LH2PoolX

LH2PoolX calculates an evidence-qualified, quasi-steady liquid-hydrogen (LH2) pool source term for one declared time window. It is an upstream input to a dispersion calculation. It is not a dispersion solver, an impact/impingement model, or a general transient pool-spreading model.

The package makes the source assumption explicit. It does not silently replace unknown liquid deposition, pool growth, inventory, or drainage with a fitted parameter.

1. Select the appropriate calculation route

Neither route predicts an unknown impingement footprint. For a downward jet, the impact, droplet, and release-to-ground handoff must be separately justified before the declared-release route is used.

2. Install

LH2PoolX supports Python 3.10 or later. CoolProp is installed automatically.

python -m pip install --upgrade lh2poolx
python -c "import lh2poolx; print(lh2poolx.__version__)"

For a reproducible calculation, record the printed LH2PoolX version, the CoolProp version, all source inputs, the selected route, and source_status.

3. Route 1: declared release and deposition fraction

Use this route only when the release-to-ground handoff is an explicit scenario input. The calculation chain is:

1. saturated-liquid, isenthalpic flash from the declared storage pressure to the declared ambient pressure;
2. liquid mass reaching the ground after the explicit deposition_fraction;
3. heat-limited evaporation flux from a semi-infinite ground-conduction relation, subject to the LH2 critical-heat-flux cap; and
4. an unconfined equilibrium area whose evaporation balances the declared liquid input.

from lh2poolx import LH2Release, evaluate_pool_source

release = LH2Release(
    rate_kg_s=0.1055,          # declared upstream LH2 release rate
    storage_pressure_barg=1.0, # storage gauge pressure
    deposition_fraction=0.50, # declared liquid fraction reaching the ground
)

source = evaluate_pool_source(
    release,
    elapsed_s=300.0,           # time since pool formation for ground conduction
    ambient_temperature_K=282.0,
)

print(f"flash fraction:     {source.flash_vapour_fraction:.3f}")
print(f"liquid to ground:   {source.liquid_to_ground_kg_s:.4f} kg/s")
print(f"pool radius:        {source.radius_m:.3f} m")
print(f"pool area:          {source.area_m2:.3f} m2")
print(f"evaporation source: {source.evaporation_rate_kg_s:.4f} kg/s")
print(f"status:             {source.source_status}")

Inputs

Read the result correctly

evaporation_rate_kg_s is the mass source to pass to a downstream dispersion route. It is not necessarily the original upstream release rate.

If source.confined is True, or source_status is confined_pool_requires_inventory_model, the declared radius cannot evaporate all of the liquid input. liquid_accumulation_rate_kg_s then identifies the unremoved liquid rate. Do not use that case as an unqualified steady source: add an inventory, spreading, drainage, or overflow model.

4. Route 2: declared observed footprint

Use this route when a record constrains pool extent more credibly than it constrains the fraction of the release that deposited on the ground. It deliberately has no release rate or deposition_fraction input.

from lh2poolx import evaluate_observed_footprint_source

# Retain both bounds when a record reports a range. Do not use a midpoint.
for radius_m in (0.5, 1.0):
    source = evaluate_observed_footprint_source(
        equivalent_radius_m=radius_m,
        elapsed_s=300.0,
        ambient_temperature_K=282.0,
    )
    print(
        f"R = {source.equivalent_radius_m:.1f} m | "
        f"A = {source.area_m2:.3f} m2 | "
        f"evaporation = {source.evaporation_rate_kg_s:.4f} kg/s | "
        f"required steady liquid inflow = "
        f"{source.required_liquid_ground_rate_kg_s:.4f} kg/s"
    )

equivalent_radius_m is the radius of a circle having the declared footprint area. It is a geometric source assumption, not a claim that the physical pool was circular.

required_liquid_ground_rate_kg_s is the quasi-steady liquid inflow that would sustain the calculated evaporation. It is not an observed deposition rate.

The FFI large-scale LH2 report describes the vertical-release pools as remaining within roughly 0.5-1.0 m of the release point while noting that fog prevented precise visual verification. That evidence can support a 0.5-1.0 m conditional radial-extent envelope. It does not support fitting a single footprint, evaporation history, or deposition fraction. See the FFI report.

5. Pass the source to a dispersion model

LH2PoolX deliberately has no dependency on SLABx, DEGADISx, or another dispersion package. Pass only the quantities accepted by the downstream route, along with the source assumption:

# source can be either PoolSource or ObservedFootprintSource
dispersion_input = {
    "source_mass_rate_kg_s": source.evaporation_rate_kg_s,
    "source_area_m2": source.area_m2,
    "equivalent_radius_m": (
        getattr(source, "equivalent_radius_m", None)
        or getattr(source, "radius_m", None)
    ),
    "source_status": source.source_status,
}

For the observed-footprint route, retain both envelope bounds in the dispersion study. The outputs are alternative, declared source assumptions; they are not a statistical confidence interval and should not be averaged into one best source.

6. Output fields

7. Substrate and physical limits

CONCRETE_CRYOGENIC is the default effective substrate parameter set. It is tied to the restricted PRESLHY E3.4 Concrete02 component check and must not be treated as a universal concrete correlation. Provide another Substrate only when its conductivity and diffusivity are independently justified for the site and temperature range.

from lh2poolx import Substrate, evaluate_observed_footprint_source

site_substrate = Substrate(
    name="declared_site_substrate",
    conductivity_W_mK=1.5,
    diffusivity_m2_s=8.0e-7,
)

source = evaluate_observed_footprint_source(
    equivalent_radius_m=0.75,
    elapsed_s=300.0,
    substrate=site_substrate,
)

LH2PoolX does not resolve jet trajectory, impingement, splashing, droplet transport, rainout, transient pool growth, liquid inventory, drainage, barriers, slope, detailed boiling transitions, or three-dimensional vapour dispersion. Use an explicit upstream model or a conditional scenario envelope when those processes govern the decision.

8. Error handling and reproducibility

The package rejects non-positive mass rates, radii, elapsed times and thermodynamic pressures, as well as deposition fractions outside 0-1. Treat an error as an input-definition issue; do not replace an invalid value silently.

To reproduce the tests from a source checkout:

git clone https://github.com/lyullee/lh2poolx.git
cd lh2poolx
python -m pip install -e .[test]
python -m pytest -q

The concrete heat-transfer implementation has a restricted component check against published PRESLHY E3.4 Concrete02 mass-loss windows. It is not validation of another substrate, a complete pool-growth model, or an unknown jet-to-ground handoff.

9. Citation and license

LH2PoolX is released under the MIT License. Cite the released software version and the experimental and heat-transfer sources used in an analysis. Citation metadata are provided in CITATION.cff.

Research reproducibility update 0.1.3

The released model equations and public API are unchanged from 0.1.2. The model equations and code map describe the actual implementation, input boundary, numerical operators and inventory accounting.