The Petrophysical Evaluation wizard carries a complete well interpretation in one window instead of one calculator at a time. Its eight tabs follow the calculation order — 1. Environment, 2. Vsh, 3. PHI, 4. Saturation, 5. Permeability, 6. Matrix, 7. QC and 8. Results. Later calculations use the selected results from earlier steps.

Every method run produces a candidate curve. Several candidates of the same property can coexist, exactly one of them is selected, and only the selected curve is consumed further down the chain. Each computed curve is written to the well in the Data manager as soon as the run finishes, so results can be reviewed, displayed and exported from there. Import and export curves through the Data manager; a curve that is not in the project yet must be imported there first. The wizard can export its QC and Results tables directly to CSV.

Launching the wizard

Open the Petrophysics ribbon bar and click Evaluation in the Properties group. The command is named Petrophysical Evaluation and opens the window of the same name. The wizard is modeless, so the main window, the views and the Data manager stay available while it is open.

The wizard window

The top bar holds the Well list, the Interval label, the Run field that names the run (Base_Run_01 by default), the current environment version, the Mode list that switches between a deterministic and a stochastic evaluation, and the Reset Environment tab (named for the active tab), Reset run, Save run, Save to workflow and Recalculate chain buttons. A draft-status message sits beside the save buttons. Under it the numbered tab strip selects the step. Every step is laid out in three panes: the settings panel on the left, the depth track in the middle, and a tab-specific crossplot or QC panel on the right. Draggable dividers separate the three panes: drag one to give the settings, the depth track or the QC panel more room. No pane can be dragged away completely, the widths you set apply to all eight tabs, and they are restored the next time you open the wizard. Widening the window itself adds the extra space to the depth track and keeps the side panes as you set them. The action button at the bottom runs the active tab; its label follows the current mode and tab. Close closes the wizard.

Rolling the mouse wheel over a numeric setting scrolls the settings panel without changing the value, even when the field has focus. Type a value, use its arrow buttons or press Up or Down to edit it deliberately.

The reset icon beside an editable numeric parameter shows its detected or default value in a tooltip; click it to restore that value. Reset Environment tab changes its name with the active tab and restores that tab's detected and default configuration for the selected well and interval. It asks for confirmation when configuration or results would be discarded, leaves the other tabs' settings in place and marks dependent calculations stale.

Reset run always asks for confirmation, then restores all tabs and stochastic settings to their detected and default state. Curves and runs already saved to the project remain available. The reset working state is saved in the temporary draft.

The eight tabs show their calculation state: ● means not calculated, ✓ current, ⚠ calculated with QC warnings, ↻ stale, ! blocked by missing or invalid input, and — not required for a skipped stochastic family. Hover over a tab for an explanation. After a successful run, a message below the tabs remains for about six seconds and the wizard stays ready for more work. QC warnings can accompany that message and remain indicated on the tab.

The depth track plots the curves of the current step against a shared measured-depth axis, and each lane is titled with the curve name, its scale and its unit. Selecting another well in the top bar clears the candidates, because they belong to the well they were computed for. Depth values on the shared track use the project depth unit, m or ft. The Interval selector above the track shows Full well, the current marker range or custom MD range, available individual zones, or All selected zones for a multi-zone selection. Custom interval... opens the 1. Environment interval controls. Choosing a zone changes the run interval and updates dependent views and calculations. The selected interval is shaded and labelled on the shared depth track on every tab, with depths in the project unit.

Candidates and selection

Tabs 2 to 5 keep lists of candidate curves. Each row carries a selection button, a colour swatch, the curve name, a PROJECT badge when the curve was taken from the project, and percentages labelled Valid, Warning and Clipped. Hover over them for the shares: Valid samples passed QC, Warning includes flagged, failed and skipped samples, and Clipped samples were clamped to the method limits. 6. Matrix shows its mineral solution separately; its third QC percentage is Negative before clamp.

Curve lists put recommended type and unit matches first, while untyped and other curves of the selected well remain available for manual choice. The choices show curve type, unit and interval coverage. Recognized mnemonics include GR and GRR for gamma ray, RHOB and RHOZ for density, NPHI and NPH for neutron porosity, PE and PEFZ for Pe, DT and DTCO for P-wave, and RT and RLLD for deep resistivity. Check the displayed curve and its unit before running a tab; the recommendation is changeable.

• Selecting a candidate – exactly one candidate per property is selected, and it is the one every downstream tab reads.
• + Run another method – calculates the method chosen in the adjacent list and adds it to the list as its own candidate. Running the same method again replaces its candidate.
• Save as copy – freezes the current candidate under a copy name, so a later run of the same method can be compared against it.
• Delete – removes a candidate. A candidate that a later tab has already consumed cannot be deleted; the wizard names the consumer and asks you to select another candidate and run again first.

Using a curve from the project

Instead of calculating a property, you can register a curve that already exists in the well — the output of another calculator, or log data imported earlier through the Data manager. Click Use from project... and complete the Use curve from project dialog:

• Curve and Origin – the curve of the selected well and where it came from.
• Declared as – what the curve is. For a porosity curve, Porosity type additionally states whether it is total or effective porosity, because that changes the calculations that consume it.
• Curve unit and Interpret as – the unit stored with the curve and the unit it is read in. If the stored unit is unknown, explicitly choose the interpretation before continuing. The mapping is remembered for later curves of the same property family carrying that unit symbol.
• Depth and Range – how the curve is resampled onto the depth step of the run, and which part of the interval it covers. Samples outside the depth range of the source curve stay undefined and are marked skipped. Values outside the property's allowed range are retained and flagged with QC warnings rather than clipped.

For porosity, Curve initially selects a Total Porosity curve when one is available, then an Effective Porosity curve or another recognized porosity curve. A known curve type can fill Declared as and Porosity type, and a recognized unit fills Interpret as. For a calculator or older curve without a useful type, set Declared as to Porosity (PHI) and choose its total or effective Porosity type before using it.

The registered curve becomes an ordinary candidate with a PROJECT badge, and its values are never modified. Because it does not depend on anything calculated inside the wizard, it never goes stale.

Staleness and Recalculate chain

Nothing is recalculated silently. When you change an input that a calculated curve consumed — for example a value on the 1. Environment tab — every curve that depends on it is marked stale, a banner names the affected steps, and the stale curves are drawn dimmed on the depth track. Recalculate chain re-runs only the stale steps, in dependency order, and reports which ones it recalculated; if nothing is stale it reports that as well. The only exception is dragging the endpoint lines on the 2. Vsh histogram, which recalculates that tab’s own candidates as you work.

Stochastic mode

The Mode list in the top bar switches the whole wizard between Deterministic and Stochastic. Deterministic is the default and works exactly as described above: every parameter is a single number. In Stochastic mode the parameters you gave an uncertainty distribution are drawn at random and the whole chain is run again for each realization, so the uncertainty of the shale volume reaches the porosity, the saturation and the permeability instead of stopping at its own tab. Switching back to Deterministic keeps the distributions you entered, and they reappear when you switch to Stochastic again.

Selecting Stochastic adds a second row to the top bar:

• Realizations – how many parameter sets are drawn: 100, 500 or 1000.
• Random seed – the number the draws start from. The same seed on the same inputs reproduces the run exactly, including after saving it and reopening the wizard; Regenerate replaces it with a new one.
• Correlation – reads Off: the parameters are drawn independently of one another.
• Save per-realization sensitivity dump – asks for a CSV file when the run finishes and writes one row per realization with the values that were drawn, so you can see which parameter drives the spread.

A small ± button and a summary appear beside every parameter that can carry a distribution, and both are hidden again in Deterministic mode. The summary reads Fixed until a distribution is set and afterwards names it with its minimum, mean and maximum. The ± button opens the Uncertainty dialog of that parameter, which offers Normal, Lognormal, Triangular and Uniform, each described by Min, Mean and Max; for the normal distribution the standard deviation derived from them, (Max − Min) / 6, is shown read-only under the fields. A parameter that has no distribution yet opens centred on the value the tab currently shows. OK stores the distribution and Set fixed removes it again; a minimum above the maximum is refused, and a mean outside the range is moved back into it.

The example below shows an Rmf Normal distribution in the Uncertainty dialog.

In Stochastic mode, the readiness panel below the tabs lists Well, Vsh, PHI, Saturation, Permeability, Mineral solve, Distributions and Random seed. Expand it to see what is ready, will be prepared, blocked or not required; use a row's Go to action to reach its input. Saturation and Permeability are not required when no resistivity curve is selected. The mineral solve is optional and is skipped without selected shale volume, components or a solvable setup. A blocked row names the missing or unusable input.

Show distributions… lists configured distributions and whether the selected methods use them. The readiness count distinguishes live distributions that will be sampled from ignored ones; at least one live distribution is needed. Prepare required candidates computes missing, preparable Vsh, PHI, Saturation and Permeability candidates in chain order with the current settings and selects them. It does not start the ensemble or repair blocked inputs, excluded candidates, invalid distributions or an empty seed. Use Regenerate for an empty seed and the indicated tab for other unresolved items.

Distributions belong to the model parameters and not to the choice of method or source. They are offered on the surface and bottom-hole temperatures and the depth the bottom-hole temperature is read at, the temperature gradient, the measured water resistivity and its temperature, the static SP deflection, the mud-filtrate resistivity and its surface temperature and the salinity on 1. Environment; on the clean and shale endpoints of 2. Vsh; on the matrix and fluid densities, the matrix and fluid travel times, KCP, PHIDSH and PHINSH on 3. PHI; on the tortuosity factor, the cementation and saturation exponents, the shale resistivity and the shale porosity of 4. Saturation; on the coefficients of each permeability method, the bulk-volume-water constant and the manual irreducible saturation of 5. Permeability; and on the endpoints of every mineral on 6. Matrix. Limits and thresholds stay single-valued and are applied unchanged to every realization: the porosity clamp PHIMAX, the permeability Floor and Cap, and the cutoffs of 8. Results.

In Stochastic mode, the bottom button reads Run stochastic chain on 2. Vsh, 3. PHI, 4. Saturation and 5. Permeability. It runs the applicable selected families once per realization, repeats the zone summation and reports progress in the Stochastic evaluation window. A solvable 6. Matrix mineral solve joins the chain; families marked not required are skipped. 1. Environment keeps Run for its single calculation. On 6. Matrix, 7. QC and 8. Results, the bottom button reads Run matrix solve, Run QC and Run deterministic summation respectively; these recompute the single interpretation rather than the ensemble. In Deterministic mode, the bottom button reads Run on every tab.

Pressing Run stochastic chain with unresolved readiness items expands the panel and highlights the first blocker without starting the ensemble. The run needs at least one distribution that the selected methods actually use and a seed to start from, and the clean and shale windows on 2. Vsh must still keep the separation the driving log requires, distributions included; whatever is missing is named instead of run. Two candidates cannot be selected in this mode, because neither carries model parameters that can be sampled: one registered with Use from project..., and the Minimum of candidates (QC floor) curve. The wizard reports that when you switch modes and selects a computed candidate in its place, or asks you to compute one first.

While a stochastic run is loaded, the depth track of tabs 2 to 5 gains a shaded P10-P90 band for that property with the deterministic curve drawn over it as a dashed line, so the single answer can be read against the spread around it. On 2. Vsh the histogram also draws each sampled endpoint distribution over the bars: the solid line is still the endpoint and carries the mean of its distribution, the dashed lines beside it are its minimum and maximum, and all of them can be dragged. 6. Matrix reports under Log responses how many realizations had to be rescaled for the mineral volumes to close on one, with the mean excess over the interval; a large fraction means the endpoint distributions are wider than the closure condition can absorb.

After Run stochastic chain finishes, 8. Results shows the ensemble percentile table: one row per summed quantity for each zone and for TOTAL, in the columns P90 (low case), P50 and P10 (high case), and Zone table → CSV exports it in that form. Because the cutoffs are not sampled, the spread reflects the parameters alone; the gross thickness of a zone does not depend on them and is therefore the same in all three columns. This table and its CSV use the same quantity names as the deterministic zone summary; depth quantities are in the project unit, m or ft, and permeability is in mD.

The 8. Results provenance badge identifies a Deterministic single interpretation or a Stochastic P90 / P50 / P10 table with its realization count and seed. Run deterministic summation shows the single interpretation even while the mode is Stochastic; Show stochastic results returns to the last ensemble while it remains in the wizard. If inputs change, the stochastic table stays readable and exportable but is greyed with a stale warning until Recalculate chain runs it again. After reopening, the ensemble table is empty because its rows are not saved with the project; No stochastic results yet offers Run stochastic chain. A Zone table → CSV export starts with the same deterministic or stochastic provenance shown by the badge.

Each family writes its percentile curves to the Data manager beside the deterministic ones, taking the mnemonic of the selected method and adding _P90, _P50, _P10, _MEAN and _SD; every mineral volume of a stochastic matrix solve gets the same set. The convention is the petroleum one and the completion message repeats it: P10 is the high case, the 90th percentile of the realizations, and P90 the low case, the 10th percentile. Editing a distribution, a parameter or an input afterwards marks the stochastic outputs stale in the same way the deterministic chain does, and Recalculate chain runs the realizations again.

The tabs

The following examples show each tab in both modes using demonstration data. Values depend on the well, interval and settings. Review the displayed warnings before accepting an interpretation: an uncertainty band does not establish that the selected model fits the measured logs.

1. Environment

Sets the well conditions that the later tabs need, and calculates the formation temperature curve.

Selecting a well prefills the available interval and temperature values from that well and recommends matching curves; you can change the selections. Origin badges such as FROM WELL, FROM PROJECT and DEFAULT show where a value came from. Depth fields, the interval and Step MD use the project depth unit, m or ft; the temperature gradient uses that unit in its denominator. Depth warnings identify values outside or inconsistent with the selected well without preventing you from completing the setup.

• Formation temperature – FT source offers Compute — ST + BHT @ depth (the default, from a surface and a bottom-hole temperature at a stated depth), Compute — ST + gradient G and Use from project — existing FT curve. Switching between the two computed forms carries the equivalent values across. Under Output, Output curve names the result (FT by default) and Unit sets the unit it is written in; internally the temperature is always kept in degrees Celsius.
• Water resistivity (Rw) – one active source out of Water analysis (lab), Rwa from wet zone, SP-derived, Catalog and Manual. Each row shows the value it produces and a MEASURED or CROSS-CHECK badge, and the fields below feed them: the measured Rw and its temperature, the static SP deflection, and the wet-zone interval the cross-check reads.
• Mud filtrate – Rmf @ surface temp, Surface temp and Salinity.
• Depth track – the Gamma Ray curve and Resistivity curve to display. The deep resistivity curve chosen here is also the one that 4. Saturation requires.
• Interval – the same depth-range selector as the other calculators on the bar (MD range, TVD range, TVDSS range, Markers range and Zones range), plus Step MD, the depth step every curve of the run is calculated on. Calculating all wells in one batch is not available in this wizard.

On the Zones range tab, tick one or more zones of a zone set to confine the whole run to those zones. The selection replaces the depth range set on the other tabs instead of narrowing it: every tab reads only the samples inside the selected zones, and the rock between two of them is left out of the calculations and of the summation. The top bar reports the zone selection and the depth range it spans. The Interval selector above the shared depth track shows the selected zone name, or All selected zones with a count for a multi-zone selection. A selected zone that has no interval in the current well is skipped and reported; if none of them resolves, the run stops and asks you to check the zone markers or to use another interval tab.

The right panel, Computed environment state, shows the resulting formation temperature at mid-interval, the water and mud-filtrate resistivities corrected to that temperature, the salinity and the fluid densities, each with the source it came from. Missing environment values never block a run: the calculation falls back to defaults and marks the output samples with a warning.

Deterministic Environment. After calculation, the depth track shows formation temperature and the right panel reports the computed environment state.

Stochastic Environment. With the uncertainty editor closed, the stored distribution summaries remain beside their parameters. The realization count and seed identify the ensemble setup.

2. Vsh

Calculates the shale volume. Under Input, Log selects the driving log (Gamma Ray, Spontaneous Potential or Density-Neutron), Curve the curve that carries it, and Histogram zone which part of the Run interval the histogram reads. The Run interval row shows the full calculation interval; choosing a histogram zone does not change it. The zone list follows the zones selected on the 1. Environment tab: its first entry is All selected zones — or Whole run interval when no zones are selected — followed by one entry per selected zone in depth order, and two zones that overlap appear as one entry carrying both names. Picking a zone scopes the histogram only: the clean and shale endpoints and the calculated shale-volume curve still cover every selected zone, and a hint under the row states that. Under Endpoints, the clean and shale values of that log are entered as numbers; a line below reports the current separation and the minimum the log requires, and warns when the methods are excluded because the endpoints are too close together.

The right panel is a histogram of the driving log with two draggable endpoint lines, clean and shale. Dragging a line updates its numeric field immediately and recalculates the candidates that use that log, so the endpoints can be picked visually against the data.

The available methods are Linear (IGR), Steiber, Clavier, Larionov, tertiary rocks, Larionov, older rocks and Minimum of candidates (QC floor). All but the linear index are calibrated on gamma ray, so the other logs offer the linear index alone. Selecting the minimum curve shows a warning that it is a quality-control floor rather than an interpretation, because the tabs that consume it inherit its optimistic bias.

Deterministic Vsh. Review the Gamma Ray endpoints against the histogram and select the calculated shale-volume candidate for downstream use. The example retains its clipping warning.

Stochastic Vsh. A consumed endpoint distribution produces a shale-volume uncertainty band after Run stochastic chain. That uncertainty is propagated to calculations that consume Vsh.

3. PHI

Calculates porosity from the density, neutron and sonic curves selected under Input. Matrix and fluid sets the rock matrix (Sandstone, Limestone or Dolomite, which fills in the matrix density and travel time), the fluid density, the upper clamp PHIMAX and the compaction correction KCP. Shale correction holds the shale endpoints PHIDSH and PHINSH, which are applied together with the shale volume selected on 2. Vsh. Gas handling offers Auto — detect ND crossover (the default), Force liquid and Force gas.

For the selected neutron log, Neutron Porosity scale offers Auto-detect, Interpret as fraction (v/v), Interpret as percent (%) and Use the stored unit. The note below the selector shows the interpretation used, including a confident detected scale for a curve with no stored unit, and warns if the stored unit conflicts with sample values or the scale is uncertain. Percent readings are divided by 100 and shown as converted to v/v for calculations. Changing the scale refreshes the crossplot and depth track with those converted values without changing the source curve. The scale choice is restored with a saved run.

The methods are Neutron-Density, Density, Neutron and Sonic (Wyllie). Each run writes three curves: the interpretation curve, an unclamped twin, and a flag curve that marks the samples the clamp changed. The right panel is a neutron-density crossplot with the sandstone, limestone and dolomite reference lines.

A calculated PHI curve is saved as total porosity when no shale-volume curve is selected. When the run consumes shale volume, the density, neutron and neutron-density methods save effective porosity; the sonic method remains total porosity because it has no shale correction. The saved curve's name and type identify which result other petrophysical tools can select.

Deterministic PHI. This Neutron-Density example consumes the selected Vsh candidate and produces effective porosity. Check the neutron scale interpretation, crossplot and warnings alongside the selected candidate.

Stochastic PHI. The porosity band combines sampled porosity settings with uncertainty inherited from the consumed shale volume. Unit and clipping warnings still need review.

4. Saturation

Calculates water saturation. A read-only Consumes banner states which environment version and which selected porosity and shale-volume curves the run will use. The methods are Simandoux, Archie and Dual Water; a hint line reports the average shale volume over the interval and suggests a method for it, but never changes the selection for you. Electrical properties holds the tortuosity factor, the cementation and saturation exponents, the shale resistivity and the shale porosity; the last two reach the shaly-sand methods only, as Archie reads the first three alone.

This is the one step that cannot fall back to a default: without a deep resistivity curve the run stops and reports Select a Resistivity curve on the Environment tab before running Saturation. The right panel is a Pickett plot with the Sw = 1 (water line) and Sw = 0.5 overlays.

Deterministic Saturation. Review the selected Simandoux candidate together with its consumed inputs, electrical properties and Pickett plot. A calculated result may still carry a convergence warning.

Stochastic Saturation. The saturation band reflects sampled settings and uncertainty propagated through the consumed environment, porosity and shale-volume inputs. Solver warnings remain visible.

5. Permeability

Calculates permeability from the selected porosity and an irreducible water saturation. The methods are Timur, Morris-Biggs and Coates, each with its own group of coefficients.

• Irreducible water saturation – SWirr source is either the Buckles relation, whose bulk-volume-water constant Fit BVW from the zone estimates over the interval you enter, or Manual constant.
• Calibration – Target type records what the result is calibrated against: uncalibrated (the default), core — air K, core — Klinkenberg or core — liquid K. An uncalibrated run keeps a warning on its output.
• Trimming – Floor and Cap bound the result in millidarcy, and the samples they change are flagged.

The right panel plots permeability against porosity on a logarithmic axis, with the trend line of the selected model.

Deterministic Permeability. Review the Timur coefficients and irreducible-water-saturation settings together with the porosity-permeability diagnostic. The example's Target type is uncalibrated, so the curve remains a model estimate.

Stochastic Permeability. Upstream uncertainty can produce a permeability band even when the model coefficients are fixed. Running an ensemble does not remove the uncalibrated warning.

6. Matrix

Solves the mineral composition of the rock from the log responses. The clay volume is fixed to the shale volume selected on 2. Vsh. In Components, the clay row's volume stays fixed but its Δt (travel time) can be entered; the mineral rows have editable endpoints, including Δt. + Add mineral adds a mineral from the supplied table, and Custom mineral duplicates the mineral above it under an editable name.

Log responses selects which measurements enter the solve — Density, Neutron Porosity, Pe and P-wave — on top of the closure condition, which is always present. A counter reports how many equations and unknowns the current setup has and whether the system is square, underdetermined or overdetermined. While there are more unknowns than equations, Run is unavailable and the counter asks you to add a response or remove a mineral. Non-porous pre-solve triggers can additionally recognise halite and anhydrite intervals from their log fingerprint before the solve runs.

Each response row has a checkbox and curve choice. Density, Neutron Porosity and P-wave share the curve selections made on 3. PHI; Pe is selected here. The wizard recommends compatible type and unit matches but leaves other well curves selectable. It checks interval coverage and reads compatible density, porosity, Pe and travel-time units in the solve's units. P-wave can be used only when its selected curve has data in the interval and every component, including clay and water, has a Δt value. If a checked P-wave row lacks an endpoint, the wizard keeps it out of the solve and names the missing value until you enter it in Components. Pe contributes volumetric cross-section as Pe multiplied by density. The equation counter explains when another response or fewer minerals are needed; P-wave is checked automatically only when another row is required.

The right panel's Diagnostic selector offers M-N with Density, Neutron Porosity and P-wave; NPHI-RHOB with Density and Neutron Porosity; Umaa-ρmaa with Density, Pe and a selected porosity candidate; Residuals after a Matrix run; and Lithology for reviewing lithotype classification from the solved mineral volumes. The first three plots use the current interval inputs and redraw when responses, curves or the interval change, even before the first Matrix run.

The default view follows that order among available views. Your chosen view is saved; if its required inputs become unavailable, the wizard shows the next available view and explains the substitution. Unavailable choices are greyed with explanations, and an unavailable message replaces the plot if no view can be drawn. The Solve and Views status reports the counted responses, equation balance and available diagnostics, with action links for an available response, a curve choice, a missing clay Δt or the PHI tab when applicable. Switching views does not rerun the solve. Remove clay (Vsh) is on by default when shale volume is selected: it subtracts the selected clay share before plotting and builds the feasible region from non-clay components; without selected shale volume it is unavailable.

Run in Deterministic mode (Run matrix solve in Stochastic mode) writes one volume curve per component plus a closure-residual curve. The depth track stacks the components and shows closure residual and a QC strip that marks warned samples. The Matrix solution reports Negative before clamp as the share of samples whose solve produced a negative component volume. At 5% or more, a warning calls the mineral solution poorly constrained; at 20% or more it calls it unreliable. The warning gives the count, percentage and worst negative value in v/v, explains that the displayed volumes were clamped to physical limits rather than fitted to the logs, and directs you to review the mineral components and selected log responses.

After a Matrix run, Residuals compares measured responses with values reconstructed from the last solved volumes. Choose a Response to see measured versus reconstructed samples against a 1:1 line and tolerance limits. The table identifies whether each response was in the solve and lists its sample count, bias, MAE, RMSE and share within tolerance; a response outside the solve provides an independent check. The depth track adds measured and reconstructed curves and residual lanes for those responses. This view uses the last solve, writes no additional curves and does not replace the separate checks and saved output of 7. QC.

Deterministic Matrix. Check the selected log responses, equation balance and calculated mineral volumes. The diagnostic choices include Lithology; solver warnings identify limitations of the mineral model.

Stochastic Matrix. Review the ensemble solve warnings as well as the mineral volumes. The bottom Run matrix solve action recalculates a single interpretation, as described above.

7. QC

Use 7. QC to check the input logs and compare measured responses with the interpretation. Under Forward model, choose the Model source (Matrix solution or PHI endpoints), review the measured curves and their roles, and adjust tolerances as needed. Data QC checks gaps, spikes, flat lines, out-of-range values, bad hole and depth offset. Click Run in Deterministic mode or Run QC in Stochastic mode, then use View to switch among Data QC, Model fit, Residuals and Zone summary. The checks report findings without editing the source curves.

QC requires a selected shale-volume and porosity interpretation. Matrix solution compares measurements with a current Matrix result; PHI endpoints uses the selected porosity model when a Matrix result is unavailable. Measured curves follow the selections on 3. PHI and 6. Matrix. A response used to fit the model is Fitted; an independent measured response is Validation; one without a usable measurement or model response is Not evaluated. Adjust each response's Tolerance in its displayed unit. In a square Matrix system fitted residuals can be near zero by construction, so use an independent validation response. Leave-one-out requires an overdetermined Matrix solution: it removes each fitted response in turn, solves without it and predicts it independently without changing the saved mineral volumes. It cannot validate a square system; add a response on 6. Matrix to make the system overdetermined.

Data QC detects gaps, spikes, flat lines, out-of-range values, unit mismatches and depth offsets; it reports a depth offset without applying it. Click a finding's table cell to highlight the affected samples on the depth track. Bad hole compares caliper with bit size for washout and checks density correction for poor pad contact. Without a caliper curve, the washout check is skipped; the density-correction check remains available when its curve is selected. Choose affected responses and whether to exclude bad-hole samples from fit statistics. The saved BAD_HOLE curve identifies samples affected by washout or poor pad contact. Status rules classify each interval as Good, Review or Poor from the within-tolerance and flagged percentages; the triggering rule is shown.

The depth track overlays solid measured and dashed synthetic responses on a shared scale. Residual lanes show the tolerance bands around zero. QC_MISFIT combines residuals in units of their tolerances; its reference line at 1 marks the tolerance limit. Hover over a sample in the QC flag strip to see all reasons for its flag. The Candidates strips trace flags to the selected upstream candidates; project curves appear neutral because they carry no calculated QC flags, rather than implying that they passed the checks.

In the right panel, Data QC lists findings by input curve, Model fit compares responses, and Residuals shows their differences. Zone summary identifies the selected candidates and their QC shares, then reports each zone's per-response share of measurable residuals within tolerance, bad-hole share and flagged share (Warning and Failed samples). Hover over Status for the rules behind Good, Review or Poor. Multiple zones have a sample-weighted TOTAL, rather than an average of zone percentages; without selected zones, one row covers the run interval. QC table → CSV exports the zone table.

Running QC writes synthetic, residual, misfit and flag curves to the Data manager, including QC_MISFIT, QC_FLAG, QC_REASON and BAD_HOLE. QC runs deterministically even in Stochastic mode. Changing QC settings marks QC stale without marking Results stale. When a current Matrix solution is available, its row in the QC summary calls the third percentage Negative before clamp.

Deterministic QC: Model fit. This example uses PHI endpoints. Select Model fit to compare the measured response with the synthetic response and review tolerances, response roles and depth flags. The actual Poor status is retained.

Deterministic QC: Residuals. Switch View to Residuals for the same interpretation. Select a Response and compare its residual histogram and statistics with the configured tolerance.

Stochastic QC. This example checks a Matrix solution and shows RHOB residuals, response statistics and depth flags. Run QC still performs a deterministic check; ensemble spread does not replace model validation.

8. Results

Applies cutoffs to the selected curves and sums the interval. Inputs consumed names the selected shale-volume, porosity, water-saturation and permeability curves and the environment version read by the summation. The cutoff set holds VSH_MAX, PHI_MIN, SW_MAX, PERM_MIN and MIN_BED_THICKNESS, and the bottom Run button applies them to a single interpretation in Deterministic mode (or Run deterministic summation in Stochastic mode). An Averaging group states the conventions the summary uses, so the basis of every number is visible.

The Input QC summary follows those selected candidates and updates when a selection changes. It reports valid, warning and clipped percentages; failed or skipped samples count in the warning share. A warning appears when either the warning or clipped share exceeds 5%.

The right panel is the zone summary table. It holds one row per zone selected on the 1. Environment tab — or a single ZONE_1 row over the whole interval when no zones are selected — followed by a bold TOTAL row. Its columns are Zone, Gross thickness, Net reservoir, Net pay, Net-to-gross, Avg. PHI reservoir, Avg. PHI pay, Avg. Sw, Avg. hydrocarbon sat., Geom. mean k and HPV, with units in the headers. The run writes the net reservoir and net pay flag curves to the Data manager. Zone table → CSV saves the table as a CSV file; run the summation first, otherwise the wizard asks you to. Hover over a column header for its metric definition. Gross thickness, net reservoir, net pay and HPV use the project depth unit, m or ft; permeability is in mD, and the CSV uses the same quantity names. On the depth track, Rock and fluid composition stacks the component shares and names them in its legend. Net reservoir and Net pay show 1 for an included sample and 0 for an excluded one; named zone or interval boundaries are highlighted across tabs.

Deterministic Results. The provenance badge identifies a single interpretation. Review the zone summary together with Input QC and the upstream warnings.

Stochastic Results. The ensemble table shows P90 (low case), P50 and P10 (high case), with the realization count and seed in its provenance. Export it with Zone table → CSV. Clicking Run deterministic summation replaces the displayed table with the single interpretation.

Saving and reopening a run

Save run in the top bar and Save run snapshot on the 8. Results tab do the same thing: they run the summation and store the state of the run on the flag curves it writes, so that a saved run always matches numbers that were actually calculated. A saved run retains its candidates, selections, parameters and zone table; when the wizard reopens, it offers an unfinished draft first if one exists. In Stochastic mode the snapshot also carries the mode, the realization count, the seed and the whole distribution table, so a reopened run is set up to reproduce itself; a run saved before the mode existed reopens as Deterministic.

The wizard automatically saves a temporary draft of the current well's setup and calculation state. When you reopen the wizard or switch wells and an unfinished draft exists, it offers Restore previous session or Start new evaluation. Restoring brings back the configuration, candidates, selections and calculation state. If a saved choice no longer exists, a warning names that curve and leaves its selector empty for you to choose another. A successful Save run saves the calculated snapshot and clears the current temporary draft.

Use Save to workflow to preserve a setup, even if it is unfinished and has not been run. Opening the workflow step restores its own saved configuration, even when the well has a newer saved run; saving edits updates that step. Open the step in the Petrophysical Evaluation wizard and run it there, because the step is not executed directly as a workflow job.

In the Data manager, right-clicking a curve that carries a saved run — the net reservoir and net pay flags, and the percentile curves of a stochastic run — offers Show evaluation settings, which opens the wizard on that well with the stored run restored.

There is no separate save step for the curves themselves. Every run writes its outputs to the well immediately and reports which curves it wrote, and they appear in the Data manager tree straight away. Exporting them to LAS, like importing new log data, is done from the Data manager. The LAS Parameter section includes per-curve provenance: the source model and version or project-curve origin, parameters, input curves, environment version and QC summary. For a curve taken from the project, it also records the unit conversion and depth resampling.

See Also

• Petrophysics
• Shale Volume
• Porosity
• Saturation
• Mineral Volume Estimation
• Data manager