Real Time Integration of Reservoir Modeling and Formation Testing. Adriaan Gisolf Reservoir Domain Champion, Norway

Transcription

1 Real Time Integration of Reservoir Modeling and Formation Testing Adriaan Gisolf Reservoir Domain Champion, Norway

2 utline: New Understanding of Reservoirs DFA what is it. DFA & Compositional Variations New Work Flows Detailed Example Analysis Conclusions

3 Key Reservoir Issues I. Compositional Grading: Reservoir fluids are often highly graded and often not in equilibrium. 1. Gravity 2. Biodegradation 3. Active charging 4. Charge history II. Reservoir Architecture Compartments, Sealing Barriers, Baffles 1. No possible physics to image compartments 2. Pressure communication does NT mean flow communication 3. Industry expectations about compartments violate geostatistics 5. Water washing 6. Convection 7. Seal leaking 8. Thermal diffusion Bottles from one column. ani Elshahawi, Shell ibernia: Colors Show Compartments.

4 Number of events Small Geophysical bjects are MUC more numerous Earthquakes Rolloff: Can t measure little ones

5 MDT Downhole Fluid Analysis Fluid Entry Sample Modules LFA / CFA IFA Pump LFA Pump DFA Tools Contamination Phase change GR Composition C2 Density Viscosity Asphaltene

6 DFA uses Downhole ptical Spectroscopy Water vs il Asphaltenes Produce Color il Peak (GR) 0.8 D C4 GR Dead il -C2- Live Crude il Wavelength 1900

7 Depth (m) Discovery well Fluid analysis Pressure Gradients g/cc g/cc g/cc Pressure (bar)

8 Curved Gradient Schematic SPE 89704, SPE Pressure 0.40 g/cc 0.52 g/cc GC Pressure Contact Depth Pressure Contact WC

9 Equation of State Fluid modeling In Equation f State modeling, sample compositions, pvt data and DFA data are used to build a fluid gradient model. Based on a gravity & chemical potential equilibrium. Fluid model can be build in a relatively short time The model contains continuous fluid composition & PVT properties as a function of depth.

10 EoS Fits DFA Data Ethane - Pentane Methane exane + C1 C2-C5 C6+ DFA DATA C1 C2-C5 C6+ C1 C2-C5 C6+ C1 C2-C5 C6+ C1 C2-C5 C6+ 10

11 Logging optimization We can use this knowledge of fluids and fluid grading of a discovery well, to: A) determine fluid connectivity to an offset well, in real time. B) Quality control the DFA data.

12 Connectivity Fluid modeling Assumptions: We assume that the fluid in the discovery well and the offset well are the same Validation Compare measured to predicted DFA compositions Agreement indicates the same fluid column. Differences indicate a fluid compartment. When a compartment / difference from the model is identified real time then the WFT program can be adjusted to investigate. Identification after rig down may leave unanswered questions.

13 Fluid Model Validation It is possible to compare compositional data (C1, C2, C3-C5, C6+, C2), asphaltene gradient, GR, Density, pressure & fluid contacts in real time. igh resolution DFA with quantified accuracy is preferred. Contamination Management

14 Depth, m DFA predictive Modeling Workflow C1 C2-C5 C6+ C on 2 C1 C2-C5 C Station 2 ff trend Measured off on on INJECTR WELL B Predicted Model utput DFA & Sample Analysis Results Fluid Model C1 C2-C Pressure, bar ES Modeling C6+ 2 Station 5 Added to Confirm Trend DISCVERY WELL A

15 Depth, m DFA predictive Modeling Workflow Station 2 ff trend C1 C2-C5 C6+ C on 2 C1 C2-C5 C6+ 1 Extra 5 off 2 on 3 on 4 Measured INJECTR WELL B Predicted Model utput DFA & Sample Analysis Results Fluid Model C1 C2-C Pressure, bar ES Modeling C6+ 2 Station 5 Added to Confirm Trend DISCVERY WELL A

16 Connectivity GCs different by 20 Meters TVD. Two Separate Gas Caps in ne Sand. Can be explained by either Compartment or Gas il Water Well B Lateral Disequilibrium. Well C Well A

17 New Approach Integrating Asphaltene Science New DFA Technology Insitu Fluid Analyzer New Asphaltene Science New Theory & Workflow N ES Flory-uggins

18 TVD, m Color Seems Continuous (neglecting lowest point!) Well A - CFA Well A - LFA Well C - CFA Well C - LFA Well B - CFA Well B - LFA d = 1.3 nm Low Concentration N Asphaltene Like Resins Molecular Distribution 3680 Well B Lowest points Model d = ~1.3 nm M = ~780 g/mol Well A Continuous color lateral connectivity ptical Density at 647 nm igh Concentration

19 Connectivity GCs different by 20 Meters TVD. Two Separate Gas Caps in ne Sand. Can be explained by either Compartment or Gas il Water Well B Lateral Disequilibrium. Well C Well A

20 Conclusions Integration of New Asphaltene Science, Models and Workflows with DFA is a powerful approach to find fluid compositional grading, compartments & connectivity and help to quality control formation tester data acquisition, in real time. The use of this newly developed workflow resulted in: Identification of one station where the flowing fluid was not representative of the native fluid. The conclusion that fluids in the discovery and offset well can not be in equilibrium if the wells both penetrate a single compartment. Identification of a continuous color gradient, suggesting a continuous reservoir in disequilibrium. Subsequent production data confirmed the reservoir is continuous.