Direct answer
Oil production forecasting software for upstream teams
PronoHorizon is production forecasting software for upstream oil and gas teams. It imports production history, generates transparent oil, gas and water decline forecasts for every well in a portfolio, produces P10, P50 and P90 forecast cases, and validates those forecasts against production history that was deliberately held back — so an engineer can see where a forecast holds and where it misses before the numbers reach a decision. It solves the problem of forecasting a portfolio that is too large to rebuild well by well in spreadsheets, without giving up visibility into the decline parameters behind each curve.
At a glance
- What it is
- Web-based production forecasting software for upstream oil and gas.
- Forecasting method
- Modified Arps decline curve analysis with visible Di, b-factor and terminal decline.
- Fluids
- Oil, gas and water.
- Forecast cases
- P10, P50 and P90 per well.
- Validation
- Historical holdout backtesting, scored with WMAPE, bias, cumulative error and P10–P90 coverage.
- Input data
- CSV, Excel or public well data, mapped and normalized into one project.
- Downstream outputs
- Reviewed forecast cases and EUR inputs for reserves, asset planning and economic evaluation.
- Coverage
- Venezuela-first positioning; usable for upstream portfolios wherever they operate.
The problem it solves
Most upstream teams already know how to forecast one well. The difficulty starts at portfolio scale: a few hundred producers rebuilt by hand in spreadsheets, each with its own decline assumptions living in a cell nobody can trace, and no practical way to ask whether last year's forecast was any good.
Two failures follow from that. Assumptions become invisible, so a reviewer cannot tell a considered b-factor from a default one. And error goes unmeasured, so nobody knows whether the portfolio forecast is systematically optimistic before it is used for reserves, planning or a transaction.
PronoHorizon is built around those two failures: keep the decline parameters visible, and make the forecast testable against history that the model was not allowed to see.
What the software does
The product moves a portfolio through four connected steps, and the assumptions stay visible at each one.
- 1. Import
- Bring in CSV, Excel or public well data. Map and normalize oil, gas, water, days on, coordinates, well headers and optional directional surveys into a single project.
- 2. Auto-forecast
- Generate oil, gas and water decline forecasts in batch, compare P10, P50 and P90 cases, and assign every well a clear review status.
- 3. Validate and edit
- Backtest from a historical cutoff, compare forecast against actual production using WMAPE and bias, surface low-confidence wells, and override curves where engineering judgment calls for it.
- 4. Complete the analysis
- Review portfolio diagnostics, resolve flagged wells, and prepare reviewed forecast cases and EUR inputs for reserves, asset planning or economic evaluation.
What makes a forecast auditable
A forecast is auditable when a second engineer can reconstruct why the curve looks the way it does. Three things make that possible here.
First, the decline model is named and its parameters are exposed: the Modified Arps fit, its initial decline (Di), its b-factor and its terminal decline are inspectable and editable per well, not hidden behind a score.
Second, data quality stays on the page rather than being silently smoothed away. Months with no production report, allocated or estimated volumes, and days-on counts remain visible, because a decline fit through an unmarked facility turnaround is a different object from a decline fit through steady production.
Third, the uncertainty is described honestly. P10, P50 and P90 are forecast cases, not a calibrated confidence interval: the spread describes three decline scenarios, and only a holdout test tells you how often reality fell inside it.
How to evaluate production forecasting software
Whatever tool you choose, these are the questions worth asking before it produces a number anyone relies on. They are listed with how PronoHorizon answers them, so the comparison is concrete rather than promotional.
- Can you see the decline parameters?
- If Di, b-factor and terminal decline are not inspectable per well, the fit cannot be reviewed. PronoHorizon exposes and allows editing of all three.
- Can you backtest it?
- A tool that cannot hide a known period and forecast into it blind cannot be checked. PronoHorizon's holdout workflow is the core of the product, not an add-on.
- Which error metrics does it report?
- A single accuracy percentage hides direction and magnitude. PronoHorizon reports WMAPE, bias, cumulative error and P10–P90 coverage together.
- Does it survive messy data?
- Real production histories have gaps, allocations and downtime. Ask whether the tool marks them or quietly interpolates them.
- Does engineer judgment persist?
- Manual overrides are worthless if the next batch run erases them. In PronoHorizon the assumptions and engineer edits stay attached to the reviewed case.
- Are the published accuracy claims reproducible?
- Ask for the dataset and the test window. PronoHorizon publishes no accuracy figure that cannot be reproduced from a named public dataset and window.
Who it is for
The same forecast foundation supports three technical workflows.
- Asset management
- Monitor producing wells, refresh forecasts, and identify material deviations from expected performance.
- Reserves and technical evaluation
- Develop traceable production forecasts with a clear path for engineering review and downstream technical evaluation.
- A&D and portfolio screening
- Screen large well sets with consistent assumptions, then focus technical diligence on the assets that merit deeper review.
Frequently asked questions
What is production forecasting software?
Production forecasting software takes the historical production of oil and gas wells and projects their future rates, usually by fitting an empirical decline model to observed history and extrapolating it. In upstream work the output feeds reserves estimates, asset planning, and acquisition or divestment screening.
What method does PronoHorizon use to forecast production?
PronoHorizon uses Modified Arps decline curve analysis. The fitted initial decline (Di), b-factor and terminal decline are visible per well and can be edited by an engineer, so the forecast can be reviewed rather than accepted on trust.
Can it forecast gas and water as well as oil?
Yes. Oil, gas and water are forecast through the same workflow, and the import step normalizes all three streams along with days on, coordinates and well headers.
How do I know whether a forecast is any good?
Hold back a period of known production, forecast into it blind from the remaining history, then reveal the actuals and score the result. PronoHorizon reports WMAPE, bias, cumulative error and P10–P90 coverage for that holdout window so the error is measured rather than asserted.
Does PronoHorizon publish accuracy numbers?
Not yet. The project's proof standard requires every published benchmark to name the public dataset and test window behind it, and results that cannot be reproduced are not shown. Instead of a headline accuracy claim, the product gives you the holdout workflow to measure error on your own wells.
Where can PronoHorizon be used?
Its initial positioning is Venezuela-first, and the public demo uses a Venezuelan portfolio. The forecasting workflow itself is not country-specific and can be used for upstream portfolios wherever they operate.