attractive for sub-sea wells, and the possibility to produce commingled in cases were Ronal
pressures are equal, but where government regulation require accounting of production from
different zones. In the latter case some means of flow measurement, either directly or
“inferred”, is of course necessary.

Smart Wells, February 2001 6/11

Figure 5: Commingled production from two stacked reservoirs.

Gas dump flooding

Figure 6 shows an example where a smart well is used to connect an oil reservoir with weak
gas cap drive to an underlying gas reservoir with a higher pressure. Pressure sensors and a
continuously variable ICV at the injection interval allow control of the “gas dump flood”. In
this example, a second well is used to drain the oil. Alternatively, the oil could be produced
trough the same well as used for the internal gas injection, using a concentric or parallel dual
completion solution.

Oil production
Gas re-injection
for pressurization

Figure 6: Pressure maintenance in an oil reservoir through controlled gas dump flooding.

Smart well research in the section Petroleum Engineering

Themes
As shown in the examples above, current applications are often extensions of conventional
completion techniques such as sequential production from different zones along the well
bore. However, smart wells may allow us to go from passive/reactive production scenarios to
active/proactive production control. This could e.g. be achieved through influencing the flow
behaviour in the reservoir by imposing a pressure profile along the well bore based on results
of down-hole measurements, and if necessary, continuously updated dynamic reservoir
models. Full development of this potential requires a more systematic analysis of reservoir

Smart Wells, February 2001 7/11

and well bore flow in terms of modern measurement and control theory. In parallel this
should be combined with a revision of conventional production scenarios, and the
development of computational tools to rapidly design and assess the value of smart well
solutions. The current research program into smart wells in the Section Petroleum
Engineering therefor focuses on the following three themes:
• Application of measurement and control concepts to reservoir engineering:
• Development of conceptual smart well solutions for improved reservoir drainage.
• Development of computational methods for smart well modelling.
We will discuss these themes in somewhat more detail below.

Application of measurement and control concepts to reservoir engineering

Figure 7 is a more detailed version of Figure 2, and represents the oil production process as a
model-based control problem. In the modelling phase a box has been added to emphasise the
role of identification, i.e. the activity of building and updating a model using measured data.
Some topics that are currently being investigated or are planned to be addressed in the near
future include:
• Development of system models of the behaviour of reservoirs, wells and surface facilities,
and in particular of a model-based control framework for reservoir engineering, including
the time-varying and non-linear aspects.
• Development of system identification techniques to estimate reservoir model parameters
from down-hole measurements. This involves low-order models focused on direct
identification and control, as a well as large scale reservoir models.
• Development of model reduction techniques and control algorithms for the systematic
optimisation of production parameters.
• Use of down hole measurement and control equipment to perform frequent down hole
well tests.
• Development of techniques to systematically take into account the effects of uncertainties
in the entire cycle of system response, measurement, modelling and control.

Input System State variables

(reservoir & wells)

CONTROL MEASURE

Control
Sensors
algorithms

Parameter
System model Measured
identifier
output

MODEL
Figure 7: Model based control.

Smart Wells, February 2001 8/11

Conceptual smart well solutions for improved reservoir drainage
Figure 8 displays the results of a first attempt to improve ultimate recovery in heterogeneous
reservoirs through the use of smart wells. This particular example simulates the drainage of a
reservoir that incorporates a high permeable streak, using a pair of horizontal production and
water injection wells with ICVs. An intuitive optimisation algorithm was used to control the
ICV settings to optimise recovery [1], [2]. This study forms the basis for a more formal
optimisation approach currently under development.

Injector Producer
Figure 8: Improved water flooding of a reservoir with a highly permeable streak. Left: top
view of the reservoir with a pair of parallel horizontal injection and production wells.
Middle: Displacement pattern at the moment of water-breakthrough using conventional
wells; black = water; white = oil. Right: Displacement pattern at the moment of water-
breakthrough using smart wells, revealing a much higher recovery.
Another area of interest is the use of smart wells to counteract the effect of pressure drop in
horizontal wells. Figure 9 (top left) illustrates the occurrence of a very uneven inflow along
the axis of a horizontal well, caused by frictional pressure drop in the well bore resulting in a
higher draw down at the heel than at the toe. This typically occurs for large-diameter, high-
rate wells producing from highly permeable reservoirs. As a result the well is prone to early
water or gas breakthrough at the heel, and the effectiveness of the well near the toe is strongly
reduced. Figure 9 (top right and bottom left) display two conceptual solutions to counteract
this effect. The first one, the smart stinger completion (SSC), employs an extended stinger
with one continuously variable ICV at the heel to flatten the draw down profile, and thus the
inflow profile [3]. The second method, the inflow switching process (ISP), uses a number of
on/off ICVs to regularly move the point of highest draw down along the well bore. Once
water or gas has broken through, that particular interval is shut off and the water or gas cone
is allowed to recede before re-opening of the interval [4]. Figure 9 (bottom right) illustrates
that both methods result in an oil production behaviour almost identical to that in case of an
ideal well without pressure drop.

Computational methods for smart well modelling

Usually, the grid block size in reservoir simulators is far to big to accurately represent the
detailed near-well bore flow. This is often accounted for through the use of a semi-analytical
correction method, often referred to as the Peaceman model. For advanced, multilateral, wells
sometimes a more accurate description is required, and recently research in the section
Petroleum Engineering addressed some aspects of improved inflow modelling [5], [6]. Other
work in the area of computational modelling involved the well bore flow in extended stinger
completions such as the SSC discussed above [7].

Smart Wells, February 2001 9/11

 Inflow / unit length Inflow / unit length

well axis well axis

Variable choke

2.00E+04

Inflow / unit length

1.50E+04 Basecase
t=1 t=2 ISP
Frictionless

Oil rate [sbbl/day]

SSC
1.00E+04

Kv /Kh = 0.1 and Kh = 2000m D

well axis
5.00E+03

0.00E+00
0 1 2 3 4 5 6 7
Time [years]

Figure 9: Smart well solutions to combat frictional pressure drop in horizontal well bores.
Top left: a conventional horizontal well, displaying decreasing inflow from heel to toe
because of pressure drop along the well bore. Top right: the smart stinger completion (SSC).
Bottom left: the inflow switching process (ISP). Bottom right: Cumulative oil production as a
function of time. The solitary line represent production for a conventional completion. The
three lines close together represent production using the SSC, the ISP and production from a
conventional well without pressure drop (i.e. the ideal situation).

Conclusion
Although the expression smart wells may be likely to disappear as suddenly as it came into
fashion, the concept of using measurement and control to optimise oil and gas production is
here to stay. Hardware is developing fast and in many directions, and in particular optical
techniques and cable-less communication are likely to lead to a dramatic increase in down
hole measuring capabilities over the coming years. Lagging behind the hardware
developments are the capabilities to use the equipment for creation of value. This is to a large
extent a matter of asset management, and the major steps to take are the development of
improved concepts for “smart” reservoir management, the handling of large amounts of data,
and increased integration between disciplines. Research in the section Petroleum Engineering
should focus on concepts, and not on hardware; not on what is possible with smart well
technology today, but on what will be possible whenever the hardware becomes available.

Smart Wells, February 2001 10/11

Acknowledgements
Research into smart wells in the section Petroleum Engineering of the Department of Applied
Earth Sciences of the Delft University of Technology (DUT) is supported by Shell
International Exploration and Production (SIEP). Shell sponsors two PhD positions, and
furthermore, the author of this paper is jointly employed (50%-50%) by DUT and SIEP. The
author acknowledges permission from Shell to publish the material and to use Figures 1 and 4
to 6, which were taken from internal Shell publications.

Glossary
DTS Distributed Temperature Sensing
DUT Delft University of Technology
ESP Electric Submersible Pump
E&P Exploration and Production
FDP Field Development Plan(ning)
ICV Inflow Control Valve or Interval Control Valve
ISP Inflow Switching Process
NPV Net Present Value
OPEX OPerating Expenditure
SIEP Shell International Exploration and Production
SSC Smart Stinger Completion
4D 4-dimensional (3 in space, 1 in time)

References
[1] Brouwer, D.R., Jansen, J.D, Van der Starre, S., Berentsen, C.W.J. and Van Kruijsdijk,
C.P.J.W.: Recovery Increase Through Waterflooding Using Smart Well Technology,
Proc. SPE European Formation Damage Conference, The Hague, 21–22 May 2001,
paper SPE 68979.
[2] Van der Starre, S.: Recovery Increase Through Waterflooding Using Smart Well
Technology, MSc Thesis, Report CTG/PW/00-017, Centre for Technical Geoscience,
Delft University of Technology, October 2000.
[3] Wagenvoort, A.W.: Application of Smart Well Technology to Combat the Effects of
Frictional Pressure Drop in Horizontal Wells - The Smart Stinger Completion, MSc
Thesis, Report CTG/PW/00-019, Centre for Technical Geoscience, Delft University of
Technology, December 2000.
[4] Droppert, V.S.: Application of Smart Well Technology to Combat the Effects of
Frictional Pressure Drop in Horizontal Wells - The Interval Switching Process, MSc
Thesis, Report CTG/PW/00-020, Centre for Technical Geoscience, Delft University of
Technology, December 2000.
[5] De Koning, M.B.F.: Application of an Advanced Well Model in a Reservoir Simulator,
MSc Thesis, Report CTG/PW/99-015, Centre for Technical Geoscience, Delft
University of Technology, September 1999.
[6] Jansen, J.D.: Expressions for the flow through grid block boundaries near wells in
reservoir models with irregular grids, Report CTG/PW/00-002, Centre for Technical
Geoscience, Delft University of Technology, September 2000.
[7] Jansen, J.D.: Numerical modeling of the flow in extended stinger completions, Report
CTG/PW/00-003, Centre for Technical Geoscience, Delft University of Technology,
December 2000.

Smart Wells, February 2001 11/11