Sensitivity Leverage Scan¶
Decision question¶
Which uncertain input should receive the next measurement dollar because it can most change the recommendation?
Baseline metric¶
Primary decision metric: variable operating cost, $/t olefin product.
One-at-a-time local sensitivity (+/-10%)¶
| Input | Delta metric (%) | Relative leverage rank |
|---|---|---|
| Steam price | +/-7.8 | 1 |
| Ethylene selectivity | -6.4 / +6.1 | 2 |
| Reactor conversion | -4.2 / +4.0 | 3 |
| Electricity price | +/-1.9 | 4 |
| Column pressure drop | +/-1.3 | 5 |
Interpretation¶
- Steam price and selectivity dominate near the current operating point.
- This matches the top risks in the Assumption Failure Map, confirming alignment between risk and leverage.
- Pressure-drop uncertainty is currently second-order.
Action choice (this week)¶
- Validate steam contract assumptions with procurement by Friday 17:00.
- Add one catalyst performance data point in the next lab run and refresh selectivity bounds before the next decision gate.
What would change this ranking¶
- If steam procurement is fixed by contract, selectivity becomes rank 1.
- If purity constraints tighten, pressure-drop and separation variables may move up.
Mindsets demonstrated¶
Rubric snapshot (example)¶
- Quantitative discipline: 4/4
- Tradeoff clarity: 3/4
- Decision defense: 3/4
Next in sequence: Decision Memo