ECSS Space Environment — SEL/SESB Rate Prediction (space-systems/ecss/e1012-sel)
Use when the task is predicting the single-event latch-up (SEL) or
single-event snapback (SESB) rate for a spacecraft electronic device
per ECSS-E-ST-10-12C §9.4.1.4–9.4.1.5 — fitting measured cross-section
data, integrating over the mission particle environment, and categorizing
the combined rate against acceptance thresholds.
Domain quick reference
- SEL (Single-Event Latch-up): a high-current state triggered in a CMOS
device when a heavy ion or high-energy proton/neutron deposits sufficient
charge in the parasitic thyristor structure to switch it on. Left
unmitigated it can destroy the device. It is characterised by a
threshold LET (heavy-ion path) or threshold proton energy (proton path)
below which the event cannot occur.
- SESB (Single-Event Snapback): an analogous high-current state in
n-channel MOSFETs triggered by impact ionisation; treated with the
same rate methodology as SEL.
- Heavy-ion cross-section vs LET is modelled by a Weibull curve:
σ(L) = σ_sat × (1 − exp(−((L − L_th)/W)^s)) for L > L_th, else 0.
Parameters (L_th, σ_sat, W, s) come from measured heavy-ion test data
(e.g. a Brookhaven or GSI heavy-ion facility run).
- Proton/neutron cross-section vs energy is modelled by the Bendel
two-parameter formula: σ(E) = B × (1 − exp(−0.18 × (18/A)^0.5 × (E−A)))^4
for E > A, else 0. Parameter A is the threshold energy (MeV) and B is
the asymptotic cross-section (cm²/device).
- The mission rate for each particle type is the integral of
(cross-section × differential particle flux) over the relevant variable
(LET or energy), computed numerically via the trapezoidal rule over the
environment spectrum supplied by the ECSS-E-ST-10-04 environment model.
- Thresholds for rate categorization (per §9.4.1.5):
- acceptable: combined rate < 1×10⁻⁷ events/device/day
- monitor: 1×10⁻⁷ ≤ rate < 1×10⁻⁵ events/device/day
- critical: rate ≥ 1×10⁻⁵ events/device/day (latch-up protection required)
Workflow
- Obtain heavy-ion test data for the device and fit a Weibull curve
(four parameters: L_th, σ_sat, W, s) to the measured cross-section vs
LET points. If only a single-parameter fit is available, derive the
remaining Weibull parameters from the given effective fluence test data.
Reject any device for which no heavy-ion characterisation data exist.
- Obtain the mission differential LET spectrum (particles/(cm²·day·
(MeV·cm²/mg))) from the approved environment model for the orbit and
solar condition. Integrate the Weibull cross-section over this spectrum
using the trapezoidal rule to obtain the heavy-ion SEL/SESB rate
(events/device/day).
- Obtain Bendel parameters (A, B) for the device from proton test data
or the literature. Obtain the mission differential proton energy
spectrum (particles/(cm²·day·MeV)) from the same environment model.
Integrate the Bendel cross-section over this spectrum to obtain the
proton/neutron SEL/SESB rate.
- Sum the heavy-ion and proton rates into a combined rate.
- Categorize the combined rate: acceptable / monitor / critical per the
thresholds in the domain quick reference. Flag any device in the
monitor or critical band; a critical result mandates latch-up current
limiting and power-cycle recovery circuitry.
- Check dominance: if one particle type contributes more than 10× the
other, flag it for review — spectrum coverage or model applicability
may need re-examination before closing the assessment.
Pitfalls
- Using a single-point saturation cross-section without fitting the full
Weibull curve: this overestimates the rate at LET values just above the
threshold (where the cross-section is still rising) and can give a
pessimistic result that invalidates an otherwise acceptable device.
- Treating the Bendel model as universally applicable to every CMOS
technology: the model was derived for bipolar devices; for advanced CMOS
nodes the direct ionisation path from proton recoils may dominate, and
the cross-section must then be measured rather than inferred.
- Summing heavy-ion and proton rates without checking which environment
model was used for each: LET spectra and proton spectra must come from
the same orbit epoch and shielding assumption, otherwise the combined
rate mixes inconsistent environments.
- Interpreting a zero computed rate as automatically acceptable: a zero
result usually means the spectrum did not extend above the device
threshold; verify that the spectrum spans the full relevant LET or
energy range before concluding the device is safe.
Behavior contract (gate 3)
The Weibull cross-section fit, Bendel proton model, trapezoidal rate
integration, severity categorization, and combined assessment logic are
exercised by the gate 3 contract test:
scripts/test_e1012_sel.py against scripts/e1012_sel_logic.py
(stdlib unittest, offline). Run:
python3 scripts/test_e1012_sel.py
Compliance
- ECSS standards are freely downloadable (ESA); cite the source and
paraphrase per standards-map.yaml.
- compliance: STANDARDS-REF, gated: false.
1---2name: e1012-sel3description: "Use when compute heavy-ion and proton/neutron single-event latch-up (SEL) and single-event snapback (SESB) rates for a spacecraft electronic device under ECSS-E-ST-10-12C §9.4.1.4–9.4.1.5: fit a Weibull cross-section curve to heavy-ion test data, integrate it against the mission LET spectrum, apply the Bendel two-parameter model to proton cross-section data and integrate over the proton energy spectrum, sum both contributions into a combined rate, and categorize the result as acceptable, monitor, or critical against per-device thresholds. Trigger: ecss, e-st-10-system-scope, sel, sesb, single-event-latch-up, single-event-snapback, heavy-ion, proton, let-spectrum, weibull, bendel."4license: Apache-2.05compliance: STANDARDS-REF6standards:7 - id: ecss8 reference-only: true9gated: false10domain: space-systems11pack: space-systems12compatibility: "agentskills.io SKILL.md; any SKILL.md host (Claude Code, Hermes, OpenClaw)"13metadata:14 source: https://skillmd.com/skills/ashfordeou/e1012-sel15 domain: space-systems16 subdomain: ecss17 tags: [ecss, e-st-10-system-scope, sel, sesb, single-event-latch-up, single-event-snapback, heavy-ion, proton, let-spectrum, weibull, bendel]18 version: 0.1.019 author: Aero Agent Skills20---2122# ECSS Space Environment — SEL/SESB Rate Prediction (space-systems/ecss/e1012-sel)2324Use when the task is predicting the single-event latch-up (SEL) or25single-event snapback (SESB) rate for a spacecraft electronic device26per ECSS-E-ST-10-12C §9.4.1.4–9.4.1.5 — fitting measured cross-section27data, integrating over the mission particle environment, and categorizing28the combined rate against acceptance thresholds.2930## Domain quick reference3132- SEL (Single-Event Latch-up): a high-current state triggered in a CMOS33 device when a heavy ion or high-energy proton/neutron deposits sufficient34 charge in the parasitic thyristor structure to switch it on. Left35 unmitigated it can destroy the device. It is characterised by a36 threshold LET (heavy-ion path) or threshold proton energy (proton path)37 below which the event cannot occur.38- SESB (Single-Event Snapback): an analogous high-current state in39 n-channel MOSFETs triggered by impact ionisation; treated with the40 same rate methodology as SEL.41- Heavy-ion cross-section vs LET is modelled by a Weibull curve:42 σ(L) = σ_sat × (1 − exp(−((L − L_th)/W)^s)) for L > L_th, else 0.43 Parameters (L_th, σ_sat, W, s) come from measured heavy-ion test data44 (e.g. a Brookhaven or GSI heavy-ion facility run).45- Proton/neutron cross-section vs energy is modelled by the Bendel46 two-parameter formula: σ(E) = B × (1 − exp(−0.18 × (18/A)^0.5 × (E−A)))^447 for E > A, else 0. Parameter A is the threshold energy (MeV) and B is48 the asymptotic cross-section (cm²/device).49- The mission rate for each particle type is the integral of50 (cross-section × differential particle flux) over the relevant variable51 (LET or energy), computed numerically via the trapezoidal rule over the52 environment spectrum supplied by the ECSS-E-ST-10-04 environment model.53- Thresholds for rate categorization (per §9.4.1.5):54 - acceptable: combined rate < 1×10⁻⁷ events/device/day55 - monitor: 1×10⁻⁷ ≤ rate < 1×10⁻⁵ events/device/day56 - critical: rate ≥ 1×10⁻⁵ events/device/day (latch-up protection required)5758## Workflow59601. Obtain heavy-ion test data for the device and fit a Weibull curve61 (four parameters: L_th, σ_sat, W, s) to the measured cross-section vs62 LET points. If only a single-parameter fit is available, derive the63 remaining Weibull parameters from the given effective fluence test data.64 Reject any device for which no heavy-ion characterisation data exist.652. Obtain the mission differential LET spectrum (particles/(cm²·day·66 (MeV·cm²/mg))) from the approved environment model for the orbit and67 solar condition. Integrate the Weibull cross-section over this spectrum68 using the trapezoidal rule to obtain the heavy-ion SEL/SESB rate69 (events/device/day).703. Obtain Bendel parameters (A, B) for the device from proton test data71 or the literature. Obtain the mission differential proton energy72 spectrum (particles/(cm²·day·MeV)) from the same environment model.73 Integrate the Bendel cross-section over this spectrum to obtain the74 proton/neutron SEL/SESB rate.754. Sum the heavy-ion and proton rates into a combined rate.765. Categorize the combined rate: acceptable / monitor / critical per the77 thresholds in the domain quick reference. Flag any device in the78 monitor or critical band; a critical result mandates latch-up current79 limiting and power-cycle recovery circuitry.806. Check dominance: if one particle type contributes more than 10× the81 other, flag it for review — spectrum coverage or model applicability82 may need re-examination before closing the assessment.8384## Pitfalls8586- Using a single-point saturation cross-section without fitting the full87 Weibull curve: this overestimates the rate at LET values just above the88 threshold (where the cross-section is still rising) and can give a89 pessimistic result that invalidates an otherwise acceptable device.90- Treating the Bendel model as universally applicable to every CMOS91 technology: the model was derived for bipolar devices; for advanced CMOS92 nodes the direct ionisation path from proton recoils may dominate, and93 the cross-section must then be measured rather than inferred.94- Summing heavy-ion and proton rates without checking which environment95 model was used for each: LET spectra and proton spectra must come from96 the same orbit epoch and shielding assumption, otherwise the combined97 rate mixes inconsistent environments.98- Interpreting a zero computed rate as automatically acceptable: a zero99 result usually means the spectrum did not extend above the device100 threshold; verify that the spectrum spans the full relevant LET or101 energy range before concluding the device is safe.102103## Behavior contract (gate 3)104105The Weibull cross-section fit, Bendel proton model, trapezoidal rate106integration, severity categorization, and combined assessment logic are107exercised by the gate 3 contract test:108 scripts/test_e1012_sel.py against scripts/e1012_sel_logic.py109(stdlib unittest, offline). Run:110 python3 scripts/test_e1012_sel.py111112## Compliance113114- ECSS standards are freely downloadable (ESA); cite the source and115 paraphrase per standards-map.yaml.116- compliance: STANDARDS-REF, gated: false.