MODULE GUIDES

Integrated forecast

Couple material balance tanks, wells and the surface network through time, with events, constraints, economics, scenario screening and P90/P50/P10 bands.

Last edited

Overview

The Integrated forecast module solves reservoir tanks, wells and the surface network together through time. Each step, every well takes its tank's pressure and producing ratios, the network is solved with constraints and a plateau target, and the delivered volumes are withdrawn from the tanks. On top of that the module runs cash flow per scenario, screens development options by NPV, and builds P90/P50/P10 forecast bands. Use it for field-development decisions where reservoir depletion and surface backpressure interact: infill timing, compression, lift-gas capacity and plateau length.

The tabs are Model links, Schedule & constraints, Run, Results, Scenarios & screening and Uncertainty.

Inputs

The forecast builds on models from other pages. Before it can run, you need:

  • Reservoir tank: an oil tank on the Material balance page with an accepted history match, which gives N and the aquifer.
  • Wells and network: wells (IPR and tubing), flowlines and a separator on the Surface network page.
  • Economics (optional): a case on the Economics page. Without one, the NPV uses typical prices and fiscal terms, marked as typical.
  • Project fluid: an oil fluid on the PVT page.
Input Notes
Surface network, Reservoir tank, Economics Linked to the module (default) or Override on this page. An override is explicit, labelled and starts blank.
Override tank Initial pressure, Oil in place, Swi, Abandonment pressure, Rock compressibility, Water compressibility, Aquifer (Fetkovich (radial) or None) and Corey end points and exponents
Forecast-only costs (added to the Economics case) Cost per well, Abandonment, Years before first oil. Facilities and pipeline come from the Economics case's capex, so they are not counted twice.
Well links For each well: the Tank it drains; Water cut / GOR mode (offset, tank or fixed); Productivity index (constant or follows liquid mobility); Choke-back allowed

Schedule & constraints

Input Unit (Field) Notes
Forecast length years 0.25 to 60
Plateau target, Field oil target, Measured at STB/d Optional plateau, measured at a separator or sink
Honour network constraints – Uses the separator limits set on the Surface network page
Choke-back rule – Worst first or Pro rata
Re-allocate lift gas (equal slope), Re-allocation interval days
Abandonment limits STB/d, fraction, scf/STB Well minimum oil rate, Well maximum water cut, Well maximum GOR, Field economic limit
Events – Well on / off, Choke change, Lift-gas rate (well), Lift gas available, Separator pressure, Add compressor, Constraint limit, Plateau target, Water injection. Each has a Year, an optional Capex and a label.

Methods

  • Coupling: each step applies due events, shuts wells past their limits, re-allocates lift gas when scheduled, gives each well its tank pressure, water cut and GOR, solves the network, and withdraws the network-delivered stock-tank volumes from the tanks. Mass is conserved to round-off whatever the convergence state.
  • Coupling scheme: Explicit (start-of-step network rate held over the step, after Hepguler et al. 1997) or Iterative (predictor-corrector, with Theta 0.5 for midpoint or 1 for implicit).
  • Timestep: optional adaptive step using a PI controller on the coupling error (Söderlind 2002), with step rejection. Steps land on every event.
  • Choke-back: when a constraint or the plateau target is exceeded, wells are choked back by a wellhead pressure-drop lever, either worst first or pro rata. Wells you mark as not controllable are left alone.
  • Tank model: the material balance equation with the linked aquifer and Corey fractional flow, as in the material balance prediction.
  • Economics: cash flow through the Economics page's own converter, so fiscal regimes such as PIA 2021 arrive unchanged.
  • Screening: a grid over well count (drilling sequence), compression timing and lift-gas availability, ranked by NPV.
  • Uncertainty: Latin hypercube sampling, default 100 realisations. Each realisation is a full coupled run with its own oil in place, aquifer strength, productivity index, skin change, Corey exponents and uptime, from split log-normal P90/P50/P10 inputs.

Workflow

  1. Build the tank on Material balance and accept a history match. Build the wells and network on Surface network. Optionally, set up Economics.
  2. Open Integrated forecast and press Build from my models. To explore first, press Open sample field for a 6-well field on one oil tank with two scenarios.
  3. On Model links, confirm the three sources, set forecast-only costs, and check each well's tank, ratio mode and productivity link.
  4. On Schedule & constraints, set the forecast length, plateau, choke-back rule, lift-gas re-allocation and abandonment limits. Use Add event for infills, compression and other changes.
  5. On Run, choose the Coupling scheme and timestep settings and press Run forecast. Watch the coupling log.
  6. On Results, read EUR, NPV, plateau and gas recovered, and the year-by-year table.
  7. On Scenarios & screening, duplicate scenarios, choose a Base scenario and compare. Press Run all changed scenarios. Turn on screening variables and press Screen n cases (n is the case count).
  8. On Uncertainty, set the P90/P50/P10 multipliers, Runs (Latin hypercube), Seed and Fidelity (Screening or Scenario), then press Run n realisations.

Interpreting results

  • Oil recovered (EUR) with recovery factor, NPV at the first discount rate of the Economics case, Plateau length and Gas recovered sum up a scenario.
  • The yearly table shows Wells flowing at year end, On production at year end, Average wells flowing (time-weighted), revenue, opex, capex, government take and net cash flow in $MM. A note names wells the plateau choke-back held shut: an infill well can start while others are shut to hold the rate.
  • On Run, check Largest mass mismatch, step cuts and the Coupling error (%) column. A Binding entry shows which constraint limited each step.
  • The scenario delta table shows Δ NPV ($MM), Δ NPV per Δ barrel ($/STB), Δ plateau (years) and Δ recovery (points) against the base.
  • Uncertainty bands come from full coupled runs. Screening fidelity is faster and coarser; use Scenario fidelity for numbers you report.

Common pitfalls

Exports

  • Send to Economics on Results sends the scenario's production profile to the Economics module.
  • Export CSV on the Run tab downloads the coupling log, one row per accepted step (coupling_log_<scenario>.csv).