The old math is dead. Since 2021, newly installed household rooftop systems in China no longer receive the national feed-in tariff (FIT) of RMB 0.37–0.45/kWh that made distributed solar a no-brainer. What remains is surplus-export pricing tied to local coal benchmark rates of roughly RMB 0.23–0.45/kWh, self-consumption savings against retail tariffs, and, increasingly, the revenue shift from daytime export to evening peak shaving via storage. A 20 kW rooftop array can still generate annual returns of 12–18%, but only if the owner sizes the system against actual consumption profiles — not against last decade's subsidy tables.

The 2021 Policy Break: Why Household FITs Collapsed

The National Development and Reform Commission's Document No. 833 (2021) removed the centralized subsidy benchmark for new distributed generation, pushing all new household projects into the "spontaneous self-use, surplus grid-export" model. Before that rule, owners in provinces like Shandong and Hebei locked in feed-in rates above RMB 0.40/kWh for 20 years. After 2022, that guarantee vanished. By 2024, the average surplus-export price paid by provincial grids was roughly RMB 0.35/kWh — and in coal-heavy provinces it fell toward RMB 0.20/kWh at certain midday hours.
This is not a failure of policy but a maturation of economics. BloombergNEF reported that global PV module prices fell more than 40% during 2023, pushing turnkey rooftop system costs in China to around RMB 3.0–3.8/W by early 2024. A 20 kW installation today costs RMB 60,000–76,000 before regional incentives. Lower capital cost partially offsets lost subsidies, yet the decisive factor is no longer the export tariff. It is how much of the generated energy replaces grid power that the household would otherwise buy at RMB 0.50–1.10/kWh.

Anatomy of a Return: Export Revenue Versus Avoided Purchase

For a typical residential customer, revenue stacks in three layers. Layer one is self-consumption, effectively worth the full retail tariff plus grid fees the homeowner avoids. Layer two is surplus export, valued by the local desulfurization coal benchmark price. Layer three — still rare — is provincial or municipal incentive payments, where they exist, such as Wenzhou's RMB 0.10/kWh municipal subsidy for distributed systems between 2022 and 2025.
Consider a 20 kW system in Hebei with 24,000 kWh annual yield (a capacity factor of roughly 13.7%). If the household self-consumes 40%, it avoids buying 9,600 kWh at an average retail tariff around RMB 0.52/kWh, creating about RMB 4,990 of savings. The remaining 14,400 kWh exported at, say, RMB 0.36/kWh adds RMB 5,180. Total annual cash benefit arrives at roughly RMB 10,200, against an RMB 66,000 capital cost — a static payback near 6.5 years, without considering degradation of roughly 0.5% per year. The return, however, is exposed to one variable owners frequently ignore: the falling midday export price.

Case Study: a 20 kW Rooftop System in Shijiazhuang, Hebei

A system installed in March 2024 by a suburban homeowner in Shijiazhuang illustrates the modern profile. The hardware included 44 × 455 W monocrystalline PERC (passivated emitter rear contact) modules and a 20 kW three-phase string inverter, at RMB 3.4/W, totaling RMB 68,000. The family's 12-month generation was 23,800 kWh. With daytime load modest — a heat pump water heater, washing machine and air conditioning shifted to mid-day — self-consumption reached 47%.
Measured export over 12 months totaled 12,600 kWh at the provincial benchmark of RMB 0.3645/kWh, worth RMB 4,593. Avoided purchases of 11,200 kWh at the Hebei residential tier-1 tariff of roughly RMB 0.52/kWh saved RMB 5,824. Combined first-year income was RMB 10,417, and with the inverter's monitoring data, the owner confirmed that 76% of all generation occurred within the 08:00–17:00 window. The project's seven-year internal rate of return (IRR) sits near 14% if export prices remain flat — and closer to 9% if the provincial grid introduces midday depression of export compensation.

When Export Prices Drop: The Case for Adding a Battery

The largest unrecognized risk in household PV economics is the "duck curve" arriving in distribution networks. Provinces such as Shandong have already discussed time-differentiated surplus-export prices, where midday solar exports earn less than morning or evening exports. A solar-only system that exports heavily between 10:00 and 15:00 may see its revenue per exported kilowatt-hour fall by 20–40% within three years. Battery storage neutralizes that cliff.
Adding a 10 kWh lithium iron phosphate (LFP) battery with a hybrid inverter raises project cost by roughly RMB 15,000–20,000 at 2025 prices, which BloombergNEF data suggests have declined about 50% since 2021. The battery shifts 6–7 kWh per day from midday export — worth RMB 0.36/kWh — to evening self-consumption displacing peak retail power at RMB 0.85–1.10/kWh. Daily arbitrage adds RMB 3.0–5.2 regardless of feed-in policy. At 300 cycles per year for 12 years, the battery's effective storage cost is about RMB 0.45–0.55/kWh of shifted energy, which beats the projected midday export price but remains higher than today's flat export benchmark. A conservative conclusion: batteries pay when the spread between peak retail and midday export exceeds RMB 0.50/kWh — true in most provinces by 2026.

How to Choose an Inverter Topology and Wiring Scheme

Inverter selection determines whether the battery can arbitrage at all. A pure grid-tied string inverter cannot charge a battery in an outage or route surplus intelligently. A hybrid inverter with at least two MPPT (maximum power point tracking) inputs splits the array so that east and west roof faces start generating earlier and end later, flattening the midday peak and increasing the share of generation that overlaps with household load — often by 4–6 percentage points compared with a single south-facing string.
Equally important is the DC/AC ratio. Designers who size an inverter at 100% of module capacity are leaving money idle. A ratio of 1.2–1.3 — for example 20 kWp of modules on a 15.4 kW inverter — clips only about 1% of annual yield on a clear-sky day but extends usable generation into low-light hours and cuts equipment cost. On a typical Hebei site, that ratio trade yields RMB 700–1,200 in inverter savings without measurable annual yield loss. Choose a hybrid unit with 10 ms switchover and islanding protection certified to GB/T 34120, and verify that the manufacturer publishes degradation curves beyond the 10-year warranty.

Sizing Against Degradation, Weather and Module Quality

Component quality shows up in the 15th year, not the first. High-tier modules with a 30-year linear power warranty degrade roughly 0.4% per year, versus 0.7% for lower-tier cells. Over 25 years, that difference compounds to a 7–8% spread in cumulative generation — on the Hebei case, that is more than RMB 13,000 in lost export and avoided-cost value. Buyers should demand datasheets listing the temperature coefficient of power (ideally below −0.34%/°C) and the first-year degradation figure of no more than 2%, verified by the China Photovoltaic Industry Association's (CPIA) type test results.
Weather is equally consequential. Hebei receives 1,300–1,600 kWh/m² of annual horizontal irradiation, but a site facing 20° east of south captures roughly 96% of optimum, while a west-facing roof produces about 5% more evening output — aligning with peak retail tariffs. The homeowner who optimizes azimuth over pure south orientation can raise effective revenue per installed watt by up to 8%. Shade from a chimney or tree branch crossing a single cell can reduce the whole string's output by 15–30% unless the inverter uses module-level rapid shutdown or distributed MPPT.

Regional Incentives and Municipal Rebates Still on the Table

Although central subsidies ended, roughly a dozen provinces and cities continue to sweeten household projects. Zhejiang's Wencheng county, for instance, introduced per-watt deployment subsidies of RMB 0.1–0.2/W for completed residential systems in select periods, and certain districts in Shandong have experimented with energy-storage capacity compensation for batteries that discharge during evening peaks. These are one-off capital grants that shorten payback by 0.5–1.5 years, but they come with administrative compliance costs. The CPIA reminded buyers in 2024 that incentive eligibility often requires equipment from an approved catalog — a constraint that should be checked before signing any equipment contract, not after installation.
The stronger long-term signal is the expansion of household participation in demand response. Provinces that pilot residential virtual power plant (VPP) aggregation allow battery-equipped homes to receive RMB 0.5–1.0/kWh for discharging during declared grid emergencies. In Jiangsu's 2024 pilot season, aggregated household batteries that discharged for 2 hours across 12 dispatched events earned participants roughly RMB 1,600 per 10 kWh system. This revenue is not guaranteed by the distribution system operator each year, but it transforms storage from a cost center into an optionality play.

The Exit Scenario: Selling Surplus in a Zero-Export Province

Some distribution networks in coastal Shandong and parts of Henan have at times restricted further household exports, citing transformer capacity limits at midday. Where export is capped to near zero, the project equation flips completely: all revenue must come from displaced retail power. In such cases, a system should be sized to the household's daytime demand plus 20–30% margin for heat pumps or electric vehicles, rather than to roof area maximization. At Hebei retail tariffs, a 10 kW array with 80% self-consumption yields roughly RMB 6,000/year on an RMB 34,000 investment — an 18% first-year return that is remarkably because it depends on no export revenue at all.
The operating philosophy shifts from "maximize kilowatt-hours" to "maximize avoided kilowatt-hours." Homeowners considering a heat pump or an EV before adding PV should size the solar array first against those future loads. Installing after the load exists defeats the purpose: you will undersize on the south roof and end up exporting at low value while buying peak power at night. Grid-tied inverters that support export limiting of 0 W, per grid instructions, are now a standard requirement in these provinces, and confirming that capability before purchase prevents a costly retrofit.

Final Checklist Before You Sign the 2025 Contract

First, simulate hourly consumption for a full year using smart-meter data; self-consumption share above 40% is the precondition for a strong solar-only investment. Second, request three quotes priced per watt and verify each includes BOS (balance of system) hardware, grid-connection application fees and commissioning — hidden logistics charges of RMB 1,000–3,000 routinely inflate otherwise identical bids. Third, specify a hybrid inverter with backup capability even if you defer the battery, because retrofitting a grid-tied string inverter later costs more than upgrading at initial installation.
Fourth, confirm the local surplus-export compensation method with the county power company before signing, and do not rely on installer brochures that quote provincial averages. Fifth, choose LFP batteries with a cycle life above 6,000 at 0.5 C discharge and a manufacturer that publishes thermal runaway test results to GB/T 38031; domestic assembly with automotive-grade cells is acceptable where the integrator holds a valid quality certification.
Every household's numbers differ, but the framework does not: at system costs below RMB 3.5/W, a 6–8 year payback remains realistic across northern China. The projects that fail in 2025 are not the ones with imperfect roofs. They are the ones that bought yesterday's subsidy logic and sold at midday, into a market that no longer rewards it.