Dry gas seal condensation and conditioning screening

DryGasSealAnalyzer screens seal-gap isenthalpic expansion, a pressure/temperature condensation grid, dead-leg cooling and condensate accumulation. It also estimates conditioning thermal duties. It does not establish a complete compressor/seal operating permissive, filter cleanliness, hydrate-free operation or protection-system adequacy.

Inputs and physical basis

Supply the composition and mixing rule through setSealGas. The analyzer keeps an owned clone. Configure cavity and primary-vent pressures with explicit units; bara is absolute. The inlet-to-vent pressure interval must be positive and ordered. Set temperatures in C or K. All setters invalidate previous analysis evidence, so rerun after a basis change.

setSealLeakageRate accepts NL/min, Nm3/hr or kg/hr. Normal volumes are ideal-gas molar equivalents at 273.15 K and 1.01325 bara, not actual volumetric flow at seal pressure or an EOS-flashed liquid volume at standard conditions. Mass leakage is converted using the supplied composition’s initialized molar mass, even if flow was entered first. Unknown leakage units, negative flow and nonfinite flow are rejected.

setGCUSupplyFlowNLmin specifies conditioning inlet throughput independently of leakage. Without it, the legacy screening assumption remains three times primary leakage. Actual supply includes the process-side labyrinth flow and other consumers; derive it from the OEM seal/support-system design rather than treating leakage as total demand.

The expansion and TP grid sum hydrocarbon and aqueous liquid volumes, including single liquid states. The grid spans the complete configured vent-to-cavity pressure range and starts 15 K below the lower of ambient and cavity temperature. Both scans report expected point counts, failed point counts and calculation_complete.

isSafeToOperate() is retained as a legacy name. A true value means only the two sampled condensation scans pass the legacy 0.01 vol-percent liquid threshold. Missing/incomplete scans, nonfinite fractions and vent-flash errors cannot pass. A grid cannot exclude a narrow unsampled condensation region; the result does not establish OEM superheat margin, continuous-path dryness, gas cleanliness or compressor protection readiness.

GCU thermal duties and composition

The legacy design screen samples dew temperatures across pressures and chooses its maximum as the cooling-pressure basis. This approximates a maximum dew temperature; it is not a rigorously traced cricondentherm and does not design the booster required if conditioning pressure is below seal supply pressure. No valid dew-temperature solution means the GCU requirement is unresolved, not proof that conditioning is unnecessary.

Cooling flashes the feed at the selected pressure. Only the gas phase survives ideal liquid separation. Reheating uses that retained composition and molar flow:

\[Q_h=\dot n_{gas}(h_{out}-h_{gas,cold})\]

Here $Q_h$ is W, $\dot n_{gas}$ is mol/s and the enthalpies are J/mol. The results expose dry_gas_composition, gas_outlet_mol_fraction, dry_gas_molar_flow_mol_s and enthalpy values so the calculation can be checked independently. Reheat target conservatively uses the feed’s sampled maximum dew temperature plus the configured margin; a full conditioned-gas downstream envelope and heater/booster design remain to be qualified.

cooling_duty_kW, reheat_duty_kW and total_thermal_duty_kW are thermal loads. The legacy total_electrical_kW key remains but is NaN: electrical demand requires a refrigeration COP, heater efficiency and booster/auxiliary power model. Consumers must handle nonfinite values; toJson() retains its existing special-floating-value behavior.

Margins in setGCUMargins are user/OEM inputs. The legacy 17 K defaults are screening assumptions and must not be described as universal API 692 requirements.

Design philosophy and validation

A complete conditioning model should connect supply/source selection or a booster, pressure regulation, cooling/separation, coalescing/particulate filtration and reheating to named NeqSim streams. Check hydrocarbon and water condensation after JT letdown, minimum ambient/dead-leg temperature, standby/startup/shutdown, hydrate conditions, filter differential pressure, heater capacity, liquid removal and component/energy balance. The current analyzer does not model these as one validated composable supply system. Its fixed-pressure dead-leg cooling and pressure-impact estimates remain simplified.

DryGasSealEvidenceTest checks water/single-liquid detection, incomplete/nonfinite evidence, normal-flow conservation, independently flashed separated composition and retained-gas reheating balance using public synthetic fluids. These are regression and conservation checks, not independent experimental qualification of the EOS or seal model.

Follow the acceptance criteria in issue #4144 and the program roadmap #4149. Primary technical references include the John Crane Type 28 AT operation guide and EagleBurgmann seal management systems. Apply the requirements for the actual supplied seal model and project.