Integrated Production and Lifecycle

For natural-language reservoir-to-products model construction including SURF, multiphase flowlines/risers, facilities, utilities and downstream processing, use neqsim-production-system-modeling with network-energy-systems. It preserves one case/interface contract for Python and Java and composes these existing screening/integrated APIs with detailed process and external reservoir workflows.

Use this guide for executable APIs that sit between the broad field-development workflow and a detailed process model. The existing field-development and economics skills remain the entry point for concept framing, cost/economic assumptions, flow assurance and reporting. This guide fills in the newer integrated well/network solver, explicit tie-in capacity planning, lifecycle product checks, concept-to-process bridge and reservoir coupling export.

When to use this

• The user needs a field-rate solution that couples simple reservoir pressure decline, precomputed well deliverability and flowline pressure loss.
• The user has measured well tests to fit into a PI or Vogel curve before a network solve.
• The user has a fixed lift-gas budget to allocate across wells using supplied lift-performance curves.
• The user needs annual base-host/satellite allocation, holdback/deferment, or process-model tie-in capacity checks.
• The user needs to determine how a lifecycle study counts deferred production, host load, facility utilization, off-spec years, or candidate modifications.
• The user needs a concept-derived process skeleton or Eclipse-style VFP keyword export, not a full reservoir simulator.

For concept screening, CAPEX/SURF cost, decline fitting, general tax/economics, standards and flow assurance, first use the existing skills in the routing table below. Keep model fidelity explicit: several classes in this area are screening surrogates, not a rigorously coupled subsurface simulator or approved design calculation.

Class map

Routing table

Build pattern

The compact example below exercises a deterministic, single-well reservoir-to-export solve. Pressure is bara; deliverability is surface rate in Sm3/day. The curve is a Vogel-like screening curve, and the drive is a material-balance surrogate. Check isConverged() before using any result.

import neqsim.process.fielddevelopment.integrated.IntegratedProductionModel;
import neqsim.process.fielddevelopment.integrated.IntegratedSolveResult;
import neqsim.process.fielddevelopment.integrated.MaterialBalanceGasDrive;
import neqsim.process.fielddevelopment.integrated.WellDeliverabilityCurve;

public class FieldNetworkExample {
  public static void main(String[] args) {
    IntegratedProductionModel model = new IntegratedProductionModel("Gas field");
    model.setExportPressure(40.0);
    model.setHydrocarbonPrice(0.25);
    model.setEnergyIntensity(0.02);
    model.setEmissionIntensity(0.04);
    model.addWell("W-1", new MaterialBalanceGasDrive(250.0, 5.0e9, 0.90),
        WellDeliverabilityCurve.fromVogel(2.0e6, 250.0));

    IntegratedSolveResult result = model.solve();
    if (!result.isConverged()) {
      throw new IllegalStateException("Network solve did not converge: " + result.getMethod());
    }
    System.out.println("Field rate Sm3/day: " + result.getFieldRate());
    System.out.println("Wells: " + result.getWellRates());
  }
}

Equivalent class lookup from Python via JPype:

from neqsim import jneqsim

IntegratedProductionModel = jneqsim.process.fielddevelopment.integrated.IntegratedProductionModel
MaterialBalanceGasDrive = jneqsim.process.fielddevelopment.integrated.MaterialBalanceGasDrive
WellDeliverabilityCurve = jneqsim.process.fielddevelopment.integrated.WellDeliverabilityCurve

model = IntegratedProductionModel("Gas field")
model.setExportPressure(40.0)
model.setHydrocarbonPrice(0.25)
model.setEnergyIntensity(0.02)
model.setEmissionIntensity(0.04)
drive = MaterialBalanceGasDrive(250.0, 5.0e9, 0.90)
curve = WellDeliverabilityCurve.fromVogel(2.0e6, 250.0)
model.addWell("W-1", drive, curve)
result = model.solve()
if not result.isConverged():
    raise RuntimeError("Network solve did not converge: " + str(result.getMethod()))
print(result.getFieldRate(), result.getWellRates())

To bridge a concept to a process skeleton, FieldConcept.gasTieback("name", tiebackKm, wellCount, ratePerWellMSm3d) creates inputs, and new ConceptToProcessLinker().generateProcessSystem(concept, ConceptToProcessLinker.FidelityLevel.SCREENING) creates a process model. FacilityBuilder.autoGenerate(concept).build() creates a FacilityConfig; it does not itself instantiate every config block into validated equipment. The linker also creates representative fluid/process inputs, so replace those with project fluid and measured data before design use.

For lifecycle work, build a FieldLifecycleConcept using NorwegianOilFieldCase examples as patterns or assemble FieldLifecycleModel and FieldLifecycleConfiguration directly; evaluate with new FieldLifecycleEvaluator().evaluate(concept). The tested lifecycle path supports user-supplied production-potential providers and existing ProcessSystem/multi-area ProcessModel mappings. For VFP export, build the table with VfpTableBuilder and a lift model (see neqsim-vfp-lift-curves); ReservoirCouplingExporter(process) with generateVfpProd(wellName, fluid, tableNumber) and getEclipseKeywords() is a placeholder correlation unless setVlpProvider(...) is called.

Result extraction

• IntegratedSolveResult.isConverged() and getIterations() report solve state and iteration count; getMethod() reports the solver method.
• getFieldRate() and getWellRates() are Sm3/day; getNodePressures() is bara. getRevenue() is the configured price multiplied by daily rate, in the caller's price-derived currency/day. getEnergyKWhPerDay() is kWh/day and getEmissionsKgPerDay() is kg/day, both based only on configured intensities.
• FieldLifecycleResult.getAnnualResults() returns annual records. The nested annual values include oil/export-gas/injection/water volumes in Sm3, annual energy in MWh, CO2 in tonnes, annual rate fields in Sm3/day and getMaximumFacilityUtilization() as a fraction. getNpvMusd(), getIrr(), getPaybackYears(), getBreakevenOilPriceUsdPerBbl() and getBreakevenGasPriceUsdPerSm3() expose economic outputs.
• Product quality is in ProductSpecificationResult: gas CO2/O2 mole percent, H2S ppm, dew points °C, GCV/Wobbe MJ/Sm3, oil RVP bara, oil BSW volume percent and oil-in-water mg/L. Use isEvaluated(), isCompliant() and getViolations(); unevaluated is not a pass.
• TieInCapacityResult exposes period results and accepted/held-back volumes. Gas totals are MSm3, oil totals are bbl and water/liquid totals are m3; periods also expose their named bottleneck and deferred-value MUSD.
• CashFlowEngine.calculate(discountRate) returns CashFlowResult; its getNpv() is MUSD, getIrr() a fraction, and getPaybackYears() years. Input units differ by commodity: oil barrels, gas Sm3 and NGL barrels per year; CAPEX and cash-flow amounts are MUSD.
• ReservoirCouplingExporter.getEclipseKeywords() returns a text keyword block. VfpTable.getFlowRates() is Sm3/day, getThpValues() bara, getWctValues() fraction and datum depth m.

Gas history match and facility limit (reservoir to export)

For a gas field with a pressure history and a compression-limited host, use these classes in process.fielddevelopment.integrated instead of the constant-Z MaterialBalanceGasDrive:

Build the capacity curve from repeated process-model throughput sweeps at several separator pressures (capacity = rate where first-constraint utilisation reaches 1.0). Workflow, hold-out testing and the traps (shut-in proxy, choke-open rows only, joint bootstrap of p/z_i and GIIP, bisection for rate-dependent flowline drop) are in the community skill neqsim-reservoir-facility-value-chain; GasValueChainSurrogatesTest has deterministic fixtures.

Gotchas

• For independently predicted produced composition from a frozen laboratory EOS, use the existing CompositionalReservoirHistory.Step.phaseWithdrawal with prescribed reservoir-phase molar totals, explicit substep refinement and Result.getProducedComponentMoles/getProducedFluid. Process the produced mixture through PvtSeparatorPath on a declared final T/P basis; retain offgas condensate and water. This is well-mixed equilibrium extraction, not a well/mobility model or surface-rate inversion. The original measured-component mode prescribes composition and cannot independently validate it. See docs/pvtsimulation/reservoir_phase_production.md and hand off through @field-development; use independent measured GOR/composition holdouts before field claims.
• MaterialBalanceGasDrive uses one constant Z, so pressure is linear in recovery; with a real p/z history (Z changes 5-10 % over depletion) use RealGasMaterialBalanceDrive.
• IntegratedProductionModel is not the full FieldLifecycleModel; it couples a simple ReservoirDrive, well-curve surrogate and branch network. MaterialBalanceGasDrive and OilTankDrive are simplified depletion models, not a compositional reservoir simulator.
• WellDeliverabilityCurve.fromVogel is an analytic Vogel-like shape. fromWellSystem samples a configured WellSystem; failed/nonphysical samples are clamped to zero and sampled rates are forced non-increasing with pressure. Check the source data and fit envelope.
• runProfile(years, dtYears) records ceil(years/dtYears)+1 points (t = 0 to the endpoint); depletion and cumulative production cover exactly years (the extra post-endpoint step was removed 2026-10).
• TieInCapacityPlanner's nameplate-only mode is a rate-allocation screen. Process-aware mode needs a real HostFacility process/model plus a correct HostTieInPoint; ProcessModel stream names should be area-qualified, such as gathering::Host Feed, when duplicates exist. Planner tests confirm the base host process is restored after trial loads.
• A tie-in stream scaled in rate keeps its composition. When GOR, water cut or field mix change over the profile, call HostTieInPoint.setFeedProvider(new RecombinedHostFeedProvider(reservoirFluid)) so the planner sets composition and molar rate from each trial load (the stream fluid must have the reference fluid's components in the same order). HostFeedProviderTest covers the recombination, the planner wiring and restoration of the stream. For yearly Centuries-based molar feeds see the enterprise enterprise-rate-to-fluid-feed skill.
• Lifecycle modification plans identify candidate scope; they do not size or cost a real equipment modification. Rebuild/rerun the facility and add verified modification CAPEX before claiming production or NPV improvement.
• Concept-generated fluids, equipment sizing and product limits are only as strong as the supplied inputs and chosen model. CashFlowEngine fiscal behavior is selected/configured; do not treat built-in defaults as current tax advice or infer a jurisdiction's rates from this guide.
• ReservoirCouplingExporter emits coupling tables/keywords; it does not run ECLIPSE, INTERSECT, OPM Flow or a full reservoir simulator.

Validation / benchmarks

• IntegratedProductionTest covers curve monotonicity/interpolation, single-well solve, reservoir-drive depletion, gas-lift budget, PI/Vogel test matching, field solve, declining profile and capacity-limited optimization. Reuse its deterministic fromVogel/fromArrays fixtures for a quick model check.
• TieInCapacityPlannerTest covers base-first and pro-rata policy, deferred backlog, process-capacity injection, multi-area qualified tie-ins and restoration of the operating point.
• NorwegianOilFieldLifecycleTest is tagged slow and covers lifecycle economics, product specs, facility capacity, host tieback, modification candidates and multi-area process linking.
• A valid network result must converge, keep rates nonnegative, and satisfy rate consistency: field rate is the sum of per-well rates. For a profile, compare recorded rates and time-step integration separately; do not accept cumulative production solely because the run returned.
• Independent reservoir/well or fiscal benchmark data is still required for calibration. Do not call a model validated because its internal unit tests pass.

Related skills

• neqsim-field-development — lifecycle framing, concept selection and integrated field workflows.
• neqsim-field-economics — cost, cash-flow, tax, NPV/IRR, uncertainty and sensitivity assumptions.
• neqsim-production-optimization — production profiles, decline analysis and reservoir surveillance.
• neqsim-subsea-and-wells — well design, SURF costs and tieback/host context.
• neqsim-agentic-process-optimization — closed-loop optimization of an already-built process model.
• neqsim-process-modeling — detailed flowsheet construction and convergence checks.
• neqsim-e300-fluid-io — Eclipse E300 fluid import/export; distinct from VFP table generation.
• neqsim-vfp-lift-curves — VFPPROD/VFPINJ lift curves from well, flowline and topside models.
• neqsim-reservoir-facility-value-chain (community) — p/z history match, well law, facility capacity and supply/capacity balance for a gas chain.

VFPPROD / VFPINJ lift curves

Deliverable lift curve tables (tubing, flowline and full-system curves, axis design, rate caps, validation and traps) are covered by neqsim-vfp-lift-curves. ReservoirCouplingExporter.generateVfpProd is a placeholder unless a VlpProvider is set with setVlpProvider(...); use VfpTableBuilder for deck output.