Sie sind auf Seite 1von 13

SPE 158268

Advanced Completion and Fracturing Techniques in Tight Oil Reservoirs in


Ordos Basin: A Workflow to Maximize Well Potential
Hai Liu, Yin Luo, Schlumberger; Xianwen Li, Yonggao Xu, Kewen Yang, Lijun Mu, Wen Zhao, Shuxun Zhou,
PetroChina Changqing Oilfield Company

Copyright 2012, Society of Petroleum Engineers


This paper was prepared for presentation at the SPE Annual Technical Conference and Exhibition held in San Antonio, Texas, USA, 8-10 October 2012.
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
In recent years, great efforts have been focused on tight oil reservoirs in Ordos Basin, which is located in North Central
China. Although historical field development in these tight oil reservoirs was still economical, average production after
hydraulic fracturing is quite low and most of wells produce just at the margins. In order to understand production potential
through appropriate completion and fracturing techniques, a pilot project was initiated recently on two horizontal wells
drilled in parallel, with three vertical wells placed between two horizontal wells along the lateral for real time fracturing
monitoring.
This paper describes how the treatment design was optimized based on detailed formation evaluation and fracture simulation
using a 3D unconventional fracturing design workflow. It presents real time fracture mapping results through a dual
monitoring well setup on total 26 stages of fracturing treatments, and also illustrates how the results were utilized to adjust
and optimize treatment design during and between treatment stages. The paper demonstrates the application of a unique
Unconventional Fracture Model (UFM) that was developed for design and evaluation hydraulic fracturing treatment in
unconventional reservoirs. The model incorporates predefined natural fracture patterns and interaction criteria for hydraulic
and natural fractures. The paper also discusses the outcome of simultaneous treatments tested in four stages compared to
other stages pumped in sequences.
Initial production tests on both wells show 124.5 m3/d and 103.2 m3/d respectively, which are significantly higher than all the
wells completed in tight oil reservoirs in the basin, of which the production varies from 5-8 m3/d in vertical wells to 32 m3/d
in horizontal wells on average. In-depth reservoir understanding, advanced design and evaluation workflow, and appropriate
completion and fracturing technique are the key of this success, which has set a milestone for developing such tight oil
reservoirs in Ordos Basin.
Introduction
Located in North Central China, Ordos Basin has a series of tight oil reservoirs. Each reservoir normally consists of a group
of stacked sand/shale deposits. Among these deposits, the majority of pay zones are Triassic Chang-6 and Chang-7 sections
of Yanchang Formation. Compared with Chang-6, Chang-7 is considered as more typical tight oil reservoir and wide
distributed in the basin, with permeability ranging from 0.05 md to 0.3 md and up to approximately 30m of net pay from
several sub-layers. In recent years, great efforts have been focused on these tight oil reservoirs. Although historical field
development in these tight oil reservoirs was still economical, average production after hydraulic fracturing is quite low and
most of wells produce just at the margins. Insufficient fracture surface area induced by traditional fracture completion
techniques has been identified as the major reason of low production. However, even by increasing the fracture treatment size
and proppant mass pumped into the formation, no obvious production incremental gains achieved with conventional fracture
design approaches. Further understanding of the fracture propagation behavior and proppant distribution mechanism is
necessary to break through the production barrier.
Therefore, in order to understand production potential through appropriate completion and fracturing techniques, a pilot

SPE 158268

project was initiated in the south-west of the basin recently on two horizontal wells drilled in parallel, with three vertical
wells placed between two horizontal wells along the lateral for real time fracturing monitoring. These two horizontal laterals
were drilled orthogonal to maximum horizontal stress orientation (NE75o) to enable propagating transverse hydraulic
fractures across the laterals. To have more precise microseismic mapping result, dual monitoring has been setup for the
project by using two out of the three monitoring wells at the same time. To keep the minimum distance between the source
events to the detectors, the VSI string in the well near to the toe were transferred to the one near to the heel when the latter
one was closer to the fractures after completing half of the lateral. The results from the two monitoring wells can be tagged
by synchronized GPS time and can be further analyzed and combined after the treatment. Figure 1 and Figure 2 show
geographic location and the well pattern of this pilot project, respectively.

Figure 1: Geographic locationof this tight oil project in Ordos Basin

Figure 2: Well pattern for the pilot project

The horizontal laterals were drilled in lower Chang-7 formation, which is very find sandstone with average permeability 0.24
md and average porosity 10.5%. The sand body normally consists of approximately 40-70% of quartz and 20-30% of
feldspar, filled with mica and illite. And both the primary porosity induced by intergranular spaces between sand grains and
the secondary porosity of the dissolved holes in feldspar are considered of same importance for the hydrocarbon storage, as
shown in Figure 3. Natural fractures are wide distributed in the whole basin; however, their nature and characteristics vary
from formation to formation and from block to block, resulting from various natural fracture mechanism and several changes
in tectonic movement during the ancient time. And currently the basin is still flat with the horizontal stress anisotropy is
approximately 3-7MPa.

Figure 3: Photomicrographics showing porosity for cores from offset wells (Left) and
XRD graphic for core from one of the target wells (Right)

These target wells are the first two horizontal wells in the reservoir, while several vertical wells have been completed. These
vertical wells can initially produce from 5 m3/d to 28.5 m3/d after conventional fracturing treatment with crosslinked gel and
approximately 50-120 ton of proppant placed. The production for these wells declined very quickly in three months to less

SPE 158268

than 3 m3/d and can hardly produce economically afterwards. It is also important to notice that the effective permeability can
be as low as 0.01 md from previous post-closure pressure decline analysis. Extremely low permeability and insufficient
fracture surface area for oil production are considered as main reason for low production.
Unconventional Reservoir to Fracture Simulation Workflow
Nowadays, rapid progress on completion and fracturing techniques enable effective development tight reservoirs, but more
fundamentally, to maximize the well potential in such tight reservoirs, maximizing the reservoir-fracture surface contact area
has been proven to be the underlying principle. However, to achieve this objective, it is necessary to have accurate
understanding about how the fracture is penetrating into the reservoir, including its complexity, effective length, fracture
containment in the pay zone and fracture conductivity for hydrocarbon production. Furthermore, the reservoir contact area
can be significantly improved with multiple transverse fractures for horizontal wells. However, even with transverse
fractures, the well potential can only be maximized with contribution from all pay intervals and proper production
distribution along the lateral.
Instead of using traditional planar fracture models based on 1D geomechanic model, a workflow based on 3D reservoir
model and 3D fracture simulation has been introduced to have better understanding on fracture propagation and thus optimize
fracture design. Starting from building 3D reservoir model from various data source, such as seismic study, cores, logs and
previous fracturing treatment data, reservoir petrophysical properties and geomechanic properties are characterized in details.
Then the staging strategy and fracturing treatment design for each stage can be optimized based on the 3D earth model
coupled with production prediction using 3D reservoir model. During the execution phase, with the real-time measurement
and interpretation from treating pressure and microseismic fracture mapping, 3D models can be calibrated and treatment
design can be optimized based on refined model. Then on-the-fly treatment adjustment can be performed according to the
real-time observation and optimized fracturing design. After the treatment, post treatment evaluation then can be done for
each stage or each job to further calibrate the models and improve the knowledge of the reservoir and its fracturing behaviors.
Figure 4 shows the unconventional workflow.

Figure 4: Unconventional Workflow Schematic

Though the underlying idea of Design-Execution-Evaluation cycle in this workflow has been introduced to the industry for
decades, this workflow is distinguished from previous works by its advance on simulating complex behavior via
unconventional models in tight reservoirs. Its features are highlighted as below:
1) 3D earth model capturing details in the reservoir. This feature has been utilizing for geologic study and reservoir
development, but now the fracture design based on complex fracture models coupled with production prediction
based on 3D reservoir model enable the fully usage of the data we feed into the complicated model.
2) Model calibration plays important roles in this workflow. Though the 3D earth models and complex fracture models
can provide details in real world dimensions with deliberate data inputs, the model accuracy needs to be improved
by fracture and production behavior calibration.
3) Based on a group of software, these workflow steps are integrated into one work platform and can be implemented
routinely for single well or multiple wells design. This feature is important to materialize the workflow, especially
when it requires some cycles of model calibration. And these calibration cycles can even be done during the fracture
treatment.

SPE 158268

Ideally, to fully power the workflow, plenty of data from various sources is necessary to capture the detailed variations of all
dimensions, and very precise modeling for fracture and production can be achieved. However, the workflow can also be
tailored according to the availability of the data, and the complexity of the 3D models should be properly selected to enable
capturing adequate details with acceptable time consumption, depending on different user scenarios.
Building 3D Geologic and 3D MEM Model
3D Geologic model used in this workflow can be derivated from seismic study with detailed structural information. They also
can come from the well correlation for structural variation and properties variation among these wells. Though not often
included in the workflows, the 3D stress field can be calculated, using the wellbore measurements as calibration points, to
ensure that both the properties and the structure are accounted for correctly in the stress field. For simpler and common case
of fracturing modeling for horizontal wells, a structure model with properties from one reference well is used.
In this pilot project, well correlations were established through the three vertical wells placed between the two horizontal
target wells. By looking at the features from several well logs, namely Gamma Ray, Neutron Porosity, Bulk Density and
Resistivity, a series of well tops, including the top and bottom of the target Chang-7 formation, were identified to reflect the
structural variation along the laterals (Figure 5). After building structural model, 3D zone model were built according to the
surfaces extracted from well correlation (Figure 7) and properties for each zone were filled based on the properties of the
monitoring wells along the laterals (Figure 6).

Figure 5: Well correlation of 3 monitoring wells between the laterals

Figure 7: Surfaces extracted from well correlation,


showing structural variation

Figure 6: 3D zone model with minimum horizontal stress

SPE 158268

The complexity of the zone model was properly selected as approximately 50x50x2m grids to enable capturing enough
variations to different dimensions while avoiding unnecessary time consumption on numerical simulation.
3D Unconventional Fracture Simulation
Traditional fracture simulation models, such as PKN, KGD, P3D, MLF_P3D, can normally utilize the reservoir properties in
vertical dimension, extending the near wellbore properties to the far end of the field. Unconventional Fracture Model (UFM)
can utilize fully three dimension properties and thus require a more complete description of the stress field than prior simple
models, [C. Cipolla et al. 2011].
The complex fracture network is largely influenced by the interaction between hydraulic fracture and the pre-existing natural
fractures. Stress anisotropy, natural fractures, and interfacial friction are critical parameters in creating fracture network
complexity, [H. Gu et al. 2011, X. Weng et al. 2011]. Some criterion has been developed to determine whether a fracture
crosses a frictional interface at various angles. Based on the criterion, UFM can numerically simulate the HF-NF interaction,
as well as the stress shadow effect from existing or ongoing offset fractures. Proppant transportation in complex fractures is
also considered.
The tectonic setting of Ordos Basin is monoclinal structure, with only approximately 1 degree of stratigraphic dip, which is
not normally deemed good conditions for natural fracture growth. However, many vertical fractures and high-angle fractures
have been found through core analysis and imaging logging interpretation. According to further geologic investigation, the
major natural fracture system is vertical diagonal fractures distributed at NNE and NW caused by the ancient tectonic
movement, and some minor natural fracture systems at various orientations are also observed. And natural fractures in southwest part of the basin, compared to those of the north and middle part of the basin, show less regular pattern and more uneven
distribution, which leads to many unexpected early water breakthrough cases for water injection programs and unexpected
well interferences during fracturing treatments.
To capture the basic characteristics and quantitative modeling of natural fracture system in this area, Discrete Fracture
Network (DFN) approach is used based on the geologic knowledge. Due to the lack of seismic study on natural fractures for
this particular area, the DFN model for natural fractures had not been established for this reservoir before. Therefore, a
manual DFN model was initially created to simulate the natural fractures. This DFN model (Figure 8) consists of two sets of
natural fractures majorly at the maximum horizontal stress orientation (NE75o) and the minimum horizontal stress orientation
(NW15o), respectively, with random angle variation within 5o, which is aligned with geologic investigation. The fracture
length and fracture spacing are ranging from 30 m to 70 m by assumption.

Figure 8: Initial DFN for natural fractures

Figure 9: Initial fracture simulation

With the natural fracture DFN model and 3D geomechanic model, the UFM simulation result (Figure 9) suggested basic
planar fractures with some simple fracture networks by opening natural fractures and creating new fracture branches during
the main fracture propagation. Several cases with various kinds of fracturing fluids have been tried to improve the fracture
complexity, however, the result showed the fracture complexity are dominated by natural fracture distribution and horizontal
stress anisotropy in this case.

SPE 158268

Fracture treatment design


To improve the economic potential, a low-cost tubing-conveyed packer technique was selected for staging by isolating one
stages treatment from the previous stages. The packer will be set by pumping at enough rate through the tubing. And to
reduce the risk of proppant settling in the annulus, which can get the packer stuck if its not properly unset as expected,
proppants were only pumped though the tubing. Based on this downhole staging tool, unconventional fracturing treatments
were designed for every stage, with the following features:
1.

2.

3.

4.

5.

2.5-3 m3/min through the tubing and 2.5-4.5 m3/min through the annulus were pumped during the job execution,
giving approximately 3 m3/min per perforation cluster. And annulus rates were slowed down at the last two proppant
stages to increase bottomhole proppant concentration and thus improve fracture-wellbore connectivity.
Immediate flowback was done for every stage. Approximately 40-60% fracturing fluids were gathered and recycled
for next stage. More than 4000 m3 of flowback fluids were pumped for each well, which took half of the whole
water supply. Massive fracturing treatments were previously impeded by water supply efficiency and sometimes
even cannot be performed due to water shortage or pad area limitation.
Flowback fluids were pumped though annulus during the treatment, and in case that there were not enough flowback
fluids for the first one or two stages, linear gel were pumped instead. In the tubing, hybrid treatments using linear gel
as pad and crosslinked gel for propped stages were pumped. Its important to notice that when the linear gel and the
weak crosslinked gel were mixed with flowback fluids including breaker residues, the fluids would be diluted and
behave like broken gel with apparent viscosity around 5-10 mPa.s.
The highest surface and bottomhole proppant concentration were approximately 600 kg/m3 and 360 kg/m3,
respectively. The average bottomhole proppant concentration was approximately 180 kg/m3 and the ratio of the total
proppant mass pumped and the total fluid volume pumped was approximately 120 kg/m3 (i.e 1 lbm/gal).
Initially continuous proppant stages were designed to assure evenly distributed proppants and adequate fracture
conductivity for oil production. As the overall fracture geometry was not as complex as expected, the proppant
stages were split into two phases by a second pad, and in this way more branches can be achieved with fewer but
still enough fracture conductivity. So two-phase pumping schedule was selected as the base case, but real-time
adjustment will be done according to hydraulic fracture geometry from microseismic mapping result.

YP1 [MD ]
MD

0 .0 0 G R

2 5 0.0 0

0.0 0TXSG _f rom _TXSP

Proposed Comp letion 1

R QD
_ i sc e
r t e
l o
g _ P ro p
o s e do
C mC
p Q
_
e
l D
ti sio cn re
1 te o
l g _
Pr o
p o se d C C
o mp e
lo sito te
n 1Q
u a il ty S o
c re _ P ro p
o s e d o
C mp e
l t o
i n1

0 .30

P ro po se d C om p l et i on 1 3 S t age s

St age 18
St age 13

Bad

BB
St age 17
GG

Good

Good

2400

St age 12
St age 16

St age 15
GG

Good

Good

2600

St age 14

St age 11

St age 13

Good

GG

St age 12

Good

GG

S tage 11

Good

GG

Good

Good

2800

3000

St age 10

GG

St age 10

St age 9

St age 9

Good
Good

For completion quality, the main concern is to find similar stress for
several perforations in the same stage to facilitate multiple fractures
development and thus more production contribution along the whole
lateral.

GG

Good

2200

Bad

For reservoir quality, petrophysical properties, such as TOC, GIP, kerogen,


porosity, permeability, should be looked at to identify the place where
more hydrocarbon fluids can be produced. As the formation properties
might vary from the near wellbore area to the far end, finding the good
reservoir quality intervals according to near wellbore logs are not enough.
Far end formation evaluation from well correlation or seismic study will be
helpful for further determination of where to put fractures to reach the
good reservoir spots.

C a si ng c ol l ar

1 :5 82 2

Good

Optimize fracture placement along the lateral


To select proper staging strategy and perforation positions, both reservoir
quality and completion quality need to be considered. Ideally, sweet spots
which have both good reservoir quality and good completion quality
should be chose as target intervals for fracture placement.

Good

GG

St age 8
St age 8

Good
St age 7

St age 7

St age 6

St age 6

St age 5

St age 5

GG

Good

Good

3200

GG

Good

Having both reservoir quality and completion quality, proper fracture


spacing arrangements are then considered to cover sweet points and to
achieve even production distribution along the lateral.

Good

GG

The above processes are used to identify stage and perforation positions in
this pilot project. For reservoir quality, the clay content, permeability,
porosity, gas logging while drilling is used to qualify good reservoir
intervals. Cut off (see TABLE 1) was set for each log and a combined good
reservoir quality required the satisfaction for every item. And the in-situ
stress, cement bond and casing collars are utilized to identify good
completion intervals (see Figure 10).
Figure 10: Staging and perforations for one
3400

BB

Bad

Bad

horizontal well

GG

Good

3652

Good

3600

St age 4

St age 4

St age 3

St age 3

St age 2

St age 2

St age 1

St age 1

SPE 158268

TABLE 1 -- CUTOFF FOR GOOD RESERVOIR QUALITY


Clay Content
< 20%
Permeability
> 0.18 md
Porosity
> 8%
Total Hydrocarbon Content (Gas Logging)
> 4%
As the result, initially 18 stages were identified for each well. During the execution phase, due to the overlap of the
microseismic mapping events for the first 5-6 stages and rather long fracture length observed, the number of stages was
reduced to 13 stages to give more appropriate fracture spacing and more focus on good reservoir quality intervals.
Real-time Optimization
Microseismic fracture mapping techniques currently are widely used to measure hydraulic fracture geometry, especially in
unconventional reservoirs. With the recent advance of the microseismic monitoring and interpretation techniques, real-time
mapping result can be delivered to stimulation engineer and on-the-fly decision on treatment adjustment can be performed
accordingly. Currently in this workflow, by integrating with 3D geologic settings and unconventional fracture simulation,
appropriate actions can be taken based on better knowledge on fracture propagation environment and mechanism.
In this pilot project, to maximize the effective fracture surface contact with the reservoir, real-time fracture treatment
optimization was done for each stage. Based on the real-time microseismic hydraulic fracture mapping result, along with the
unconventional fracture simulation during the treatment, the fracture geometry for each stage was tailored to achieve
adequate propped fracture length or to put effort for increasing fracture complexity. During the execution phase, the
observation from microseismic hydraulic mapping shows limited fracture complexity, with several short branches sometimes.
Instead of creating fracture complexity, increasing pad and fluid volume normally led to rather long hydraulic fracture halflength, which could be more than 500 m in some stages. Also, from the microseismic mapping result, the fractures were
basically well contained in the target formation. By matching the result of microseismic mapping and unconventional fracture
simulation during pumping, the simulated propped fracture half-length could hardly exceed 300-400 m with low-viscosity
fracturing fluid in most cases, and the major proppant distribution was within 200 m. Beside, due to the 600 m spacing
between these two horizontal laterals, the optimized fracture half-length was set approximately 300 m unless growing
fracture complexity was observed.
The followings are two instances for real-time adjustment.
Instance 1:
In this stage, the fracture was approaching to the monitoring well and some degree of fracture complexity was observed
with adequate fracture half length. The original designed second pad to increase fracture complexity was cancelled on-the-fly
to control the fracture length and keep better conductivity in the fracture network. Figure 11: Microseismic result showing
some degree of fracture complexity and enough fracture length (Left) and treating plot (Right)shows the microseismic
mapping result and treating plot.

Figure 11: Microseismic result showing some degree of fracture complexity and enough fracture length (Left) and treating plot
(Right)

SPE 158268

Instance 2:
In this case, the opposite two stages were pumping simultaneously to observe the fracture interference. From the
microseismic monitoring result, after the first pad was deplete, fracture stopped growing in length, and thus limited fractures
interference observed. Therefore, second pad was pumped and enlarged to increase the fracture surface contact. As the result,
fracture complexity increased to some extent afterwards. Figure 12 shows the microseismic mapping result and treating plot
for this instance.

Figure 12: Microseismic result showing fracture complexity increasing (Left) and treating plot (Right)

Other parameters, like pumping rate, proppant concentration were adjusted also. The pumping rate was decreased
eventually to reduce the downhole tool failure risk with limited influence on the fracture surface area created. As the result of
real-time adjustment which normally leads to saving fracturing fluids from the base design case, and stage number reduction
due to microseismic events overlap, only about half of the designed fracturing fluids were pumped to achieve desired fracture
surface contact area. Figure 13 below shows the fluid volume, proppant concentration and pumping rates were adjusted every
stage.

Figure 13: Fracture Parameters for 13 stages of one horizontal well

Simultaneous Fracturing
As mentioned in the previous section, most of the observed fracture geometries from early stages were related simple and
tend to be planar fractures. Anticipating more fracture surface area and better chance to create complex fracture network, [G.
Waters et al. 2009], four stages from both well were treated simultaneously to try if the opposite fractures interference can
increase the fracture complexity.
From the microseismic mapping result, no obvious fracture complexity increase can be seen when the opposite fractures
approaching to each other. The fracture geometry stayed planar with some degree of minor branches. Though the band width
of microseismic events are slightly bigger than most of other stages, by looking at the fracture propagation history, they are
more likely to be parallel planar fractures approaching oppositely, with some slight interference of fracture orientation when
they hit each other. Figure 14 shows the microseismic result for the four stages of simulatenous fracturing and the matched
fracture geometry.

SPE 158268

Figure 14: Microseismic result for four stages of simultaneous fracturing (Left) and matched fracture geometry (Right)

However, microseismic events results indicate that the heights of fractures from simultaneous fracturing are slightly greater
than the rest of stages (Figure 15), but not helping on increasing the surface contact within the target formation in this case.
Therefore, without achieving its objective of increasing fracture complexity, simultaneous fracturing was cancelled after
these four stages. And the results also suggest that fewer natural fractures might exist than we initially modeled, hence with
even greater energy when the fractures approaching to each other, fracture complexity was not facilitated by hydraulic
fractures and natural fractures interactions.

Figure 15: Vertical profile of microseismic events for 13 stages for one horizontal well. Target pay zone are shown in orange and
the 4 stages of simulatenous fracturing are shown in blue

Post-Treatment Evaluation
After finish total 26 stages, the microseismic result from dual monitoring wells has been re-processed. By comparing the
events arrival time and P/S-wave characteristics, the same event detected from two sets of receivers are identified and
combined accordingly. And the velocity model is calibrated to match the dual event arrival time. A refined microseismic
result has been delivered for post-treatment fracture evaluation.
More deliberate and time-consumption matching was done after the job. Compared with the real-time interpretation result,
the re-processed dual monitoring result (Figure 16 Left) can show more clearly for the fracture skeletons, which are more
likely planar fractures with a certain degree of complexity in some areas (Figure 16 Right). And both the real-time and reprocessed monitoring results show much intensive microseismic activities between the two horizontal laterals compared with
those at the outer area. The major reason is that the effective detecting range for the monitoring tool is really limited in some
part of this formation. However, the root cause is still under investigation.

10

SPE 158268

Figure 16: Microseismic result (Left) and matched fracture geometry (Right) for all stages

As the result of fracture geometry matching with the observed microseismic mapping result (Figure 17 Left), the natural
fractures model (Figure 17 Right), and horizontal stress anisotropy was calibrated. Compared to the previous natural fractures
model, the calibrated model shows fewer natural fractures in the area and much shorter fractures exist in the minimum
horizontal stress direction. However, long natural fractures might exist in some regions to match the fracture complexity
observed. On the other hand, the horizontal stress anisotropy is calibrated as only 30-50% of the original setting from the
tectonic knowledge. The calibrated result means the fracture complexity can be increased through HF-NF interaction as long
as natural fractures are distributed along the main fracture propagation routes. Furthermore, with the designed low-viscous
fracturing fluids and 3m3/min pumping rate per cluster, proper and relatively even proppant distribution can be achieved
along the fractures or fracture branches in the complex fracture networks.

Figure 17: Map view of all fractures matched with microseismic result (Left) and calibrated DFN for natural fractures (Right)

Production Result
Initial production tests on both wells show 124.5m3/d and 103.2m3/d respectively, which are significantly higher than all the
wells completed in tight oil reservoirs in the basin, of which the production normally varies from 5-8m3/d in vertical wells to
32m3/d in horizontal wells on average. Figure 18 below shows the initial production comparison between the two target
wells and the vertical well on the same pad. The initial production tests for these two pilot wells show 4.4 and 3.6 times of
production with only half of reservoir-to-wellbore pressure drawdown calculated from dynamic wellbore fluid level. And it
also needs to mention that the production from the vertical well is actually from Chang-7 and Chang-8 together, which could
be 100% more than production from Chang-7 individually, as Chang-8 formation has better quality and normally produce
more than Chang-7 in the area.

SPE 158268

11

Figure 18: Initial production for both horizontal wells (Well A and Well B) compared with the vertical well on the same pad

This initial result also shows significant production increase compared with the production from Chang-6 formation of the
nearest offset reservoir (Figure 19), which actually has slightly better reservoir quality. Compared with the best horizontal
well, the initial results show 80-100% production increase.

Figure 19: Production comparison with Chang-6 formation in offset reservoir

The stabilized production rates after 3 month are 5.5 and 3.9 times of the same vertical well. The results also show
approximately more than 50% increase compared with the horizontal well completed in the same formation in the north part
of the basin.
Conclusion
After multistage massive hybrid fracture treatments for each well, significant production increase achieved with only half
fluid volume compared with the initial design volume. In-depth reservoir understanding, advanced design and evaluation
workflow, and appropriate completion and fracturing technique are the key of this success, which has set a milestone for
developing such tight oil reservoirs in Ordos Basin.
Successful application of the unconventional fracturing design workflow led us to achieve the goal of this pilot project. The
key steps implemented in this workflow are identified as below:
1. Well correlations are established through the three vertical wells placed between the two horizontal target wells,
which reflect the detailed formation variation along the laterals.
2. A 3D zone model is then created based on the well correlations. Geologic structure and reservoir property variation
are captured in the 3D model.
3. A Discrete Fracture Network (DFN) model for natural fracture modeling is created based on the knowledge from
geologic knowledge and core observation, and can be further calibrated according to microseismic fracture mapping
result.
4. Unconventional Fracture Model (UFM) can explicitly model fracture propagation in 3 dimensions. HF-NF
interaction, stress shadow effect between nearby fractures, proppant transportation in fracture networks, lateral stress
and property variations are considered in this model.

12

SPE 158268

5.
6.

7.

Fracture placement with proper completion technology is optimized by looking at both reservoir quality and
completion quality.
Fracturing treatments are further optimized according to the microseismic fracture mapping results showing
observed fracture geometry, especially the fracture complexity and fracture penetration. This can be done during
every stage and between stages. Further adjustment for fracture placement along the laterals can also be done to
assure an efficient completion. Compared to optimization with microseismic result, unconventional fracture
simulation can give more information about proppant transportation and other fracture behaviors.
Post-job evaluation has been done based on dual monitoring microseismic result. Better knowledge of natural
fracture model, horizontal stress anisotropy and other fracture properties are updated for future fracturing design or
water injection program.

This pilot project also make four stages tested with simultaneous fracturing, which only show more height growth, but not
helping on increasing the surface contact within the target formation in this case.
Acknowledgement
The authors would like to thank PetroChina Changqing and Schlumberger for supporting this work and for permission to
publish this paper.
SI Metric Conversion Factors
bbl
x 1.589 874
cp
x 1.0
ft
x 3.048
o
F
(oF-32)/1.8
lbm/gal
x 1.198 264
psi
x 6.894 757

e-01
e+00
e-01
e+02
e-03

=
=
=
=
=
=

m3
mPa.s
m
o
C
kg/m3
MPa

References
H. Gu, X. Weng, J. Lund, M. Mack, U. Ganguly, and R. Suarez-Rivera. 2011. Hydraulic Fracture Crossing Natural Fracture
at Non-orthogonal Angles, A Criterion, Its Validation and Applications, paper SPE 139984, presented at the SPE Hydraulic
Fracturing Technology Conference and Exhibition, The Woodlands, Texas, USA, 24-26 January.
C. Cipolla, X. Weng, M. Mack, U. Ganguly, H. Gu, O. Kresse, and C. Cohen. 2011. Integrating Microseismic Mapping and
Complex Fracture Modeling to Characterize Fracture Complexity, paper SPE 140185, presented at the SPE Hydraulic
Fracturing Techonology Conference and Exhibition, The Woodlands, Texas, USA, 24-26 January.
X. Weng, O. Kresse, C. Cohen, R. Wu, and H. Gu. 2011. Modeling of Hydraulic Fracture Network Propagation in a
Naturally Fractured Formation, paper SPE 140253, presented at the SPE Hydraulic Fracturing Techonology Conference and
Exhibition, The Woodlands, Texas, USA, 24-26 January.
C.L. Cipolla, M.J. Williams, X. Weng, M. Mack, and S. Maxwell. 2010. Hydraulic Fracture Monitoring to Reservoir
Simulation: Maximizing Value, paper SPE 133877, presented at the SPE Annual Technical Conference and Exhibition,
Italy, 19-22 September.
O. Kresse, C. Cohen, X. Weng, R. Wu, and H. Gu. 2011. Numerical Modeling of Hydraulic Fracturing in Naturally
Fractured Formations, paper ARMA 11-363, presented at the 45th US Rock Mechanics / Geomechanics Symposium, San
Fransisco, California, USA, 26-29 June.
W. Xu, J. LeCalvez, and M. Thiercelin. 2009. Characterization of Hydraulically-Induced Fracture Network Using Treatment
and Microseismic Data in a Tight-Gas Formation: A Geomechanical Approach, paper SPE 125237, presented at the SPE
Tight Gas Completions Conference, San Antonio, Texas, USA, 15-17 June.
J.H. Frantz Jr., J.R. Williamson, W.K. Sawyer, D. Johnston, G. Waters, L.P. Moore, R.J. MacDonald, M. Pearcy, S.V.
Ganpule, and K.S. March. 2005. Evaluating Barnett Shale Production Performance Using an Integrated Approach, paper
SPE 96917, presented at the SPE Annual Technical Conference and Exhibition, Dallas, Texas, USA, 9-12 October.
A.A. Ketter, J.L. Daniels, J.R. Heinze, and G. Waters. 2006. A Field Study Optimizing Completion Strategies for Fracture
Initiation in Barnett Shale Horizontal Wells, paper SPE 103232, presented at the SPE Annual Technical Conference and
Exhibition, San Antonio, Texas, USA, 24-47 September.

SPE 158268

13

J. Daniels, G. Waters, J. LeCalvez, J. Lassek, and D. Bentley. 2007. Contacting More of the Barnett Shale Through an
Integration of Real-Time Microseismic Monitoring, Petrophysics and Hydraulic Fracture Design, paper SPE 110562,
presented at the SPE Annual Technical Conference and Exhibition, Anaheim, California, USA, 12-14 October.
T.N. Olsen, T.R. Bratton and M.J. Thiercelin. 2009. Quantifying Proppant Transport for Complex Fractures in
Unconventional Formations, paper SPE 119300, presented at the SPE Hydraulic Fracturing Technology Conference, The
Woodlands, Texas, USA, 19-21 January.
G. Waters, B. Dean, R. Downie, K. Kerrihard, L. Austbo, and B. McPherson. 2009. Simultaneous Hydraulic Fracturing of
Adjacent Horizontal Wells in the Woodford Shale, paper SPE 119635, presented at the SPE Hydraulic Fracturing
Technology Conference, The Woodlands, Texas, USA, 19-21 January.

Das könnte Ihnen auch gefallen