Using Seismic Inversion and Net Pay to Calibrate Eagle Ford Shale Producible Resources

Bo Chen1, Dhananjay Kumar1, Anthony Uerling2, Sheryl Land2, Omar Aguirre2, Tao Jiang1, and Setiawardono Sugianto1

1BP America, 501 Westlake Park Blvd., Houston, Texas 77079

2BP America, 737 N. Eldridge Pkwy., Houston, Texas 77079

GCAGS Explore & Discover Article #00104* http://www.gcags.org/exploreanddiscover/2016/00104_chen_et_al.pdf Posted September 13, 2016.

*Abstract published in the GCAGS Transactions (see footnote reference below) and delivered as an oral presentation at the 66th Annual GCAGS Convention and 63rd Annual GCSSEPM Meeting in Corpus Christi, Texas, September 18– 20, 2016.

ABSTRACT

The production outcome of an Eagle Ford well depends on many factors including reservoir properties. This paper discusses an utilization of seismic attributes to predict them and identify the high production potential areas. Multiple studies have identified the sweet spots in unconventional plays, however not many have been directly correlated with and confirmed by the production data. We studied the production data and analyzed the reservoir characteristics associated with it taking into account the operation procedures. It was found that the producible resources lie in the thicker and more porous intervals. Both thickness and porosity can be determined with seismic data. Seis- mic inversion is utilized for porosity prediction and seismic net pay for ‘net’ thickness prediction with higher porosity. The seismic net pay, confirmed by the blind well log data, has been used to predict the producible resources for future wells in the Eagle Ford shale in South Texas. The study includes the following four steps. First, a correlation between the petrophysical net pay and the estimated production volume is established. Well logs are analyzed to establish a method to predict net pay from seismic data. Sec- ond, post migration seismic data conditioning is applied to improve seismic data quality, by attenuating noises, flattening the gathers, and balancing the frequency spectrum and amplitude across offsets. Good quality seismic data are required for seismic net pay estimation. Third, using the conditioned seismic data, colored inversion is applied to invert the reflectivity data to relative acoustic impedance. Acoustic impedance is in- versely proportional to porosity and is used to predict porosity in the lower Eagle Ford Shale. Finally, seismic net pay is calculated by detuning the relative acoustic impedance and integrating over the gross thickness intervals. To quality control (QC) the results, the predicted seismic net pay is compared with well log data and estimated production data. We found that seismic net pay in the lower Eagle Ford as an indicator of its reservoir quality. The reliable estimation of seismic net pay requires an understanding of the rock properties, good quality well data, seismic data conditioning, well calibrated hori- zons, and accurate seismic inversion for impedance followed by porosity prediction.

Originally published as: Chen, B., D. Kumar, A. Uerling, S. Land, O. Aguirre, T. Jiang, and S. Sugianto, 2016, Using seismic inversion and net pay to calibrate Eagle Ford Shale producible resources: Gulf Coast Association of Geological Societies Transactions, v. 66, p. 927. 1 Using Seismic Inversion and Net Pay to Calibrate Eagle Ford Shale Producible Resources

B. Chen, D. Kumar, A. Uerling, S. Land, O. Aguirre, T. Jiang, S. Sugianto BP America Inc.

1 Objectives

• Identify key petrophysical properties driving production

• Predict such rock properties from seismic

2 Outline

• Studied area • Petrophysical analysis • Seismic data conditioning • Seismic (coloured) inversion • Seismic net pay • Conclusions

3 Production Potential Prediction

R2 =0.726 =0.728

)

bcf

EUR ( EUR EUR ( EUR

Log High Quality Shale Thickness (ft) Seismic Net Pay (ft)

Log and seismic derived high quality shale (HQS) thickness correlate well with EUR when operation is the same.

4 Studied Area

Studied area is located in the dry gas window,

Seismic Survey western portion of Eagle Ford play.

Petrophysical Analysis

) Colored by wells

1000 10 nD

100 10 1 5

0.1 (%) TOC 0.01 0.001 Core Permeability Permeability Core ( 0 5 10 15 0 5 10 15 20 Porosity (%) Porosity (%)

High quality shale (HQS) interval: shale intervals with higher porosity, higher permeability, higher TOC, higher gas saturation, lower clay content.

6 Rock Quality versus EUR

Gross intervals

R² = 0.38

) bcf

EUR ( EUR HQS intervals

R² = 0.73

TGIP (bcf/section)

) Seismic shows more porous and

thicker rocks than bcf

nearby pilot EUR ( EUR Resources in High Quality Shale form stronger correlation with EUR with the same completion designs. Close to fault TGIP (bcf/section)

7 Petrophysical Analysis Well A Well B Well C Well D Gamma HQS Gamma HQS Gamma HQS Gamma HQS Ray PHIT Perm flag Ray PHIT Perm flag Ray PHIT Perm flag Ray PHIT Perm flag

50ft

154 68 76 165 TGIP (bcf/section) 137 22 74 134

Gross Interval

8 HQS Interval Seismic Data Conditioning Before (Hz) 20 4 0 6 0 8 0 Before (dB) Before - 10 LEF Instantaneous Amplitude 100 ms - 20 high - 30 UEF low LEF 3km

Buda Mismatch 1km After (Hz) 20 4 0 6 0 8 0 LEF Instantaneous Amplitude AfterAfter (dB) - 10 high

100 ms - 20 - 30 low UEF 3km LEF Post-migration seismic data Buda Doublet 1km conditioning reduced noise, increased the resolution and improved the Red: trough amplitude fidelity.

9 Blue: peak Rock physics relationships

R² = 0.78

0.12

0.08 Porosity

0.04

25,000 35,000 45,000 P-wave Acoustic Impedance ((ft/sec)*(g/cc))

Porosity forms a good correlation with P-wave acoustic impedance (AI), which can be derived confidently from seismic data.

10 Seismic (coloured) Inversion Inversion operator

Coloured inversion (CI) provides bandlimited (relative) impedance

(Lancaster & Whitecombe, 2000)

MBI CI 20k 60k -12k 12k Inversion QC with wells Coloured inversion (CI) and model-based inversion (MBI) generated similar results.

Log: red Seismic: blue (Kumar et al., 2014) 11 Seismic Net Pay Gamma Seismic Modeled Ray Porosity AI Log AI 100% net AI

HQS thickness Petrophysical net-to-gross = gross reservoir thickness average impedance value over reservoir Geophysical net-to-gross = impedance value at 100% net (HQS)

(Connolly, 2010) 12 Seismic Net Pay

Wedge model

Apparent ( Thickness

Top 1 Max seismic net to gross 0 Average band-limited impedance

Apparent thickness 150 ) 50 0.8 ms 0.6 100 100 150 0.4 Time ( Time 50

200 Base 0.2 Seismic Seismic to Net Gross data range data range ms 0 0 50 100 150 200 ) 0 50 100 150 200 Time Thickness (ms) Time Thickness (ms)

average seismic impedance Seismic net-to-gross = average modeled impedance for 100% net (HQS)

true thickness (ms) × apparent thickness (ms)

Seismic net pay = seismic net-to-gross × apparent thickness

(Connolly, 2007) 13 Seismic Maps Time thickness Inverted porosity

0 0 10 5 20 30 10 3km 3km 40 (ms) 15 (%)

Seismic net pay

Seismic net pay more directly 0 characterizes the sweet spot 50 distributions than thickness or 100 porosity maps alone. 150 3km 200 (ft)

14 QC with Well Logs 0.15

0.12

0.09 Log porosity

0.06 Seismic porosity Porosity 0.03 0 0 2 4 6 8 10 12 14 Well index

) R =0.79 ft

Seismic estimated porosity and net pay match with well log HQS Thickness Thickness HQS ( values with high accuracy. Seismic Net Pay (ft)

15 Seismic Prediction of EUR

R2 =0.726 =0.728

)

bcf

EUR ( EUR EUR ( EUR

Log HQS thickness (ft) Seismic Net Pay (ft)

The seismic net pay correlation to EUR is as good as the correlation with well log HQS thickness.

16 Conclusions

• TGIP in HQS (high quality shale) intervals correlates well with EUR.

• Seismic (coloured) inversion and seismic net pay methods generate reliable porosity and HQS thickness.

• Seismic derived HQS can predict EF shale sweet spots.  Plan new wells  Identify refrac candidates  Appraise completion trials

17 Acknowledgements

• BP Lower 48 and BP Upstream Technology

• Numerous BP colleagues’ suggestions and feedback, including C. Sullivan, W. Simpson, C. Lazos, A. Cunningham, J. Zhang, S. Davis, S. Li, L. Sarle, M. Blome, etc.

18 References

• Connolly, P., 2007, A simple, robust algorithm for seismic net pay estimation: The Leading Edge, 26, 1278-1282, doi: 10.1190/1.2794386 • Connolly, P. A., 2010, Robust workflows for seismic reservoir characterization: SEG Distinguished Lecture. • Kumar, D., Sugianto, H., Li, S., Patel, H. and Land S., 2014, Using relative seismic impedance to predict porosity in the Eagle Ford shale: SEG Technical Program Expanded Abstracts, 2688-2692, doi: 10.1190/segam2014-0473.1 • Lancaster, S. and D. Whitecombe, 2000, Fast-track Coloured Inversion: SEG expanded abstract, doi: 10.1190/1.1815711

19 Questions