Full feed-in from 2027: What changes for solar surplus users?

Anyone running or planning a PV system faces a new calculus for surplus solar power once Germany's EEG 2027 enters into force. The government draft, approved by the federal cabinet on 29 July 2026, abolishes the fixed feed-in tariff for new installations, drops the full-feed-in bonus, and replaces the familiar 60% cap with a permanent limitation of feed-in power to 50% of installed capacity. For operators who currently sell their surplus — or who are weighing alternatives such as mining — the comparison numbers shift considerably.

One caveat up front: as of October 2026 the EEG 2027 has not been passed. The Bundestag held its first reading of the draft on 24 September 2026, and the economics committee's hearing took place on 5 October 2026. The Bundesrat has, among other things, proposed deleting the 50% cap. Everything below refers to the current draft and may still change. Existing installations keep grandfathered protection.

What the EEG 2027 draft actually proposes

The core of the reform is the end of the 20-year guaranteed feed-in tariff for new installations from 1 January 2027. In its place comes a tiered system:

Transition payment: Installations commissioned in 2027 with less than 50 kW of installed capacity receive a payment of a nominal 5.2 ct/kWh for up to 36 months (1 ct/kWh is deducted from the uniform reference value of 6.2 ct/kWh). Access then narrows step by step: in 2028 only for systems under 25 kW, from 2029 only for systems under 7 kW.
Direct marketing as the default: After the transition phase, electricity is to be sold on the exchange, with a temporarily tiered direct-marketing premium.
End of the full-feed-in bonus: The existing surcharge for systems that feed in everything is dropped. A single, size-independent rate applies.

In practice: registering by 31 December 2026 locks in the current rates for 20 years — since 1 August 2026 that is 7.70 ct/kWh for partial feed-in and 12.22 ct/kWh for full feed-in on systems up to 10 kWp. Anyone installing from 2027 gets a guaranteed 5.2 ct/kWh for at most three years, and exchange prices after that.

From 60% to 50%: the new cap on feed-in power

Since the "Solarspitzengesetz" (in force since 25 February 2025), PV systems between 2 and 100 kWp without a smart meter may feed in at most 60% of their installed capacity. With a smart metering system and control unit, that limit falls away — full feed-in is then possible, but is unpaid during hours with negative exchange prices.

The EEG 2027 draft reverses the logic: new second-segment systems under 100 kW are to have their active power feed-in at the grid connection point permanently capped at 50% of installed capacity — regardless of smart-meter status. Full feed-in in today's sense would effectively no longer be possible; the non-feedable remainder must be self-consumed, stored, or the inverter curtailed. At the same time, the draft lowers barriers for so-called zero-feed-in systems that keep their surplus entirely — storing or curtailing instead of feeding in.

For the surplus comparison this is the decisive point: the cap applies to feeding into the grid. It does not apply to self-consumption. A miner running in your house is legally self-consumption — the 50% limit does not touch it directly, but it does change the alternative ("feed in") that mining competes against.

Negative prices: why feed-in revenue is under pressure

Independent of the cap, the value of fed-in electricity erodes around midday. The number of hours with negative day-ahead prices in Germany has climbed for years: 301 hours (2023), 457 (2024), 573 (2025) — about 6.5% of all annual hours. The first half of 2026 already counted 299 such hours; full-year forecasts range from 700 to 900. On 1 May 2026 the day-ahead price fell to −499.99 €/MWh, after which EPEX Spot lowered the price floor to −600 €/MWh. The average negative price in 2026 so far is around −28.65 €/MWh.

For new installations commissioned since 25 February 2025 the rule is: if the spot price drops below zero in a quarter-hour, the payment is cut to zero from the very first quarter-hour (§ 51 EEG). The missed periods are appended to the funding period under § 51a EEG — but for solar systems only at half weight, and with up to 20 years of delay. Expert opinions on the EEG reform assume that 15–20% of annual hours, in some cases up to 30%, will carry negative exchange prices over the next five years.

Applied to a full-feed-in system: if 600 hours of negative prices occur in a year, roughly 9% of the annual feed-in revenue is initially lost, depending on the model. That is exactly the gap eating into the apparently safe feed-in business — and it hits full-feed-in operators disproportionately, because their entire revenue depends on the tariff.

What this means for the feed-in vs. mining comparison

The shift is visible in the numbers:

Until 31.12.2026 (register in time): 12.22 ct/kWh full feed-in or 7.70 ct/kWh partial feed-in, guaranteed for 20 years — a plannable benchmark every alternative must beat.
From 2027 (draft): 5.2 ct/kWh for at most 36 months, then direct marketing at exchange prices — with the risk of negative-price hours where revenue is zero. The market value of midday solar power sits near zero in many hours.

On the mining side, SolarMiner's live data (retrieved 8 October 2026) shows: Bitcoin price €73,675, network difficulty 1.327 × 10¹⁴, current block reward 3.14 BTC. An Antminer S19k Pro runs at roughly 23.9 J/TH. That yields a mining revenue per kWh directly comparable to whichever feed-in tariff applies — plug your own system values into the revenue calculator. The underlying comparison is laid out in detail in solar surplus and mining 2027.

The structural difference: feeding in delivers a guaranteed or exchange-dependent payout per kWh with no extra hardware. Mining converts the same kWh into revenue that depends on the Bitcoin price, difficulty, and hardware efficiency — but the miner counts as self-consumption and is untouched by the 50% cap. As the feed-in tariff falls to 5.2 ct/kWh and to zero in negative-price hours, the bar that mining has to clear comes down.

Where mining is still not the better choice in 2027

The honest side of the ledger:

Registering in 2026 is a strong deal. 12.22 ct/kWh for full feed-in, guaranteed for 20 years, is a bond-like yield with no hardware risk, no difficulty risk, and no price risk. Mining has to beat that — and often loses once purchase price, maintenance, and residual value are included.
Existing installations are unaffected. If you already feed in under a tariff, nothing changes for you. The feed-in vs. mining question only arises for additional surplus.
Noise has a price. An ASIC drawing around 3 kW continuously is unbearable in living space without serious acoustic decoupling. Quiet installation eats budget and space.
A battery is often the sturdier answer. If the goal is backup power, higher self-consumption, or protection from negative-price hours, storage delivers those directly — mining does not.
Mining carries market risk. The Bitcoin price can fall and difficulty rises structurally. Comparing a 20-year guaranteed tariff to a revenue-variable business means comparing different risk classes.

In short: the EEG 2027 makes feeding in less attractive for new systems and improves the relative position of surplus mining — but mining is no guarantee because of it, and for grandfathered systems, feeding in usually remains the sounder base.

Sources

Not investment or tax advice. Legal status: 8 October 2026; the EEG 2027 is still in the legislative process.