What is conversion efficiency (η)?
When molten metal contacts water, the thermal energy stored in the melt can convert into a mechanical pressure wave — a steam explosion or Fuel-Coolant Interaction (FCI). The energy conversion efficiency (η) is the fraction of available thermal energy that actually becomes destructive blast work:
η = Emechanical / Ethermal (expressed as %)
Why does it matter so much?
This single parameter has the largest impact on predicted consequences. At 100% efficiency, the model predicts maximum possible damage — the theoretical upper bound where all thermal energy converts to blast. In reality, most energy dissipates as steam venting, radiation, and heating of surrounding structures. The difference is enormous:
| Efficiency | TNT-eq (1000kg steel, 200L water) | Use Case |
|---|
| 100% | ~Full calculated value | Safety case / regulatory worst-case |
| 1.5% (steel best-estimate) | ~1/67th of full value | Realistic scenario analysis |
| 3.07% (max ever observed) | ~1/33rd of full value | Conservative-realistic upper bound |
What does the experimental evidence say?
No FCI experiment has ever exceeded ~3% conversion efficiency. Key findings from major international research programmes:
| Programme | Facility | Max η Observed | Notes |
|---|
| SERENA Phase 2 | OECD/NEA (multi-lab) | 3.07% | Absolute maximum across all international programmes |
| KROTOS | CEA Grenoble | ~1.5% | Alumina (Al₂O₃) melt, external trigger |
| TROI | KAERI Korea | ~0.5% | UO₂-ZrO₂ prototypic corium |
| FARO/KROTOS | JRC Ispra | ~0.3% | Large-scale (up to 225 kg) prototypic melt |
| SNL OG/FITS | Sandia National Labs | 1–3% | Fe-Al₂O₃ thermite — closest analogue to molten steel |
| ZREX | ANL Argonne | ~2% | Zr/Zr-Fe metallic melt |
Important: No direct FCI experiments with industrial molten steel exist in the published literature. Values for steel are extrapolated from Fe-Al₂O₃ thermite data (Sandia OG series), which is the closest available analogue at similar temperatures and densities.
How to choose the right mode:
100% Conservative — Use for: safety cases, regulatory submissions, facility siting studies (API RP 752/753), insurance assessments, any context where you must demonstrate worst-case consequences. This is the Baker (1983) upper-bound method used by most blast analysis standards.
Variable (Research-Based) — Use for: realistic scenario analysis, comparing relative severity of different scenarios, operational risk assessment, incident investigation reconstruction, emergency planning zone estimation. The model automatically calculates η based on your material, confinement, contact mode, and superheat using published correlations.
Custom Value — Use when: you have site-specific experimental data, corporate engineering standards specify a value, you are conducting sensitivity analysis, or you need to match a value used in a peer calculation for comparison. If unsure, use 3% as a conservative-realistic upper bound (maximum ever observed in experiments).
Key physical factors that affect real efficiency:
1. Melt fragmentation — Fine particles (<250 μm) transfer heat faster, increasing efficiency. Coarse pouring → larger fragments → lower η. (Ciccarelli & Frost, Nuclear Engineering and Design, 1994)
2. Confinement — Enclosed spaces reflect pressure waves, amplifying mechanical coupling. Open-air spills dissipate energy into the atmosphere.
3. Contact mode — Melt falling into a deep water pool (stratified) vs. water trapped under melt (trapped) vs. water sprayed onto melt surface produce very different interaction geometries.
4. Superheat — Efficiency peaks at moderate superheat (~370°C above liquidus) then declines at extreme temperatures due to Leidenfrost vapour film stabilisation (SERENA Phase 2 data).
5. Water volume ratio — A minimum coolant-to-melt mass ratio is needed for sustained interaction; too little water flashes to steam immediately with minimal pressure buildup.
Key References: Corradini, M.L. et al., "Vapor Explosions in Light Water Reactors: A Review of Theory and Modeling," Progress in Nuclear Energy, 22(1), 1988. | OECD/NEA SERENA Programme Reports, Phase 1 (2006), Phase 2 (2014). | Sehgal, B.R., "Nuclear Safety in Light Water Reactors: Severe Accident Phenomenology," Academic Press, 2012. | SNL STEX-II Database (Fe-Al₂O₃ test series). | Ciccarelli & Frost, Nuclear Engineering and Design, Vol. 149, 1994.