Parameter sweep and batch comparison
Pick device parameters and a range, let the engine sweep the whole grid with a real solver call in every cell, then read the comparison table and the best cell. Compare designs on two or three objectives at once (Pareto front with a knee), and rank the economics by sensitivity. Nothing is interpolated, and anything the engine does not return is shown as "—", never as zero.
Example case — every figure below is computed by the engine, none is entered by hand. In-memory reference case: 10 kV grid (100 MVA, X/R 5), one 630 kVA Dyn11 transformer, LV busbar, one 180 kW static load, one 100 kWp PV plant and one 150 kVA / 300 kWh BESS. All cables are explicitly modelled (copper), so the cable-section sweep really changes the result. Opening this page costs 0 engine requests.
Example case — every figure below is computed by the engine, none is entered by hand.
Example case — every figure below is computed by the engine, none is entered by hand.
Basis: DOE grid traversal with a real solver call per cell (engines/hosting-capacity.js load flow plus IEC 60909 short circuit, engines/lcc-evaluate.js economics); Pareto front by non-dominated sorting over real per-candidate solver calls; sensitivity by perturbing each economic variable around its own base value with the same LCC engine. An over-large grid or an insufficient solver budget is refused with a suggestion instead of being silently truncated, and a cell that cannot be solved is reported as not computable rather than invented.