PRESENTATION TITLE · 2019. 6. 19. · North (ft) Plan View: Gabriel A 1H 16772 ft 17711 ft 18368 ft EOW 3400 3600 3800 4000 4200 4400 4600 4800 5000 5200-6800-6600-6400-6200-6000-5800-5600-5400

Industry Steering Committee onWellbore Survey Accuracy

Wellbore Positioning Technical Section

48th General MeetingSept 27th, 2018Dallas, USA

PRESENTATION TITLE

1

• Declination Error at Depth: A Comparison Study of Gyro vs. MWD Surveys

• Chad Hanak

Industry Steering Committee onWellbore Survey Accuracy

Wellbore Positioning Technical Section

48th General MeetingSept 27th, 2018Dallas, USA

• Chad Hanak• President• September 27, 2018• Superior QC

Speaker Information

2

Speaker Bio• President of Superior QC

• A Patterson-UTI Company• Survey FDIR

• Past Experience: • NASA• Baker Hughes

• University of Texas• Ph.D. in Aerospace Engineering• Specialized in Guidance, Navigation, and Control

• Based in Houston• Expertise

• Wellbore navigation (survey correction)• Automation• Machine Learning

3

https://www.superiorqc.com/

48th General MeetingSept 27th, 2018Dallas, USA Wellbore Positioning Technical Section

Why Are You Re-correcting Surveys!?

4


35° Turn in Lateral Results in Back Corrections5

-4000 -2000 0 2000 4000 6000 8000

East (ft)

-9000

-8000

-7000

-6000

-5000

-4000

-3000

-2000

-1000

0

Nor

th (f

t)

Plan View: Gabriel A 1H

16772 ft

17711 ft

18368 ft

EOW

3400 3600 3800 4000 4200 4400 4600 4800 5000 5200

East (ft)

-6800

-6600

-6400

-6200

-6000

-5800

-5600

-5400

Nor

th (f

t)

Plan View: Gabriel A 1H

16772 ft

17711 ft

18368 ft

EOW


AMI & Twist Estimates Gain Observability in Turn6

1.65 1.7 1.75 1.8 1.85 1.9 1.95 2 2.05 2.1 2.15

MD (ft) 10 4

-700

-650

-600

-550

-500

-450

-400

-350

-300

-250

Axia

l Mag

netic

Inte

rfere

nce

(nT)

Axial Magnetic Interference Estimate Thru the Turn

Estimate

3- Uncertainty

1.65 1.7 1.75 1.8 1.85 1.9 1.95 2 2.05 2.1 2.15

MD (ft) 10 4

-0.1

-0.05

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

Twis

t (de

g)

Twist Estimate Thru the Turn

Estimate

3- Uncertainty

Median Estimate from QDC Roll Test

Turn Straight Hole

Turn Straight Hole


A 5° Lateral Turn Example Using Multi-Station Analysis (MSA)7


Declination Error at DepthHow Do the IFR and BGGM Error Models Fare?

8


Standard MWD9

Azimuth Referenced to

True North


Grid North

MWD Sensor

Raw Magnetic Azimuth

Azimuth Referenced to Magnetic North

Grid Convergence+= Declination -

IFR

Reference Btotal ~ 50 nTDip ~ 0.1°

Reference Dec ~ 0.16°


Magnetic SurveyCorrections

10


True North


Grid North

MWD Sensor

Raw Magnetic Azimuth

Azimuth Referenced to Magnetic North

Grid Convergence+= Declination -

IFR

Reference Btotal ~ 50 nTDip ~ 0.1°

Reference Dec ~ 0.16°

Survey Correction

Corrected Magnetic Azimuth


How Accurate is IFR Data at Depth?11

IFR data frequently comes from aero-mag surveys, hundreds of feet above the ground.

May be checked at ground level for accuracy.

IFR error level at depth is not well known.


Raw MWD, Corrected MWD, and Gyro Final Positions (Downhole View)

12

Gyro TD MWD TD

Upper Left 12,545 ft 17,886 ft

Upper Right 16,907 ft 19,016 ft

Bottom 14,976 ft 19,957 ft• Blue represents raw MWD surveys• Orange represents FDIR corrected surveys• Green represents gyro surveys


Disagreement Between Gyro and Corrected MWD Can be Attributed to Three Sources

13

MWD Correction

Error

Magnetic Declination

Error

GyroError

Only one error source is global to all the wells on a pad


After Removing Declination Error14

Dec. Diff. from HDGM

HDGM 4.98 deg --

IFR 4.95 deg -0.03 deg

Estimated from Gyro

5.23 deg 0.25 deg

• Orange represents MWD Corrections with declination error removed based on downhole measurement with gyro comparison

• Green represents gyro surveys

• IFR Declination error estimated at 0.28°


Declination Error Study (5 Multi-Well Pads & 3 Individual Wells)15

Well/Pad Number of Wells/Gyros

Azimuth IFR Dec. Error

BGGM Dec. Error

Dec. Agreement (IFR-BGGM)

Pad 1 3 290° 0.29° 0.55° -0.33°

Pad 2 3 280° 0.28° 0.24° -0.11°

Pad 3 4 305° 0.01° -0.17° -0.03°

Pad 4 3 90° -0.59° -0.64° -0.04°

Pad 5 2 320° 0.17° 0.22° -0.14°

Well 1 1 165° 0.16° 0.21° -0.05°

Well 2 1 90° -0.11° -0.02° -0.12°

Well 3 1 270° -0.79° -1.36° 0.08°

IFR Error Model Dec. 1-σ

BGGM Error Model Dec. 1-σ

0.16° 0.42°


Are the Downhole Results Consistent with the Error Model Declination Magnitudes

16

• Chi-Square metric with known mean and variance• Rejection of the Null Hypothesis at 1% probability or less

Chi-Square Metric

Probability of Agreement with

Error Model

Full Data Set 46.73 0.000017%

Worst Point Removed 22.35 0.22%

IFR: from error model, μ = 0, σ = 0.16°

Chi-Square Metric

Probability of Agreement with

Error Model

Full Data Set 15.79 4.5%

Worst Point Removed 5.13 64%

BGGM: from error model, μ = 0, σ = 0.42°

Downhole declination disagreement with BGGM error model not statistically significant

Downhole declination disagreement with IFR error model is statistically significant


Fitting the Data to a Gaussian Cumulative Distribution Function (CDF)

17

-0.8 -0.6 -0.4 -0.2 0 0.2 0.4

Declination Error (deg)

-1.5

-1

-0.5

0

0.5

1

1.5

yk

, Inv

erse

CD

F (d

eg)

Inverse Gaussian CDF vs. Sample Declination Errors

Sample Declination Error

Best Fit Line

IFR


Comparison to Error Model Magnitudes Based on CDF Fit18

IFR Declination1-σ

BGGM Declination1-σ

Error Model Value 0.16° 0.42°

Estimated Value 0.54° 0.82°

% of EM Value 3.4x 2.0x

CDF Fitting Results

IFR Declination1-σ

BGGM Declination1-σ

Error Model Value 0.16° 0.42°

Estimated Value 0.43° 0.50°

% of EM Value 2.7x 1.2x

CDF Fitting Results, Excluding Worst Point

IFR error model appears optimistic in terms of declination by about 3x

BGGM error model may be somewhat optimistic in terms of declination

IFR shown to be an improvement over BGGM (recent BGGM improvements not considered)


Survey Corrections are Used to Fix Spacing Between Parallel Wells Originating from the Same Surface Location

19

Standard Magnetic Reference Field Model + MSA or FDIR

Standard Magnetic Reference Field Model

No CorrectionsSurvey Corrections Applied


IFR is Used to Fix Global Rotational Shift of All Wells Originating from the Same Surface Location

20

Standard Magnetic Reference Field Model

IFR Magnetic Reference Field Model

IFR Corrections AppliedStandard Geomagnetic Model


21

Conclusion Regarding Downhole Declination Uncertainty

• IFR error model• Appears to be optimistic based on downhole data• Statistically significant result• Uncertainty may be 3x the modeled value

• BGGM error model• No statistically significant disagreement with downhole data• May still be somewhat optimistic

• Results call into question anti-collision scans (IFR especially)• IFR still shown to be more accurate than BGGM

• Recent BGGM improvements not considered

PRESENTATION TITLESpeaker InformationSpeaker BioWhy Are You Re-correcting Surveys!?35° Turn in Lateral Results in Back CorrectionsAMI & Twist Estimates Gain Observability in TurnA 5° Lateral Turn Example Using Multi-Station Analysis (MSA)Declination Error at DepthStandard MWDMagnetic Survey�CorrectionsHow Accurate is IFR Data at Depth?Raw MWD, Corrected MWD, and Gyro Final Positions �(Downhole View)Disagreement Between Gyro and Corrected MWD �Can be Attributed to Three SourcesAfter Removing Declination ErrorDeclination Error Study (5 Multi-Well Pads & 3 Individual Wells)Are the Downhole Results Consistent with the Error Model Declination MagnitudesFitting the Data to a Gaussian �Cumulative Distribution Function (CDF)Comparison to Error Model Magnitudes Based on CDF FitSurvey Corrections are Used to Fix Spacing Between �Parallel Wells Originating from the Same Surface LocationIFR is Used to Fix Global Rotational Shift of All Wells Originating from the Same Surface LocationConclusion Regarding Downhole Declination Uncertainty

Documents

PRESENTATION TITLE · 2019. 6. 19. · North (ft) Plan View: Gabriel A 1H 16772 ft 17711 ft 18368 ft EOW 3400 3600 3800 4000 4200 4400 4600 4800 5000 5200-6800-6600-6400-6200-6000-5800-5600-5400