href="#ulink_046a5627-3119-573a-b609-6199c6fe80b0">Figure 3.2 Expected versus posterior false discovery rate for two sizes of the t...Figure 3.3 Histograms of area under posterior ROC Curve (AUC) for three anomaly ...Figure 3.4 AUC histograms and quantile regions calculated from 1000 pairs of tra...Figure 3.5 Median of posterior AUC for three AD classifiers as a function of the...Figure 3.6 Quantile width of AUC distribution calculated on anomaly detection re...Figure 3.7 Sparsity values on the records of the training and test sets. Horizon...Figure 3.8 Anomaly detection. Histograms of posterior true discovery rate (TDR) ...Figure 3.9 Histograms of anomalyIndicator for three types of classifiers. (a)-ad...4 Chapter 4Figure 4.1 Type of Kerogen present in source rocks.Figure 4.2 It shows kerogen conversion and maturity (Tmax).Figure 4.3 Bulk compositions of oils and bitumen studied.Figure 4.4 GC-Fingerprint of a carbonate-derived oil.Figure 4.5 GC-Fingerprint of a shale facies derived oil.Figure 4.6 GC- Fingerprint of Tertiary Oil No. 3 which is severely biodegraded.Figure 4.7 GC- Fingerprint of Cretaceous Oil No. 3 which is not biodegraded.Figure 4.8 Triterpane fingerprints of carbonate-derived oil.Figure 4.9 Triterpane fingerprint of shale-derived oil.Figure 4.10 Cross plot of dibenzothiophene/phenanthrene versus pristane/phaytane...
5 Chapter 5Figure 5.1a Prototype single-siren tool (assembled).Figure 5.1b Prototype single-siren tool (disassembled).Figure 5.2a Example MWD collar used for siren frequency and source placement opt...Figure 5.2b Drillpipe p/Dp to 12 Hz.Figure 5.2c Drillpipe p/Dp to 50 Hz.Figure 5.2d Drillpipe p/Dp to 100 Hz.Figure 5.3a Three step pulse recovery in noisy environment (pressure, vertical; ...Figure 5.3b Three step pulse recovery (very noisy environment).Figure 5.4a Early 1980s “stable closed’ siren (left) and improved 1990s “stable-...Figure 5.4b Streamline traces for erosion analysis.Figure 5.4c Velocities for erosion and pressure analysis.Figure 5.5a Short “hydraulic” wind tunnel system.Figure 5.5b Very long “acoustic” wind tunnel.Figure 5.5c A pair of ganged or tandem mud sirens.Figure 5.5d Some sirens tested in wind tunnel.Figure 5.5e Evaluation of hub convergence effects on signal strength and torque.Figure 5.5f Flow straighteners (PVC tubing) for upstream and downstream use.Figure 5.5g Flow meter.Figure 5.5h Siren test section with differential transducers.Figure 5.5i Real-time data acquisition and control system.Figure 5.5j Torque, position and rpm counter.Figure 5.5k Short wind tunnel, “bird’s eye” view.Figure 5.5l Test shed window overlooking long wind tunnel.Figure 5.5m Piezoelectric transducer closest to siren for constructive interfere...Figure 5.5n Distant multiple transducer array setup.Figure 5.5o Fireworks for low frequency noise generation, when all else is unava...Figure 5.6a Siren Dp vs ω at with flow rate fixed.Figure 5.6b Low-frequency (10 Hz) long wind tunnel data.Figure 5.6c Low-frequency (10 Hz) signal recovery.Figure 5.6d High-frequency (45 Hz) long wind tunnel data.Figure 5.6e High-frequency (45 Hz) signal recovery.Figure 5.7a Method 4-3, A-002 (8 feet).Figure 5.7b Method 4-3, A-003 (12 feet).Figure 5.7c Method 4-3, A-004 (16 feet).Figure 5.7d Method 4-3, A-005 (20 feet).Figure 5.7e Method 4-3, B-002 (8 feet).Figure 5.7f Method 4-3, B-004 (16 feet).Figure 5.7g Method 4-3, A-002 (8 feet).Figure 5.7h Method 4-3, A-003 (12 feet).Figure 5.7i Method 4-3, A-004 (16 feet).Figure 5.7j Method 4-3, A-005 (20 feet).Figure 5.7k Method 4-3, B-003 (12 feet).Figure 5.7l Method 4-3, B-006 (24 feet).Figure 5.7m Method 4-3, A-002 (8 feet).Figure 5.7n Method 4-3, A-004 (16 feet).Figure 5.7o Method 4-3, A-006 (24 feet).Figure 5.7p Method 4-3, B-006 (24 feet).Figure 5.7q Method 4-3, B-008 (32 feet).Figure 5.8a Method 4-4, Run C-1.Figure 5.8b Method 4-4, Run C-2.Figure 5.8c Method 4-4, Run C-3.Figure 5.8d Method 4-4, Run C-4.Figure 5.8e Method 4-4, Run C-5.
6 Chapter 6Figure 6.1 Histograms of the number of clusters and the number of records in ind...Figure 6.2 Clustering of a randomized test set that includes regular and anomalo...Figure 6.3 Number of appearances of individual records in different clusters of ...Figure 6.4 Histograms of three prior anomaly parameters.Figure 6.5 Histograms of three posterior parameters characterizing presence or a...Figure 6.6 Anomaly indexes of individual records. High permeability anomaly.Figure 6.7 Histograms of three prior anomaly indexes.Figure 6.8 Histograms of three posterior anomaly indexes.Figure 6.9 Values of anomaly index of individual records in gas-sand and brine-s...
7 Chapter 7Figure 7.1 Dissimilarities for 14 parameters in four combinations of conditions....Figure 7.2 Histograms of the values of the Group_A parameter calculated for gas-...Figure 7.3 Histograms of the Group_A parameter calculated for brine-sand and sha...Figure 7.4 ROC curves for four high performing parameters. Method: LDA.Figure 7.5 Mean area under ROC curve for 14 parameters. Area under Rock Curve av...Figure 7.6 Group_A vs. Poisson’s ratio cross section of gas sand vs. brine sand ...Figure 7.7 Vp/Vs versus λ cross section of gas-sand vs. brine-sand or shale.Figure 7.8 ROC Curves for KNN and LDA classification techniques for 2D predictor...
8 Chapter 8Figure 8.1 Comparison of solutions for n = 0.3.Figure 8.2 Comparison of solutions for n = 0.8.Figure 8.3 Determining n for Example 4.Figure 8.4 Matching cart for Example 1.Figure 8.5 Matching chart for Example 2.Figure 8.6 Matching chart for Example 3.Figure 8.7 Matching chart for Example 4.Figure B1 Pressure and pressure derivative type c.Figure B2 Pressure and pressure derivative type curve for n=0.4.Figure B3 Pressure and pressure derivative type curve for n=0.5.Figure B4 Pressure and pressure derivative type curve for n=0.6.Figure B5 Pressure and pressure derivative type curve for n=0.7.Figure B6 Pressure and pressure derivative type curve for n=0.8.Figure B7 Pressure and pressure derivative type curve for n=0.9.Figure B8 Pressure and pressure derivative type curve for n=1.0.
9 Chapter 9Figure 9.1 Plot of sorted values of individual permeability forecasts generated ...Figure 9.2 Diagram of the structure of the second level committee machine. Weigh...Figure 9.3 Mean absolute bias of permeability forecast with additive multiplicat...Figure 9.4 Individual forecasts, output of the first level committee machine, an...Figure 9.5 Bias of permeability forecast by the first level committee machine. O...Figure 9.6 Bias of permeability forecast by the first level committee machine. O...Figure 9.7 Permeability forecast with multiplicative, exponential, and additive ...Figure 9.8 Permeability forecast with second level committee machine.Figure 9.9 Workflow for permeability prediction.
10 Chapter 10Figure 10.1 Distribution of elastic deformation modules of coalbeds: (a) distrib...Figure 10.2 Geostatic pressure in wells.Figure 10.3 Distribution of the lateral thrust of coalbeds.Figure 10.4 Gas flow rates of Taldinskaya area wells.Figure 10.5 Results of experimental studies using the multiwave VSP technique: (...
11 Chapter 11Figure 11.1 Histogram of a number of individual forecasts in Monte Carlo cycle o...Figure 11.2 Bias of the forecast with permeability Models 1 and 3. Bias is calcu...Figure 11.3 Ordered values of permeability and their forecasts with permeability...Figure 11.4 Forecast bias as a function of permeability values for two machine l...Figure 11.5 Individual and committee machine permeability forecasts in compariso...Figure 11.6 Instability indexes of individual permeability forecasts by support ...Figure A2.1 Configuration of a Committee Machine.Figure A2.2 Multilayer Perceptron (a) without and (b) with short cut connections...
