SocletYof Petroleum Ermineers

SPE 50669

Screening criteria for evaluating the development potential of remaining oil in mature
fields

K.J.Weber,SPE, and H. Dronkert (Delft, University of Technology).

Copyright 1998, Society of Petroleum Engineers Inc.

viable targets.
This paper was prepared for presanfation at the 1998 SPE European Petroleum Conference held
in The Hague. The Netherlands, 20-22 October 1998.
Introduction
This paper was selectad for presentation by an SPE Program Committee following review of
information contained in an abstract aubmittad by the author(s). Contents of tha paper. as
Numerous oil fields which have been in production for many
presented, have not baan reviewed by tha Society of Petroleum Engineers and are subject to years are being reviewed to evaluate options for increasing the
cmrsction by the author(s). The material, as presanted, does not nacesaarily reflect any position
of the Society of Petroleum Engineers, its officers, or mambars. Papers presented at SPE ultimate recovery. The task involves the determination of the
meetings are subject to publication raview by Editorial Committees of the Society of Petroleum
Engineers. Elac.tronic reproduction, distribution. or storage of any part of this paper for
volume and location of mobile remaining oil and subsequently
commercial purposes without the written consent of the Society of Petroleum Engineers is the technical and economic assessment of methods to recover
prohibited. Permission to reproduca in print ia restricted to an abstract of not mora than 3tXl
wcfds illustratic+m may not be copied. The abstract must contain conspicuous acknowledgment this oil.
of where and by whom the paper was presented. Write Librarian, SPE, P.O. Box 833838.
Richardson, TX 75063-3836. U. S.A., fax 01-972-952-9435,
The first part of this task is often difficult because of the poor
data quality associated with old fields. This is particularly true
for eastern Europe, former Russian republics and South
America where many fields are on offer for farm-in type
Abstract
ventures. Certain basic data are usually available and the
Continuing reservoir management at mature stages often purpose of this paper is to provide first round screening criteria
concentrates on delineating pockets of remaining mobile oil. based on these data in order to select those reservoirs for which
This is becoming a major task for reservoir geologists and re-development schemes are likely to be economic. For the
petrophysicists. Many old fields are coming up as investment selected reservoirs further study and some additional data
targets for reactivation and there is general consensus that acquisition is warranted.
modern techniques can lead to additional recovery of between 10 The data which may be expected to be present are well logs,
and 20 percent. However, there are many pitfalls which can limited core measurements, basic facies descriptions, original
render an apparently attractive project into an economic failure. oil-in-place and cumulative production figures, contour map and
This paper will discuss the non-technical screening criteria well positions. Essential is also to have access to well
related to reservoir architecture, accumulation condition and completion data. Individual well performance data are often
production history. difficult to obtain.
Mobile remaining oil can be found in a number of predictable The proposed screening scheme is based on a classification
locations in reservoirs depending on their structural style and of types of remaining oil configurations. Once such a potential
facies. oil pocket has been recognized one attempts to assess its
Attic oil along faults is perhaps the most simple configuration economic value by estimating a number of parameters to a
but sizeable volumes of remaining oil can also occur as a limited degree of accuracy. Dip, original accumulation
fimction of reservoir stratification and lateral discontinuity. A conditions, bedding thickness, reservoir profile, porosity
systematic overview of the different types has been compiled distribution and original oil saturation can often be determined
based on structural or stratigraphic lateral continuity and vertical satisfactorily. More detailed reservoir architecture and
reservoir connectivity. This leads to four main types with some particularly permeability distribution are more difficult to
sub-groups for each of which screening criteria have been obtain.
derived on the basis of field examples and models. The classification scheme for mobile remaining oil pockets
The screening criteria specifi minimum conditions which may consists of a split into reservoirs with a high and a low vertical
lead to economic re-development with horizontal side-tracks permeability / connectivity and a fimther subdivision into types
from existing wells. In addition recommendations are given with with a high and low horizontal connectivity.
respect to data gathering to confm the presence of economically

345
2 K.J.WEBER & H. DRONKERT SPE 50669

This classification can usually be carried out. In this paper four are separated by impermeable intercalations (C). This situation
major types of mobile remaining oil configurations, representing is quite common in layer cake reservoirs. Even without the
the four combinations of high and low vertical and horizontal impermeable separations cross-flow may be limited if the
conductivity are being discussed. The screening criteria vertical permeability of the low permeable layer is low. This
presented are based on re-development with pairs of horizontal situation Ii-equentlyoccurs in fluviatile labyrinth reservoirs and
side-tracks from existing wells. A cost of one million dollars has this can occur in combination with the configuration B1 and D.
been assumed per job of recentering the hole, milling the casing The cases selected for this paper are the A type oil-rim
and drilling and completion of the two side-tracks of 300 meter reservoir, the attic oil type B 1, the layer cake type B2 with
length each. This is based on various cost estimates obtained for varying permeability contrasts and bed thicknesses and the
land operations, which form the bulk of the expected projects. fluviatile type reservoir combining the B2 and C characteristics
The economic analysis based on this method and cost level while also taking into account the internal heterogeneity of the
shows a remarkably large scope for re-development of channel-fill bodies. Except for the oil-rim reservoirs, the cases
reservoirs with oil-rims, attic oil cases in faulted reservoirs and studied are light oil reservoirs produced by waterdrive.
layer cake reservoirs with beds of contrasting permeability. Simulation of jigsaw puzzle type reservoirs with adjacent
Fluviatile labyrinth type reservoirs’ are much more difficult to re- sand bodies with large permeability contrasts has shown that
develop but a number of observations are being made to indicate retention of significant volumes of mobile oil in the low
the more favorable configurations. permeable bodies requires very large contrasts which rarely
occur in naturez. Thus, such configurations have been left out of
Classification of Mobile Remaining Oil Configurations consideration in this study.
The retention of mobile oil in sufficiently large volumes to allow
economic re-development is largely controlled by the presence Oil Rim Reservoirs. Producing oil rim reservoirs with vertical
of heterogeneous pressure distribution, and the fluid density and wells will always leave behind considerable volumes of
viscosity contrasts. This paper is only concerned with sandstone remaining movable oil as a result of cusping and coning. Away
reservoirs containing light oil with have been developed with from the drainage points the oil column may be large enough to
vertical wells, produced under reasonable draw-down conditions. be produced economically with horizontal side-tracks. If
In view of the potential for recompletion and infill drilling, the conveniently positioned shale breaks are present and the dip is
most important heterogeneities are faults, boundaries of genetic low, individual vertical wells can sometimes produce oil rims
units, large permeability contrasts and baffles to flow such as quite effectively. A small well spacing obviously leads to a
shale intercalations. Following the subdivisions of elastic higher recovery factor than a large spacing. Data from 29 oil
reservoirs into layer cake, jigsaw puzzle and labyrinth types one rim reservoirs, mostly with fairly high vertical conductivity and
can already predict a number of typical oil displacement patterns. well spacings of 600 to 800 m were collected to form a basis of
Considering the major large scale heterogeneities we can comparison for our theoretical models.
subdivide the reservoirs into types with a high vertical Considering a homogeneous anisotropic reservoir produced
conductivity and those in which the stratification and low at a moderate supercritical rate with a given number of wells,
permeable intercalations result in a low vertical conductivity. the final configuration of the remaining mobile oil may be
Next we can fi.wther make a distinction between layer cake assumed to be independent of the original oil column. Using a
reservoirs with a high degree of lateral continuity of the beds and detailed reservoir simulation of a field as example3, a theoretical
reservoirs in which the lateral continuity is limited by faults or model was defined by iteration to fit the observed trend in
pinch-outs of the sand bodies. This leads to a scheme as shown recove~ factors (Fig. 2a). Under condition of a strong
in Fig. 1. To the first category A we can attribute oil rim waterdrive, the gascap essentially stays in place except for the
reservoirs with a high vertical conductivity in which formation of cusp towards the producing well. The water will
unproducible oil columns are left between the vertical wells as move up until the top of the water cone and/or the base of the
a result of cusping and coning. gascone will reach the perforated zone. With a cusping zone of
Poor lateral continuity can be caused by a normal fault (B 1) 5 m in which the recovery coefficient is 0.2 and a coning zone
which even, when non-sealing over the juxtaposed reservoir of 10 m with a recovery coefficient of 0.1, we obtain overall
interval, traps oil in the up-thrown block against the caprock in recoveries that follow the trend of the 29 reservoirs (Fig. 2b).
the down-thrown block. Depending on the throw of the faul~ the The actual recovery is somewhat higher for most fields which
structural dip and the distance of the vertical wells from the fault, may be attributed to the presence of shale breaks holding back
a volume of oil will remain behind when the wells water out. the coning effect4 and to careful monitoring of the production
In labyrinth reservoirs one finds updip stratigraphic traps (B2)
rate during the final stages before closing-in the wells.
especially in low net/gross cases. In such cases we will also
This result gave encouragement to set up a general
encounter poor sweep efficiency unless the well spacing is small
simulation model to predict firstly the recovery fi-oma given oil
(D). Poor sweep can also result from stratification with large
rim thickness and secondly the recovery from a pair of
permeability contrast between the beds, particularly when these
horizontal side-tracks drilled into the remaining oil zone (Fig.

346
SPE 50669 SCREENING CRITERIA FOR EVALUATING THE DEVELOPMENT POTENTIAL OF REMAINING OIL IN MATURE FIELDS 3

2c). The model consisted of a homogeneous anisotropic volume is a fimction of the throw of the fault, the dip, the
sandstone with a horizontal permeability of 1000 mD and a- thickness of the reservoir, the permeability profile and the
vertical permeability of 100 mD. The porosity is 0.25 and N/G position of the wells together with the production policy. Thus
is 0.80. The dimensions are 800m in the strike direction and it is necessary to study the relative influence of a number of
1200m in the dip direction, thickness is 40 m and the dip is 5 parameters to obtain a realistic idea of the remaining mobile oil
degrees. The model is subdivided into variable size blocks of volumes.
10x1Oto 80x100 in the x-y directions with a constant thickness To get an impression of the volumes that are commonly left
of 5 m. Oil viscosity is 0.3 cp (0.3 * 10-3Pas). The oil rim height in attic traps, a study was made of numerous fields with
was varied from 20 to 100 m. Simulation has been carried out structures like faulted anticlines, tilted blocks, and those formed
with the IMEX advanced reservoir simulator (Computer by salt domes or wrench fault tectonics. Assuming a horizontal
Modelling Group). The grid size is reduced around the central oil-water contact at the level of the intersection of the well and
vertical well and along the paths of the subsequent horizontal the top of the reservoir one usually observes oil volumes in the
side-tracks. range of 100000 to 300000 m3for distances to the faults of 100
The resulting graph of oil recovery percentage versus oil to 300 m and dips of 1 to 10 degrees. However these volumes
column height for vertical wells shows a reasonable but rather are in reality much higher when coning around the producing
optimistic fit with the field data (fig. 2c). This is partly due to a wells is taken into account as shown in Fig. 3a.
careful choice of production rate as a function of oil column and A series of reservoir simulations were carried out with the
permeability based on the studies of Schols 5.The vertical wells IMEX black oil simulator, using a light undersaturated crude
were abandoned when a 90 percent water cut or a GOR of 1000 and the series of parameter permutations shown in Table 1. The
m3/m3was reached. model consisted of a rectangular grid with 19x20x 10 blocks in
After the abandonment of the vertical well the simulation is the x-y-z directions with a size of 920 m in the x (strike) and
continued with the production of the pair of side-tracks as shown 600 m in the y (dip) direction. As detailed in Table 1, the
in Fig. 2c. The abandonment conditions were the same as for the vertical thickness and permeability profile were varied to
vertical wells. This results in recoveries which are from 8 to 10 represent an average barrier bar, fluviatile or homogeneous
percent higher than those achieved with vertical wells only (Fig. anisotropic reservoir. Firstly the production was modeled of a
2b). An economic assessment of these results for an oil price of vertical well situated in the centre of the grid at distances of 50
15 US$/bbl, an investment of 1 million US$ and average figures to 250 m from the fault forming the upper no-flow boundary.
for operating cost and government take, shows that side-tracking The production was stopped when a water cut of 90 percent was
into remaining oil zones of oil rim reservoirs is attractive for a reached. Well production levels were optimised to obtain
wide range of cases. For a 5 year production period of the maximum recovery.
horizontal wells, net present values of from 1.5 to nearly 4 MM After a 200 day time lapse to stabilise reservoir pressure, the
US$ were calculated for respectively a 20 and a 60 m thick simulations were carried on with the production of a pair of
original oil rim. horizontal side-tracks along the fault, each with a perforated
The conclusion is that for reservoirs that are clearly quite length of 280 m. Again production was stopped when a 90
homogeneous, side-tracking is almost certainly a viable project. percent water cut was reached. A strong natural waterdrive was
Oil rim reservoirs with low ultimate recovery relative to the assumed.
results shown in Fig. 2b may also be interesting. Apart from poor The results of the simulations of the vertical well
production practice this low recovery may be caused by performances and the remaining oil volumes prior to drilling the
particular types of heterogeneity such as combinations of thin horizontal side-tracks show a large discrepancy between the
high permeability streaks coupled with dips above 5 degrees. For volumetric estimates made with the planimeter. Especially when
such cases specially designed horizontal side-track trajectories dips are low the assumption of a horizontal oil-water contact
can be designed. In this context it is usefid to study a number of results in a significant underestimation of the remaining oil (Fig.
recent case histories of the development of oil rim reservoirs 3b). The error can be as large as a factor of 100 for low dips and
with horizontal wells 67,8For
. the more homogeneous reSWVohX small distances to the fault. To obtain an estimate of recoverable
the screening activities should be concentrated on the oil for the pair of side-tracks one has first to determine the
permeability profile to ensure that the remaining oil distribution volume between well and fault using a horizontal contact, next
resembles the modeled configuration. one applies the correction indicated by Fig. 3b (different for
each reservoir type). Finally one uses the relationship between
Attic Oil. The next configuration for which screening criteria distance to fault, dip and recovery factor for the appropriate type
were derived is the attic oil type (fig. 3a). Numerous oil fields (Fig. 3c).
are either bounded by faults or are intersected by crestal normal The actual screening procedure should include a critical
faults. If oil occurs adjacent to the faults a certain volume of oil review of the accuracy with which the fault position is known.
remains trapped between the fault and the nearest wells. This If the fault has not been penetrated near the attic trap, this

347
 4 K.J.WEBER & H. DRONKERT SPE 50669

accuracy may be quite low. Seismic positioning of faults for land thickness and permeability ratio. Surprisingly it appears that
data is occasionally as accurate as about 25 m but more even in beds of only 2 m. economic projects may be carried out
commonly one must reckon with values of 50-100 m. For older if the permeability contrast is high. For thicker beds there is
marine data the accuracy may be much poorer. A second considerable leeway before projects become uneconomic
important matter is the type of fault. In this study single well provided the sidetracking can be carried out for the assumed
defined fault planes have been assumed. However, tilted blocks figure of 1 MM US$.
are frequently bounded by complex fault zones. Wrench faults Sidetracking into thin beds of moderate to low permeability
are also commonly of a complex nature. Diagenesis around fault in mature fields is already planned12 and carried out*3.In the
zones is another complicating factor. A wise precaution against Venezuelan case (Fig. 4c) logging in an infill well clearly
unfavorable surprises is the drilling of a pilot hole penetrating showed the high oil saturation in several lower coastal plain bars
the fault plane. This also allows designing optimally placed while an underlying highly permeable fluviatile sandstone was
trajectories for the two side-tracks. watered out. Horizontal wells are planned for these undrained
The recovery factors obtained in the simulations vary widely. beds for which recovery per well in the rangeof50000 to 200
As can be expected they are highest for the barrier bar type for 000 m3 is expected from beds from 7 to 10 m thick with
which, even for low dip and distances between fault and vertical permeability ratio’s in the range of 0.01-0.1. This fits rather
well of 100 m, a recovery factor of about 20 percent is reached well with the values which can be predicted with the graph in
(Fig. 3c). For the homogeneous anisotropic reservoir one needs figure 4b.
distances of 200 m to reach this level. Nevertheless this can still
amount to a volume of100 000 m3or more. Thus it is no surprise FIuviatile Labyrinth Reservoirs. This type of reservoir is
that attic oil recovery with horizontal wells is already common commonly associated with low recovery factors even for light
practice 9’10. undersaturated crude and high horizontal permeability levels.
This is attributed to a number of different causes. Firstly, there
Layer cake Resewoirs. The study of layer cake reservoirs was
is the discontinuous nature of the composing sand bodies,
especially when netigross ratio’s are below 0.5. Secondly, there
started with a survey of sedimentological configurations leading
to a combination of thick high permeability beds and thinner is usually a strong stratification within the sand bodies with the
lower part of the channels being filled with high permeability
layers with considerably lower permeability. The results show
cross-bedded sandstone and the upper part with tow energy
the very common nature of this combination (Table 2). We are
flaser bedded sandstone with low vertical permeability. Thirdly
particularly interested in cases for which there is no cross-flow
there is the phenomenon of capillary trapping in the coarse
between the high and low permeability beds. This is often the
laminae of the cross-bed sets which may account for 10-15
case in dekaic offshore and marine sediments with their very
percent of the remaining oil[4. Clearly it will be difficult to
continuous shale intercalations”. However, lower coastal plain
predict the actual distribution of remaining mobile oil in these
crevasse splays and lagoonal mouth bars can be separated from
reservoirs.
very permeable fluviatile sand bodies by clayey soils and coals.
To study the effects of the macroscopic parameters a study
Even if a reservoir has been perforated over its entire height,
was made of reservoir architecture and sand body size for so-
the low permeable zones may be left poorly drained because of
called non-migrating distributary channel type reservoirs. A
the preferential flow through the highly permeable “thief’ zones.
number of deterministic prototype models were available from
To model this situation a series of simulations were carried out
fields in Venezuela, Borneo and the Gulf Coast. In addition
with the IMEX black oil simulator. The highly permeable bed
much outcrop data has been collected by the Applied
was always 10 m thick with an isotropic permeability of 1000
Geoscience Department at the Delft, University of Technology.
mD. The thin beds were from 2 to 8 m thick with permeabilities
Statistical analysis was carried out on size and shape of
ranging from 50 to 200 mD. Models were made with the low
distributary channels. A graph relating the expectation of the
permeability layer on top or below the high permeability bed. For
widthlthickness ratio as a function of the thickness was
comparison runs were also made without impermeability barriers
constructed. In addition, a study was carried out of the radius of
between the beds. The dip was kept at 5 degrees.
curvature distribution in such reservoirs for which recent
Like in the previous cases, the production of the vertical wells
examples of coastal plains were also used. This provided a
was modeled first up to the point of raeching a 90 percent water
graph showing the average cumulative fkequency distribution of
out (Fig. 4a). A light undersaturated crude was used and a strong
the radius of curvature normalized by dividing by the channel
water drive was assumed. As can be seen, the upper low
width. These two graphs can be used to construct reservoir
permeable bed remains poorly drained. The subsequent modeling
models that fit the natural distribution of this geometrical
of the production of twin horizontal side tracks of 300 m each
parameter. An example which has been used for simulation
leads to the graph shown in Fig. 4b. The relationship is shown
studies is shown in Fig. 5a.
between the cumulative production, the ratio of low to high
The model shown has a net to gross ratio of about 0.5. At
permeability, and the thickness of the thin bed. Also shown is the
that level the interconnectivity of the system is quite high’s.
net present value (NPV) related to projects as a function of bed

348
SPE 50669 SCREENING CRITERIA FOR EVALUATING THE DEVELOPMENT POTENTIAL OF REMAINING OIL IN MATURE FIELDS 5

Already at levels of just over 0.3 the fluviatile reservoirs are Conclusions
usually interconnected although obviously many dead ends exist. The major conclusion from this scouting study of screening
An important matter is the erosion of younger into older channel- criteria is that there is considerable scope for horizontal
fills. This process removes all or part of the upper low energy sidetracking for many remaining oil configurations. If these
part of the channel-fills. As shown in Fig. 5b the percentage of projects can be carried out for costs of the order of I MM US$
low energy fill shows a sharp decline when the net to gross ratio for a pair of 300 m long horizontal holes, the economics look
increases above the 0.7 range. However many fluviatile robust for a wide range of oil rim, attic oil and layercake type
reservoirs have much lower n/g ratio’s. A study of the Daqing reservoirs. For the fluviatile reservoirs the situation is more
field in China has shown the importance of the low energy fill complex and much additional work is required to obtain a
with respect to the vertical sweep in the channel-fills ‘G.The low reliable system of predicting potential oil production from
vertical permeability prevents the water from sweeping horizontal sidetracks.
significant volumes of oil from the low energy fill even though It appears that horizontal sidetracking is already becoming
horizontal permeability is high. common practice and it can be expected that much actual
It was decided to carry out simulations in the prototype model performance data will become available soon. There are also
(Fig.5a) to get an idea of the relative importance of the indications of lowering cost and improving technical feasibility
stratigraphic trapping in such a reservoir as a result of bends and and performance.
discontinuities and the effect of the stratification within the The screening procedure recommended is to use graphs of
channel-fills. The upper part of the channels consists of low the type incorporated in this paper as optimistic base case
energy fill with a Kh of 200 and a Kv of 5 mD while the lower predictions. With a set of graphs of this kind and limited field
half consists of high energy fill with a Kh of 1000 and a Kv of data one can already carry out a first round selection of likely
50 mD. The capillary phenomena were left out of the simulation. prospects for re-development. This reduces the initial work
The model consists of a 2000x2000x 12 m size reservoir with a significantly and leads to a much more rational selection of
dip of 2 degrees. The grid is IOOX100 m in the x-y direction, with cases for more detailed evaluation. The ideal approach may be
6 layers of 2 m thickness in the z direction. The oil is again a the construction of a computerized expertise system which
light undersaturated crude and a strong water drive is assumed. allows a quick screening in a data room and which provides a
A grid of wells is drilled with 400 m spacing. Two range of economic results for rather basic input which can be
configurations have been modeled, one with the channel trend to estimated readily by the specialist examining the maps and files.
the dip direction and one with the trend parallel to the strike. Developing screening criteria for re-development projects is
Well productivities were chosen which reflect the permeability considered a very useful research project which may be very
profile of each well. A 25 year production life was modeled. suitable for a university in combination with an oil company
The results show a recovery factor of 5 I percent for the model providing realistic cost data and field examples.
with the channels running parallel to the dip direction and 46
percent for the other case. This implies that the impact of the Acknowledgements
stratigraphic trapping is rather limited. Of much larger During 1998 a new research project was started at the Applied
importance is the retention of oil in the upper low energy part of Geosciences Department of Delft, University of Technology,
the channel-fills. In Fig. 5C this can clearly be seen when we called “RoRo” or Recovery of Remaining Oil. Four students,
look at the high oil saturation left in the upper layer of the model D.E. Van Bourgonje, A.Verwey, J.Buter and O.KUYUCU have
which consists entirely of low energy material. been engaged on this project and their cooperation is gratefully
The main conclusion is that it is likely that much of the acknowledged
remaining mobile oil is situated in the upper parts of the channel-
fills. Whether there is scope for retrieval of this oil is difficult to References
assess. From the study of the Iayercake reservoirs for the case 1. Weber,K.J.and van Geuns, L.C. : “Frameworkfor Constructing
without impermeable barriers between the beds it could be Clastic Reservoir Simulation Models”, J. Petr. Techrz., VOI.42,
no. 10, 1248-1253, 1296-1297, 1990.
concluded that there is the possibility of producing significant
2. Barthel, R. : “The Effect of Large-Scale Heterogenieities on the
volumes of oil via horizontal wells in the poorly swept zones Performance of Waterdrive Reservoirs”, SPE paper 22697,
even where these are only a few meters thick. presented at the 66* Annual Techn. Conf.& Exh., Dallas, TX,
It is recommended to continue the study of the possible Ott.6-9, 1991.
production from horizontal sidetracks into the upper part of 3. Weber, K.J., Klootwijk, P.H., Konieczek, J. and van der Wugt,
channel-fills. From the simulation data a thickness of 3 m or W.R.: “Simulation of Water Injection in a Barrier-Bar-Type, Oil
more may be enough for an economic project. Other processes Rim Reservoir in Nigeria”, J. Petr-Teck, November 1978, 1555-
for improving recovery from these fluviatile reservoirs may well 1565.
4. Richardson, J.G., Sangree, J.B. and Sneider, R.M.: “Coning”,
be more attractive but these are beyond the scope of this paper.
J. Petr. Techn., August 1987, 883-884.
5. Schols, R.S.: “An Emperical Formula for the CriticaI Oil
Production Rate”, Erdoel-Erdgas, VOL88, no.1,6-11, 1972.

349
6 K.J.WEBER & H. DRONKERT SPE 50669

6. Harst, A.C. van der: “ErbWest: An Oil Rim Development with 11. Weber, K.J.: “Influence of Common Sedimentary Structures on
Horizontal Wells”, SPE paper 22994, presented at the SPE AEia- Fluid F1OWin Reservoir Models”, JPetr.Techn., March 1982,665-
Pacific Conf. Perth, W.Australi~ 4-7 Nov. 1991. 672.
7. Kossack, C.A. and Kleppe, J.: “Oil Production from the Troll Field: 12. Bryant, I. D.,Herron, M.M. and Garnero de Villarroel, H.:
A Comparison of Horizontal and Vertical Walls”, SPE paper 16869, “Application of Sequence Stratigraphic Reinterpretation of Lower
presented at the 62ndAnnual Techn.Conf. And Exh., Dallas, TX, 27- Lagunillas Member to Futher Development of Bloque IV, Lake
30 Sept., 1987. Maracaibo, Venezuela”, SPE paper 36538, presented at the 71th
8. Pelgrom, J.J., Mann, J.P., Davidson, C.J., Tickle, A.J., Blok, R.H., Annual Techn. Conf. And Exh., Denver, Colorado, 6-9 Sept.,
Mattjes, G.R. and Risseeuw, A. S.: “Improved Well Design 1996.
Maximises Oil-Rim Profitability”, SPE paper 27737, presented at 13. Faure, A.M., Zijlker, V.A., van Elst, H. and van Molsen, R.J.:
the European Prod. Op. Conf. And Exh., Aberdeen, U.K., 15-17 “Horizontal Drilling with CoiIed Tubing: A Look at Potential
March, 1994. Application to North Sea Mature Fields in Light of Experience
9. Martins, E.J. and Larez, N.J.: “Recovery of Attic Oil Through Onshore the Netherlands”, SPE paper 26715, presented at the
Horizontal Drilling”, SPE paper 26334, presented at the 68ti Annual Offshore Europe Conf., Aberdeen, U.K., 7-8 Sept., 1993.
Techn. Conf, And Exh., Houston, TX, 3-6 Oct., 1993 14. Lingen, P.P. van and Knight, S.: “Evaluation of capillary
10. Vo, D.T., Marsh, E.L., Sienkiewicz, L.J. and Mueller, M.D.: “Gulf entrapment within reservoir flow units”, SPE paper 38934,
of Mexico Horizontal Well Improves Attic Oil Recovery in Active presented at the 72ndAnnual Techn. Conf. And Exh., San Antonio,
Water Drive Reservoir”, SPE/DOE paper 35437, presented at the TX, Oct. 1997.
10* SPE/DOE Symp. On Improved Oil Recove~, Tuls% OK, 21-24 15. Knight, S.: “Stratigraphic and Structural Controls on Reservoir
April, 1996. Connectivity: a 3D Modelling Approach”, PhD Thesis, University
of Liverpool, U.K., 1996.
16. Zhiwu, W., Qimin, W., Bohu, L., Chengjing, L. and Xiangzhong,
L.: “Ways to improve development efficiency of Daging Oilfield
by Waterflooding”, in Oil Field Development Techniques: proc. of
the Daging Internat.Mtg., Mason, J.F. and Dickey, P.A.(eds),
AAPG Studies in Geology N0,28, 1982.

Table l-Reservoir simulation scheme with 36 combinations of parameters for performed simulations.

Reservoir dip

2-5-10-20 degrees

Distance to fault

50-110-250 meter

[
Homogeneous Anisotropic Barrier Bar Fluviatile

thickness Kh Kv/Kh thickness Kh KvlKh thickness Kh Kv/Kh

m mD m mD m mD

50 500 0.05 15 1000 0.1 5 250 0.05

5 500 0.05 10 1000 0.05

5 250 0.05

350
 SPE 50669 SCREENING CRITERIA FOR EVALUATING THE DEVELOPMENT POTENTIAL OF REMAINING OIL IN MATURE FIELDS 7

Table 2- Ovewiew of layer cake facies and dimensions, H=common thickness range

Environment of deposition Thin bedded sandstones H Associated sandbodies H

fluvio / Iacustrine sheet sands 0.2-2 channel fills 5-20

crevasse splays 1-3 point bars 5-20

deltaic inland crevasse splays 1-3 channel fills 5-20

lagoonal mouth bars 1-5 point bars 5-20
Iagoonal sheet sands 0.5-2 barrier bars 10-25
tidal flats 0.5-2 tidal channel fills 10-40

deltaic offshore, river dominated mouth bars 1-15 channel fills 5-20
transgressive sandstones 1-1o

deltaic offshore, wave dominated upper shore face 3-15 channel fills 5-20
lower shore face 3-1o barrier bars 10-25
tidal flats 0.5-2

deltaic offshore, tide dominated tidal bars 2-1o tidal channel fills 10-40
tidal flats 0.5-2

marine, shallow bars 1-1o ridges 10-40

storm deposits 0.1-1

marine, deep turbidites 0.5-10 Channel fills 1o-1oo

overbank splays 1-3 thick bedded fan 1o-1oo
distal fan 0.1 -0.5 turbidites

High Kv Low K ~
Good continuity Poor continuity Good continuity Poor continuity
updip permeability
unproducible columns attic oil stratigraphic trap contrasts poor sweep

A BI B2 c D

Fig. 1- Classification scheme of remaining oil types.

351
 K.J.WEBER & H. DRONKERT SPE 50669
8

A Theoretical model

——.———— 7A
-----A
U. R.= (hl*O.2+h2*0.l +h3*0.55)/H

CUSPING ZONE R. C.= 0.2 ; %.illllll

‘---CONING ZONE R. C.=0.1

!?!1
“:
1[ i

I U.R. = Ultimate
R.C. = Recovery
recovefy
coefficient

70

60

50

40

30

20 i: +
.i~eh
.
----
. . . . .
. . . . ...--”
-. . ..-.

~
. ...---””
------

.. . . .. . .

++

.. . . . . . . . . Trend

Calculated
. . . . . . . . . . ..-

Field Data
(see A)
●

---- Trend IMEX vertical

10 ●
. ...0 Trend side-tracks
t
o
0 20 40 60 80 100 120

Oil Column Height (III)

c
II
30C

Fig. 2- Oil Rim Reservoirs. A: Theoretical model indicating parameters used by simulating coning and cusping configurations.
B: Recovery factor and oil column height of primary production, IMEX simulation and theoretical model calculations. C: Sketch
of oil saturation distribution based on the simulation result with two side-tracks of 300m.

352
SPE 50669 SCREENING CRITERIA FOR EVALUATING THE DEVELOPMENT POTENTIAL OF REMAINING OIL IN MATURE FIELDS 9

MULTIPLIER AS FUNCTION OF
B RESERVOIR DIP AND DISTANCE TO FAULT

10 100 1000
Distance-to-Fault [m]

c RECOVERY FACTOR AFTER FIVE YEAR

PRODUCTION IN BARRIER TYPE RESERVOIR

35 \
35% 40% 45%
g 30
.+ovo \
Fig. 3- Attic configurations. A: Distribution of ~ 25
oil saturation after drilling the pilot hole but & ’20
before side-track production. B: Correction factor
(multiplier) to be used for different dips and distances \
to fault in the case of a homogeneous reservoir.
The multiplier corrects the planimetered volumes \ \
20%
to simulated volumes. C: Recovery factors for a \
barrier type reservoir depending on reservoir dip \ \
\
and distance to fault.
o 100 200 300
Distance-to-Fault [m]

353
10 K.J.WEBER & H. DRONKERT SPE 50669

A B
1. Initial situation
Bottom layer 10 m and 1000 mD, “ z.. = NPV (mIn. $)
top layer 4 m and 100 mD
- 200000
““ L — = top layer thickness (m)

“*-
. .

-- . . . . . . ----

-------- . . ------- -------

2. Vertical well with 90?40watercut

0 ‘! (

Vertical well with 90% watercut is initial situation for

horizonal sidetracks 0.05 0.1 0.15 0.2
K Top/K Bottom

3. After sidetrack production

Horizontal sidetracks with
a watercut of 90°A

no flow boundary
Drained oil

Fig. 4- Layer cake reservoirs. A: Profiles of simulation runs indicating the remaining oil in the upper layers of layered low Kh/Kv
reservoirs. B: Net present value and top layer thickness for different cumulative production and permeability ratio between
Kvtop/Kvbot. Already with a 2m thick sand layer and a Kvtop/Kvbot <0.01 recovery is profitable. C: Example of the Lower
Lagunillas in Venezuela’z where a low permeable layer lies on top of high permeable sands.

354
SPE 50669 SCREENING CRITERIA FOR EVALUATING THE DEVELOPMENT POTENTIAL OF REMAINING OIL IN MATURE FIELDS 11

PERCENTAGE NET /GROSS and PERCENTAGE LOW ENERGY TOP CHANNEL FILL
R
u

NIG-O.85 percentage Iowenargy M 0.27

Top slice
c
1:
producing
well

Second slice
Fig. 5- Fluviatile labyrinth reservoirs. A: 3D model of fluviatile channels with lower permeable upper parts. The N/G is 0.5. B:
Simulation profiles with low permeable upper parts. The effect of the N/G has a more than linear impact on the volumes of the
low energy low permeable top layers. C: Oil saturation distribution from simulation results shows producing well positions and
the two top horizontal slices displaying the oil trapped in the up dip upper part of the low energy channel tops.

355