IPTC 11691

Downloaded from http://onepetro.org/IPTCONF/proceedings-pdf/07IPTC/All-07IPTC/IPTC-11691-MS/1802076/iptc-11691-ms.pdf/1 by Middle East Technical University Ankara user on 23 October 2021
Evaluating Connected Reservoir Volume for Optimizing Reservoir Management in
Farragon Field, an Offshore North Sea New Development
Julio Herbas, SPE, Munawar Usman, SPE, Ronnie Parr and Jordy Buter, BP Exploration Operating Company Limited

Copyright 2007, International Petroleum Technology Conference

that drove the implementation of optimized reservoir
This paper was prepared for presentation at the International Petroleum Technology management strategies.
Conference held in Dubai, U.A.E., 4–6 December 2007.

This paper was selected for presentation by an IPTC Programme Committee following review
of information contained in an abstract submitted by the author(s). Contents of the paper, as
Introduction
presented, have not been reviewed by the International Petroleum Technology Conference Preliminary classical reservoir engineering studies performed
and are subject to correction by the author(s). The material, as presented, does not
necessarily reflect any position of the International Petroleum Technology Conference, its on the Farragon field during the first year of oil production,
officers, or members. Papers presented at IPTC are subject to publication review by Sponsor
Society Committees of IPTC. Electronic reproduction, distribution, or storage of any part of this
suggested a bigger than originally estimated volume of initial
paper for commercial purposes without the written consent of the International Petroleum oil in place. This was in-line with the production performance
Technology Conference is prohibited. Permission to reproduce in print is restricted to an
abstract of not more than 300 words; illustrations may not be copied. The abstract must which was consistently above the initial expectations.
contain conspicuous acknowledgment of where and by whom the paper was presented. Write
Librarian, IPTC, P.O. Box 833836, Richardson, TX 75083-3836, U.S.A., fax 01-972-952-9435.
Basic Decline Curve Analysis, Pressure Transient Test
Abstract interpretations, material balance and numerical simulation
The Farragon field, discovered in April 2003, is a low relief, models, demonstrated consistency with field trends and with
pancake shaped reservoir located in the UK sector of the analogous fields, although the production period by the time
North Sea. This relatively small offshore field was developed the first studies were performed was quite short.
with two sub-sea horizontal wells tied to existing production
facilities. Initially the light 34 °API oil was produced by Transient test data were interpreted to determine the reservoir
natural flow and subsequently gas-lift has been used for properties and to estimate the connected volumes. Material
artificial lift. A few months after first oil, the field was balance provided oil in place results consistent with those
consistently achieving production rates higher than predicted from the pressure transient tests, a quantitative insight into the
which led to a decision to enhance the application of early production mechanism and water front displacement.Water
extensive reservoir engineering studies aiming to better breakthrough occurred in the deeper well # 2, 6 months after
understand the reservoir mechanism, volumes in place and 1st oil. The water cut was matched and, the water breakthrough
their implications for field depletion plans. in the shallower well # 1 has been predicted to occur after
some 20 months of the field production start.
Data from permanent pressure gauges installed in the two
open-hole gravel-packed horizontal producers was analysed to Full field numerical reservoir simulation studies were then
improve understanding of the reservoir. Average reservoir performed confirming previous results from the classical
pressure, productivity index and connected volume were reservoir engineering analyses. This provided the basis to
interpreted from build-up & drawdown tests. The calculated justify an important increase in the field volumes and
oil in place volumes were history matched with multiple recoveries and for the initiation of further studies aiming to
Material Balance runs which used BP’s TDRM, Top Down investigate alternate exploitation schemes to increase the final
Reservoir Modelling1 process. This powerful tool is designed recovery further.
to evaluate reservoir engineering uncertainties in a wide range
of scenarios considering multiple variables providing much Geology and Reservoir Description
greater confidence in the obtained results The Farragon reservoir lies in sands of Palaeocene age, close
to several existing fields: the mature Andrew and Cyrus fields
Full field multidimensional numerical simulation models with to the South-West and the Arundel discovery to the North-
updated volumes from new 3D seismic reinterpretations, West. The sediments were deposited in a submarine fan
history matched the field performance, confirming previous environment consisting of distal turbiditic facies. The
classical reservoir engineering conclusions. reservoir thickness is up to 55 m gross (40 m net), limited at
the top by a structural seal and at the bottom by a large aquifer
The outcome before the first 18 months of field life of regional extent. The oil water contact is at 2556.5m TVDss.
demonstrated the value of extensive application of reservoir Figure 1 shows a depth map of one realisation of the four-way
and petroleum engineering techniques in a very early stage of dip closed structure.
the field life as it resulted in a 25% increase in estimated oil
recovery. This paper outlines the general methodology applied The reservoir consists of two principal oil bearing sand bodies
defined as Upper and Lower Sands with different
2 IPTC 11691

petrophysical characteristics, separated from the aquifer by an Table 1: Reservoir Properties

extensive, correlatable shale that controls the water Initial Reservoir Pressure, psi 3570
encroachment. The Upper Sand exhibits better petrophysical Saturation Pressure, psi 1100
Current Pressure, psi 3300
characteristics and is interpreted by seismic stratigraphy to be Reservoir Temperature, °F 228
more confined than the Lower Sand. The fault system Oil Gravity, °API 34
comprises a number of normal faults across the field, with Solution Gas Oil Ratio, scf/stb 295

Downloaded from http://onepetro.org/IPTCONF/proceedings-pdf/07IPTC/All-07IPTC/IPTC-11691-MS/1802076/iptc-11691-ms.pdf/1 by Middle East Technical University Ankara user on 23 October 2021
small displacement (<20m) and no apparent sealing Brine Salinity, ppm 4200
characteristics. Permeabilities range from 200 to 870 mD, with Formation Volume Factor, vol/vol 1.24
Gas Gravity 0.77
a Dykstra Parson’s coefficient of 1. The average porosity is
24% and Net:Gross varies in the range of 40-80%.
Wells and Production Performance
Two long horizontal production wells were drilled in the area
with best productive sands (Figure 2). Well #1 located at the
top of the structure to drain the Upper Sands and Well #2 at
the bottom principally to drain the Lower and B3 Sands (see
fig. 2). Rapid water encroachment and early breakthrough in
both the wells was predicted. However water breakthrough
occurred in the deeper Well #2 only after production of over 4
million barrels from the field. It is still awaited in Well #1
after production of over 10 million barrels in total. The oil
production started to decline just after the water breakthrough
in Well #2. (Figure 3).

Produced water in the field comes from the bottom and lateral
aquifers and is interpreted as the main production mechanism
that provides pressure support. The water production is closely
monitored to detect breakthrough because of its impact on the
recovery factor. It has reached near 35 % after production of
circa half of the original estimated recovery. Material balance
Figure 1: Top Reservoir Interpretation techniques with relative permeabilities and vertical volumes
distribution were used to predict the water front movement
and to implement strategies to maximize oil recovery. A
significant factor which delayed water encroachment is
considered to be the presence of laterally continuous shales.
For example, the B3 shale observed twice in well #2 and in the
vertical discovery well indicated in Figure 2.

Figure 2: Seismic Cross Section

Figure 3: Wells Production Performance

The reservoir contains under-saturated light oil with initial
pressure nearly twice the bubble point pressure (monophasic
fluid with no gas cap). Table 1 below shows some basic Decline Curve Analysis
reservoir properties. Decline Curve Analysis is a very useful technique to forecast
well performance and reserves. In single phase flow regimes a
strong relationship between the decline trends and some
reservoir parameters like permeability and drive mechanism
IPTC 11691 3

has been established. This is the case in the Farragon field, The Exponential fit decline parameter was found to be in the
where oil is being produced above saturation pressure2. The 0.28 to 0.34 range which is consistent with the decline fit from
applied single-phase liquid exponential decline equation for the analogue field of 0.36. It was concluded that use of the
under-saturated oil is based on material-balance, pseudo analogue Exponential fit parameter would give reliable
steady-state and Productivity Index equations2: predictions; this assumption was afterwards verified with 18
months of production data.

Downloaded from http://onepetro.org/IPTCONF/proceedings-pdf/07IPTC/All-07IPTC/IPTC-11691-MS/1802076/iptc-11691-ms.pdf/1 by Middle East Technical University Ankara user on 23 October 2021
Material-balance:
Pr2= (Pri2/Npu) Np+Pri2) ……………... (1) Pressure Transient Tests
Transient Test analysis was used right from the production
Pseudo steady-state: start up of the Farragon field which was a very useful tool in
Qo=Jo’(Pr2/Pri)-(Pr2Pwf2)n………………… (2) the determination of average permeabilities, productivity
index, connected volumes and skin factors. This was possible
Productivity Index as the two wells are equipped with down hole pressure gauges
Qo=Jo(Pr–Pwf) …………………………… (3) for continuous transmission of pressures that are recorded and
archived in a field data base.
Decline equation
q(t)=qi /edit .…………………………………..(4) The main target from the interpretation of the transient tests
was the determination of a field average reservoir pressure
Where qi: initial pseudo steady-state rate trend to be used as history match parameter in material
Exponential parameter d =1-e-a …………(5) balance and in the full field dynamic simulations. The
a: negative slope of natural logarithm of the oil rate transient tests interpretations were also focused to determine
Hyperbolic parameter di= 1-(1+nai) 1/n …… (6) and quantify the connected volumes to the two producer wells.
The referential parameters used in the interpretations are
DCA Methodology. summarized in Table 2.
The available data was examined and it was found that a wide
range of choke sizes from 10 to 100% had been applied to Table 2: Static Data for Transient Test Analysis
produce these wells. To ensure reliability of the decline curve Formation Volume Factor, vol/vol 1.2
analysis results, points with identical choke sizes where Reservoir Thickness, ft 100
Viscosity, cp 0.6
grouped and used to define the decline parameters. Oil Compressibility, psi-1 10.5 x 10-6
Water Compressibility, psi-1 3.13x 10-6
Exponential and Hyperbolic decline curves were applied on a Formation Compressibility, psi-1 3.6x10-6
well by well and field basis to determine the best fit Total Compressibility, psi-1 1.2 x10-5
parameters. The obtained decline parameters were compared Porosity, % 22
Water Saturation, % 28
and tuned with an analogue field selected based on reservoir
Wellbore Radius, ft 8000
similarities (Figure 4). The Hyperbolic decline showed that
variations in the back pressure exponent (n) between 0.005 to
Transient Test Interpretations.
0.5 resulted in slight changes in the fit of the curve, but
The transient data was interpreted3 with classical techniques:
substantial changes in total reserves. These variable values for
Type Curves models, Derivative plots, Superposition, and
the back pressure exponent (n) are explained by the fact that
Horner plots. Figures 5 and 6 show an example of the pressure
it is an expression of the decline exponent (b) for wells
transient and production data of one of the Farragon wells.
producing at a low flowing pressure Pwf = 0. However, for the
wells in this study Pwf is relatively high from 30 to 60 Bar.
An important event for interpretation was a shut in period of
Given this uncertainty, the hyperbolic decline curves were not
nearly 20 days after the first 8 months of production when the
used for further analysis3.
host platform was shut down for repairs and maintenance.
Pressure data for 115 days prior to the build-up was available
for the analysis. A brief description of the applied
methodology follows:

Based on the available reservoir information, two “Type

Curve Models” were selected for evaluation in each well:

1. Horizontal Well Homogenous Radial Flow.

2. Horizontal Well Two Layers Reservoir with Cross
Flow.

Figure 4: Decline Curve Analysis fit Analogue Field

4 IPTC 11691

• Kv/kh for the Upper and Lower sands 0.47 and 0.15
respectively.

Superposition Plot
• Extrapolated pressure 2714 psi.

Downloaded from http://onepetro.org/IPTCONF/proceedings-pdf/07IPTC/All-07IPTC/IPTC-11691-MS/1802076/iptc-11691-ms.pdf/1 by Middle East Technical University Ankara user on 23 October 2021
Horner Plot Extrapolated pressure was estimated at 2970 psi at
measured depth.

In conclusion for Well #2, the Horizontal Well Two Layers

Reservoir with Cross Flow Model, was considered the better
model based on its agreement with petrophysical
characteristics including Kv/Kh and RFT data that indicated
cross flow between the upper and lower sands.

Figure 5: Pressure and Production data Well #1

Well #1. The Homogeneous Radial Flow model

produced the following results for this well, the highest on the
structure:

Derivative Plot
• Average reservoir permeability 290 mD, close to the
average value of 400 mD from cores.
• Mechanical skin -7.
• Kv/Kh 0.07, which is lower than the range of
permeability anisotropy from logs: 0.1 - 0.4

Superposition Plot Figure 6: Horner Plot Well #1, Flow Period 2

• Extrapolated pressure estimated at 2720 psi at
measured depth Deconvolution techniques were applied to determine initial
pressures4 (Figure 7) and estimate connected volumes, which
MBH analysis for Well #1 using extrapolated pressures from yielded values slightly higher than those from static models
superposition plots and approximate reservoir dimensions was for the core area suggesting the possibility of hydraulic
used to determine the average reservoir pressure estimated at communication with neighbouring areas.
3300 psi.

The Horizontal Well Homogenous Radial Flow model was

selected as more suitable for well #1 as the results are
consistent with core and log data. This model also fits the
seismic stratigraphy interpretation that indicates
homogeneous, high permeability, confined sands.

Well #2. For this well located at the bottom of the

structure draining the Lower sands that exhibit a water oil
contact, the same two models were evaluated. The second
model Horizontal Well Two Layers Reservoir with Cross
Flow provided the following results which were considered
more coherent compared to the ones obtained with the
Homogeneous Radial Flow model:

Derivative Plot
• Average permeability for the upper and lower sands:
158 and 54 mD respectively. Figure 7: Deconvolution Plot Well #2
• Mechanical skin 3.
IPTC 11691 5

Material Balance MBAL program.

The Oil in Place plays a key role in the definition of the field
depletion plan and field life extension options. Based on that During a history match calculation, the tank pressure changes
premise a material balance study was undertaken to review the are calculated using the historic field production rates and the
Oil in Place, utilizing production data and reservoir parameters predefined STOIIP and aquifer parameters that are estimated
determined from the Pressure Transient Test analysis. A multi from the static reservoir and aquifer models. Then multiple

Downloaded from http://onepetro.org/IPTCONF/proceedings-pdf/07IPTC/All-07IPTC/IPTC-11691-MS/1802076/iptc-11691-ms.pdf/1 by Middle East Technical University Ankara user on 23 October 2021
tank Material Balance model with two wells was set-up to iterations are performed until the observed pressures are
account for differences in properties of the Upper and Lower matched. Multiple history matches of field performance data
Sands (Figure 8). were obtained using BP's TDRM, Top Down Reservoir
Modelling process. TDRM was applied in order to investigate
a range of possible STOIIP history matches for the wide range
of subsurface uncertainties3, outlined in Table 4.

Table 4: Reservoir Parameters Uncertainty

Parameter Range
Reservoir Allocation 0 -1
Original Oil in Place, MMSTB 40 - 120
Transmissibility, rb/day*cp/psi 0 – 1000
Aquifer Volume, MMRB 1000 - 200000
Aquifer Permeability, mD 10 -500
Reservoir Width, feet 5000 - 50000

Figure 9 below shows results of the multiple Material Balance

runs, represented by STOIIP Histograms in 13 discrete groups.
The grey values show all the runs and the red values show the
best quality matches. An outcome from this study was the
definition of a range of STOIIP variation that suggested
Figure 8: Schematic Material Balance Model
volumes bigger than initial estimations.
The PVT data from the laboratory was calibrated to reliably
predict pressure and saturation changes which may occur
within the reservoir, the main parameters are in Table 3.

Table 3: General Fluid Properties

Saturation Pressure, psia 1100
Oil Volumetric Factor, RB/STB 1.24
Oil Viscosity, cp 0.6
Gas Gravity 0.77

The material balance study identified aquifer drive as the main

production mechanism that influences the Farragon field
performance. Based on regional studies the aquifer was
represented by Hurst-Van Everdingen-Dake Edge Water drive
with Sealed Boundary, represented with equation 7:

We(t)=10-6∑ UΔPjWD(α (tn-tj), Rd )……………….. (7)

This aquifer model assumes that the speed at which the

pressure disturbance travels through the aquifer and reservoir
system depends on the pressure diffusivity of the system. The
form of the influx depends on the boundary conditions at the
outer aquifer boundary; for a bottom drive aquifer, the influx
is a function of the dimensions and connectivity of the surface
at which the influx is taking place; the pressure diffusivity of
the system is calculated as a function of the length Figure 9: TDRM generated histogram of 866 MBAL runs
perpendicular to the surface. The WD function is the
dimensionless aquifer function and depends on dimensionless Tracking Aquifer and the Water Oil Contact.
time and the size of the aquifer relative to the reservoir. U is Aquifer influx was interpreted by material balance as the main
defined as an aquifer constant; ά is the pressure diffusivity of production mechanism as shown in Figure 10. The water oil
the system and is called the tD constant in the material balance contact interpreted from the logs and pressure data was
6 IPTC 11691

considered a critical factor in the production performance of The model was calibrated with results from classical reservoir
the two producer wells. Material balance was applied with engineering studies and history matched with field history.
water oil relative permeability data from reservoir cores to The main matching parameters were the reservoir pressures
predict the oil-water contact movement, as shown in Figure and the water production profile of Well #2 located near the
11. Later on this interpretation was confirmed by well oil-water contact. In the model the aquifer was represented by
production tests that indicated an uneven water movement in a large water volume under the oil zone and complemented

Downloaded from http://onepetro.org/IPTCONF/proceedings-pdf/07IPTC/All-07IPTC/IPTC-11691-MS/1802076/iptc-11691-ms.pdf/1 by Middle East Technical University Ankara user on 23 October 2021
the lower part of the reservoir and gentle movement toward with 4 analytical Carter Tracy aquifers in directions North,
the top well. South, East and West. The water production history match
procedure was affected mainly by the connections between the
oil zone and the aquifer through the non-reservoir shale
located above the aquifer; the capillary pressures that play a
role in the water saturations distributions and the relative
permeabilities that were tuned by Corey exponents honouring
the end points measured in the laboratory.

The structure was developed from seismic interpretations

updated with revised time depth conversion parameters. In-
place volumes from the simulation model agreed with the
static model and with those from material balance and
transient test interpretations, confirming an increase compared
to the initial estimations. The model was initialized, history
matched and used in predictions to investigate the field
performance under several different development scenarios.

Figure 10: Production Mechanisms Distribution

Well No. 1 - 82 feet above WOC

Well No. 2 - 49 feet above WOC

Water Oil Contact (Feet)

Figure 12: Full Field Model Cross section

Lifting curves for both wells were calibrated with actual data
from well tests and from the downhole pressure gauges.
Figure 13 shows historic and predicted oil rates and water cuts
(scales are excluded intentionally).

QOP WCT WCUTH

QOP FN THP-Opt-3-Cw1-5_short_ WCT FN Optim-3-Cw1-5_ WCT FN THP-Opt-3-Cw1-5_short_ WCUTH FN Prod-test

QOP FN Optim-3-Cw1-5_

14000 1.0

0.9

12000

Figure 11: Tracking the Water Oil Contact 0.8

10000
0.7
OIL PRODUCTION RATE (STB / DAY)

Full Field Dynamic Simulation

WATER CUT (FRACTION)

0.6

A Full field multidimensional dynamic simulation model was

8000

0.5

built to predict Farragon field performance. The model has 6000

0.4

137,500 grid blocks with 55x50x50 cells in the X, Y & Z 0.3

directions. Dimensions are 108 feet in both the X and Y

4000

0.2

directions with 3 feet thickness in the oil zone and 6 feet in the 2000

0.1

water zone. Fluid description in black oil mode comes from 0 0

available PVT data. An oil-water contact was is set at 2557 m

11/05 03/07 08/08 12/09 05/11 09/12 01/14 06/15

TVDss. Initial and bubble point pressures are 3570 psia and
Figure 13: WC% History Match and Predicted Well Performance
1100 psia respectively.
IPTC 11691 7

Future Work ΔPj: Reservoir Pressures at time j

Future plans include studies of alternate exploitation schemes Wd: dimensionless aquifer function
including additional wells and feasible EOR schemes. The α: pressure diffusivity function of permeability, porosity,
application of gel treatments5 to control water encroachment is viscosity and compressibility
another option that can be investigated. Rd: Outer/Inner radius ratio

Downloaded from http://onepetro.org/IPTCONF/proceedings-pdf/07IPTC/All-07IPTC/IPTC-11691-MS/1802076/iptc-11691-ms.pdf/1 by Middle East Technical University Ankara user on 23 October 2021
Results and Conclusions References
1. Classical Reservoir engineering techniques such as 1. Williams, G.J.J. et al.; “Top-Down Reservoir Modelling,” paper
Decline Curve Analysis, Pressure Transient Test SPE 89974 presented at the 2004 SPE Annual Technical
analysis and Material Balance applied in conjunction Conference and Exhibition, Houston.
with the TDRM, Top Down Reservoir Modelling1
2. Fetkovich M.J., Fetkovich E.J. & Fetkovich M.D.: “Useful
process at an early stage of the productive life of the Concepts for Decline Curve Reserves Estimation and Analysis”,
field proved to be very powerful in understanding the paper SPE 28628-PA. February 1996
reservoir and in optimizing reservoir management.
3. Kobani J.: A General Review of the Oil in Place and
2. The application of the above techniques in the Prefeasibility Study of gas Injection in The Farragon Field,
Farragon field allowed the identification of greater Thesis for Petroleum Engineering Master Degree at the IFP
resources than initially estimated and the initiation of performed at BP Exploration Company Aberdeen Scotland
further studies that supported the timely field November 2006.
recovery updates resulting in a significant increase in
4. Levitan, Michael, Crawford Gary & Hardwick Andrew:
field value. “Practical Considerations for Pressure-Rate Deconvolution of
Well-Test Data”, paper SPE 90680 March 2004 presented at the
3. Connected Pore Volume estimated from Transient Annual Technical Conference and Exhibition held in Houston,
Test provided valuable support to update the field Texas, USA, 25-29 September 2004.
volumes and promoted the execution of seismic
interpretation updates with revised time depth 5. Herbas J., Moreno R., Romero MF, Combe D. & Serna A: Gel
conversion parameters. Performance Simulations and Laboratory/Field Studies to
Design Water Conformance Treatments in Eastern Venezuelan
4. Material balance studies provided a valuable insight HPHT Reservoirs. SPE 89398 presented at the SPE/DOE IOR
2004 Symposium Tulsa Oklahoma April 17-21.
into reservoir features that facilitated the history
match of the full field dynamic simulation model.

5. The Decline Curve analysis technique provided

predictions for the recoverable oil volumes consistent
with the volumes obtained from material balance and
dynamic simulations.

Acknowledgments
The Authors will like to thank BP, ENI and Nexen, Joint
Owners of the Farragon field for their permission to write this
paper and for its support in its presentation.

The results and opinions expressed in the paper do not

necessarily reflect the views of the Joint Owners.

Nomenclature
STOIIP: Stock Tank Oil Initially in Place
mD: milliDarcies
Pr: reservoir pressure
Pri: initial reservoir pressure
N: initial oil in place
Np: cumulative produced oil
Qo : Oil rate
Jo : Productivity Index
Pwf : Flowing well pressure
qi: initial pseudosteady-state rate
d: decline exponent
We(t): aquifer model as function of time
U: Aquifer constant function of porosity, thickness and
compressibility