ECLSS Monte Carlo Reserve-Margin Simulator (2D)

Mission Parameters

Crew size 4

Mission duration 180 d

Parameter uncertainty 18%

Monte Carlo runs 600

Live Statistics

Mean reserve margin

— %

Stockout probability

— %

Dominant uncertainty driver

—

Runs simulated

How it works

Each run integrates a mass-balance ODE for the O₂ reserve buffer R(t), in units of the initial reserve R₀:

dR/dt = G·(1 − δ) − c·n·(1 + εc) − ℓ·(1 + εl)

G  = nominal regenerator output
δ  = regenerator efficiency loss  ~ U(0, σ/2)
c  = per-crew consumption rate (0.84 kg O2/day)
n  = crew size
εc = consumption noise ~ N(0, σ)
ℓ  = cabin/EVA leak overhead
εl = leak noise ~ N(0, σ)

σ is the uncertainty slider. Each of the N sampled runs is stepped forward with a clamped Euler integrator (Δt derived from the mission length) for the full mission duration; R is clamped to zero the moment it would go negative — that run is flagged a stockout. The top panel plots every visualised run's own reserve trajectory as a line sweeping left→right (time) and up→down (reserve fraction), colored green if it finishes the mission and red if it stockouts, with a blue mean line and shaded ±1σ envelope. The bottom-left panel is a histogram of final margins across the whole batch; the bottom-right panel ranks the three stochastic inputs by |Pearson correlation| with the final margin — the same one-at-a-time sensitivity ranking used to prioritize ECLSS hardware margins before flight. Drag inside the trajectory panel to pan the time axis, scroll/pinch to zoom it.

Verified against the 3D source: the nominal (noise-free) drift is G−c·n−ℓ = −0.05·(c·n), and R₀ is sized to 0.55·(c·n)·duration, so a zero-noise run ends at margin 1 − 0.05/0.55 ≈ 90.9% regardless of mission length — confirmed numerically here, matching the "3% regenerator shortfall + 2% leak overhead" design note in the original.

Run completes mission Run stockouts Mean ± 1σ envelope

Drag to pan · scroll to zoom