Horizontal Well Stimulation Optimization - Barree · Horizontal Well Stimulation ... • Can...
65
Horizontal Well Stimulation Optimization R. D. Barree Barree & Associates LLC
Horizontal Well Stimulation Optimization - Barree · Horizontal Well Stimulation ... • Can vertical wells drain the reservoir ... the orientation of the well and completion and
• Can vertical wells drain the reservoir (contact the required volume)?
• Can vertical infill wells be drilled economically?
• Can field development economics be improved by switching to horizontals?
Presenter
Presentation Notes
When considering a horizontal well development, it is helpful to consider the potential risks and rewards of horizontal wells compared to vertical wells. One of the most important decisions is whether vertical wells can be effectively stimulated. If effective fracture lengths are short in vertical wells, they can be expected to perform similarly in horizontal transverse fracture systems. The typical reason horizontal well developments are employed is to increase drainage and EUR while limiting the number of wells. This requires accurate description of the reservoir and knowledge of drainage area and aspect ratio of vertical wells, as well as their efficiency. Horizontal wells may also be unfavorable in stacked pay or large vertical reservoir sections.
Selection of Horizontal Well Completion/Stimulation Method
• Know the reservoir– Deliverability (kv, kh, Pi, GLR)– Stress field magnitude and orientation– Drainage area shape and size
• Define the well– Maximum possible length– Limitations on orientation– Expected skin damage– Options for completion
• Cased/cemented; open‐hole; uncemented liner
Presenter
Presentation Notes
The decision to opt for horizontals, and the proper design of horizontal wells, is based on understanding the reservoir. The flow capacity, directional permeability, stress field, and compartmentalization of the reservoir must all be well known. There may also be limits imposed by mechanical stability or drilling limits (TD, torque and drag, hole stability, mud weight limitations) that can affect development options. Finally, the orientation of the well and completion and stimulation methods must be decided to maximize recovery and economic return.
• Longitudinal propped frac– Longitudinal frac with “hair” on it
• Transverse propped fracs (multiples)
Presenter
Presentation Notes
Depending on the deliverability of the reservoir and the damage done to the well while drilling, various completion and stimulation options are available. In almost all cases, regardless how careful the drillers are, a horizontal well will be damaged (have positive skin) after drilling and will benefit substantially from stimulation. The degree of benefit and costs associated with stimulation, including wellbore configuration, must be considered. Orientation of the well with respect to stress field can determine which completion options are available.
Factors Affecting Horizontal versus Vertical Well Performance
• Flow path in proppant pack
• Near‐well tortuosity and flow restrictions
• Reservoir drainage area
• Interference of multiple transverse fractures
• Drainage aspect ratio
Presenter
Presentation Notes
Various parameters affect overall production performance in horizontal wells. Predict-K is primarily focused on the proppant pack conductivity and cleanup. Production forecasting and design was not originally intended to be the goal of the program, but it has been moving steadily in that direction. This development continues and extends that progression. Many of the controlling parameters are dependent on accurate reservoir characterization and are not determined by proppant pack conductivity.
• Reservoir permeability• Frac fluid cleanup• Producing water‐cut and condensate yield• Applied drawdown• Created fracture length• Average proppant concentration• Applied closure stress• Reservoir rock “hardness”
Effective frac length is usually much shorter than expected
Presenter
Presentation Notes
As previously discussed, the maximum effective producing length of an induced fracture depends on many factors associated with reservoir deliverability and producing character. The achievable frac length in a vertical well will control the behavior of a frac in a horizontal well. All frac cleanup is controlled by the same physics. It is necessary to avoid overestimation of stimulation benefit when designing horizontal well programs.
Stepping back from the fracture gives a better true-scale perspective of the fracture-to-wellbore connection. In the figure, there is a vertical fracture covering an 80 ft pay height with a half-length of 1000 ft. The wellbore diameter is 8” and the frac was placed through a frac port. On this scale the frac port and the entire wellbore is not visible. Is the well vertical, horizontal transverse, or horizontal and longitudinal with the frac? In terms of the frac geometry and production performance, it is indeterminate from this perspective. What matters is the near-well connection, degree of convergent radial flow, non-Darcy effects near the well, and tortuosity or offset between the frac plane and the flow port. If this is true, then the Predict-K cleanup and conductivity calculations should apply equally well to fracs on horizontal wells if we can define the geometry of the connection.
Assumed Fracture Geometry for Productivity Estimates: Vertical Well
Gross Fracture Height Open to Flow Constant Conductivitywith Length and Height
Presenter
Presentation Notes
To model vertical fractured well productivity, some common assumptions are made. In Predict-K, the case of a single planar vertical fracture intersecting a well in a closed, bounded system is considered. The fracture is assumed to be producing over the same height as the producing formation (net pay), and to communicate with the wellbore over the entire thickness of the producing formation (gross frac height). The fracture is also assumed to have uniform conductivity over its flowing length and height. As has been previously shown, determining the actual effective producing conductivity in multiphase non-Darcy flow is critical. Determining the effective fracture length contributing to production is also critical. These two input parameters have been the primary focus of Predict-K.
Assumed Fracture Geometry for Productivity Estimates: Horizontal Transverse
Gross Fracture Height Open to Flow Constant Conductivitywith Length and Height
Presenter
Presentation Notes
For vertical fractures that are transverse to a horizontal wellbore axis, the flow geometry for the fracture is essentially the same as a vertical fracture on a vertical well with a small perforated interval. The frac covers the height of the pay zone and flow through the fracture surface must flow down the length of the fracture to the connection with the well at the perfs. The closure stress, pack width, proppant concentration, and cleanup behavior are the same as for the vertical well case.
• Difficult and expensive to place in desired locations
SIDE VIEW
HYDRAULICFRACTURE
L f
Presenter
Presentation Notes
The image of equally spaced transverse fractures along a horizontal well gives the impression that a large reservoir volume can be accessed and connected to the well. This geometry allows drilling across reservoir flow boundaries and connection to a larger reserve base. When the effective frac length that can be generated is large, this geometry can be very effective. If a well is drilled to intersect existing conductive fractures, and they are not damaged by drilling and completion, recovery can be significantly enhanced.
Assumed Fracture Geometry for Productivity Estimates: Horizontal Longitudinal
Gross Fracture Height Open to Flow Constant Conductivitywith Length and Height
Presenter
Presentation Notes
For a vertical frac along the axis of a horizontal well, the effective flow path length of fluid from the formation to the wellbore can be as short as half the pay height (for an open-hole well or uncemented annulus) and up to half the spacing between the clustered perf or frac port locations. The impact of frac conductivity and ultimate production performance depends on the specifics of the completion design.
• Thin reservoir sections• Low kv/kh in thick reservoirs
• Better when effective Xf is low
• Better in continuous reservoirs
• Large volume treatments
• Easy to place and pump effectively
H O R IZ O N TA L S E G M E NPA R A L L E L
TO L E A S T S T R E S S
Lf
Presenter
Presentation Notes
Longitudinal fractures are not subject to the same cleanup problems as vertical or transverse fracs. If the longitudinal is connected along the lateral length, then the flow path is no more than the half-height of the fracture or pay. In this case the effective length of the fracture can be short and conductivity is not critical. Vertical coverage is essential and vertical permeability must be maintained across the pay height. This can be a problem in any fracture geometry from horizontal wells. In general, longitudinal fracs perform more efficiently in matrix-perm dominated production systems where contacting natural joints and fractures is not critical to production. Because of the stress concentration around the wellbore, the breakdown and treating pressures in this geometry are often significantly lower than for transverse fracs.
• Usually results from intended longitudinal fracs
• More of an accident or “gift” than an intentional result
Presenter
Presentation Notes
When generating longitudinal fracs the formation almost always generates complex secondary shear fractures that result in a complex, non-planar geometry. The secondary fractures may contribute to production but are difficult to predict or control. Field evidence strongly suggests that the secondary fractures can have a large extent.
• Controlled by in‐situ stress field– Magnitude of all three principal stresses
– Principal stress axis
– Degree of overpressure or depletion
• Direction of well (azimuth)
• Deviation of well from vertical
Presenter
Presentation Notes
In horizontal wells, as in vertical wells, the fracture azimuth and dip are controlled by the in-situ stress tensor. For horizontal wells, the azimuth and deviation of the well can be planned to control the fracture orientation relative to the wellbore axis. The relative deviation between the well and stress field also affects breakdown pressure, treating pressure, and near-well connection efficiency.
Radial Coordinate System of Stresses around a Wellbore
σr
σt
σvσv
σx
σy
Presenter
Presentation Notes
The stresses discussed so far refer to undisturbed rock which makes up the bulk of the reservoir. These are called “far field” stresses. However, near a wellbore the stress state is disturbed by removal of some rock material during drilling. The removed rock originally supported some of the in-situ stress. �To compensate for the loss of support caused by removing this rock a new stress distribution is established around the wellbore, resulting in a stress concentration at the borehole wall. Around the wellbore, the stresses can be represented in a radial coordinate system, rather than in the rectangular system used for far field stresses. Near the well stresses must be resolved into vertical (v), radial (r) and tangential (t) components rather than the normal x, y, and z direction stresses applicable in the undisturbed reservoir. At some distance from the well the stresses return to their undisturbed far field values.
Stresses Vary with Directionand Distance from Well
( )
( )
σ σ σ σ σ θ α
σ σ σ σ σ θ α
σ α
rh H w h H w w w
w o
th H w h H w w
w o
v ob o
rr
rr
rr
rrP P
rr
rr
rrP P
P P
=+
−⎛⎝⎜
⎞⎠⎟ +
−− +
⎛⎝⎜
⎞⎠⎟ + −
=+
+⎛⎝⎜
⎞⎠⎟ −
−+
⎛⎝⎜
⎞⎠⎟ − −
= −
21
21 4 3 2
21
21 3 2
2
2
2
2
4
4
2
2
2
2
4
4
2
2
cos
cos
Po = far field pore pressurePw = wellbore fluid pressurePob = overburden pressure
r = distance from wellboreσH = maximum horizontal stressσh = minimum horizontal stressθ = angle from direction
of minimum stress
σH
σh
Presenter
Presentation Notes
The magnitudes of the induced radial and tangential stresses can be calculated as a function of radius from the wellbore, and angle from one of the principal horizontal stress directions. The equations shown are for the general case of two unequal horizontal stresses, where the direction angle (q) is based on the angle from the minimum stress direction. Note that the vertical stress near the wellbore is only a function of the overburden and pore pressures. Both the radial and tangential stresses decline rapidly as distance from the well (r) increases. A few borehole diameters away from the well, the tangential stress declines to its undisturbed far-field value, while the radial stress increases to the far-field value. So, as r approaches infinity, the equations simplify greatly to:
Tangential Stress Distribution Around a Horizontal Well
S1=6000
S1=6000S2=6000S3=4200
IncS1=0AzSH=70
Azi=70Dev=90
The wellbore acts asa tunnel arch:Vertical stress istransmitted to thesides of the hole
Presenter
Presentation Notes
The tangential stress distribution around the well can be visualized using color to represent stress magnitude. In the figure the red areas show high compressive stress and the blue area low stress. For a normal stress field with the horizontal stress much less than the vertical stress the maximum compressive stresses occur at the sides of the hole. At the top and bottom the tangential stress is low. As horizontal stress increases, the stress at top and bottom increase and the variation in stress around the hole diminishes.
Fracture Initiation at Point of Minimum Tangential Stress
+ =
σh max
σh min
well pressure
Failure whentangential stressreaches zero
Presenter
Presentation Notes
When a well is drilled, uneven tangential or “hoop” stresses are generated around the wellbore. These stresses are compressive, and higher than the far-field in-situ stress. Wellbore fluid pressure acts equally in all directions as it is raised above the reservoir pressure. The superposition of the wellbore tangential stresses and borehole fluid pressures tends to decrease the magnitude of the compressive tangential stress around the well. When the wellbore pressure is raised high enough, the tangential stress will approach zero at some azimuth around the well. With enough additional increase in well pressure to overcome the rock tensile strength, the formation will break, or “part” in this preferred stress orientation.
Min Horizontal far-field StressTangential Near-Well Stress
Axial Stress
Radial Stress
Presenter
Presentation Notes
As the well is deviated from vertical, the orthogonal projection of stresses to the wellbore changes. The tangential stress is driven by the three principal stresses and deviation of the well, along with its azimuth relative to the horizontal stresses.
When fracturing deviated wells, a situation similar to that encountered with perforation orientation arises. Because a deviated hole does not lie in the plane of the vertical principal stress, additional shear stresses exist which must be overcome to initiate a fracture. As the angle of well deviation from the principal stress axis increases, the pressure required to initiate a fracture also increases. The figure shows that the maximum breakdown pressure occurs at a critical deviation angle of about 45 degrees. Also, the orientation of the wellbore axis with respect to the plane of the fracture (a) influences the observed breakdown pressure. This is most easily visualized considering a horizontal well with a deviation angle of 90 degrees from the assumed vertical maximum principal stress direction. If the well is drilled in the plane of the fracture (a = 0) so that a longitudinal fracture will be initiated along the wellbore axis the breakdown pressure gradient will be approximately 0.65 psi/ft for the conditions shown. However, if the same well were drilled normal to the fracture plane (a = 90) so that a transverse fracture will initiate, the breakdown gradient would be approximately 0.94 psi/ft. Because of this high breakdown pressure, it is more normal for a longitudinal fracture to initiate in a horizontal well, regardless of orientation. The created fracture may then split into multiple plumes which re-orient to fit the dominant stress field away from the well.
Preferred Fracture Direction for a Normal Fault Environment
Presenter
Presentation Notes
The stress tensor can be derived from breakouts and direct measurements of closure stress (minimum in-situ). The complete stress tensor is applied to the Geo-Mechanics International SFIB (Stress and Failure in Inclined Boreholes) model, along with an estimated rock tensile strength. The formation parting pressure can be predicted for various well azimuths and deviations using a solution of the Kirsch equations. The plot on the right is a lower hemisphere representation of well azimuth and deviation. A well drilled due north points to the top of the plot. A well drilled vertically is at the center of the circle. The first concentric ring is 30 degrees deviation from vertical. The second ring is 60 degrees deviation. A horizontal well drilled to the east would lie on the outer rim at the right side of the plot. Color contours represent the overbalance pressure (in psi) required to initiate a fracture in any wellbore configuration. The overpressure is shown with respect to pore pressure. So, a horizontal well drilled to the SW will fracture with a wellbore pressure of 2800 psi above pore pressure. The plot shows that wells drilled at 45 degree deviation to the SE and NW will prove the most difficult to frac. These wells will require more than 5000 psi overbalance.
Preferred Frac Direction for a Thrust‐ or Strike‐Slip Fault Environment
Presenter
Presentation Notes
This is an SFIB analysis for fracture initiation in a strike-slip fault regime. Color contours represent the overbalance pressure (in psi) required to initiate a fracture in any wellbore configuration. The overpressure is shown with respect to pore pressure. So, a horizontal well drilled to the SW will fracture with a wellbore pressure of 800 psi above pore pressure. The plot shows that wells drilled at 90 degree deviation slightly west of north and east of south will prove the most difficult to frac. These wells will require more than 5000 psi overpressure. The azimuth of the well relative to the direction of the thrust is not critical. Breakdown and treating pressures will be similar for any well drilled in a +/- 30 degree window.
– Original Gas in Place (OGIP) contacted by vertical
• Generate a metric to evaluate completion efficiency of multiply fractured wells
• Determine the number of fractures to place in a lateral for effective reservoir drainage
Presenter
Presentation Notes
To effectively design a horizontal well completion with multiple transverse fractures, a method is needed to accurately determine the effective recovery and rate delivered by any combination of drilled lateral length and imposed fracture system. Production must be tied to a reasonable analog for reservoir performance. The best available analog is the production of a nearby offset vertical well with an effective fracture treatment. This gives reservoir flow capacity, a measure of effective fracture length, and a (hopefully) statistical representation of the compartmentalization of the reservoir (area drained by a single well). Given this information, it is possible to determine the optimum spacing and frac size, and to evaluate the efficiency of current completions.
Horizontal Well Production Characterized by Vertical Well Performance
• Vertical well performance is needed to determine reservoir properties:– KH, P*, Xf, A, L/W
• Vertical transverse fracs on the lateral are assumed to clean‐up and perform like the vertical‐well frac
• The reservoir is assumed to be isotropic and homogeneous along the lateral
W
L
Presenter
Presentation Notes
A fractured vertical well establishes a drainage pattern based on fracture effective length and permeability anisotropy, along with physical flow barriers in the formation. Analysis of production gives KH, A, and L/W for a vertical well. Once several wells have been analyzed, the data should be sufficient to characterize the average reservoir performance and expected range of variance.
When combined, a fractured horizontal well produces as if the drainage area of a vertical fractured well (the outer limbs) is added to the drainage area and production of a vertical fracture in a thin reservoir slice, multiplied by the number of transverse fractures, including the end-fracs. The inner face of each end-frac contributes the production of an equivalent internal frac. The result of all this is that the internal transverse fracs interfere with each other at a time determined by spacing and reservoir permeability. The benefit of the internal transverse fracs is primarily felt at early time and diminishes after interference is established. At late time, the bulk of the production occurs from the drainage areas of the outer fracs. At very long time, the entire composite wellbore can be modeled as an infinite-conductivity horizontal well of length equal to the total distance between the outermost fractures. The overall stimulation benefit can be represented by an equivalent wellbore radius equal to half the infinite-conductivity effective frac length established by cleanup. With more competing fractures the velocity within each frac decreases, if the entire well is production limited or if BHFP increases due to friction or loading. This may decrease the effective length of the fractures because the energy available for cleanup is reduced.
• Single Phase Gas simulator with separate fractures contributing to flow against a constant surface pressure
• Analyze total well response as a single zone to determine composite properties
• Case 1 – Variable kh in each fracture
• Case 2 – Variable Xf in each fracture
Presenter
Presentation Notes
A model was developed to calculate the production from internal and external fractures, including interference effects. Two cases are presented to investigate the accuracy of the model when applied to analyzing actual field performance. In the first case, the formation kh is varied along the length of the well. In the second case, the effective frac length is varied with constant kh.
The total well production can be analyzed as a rate transient and fitted to a Pwd-Tda type curve. The points on the plot are the production data and the lines are the type curve model generated to match it. A unit slope of the pressure derivative (bottom line) indicates boundary dominated flow.
• OGIP = 6.61 BCF (actual 13.2 BCF)– Half the net pay does not contribute to production
• kh = 2.18 mdft (actual kh 2.18 mdft)
• Flow capacity from multiple fractures is accurately measured by a single composite analysis
∑∑=
i
ii
hhk
kh
Presenter
Presentation Notes
Analysis of the results show that the apparent kh is the sum of the total kh of the system or the sum of the kh contacted by each frac. The OGIP is underestimated because half the total pay height in the example has such a low perm that early contribution to production is minimal. The apparent fracture length underestimates the input value.
Semilog Plot k = 0.0500 md and Xf = 345.3 feet - Case 2
262.
76
Firs
t Bou
ndar
y0
200000
400000
600000
800000
1000000
1200000
1400000
0.1 1 10 100 1000 10000
Ta (days)
ΔM
(P)/Q
Presenter
Presentation Notes
The apparent kh is the total of the system. In this case, the correct permeability is returned when the total of all frac heights is used as formation thickness. If the height of the zone was used then the apparent perm would be the actual perm times the number of fractures contributing.
• Xf = 342 ft (semilog plot shows 345 ft)• For a system with little permeability variation the Xf
will be the height averaged individual Xf of each of the fractures. As the variance in permeability increases the measured Xf decreases
i
fif h
XhX
∑∑=
Presenter
Presentation Notes
The total flow capacity returned is correct for the entire well. Because the perm feeding each fracture is the same, the OGIP returned is correct. The apparent Xf is the height weighted average of all fractures contributing.
• As more fractures are added to a lateral the total well kh should increase
• Assuming a constant kh for each fracture
• Average kh should be measured from a single vertical fracture
Average
Horizontal
(kh)n (kh) Efficiency Fracture =
Presenter
Presentation Notes
Based on this analysis, it is possible to determine the efficiency, in terms of number of fractures usefully contributing production, if the average kh of a vertical well is known. The apparent kh of the total well should be the average vertical kh multiplied by the number of fracs. The efficiency is therefore the ratio of the total horizontal well kh to the vertical well kh times number of fracs placed.
Production Rate Decline for Internal and External Fractures
Presenter
Presentation Notes
The plot shows the production decline rates for interior and exterior fractures on the same length lateral with different fracture spacing or number of fracs placed. When fractures are too close together the production transients interfere quickly and the interior frac rate declines rapidly. The goal is to space the fractures to maximize the flush production from each before interference dominates flow.
Based on economic analysis and job costs, designing for an interference time of 6 months appears to give the maximum economic value for a generic lateral completion over a wide range of permeabilities. Earlier interference contributes little incremental reserves and increases cost of completion. The plot indicates that a spacing of 750 feet is adequate for 0.01 md reservoirs while a spacing of 250 feet can be used for 0.001 md systems. The permeability refers to the large-scale system effective perm during production. Cross-cutting joints and fractures will decrease time to interference and push to wider frac spacing.
Field Example – Vertical WellVertical Well Volumetric OGIP = 1.82 BCF
Gas in Place = 1.77 BCF - Equivalent Area 27.22 acres - L/W ratio 2.72 - Vertical Well
90.2
7Bo
unda
ry
0
500000
1000000
1500000
2000000
2500000
3000000
3500000
4000000
4500000
5000000
0 200 400 600 800 1000 1200 1400 1600 1800 2000
Ta (days)
ΔM
(P)/Q
0.1
1
10
100
0.00001 0.0001 0.001 0.01 0.1 1 10 100
TDA
Pw
D
k = 0.0303 md and Xf = 101.9 feet
Boun
dary
0
200000
400000
600000
800000
1000000
1200000
1400000
1600000
1800000
2000000
0.1 1 10 100 1000 10000 100000
Ta (days)
ΔM(P
)/Q
0
200
400
600
800
1000
1200
1400
1600
1800
0 100 200 300 400 500 600 700 800
Time (Days)
Rat
e (M
SC
F/da
y)
0
1000
2000
3000
4000
5000
6000
7000
Pres
sure
(psi
a)
Presenter
Presentation Notes
The summary analysis of a fractured vertical well sets the stage for analysis of a horizontal producer with multiple fractures. The vertical well is in boundary dominated flow in a volumetric reservoir.
• Vertical Well Production Analysis– kh ‐ 2.4 md.ft
– Infinite conductivity half length 102 ft
– Gas in place 1.77 BCF
Presenter
Presentation Notes
The vertical well produces from 2.4 md-ft of flow capacity and contacts 1.77 BCF in a 27 acre area with 2.7:1 aspect ratio. The effective fracture length is 102 feet.
• Horizontal Well – 12 fractures in a 4000 ft lateral– 363 ft actual spacing– Kh 6.32 md.ft– Infinite conductivity Half Length 135 ft– Gas in place 2.90 BCF
• Fracture Efficiency = 6.32 / 12*2.4– 22% (3 out of 12 fractures contributing)
• Fracture Spacing– Based on 6 month interference time a fracture spacing of 1360 ft is more appropriate ( 3 fractures)
– The fractures interfere after 12.6 days at a spacing of 363 ft
Presenter
Presentation Notes
The horizontal well stimulation has been grossly over designed and equal production could have been obtained with ¼ the total number of fracs placed.
• Flow capacity (kh) from production analysis of multiple fractures is accurate
• Xf from production analysis is a poor method to evaluate completion efficiency
• For an efficient completion kh from each layer should add (total kh = number fracs * average single layer kh)
• Increasing lateral length and distance between the bounding fractures will increase contacted reserves
• Placing more fractures in a lateral than the permeability can support leads to destructive interference
Presenter
Presentation Notes
Performance of a complex horizontal well with multiple fractures can be graded based on offset vertical well performance. Without knowing the producing character of a representative vertical well, and the variance expected in a given reservoir, it is not possible to optimize a horizontal well program.
Effect of Producing Lateral Length on Production Rate
Presenter
Presentation Notes
Lateral well length has a dominant effect on well production rate. The curves show that instantaneous rate is directly proportional to well length. A doubling of Lh affects a doubling in rate at any time.
The rate forecasts at constant BHFP support the indications of the type-curves. Rwa strongly influence the rate of production decline at early times. All these runs are made assuming a maximum flow rate of 2000 bpd, otherwise the well IP would increase for the larger Rwa values. At late-time the well production rates tend to converge and the decline is dominated by reservoir geometry rather than near-well stimulation.
• If KH is additive for each transverse frac– Does the number of producing fracs times effective length exceed drilled lateral length?
– Is the production dominated by secondary joints and fractures?
– Can a longitudinal frac expose more area?
Presenter
Presentation Notes
When comparing transverse and longitudinal fracture systems, the effective length of the transverse fracs and the number of producing fractures (including interference effects) must be compared to the drilled length of the lateral. If the production is dominated by matrix permeability, and not an extensive conductive fracture network, then the total of the effective frac lengths (tip to tip) times the number of contributing fracs should exceed the drilled lateral length or a longitudinal frac will likely perform better. This assumes that a frac can be propagated along the entire drilled lateral length.
Comparison of Predicted Transverse and Longitudinal Frac Performance in Low Perm System
Presenter
Presentation Notes
The plot shows the transverse frac field data (magenta line) and the Predict-K match of the production with multiple transverse fracs, and the results for three longitudinal fracture cases. The yellow points show the production for 100% of a 2000’ lateral stimulated over the entire pay height. The green points, which cross the transverse frac case, are for 50% of the lateral length stimulated. The blue points show the results is only 7 20% of the lateral is frac’d.
• Establish communication with the reservoir across entire well length
• Extensive stimulation outside the well (frac‐length) is not critical
• Propagation of secondary fractures (off longitudinal) may be beneficial (not clear from production response)
Presenter
Presentation Notes
General guidelines for horizontal well stimulation stress maximizing contributing lateral length and surface area of reservoir connected to the well. Pumping excessively large jobs to generate long created lengths of low conductivity, or massive conductivity away from the well, generally have poor return on investment.
Cemented vs. Uncemented Liners and Open‐hole Completions
• Cemented liners and/or casing– Only for matrix‐flow systems– Only when borehole stability is questionable– When zonal isolation for stimulation is absolutely necessary
• Uncemented liners (with limited perfs)– Longitudinal fracs (acid and propped)– Re‐entering or re‐activating existing fractures
• Open‐hole completions– Only in hard, competent rock– Good for surface acid‐wash completions– Acid and water‐fracs for opening existing natural fractures
• Poor control of fluid entry• Difficult to ensure multiple fractures treated
Presenter
Presentation Notes
Many choices are available concerning types of completion and methods to develop diversion for frac placement. Cemented liners are only recommended in certain cases where secondary fracture production is insignificant and when zonal isolation is needed. Uncemented liners can be used in many cases to provide long-term access to the well and to provide diversion for frac placement. Open-hole completions are suitable only for stable formations where the risk of hole collapse is minor. Frac placement in open-hole completions offers special challenges.
Long flow path in fracture requires high conductivity and long effective Xf, with similar behavior to vertical well
No production allowed through cemented section
Presenter
Presentation Notes
In fracturing of deviated wellbores, the deviation between the well azimuth and fracture orientation can cause additional problems after initiation. Because fractures may curve out of the plane of the wellbore to propagate in the direction of the principal stress field the near wellbore portion of the fracture may have a restricted width. In these cases, severe near wellbore tortuosity may exist which causes high treating pressure, especially at low flow rates. During the treatment, the danger of bridging and early screenout also exists. If the fracture treatment is successfully pumped, it may perform badly because of converging flow effects or poor near wellbore conductivity. Converging flow effects are caused by forcing all the fluid produced through a narrow perforation interval connected to the wellbore. In a vertical well this is analogous to a limited partial penetration completion. Low final fracture conductivity can be caused by the reduced fracture width realized during pumping, which limits the maximum proppant concentration. In addition, the proppant placed near the wellbore will be subjected to higher closure stresses than expected, which can induce more crushing and lower final proppant pack permeability. Under the best of conditions, this kind of frac acts as a fracture in a vertical well. It is required to carry fluid from the formation to the well down a long flow-path. For this reason, the same high conductivity is required as in a vertical well frac. All the same damage mechanisms apply, with the possible addition of near-well tortuosity and pinching. If the effective producing length of the vertical well fracture is small, the same can be expected of transverse fracs from a cemented liner.
Induced deformation (strain) generates stress in proportion to YME
Second frac:May be Axial
Third frac:May be oriented arbitrarily
Presenter
Presentation Notes
Stress shadowing and induced stress diversion have been mentioned previously. A common observation in fracture mapping is that a primary transverse frac can be generated if the well is drilled and completed in the correct orientation. Often the generation of the first frac causes enough strain (related to fracture width) to alter the stress field in the surrounding rock. In high modulus rocks, or reservoirs with low stress anisotropy, this can be sufficient to change the principal stress orientation and the direction of fracture growth. The second fracture generated may be axial to the well and perpendicular to the first frac. Later fracs may take arbitrary orientations or can be horizontal.
Typical Observation of Multiple Fracs in Horizontal Wells
Presenter
Presentation Notes
The illustration of a mapped fracture system from the Rose Field is taken from Pinnacle Technologies mapping data in a companion SPE paper. In this well, and several offsets, the primary fracture was longitudinal. A second transverse fracture appeared at the toe of the well, followed by a horizontal fracture at the heel. These are the fractures that could be resolved from the mapping and other multiple fractures may have been formed.
Interaction of Fractures: Simultaneous Injection into 11 Fractures
Presenter
Presentation Notes
Map view of hydraulic fracture propagation patterns from 11 simultaneously pressurized injection points for initial fracture spacings of a) 100 m, b)50 m, and c) 25 m. All fractures are vertical have the same height of 100 m. The velocity exponent for propagation was n=1, and the remote horizontal stress was isotropic. As fracture spacing gets smaller, mechanical fracture interaction becomes stronger, evidenced by the non-planar propagation paths of the interior fractures of the arrays. The illustration is taken from ARMA 08-327, by J. E. Olson.
Hydraulic fracture pattern for a sequential injection from left to right, where each fracture injection is started after the previous one reaches its full extent (intersecting the 300 m boundary). Fracture spacings are a) 100 m, b) 50 m, and c) 25 m for fractures contained within a 100 m layer. Each case had a moderate differential compression in the y-direction (40% of net pressure). When initial fracture spacing is less than layer thickness (b and c), there is significant mechanical interaction between fractures, and propagation paths are substantially diverted from orthogonal to the wellbore trend. (Olson, op cit).
Achieving Diversion and Fracturing the Entire Lateral
• Cemented liner with limited‐entry perforating– Limit number and size of perforations
• Multiple plug and perf stages
• Open‐hole with liner and external packers– Chemical “swell” packers (Halliburton)
– Hydraulic‐set packers
– Inflatable OH packers
• Sliding sleeve and frac‐ports for individual stages
• Dynamic diversion with uncemented liners
Presenter
Presentation Notes
If the goal is to stimulate the longest possible section of the lateral, or to control the initiation of multiple transverse fracture, some means for diversion or isolation must be utilized. Not all the techniques listed or now being used can be discussed in detail. A few examples will be presented. Conventional limited-entry fracturing through a cased, cemented, and perforated completion requires a small number of holes, ensured breakdown, and high treating rates and pressures. Jobs have been pumped at rates in excess of 100 bpm down large casing. Perforation breakdown, erosion, and cement damage are frequently problems with these completions. High breakdown pressures, inability to establish a fracture, high treating pressures, and early screenouts are reported. Mechanical isolation systems using external packers in open holes are becoming more common. The degree of mechanical complexity varies. External packers and sliding sleeves are sometimes utilized to complete multiple laterals out of a single vertical wellbore. The dynamic diversion technique that has been used very successfully in several reservoir will be described in detail.
Packers Plus Mechanical Isolation System for Open‐Hole Stacked‐Fracs
5-1/
2" 2
0 pp
f P-1
10 C
asin
g Sh
oe
5-1/
2" P
acke
r Plu
s P
erm
a-Pl
us L
iner
Han
ger P
acke
r
4.62
5" o
pen
hole
2-7/
8" 6
.5 p
pf 8
rd li
ner b
ox x
pin
con
nect
ions
Pac
kers
Plu
s 4-
1/2"
x 2
-7/8
" Roc
kSea
l II o
pen
hole
pac
ker s
yste
m
2-7/
8" 6
.5 p
pf 8
rd li
ner b
ox x
pin
con
nect
ions
Pac
kers
Plu
s 2-
7/8"
Dril
labl
e Fr
acPo
rt
2-7/
8" 6
.5 p
pf 8
rd li
ner b
ox x
pin
con
nect
ions
Pac
kers
Plu
s 4-
1/2"
x 2
-7/8
" Roc
kSea
l II o
pen
hole
pac
ker s
yste
m
2-7/
8" 6
.5 p
pf 8
rd li
ner b
ox x
pin
con
nect
ions
Pac
kers
Plu
s 2-
7/8"
Dril
labl
e Fr
acPo
rt
2-7/
8" 6
.5 p
pf 8
rd li
ner b
ox x
pin
con
nect
ions
Pac
kers
Plu
s 4-
1/2"
x 2
-7/8
" Roc
kSea
l II o
pen
hole
pac
ker s
yste
m
2-7/
8" 6
.5 p
pf 8
rd li
ner b
ox x
pin
con
nect
ions
Pack
ers
Plus
4-1
/2" x
2-7
/8" R
ockS
eal I
IS o
pen
hole
pac
ker s
yste
m
Roc
kSea
l Cen
traliz
er
2 7/
8" 6
.5pp
f 8rd
linn
er b
ox x
pin
con
nect
ions
Roc
kSea
l Cen
traliz
er2-
7/8"
Hyd
raul
ic F
racP
ort w
ith a
+/-3
,720
psi
ope
ning
pre
ssur
e
Roc
kSea
l Cen
traliz
er
2-7/
8" 6
.5 p
pf 8
rd li
ner b
ox x
pin
con
nect
ions
2-7/
8" T
oe C
ircul
atin
g Su
b
2-7/
8" F
loat
Col
lar
0
2-7/
8" F
loat
Sho
e
Open-Hole hydraulic-Set Packers
Sliding-Sleeve Frac Ports
Presenter
Presentation Notes
The diagram shows an example Packers Plus installation. In this case, the frac is pumped down casing and through tubing into an open-hole lateral. The lateral can be divided into multiple sections by open-hole packers. Each frac port is opened in sequence by dropping sized balls and sliding the pot sleeve open by hydraulic pressure. The system can be used for multiple longitudinal or transverse fracs. If a longitudinal frac initiates, and sand is placed across a packer and behind a sliding sleeve, some mechanical problems can be encountered. With hydraulic set packers, a screenout or high pressure spike can set all packers exposed to the live pressure and may break-down the open hole at the packer element. The supplier is continuing to update the hardware.
Wellbore Tangential Stress Exerted by Packer Setting Pressure
6000 psi
2000 psi
50 psi
Presenter
Presentation Notes
The curves on the plot show computed net effective tangential stress at the borehole wall as a function of angle from side of hole. Axial stress is computed for non-penetrating fluid. A positive stress means that a fracture cannot initiate. The net stress must be negative enough to overcome any rock tensile strength. The conditions used for the calculations are a horizontal well at 10,000’ TVD drilled for transverse fracture initiation with a 400 psi stress anisotropy. The curves show the net stress remaining after inflation of three types of packers with different element stresses (presumed to mimic the differential pressure rating). Mechanical packer: 6000 psi net stress Inflatable packer: 2000 psi net stress Swellable packer: 50 psi net stress The calculations show that setting a packer can cause fracture initiation. If the packer stress is high enough to exceed fracture breakdown pressure and contain the frac, then it is high enough to initiate the frac before you start pumping.
• Limited‐entry perfs (pre‐drilled or shot in‐place)– Few holes– Large perf spacing
• Perf Balls– Only plug perfs– No diversion in annulus
• Large sand (gravel) in annulus and perfs– Bridges in frac opening– Diverts in annulus
• Dynamic pressure gradient in annulus• Combination of balls and gravel is recommended
Presenter
Presentation Notes
Uncemented liner completions can be very cost effective, depending on the perforation and diversion method employed. Conventional limited entry treatments can be pumped with either pre-drilled holes or holes shot in-place. If additional holes are shot or drilled, perf balls can be used for diversion if rate and differential pressure are sufficient to seat the balls. Sand placed in the open annulus can create bridges and annular packs that act as external packers. The pressure drop through a sand pack is hundreds of times higher than in the annulus. Even without packing, the ratio of the pressure gradient in the annulus to the gradient in the liner is large. The liner and annulus can be used as concentric paths for effective frac placement.
Various hole-size and liner configurations were evaluated to determine the balance of internal and external pressure gradients. The pressure drop in the liner is calculated using the pipe fiction algorithm currently employed in GOHFER. This method has been field tested in hundreds of frac treatments and gives accurate estimates of pipe friction for various fluid systems. The calculations for these evaluations are made assuming a gelled aqueous polymer fluid with a gel loading of 30-40 lb/Mgal. The annular pressure drop is calculated using the same method where the pipe diameter is replaced by the equivalent hydraulic diameter. This is calculated from the wetted perimeter and cross-sectional flow area of the annulus. The liner is assumed to be un-centered for calculation of effective wetted perimeter. To account for the “pinch point” where the liner and hole are in contact, the wetted perimeter of the hole is multiplied by 0.66 (loss at 120 degrees of arc) and the perimeter of the liner is multiplied by 0.58 (209 degrees of arc).
The pressure drop down the pipe and in the annulus is shown in the plot for various combinations of hole size and liner size. To maintain fluid diversion, and distribute injected fluid down the liner, the pressure drop in the annulus should exceed the pressure drop in the liner at the expected ratio of the inside and outside flow rates. For example, 40 bpm injection in the 4.5” liner generates a pressure drop of about 18-19 psi/100 ft. The pressure drop in the 4.5x6.125 annulus is greater than that for any flow rate down to 12 bpm. So, with no perforation diversion at all, roughly 80% of the injected fluid (at a total pump rate of 50 bpm) will be carried by the liner and only 20% by the annulus. A smaller annulus will increase the effective diversion, if the liner can be placed.
Fluid flow in annulus is always toward fluid exit (open fracture).High dP in annulus allows diversion of fluid from different perfs.Screenout of first-open fracs builds flow resistance. Later slurry diverted to new fractures.Short effective flow path and low conductivity needed.
Presenter
Presentation Notes
The diagram shows a possible mechanism for dynamic diversion fracturing in uncemented liners. Fluid exist open perfs and pressurizes the annulus. The pressure along the annulus and liner should tend to equalize, with possible crossflow at perforations. A fracture may initiate anywhere along the annulus and take fluid until it builds net pressure or sands off. At pack-off an annular bridge forms that diverts fluid to other parts of the lateral. Fracturing continues until the annulus is packed.
A dynamic diversion job pumped on a 4000’ uncemented liner is shown in the treating plot. The multiple sand ramps and high concentration diversion stages are intended to propagate, pack, and divert multiple fracture initiation sites. Each ramp is followed by a clean pad to re-initiate a new frac. While the method does not allow control of initaition sites, it does allow the entire lateral to be treated. Production analysis indicates that wells treated this way can be very efficient compared to other stimulation techniques. Currently some operators are combining this method with external packers and point injections, using sliding sleeves or plug/perf jobs.
• The selection of a stimulation/completion design cannot be made without an accurate reservoir description
• Reservoir geometry and drainage pattern may control well performance
• Reservoir response to stimulation in vertical wells should be understood
• In almost all cases, maximizing productivelateral length is the primary goal
Presenter
Presentation Notes
The decision to undertake a horizontal well development must be made based on a thorough understanding of reservoir properties and production mechanism. This can only be gained by understanding stimulated vertical well performance. There is only so much that fracturing can do, an frequently the larger scale reservoir structure determines drainage area, EUR, and optimum well design. In general, the goal of any successful horizontal development is to maximize contact with the reservoir (both volume and area).
• If the reservoir has high matrix kh– Attempt minimal drilling damage or surface wash– Well azimuth is not critical
• With pre‐existing natural fractures (open or mineralized)– Drill across fractures and attempt to re‐open– Acid‐frac or prop‐frac through uncemented liner
• Low matrix kh or low kv/kh– Must design for propped frac– Longitudinal fracs where vertical well Xf is poor– Transverse fracs only when Xf>250 ft is possible
Presenter
Presentation Notes
For high kh fields, it may be sufficient to drill a clean hole and remove as much drilling damage as possible. In this case, the well need not be fractured and well azimuth is not critical. If the primary production is through joints and fractures, then the well must be oriented to intersect as many as possible and maintain connection to them. In low perm systems with limited existing fracture conductivity, a propped fracture must be used to sustain production.
• Different ideas about when and where it can work– Nearly isotropic horizontal stresses
• Open and extend fractures in all directions• Pumped at high frac rate to maintain injection above leakoff rate
– High horizontal stress anisotropy• Elevate pore pressure and relieve net stress causing massive shear failure over large area
• Requires injection “below frac pressure” and access to large reservoir volume for pore pressure change
• Simul‐frac designs– Enhance shear stimulated volume by pumping into parallel wells simultaneously
Presenter
Presentation Notes
When secondary fractures are the primary target, as in many gas-shale plays, other methods are now frequently being employed. These are not directed at establishing conventional propped fractures, but at enhancing shear slip and large-scale stimulated reservoir volumes. These methods require that unpropped fractures filled with frac fluid must remain conductive during production (at low BHFP) over a long time. In some formations this can be achieved through a combination of the right stress state and high “brittleness” of the rock. These designs are not applicable in all cases.
