A 20 kW rooftop system in Hebei still earns roughly 10,000 RMB a year — but an identical array commissioned after 31 May 2025 no longer holds a fixed export price and carries market risk instead. Model it before you sign, not after.

Why the 2025 Pricing Reform Reset Household Solar Economics

On 9 February 2025, the National Development and Reform Commission (NDRC) and the National Energy Administration (NEA) issued Document 136, effective 1 June 2025. New wind and solar projects must sell power at market prices; projects already operating keep a stabilised mechanism price. For household systems, the practical consequence is blunt: the export side of the meter is now volatile.
The subsidy era ended earlier. China paid a 0.03 RMB/kWh national subsidy on residential PV in 2021 and 2022, and stopped it in 2023. Installations responded exactly as you would expect — residential additions peaked at 43.5 GW in 2023, then fell to roughly 29 GW in 2024, according to China Photovoltaic Industry Association (CPIA) data, even as total national additions hit a record 277.57 GW that year per the NEA.
Cheaper hardware partly offsets weaker tariffs. Module prices fell below 0.70 RMB/W in 2024, and BloombergNEF logged spot module prices near USD 0.09–0.10/W for Chinese TOPCon product. A residential system that cost 3.5 RMB/W in 2022 now installs for 2.6–3.1 RMB/W. The tariff cut and the capex cut are running in opposite directions.

Three Grid Revenue Models Compared

Household PV in China settles under three structures, and the choice dominates everything else. Full export sells 100% of generation at the grid price. Self-consumption with surplus export offsets retail electricity first. Self-consumption plus storage adds a battery to push the self-consumption ratio higher.
| Model | Effective value per kWh | Annual yield per kW | Typical payback |
|---|---|---|---|
| Full export | 0.30–0.39 RMB | 350–500 RMB | 7–9 years |
| Self-consumption + surplus export | 0.42–0.58 RMB blended | 500–700 RMB | 6–8 years |
| Self-consumption + LFP storage | 0.55–0.75 RMB blended | 700–1,000 RMB | 8–11 years |
The logic is simple arithmetic: every kilowatt-hour you avoid buying is worth the retail price, typically 0.52–0.82 RMB/kWh on tiered residential tariffs, while every kilowatt-hour exported is worth the benchmark coal-fired price or less. Self-consumption is worth roughly 1.5 to 2 times export. That single ratio drives the payback spread above.

What a 20 kW Household System Actually Earns: A Hebei Case Study

Take a real configuration: a 20 kW rooftop array on a two-storey village house near Shijiazhuang, commissioned in early 2024, with 36 TOPCon modules, a 20 kW three-phase string inverter and no battery. All-in cost was 62,000 RMB, or about 3.1 RMB/W including mounting, cabling, grid connection and metering.
Annual generation measured 24,100 kWh, a capacity factor near 13.8% and roughly 1,200 full-load hours — normal for Hebei's irradiation. The household consumes about 9,000 kWh a year; 8,000 kWh is self-consumed and 16,100 kWh is exported. At Hebei's benchmark price of 0.3644 RMB/kWh, export earned 5,867 RMB. Self-consumption offset 8,000 kWh at an average 0.52 RMB/kWh, saving 4,160 RMB. Total first-year benefit: 10,027 RMB, a 6.2-year simple payback.
Two caveats matter. First, module degradation of about 0.5% a year trims lifetime output. Second, if the same project had been connected after 1 June 2025, its export volume would clear at market prices — and Shandong and other provincial spot markets have recorded midday prices below 0.10 RMB/kWh, with negative pricing episodes during high-solar hours, according to provincial exchange disclosures. Same roof, materially different risk.

Reading the Tariff Stack: Benchmark Prices and Provincial Spreads

The benchmark coal-fired price is the anchor for legacy export contracts, and it varies meaningfully by province: 0.4530 RMB/kWh in Guangdong, 0.4153 in Zhejiang, 0.3949 in Shandong, 0.3644 in Hebei, and 0.3598 in Beijing, per provincial development and reform commission schedules. A 12% spread between Guangdong and Beijing changes payback by roughly eight months on a 20 kW system.
Retail tariffs matter more than export prices for self-consumption-heavy systems. Residential tiered pricing usually starts near 0.52–0.55 RMB/kWh and rises above 0.80 RMB/kWh in the third tier, so households with air-conditioning, heat pumps or electric vehicle charging capture the highest avoided cost. Households with tiny consumption and a large roof should not pretend to be self-consumption projects; they are exporters.
Provincial policy is where the real differentiation hides. Several provinces have used local feed-in floors, capacity-based incentives or green certificate channels to soften the market transition, and these expire. Always verify the current provincial document rather than a 2023 quotation sheet from a salesperson.

Battery Storage Changes the Math — Above a Threshold

LFP (lithium iron phosphate) is now the default household chemistry: 6,000+ cycles at 80% depth of discharge, round-trip efficiency of 90–95%, and a thermal profile that suits garages and utility rooms. Compare that specification directly against lithium battery options before accepting a quote built around older chemistry.
Storage pays when three conditions overlap: the export price falls below roughly 0.20 RMB/kWh, self-consumption without a battery sits under 40%, or the household needs outage resilience. A 10 kWh battery adds 25,000–35,000 RMB installed in 2025 pricing — roughly 40% more capex on a 20 kW system — and buying one to chase a 0.36 RMB/kWh export tariff is usually a losing trade.
Where storage wins is on avoided import. Shifting 3,500 kWh a year from a 0.60 RMB/kWh retail tier to stored solar saves 2,100 RMB annually, cutting battery payback to seven or eight years — inside warranty life, but not by much. Design the array and the inverter for storage from day one even if you buy the battery later. The engineering detail is covered in battery storage system design.

Inverter, Module and BOS Selection: Where Systems Lose Money

The inverter is the component most often underspecified in household tenders. Insist on at least two independent maximum power point tracking (MPPT) channels on a 20 kW unit — roof planes facing east and west behave as different systems, and a single-MPPT inverter throws away 4–8% of yield. A DC/AC ratio of 1.2–1.3 is healthy for Hebei and Shandong; 1.5 clips in April sunshine. Compare hybrid inverter architectures if storage is on the roadmap.
Module selection has narrowed to a quality question rather than a technology question. TOPCon modules at 22–23% efficiency with 25-year performance warranties are standard, and the difference between tier-one manufacturers is now measured in warranty service, not watts. Spend the negotiation energy on residential panel specification and warranty terms rather than on chasing an extra 5 W per module.
Balance of system (BOS) — mounting, DC cabling, combiner boxes, earthing and labour — accounts for 30–40% of installed cost, and it is where cheap quotes quietly strip value. Aluminium rail thickness, stainless fasteners and waterproofing details determine whether the roof leaks in year three and whether the array survives 25 years of wind loading. Walk through completed household installations to see how these details are specified in practice.

Five Rules Before You Sign a Grid Connection Contract

Model the export price you will actually receive. Ask the installer for the last twelve months of clearing prices in your province, not the benchmark price from 2023. If export exceeds 60% of generation, stress-test the payback at 0.15 RMB/kWh.
Maximise self-consumption before adding storage. Shift EV charging into daylight hours and run the heat pump off the array. Raising self-consumption from 35% to 55% is free; a battery costs 25,000–35,000 RMB.
Cap the DC/AC ratio at 1.3 and demand two MPPT channels. These two parameters alone separate systems that hit their modelled output from systems that miss by 10%.
Verify the commissioning date against policy deadlines. The 30 April and 31 May 2025 cut-offs determined which projects kept fixed pricing; similar provincial windows will appear again, and a four-week delay can change a contract's value.
Insist on measured baseline data. Require the installer to record generation, self-consumption and export for the first 90 days and compare against the model. A proposal without a monitoring plan is a guess.
Household PV in China remains a 6–9 year payback business for well-designed, self-consumption-heavy systems — but the margin for error has narrowed since the tariffs went to market. Send us your roof dimensions, annual consumption and province, and we will return an hour-by-hour production and revenue model built on your local clearing data. Talk to our engineering team before you sign the grid connection agreement.