FLOWBACK WATERCHEMISTRY - LBCG...TAQA N ORTH L TD. CANADIAN BUSINESS CONFERENCES C ANADIAN S HALE W ATER R ECYCLING AND R EUSE C ONGRESS 2015 16-17 S EPTEMBER 2015 • C ALGARY, A

FLOWBACK WATER CHEMISTRY

FORECASTING FLOWBACK WATER RECOVERY AND ITS SALINITYRESPONSE TO ADEQUATELY PREPARE FOR WATER RECYCLE AND

REUSE

JESSE D. WILLIAMS-KOVACS, UNIVERSITY OF CALGARY AND TAQA NORTH LTD.

CANADIAN BUSINESS CONFERENCES CANADIAN SHALE WATER RECYCLING AND REUSE CONGRESS 201516-17 SEPTEMBER 2015 • CALGARY, ALBERTA

Slide 2

DISCLAIMER

Please note that methods and discussions presented in this presentation

are related to work completed at the University of Calgary and are

therefore not representative of the views of TAQA North Ltd. Further,

examples presented in this presentation are not taken from TAQA North

Ltd. operations and were conducted external to the company.

QUANTITATIVE FLOWBACK ANALYSIS/SALINITY MODELING • JESSE WILLIAMS-KOVACS

Slide 3

OUTLINE

1. Quantitative Flowback Analysis Conceptual Model of Flowback and Analysis Procedure Case Study #1: Commingled Shale Gas Flowback

2. Flowback Water Salinity Factors Effecting Flowback Water Salinity Salinity Response of Select Western Canadian Unconventional Plays

3. Introduction of New Flowback Water Salinity Model4. Key Takeaways5. Discussion Points/Reference Material

6. Flowback Data Gathering and Assessment7. Case Study #2: Commingled Tight Oil Flowback8. Salinity Response of Additional Plays9. Additional Salinity Model Slides10. Load Fluid Recovery11. Impact of HC Breakthrough and Shut-Ins During Flowback



Slide 4

CONCEPTUAL MODEL OF FLOWBACK


CONCEPTUAL MODEL OF FLOWBACKCONCEPTUAL MODEL OF FLOWBACKSlide 5

Simple Bi-Wing Planar

Fractures

Complex Fracture Network

Common Frac Shapes:1. Circular2. Elliptical3. Rectangular

Slide 6

CONCEPTUAL MODEL OF FLOWBACK



Slide 7

ANALYSIS PROCEDURE

• 2 PDA methods used in flowback analysis:1. Analytical Simulation History-match flowback data to estimate key frac parameters before

and after formation fluid breakthrough and estimate frac water recovery2. RTA and Type-Curves BBT single-phase: estimate key hydraulic fracture properties Fracture conductivity (radial flow) BBT half-length (fracture depletion)

• Analysis Procedure:

Deterministic Match

Stochastic Match

Comparison With Other

Data Sources

Diagnostic Plots and BBT RTA

Stochastic Output

• Marcellus Shale dry gas window

• Cased hole completion

• Hydraulically fractured with slickwater in 12 stages, with 3 perf clusters per stage

• Perf clusters spaced at ~ 100 ft

• 6,800 STB of frac fluid and 115 T pumped per stage

• Microseismic on off-setting wells suggests circular fractures with complex geometry (minimal overlap between stages) – Frac Geometry #3

CASE STUDY #1 – OVERVIEWSlide 8


a)

CASE STUDY #1 – RAW DATA AND DIAGNOSTICSSlide 9

Multi-Phase Depletion

b)

Linear Flow Becomes Dominant


Gas production from onset of flowback

CASE STUDY #1 – DETERMINISTIC MATCHSlide 10

Frac FIP ~ 6,700 STB → Maximum

Load Fluid Recovery (~ 8%)

• Good history-match to both phases throughout the flowback period

• Stochastic simulation could be applied to assess the uncertainty associated with our key parameter estimates


c)

Good agreement With flowback half-length estimate

CASE STUDY #1 – ONLINE PRODUCTION DATA AND RTASlide 11

End of water rate measurements


Pf

Slide 12

Minimal water production beyond ~100 days

Single-phase model over-predicts long-term production

CASE STUDY #1 – FORWARD FORECAST

Single-phase model over-predicts cum production by ~20%


Comparable match between 7 and 200 days

Predicted Water Recovery ~ 5,900 STB (~ 7.2% Recovery, ~ 88% of Max)

CASE STUDY #1 – FORWARD FORECASTSlide 13

Multi-Phase Flowback Data CAN Be Quantitatively Analyzed to Estimate Key Frac Properties and Water

Recovery

CASE STUDY #1 – KEY TAKEAWAYSlide 14

Dissolution of salt at the fracture face or in the matrix From deposition or hydrogeological process causing crystallization

Diffusion/Dispersion of ions from high salinity matrix or secondary fractures Secondary fractures typically have higher surface area than primary fractures

(enhanced diffusion)

Advection due to bulk motion of fracture and formation fluids

Precipitation of salt within the fractures i.e. calcium carbonate – influenced by partial pressure of carbon dioxide

Production of high salinity formation water

Measuring individual ion ratios can indicate dominant mechanisms and lab experiments can help determine transport coefficients

Slide 15

FACTORS AFFECTING FLOWBACK WATER SALINITY


Formation Breakthrough


Slide 16

SALINITY RESPONSE – NE BC MONTNEY TIGHT OIL

Significant Formation Water

Diffusion controlled BBT

Mixing controlled ABT

Frac complexity?

• WOR signature suggest formation water production

• Early-time salinity response suggests potential fracture complexity

• BBT salinity response diffusion dominated

• ABT salinity response dominated by mixing with hyper-saline formation water

• No definitive plateau reached

Approaching plateau at ~150,000 ppm?

Well 3 had 2-3x load fluid pumped leading to lower slope and later

plateau


Sharak et al. (2014), SPE 168598-MS

Similar as above

Ghanbari et al. (2013), SPE 167165-MS

Slide 17

SALINITY RESPONSE – HORN RIVER SHALES

• Formation thought to produce minimal formation water

• Diffusion appears to be the dominant mechanism for mass transfer

• The salinity profile and plateau are similar for the Muskwa and Otter Park

• Salinity response in Evie may be due to higher fracture complexity and appears to plateau significantly higher

Plateau reached ~35,000 ppm

Diffusion controlled throughout

Approaching Plateau of ~70,000 ppm?



Slide 18SALINITY MODEL DEVELOPMENT – FULLY COUPLEDAPPROACH (RADIAL FRACTURES)

Repetitive Element

Matrix

Matrix

Matrix

Matrix

A-D-RMass Transfer

Fracture

A-D-RMass Transfer

A-D-RMass Transfer

A-D-RMass Transfer

A-D-RWith Matrix

Inflow

No

Flux

No

Flux

No

Flux

No

Flux

ye/2 wf/2z = 𝟎𝟎

xf

r = rwxe



Matrix: Linear Advection-Diffusion-Reaction Equation𝜕𝜕𝐶𝐶𝑚𝑚 𝑧𝑧, 𝑡𝑡

𝜕𝜕𝑡𝑡= 𝐷𝐷𝑚𝑚 𝑡𝑡

𝜕𝜕2𝐶𝐶𝑚𝑚𝜕𝜕𝑧𝑧2

−𝑞𝑞𝑤𝑤,𝑚𝑚 𝑡𝑡

𝐴𝐴𝑚𝑚,𝑤𝑤 𝑡𝑡

𝜕𝜕𝐶𝐶𝑓𝑓𝜕𝜕𝜕𝜕

+ 𝑘𝑘𝑘 𝑡𝑡 𝐶𝐶𝑚𝑚𝑚𝑚𝑚𝑚 − 𝐶𝐶𝑚𝑚Accumulation Diffusion/

DispersionReaction/Dissolution

Fracture: Radial Advection-Diffusion-Reaction-InflowAdvection

𝜕𝜕𝐶𝐶𝑓𝑓 𝑟𝑟, 𝑡𝑡𝜕𝜕𝑡𝑡

=𝐷𝐷𝑓𝑓 𝑡𝑡𝑟𝑟

𝜕𝜕𝜕𝜕𝑟𝑟

𝑟𝑟𝜕𝜕𝐶𝐶𝑓𝑓𝜕𝜕𝑟𝑟

−1𝑟𝑟𝑞𝑞𝑤𝑤,𝑓𝑓 𝑡𝑡𝐴𝐴𝑓𝑓,𝑤𝑤 𝑡𝑡

𝜕𝜕𝐶𝐶𝑓𝑓𝜕𝜕𝑟𝑟

− 𝑘𝑘 𝑡𝑡 𝐶𝐶𝑓𝑓 +∅𝑚𝑚,𝑤𝑤 𝑡𝑡∅𝑓𝑓,𝑤𝑤 𝑡𝑡

𝐷𝐷𝑚𝑚⁄𝑤𝑤𝑓𝑓 2

𝜕𝜕𝐶𝐶𝑚𝑚𝜕𝜕𝑧𝑧

𝑧𝑧= ⁄𝑦𝑦𝑒𝑒 2

Accumulation Diffusion/Dispersion

Matrix-Fracture TransferAdvection Adsorption/Precipitation

Initial Condition:𝐶𝐶𝑚𝑚 𝜕𝜕, 𝑡𝑡 = 0 = 𝐶𝐶𝑚𝑚𝑚𝑚Boundary Conditions: 1) No Flux at Center of Matrix Block − �𝜕𝜕𝐶𝐶𝑓𝑓

𝜕𝜕𝑚𝑚 𝑧𝑧=0= 0

2) Contnuity at M � F Int − 𝐶𝐶𝑚𝑚 𝑧𝑧 = 𝑦𝑦𝑒𝑒2

, 𝑡𝑡 = 𝐶𝐶𝑓𝑓

Matrix Initial and Boundary Conditions

Initial Condition:𝐶𝐶𝑓𝑓 𝜕𝜕, 𝑡𝑡 = 0 = 𝐶𝐶𝑓𝑓𝑚𝑚Fracture Initial and Boundary Conditions

Boundary Conditions: 1) MB at Well − 𝑉𝑉𝑤𝑤,𝑤𝑤 𝑡𝑡 �𝜕𝜕𝐶𝐶𝑓𝑓𝜕𝜕𝑡𝑡 𝑟𝑟=𝑟𝑟𝑤𝑤

= −𝐴𝐴𝑓𝑓,𝑤𝑤 𝑡𝑡 𝐷𝐷𝑓𝑓 𝑡𝑡 �𝜕𝜕𝐶𝐶𝑓𝑓𝜕𝜕𝑟𝑟 𝑟𝑟=𝑟𝑟𝑤𝑤

2) No Flux or Constant Point Source at Frac Tips− �𝜕𝜕𝐶𝐶𝑓𝑓𝜕𝜕𝑟𝑟 𝑟𝑟=𝑚𝑚𝑓𝑓

= 0 OR 𝐶𝐶𝑓𝑓 𝜕𝜕𝑓𝑓, 𝑡𝑡 = 𝐶𝐶𝑚𝑚𝑚𝑚



System Cannot Be Solved Analytically Without MAJORSimplifications – Assuming No Mobile Water in The

Matrix, Constant Coefficients and Linear Fractures The Solution Would Not Fit on A Slide in 10 pt Font and

Likely Could Not Be Applied in the Real World!!!

So What Can We Do to Find a More Practical Solution?

Dimensionless Variables

Concentration: For BC 1a − 𝐶𝐶𝐷𝐷𝑓𝑓 =𝐶𝐶𝑓𝑓 − 𝐶𝐶𝑓𝑓𝑚𝑚𝐶𝐶𝑚𝑚 − 𝐶𝐶𝑓𝑓𝑚𝑚

For BC 1b − 𝐶𝐶𝐷𝐷𝑓𝑓 =𝐶𝐶𝑓𝑓 − 𝐶𝐶𝑓𝑓𝑚𝑚𝐶𝐶𝑚𝑚𝑚𝑚𝑚𝑚 − 𝐶𝐶𝑓𝑓𝑚𝑚

Time: 𝑡𝑡𝐷𝐷 =𝐷𝐷𝑡𝑡

⁄𝑤𝑤𝑓𝑓 2 2 =4𝐷𝐷𝑡𝑡

𝑤𝑤𝑓𝑓2

Distance: 𝜕𝜕𝐷𝐷 =𝜕𝜕⁄𝑤𝑤𝑓𝑓 2

Slide 21SIMPLIFIED (UNCOUPLED) MODEL DEVELOPMENT –DIMENSIONAL FORM

Fractures: Linear Advection-Diffusion-Reaction Equation

Conceptual Model – Uncoupled Approach

x = ⁄𝑤𝑤𝑓𝑓 2x = 0

𝐶𝐶𝑓𝑓 𝜕𝜕, 𝑡𝑡 = 0 = 𝐶𝐶𝑓𝑓𝑚𝑚

Matrix MatrixPrimary Fracture


∅𝜕𝜕𝐶𝐶𝑓𝑓 𝜕𝜕, 𝑡𝑡

𝜕𝜕𝑡𝑡= 𝐷𝐷∅

𝜕𝜕2𝐶𝐶𝑓𝑓𝜕𝜕𝜕𝜕2

−𝑞𝑞𝑤𝑤𝐴𝐴𝜕𝜕𝐶𝐶𝑓𝑓𝜕𝜕𝜕𝜕

− 𝑘𝑘𝐶𝐶𝑓𝑓Accumulation Diffusion/

DispersionAdvection Adsorption/

Precipitation

Assuming no formation water production for now and constant coefficients

Initial Condition:𝐶𝐶𝑓𝑓 𝜕𝜕, 𝑡𝑡 = 0 = 𝐶𝐶𝑓𝑓𝑚𝑚Boundary Conditions: 1a) 𝐶𝐶𝑓𝑓 𝜕𝜕 = 0, 𝑡𝑡 = 𝐶𝐶𝑚𝑚

1b) �𝜕𝜕𝐶𝐶𝑚𝑚𝜕𝜕𝑡𝑡 𝑚𝑚=0

= 𝑘𝑘′ 𝐶𝐶𝑚𝑚𝑚𝑚𝑚𝑚 − 𝐶𝐶𝑚𝑚

2) �𝜕𝜕𝐶𝐶𝑓𝑓𝜕𝜕𝑚𝑚 𝑚𝑚=

𝑤𝑤𝑓𝑓2

= 0

Initial and Boundary Conditions

2 Simple BC @ M-F Interface

Slide 22SIMPLIFIED (UNCOUPLED) MODEL DEVELOPMENT –DIMENSIONLESS FORM

Conceptual Model – Uncoupled Approach

xD = 1𝜕𝜕𝐷𝐷 = 0

𝐶𝐶𝐷𝐷𝑓𝑓 𝜕𝜕𝐷𝐷, 𝑡𝑡𝐷𝐷 = 0 = 0

Matrix MatrixPrimary Fracture

Initial Condition:𝐶𝐶𝐷𝐷𝑓𝑓 𝜕𝜕𝐷𝐷, 𝑡𝑡𝐷𝐷 = 0 = 0Boundary Conditions: 1a)𝐶𝐶𝐷𝐷𝑓𝑓 𝜕𝜕𝐷𝐷 = 0, 𝑡𝑡𝐷𝐷 = 𝐶𝐶𝐷𝐷𝑚𝑚 = 1

1b) �𝜕𝜕𝐶𝐶𝐷𝐷𝐷𝐷𝜕𝜕𝑡𝑡𝐷𝐷 𝑚𝑚𝐷𝐷=0

= 𝐷𝐷𝑚𝑚,2 1 − 𝐶𝐶𝐷𝐷𝑚𝑚

2) �𝜕𝜕𝐶𝐶𝐷𝐷𝑓𝑓𝜕𝜕𝑚𝑚𝐷𝐷 𝑚𝑚𝐷𝐷=1

= 0

Initial and Boundary Conditions

Dimensionless Groups

𝑃𝑃𝑒𝑒 =𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝑡𝑡𝐴𝐴𝐴𝐴𝐴𝐴𝐷𝐷𝐴𝐴𝐷𝐷𝐷𝐷𝐷𝐷𝐷𝐷𝐴𝐴𝐴𝐴𝐴𝐴 =

𝑞𝑞𝑤𝑤𝑤𝑤𝑓𝑓2∅𝐷𝐷𝐴𝐴

𝐷𝐷𝑚𝑚,1 =𝐴𝐴𝐴𝐴𝐷𝐷𝐴𝐴𝑟𝑟𝐴𝐴𝑡𝑡𝐴𝐴𝐴𝐴𝐴𝐴𝐷𝐷𝐴𝐴𝐷𝐷𝐷𝐷𝐷𝐷𝐷𝐷𝐴𝐴𝐴𝐴𝐴𝐴 =

𝑘𝑘𝑤𝑤𝑓𝑓2

4∅𝐷𝐷

𝐷𝐷𝑚𝑚,2 =𝑆𝑆𝐷𝐷𝑟𝑟𝐷𝐷𝑆𝑆𝐴𝐴𝐴𝐴 𝑅𝑅𝐴𝐴𝑆𝑆𝐴𝐴𝑡𝑡𝐴𝐴𝐴𝐴𝐴𝐴

𝐷𝐷𝐴𝐴𝐷𝐷𝐷𝐷𝐷𝐷𝐷𝐷𝐴𝐴𝐴𝐴𝐴𝐴 =𝑘𝑘𝑘𝑤𝑤𝑓𝑓2

4𝐷𝐷


𝜕𝜕𝐶𝐶𝐷𝐷𝑓𝑓 𝜕𝜕𝐷𝐷 , 𝑡𝑡𝐷𝐷𝜕𝜕𝑡𝑡𝐷𝐷

=𝜕𝜕2𝐶𝐶𝐷𝐷𝑓𝑓𝜕𝜕𝜕𝜕𝐷𝐷

2 − 𝑃𝑃𝑒𝑒𝜕𝜕𝐶𝐶𝐷𝐷𝑓𝑓𝜕𝜕𝜕𝜕𝐷𝐷

− 𝐷𝐷𝑚𝑚,1𝐶𝐶𝐷𝐷𝑓𝑓Accumulation Diffusion/

DispersionAdvection Adsorption/

Precipitation

Assuming Cm >> Cfi –frac water is typically quite fresh

Fractures: Linear Advection-Diffusion-Reaction Equation

Slide 23

SOLUTION PROCEDURE

• Utilizing Laplace Transform Method:

Solve Resulting ODE in Laplace

Space

Average Spatial Solution Over

Domain of Interest

Apply Forward Laplace Transform

to PDE

Numerically Invert Back to Time Domain Using

Stehfest Algorithm

Laplace Domain Solutions For All Boundary Conditions are UGLY!


Slide 24UNCOUPLED SALINITY MODEL – ADVECTION-DIFFUSION-REACTION WITH STATIC BOUNDARY CONDITION

Pure diffusion is a straight-line on the square-root of time plot

passing through the origin


Type-Curve Set For Da = 0 (Advection-Diffusion Only)


Pure diffusion is a straight-line on the square-root of time plot

Advection increases mass

transfer rate


Multi-phase flowback data can be quantitatively analyzed Estimate key frac properties and load fluid recovery

Salinity response from play to play can be highly variable Can also be variable for wells on the same pad

A new model was proposed for matching the salinity response of produced water during flowback Multiple mechanisms considered Simplified approach used to allow analytical solution

Future work will continue with development/application of salinity model More advanced/realistic boundary conditions Investigate additional approximate solution methods Additional simplifications to the fully coupled model

Slide 26

KEY TAKEAWAYS


ACKNOWLEDGEMENTS / THANK YOU / QUESTIONS / DISCUSSION POINTSJesse Williams-Kovacs would like to thank the University of Calgary, TAQA North

Ltd. AND Dr. Christopher Clarkson at the University of CalgaryDISCUSSION POINTS:1. Impact of shutting in wells during flowback and methods to avoid this2. Causes of high salinity of produced water3. Benefits of detailed water analysis

REFERENCE MATERIAL:1. Shale Gas: SPE 162593, SPE 164550, URTeC 21491832. Tight Gas Condensate: SPE 1672313. Tight Oil: SPE 166214, SPE 167232, SPE 175983, SPE 175893 (SPE URC-

Canada, Oct. 2015)4. Stage-by-Stage and Multi-Well Flowback: SPE 1715915. Tight Oil Case Study (flowback + long-term production): The Leading EDGE,

October 2014

CANADIAN BUSINESS CONFERENCES CANADIAN SHALE WATER RECYCLING AND REUSE CONGRESS 201516-17 SEPTEMBER 2015 • CALGARY, ALBERTA

Slide 28

FLOWBACK DATA GATHERING AND ASSESSMENT


Slide 29

FLOWBACK DATA GATHERING AND ASSESSMENT

Data Gathering

Flowback Specific:1) High-frequency rates and flowing pressures (hourly or more frequent – every 15-30 minutes desirable)2) Initial fracture pressure (frac modelling, other estimate – typically 20-50% greater than Pi)3) Indication of fracture geometry (microseismic, frac modelling, experience, etc.)

Optional:1) Method to allocate data to individual stages (i.e. spinner & tracer logs, fibre-optic techniques) 2) Sandface flowing pressures (downhole gauges)3) Salinity of frac fluid4) Estimate of fracture relative permeability curves (lab experiments)

Diagnostics

Quantitative Assessement

Other:1) Wellbore schematic, deviation survey and stimulation information 2) Fluid properties (oil, water and gas analysis)3) Formation temperature4) Estimate of initial reservoir pressure and matrix permeability (DFIT, pre-frac welltest, core)5) Estimates of net pay, porosity and fluid saturation (logs and core)

Optional:1) Detailed PVT analysis2) Estimate of matrix relative permeability curves (lab experiments)3) Offset well analysis

Base Plots:1) qw, qo and qg vs. t (stage-by-stage if available)2) Flowing pressure (surface, downhole or converted) and choke setting vs. t

Primary Diagnostic Plots:1) RNP vs. t or tca (water typically most diagnostic)2) RNP Derivative vs. t or tca (water typically most diagnostic)3) GWR vs. Gp (gas specific)

Other Diagnostic Plots:1) RNP vs. cumulative production (all phases)2) pcf vs. qw, qo and qg3) Gp or Np vs. Wp

Hydraulic Fracture Property Determination and Forecast:1) Rate-transient analysis (radial flow analysis, flowing material balance, Fetkovich type curve)2) Analytical simulation (history-match)3) Forecast long-term production


• Tight oil reservoir in the WCSB (NE BC)

• Cased hole completion

• Hydraulically fractured with hybrid water fracs in 18 stages

• Frac stages spaced at ~ 330 ft

• 1,350 STB of frac fluid and 45 T pumped per stage

• Microseismic suggests circular fractures with bi-wing planar geometry – Frac Geometry #1

CASE STUDY #2 - OVERVIEWSlide 30


CASE STUDY #2 – RAW DATA AND DIAGNOSTICSSlide 31



FcT ~ 150 md-ft

BBT FIP ~ 23,000 STBxf_BBT = 421 ft/stage

FcT ~ 150 md-ft

xf_BBT = 421 ftFcT ~ 150 md-ft

CAST STUDY #2 – RATE TRANSIENT ANALYSISSlide 32

Frac FIP ~ 23,000 STB → Maximum Load Fluid Recovery (~ 92%)

CASE STUDY #2 – DETERMINISTIC HISTORY-MATCHSlide 33

• Good match of water, oil and gas rate prior to gas breakthrough (~8 days)

• Significant uncertainty in analysis due to number of parameters being adjusted to achieve history-match (i.e. xf, fracture permeability, thickness, etc.)

Gas breakthrough into the fractures –oil and water rate over-predicted after this point

Cum

Gas

Pro

d (M

SCF)



CASE STUDY #2 – STOCHASTIC HISTORY-MATCHSlide 34


Slide 35

CASE STUDY #2 – STOCHASTIC HISTORY-MATCH

CASE STUDY #2 – PARAMETER DISTRIBUTIONSSlide 36


CASE STUDY #2 – FORWARD FORECASTSlide 37

Water recovery takes > 1,200 days!




Slide 38SALINITY RESPONSE – NW AB MONTNEY LIQUIDS-RICHTIGHT GAS

• Desiccated formation – possible to have salt precipitated prior to stimulation

• WGR signature suggest negligible formation water production

• Possible salinity measurement error early in time

• Variable plateau reached – possibly due to fracture complexity or longer shut-in in Well 2

Minimal Formation Water

Diffusion/ Reaction

Measurement Issues?

Variable plateau reached

Well 1 ramps up faster




Slide 39

SALINITY RESPONSE – CENTRAL AB MONTNEY TIGHT OIL

• WOR signature suggest formation water production

• BBT salinity response diffusion dominated

• ABT salinity response dominated by mixing with hyper-saline formation water

• Plateau appears to be reached at ~ 150,000 ppm


Diffusion controlled

BBT

Mixing controlled

ABT

Approaching plateau at ~150,000 ppm

Well 2 had 25% more load fluid pumped Leading to lower slope and

later plateau



Slide 40SALINITY RESPONSE – SOUTH PEMBINA CARDIUMTIGHT OIL

• No online water reported – thought to be negligible from off-sets

• Salinity response appears diffusion dominated both BBT and ABT

• Interpretation limited by data quality

• Plateau appears to be reached at ~ 20,000 ppm – likely below connate water salinity

Diffusion controlled throughout

Plateau reached ~20,000 ppmPoor data?

Well 1 ramps up Faster




Slide 41SALINITY RESPONSE – HOADLEY GLAUCONITETIGHT GAS

• WGR signature suggest formation water production

• Early-time salinity response suggests possible measurement issues – missing BBT diffusion?

• ABT salinity response dominated diffusion

• Variable plateau reached – possibly due to fracture complexity and extent of formation water


Measurement issues? –missing impact of BBT

diffusion

Variable plateau reached

Mixing likely dominant ABT

Well 1 ramps up faster



Slide 42SALINITY MODEL DEVELOPMENT – FULLY COUPLEDAPPROACH (LINEAR FRACTURES)

Repetitive Element

Matrix

Matrix

Matrix

Matrix

A-D-RMass Transfer

Fracture

A-D-RMass Transfer

A-D-RMass Transfer

A-D-RMass Transfer

A-D-RWith Matrix

Inflow

No

Flux

No

Flux

No

Flux

No

Flux

ye/2 wf/2z = 𝟎𝟎

xf

x = 𝟎𝟎 xe


Slide 43SALINITY MODEL DEVELOPMENT – FULLY COUPLEDAPPROACH (LINEAR FRACTURES)

Matrix: Linear Advection-Diffusion-Reaction Equation𝜕𝜕𝐶𝐶𝑚𝑚 𝑧𝑧, 𝑡𝑡

𝜕𝜕𝑡𝑡= 𝐷𝐷𝑚𝑚 𝑡𝑡

𝜕𝜕2𝐶𝐶𝑚𝑚𝜕𝜕𝑧𝑧2

−𝑞𝑞𝑤𝑤,𝑚𝑚 𝑡𝑡

𝐴𝐴𝑚𝑚,𝑤𝑤 𝑡𝑡


+ 𝑘𝑘𝑘 𝑡𝑡 𝐶𝐶𝑚𝑚𝑚𝑚𝑚𝑚 − 𝐶𝐶𝑚𝑚Accumulation Diffusion/

DispersionReaction/Dissolution

Fractures: Linear Advection-Diffusion-Reaction-InflowAdvection

𝜕𝜕𝐶𝐶𝑓𝑓 𝜕𝜕, 𝑡𝑡𝜕𝜕𝑡𝑡

= 𝐷𝐷𝑓𝑓 𝑡𝑡𝜕𝜕2𝐶𝐶𝑓𝑓𝜕𝜕𝜕𝜕2

−𝑞𝑞𝑤𝑤,𝑓𝑓 𝑡𝑡𝐴𝐴𝑓𝑓,𝑤𝑤 𝑡𝑡


− 𝑘𝑘 𝑡𝑡 𝐶𝐶𝑓𝑓 +∅𝑚𝑚,𝑤𝑤 𝑡𝑡∅𝑓𝑓,𝑤𝑤 𝑡𝑡

𝐷𝐷𝑚𝑚⁄𝑤𝑤𝑓𝑓 2

𝜕𝜕𝐶𝐶𝑚𝑚𝜕𝜕𝑧𝑧

𝑧𝑧= ⁄𝑦𝑦𝑒𝑒 2

Accumulation Diffusion/Dispersion

Matrix-Fracture TransferAdvection Adsorption/Precipitation

Initial Condition:𝐶𝐶𝑚𝑚 𝜕𝜕, 𝑡𝑡 = 0 = 𝐶𝐶𝑚𝑚𝑚𝑚Boundary Conditions: 1) No Flux at Center of Matrix Block − �𝜕𝜕𝐶𝐶𝑓𝑓

𝜕𝜕𝑚𝑚 𝑧𝑧=0= 0

2) Contnuity at M − F Int − 𝐶𝐶𝑚𝑚 𝑧𝑧 = 𝑦𝑦𝑒𝑒2

, 𝑡𝑡 = 𝐶𝐶𝑓𝑓

Matrix Initial and Boundary Conditions

Initial Condition:𝐶𝐶𝑓𝑓 𝜕𝜕, 𝑡𝑡 = 0 = 𝐶𝐶𝑓𝑓𝑚𝑚Fracture Initial and Boundary Conditions

Boundary Conditions: 1) MB at Well − 𝑉𝑉𝑤𝑤,𝑤𝑤 𝑡𝑡 �𝜕𝜕𝐶𝐶𝑓𝑓𝜕𝜕𝑡𝑡 𝑚𝑚=0

= −𝐴𝐴𝑤𝑤,𝑤𝑤 𝑡𝑡 𝐷𝐷𝑓𝑓 𝑡𝑡 �𝜕𝜕𝐶𝐶𝑓𝑓𝜕𝜕𝑚𝑚 𝑚𝑚=0

→ �𝜕𝜕𝐶𝐶𝑓𝑓𝜕𝜕𝑚𝑚 𝑚𝑚=0

= 0 for 𝑉𝑉𝑊𝑊 = 0

2) No Flux or Constant Point Source - �𝜕𝜕𝐶𝐶𝑓𝑓𝜕𝜕𝑡𝑡 𝑚𝑚=𝑚𝑚𝑓𝑓

= 0 OR 𝐶𝐶𝑓𝑓 𝜕𝜕𝑓𝑓, 𝑡𝑡 = 𝐶𝐶𝑚𝑚𝑚𝑚


Extent of adsorption in fracture impacts max concentration


Slide 45UNCOUPLED SALINITY MODEL – ADVECTION-DIFFUSIONWITH REACTIVE BOUNDARY CONDITION

Increasing rate of surface reaction causes solution

to approach diffusion with static BC at CDmi (0,t) = 1


Slide 46UNCOUPLED SALINITY MODEL – ADVECTION-DIFFUSIONWITH REACTIVE BOUNDARY CONDITION

Increasing rate of surface reaction causes salinity to

increase more rapidly


Slide 47

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

0 100 200 300 400 500 600

WO

R (S

TB/S

TB)

Time (days)

NE BC Montney Long-Term WOR vs Time

NE BC MNTN Hybrid 1 NE BC MNTN Hybrid 2 NE BC MNTN Hybrid 3


WATER RECOVERY – NE BC MONTNEY TIGHT OIL


0

10

20

30

40

50

60

0%

20%

40%

60%

80%

100%

120%72

Hou

rs

7 D

ays

14 D

ays

30 D

ays

90 D

ays

72 H

ours

7 D

ays

14 D

ays

30 D

ays

90 D

ays

72 H

ours

7 D

ays

14 D

ays

30 D

ays

90 D

ays

NE BC MNTN N2 Hybrid 1 NE BC MNTN N2 Hybrid 2 NE BC MNTN N2 Hybrid 3

Oil

Rec

over

y (M

STB

)

Wat

er R

ecov

ery

(%)

NE BC Montney Fluid Recovery

Water Recovery % Oil Recovery Volume

Slide 48

18 StagesSlickwater/Linear Gel Hybrid

22.9 MSTB Load10 Day Shut-In

Planar Fractures



Planar Fractures


61.3 MSTB Load4.5 Day Shut-InPlanar Fractures

WATER RECOVERY – NE BC MONTNEY TIGHT OIL


0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

0 50 100 150 200 250 300 350 400 450 500

WG

R (S

TB/M

MSC

F)

Time (days)

NW AB Montney Long-Term WGR vs Time

NW AB MNTN N2 Foam 1 NW AB MNTN N2 Foam 2NW AB MNTN N2 Xlink 1 NW AB MNTN N2 Xlink 2

WATER RECOVERY – NW AB MONTNEY LIQUIDS-RICHTIGHT

Slide 49

Minimal Formation WaterPossible Frac

Communication With Water-Bearing Formation


0100200300400500600700800900

0%10%20%30%40%50%60%70%80%90%

72 H

ours

7 D

ays

14 D

ays

30 D

ays

90 D

ays

72 H

ours

7 D

ays

14 D

ays

30 D

ays

90 D

ays

72 H

ours

7 D

ays

14 D

ays

30 D

ays

90 D

ays

72 H

ours

7 D

ays

14 D

ays

30 D

ays

90 D

ays

NW AB MNTN N2Foam 1

NW AB MNTN N2Foam 2

NW AB MNTN N2XLink 1

NW AB MNTN N2XLink 2

Gas

Rec

over

y (M

MSC

F)

Wat

er R

ecov

ery

(%)

NW AB Montney Fluid Recovery

Water Recovery % Gas Recovery Volume

WATER RECOVERY – NW AB MONTNEY LIQUIDS-RICHTIGHT GAS

Slide 50

12 StagesN2 Foam


Planar Fractures

12 StagesN2 Foam


Planar Fractures

13 StagesN2 XLink Gel


13 StagesN2 XLink Gel


Planar Fractures


0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

0 50 100 150 200 250 300 350 400 450

WO

R (S

TB/S

TB)

Time (days)

Central AB Montney Long-Term WOR vs Time

Central AB MNTN Foam 1 Central AB MNTN Foam 2

WATER RECOVERY – CENTRAL AB MONTNEY VOLATILETIGHT OIL

Slide 51


Spikes in water production –measurement problems or frac’d

into water-bearing zone


0

20

40

60

80

100

120

140

160

0%

20%

40%

60%

80%

100%

120%

140%

160%

72Hours

7 Days 14Days

30Days

90Days

72Hours

7 Days 14Days

30Days

90Days

Central AB MNTN N2 Foam 1 Central AB MNTN N2 Foam 2

Oil

Rec

over

y (M

STB

)

Wat

er R

ecov

ery

(%)

Central AB Montney Fluid Recovery


WATER RECOVER – CENTRAL AB MONTNEY VOLATILETIGHT OIL

Slide 52

23 StagesN2 Foam


Planar Fractures

20 StagesN2 Foam



WATER RECOVERY – SOUTH PEMBINA CARDIUM TIGHT OILSlide 53

Long-Term Water Data Not Gathered – Appears Minimal Based on Offset Production


0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

0%

5%

10%

15%

20%

25%

30%

35%

72 Hours 5 days 72 Hours 5 days

South Pembina CRDM Foam 1 South Pembina CRDM Foam 1

Oil

Rec

over

y (M

STB

)

Wat

er R

ecov

ery

(%)

South Pembina Cardium Fluid Recovery


WATER RECOVERY – SOUTH PEMBINA CARDIUM TIGHT OILSlide 54

14 StagesN2 Foam


Planar Fractures

14 StagesN2 Foam


Planar Fractures


0

20

40

60

80

100

120

140

160

180

200

0 100 200 300 400 500 600 700 800 900

WG

R (S

TB/M

MSC

F)

Time (days)

Hoadley Glauc Long-Term WGR vs Time

Hoadley Glauc N2 Foam 1 Hoadley Glauc N2 Foam 2

WATER RECOVERY – HOADLEY GLAUCONITE TIGHT GASSlide 55


Spikes in water production –measurement problems or frac’d

into water-bearing zone


0

10

20

30

40

50

60

70

0%

20%

40%

60%

80%

100%

120%

140%

72Hours

7 Days 14 Days30 Days90 Days 72Hours

7 Days 14 Days30 Days90 Days

Central AB MNTN N2 Foam 1 Central AB MNTN N2 Foam 2

Gas

Rec

over

y (M

MSC

F)

Wat

er R

ecov

ery

(%)

Hoadley Glauc Fluid Recovery


Slide 56

12 StagesN2 Foam


Planar/Complex Fractures

12 StagesN2 Foam


Planar/Complex Fractures

WATER RECOVERY – HOADLEY GLAUCONITE TIGHT GAS


0

5

10

15

20

25

30

35

40

45

50

0 10 20 30 40 50 60 70

WG

R (S

TB/M

Msc

f)

Time (days)

Marcellus Shale Long-Term WGR vs Time

Marcellus Slickwater 1 Marcellus Slickwater 2

WATER RECOVERY – MARCELLUS SHALE GASSlide 57

Negligible Formation Water


0

50

100

150

200

250

300

350

0%

1%

2%

3%

4%

5%

6%

7%

72Hours

7 Days 14 Days30 Days60 Days 72Hours

7 Days 14 Days30 Days60 Days

Marcellus Shale Slickwater 1 Marcellus Shale Slickwater 2

Gas

Rec

over

y Vo

lum

e

Wat

er R

ecov

ery

(%)

Marcellus Shale Fluid Recovery


Slide 58

12 StagesSlickwater


Complex Fractures

12 StagesSlickwater


Complex Fractures

WATER RECOVERY – MARCELLUS SHALE GAS


Slide 59

UNDERSTANDING LOAD FLUID RECOVERY

1. Long shut-ins reduce recovery Loss of frac energy Gravity segregation in fractures

2. Foam fracs increase water recovery Maximum frac energy Minimum water requirement Slickwater typically has lowest energy/recovery

3. Fracture complexity reduces recovery Water retention in secondary fractures (low conductivity)

4. Shales reservoirs have low load fluid recovery relative to sandstone/siltstone reservoirs Water imbibition into matrix due to capillary pressure Water intake by clays

5. High drawdowns typically increase load fluid recovery May damage fractures and impact long-term performance



IMPACT OF HYDROCARBON BREAKTHROUGH AND SHUT-INSSlide 60

Single-phase flow

Gas Break-through reduces xf,eff

Shut-in reduces xf,eff

Example From SPE 119894

Tight Gas

Shut-in reduces xf,eff

Tight Gas Condensate

a)

Gas Break-through reduces xf,eff

Oil Break-through reduces xf,eff

Tight Oil

b)

c)

d)

Slide 61

Flowback Operations Impact Long-Term Productivity

IMPACT OF HYDROCARBON BREAKTHROUGH AND SHUT-INS

Documents

FLOWBACK WATERCHEMISTRY - LBCG...TAQA N ORTH L TD. CANADIAN BUSINESS CONFERENCES C ANADIAN S HALE W ATER R ECYCLING AND R EUSE C ONGRESS 2015 16-17 S EPTEMBER 2015 • C ALGARY, A