SPE-207611-MS

Downloaded from http://onepetro.org/SPEADIP/proceedings-pdf/21ADIP/3-21ADIP/D031S085R003/2540576/spe-207611-ms.pdf/1 by University of Petroleum & Energy Studies - India user on 18 February 2024
Uncertainty and Sensitivity Analysis on Clastic Systems: Integrated
Workflows for Understanding and Managing the Risk, A Case Study in
Tuwayil Formation UAE

Mahmud Ebeid, Humberto Parra, and Dipankar Ghosh, ADNOC; Jeonggil Kang and Kwangwon Seo, KADOC

Copyright 2021, Society of Petroleum Engineers

This paper was prepared for presentation at the Abu Dhabi International Petroleum Exhibition & Conference to be held in Abu Dhabi, UAE, 15 – 18 November 2021.
The official proceedings were published online on 9 December 2021.

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
of the paper have not been reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material does not necessarily reflect
any position of the Society of Petroleum Engineers, its officers, or members. Electronic reproduction, distribution, or storage of any part of this paper without the written
consent of the Society of Petroleum Engineers 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 contain conspicuous acknowledgment of SPE copyright.

Abstract
This study has been done on Late Cretaceous tight clastic reservoir located south west of Abu Dhabi city
with the border with Saudi Arabia. The field was discovered in the 1960s and a few wells were drilled
subsequently. The Tuwyail clastic reservoir is characterized as tight with average permeability below 1 mD.
The trap is identified as structural trap as north south anticline with gentle dip in both sides. Total of six
wells were drilled targeting Tuwyail reservoir which part is of Wasia group.
However, assessing potential of this accumulation poses a great challenge not only in terms of
understanding of the depositional system which still unknown before but also in terms of quality of the
legacy data like well data that impact the modeling studies. The aim of this paper is to provide an insight on
integrated workflows for assessing the different uncertainties on clastic systems with limited data, focused
on the most important sensitivities parameters impacting the oil in place, like reservoir proportions, free
water level [FWL] and lateral distributions of the sedimentary elements within the area of interest which
playing a big rules in future developing of the field.
Before moving to full field development a full uncertainty and sensitivity analyses were conducted for
the Tuwayil reservoir to find the highly uncertain parameters that impacting the future development of the
reservoir, in the same time the main challenges is the limited data with low quality as the wells had been
drilled in 60s with limited technology at that time and the core data were left in a bad conditions since the
filed was left behind.

Introduction
This study has been done on Late Cretaceous tight clastic reservoir located south west of Abu Dhabi city
with the border with Saudi Arabia. The field was discovered in the 1960s and a few wells were drilled
subsequently. The Tuwyail clastic reservoir is characterized as tight with average permeability below 1 mD.
The trap is identified as structural trap as north south anticline with gentle dip in both sides. Total of six
wells were drilled targeting Tuwyail reservoir which part is of Wasia group.
However, assessing potential of this accumulation poses a great challenge not only in terms of
understanding of the depositional system which still unknown before but also in terms of quality of the
2 SPE-207611-MS

legacy data like well data that impact the modeling studies. The aim of this paper is to provide an insight on
integrated workflows for assessing the different uncertainties on clastic systems with limited data, focused
on the most important sensitivities parameters impacting the oil in place, like reservoir proportions, free
water level [FWL] and lateral distributions of the sedimentary elements within the area of interest which

Downloaded from http://onepetro.org/SPEADIP/proceedings-pdf/21ADIP/3-21ADIP/D031S085R003/2540576/spe-207611-ms.pdf/1 by University of Petroleum & Energy Studies - India user on 18 February 2024
playing a big rules in future developing of the field.
Before moving to full field development a full uncertainty and sensitivity analyses were conducted for
the Tuwayil reservoir to find the highly uncertain parameters that impacting the future development of the
reservoir, in the same time the main challenges is the limited data with low quality as the wells had been
drilled in 60s with limited technology at that time and the core data were left in a bad conditions since the
filed was left behind.
As a results of that study, a complete first pass static model was built by integrated all available data
incudes and not limited to seismic and well data were used, then a comprehensive uncertainty and sensitivity
analysis were conducted using more than 10000 runs and a list of uncertainties were defined based on its
impacts on the hydrocarbon in place and future mitigation plan were put that include drill more wells and
make the data gathering plan to help to reduce that risks.

Static modeling
Data avalbility and data quality
All the wells had been drilled on 60s with limited technology that time, 6 wells are existing in the study
area which include the openhole logs, no deviation survey and limited core data to the reservoir intervals.
Old 2 D seismic data are available but 3D seismic were conducted in 2015 with a good quality that give
a clear picture of the structure and the reservoir can be mapped from the seismic refractors includes faults
and horizons. All the wells have final geological well report as scanned copy that help on understanding
the history of the well drilling operations.
Data quality is the big issue in the study area as the wells were drilled long time ago and the existing core
is in a very bad condition and some of intervals look like dust, although the wells are vertical the limited
deviation surveys is another challenge for depth uncertainty for picking the right depth of the formation tops
depth. Unfortunately some of the scanned copy of the well reports cannot be read and the original copies
were lost. In some wells the density logs are not realistic due to the whole condition and this imposing a
challenges to get correct porosity log estimations which lead to uncertainties in the permeability and water
saturation calculations. All of this has a big impact on the hydrocarbon volume in place that is the key for
the economy of the field.
Since Modelling is to be considered as the blender into which all Geosciences and Engineering data
are integrated, data QC and enhancement processes are the first step to be taken for all data; Seismic
through Production Testing. Analog Data from nearby fields is an important, if temporary aid to close the
gap. For Structural Modelling, good seismic horizon interpretation is the first target to accomplish. Further
adjustment and fine-tuning are applicable as additional control points become available. Algorithm options
for populating data in Static Models can focused toward achieving the best match possible around the
wellbore for porosity, permeability, and saturation.
Straigraphic model was created, the main criteria was used the open hole well logs specially the gamma
ray (GR), Neutron/Density logs to identify different cycles of sea level change during carbonates deposited
in the field.
In fig [2] it shows example of the existing core however the core doesn’t slapped but it is in a bad
conditions as some parts cannot take a plugs for analyses and only core descriptions can be done also
this added uncertainty in the core depths, the example of the logs showed bad hole coditions and wrong
estimations of the porosity logs, scanned copy of the final well report shows missed information as well
SPE-207611-MS 3

Downloaded from http://onepetro.org/SPEADIP/proceedings-pdf/21ADIP/3-21ADIP/D031S085R003/2540576/spe-207611-ms.pdf/1 by University of Petroleum & Energy Studies - India user on 18 February 2024
Figure (1)—UAE onshore and offshore gas and oil fields location map

Figure (2)—left is the existing core for Tuwayil S.S, Right is the log quality

Conceptual depositional model

The Tuwayil Formation comprises part of the Upper Wasia Group, which is divided into four formations
(the Shilaif, Mishrif, Tuwayil and Ruwaydha Formations in ascending order). Initial geological and seismic
analyses by Azzam (1994) and Azzam and Taher (1993), suggest the succession is divided into three
main sequences (I: Mishrif/Shilaif; II: Tuwayil; and III: Ruwaydha). The Mishrif/Shilaif has subsequently
been further divided, with the Tuwayil and Ruwaydha Formations incorporated into a single sequence
(Al-Zaabi et al. 2010). However, both sequence stratigraphic approaches invoke eastward progradation of
the monotonous peloidal grainstone shoals and shelf-edge complexes of the Mishrif into time-equivalent
argillaceous and bituminous basinal wackestones of the Shilaif Formation during high stand conditions,
together with development of a relatively steep, N-S orientated shelf edge boundary (Fig. 3). A subsequent
fall in sea level during the Turonian exposed the Mishrif shelf and resulted in deposition of the Tuwayil
Formation as a low-stand wedge within the basinal areas of the former Mishrif/Shilaif complex. Initial
Tuwayil deposition comprised shallow water carbonates, but an influx of siliclastic sediments in response
to uplift and erosion of the Arabian Shield hinterland to the south and west resulted in final filling of
the accommodation space in the basinal areas (Al-Zaabi et al. 2010). A key implication that should be
considered based on this regional model is the potential for clastic-filled incised valleys or caverns within the
Mishrif shelf area to the west of the study area. Finally, carbonate facies of the Ruwaydha Formation were
deposited during high stand conditions following a major, rapid sea level rise; limited thickness variations
4 SPE-207611-MS

of this formation across the area suggest that the topographical low was largely filled during Tuwayil times
(Taher et al. 2010; Al Zaabi et al. 2010).

Downloaded from http://onepetro.org/SPEADIP/proceedings-pdf/21ADIP/3-21ADIP/D031S085R003/2540576/spe-207611-ms.pdf/1 by University of Petroleum & Energy Studies - India user on 18 February 2024
Figure (3)—Tuawyail conceptual deposition model using esxisting core and openhole logs

Figure (4)—property model summery, including Facies, porosity, permeability and water saturation
SPE-207611-MS 5

Downloaded from http://onepetro.org/SPEADIP/proceedings-pdf/21ADIP/3-21ADIP/D031S085R003/2540576/spe-207611-ms.pdf/1 by University of Petroleum & Energy Studies - India user on 18 February 2024
Figure (5)—different scenarios for the Facies model distributions

Figure 6—tornado chart showing that the structure and saturations have big impact on the STOIP uncertainties

A hierarchical approach to core descriptions, together with the integration of new RCA data and
petrographical analyses from the historic core dataset, and a detailed core-to-log calibration facilitating log-
based evaluation of the uncored intervals, reveals a more complex depositional and reservoir model.
Deposition is considered to have occurred within a range of paralic to offshore depositional settings:
delta plain (ie. fluvially-dominated and distributary-dominated), delta-front to prodelta, open marine clastic
6 SPE-207611-MS

and carbonate settings); which, when placed in context of a revised high-resolution sequence stratigraphic
framework, appear to vary in relative extent and distribution through time. Whilst geometric mean
permeabilities are low (<0.1mD), permeabilities range up to 800mD, with porosities up to 28%, and there
appears to be a strong depositional control on these properties, specifically grain size and ductile content,

Downloaded from http://onepetro.org/SPEADIP/proceedings-pdf/21ADIP/3-21ADIP/D031S085R003/2540576/spe-207611-ms.pdf/1 by University of Petroleum & Energy Studies - India user on 18 February 2024
with the better properties hosted within the coarser, cleaner deposits, typically found within the fluvial and
distributive channel systems.
The Tuwayil has been divided into four intervals (in ascending order): A: follows incision into the
underlying Mishrif shelf during a major low-order sea level fall and comprises of A:N/NE-directed
progradation of poor quality low-stand deltas into the ‘Mishrif-Shilaif’ basinal low; B: a further sea level
fall (interpreted sequence boundary in core) results in a high-frequency basin-ward shift in clastic facies
to coarser, better quality potentially laterally well-connected, fluvial-deltaplain channel and sandsheet
deposits; C: a rise in sea level results in a flood back of this fluvial system and development of a more
heterogeneous, variable quality distributive channel-mouthbar complex, with interbedded poor quality
swamps, interdistributary bay mudrocks and lagoonal carbonates; D: finally, transgressive erosion and
development of poor quality, marine shore face and offshore conditions prior to the Ruwaydha carbonates.

Property modeling
The sequence of property model workflow started with Facies, porosity, permeability and water saturations
models include saturation height function approach keeping the scope for any future update open if new
data has acquired Facies model was built with the support of the geological conceptual deposition model
based on the existing core data analyses using hierarchical and Truncated Gaussian Simulation algorithm
[TGS], the geometries of the depositional model implemented using the width and thickness of fluvial
channel bodies and valley fills in the geological record : a literature compilation and classification, Journal
of Sedimentary Research 2006.
Porosity is one of the first and most important components in property modeling, because distribution of
Porosity controls the distribution of pore volume, which ultimately contributes to STOIIP and determines
production by interconnectivity matrix of reservoir. To get better and accurate porosity model, Porosity
data for vertical well were loaded into the project and quality checked, unrealistic and up normal values
eliminated. After scaled up the porosity data. Analysis done using Porosity logs for one vertical well and
analogue data for the horizontal variogram. Porosity values for cells penetrated by a well honor the well
data. Porosity model was created using Gaussian random function simulation in PETREL conditioned to
the Facies model. Porosity ranges and variogram analyses were different for each Facies code,
Due to absence of the Permeability logs for all the wells a Porosity permeability relationship for each
Facies code were created using a cross plot function that coming from the core data [RCA], to build the
permeability in the 3D Collocated co-kriging algorithm used to build the permeability in the 3D with
the support of the cross plot and porosity/Facies as secondary trend to control the distributions of the
permeability.
Water Saturation model is one of the most important and challenged components in property modeling,
because it has contributes to STOIIP and producibility of reservoir. To get better and accurate saturation
model different SHFM approach used to calculate water saturation model for all reservoirs.
Total of 19 MICP plug data for existing core are available and cover most of the Facies codes, the related
MICP curve per each facies were used directly with the function of the height above the FWL to calculate
the water saturation and calibrated with SW log and well testing results, the more weight going to the testing
results as less acurety in the SW log.
The base case of the free water level were estimated based on the SW log and well testing results, the
absence of the pressure points lead to big uncertainty in define the free water level in the study area.
SPE-207611-MS 7

Uncertainty analysis
Due to the limited and quality of the data the uncertainty is huge in the study area including and not limited
to Structure, Facies, porosity, permeability, water saturations, free water level etc. a comprehensive and
detailed uncertainty and sensitivity analyses were conducting for Tuwyail reservoir as shown in table [1]

Downloaded from http://onepetro.org/SPEADIP/proceedings-pdf/21ADIP/3-21ADIP/D031S085R003/2540576/spe-207611-ms.pdf/1 by University of Petroleum & Energy Studies - India user on 18 February 2024
three scenarios for the facies model, High [More sand and more reservoir], middle [Base case scenario] and
low sand [less reservoir], all scenarios are respect the well data means the percentage of the sand at the wells
are kept no changes only away from the well had been adjusted based on the scenario that used.

Table (1)—uncertainty parameters ranges used in the sensitivity and uncertainty workflow

Table 2—uncertainties and it is migration plan at the static model level.

8 SPE-207611-MS

The structures uncertainty parameters were identified using the different velocity model errors and used
50 feet as standers deviations which will be added to the error surface for each run.
The free water level ranges were identified using the well test intervals that the value cannot exceed that
that level of the oil intervals and cannot go below the structure spill pint [last closed contour] Saturation

Downloaded from http://onepetro.org/SPEADIP/proceedings-pdf/21ADIP/3-21ADIP/D031S085R003/2540576/spe-207611-ms.pdf/1 by University of Petroleum & Energy Studies - India user on 18 February 2024
height function ranges used ±0.15 based on the difference between the MICP data and SW log, for the
porosity different seed number were used for each case to come up with different realizations.
The uncertainty in the facies geometry were used the ranges based on the analogs data and using different
seed number for each cases.
The most critical parameter is the Facies model which has a big impact on the hydrocarbon volumes for
example three scenarios were created based on the geometry of the sand, low, middle and high cases as the
below figure, more sand will give high volume of the hydrocarbons.
After run more than 10000 runs for sensitivity analyses the tornado chart showing that the Facies, porosity,
water saturations, free water level and structure are the most uncertain parameters that impact the volume in
place and I mitigations plan is required to decrease the risk of that parameters, Facies is the most important
parameter which mean good Facies model will lead to high certainty of the oil in place and it is affected
the whole development of the field.
As mitigation plan for the structural variance a velocity model update after appraisal well is required and
it will reduced the uncertainty between well prognosis before dill the well and actual well tops after drilling.
Drill more wells and gather more quality data that can help in understand the reservoir heterogeneity
and reservoir fluids.
To reduce the uncertainty of the lateral reservoir heterogeneity a detailed core analyses is required with
detailed facies study include well data, cores, testing is required to reduce the uncertainty in the free water
level and water saturation estimations.

Conclusion and recommendations

It is important to integrate all the available data for the target reservoir regardless the quality of that data and
identify the risk and the uncertainties parameters before moving to the full field development or maybe can
help in identify the appraisal plan for better filed economic evaluations In the study area Tuwyail reservoir
static model first pass was created using historical well data and RCA for the existing core data which helps
in building the depositional conceptual model that control the reservoir geometry. The historical well data
which have very huge uncertainty especially due to quality of the core which considered during the study.
Tuwyail S.S is clastic reservoir and big challenge is to understand the complex depositional system which
were controlling the geometry of the reservoir that will impact the strategy for appraisal and develop that
reservoir.

References
WIDTH AND THICKNESS OF FLUVIAL CHANNEL BODIES AND VALLEY FILLS IN THE GEOLOGICAL
RECORD: A LITERATURE COMPILATION AND CLASSIFICATION, Journal of Sedimentary Research, 2006, v.
76, 731–770 Research Article,DOI: 10.2110/jsr.2006.060
Reservoir Model Design ‘A Practitioner’s Guide, Philip Ringrose & Mark Bentley, 2015
Al-Jenaibi, Faisal Salah, Lacroix, Sebastien, Meziani, Said, ADNOC and Bughio, Shams Udin, Ben Salem, Ryadh,
Monfared, Hashem, SPT Group. Fast-Track Modeling Approach: An Alternative Workflow For Reservoir Simulation
Modeling. Presented at ADIPEC, Abu Dhabi, 16–18 September 2013. SPE-166004-MS. https://www.onepetro.org/
download/conference-paper/SPE-166004-MS?id=conference-paper%2FSPE-166004-MS
Developing the Obaiyed Tight Gas/Condensate field, Egypt B A case study
Eilard Hindericus Hoogerduijn Strating; Sameh Amin; Rady Abdel Samiee; Mahmoud Abu Shanab; Mahmoud Ebied;
Alaa Abdel Samaie; Nasser Mohamed Ali Paper presented at the SPE Europec/EAGE Annual Conference, Madrid,
Spain, June 2005. Paper Number: SPE-94106-MS
SPE-207611-MS 9

AZZAM, I. 1994. Sequence stratigraphy of Middle Cretaceous silicilastic sandstone (Tuwayil Formation) in West Abu
Dhabi: A model Approach to Oil Exploration. In: AL HUSSEINI, M.I. (ed) GEO 1994 - The Middle East Petroleum
Geosciences, 1, 155 - 165.
AZZAM, I. AND TAHER, A. 1993. Sequence stratigraphy and source rock potential of Middle Cretaceous (upper Wasia
Group) in West Abu Dhabi. Society of Petroleum Engineers Middle East Techncial Conference and Exhibition,

Downloaded from http://onepetro.org/SPEADIP/proceedings-pdf/21ADIP/3-21ADIP/D031S085R003/2540576/spe-207611-ms.pdf/1 by University of Petroleum & Energy Studies - India user on 18 February 2024
Bahrain, 475 - 487.
Al ZAABI, M. TAHER, A., AZZAM, I, WITTE, J. 2010. Geological overview of the Middle Cretaceous Mishrif
Formation in Abu Dhabi. International Petroleum Exhibition and Conference, Abu Dhabi, Society of Petroleum
Engineers 137894.