Activation Logging Interpretation Methods Zoran Markovic Chombe Hotwell

Activation Logging Interpretation Methods Zoran Markovic - Chombe Hotwell Handelsges. m. b. H Oedenburger Strasse 6 7013 Klingenbach, AUSTRIA Tel. : +43 (0) 2687 - 48058 | Fax: +43 (0) 2687 – 48059 office@hotwell. at | www. hotwell. at best components | simple design | rigorous testing | no sacrifice to performance

Interpretation Methods • Hotwell Interpretation Method • Chinese Interpretation Method

Hotwell Interpretation Method • Based on cross-correlation between all detectors readings • 16 correlations are automatically done each to each detector including each to itself as a base curve and as an offset curve • All cross-correlations are also supported by regression analysis for each step of curves shifting

Cross-Correlation • Based on multiplication and expressed through: • where: x - base curve value y - offset curve value t - time index l - time shift value tm - time at maximum multiplication

Linear Regression Analysis • in the same time as multiplication in each step it is performed the regression type correlation between base and offset curves with calculation of regression coefficient (r) • Regression coefficient is calculated using following Formula

Correlogram Log • After Multiplication and correlation so called "Correlogram Log" is constructed as on the Figure: • Perfect Correlogram.

Correlogram Log - Complex • Complex Correlogram. • Correlogram surface in the zone of maximum correlation coefficient down to 20% of maximum value is integrated according time.

Velocity Calculation • Velocity (m/min of ft/min) is simple calculated by: Distance Between Two Detectors (meters of feet) Time Between Two Detectors Pick Arrivals (minutes)

Cross-Correlation Matrix • Same kind of calculation is repeated 16 times for each correlations combination of four gamma ray curves. These 16 correlations are filling so called correlation matrix:

Velocity Correction • The calculated velocity is subject to additional correction for the type of flow using either modeled empirical formulas (for 124 mm ID casing) or modeling via Reynolds number. The details about these corrections will be explained in next presentation:

Chinese Interpretation Method • Based on time integration of activated radioactivity picks on particular detector and the distance from the neutron generator, but taking into account the generator fire time as well. • The method is implementation of Chinese algorithm but with some additions as well in case user chooses to apply. • The additions do not change the original algorithm in basic but adds some parameters.

Fire Points Summation • Prior to application of the method all fire points at one stop check points are accumulated. The example of such accumulated curve is on following figure.

Integration Method • The integration is performed from starting time (a - user selectable) to end time (b - user selectable) according to following formula:

Correction Factor Application • Time tm is corrected by subtraction for 2 x. Correction. Factor. The Correction Factor is a meter of experience, of the time of fire and of the detector used. The difference according to original Chinese method is that time here starts fro fire begin and in original method at fire end. So it is necessary also to subtract time of fire from tm. So it follows that: Where: tm - estimated time cf - correction coefficient (0. 35 - changeable by user) tf - fire duration (typically 0. 5 -5 seconds)

Velocity Calculation • Velocity (m/min of ft/min) is simple calculated by: Distance Between Detector and Generator (meters of feet) Corrected Integrated Time (minutes)

Chinese Method Summary • Same kind of calculation may be repeated for all four detectors if sufficient count rate. The average velocity may be taken into account as the velocity for particular stopcheck. • The method is subjected to user selection of correction factor which is dependable on experience for different flowrates and different detectors. • It can be parallel used for the interpretation of flow through different pipes. For example for flowrates inside and outside tubing if dual injection is used.

Case Study – Hotwell Method • • • • • • AFL Processing Report Listing: Date: Thursday, August 21, 2008 ======================== ---------------------Well: Well-1 Field: N/A Company: Hotwell Country: Austria ---------------------======================== --------------------AFL Passes/Stop-Checks Processing Results Chinese Interpretation Method ---------------------Depth Fluid Velocity Flow-Rate m m/min m 3/D 990 19. 81527 303. 1474 1010 12. 46587 190. 7112 1030 8. 269562 126. 5134 1050 3. 651532 55. 86361 1070 2. 044897 31. 28422 1100 0. 813107 12. 43946 ========================

Case Study – Chinese Method • • • • • • AFL Processing Report Listing: Date: Thursday, August 21, 2008 ======================== ---------------------Well: Well-1 Field: N/A Company: Hotwell Country: Austria ---------------------======================== --------------------AFL Passes/Stop-Checks Processing Results Hotwell Interpretation Method --------------------Depth Fluid Velocity Flow-Rate m m/min m 3/D 990 20. 10181 307. 5312 1010 12. 24931 187. 3982 1030 8. 135475 124. 462 1050 3. 786518 57. 92871 1070 2. 058471 31. 49188 1100 0. 7118677 10. 89063 ========================

Case Study – Both Methods

Case Study – Both Methods Results COmparison

Summary • Two different methods are applied in Hotwell HWAFLProc activation flow logging interpretation package. • One is based on cross-correlation between detector readings (Hotwell) • Other on integration and distance from Generator of each detector separately (Chinese).

Summary Continued • Both methods may be used for interpretation and it can be used average result as the final one. • Cross-correlation method is semiautomatic and offers better control while it can not interpret multiple flows (through tubing and through annular space) in same time. • Chinese method can be used for multiple flows while good knowledge and experience regarding selection of parameters is required.

Summary • Shown case study is interpreted using both methods and is showing good results correlation confirming usability of both methods.

Thank You Very Much