Skip to main content

FIP calculation

Background​

The feed-in premium (FIP) is the main subsidy scheme in Japan. The FIP premium depends on the plant's own subsidy rate and on electricity market prices, both current ones and those of the previous financial year. The FIP often makes up a significant share of a solar plant's revenue, so it has to be calculated to estimate that revenue.

Tensor Cloud calculates the FIP for each asset from actual prices for the past and from price forward curves for the future.

FIP period​

The FIP is paid for 20 years. The period starts on the asset's COD, but no earlier than the FIP start date and no later than 3 years after it. For an asset that moves from FIT to FIP, the FIP applies from the FIP start date until 20 years after the start of the FIT period.

When a PPA names the offtaker as the FIP beneficiary, the simulation books no FIP revenue for the IPP.

FIP premium calculation logic​

For fiscal year YY (April to March) and month MM, and subsidy rate RR:

  1. Calculate PavgP_{avg}: the average of the spot price PsP_s and the intraday price PiP_i in each time slot, weighted by their volumes Vs,ViV_s,V_i. For historical premiums we use this value as published by JEPX. Simulations do not forecast intraday prices, so they use the spot price PsP_s as PavgP_{avg}. Pavg=(Ps⋅Vs+Pi⋅Vi)/(Vs+Vi)P_{avg} = (P_s \cdot V_s + P_i \cdot V_i)/(V_s + V_i)
  2. Calculate Pavg(Y−1)P_{avg}(Y-1): average of PavgP_{avg} weighted over total PV generation in the area for year Y-1
  3. Calculate Pavg(M,Y−1)P_{avg}(M,Y-1): average of PavgP_{avg} weighted over total PV generation in the area for month M in year Y-1
  4. Calculate Pavg(M,Y)P_{avg}(M,Y): average of PavgP_{avg} weighted over total PV generation in the area for month M in year Y
  5. Calculate Preference,raw=Pavg(Y−1)+(Pavg(M,Y)−Pavg(M,Y−1))P_{reference,raw} = P_{avg}(Y-1) + (P_{avg}(M,Y) - P_{avg}(M,Y-1))
  6. Take PnfveP_{nfve}, the non-fossil value price: a fixed schedule derived from the results of the quarterly auctions for non-fossil certificates (see Assumed values)
  7. Calculate PreferenceP_{reference}: if Preference,raw+Pnfve≤0P_{reference,raw} + P_{nfve} \leq 0 then Preference=0P_{reference} = 0 else Preference=Preference,rawP_{reference} = P_{reference,raw}
  8. Calculate Preference,total=Preference+Pnfve−PbalancingP_{reference,total} = P_{reference} + P_{nfve} - P_{balancing}, with the balancing cost PbalancingP_{balancing} taken from a fixed schedule by fiscal year and FIP start year (see Assumed values)
  9. Calculate adjustment factor FF: total PV generation during the month / total PV generation during slots when prices are >0.01 JPY/kWh in the month.
  10. Calculate Ppremium,raw=R−Preference,totalP_{premium,raw} = R - P_{reference,total}
  11. Calculate Ppremium=Ppremium,raw⋅FP_{premium} = P_{premium,raw} \cdot F
  12. Ppremium=max⁡(Ppremium,0)P_{premium} = \max(P_{premium}, 0)

In simulations, the area PV generation used as weight is reduced by the zonal curtailment of the scenario.

note

The FIP premium is constant throughout each month, except that it is zero in slots where prices are at or below 0.01 JPY/kWh (price floor). If the asset has a FIP rate schedule, the rate RR changes on the date of each scheduled change, so the premium can also change within a month.

Assumed values​

Two inputs to the reference price are not derived from market data at calculation time, but taken from fixed schedules.

Non-fossil value price​

PnfveP_{nfve} follows a step schedule in JPY/kWh. Each value applies from the month shown until the next change, and the earliest value also applies to any month before it.

PeriodPnfveP_{nfve} (JPY/kWh)
up to 2023-010.6
2023-02 – 2023-040.775
2023-05 – 2024-010.95
2024-02 – 2024-040.775
2024-05 onwards0.6

Balancing cost​

The Agency for Natural Resources and Energy (ANRE) includes a balancing cost in the premium for FIP solar and wind plants. In the calculation it is subtracted from the reference price (step 8), which raises the premium.

PbalancingP_{balancing} is set per Japanese fiscal year (April to March) in JPY/kWh, as the sum of two parts:

  • Base amount: depends on the fiscal year in which the plant started operating under FIP (for a move from FIT to FIP, the year of the move). Plants that started in FY2023 or earlier, or in FY2027 or later, are on the regular schedule. Plants that started in FY2024, FY2025 or FY2026 get a start-year amount instead: 1.0 JPY/kWh in their first FIP year, stepping down each year to join the regular schedule in the fourth. Tensor Cloud takes the start year from the first day of the asset's FIP period: the FIP start date for a move from FIT to FIP, otherwise the COD, but no earlier than the FIP start date and no later than 3 years after it. For example, an asset with a FIP start date in March 2025 (FY2024) and a COD in June 2025 gets the FY2025 amounts.
  • Top-up α: an extra amount for every FIP solar and wind plant, whatever year it started FIP. ANRE's Procurement Price Calculation Committee sets each fiscal year's α in the January before it. α becomes zero from the fiscal year after FIP's share of FIT and FIP reaches 25%.

Values are in JPY/kWh. PbalancingP_{balancing} is the base amount in the column for the plant's FIP start year, plus α.

Fiscal yearRegular scheduleFIP start FY2024FIP start FY2025FIP start FY2026Top-up α
before FY202200
FY2022 – FY20231.00
FY20241.01.00
FY20250.80.861.01.0
FY20260.70.730.831.01.0
FY20270.60.60.660.80
FY20280.50.50.50.60
FY20290.40.40.40.40
FY20300.30.30.30.30
FY20310.20.20.20.20
FY20320.10.10.10.10
FY2033 onwards00000

For example, a plant that moved from FIT to FIP in FY2026 has PbalancingP_{balancing} = 1.0 + 1.0 = 2.0 JPY/kWh in FY2026, 0.8 JPY/kWh in FY2027 and 0.6 JPY/kWh in FY2028, then follows the regular schedule.

  • FY2022 to FY2024: the regular schedule is 1.0 JPY/kWh, which matches the actual reference prices for those years.
  • Regular schedule from FY2031: ANRE has published the regular schedule up to FY2030; after that it moves to a final level, the imbalance risk unit price, which is not published yet. Tensor Cloud assumes it keeps falling by 0.1 JPY/kWh a year until it reaches zero in FY2033.
  • α from FY2027: ANRE has not set it yet; the committee is expected to decide FY2027's α in January 2027. Until then Tensor Cloud uses 0, so that results do not count on a top-up that has not been decided. We update the value after each January decision.
  • Realized revenue and the FIP premium calculator: simulations and battery optimization take each asset's start year as described above, so they apply the start-year amounts. The FIP premium in realized revenue and in the FIP premium calculator is calculated per area and FIP rate rather than per asset, so it uses the regular schedule plus α for every plant. For a plant that started FIP in FY2024, FY2025 or FY2026, it is up to 0.3 JPY/kWh below the plant's actual premium: for a FY2026 start, 0.3 JPY/kWh in FY2026, 0.2 JPY/kWh in FY2027 and 0.1 JPY/kWh in FY2028.

Sources: ANRE's materials for the Procurement Price Calculation Committee, 101st meeting, material 2 (January 2025: regular schedule, start-year amounts and FY2025 α on p.13 and p.16) and 113th meeting, material 2 (January 2026: FY2026 α on p.17).

FIP compensation calculation logic​

Once the monthly FIP premium is known, the FIP compensation is the premium multiplied by the energy eligible for FIP.

The formula below covers the general case of an asset with a battery and grid charging enabled.

This logic is based on the official METI documentation. For each time slot tt within month MM:

RFIP(t)={Ppremium,M(Esolar-to-grid(t)+λMEbattery-to-grid(t))Pday-ahead(t)>0.010Pday-ahead(t)≤0.01R_{\text{FIP}} (t) = \begin{cases} P_{\text{premium},M} \Big( E_\text{solar-to-grid}(t) + \lambda_M E_\text{battery-to-grid}(t) \Big) & P_{\text{day-ahead}}(t) > 0.01\\ 0 & P_{\text{day-ahead}}(t) \leq 0.01 \end{cases}

With λM\lambda_M the fraction of battery energy coming from solar during the month:

λM=∑t∈MEsolar-to-battery(t)∑t∈M(Esolar-to-battery(t)+Egrid-to-battery(t))\lambda_M = \frac{\sum_{t \in M}{E_\text{solar-to-battery}(t) } }{ \sum_{t \in M}({E_\text{solar-to-battery}(t) + E_\text{grid-to-battery}(t)})}

For DC-coupled batteries, λM\lambda_M is calculated from the energy flows on the DC side of the site inverter. In a month without any battery charging, λM\lambda_M is zero.

note

For assets without a battery, Ebattery-to-gridE_\text{battery-to-grid} is zero, so only solar contributes. For assets with a battery but without grid charging, Egrid-to-batteryE_\text{grid-to-battery} is zero, so λM=1\lambda_M=1 and all battery discharge contributes to FIP revenue.

The battery is optimized against the day-ahead price plus the FIP premium, so it values discharge in slots where the premium is paid.

Further reading​