Release notes for Flexa Analytics

Nine new problem types for oil and gas

Monte Carlo moves from a fixed set of built-in problems to a shared quantitative engine. Sampling, correlation, sensitivity and scenarios are now common to every problem, and nine new problem types are built on top of it — covering reserves, project economics and field operations.

• addedReserves — Probabilistic volumetrics reported as P90, P50 and P10, for in-place and recoverable volume.
• addedProbabilistic Economics — NPV, IRR, payback and breakeven price as distributions, with a depreciation shield.
• addedDecline Curve — Estimated ultimate recovery from an Arps decline with uncertain initial rate, decline rate and b-exponent.
• addedExploration EMV — Risked value, expected monetary value and the probability of a positive outcome for a drill-or-drop decision.
• addedPrice Deck — Brent, WTI and Henry Hub sampled together through a correlation matrix, with a mean-reverting variant.
• addedHedging Payoff — Swap, put or collar compared against the unhedged price, so the cost of protection can be weighed against the downside it removes.
• addedDrilling AFE — Well cost from planned days, day rate, non-productive time and flat cost.
• addedCarbon Cost — Emissions liability above a free allowance, with both volume and carbon price uncertain.
• addedSensitivity analysis — Tornado and spider charts showing which input actually drives the result.
• addedScenario manager — Name several scenarios and compare them against the same model.
• addedConvergence panel — See whether a run has settled, instead of guessing at an iteration count.
• addedLatin Hypercube sampling alongside standard Monte Carlo, on a seeded generator so a run can be reproduced exactly.
• addedInputs can be correlated with each other, so variables that move together in reality move together in the simulation.
• addedValue at Risk and Conditional VaR in the summary statistics.
• addedParameters can be edited from the results view, without going back to setup.
• improvedThe distribution chart centres on the spike and labels the peak, so the most likely value is readable without hovering across the histogram.
• improvedPrice Deck mean-reversion and the Economics depreciation shield are now adjustable in the interface rather than being fixed internally.
• fixedCorrections to the normal cumulative distribution, portfolio covariance and Expected Shortfall. Statistics produced by earlier builds should be re-run.
• fixedValue at Risk and Conditional VaR now use the loss convention.
• fixedThe Monte Carlo tab, and its entry points in the top bar and the published view, were missing after an earlier update.
• fixedSimulation settings are kept per layout and restored on reload, instead of resetting or leaking between layouts.
• fixedA parameter-driven simulation could take the whole visual down; long runs no longer freeze the report while they compute.
• fixedA Portfolio Risk simulation could load forever instead of finishing.
• fixedOption Pricing ran several times from a single click on Run, multiplying the wait.
• fixedDisplay options now list only entries that have data, populate the Chart group, and survive publishing when every option is unticked.
• fixedClear resets the entered parameters, not just the results.
• fixedTornado value labels and percentile markers no longer print on top of each other.

Plus calendar-based demand seasonality, a unified portfolio return model, input validation and units across the parameter-driven problems, and a layout that holds up in a narrow Power BI panel.