SPE

International

Society of Petroleum Engineers

SPE 98818

Intelligent Sand Management

M. Babs Oyeneyin, SPE; Chris Macleod; Gbenga Oluyemi, SPE; The Robert Gordon University
and Anthony Onukwu, SPE; Vetco-Gray Limited

Copyright 2005, Society of Petroleum Engineers Inc.

In this paper attempts have been made to address the
This paper was prepared for presentation at the 29th Annual SPE International key issues with sand management. The paper goes
Technical Conference and Exhibition in Abuja, Nigeria, August 1 -3, 2005.
further to provide a case for lifecycle intelligent sand
This paper was selected for presentation by an SPE Program Committee following
review of information contained in an abstract submitted by the author(s). Contents management including the current technological
of the paper, as presented, have not been reviewed by the Society of Petroleum
Engineers and are subject to correction by the author(s). The material, as
innovations and future challenges in terms of
presented, does not necessarily reflect any position of the Society of Petroleum technology gaps
Engineers, its officers, or members. Papers presented at SPE meetings are subject
to publication review by Editorial Committees of the Society of Petroleum Engineers.
Electronic reproduction, distribution, or storage of any part of this paper for
commercial purposes without the written consent of the Society of Petroleum
Highlights of some of the development activities
Engineers is prohibited. Permission to reproduce in print is restricted to an abstract ongoing at The Robert Gordon University are also
of not more than 300 words; illustrations may not be copied. The abstract must
contain conspicuous acknowledgement of where and by whom the paper was presented
presented. Write Librarian, SPE, P.O. Box 833836, Richardson, TX 75083-3836,
U.S.A., fax 01-972-952-9435.
The paper ends with an introduction of the Sand
ABSTRACT Management Network – A Network of upstream oil and
gas operating companies and service providers with
More than 70% of oil/gas reservoirs worldwide are in interest in sand management. The Network purpose is
various degrees of unconsolidation requiring some to share successful methods and technologies as well as
elements of sand management. This is more so promote best practices in the management of sand.
especially in the Niger Delta as activities extend to the
more challenging deepwater environments of offshore
West Africa.
INTRODUCTION
Debate on the most effective sand management strategy For many of the clastic reservoir sands around the
has been intense for many years, with different world sand production is a major issue and the
companies adopting different strategies. successful development of these assets requires having
in place an effective sand management strategy. Where
Development of appropriate strategy for the sand production is considered a risk the conventional
management of reservoirs with sanding problems is approach is usually to initiate downhole control to
rather complex and requires an integrated approach to exclude sand production using one of a number of
finding the optimum solution to solving the problem. traditional techniques such as gravelpacks, standalone
This requires integration of key aspects of reservoir screens many of which have been known to fail
characterisation, drilling, completion and production resulting in productivity impairment over time. This
technologies including sand tolerances (Seabed concern has led to new concept of total sand
wellhead/flowlines, topside facilities) . Mitigation of management in terms of whole development strategy
sanding requires reliable sanding prediction, precise covering reservoir response to failure, completion,
well design, accurate technology selection as well as monitoring and topside management. This total
optimum completion strategy. These are transient management strategy requires an integrated approach
phenomena requiring a life cycle real time that can only result in significant benefits of increased
management. productivity and reduced capital and operational costs
via reduced remedial action. The challenges here are
in how to:
2 M.B.Oyeneyin; Chris Macleod; Gbenga Oluyemi and A. Onukwu SPE98818

To ensure that reservoirs and wells with sanding

• maximise the well and development problems offer maximum deliverability at minimum
performance cost requires an integrated approach to sand
• minimise well cost management. The conventional process of sand
• maintain well and asset integrity management [Fig 1] includes:

These would require: • Sanding risk assessment through sand failure

and production prediction. This involves if
• Knowing if and when a clastic reservoir will and when a reservoir sand will fail.
produce sand • Selection of fit-for-purpose sand control
• Knowing what volume of sand is likely to be option. This can be either passive or active
produced through monitoring sand control. There are options of surface or
• Applying the appropriate fit-for-purpose downhole control
downhole sand control where and when • Downhole sand grain and produced volume
necessary monitoring
• Topside Management.
Applying appropriate surface control and managing the
sand production on surface where feasible Many of these strategic steps appear to ignore fines
Optimising the design and installation of the sand management in the overall equation.
control completion system and topside facility.

The eventual goal is to ensure that the wells offer A Case for Life Cycle Intelligent Sand Management
maximum deliverability at minimum life time cost in Every step of sand management is a dynamic process
order to maximise individual well NPV. that requires real-time monitoring and implementation
from ‘’cradle-to-grave’. The key areas include:
Sand failure, migration and eventual production are
transient phenomena. Therefore any successful sand • Fines Migration Mechanisms in Darcy Sands
management strategy will require a life cycle a ‘’cradle- • Pore Blocking Mechanisms
to-grave’’ approach for better results. • Grain Size Profiling with LWD/PLT System
• Real Time Tracking of solids
Mitigation of sanding requires reliable sanding mobilisation/migration
prediction, precise well design, accurate technology
• Screen Systems Performance for HW
selection as well as optimum completion strategy.
Applications
These are transient phenomena requiring a life cycle
• Coupled Reservoir Simulator-CFD Analysis
real time management. This is the strategy that has been
adopted by the Well Engineering Research Group at • Real-Time Prediction of Sanding Potential
The Robert Gordon University . The specific focus of • Intelligent Sand Control Completions
this strategy, the subject of this paper includes: • Development of a new VSSTM Intelligent
Screen System
• Real-Time Grain size profiling while drilling
or during production and application to This paper however is focussed on real-time sanding
sanding potential prediction and sand volume prediction as well as intelligent sand
• Prediction of corresponding sand volume control completions.
migrating/produced .
• Real-Time tracking of particle migration for
SANDING POTENTIAL PREDICTION
impairment profile analysis
• Development of appropriate intelligent
downhole control system with capability to .The first stage of a sand management strategy in the
respond to changing sand production and well development of any reservoir with sanding problems is
operating conditions. to determine whether there is any risk of sand
production. The focus is on load bearing sands while
fines are usually ignored. The general presumption is
INTEGRATED SAND MANAGEMENT that fines can readily be produced without due
PHILOSOPHY consideration of the after effects.

The sanding prediction normally involves:

3 Intelligent Sand Management SPE 98818

management can also be made much easier. Real time

• Developing formation strength profiles based sand prediction is also bound to offer a great potential
on log data obtained from pilot holes for optimized drilling and production, providing safety
• Calibrating the log derived strength data and efficiency. The overall impact of all these are
against laboratory test data on core materials reduced life-cycle costs and improved field economics.
obtained from the pilot wells
• Evaluating the calibrated formation strengths Theoretical Background
against the forecasted stresses acting on the
formation in order to determine the probability There are a plethora of models, which relate mean grain
of sand failure size with other textural, and petrophysical parameters
• Reviewing well test data for indications of like porosity, permeability, irreducible water saturation,
sand production. grain packing, grain shape etc.[1, 2, 3, 4, 5, 6, 7, 8, 9] .
Similarly models abound which relate these textural
The results generally offer some useful guide for and petrophysical parameters with UCS and other
assessing whether downhole sand control would be geomechanical properties of rock such as internal
required as part of initial completion during production friction angle, Poisson ratio, bulk compressibility[10, 11]
start-up. etc, which are indicative of sanding potential in
reservoir rocks. This cross relationship between grain
Need for Real Time Sanding Prediction size and geomechanical properties is the methodology
that has been adopted for real time sanding potential
Sanding in well/reservoir is a life-cycle transient prediction and the corresponding produced sand
‘’cradle-to-grave’’ phenomena which changes with volume.
time due to constraints imposed by operational factors
and reservoir diagenesis right from exploration, This technique builds on others already commonly
drilling, completion production and workover phases of available in the industry. Formation strength logs from
development through to subsequent abandonment. The pilot holes are now routinely produced from pilot holes
sanding phenomena are strongly influenced by the based on core data and DSI sonic data. From these two
reservoir stress path, overburden/reservoir pressure, the formation failure modes can be analysed:
drawdown/depletion profile, type of reservoir fluid, Shear Compressive failure which is a function of
completion and production strategy, drive mechanism, massive, continuous sand production. It is dependent
and especially onset of water production. Thus the on the formation strength and the in-situ effective
failure envelope and failure mode would continue to strength
change especially with the presence of fines which can Tensile failure which is the transient failure mode and
readily migrate. . Predictions made prior to drilling a is the result of excessive tensile stresses applied to the
well cannot be said to be valid during or after drilling sand face due to high drawdown. This is defined by the
and during production operations. It is therefore UCS[Uniaxial Compressive Strength]
important to have sanding prediction capability to
accurately quantify the sanding potential at any time Thus any real-time strength evaluation an strength log
during the lifecycle of a development. This is only generation can be made using sonic, density, porosity
possible through integration of initial prediction prior to and clay content.
well construction, and real time monitoring/evaluation
whilst drilling and producing to update the prediction Likewise Formation strength logs can be obtained from
earlier made in the pre-drilling stage of field typical PLT systems for real-time evaluation. This
development. approach has been used successfully in the past to
identify perforation intervals based on shear and tensile
Why real time? failures[12].

With real-time prediction of sanding potential, it will be

possible to track the changes in failure envelope of
producing reservoir sand and also predict the Use of Neural Network
corresponding sand volume
Timely economically wise decision regarding sand Grain size profiling is conventionally done by
control and type of sand control to be adopted, will performing sieve analysis on core sample. Grain size
become easy for any reservoir management teams. distribution parameters are then estimated using various
Decisions regarding completion type and design, screen formulae and relations. Alternatively linear regression
deployment, gravel pack design and topside analysis and functional formulae, which relate
4 M.B.Oyeneyin; Chris Macleod; Gbenga Oluyemi and A. Onukwu SPE98818

petrophysical parameters to grain parameters, have also Tensile failure which is the transient failure mode and
been used. These techniques use simplified equations to is the resultr of excessive tensile stresses applied to the
express complex correlative relationships between sand face due to high drawdown. This is defined by the
these parameters [13]. Neural networks have the UCS[Uniaxial Compressive Strength]
potential of accurately describing the behaviour of any
complex systems once appropriate input data, topology Thus any real-time strength evaluation an strength log
and network parameters are used for the neural generation can be made using sonic, density, porosity a
implementation of analysing such systems and have and clay content.
been successfully applied
Both neural networks coded in C++ and multivariate Likewise Formation strength logs can be obtained from
techniques have been used to model grain size typical PLT systems for real-time evaluation. This
distribution and their performances compared. Results approach has been used successfully in the past to
show that neural networks perform better than the identify perforation intervals based on shear and tensile
multivariate model. failures[11].

Equations for obtaining petrophysical parameters

from logs Use of Neural network

Real time sanding potential prediction requires that Grain size profiling is conventionally done by
relevant information and data are available at all times. performing sieve analysis on core sample. Grain size
Information on petrophysical data which serve as input distribution parameters are then estimated using various
to the network can be obtained directly from various formulae and relations. Alternatively linear regression
sources. Examples of such sources, as earlier indicated, analysis and functional formulae, which relate
include LWD, wireline logs, NMR, petrophysical parameters to grain parameters, have also
RPMWD[Reservoir Pressure Measurement While been used. These techniques use simplified equations to
Drilling], and MDT. Empirical and semi empirical express complex correlative relationships between
models can also be used to calculate the values of these these parameters [12]. Neural networks have the potential
parameters from the logs. A typical model being used of accurately describing the behaviour of any complex
is: systems once appropriate input data, topology and
network parameters are used for the neural
implementation of analysing such systems and have

2
d50 =
C *φ (
e
5 − 3m )
[(1 − Swirr)/(Swirr)]*m 2 been successfully applied
Both neural networks coded in C++ and multivariate
1
k 2 *S2 techniques have been used to model grain size
distribution and their performances compared. Results
(1)
show that neural networks perform better than the
where:
multivariate model.
d50 = 50 Percentile size of the formation grain
k = permeability
Analysis of Findings
φ e = effective porosity
S = specific surface area of grains Figures 2-4 show correlation of various petrophysical
Swirr = irreducible water saturation and textural parameters. Figures 5-7 show the
m = cementation factor correlation of petrophysical property of porosity with
C is an empirical constant which depends on the various geomechanical parameters. All these figures
hydraulic unit and stress path. indicate some sort of correlation between these
parameters but the nature of the relationship is not
This technique builds on others already commonly clearly understood.
available in the industry. Formation strength logs from
pilot holes are now routinely produced from pilot holes Figure 8 is a plot of neural network error versus epoch
based on core data and DSI sonic data. From these two during the training of optimised network used for the
formation failure modes can be analysed: grain size distribution modelling.
Shear Compressive failure which is a function of
massive, continuous sand production. It is dependent Figure 9 compares the performance of neural network
on the formation strength and the in-situ effective model with that of multivariate model
strength
5 Intelligent Sand Management SPE 98818

Figure 10 is a continuous plot of grain size along the To estimate the actual sand volume during drilling or
axial and horizontal direction. A relationship can be production, a relationship between sand grain volume
developed between grain size profiles and UCS based and porosity was established.
on the correlation in Figure 5. This relationship can
then be utilised to construct a continuous strength This concept was based on the simple principle that
profile similar to the one shown in Figure 11, for the sand detachment from rock formation causes a decrease
interval of interest. The development of more robust in sand grain volume and thus an increase in porosity
semi-empirical model to define the complex i.e. porosity at time ti will increase at time tn, if sanding
relationship has been initiated as part of on-going work starts to occur. However, this does not account for the
at RGU[13] effect of sand failure with regards to sand pack
structure.

REAL TIME SAND VOLUME EVALUATION To get this relationship to work in real time, the use of
various well logs/production logs were adopted. The
Unconsolidated reservoirs will definitely produce sand, advantage of the use of well logs and production logs is
but the main issue is actually what production is availability of dynamic information throughout the
considered unbearable, this is the sand tolerable level of well, its frequency and continuity in real time.
the reservoir. This is where sand volume measurement
becomes an utmost necessity in intelligent sand Well logs/production log selected for this purpose are:
management.
• Sonic log using (BHC) Bottom Hole
Inaccuracy in measurement of sand produced from Compensated logging tool
formations, from available sand • Density log using (FDC) Formation Density
monitoring/measurement systems have threatened the Compensated logging tool
efficiency of existing sand management strategy. This • Neutron log using (CNL) Compensated
is due to lack of accuracy in existing topside and Neutron Logging tool
downhole sand monitoring systems. These systems
have introduced errors in terms of volume differences Sonic measurement during production was obtained
between actual and surface or subsea sand production from production logging tool
volume measurements. There is a gap left in
measurement i.e. this is what is left in the perforations Computation of real time actual sand volume
or wellbore that did not get to the surface and what is measurement
measured at the surface with existing sand The premise here is currently based on the basic
monitoring/measurement tools. This is one technology relationships between porosity , pore size and pore
gap that intelligent sand management looks forward to volume as well as bulk volume here given as
proffering solutions to.
Vp
φ= (2)
However, with an intelligent approach, actual sand Vb
volume from sand prone or producing formations could
Where Vp is pore volume and Vb is bulk volume of
be measured accurately. This includes real time sand
rock formation and φ is porosity
monitoring and measurements that will help in making
quick decisions on sand production issues. Few
From Blake-Kozeny Correlation:
examples of benefits of measuring sand volume
accurately and in real time are effective topside
facilities sand management that mitigates against d *φ
D pore= 50 (3)
potential safety and environmental risks resulting from 3(1 − φ )
production tubing or flowline erosion and failures,
perforation/screen blockages, and sand
handling/disposal issues, other benefits are effective
Porosity from validated log response provided a
wellbore sand cleanout process, optimization of
continuous representation of pore volume as a function
production in sand prone/producing formations.
of depth in a well for a given hydraulic unit.
Measurement of sand volume in real time
Establishing a relationship between φ and Vg, equation
2 was re-arranged as
Vg = Vb (1- φ ) (4)
6 M.B.Oyeneyin; Chris Macleod; Gbenga Oluyemi and A. Onukwu SPE98818

• What are the sources and effects of Fines

Equations 3 and 4 in simple terms indicate that a • What are the Fines Management Issues with
change in grain size profile within an unconsolidated respect to
reservoir lithology has effect on the porosity, pore o Mitigation
volume and corresponding instu volume of sand grain. o Reservoir Control?
o Wellbore Control?
From the equation (3) above, volume of sand grain was o Topside Control?
computed from change with porosity with time in sand • What are the strategies for Controlling Fines
prone/producing formation. Mobilisation/Migration?

However, the variable in the equation 3 are Vg and φ , What are Fines?
Vb was assumed to be constant and was estimated from
log data. Fines can be defined as solids of the order of less than
10 m p to about 100 m Solids[14] . They can be
Equation 3 & 4 can be modelled as present as Drilling/Completion Fluid Solids, Filter
Cake Debris or Injection Water Solids. From the
Qs = U f ( φ ) (5) reservoir perspective fines can be load-bearing sands as
depicted in PSD profile of some Adriatic sands[ Fig.
11 ] or ‘’insitu’’ fines within a reservoir sand body.
Where Qs represented actual sand volume, U is
constant and is the bulk volume of sand
Sources and effects of fines
prone/producing formation volume, φ is porosity value
at specific time in sand prone/producing formation. The sources of fines can be

The equation 3 above can be represented in this form ‘’External’ fines emanating from:
Vgn = Vb (1- φ n) (6)
n here represents changes in the variable at time n. • Solids in injection water
• Drilled solids
Thus with this concept, the produced volume and grain • Low gravity solids in drilling/completion
size distribution f produced sand can be computed [ fluids
Figure 12 ] • Filter cake debris
• Wellbore debris

SAND CONTROL AND TOPSIDE MANAGEMENT Internal fines

These are mobilised fines within the reservoir which
Following sanding risk assessment the subsequent sand can be free or associated fines. Examples include
management steps include: silica, feldspar, mica, calcite, dolomite, siderite,
chlorite.
(i) Downhole sand control.- Active or Passive Sand
Control Completion Options
Effects of Fines/Sand Production
(ii) Downhole Sand Monitoring/Management
(i) Reservoir Effect
(iii) Surface or Topside Management • Fines are very unforgiving to Darcy Sands
• Productivity Impairment – This can be
A key part of the outlined management strategies is the aggravated by multiphase production
effective management of fines the details of which are especially with increased of water
highlighted below: breakthrough evidenced by increase in BS&W

Challenges of Fines Management (ii) Sand Control Completions especially

Gravelpacks and Fracpacks
Some of the most fundamental questions with respect to
fines management are : • Very High Impairment
• Screen Systems evidenced by plugging and
• What Are Fines? hot spots
7 Intelligent Sand Management SPE 98818

• Completion String/Flow line • Oriented

• Line erosion Perforating/Fracpacking/Consolidation
• Line Blocking(Critical with Subsea tiebacks) The success of these depends on a good
• Topside System Blocking knowledge of the stress path
• Water Shut-Off Strategies including
The key impairment drivers[15] are: depressurisation of Aquifer or reservoir
• Grain Size/Size Distribution chemical treatment
• Host Media Pore Size Distribution • Limiting Drawdown
• Concentration – NTU level appears to be the • Use of ESS or ESL with Effective Compliance
most important. This can also be aggravated level to create a near wellbore compaction that
by higher electrostatic attractive forces can promote ‘’Secondary Filtration’’ of
between fine particulates in viscous environ fines/migrating load bearing sands?
and can be reduced by the use of Dispersants
• Shape . Active Sand Control
• Production Rate Choices can include
• Type of Fluid
• Type of Completion • Downhole Control
• Pore Blocking Mechanism • Oriented Perforating + Frac & Pack with
o Internal Pore Filling resin-coated Proppants
o Shallow Internal Bridging • Fracpacking with medium size synthetic
o External Cake proppants behind perforations plus
• Need for Secondary Filtration. Shallow conventional annular gravelpack will control
Internal bridging appears to be better[ 16] fines.
• Gravelpacking
Fig. 12 shows the effect of completion(open hole • Standalone screens[ Slotted liners, Wire-
versus perforated completions subsequently enhanced wraps, Premium screens, Expandable screens
by frac & pack using synthetic proppants. As can be and Expandable slotted liners]
observed, open hole completion, as expected has a • Flow Control
higher flow efficiency than perforated completions. • Reservoir chemical treatment via
However the placement of synthetic proppants behind screenless completions
perforations via frac & pack operations not only • Surface Control with high rate production and
improves the near wellbore flow efficiency but also the subsequent topside separation of solids. .
bridging effectiveness of the synthetic proppant in the Such surface control techniques include:
control of fines migration. • Topside Management using Subsea Well Head
separation systems such as Hydrocyclones
Benefits of the proppants include : • Topside Separation with the use of Topside
Strainers
• Highly Spherical
• Smaller Pore Size /Pore Size Distribution To ensure quality success there is the need for
• Better Filtration Capability laboratory analysis that include:
• Better Bridging Efficiency
• Higher Permeability • Effective grain size analysis - Important to
include fines in the overall size distribution,
Control Methods shape and angularity
• Disc Coupon
These can be : • Disc Coupon + Proppant back Flow Test
• Core Flood return perm analysis which is
Passive Control involving but not limited to : usually limited to liquid flow tests for safety
reasons
• Horizontal Wells Completion as long as it is • Sampling
targeted away from OWC and the hydraulics • Pore Size Distribution Analysis [16-17]
optimisation is right in relation to the IPR- • Pore Blocking Mechanism [16-17]
wellbore performance relationship as well as • Need for Coupled Reservoir Simulator-CFD
optimum length of the lateral section. risk assessment for fines/solids migration [14] .
8 M.B.Oyeneyin; Chris Macleod; Gbenga Oluyemi and A. Onukwu SPE98818

The selection of a fit-for-purpose control method

depends on the gathering and analysis of key reservoir The benefits of the VSSTM include:
and well data as well as rigorous laboratory analysis (i) Application to any type of reservoir sands different
which do add to well costs. The process is elaborate grain size distribution
the success of which depends on reservoir quality index (ii) Extended screen life,
and operating conditions. (iii) Adjustable screen performance over the life of the
This is why different companies have adopted different well as the well depletes
strategies with mixed results. The key decision drivers (iv) Routine information collection on reservoir
adopted are: performance and inflow performance along the length
of the sand face.
• Economic (v) Flow control
• Rate/Recovery/Facilities
• Completion As a system the technology has the potential to extend
• Subsea vs. Platform and improve field life in mature and marginal fields.
• HSE & Regulatory Restrictions Even though the unit development cost for such wells
• Results of Risk Assessment may be slightly higher, the overall well/production cost
per barrel will be much lower as a result of higher rates,
higher recovery factor, minimum workover, etc.
A CASE FOR INTELLIGENT SCREEN SYSTEM
As sub-systems the technology could be readily
All current sand screens have fixed slot sizes, are prone adapted for use in other aspects of intelligent wells that
to the same failure mode of plugging by fines and the might include intelligent control of inflow-control
development of “hotspots” along the screen leading to devices particularly in multi-zone or multilateral wells.
screen erosion and localised sand production. Localised Likewise it could be deployed as a topside strainer unit
screen failure is a major and unpredictable problem. for intelligent topside management.
To overcome these limitations a new intelligent screen
system known as the VSSTM is being developed by the
Sand Management Research Group at the Robert THE SAND MANAGEMENT NETWORK
Gordon University.
Until recently sand management was seen as an art
The VSS is the product name for a new screen system which has continued to develop technologically but
being developed for intelligent sand with different companies continuing to adopt different
management/control. The VSS is designed to react to strategies in the development of reservoirs/fields with
changing well conditions so that inflow performance, sanding problems. As sanding issues are beginning to
screen longevity and performance are optimised. The dominate the development of not only matures fields of
VSS also extends the principles of “intelligent well the North Sea but also the deepwater environments of
design to include control of the sand face. Existing the Atlantic Margin, it has become essential to create a
intelligent well technology is focused on inflow knowledge-sharing technology forum as a way of
through a single point of control. reducing potential risks and create a win-win situation.
This is the rational behind the initiation of the joint
The Intelligent Variable Slot Screen System (VSSTM) industry Sand Management Network with special focus
has been conceived as a way to remotely adjust the on the North Sea.
screen slot size based on distributed sensing of, for
instance, temperature, solids loading and pressure The network is an expanding network of upstream oil
drop, and avoid plugging of the screen, localised and as related companies and technology providers that
erosion of the screen and localised sand production. have come together to tackle sand management related
issues.
A Proof of Concept award funded by Scottish
Enterprise is currently in place and has been used to The joint industry network initiated in 2002 by The
fund the development of a demonstrator version of the Robert Gordon University has been specifically
VSSTM in order to prove the concept. The prototype established to:
version of the VSSTM is now built ready for initial
functionality testing. Initial market survey as part of • Exchange ideas on sand management
the funded project objective has shown evidence of technology use
interest among operators and service providers with • Communicate to its membership and
interests in sand management and filtration. management strategies and technology gaps
9 Intelligent Sand Management SPE 98818

and developments, required R&D for field 2 Coates, G. R. & Denoo, S., 1981, “The Producibility
development and operation in order to reduce Answer Product,” The Technical Review,
North Sea costs per barrel by either improving Schlumberger, Houston (June) 29 (2), 55-63
recovery or reducing well/operating costs. 3 Kozeny, J. 1927, “Uber Kapillare Leitung Des
• To develop a high level sand management Wassers Im Bodem (Fustier Ver Sikeung Und
selection process using best –in-class Anwendung Auf Die Bemasserung).” Sitzungsber
techniques, procedures and equipment Akad., Wiss, Wein, Math-Naturwiss, KL, Vol 136
(IIa) 1927.
The network which consists of eight operators, eight 4 Krumbien, W. C. and Monk, C. D. : “Permeability as
service providers , the UK Department of Trade and a Function of the Size Parameters of
Industry[DTI] the RGU and two other universities., Unconsolidated Sand.” Trans. AIME, Vol. 151, pp
conducts its knowledge-sharing activities through 153-162.; 142
dedicated technology Forums covering key sand 5 Revil, A & Glover, P.W.J., 1997, “Theory of Ionic
management issues and practices. Surface Electrical Conduction in Porous Media,
A dedicated SMN website www.sandmanagement.com Phys. Rev.B, 55 (3), 1757-1773
provides a unique avenue for continuous discussions, 6. Slichter, S. C. : “Theoretical Investigation of the
permanent records of activities and is open to all Motion of Groundwater.” US Geol. Surv., NTIS,
members of the network. Springfield, VA, 19th Annual Report, 1989, pp
295-384.
Similar Networks would be desirable for all other oil 7 Willie, M. R. and Rose, W. D. : “Some Theoretical
producing regions of the world as the benefits of Considerations Related to the Quantitative
commonality of purpose are shared by operators and Evaluation of the Physical Characteristics of
service/technology providers accordingly through Reservoir Rock from Electric Log Data.” Tran.
network activities. AIME, Vol. 189, 1950, pp 195-118.
8 Carman, P. C.: “Fluid Flow Through Granular Beds”
Trans. Inst. Chem. Eng., Vol.15, pp150-166, 1937
CONCLUSIONS 9 Han, G. and Dusseault, M. B. : “Quantitative
Analysis of Mechanism for Water-Related Sand
This paper has attempted to present a holistic view of Production”. SPE Paper No. 73737, 202
integrated sand management from reservoir through to 10 Tokle, K. et al. : “Predicting Uniaxial Compressive
sand control completions up to the topside facilities Strength from Log Parameters.” SPE Paper No.
required for optimum performance. It goes further to 15645, 1986
make a case for Life cycle intelligent sand management 11. Beesley, M.; Keith Munday and Craig Paveley:
based on the premise that every aspect of sand ‘’An integrated approach to sand management’’,
management is a dynamic, ‘’cradle-to-grave’’ Proceedings to the 1st European Sand Management
phenomenon. Highlights of the ongoing work at RGU Forum, Aberdeen, March, 2004.
have been presented with focus on real-time sanding 12 Zhou, C and Wu, X : “Neural Network Formation
prediction and the development of a new VSS system. Parameters Estimation From Well Logs in
Quantitative Log Analysis: A Comparative Study.
Knowledge sharing among operators and service SPE Paper No. 25359, 1993
providers with interests in sand management is 13. Oluyemi, Gbenga: ‘’Grain Size profiling using
inevitable. This is evidenced by the deliverables from Neural Network and application to real-time
joint industry programmes of the Sand Management sanding potential prediction’’, On-going PhD
Network initiated for the North Sea. Creation of Research programme at RGU.
similar Technology Networks especially for the new 14. Said S. Mufarji; M.B.Oyeneyin; and Shah Jihan:
development areas such as offshore West Africa where ‘’Optimising productivity from fracture damaged
sanding is a major development problem will go a long reservoirs’’, SPE Paper No. 91971, 2004.
way in evolving a ‘best-in-class’’ sand management 15. Oyeneyin, M. B: ‘’Fines Management’’, Technical
strategy for each oil producing region. Presentation to the Sand Management Network,
Jan. 2005
REFERENCES 16. Oyeneyin, M.B; .J.M. Peden; Ali Hosseini; and G.
Ren: ‘’Factors to consider in the effective
1 Berg, R. R., 1970,: “Method for Determining management/control of fines migration in high
Permeability from Reservoir Rock Properties” permeability sands’’, SPE Paper No. 30112, 1995.
Trans. Gulf Coast Ass. Geol. Soc., Vol. XX, 17. Oyeneyin, M.B; .J.M. Peden; G. Ren; Y. Bigno and
pp303-317 Ali Hosseini: ‘A new gravelsizing package for
 10 M.B.Oyeneyin; Chris Macleod; Gbenga Oluyemi and A. Onukwu SPE98818

effective sand control design and analysis’’, SPE

Paper No. 26219, 1993. 60
50
40

Packing coefficient
Risk Assessment 30
For Sanding Potential Prediction
20
10
0
0.00 0.10 0.20 0.30 0.40 0.50 0.60

Control Topside Mean grain size (mm)

Monitoring
Management
Fig. 4 : Correlation between packing coefficient
and grain size

Active Passive Sand

Transport

200
Remedi Sub-Surface Surface Facilities:
Separatio 150
n
100
Fig. 1: Sand Management Strategy 50

0
0 20 40 60 80 100
70 Porosity (%)
60
Porosity (%)

50
40 Fig. 5 : Correlation between UCS and porosity
30
20
10
0
0.00 0.10 0.20 0.30 0.40 0.50 0.60 C orrelation betw een angle ofinternalfriction and porosity
Mean grain size (mm)

Fig. 2:Correlation between porosity and grain size 50

40

30

20
300
250 10
Permeability (mD)

200
0
150
0 10 20 30 40
100 P orosi
ty (% )

50
0
0.00 0.10 0.20 0.30 0.40 0.50 0.60 Fig. 6 : Correlation between angle of internal
friction and porosity
mean grain size (mm)

Fig. 3 : Correlation between grain size and

permeability
11 Intelligent Sand Management SPE 98818

0.3
0.25
0.2
Poisson ratio

0.15
0.1
0.05
0
0 10 20 30 40
Porosity (%)

Fig. 7 : Correlation between Poisson ration

and porosity

0.03

0.025

0.02

0.015
error

0.01

0.005

0
0 50000 100000 150000
epoch
Fig. 10 : Continuous grain size prediction and
Fig. 8 : Neural network training measured grain size

0.2

0.18
Predicted grain size (mm)

0.16

0.14

0.12

0.1
0.1 0.12 0.14 0.16 0.18 0.2
Measured grain size (mm)
NN predicted Multivariate predicted

Fig. 9: Comparison of Neural network model with Fig. 11 : Example of UCS profile log
multivariate model
12 M.B.Oyeneyin; Chris Macleod; Gbenga Oluyemi and A. Onukwu SPE98818

SAND VOLUME SAND GRAIN SIZE

6.0000

5.0000
Sand Grain Size vs Volume

4.0000

3.0000

2.0000

1.0000

0.0000
9985
9970
9955
9940
9925
9910
9895
9880
9865
9850
9835
9820
9805
9790
9775
9760
9745
9730
9715
9700
9685
9670
9655
9640
9625
9610
9595
9580
9565
9550
9535
9520
9505
10000

Depth (ft)

Fig. 12: Sand Volume vs. Grain Size Distribution

100

90

80
SAND A
SAND B
70

60
Cum. WT; %

50

40

30

20

10

0
1 10 100 1000

SIze, Microns

Fig. 13: PSD of Adriatic Sands

1.2

Open Hole Perforation Synthetic Proppant

0.8
k/ki

0.6

0.4

0.2

0
0 20 40 60 80 100 120 140 160 180

Fines Concentration, gm/l

Fig. 14: Impairment Profile