Iron-sulfide wall inventory and elemental-sulfur source

IronSulfideWallInventory and IronSulfideOxidationSource represent a source that a fluid-only equilibrium calculation cannot retain: corrosion product accumulated during an earlier wet or sour period. The source can be connected directly to NeqSim’s existing S8 solid flash, sulfur filter, and compressor deposit models.

The intended root-cause chain is:

water/seawater + carbon steel + H2S -> wall FeS inventory
wall FeS + later O2 ingress          -> S0 source represented as S8
S8 + entrained condensate            -> transport to compressor
warm local shaft/seal surface         -> condensate evaporation and solid-S8 deposition

This is a screening and inventory model, not a universal corrosion-rate correlation. Mineralogy, scale ageing, porosity, water coverage, oxygen mass transfer and competing sulfate/thiosulfate formation can change the rate and elemental-sulfur selectivity by orders of magnitude. Consequently, all kinetic defaults are zero and an uncalibrated calculation reports low/base/high S0 rates.

Model basis

The implemented bookkeeping reactions are:

[ \begin{aligned} \mathrm{Fe + H_2S} &\rightarrow \mathrm{FeS + H_2}
\mathrm{FeCO_3 + H_2S} &\rightarrow \mathrm{FeS + CO_2 + H_2O}
\mathrm{Fe_2O_3 + 3H_2S} &\rightarrow \mathrm{2FeS + S^0 + 3H_2O}
\mathrm{4FeS + 3O_2} &\rightarrow \mathrm{2Fe_2O_3 + 4S^0} \end{aligned} ]

The last equation is the default Fe2O3-equivalent screening branch. The oxygen stoichiometry and fraction ending as S0 are configurable because partially oxidised sulfur and iron products are expected in real scales. Iron oxide is stored on an Fe2O3-equivalent iron-inventory basis; it does not assert that hematite is the only wall phase.

The scientific basis is the established low-temperature Fe-S reaction network and its metastable phases: Rickard and Luther’s review of iron-sulfide chemistry (doi:10.1021/cr0503658), Benning, Wilkin and Barnes’ experimental reaction pathways below 100 °C (doi:10.1016/S0009-2541(99)00198-9), and the structural transformation measurements of Csákberényi-Malasics et al. (doi:10.1016/j.chemgeo.2011.12.009). These sources support retaining a polymorph/reactivity category and explicit uncertainty; they do not justify one plant-independent oxidation rate.

Java example

IronSulfideWallInventory wall = new IronSulfideWallInventory(250.0, 80.0, 20.0);
wall.setPipeSurfaceAreaM2(1200.0);
wall.setPorosityFraction(0.35);
wall.setWettedFraction(0.20);
wall.setSurfaceAreaMultiplier(2.0);
wall.setFeSPhase(IronSulfideWallInventory.FeSPhase.MIXED);
// Alternative to direct masses: total thickness + skeletal density + lab mass fractions.
// wall.setMassesFromMeasuredScaleThickness(0.001, 4300.0, 0.70, 0.20, 0.10);

IronSulfideOxidationSource source =
    new IronSulfideOxidationSource("pipeline wall source", oxygenContainingGas, wall);
source.setFeCO3SulfidationFractionPerHour(1.0e-4);
source.setFeSOxidationFractionPerHour(2.0e-3);
source.setOxygenTransferFraction(0.10);
source.setElementalSulfurYieldFraction(0.50);
source.setElementalSulfurYieldBounds(0.10, 1.00);
source.setKineticUncertaintyFactor(3.0);
source.run();                         // rate calculation; inventory is unchanged

double s0 = source.getElementalSulfurRateKgPerHour();
String report = source.toJson();

// Advance a one-hour oxygen-containing nitrogen purge and retain its history.
IronSulfideWallInventory.ExposureEvent purge =
    new IronSulfideWallInventory.ExposureEvent(
        IronSulfideWallInventory.ExposureType.NITROGEN_PURGE,
        1.0, 30.0, 0.0, "Nitrogen purge with measured residual oxygen");
source.runExposure(purge, UUID.randomUUID());

run() deliberately does not mutate the historical wall inventory. Use runTransient(dt, id) or runExposure(event, id) to consume/form material over time.

Condensate and warm-shaft coupling

The outlet contains generated elemental sulfur as the existing S8 component. It can therefore feed SolidFlashDepositSource. Include separator carry-over or another estimate of entrained condensate in the connected stream. A local temperature can be supplied explicitly or read from a solved compressor thermal node:

SolidFlashDepositSource deposit = new SolidFlashDepositSource(
    source.getOutletStream(), "S8", DepositMechanism.SULFUR_S8, 0.5);
deposit.setThermalNode(compressor, CompressorThermalModel.INLET_SHAFT);

double precipitated = deposit.getPrecipitationRate("kg/hr");
double deposited = deposit.getDepositRate("kg/hr");
double evaporatedCondensateFraction = deposit.getLastLiquidEvaporatedFraction();

The evaporation metric compares equilibrium liquid flow at the bulk inlet with liquid remaining at the local P-T condition. It indicates whether a warm shaft can remove the liquid carrier at that location; the capture fraction still needs plant evidence or a sensitivity range.

Recommended RCA inputs

scale mass or thickness and sampled mineralogy (XRD/SEM-EDS/XPS where available);
pipe internal area, wet fraction, porosity, roughness/reactive-area multiplier;
water, seawater, hydrotest, shutdown, purge and restart history;
measured O2 concentration and duration, not only the nitrogen specification;
entrained-liquid rate and condensate composition at compressor suction;
local shaft/seal/bearing temperatures or compressor thermal-model estimates;
deposit composition and sulfur oxidation state;
uncertainty cases for kinetics, S0 yield, oxygen transfer and capture efficiency.

Avoid interpreting the maximum sulfur contained in FeS as an hourly rate. It is an inventory ceiling; the released rate is the minimum of reaction kinetics, available FeS and oxygen transferred to the wall.

Example notebook

examples/notebooks/IronSulfideWallSulfurDeposition.ipynb uses the reported 10 kg/h nitrogen stream with 2% oxygen as a screening example and connects the wall source to a local warm-shaft solid flash.