MICP QC & Interpretation Workflow
- 1. LR Senergy Mercury Injection Capillary Pressure (MICP) Data Evaluation and Interpretation
- 3. Background • MICP is used extensively for Sw-height and rock typing • LR-Senergy has processed, QC and interpreted over 3000 samples & datasets (no limitation in number of samples) • High Pressure Mercury Injection (HPMI up to ~ 60,000 psi Hg- Air): • significant challenges due to small sample size and high pressures potentially disrupting pore system • Data is in general not quality controlled and difficult to handle efficiently given the extensive data sets • LR-Senergy has developed best practice workflows and QC discrimination criteria to ensure only best quality data used for Sw-Height / Rock Typing Typical sample size 4 – 6 ml bulk volume
- 5. Data Evaluation QC Considerations and Issues 1. Data inventory and loading (input routines customised for every data set type (each lab has several formats)) 2. Sample poroperm distributions check 3. Data availability: parent plug data, raw injection data, chip and parent plug helium porosity, plug permeability, etc 4. Conformance correction verification 5. Injection sample pore volume • must be greater than 0.5 ml • impact of volume errors on Sw much greater 6. Injection sample versus parent plug porosity 7. Curve shape parameters checks • capillary pressure curves • Swir vs. controlling parameters, etc. 8. LR-Senergy has developed software to automate as many of the processes as possible 9. Samples which do not meet the defined criteria are either rejected or flagged as suspect 1 Use data with confidence 2 Sample Suspect - use data with care 3 Reject Sample - unacceptable data or uncertainty Parent plug End trim (injection sample)
- 6. Data Evaluation Workflow Data Inventory Lab Data Availability Check (Data Availability Flags) Independent Conformance Correction estimation Values Analysis (Value Flags) Curve Shape/Trend analysis (Shape Flags) Overall Flags – Final QC Data Base MICP Interpretation*
- 7. Data Evaluation Poroperm Distributions • Ensure MICP samples are representative of net intervals and selection is not biased
- 8. Data Evaluation Conformance Correction • Attempt to minimise subjectivity in conformance and entry pressure selection and consequent injection volume correction Lab Pc vs Sw Cumulative dV/dPc vs Shg Incremental Volume vs Pc Cumulative dV/dPc vs Pc
- 9. Data Evaluation Porosity Comparison: Chip vs. Plug • Reject if injection sample (Hg) porosity and parent plug (He) porosity differ by > ±1.0 p.u. • Should be ±0.5 p.u. (API RP40) or better, but allow for some heterogeneity • Hg may not penetrate all pore throats penetrated by helium molecules Hg He
- 10. Data Evaluation Pc vs. Sw curves and endpoints “Swir” vs k “Swir” vs Pc vs Sw
- 11. Data Evaluation Overall QC Flags (example) • Reject 10% of samples that failed QC • Low rejection rate - generally much higher • Remaining 90% treated with caution as full data lacking • e.g. no He porosity on injection chip for direct comparison with Hg porosity
- 13. Data Interpretation Pc vs. Sw • Based on QC data • Test Swanson (and Thomeer) permeability correlations • Evaluate R35 for rock typing (or any possible R (R20, R25, R30, R40, etc.)) • Evaluate potential strategies for saturation-height modelling and rock typing
- 14. Data Interpretation Field-Tuned Swanson Permeability • (Sb/Pc)A at the apex is the maximum Sb/Pc ratio calculated from all points on the Pc curve. • Assumed to “represent the point where all of the major connected pore spaces controlling permeability have been intruded by mercury”. • Used to predict permeability in MICP samples where permeability measurement not possible 64.1 .427 APc Sb Ka 691.1 399 APc Sb Ka Swanson (1981) (generic coefficients) Tuned to Field Pc data
- 15. Data Interpretation R (pore throat radius) Binning • Traditional Winland R35: pore throat radius corresponding to pressure at which mercury saturates 35% of pore volume. “Point where the pore network becomes interconnected” • LR-Senergy can calculate R from field MICP data at any required mercury saturation, not sticking just to the traditional Winland R35 • Multivariable regression can be used to determine which R produces the best fit for the actual MICP field data • MICP sample Hg porosity and Swanson permeability can be used in the regressions – not just parent plug values 1.0 – 3.0 mm 3.0 – 8.0 mm > 8 mm < 0.1 mm 0.1 – 0.3 mm 0.3 – 1.0 mm R35 Bins (based on field MICP data) )(7796.0)(5783.05699.035 LogLogKRLog air
- 16. Data Interpretation Pore Throat Size Distribution (PSD) by R35 Bins • Pore throat size modality and size distribution discriminated by R35 bins (any R can be modelled)
- 17. Data Interpretation Pc vs. Sw Modelling: J-function • J function improved by • using MICP sample porosity and Swansom permeability • using normalised water saturation (Swn = 1-Swir) • Swir = Sw at 640 ft HAFWL • J function relationships effectively discriminated by R35 bins
- 18. Data Interpretation Thomeer Curve Fitting: Rock Typing • Thomeer model is extensively used to fit the Pc curves (hyperbolic function) for saturation height modelling and rock typing • Fitted curves are defined by the pore structure factor (G), threshold pressure (Pd) and the total interconnected pore volume (Sb∞) (Bulk Volume saturation at infinite pressure ~ Porosity) • LR-Senergy workflow can extract G, Pd and Sb∞ from each one of the QC MICP curves and use the data to check for trends and potential clusters Sb∞: Total interconnected PV at inf. Pc ( ~ ) Pd: Hg/A displacement pressure G: pore geometrical factor K: air permeability (mD) Thomeer, J. H. (1983, April 1). Air Permeability as a Function of Three Pore-Network Parameters. Society of Petroleum Engineers. doi:10.2118/10922-PA
- 19. Data Interpretation Thomeer Curve Fitting: Rock Typing •Thomeer Based Clusters (G, Pd and Sb∞ extracted from field MICP data) G Pd Sb∞