Battery or More Panels? What Actually Raises Self-Consumption

Method note

Worked example for an 8 kW south-facing rooftop at Tehran coordinates (35.6892°N, 51.3890°E). Monthly AC production comes from a JRC PVGIS PVcalc run dated 2026-07-23 (8 kWp and 10 kWp, tilt 33°, south, 14% system loss). An independent hourly pvlib 0.11.2 cross-check on PVGIS TMY for the same 8 kW setup (run 2026-07-23) yields about 12,639 kWh/year, within about 2% of the PVGIS PVcalc annual total used in the tables below. Self-sufficiency rates without and with storage follow published residential ranges from IEA PVPS Task 12 LCA work and related self-consumption literature (about 30–37% SSR without a battery; higher with 5–10 kWh storage). Annual household use is a stated 5,400 kWh design load with a mild summer peak. Export credit is a labeled variable (`export_weight`), not a utility promise. Rebuild with your bills before decisions. Method companion: solar ROI payback method.

Direct answer

On this Tehran climate run the 8 kW array yields about 12,876 kWh/year (~1,609 kWh/kW). With a 5,400 kWh home load and ~35% self-sufficiency (no battery), only about 1,890 kWh is used on site and the rest exports. Adding a 10 kWh LFP pack and raising SSR to about 57% (IEA Task 12 scale for a ~10 kWh residential pack) lifts self-consumed energy to about 3,078 kWh (+1,188 kWh). Upsizing the array to 10 kW without storage (~16,095 kWh/year) only nudges self-consumption to about 2,052 kWh (+162 kWh) because midday surplus grows faster than evening load. When export credit is weak (`export_weight` 0.2), the battery scenario also wins on annual benefit index. When export is credited near retail, extra PV wins on index even if self-consumed kWh barely moves. That is the 2026 storage lesson in one table: batteries buy timing, not nameplate watts. Background: self-consumption explained and battery storage guide.

Hypothesis

If an 8 kW south-facing Tehran rooftop serves a 5,400 kWh/year household with evening-heavy use, then a modest LFP pack raises self-consumed kWh more than spending the same design step on extra PV capacity when export credit is low, because storage shifts midday surplus into night hours while upsizing mainly increases export.

Inputs

Monthly production (8 kW, PVGIS)

(Month rows round to whole kWh; PVGIS annual total is 12,876 kWh.)

Figure: hourly pvlib monthly shape for the same 8 kW setup (~12,639 kWh/year). The table above stays on PVGIS PVcalc values used in the scenarios.

Monthly energy-match offset (teaching upper bound)

Consumption shape is a stated mild summer peak totaling 5,400 kWh. Production is the 8 kW PVGIS row. Formula: self-consumed = min(production, consumption); exported = max(0, production − consumption); imported = max(0, consumption − production). This monthly energy match ignores hour-level timing, so imports look near zero. Real homes still import after sunset; the SSR scenarios below are the realistic teaching layer.

Scenario comparison: self-consumed kWh

Self-consumed lift from a 10 kWh pack (+1,188 kWh) dwarfs the lift from +2 kW of PV without storage (+162 kWh). Upsizing still adds ~3,219 kWh of annual production, almost all as export.

Annual benefit index

Index = self-consumed × 1.0 + exported × `export_weight`.

Installed cost is omitted as a single currency figure. Compare quotes by asking what you pay per extra self-consumed kWh versus per extra exported kWh. See complete guide to solar ROI and solar proposal checklist.

What to do with these tables

1. Pull 12 months of billed kWh and sketch daytime vs evening share. 2. Run PVGIS or PVWatts for your coordinate; keep the run date. 3. Apply a no-storage SSR near 30–40% unless you have monitored hour data. 4. Size storage from evening deficit, not from total annual export. Hybrid inverters and DC vs AC coupling change how that pack connects. 5. Stress-test `export_weight` before choosing upsize vs storage.

Closing

For this Tehran PVGIS climate and IEA-scale self-sufficiency bands, a 10 kWh LFP pack buys far more on-site kWh than adding 2 kW of PV when the home already exports heavily. Extra modules still win when every exported kWh is worth nearly as much as a self-consumed one. Decide with your tariff in writing, then rebuild the index with your own load shape.