Energy flow (off-grid)

What gets simulated

At each time step the simulator works out a strict priority order. Load is served first, drawing on whatever solar is available right now and topping up the deficit from the battery. When solar exceeds load, the surplus charges the battery up to its usable capacity. Once the battery is full, any remaining solar is curtailed (dumped) — there is nowhere else for it to go.

When the combination of solar plus battery still can’t cover the load — usually a stretch of overcast days where the battery has already drained to its minimum reserve — the simulator records a shortfall. In real life this is where the lights dim, the fridge cycles off, or you have to fall back to a generator / shore power / powered site.

The battery is modelled with a real-world round-trip efficiency, a minimum reserve (so the chemistry doesn’t drop into damaging deep-discharge territory), and a maximum charge / discharge rate. Off-grid mode never makes export-vs-import trade-offs because there is no grid — the simulator’s job is to surface daily headroom (how much extra load the budget can absorb) and shortfall risk (how far the load exceeds the daily budget).

What each series means

• Solar (orange filled area) — the instantaneous solar production from your panel groups, after temperature derate and system efficiency.
• Load (black line) — the modelled site / van / off-grid home demand at that time of day.
• Battery charge (blue line, right axis), in kWh. Bounded by usable capacity and the configured minimum reserve.
• Curtailed (dashed purple, only when present) — solar the system couldn’t store because the battery was already full.
• Shortfall (bold red, only when present) — load the system couldn’t meet from solar plus battery. This is where you would fall back to a generator or move to a powered site.

Forecast uncertainty band — what it means off-grid

The faint amber band around the solar curve is the forecast uncertainty. It’s symmetric (real weather can surprise either way) and widens with how far ahead the date is — today’s band collapses to a point because the day’s weather is already in. The band is clamped on the lower side at the maximally-cloudy day’s diffuse-only floor (≈ 10 % of clear-sky) and on the upper side at clear-sky production, so the visualisation never claims a number outside what physics allows.

Off-grid users should read the band as shortfall risk, not just as “how much production might vary”. Grid-tied users mostly care about kWh totals translating to dollars; off-grid users care about whether a cloudy day flips the system from “comfortable” to “not enough to cover load by sunset”. When the lower edge of the band on a forecast day dips below the load curve for long enough to drain the battery to its minimum reserve, that’s a marginal-shortfall day — the forecast-confidence number controls how aggressive the simulator is about flagging those.

For the multi-day picture (Date range, Week ahead), the per-day band drives the Low / High scenarios in the Action plan and Week-ahead summary. “Low” is what happens if every day comes in at the bottom edge of its band; “High” is the top edge. Marginal-shortfall days are the ones that show up as shortfall in “Low” but not in “Mean” — those are where you’d want a generator-run plan or a powered-site fallback.