Analysis · Netherlands
Grid bottlenecks and declining feed-in rates are encouraging customers to switch to solar+storage systems in the Netherlands
Grid bottlenecks and declining feed-in rates are encouraging customers to switch to solar+storage systems in the Netherlands
The Netherlands ranks among the top-three countries in the world for the amount of solar PV installed per capita (1). And yet, as the electricity system becomes flooded with solar during the daytime, it is becoming increasingly challenging to get a grid connection for new solar projects. At the same time, the feed-in payments on offer have declined drastically, making behind-the-meter configurations coupled with battery storage increasingly popular. This short brief looks at the case of a Dutch cold-storage distribution centre trying to make the right decision amidst a complicated regulatory and pricing environment.
The Netherlands ranks among the top-three countries in the world for the amount of solar PV installed per capita (1). And yet, as the electricity system becomes flooded with solar during the daytime, it is becoming increasingly challenging to get a grid connection for new solar projects. At the same time, the feed-in payments on offer have declined drastically, making behind-the-meter configurations coupled with battery storage increasingly popular. This short brief looks at the case of a Dutch cold-storage distribution centre trying to make the right decision amidst a complicated regulatory and pricing environment.
3.5 MWp
3.5 MWp
Solar size where value stops rising without export
€0.009/kWh
€0.009/kWh
What exported solar would have earned in 2025
582
582
Number of hours with negative wholesale prices in 2025
The case models a 25,000 m² food distribution centre in Zeewolde, just outside of Amsterdam, with an annual refrigeration-related demand of 7.9 GWh (2, 3) and room for 4.5 MWp solar array on the roof. Although it technically has a grid connection, feed-in surplus solar power has recently been refused by the local operator. TenneT declared feed-in congestion here in 2021 (4), and the timeline for a new high-voltage substation being developed in Utrecht North has slipped from 2029 to 2033, and further setbacks cannot be ruled out (5). Until then, large solar producers and large consumers are forced to wait, or to seek out alternatives for soaking up their surplus daytime power.
The case models a 25,000 m² food distribution centre in Zeewolde, just outside of Amsterdam, with an annual refrigeration-related demand of 7.9 GWh (2, 3) and room for 4.5 MWp solar array on the roof. Although it technically has a grid connection, feed-in surplus solar power has recently been refused by the local operator. TenneT declared feed-in congestion here in 2021 (4), and the timeline for a new high-voltage substation being developed in Utrecht North has slipped from 2029 to 2033, and further setbacks cannot be ruled out (5). Until then, large solar producers and large consumers are forced to wait, or to seek out alternatives for soaking up their surplus daytime power.
The value of exported solar has declined steadily
The value of exported solar has declined steadily
The value of exported solar has declined steadily
Solar used on site cuts the electricity bill by €0.161/kWh based on the savings from electricity, reduced network charges and a reduction to the amount of energy tax owing (6, 7, 8). Exported solar sells at the day-ahead price in the hours most other roofs are exporting too: in fact, fully 94% of 2025’s 582 negative-price hours fell during daytime hours (9). In 2025, a 3 MWp solar array’s exports would have fetched €0.009/kWh, less than 1 cent/kWh.
In this scenario, large power users are increasingly being driven to find ways to increase their self-consumption ratio (the amount of self-generated electricity they can productively use themselves onsite).
Solar used on site cuts the electricity bill by €0.161/kWh based on the savings from electricity, reduced network charges and a reduction to the amount of energy tax owing (6, 7, 8). Exported solar sells at the day-ahead price in the hours most other roofs are exporting too: in fact, fully 94% of 2025’s 582 negative-price hours fell during daytime hours (9). In 2025, a 3 MWp solar array’s exports would have fetched €0.009/kWh, less than 1 cent/kWh.
In this scenario, large power users are increasingly being driven to find ways to increase their self-consumption ratio (the amount of self-generated electricity they can productively use themselves onsite).
Figure 1. The full 4.5 MWp roof on 12 July 2025, no export, with and without a 1 MW/2 MWh battery.
Finding the optimal system size can be challenging
Finding the optimal system size can be challenging
Figure 2 shows the point beyond which increasing the PV system size starts to worsen the project's NPV. Without the right to export surplus to the grid, the array’s value reaches a plateau at about €3.4m: beyond that point, extra panels mostly generate electricity that the owner cannot use, with the result that a growing share of the output gets curtailed.
Figure 2 shows the point beyond which increasing the PV system size starts to worsen the project's NPV. Without the right to export surplus to the grid, the array’s value reaches a plateau at about €3.4m: beyond that point, extra panels mostly generate electricity that the owner cannot use, with the result that a growing share of the output gets curtailed.
Figure 2. Net present value over 25 years by array size, with and without export.
All-equity, discounted at 5.2% (11). Electricity prices rising 1% a year above inflation.
In order to extract the most value from the investment, the owner is faced with two options: (A) develop a solar-only system, under-sized to avoid excessive curtailment, or (B) develop a solar+storage system that makes use of the full roof and feed that extra electricity into an onsite battery. Naturally, adding a battery to the system also adds to the total cost (see Table 1).
In order to extract the most value from the investment, the owner is faced with two options: (A) develop a solar-only system, under-sized to avoid excessive curtailment, or (B) develop a solar+storage system that makes use of the full roof and feed that extra electricity into an onsite battery. Naturally, adding a battery to the system also adds to the total cost (see Table 1).
Table 1. Weighing up the two options from a facility-management perspective
Option A: 3.5 MWp solar only
Option A: 3.5 MWp solar only
Option B: full roof (4.5 MWp) + 1 MW/2 MWh battery
Option B: full roof (4.5 MWp) + 1 MW/2 MWh battery
At a glance
Costs €1.8m upfront
Pays back in 6.5 years
NPV of roughly €3.4m over 25 years
At a glance
Costs €1.8m upfront
Pays back in 6.5 years
NPV of roughly €3.4m over 25 years
At a glance
Costs €2.9m upfront
Pays back in 8.2 years
NPV of roughly €3.4m over 25 years
At a glance
Costs €2.9m upfront
Pays back in 8.2 years
NPV of roughly €3.4m over 25 years
Pros
Low complexity: panels need little beyond periodic inspection and cleaning.
Work stays on the roof: no battery plot or extra fire-safety measures needed at ground level.
The smaller budget is easier to fund from the company's own cash, or from retained earnings.
Pros
Low complexity: panels need little beyond periodic inspection and cleaning.
Work stays on the roof: no battery plot or extra fire-safety measures needed at ground level.
The smaller budget is easier to fund from the company's own cash, or from retained earnings.
Pros
The whole roof is used.
The site buys less of its power from the grid, which means the customer's electricity bill is less exposed to rising prices.
In the event of a grid outage, the BESS system takes over, providing continuous power supply without needing fuel, and without increasing the site's carbon footprint.
Pros
The whole roof is used.
The site buys less of its power from the grid, which means the customer's electricity bill is less exposed to rising prices.
In the event of a grid outage, the BESS system takes over, providing continuous power supply without needing fuel, and without increasing the site's carbon footprint.
Cons
About a fifth of the roof stays empty.
A greater share of the site’s power still comes from the grid, so the bill keeps tracking grid prices.
The site has no clear back-up if the grid fails, and if the outage proves longer, the site would need to rely on a back-up generator to keep the cold storage operating.
Cons
About a fifth of the roof stays empty.
A greater share of the site’s power still comes from the grid, so the bill keeps tracking grid prices.
The site has no clear back-up if the grid fails, and if the outage proves longer, the site would need to rely on a back-up generator to keep the cold storage operating.
Cons
More upfront capital and a longer payback.
More complexity: needs space on site, fire-safety approval, a permit, battery cell replacement after year 7-10, and insurance.
A second system to monitor and maintain, with alarms and software updates.
Cons
More upfront capital and a longer payback.
More complexity: needs space on site, fire-safety approval, a permit, battery cell replacement after year 7-10, and insurance.
A second system to monitor and maintain, with alarms and software updates.
Figures from the model, with electricity prices rising 1% a year above inflation. Solar and battery costs (12, 13). Outage backup was not modelled.
This is a complex decision
This is a complex decision
This is a complex decision
The choice for a facility like this is rarely ever black and white. Option A asks almost nothing new of the facility team: panels on the roof, regular monitoring and a yearly inspection. Option B adds an asset with additional safety requirements and a higher price tag. Someone has to own the battery’s alarms, service contract and software updates, agree fire-safety measures, and pay higher insurance premiums. In addition, the owner has to ensure that enough capital is available to pay for the eventual battery cell replacement and end-of-life disposal.
The second question relates to the company's time horizon. Option A is the safer bet if today’s conditions hold. Option B pulls ahead only if things go well: grid prices rising faster than assumed, battery prices falling, or time-dependent network tariffs arriving in 2028 and rewarding a battery that trims evening peaks (14). None of those elements is certain, and a battery cannot be un-purchased. A middle path is to build option A now, keep space and a cable route free for storage, and revisit the battery when prices or tariffs move.
The choice for a facility like this is rarely ever black and white. Option A asks almost nothing new of the facility team: panels on the roof, regular monitoring and a yearly inspection. Option B adds an asset with additional safety requirements and a higher price tag. Someone has to own the battery’s alarms, service contract and software updates, agree fire-safety measures, and pay higher insurance premiums. In addition, the owner has to ensure that enough capital is available to pay for the eventual battery cell replacement and end-of-life disposal.
The second question relates to the company's time horizon. Option A is the safer bet if today’s conditions hold. Option B pulls ahead only if things go well: grid prices rising faster than assumed, battery prices falling, or time-dependent network tariffs arriving in 2028 and rewarding a battery that trims evening peaks (14). None of those elements is certain, and a battery cannot be un-purchased. A middle path is to build option A now, keep space and a cable route free for storage, and revisit the battery when prices or tariffs move.
Conclusion
Conclusion
Conclusion
Over 25 years both paths end up with a nearly identical NPV. Option A keeps the facility simple and the capital requirement smaller; option B buys more independence from the grid at the price of a new system to run (and insure). But with a longer payback time, it is difficult for a cold storage operator in a market like the Netherlands with high grid reliability to justify the higher Capex: ultimately, the thermal storage it already has likely provides enough security, and avoids needing to spend over €1 million more in initial Capex. Arriving at this decision may not be what the installer or EPC company wants to hear - but in some cases, a more prudent use of capital is warranted.
Over 25 years both paths end up with a nearly identical NPV. Option A keeps the facility simple and the capital requirement smaller; option B buys more independence from the grid at the price of a new system to run (and insure). But with a longer payback time, it is difficult for a cold storage operator in a market like the Netherlands with high grid reliability to justify the higher Capex: ultimately, the thermal storage it already has likely provides enough security, and avoids needing to spend over €1 million more in initial Capex. Arriving at this decision may not be what the installer or EPC company wants to hear - but in some cases, a more prudent use of capital is warranted.
Assumptions and sources, 12 rows
What the model leaves out: the battery charges from the roof only; grid charging, balancing services and trading are excluded, and would change its value. Battery cost is priced by energy size for every duration; longer systems carry less power electronics per kWh and may cost somewhat less. Network charges are fixed per kWh until time-dependent tariffs arrive, planned for 202814. No SDE++ subsidy is assumed.
| Input | Value | Source |
|---|---|---|
| Site | Horsterparc, Zeewolde, Liander network | 4 |
| Load | 7.9 GWh/yr, peak 1,345 kW; 120,000 m³ refrigerated at 65.9 kWh/m³ | 2,3 |
| Array | East-west at 10°, 929 kWh/kWp in 2025 after 14% system loss; 2025 irradiance scaled to PVGIS (×0.899) | 10,15 |
| Value of solar used on site | €0.161/kWh: energy €0.101, network €0.0226, energy tax €0.03735 | 6,7,8 |
| Value of exported solar | 2025 day-ahead price less €7/MWh; nothing below zero | 9,11 |
| Monthly peak charge | €3.57/kW/month; the contracted-kW and standing charges are fixed and left out | 7 |
| Solar capex and O&M | €509/kWp; €11.7/kWp/yr fixed, €0.0089/kWh variable | 12 |
| Battery | 0.5–8 MWh at 2, 4, 6 hours, charged from the roof only; installed cost by energy size from a vendor guide; O&M 1.5% of capex/yr; degradation 2%/yr, replaced at 80% for 30% of capex, not in the final year | 13,16 |
| Tax incentive | EIA 40% on the battery only | 17 |
| Electricity price rise | 1% a year above inflation, a judgement call given continued reliance on gas, a planned nuclear build-out and years of grid congestion. Applied to the savings only; O&M and replacements rise with inflation | |
| Discount rate, tax, inflation | 5.2% nominal after tax; 25.8%; 2% | 11 |
| Horizon | 25 years, all-equity, no SDE++ subsidy |
Sources
Sources
Sources
1. Solar PV per capita. IEA PVPS, Snapshot of Global PV Markets 2026 (Report T1-49:2026, April 2026). PV per capita at the end of 2025: Australia 1,629 W, the Netherlands 1,601 W, Germany 1,404 W. https://iea-pvps.org/snapshot-reports/snapshot-2026/
2. Load profile. ELMAS dataset (Mines Paris for Enedis; Scientific Data, 2023), CC BY 4.0. Hourly smart-meter data from French business customers, 2018. Cluster 3: supermarkets, food wholesale and warehousing, including 719 medium-voltage warehousing sites. https://figshare.com/articles/dataset/ELMAS_dataset/23889780
3. Cold-store electricity use. Evans, J.A. et al., Specific energy consumption values for various refrigerated food cold stores, Energy and Buildings 74 (2014) 141-151. European mean of 65.9 kWh per refrigerated m3 per year for mixed chilled and frozen stores. https://www.sciencedirect.com/science/article/abs/pii/S0378778813007925
4. Feed-in congestion at the site. Gemeente Zeewolde, Beleidsvisie netcongestie Zeewolde (October 2025). Records TenneT’s declaration of feed-in congestion for Utrecht, Flevoland and Gelderland on 12 October 2021, now expected to last until 2033-2035. https://www.duurzaamzeewolde.nl/fileadmin/nieuwbouw-zeewolde/documenten/duurzaam/Beleidsvisie_netcongestie_Zeewolde_Def.pdf
5. Grid expansion delay. TenneT, Major expansion projects for the high-voltage grid in Utrecht, Flevopolder and Gelderland are taking longer than planned (12 June 2025). Completion moves from 2029 to 2033, or 2035 with further setbacks; splitting the regional grid into four sub-networks depends on a new high-voltage substation in Utrecht North. Large consumers and larger wind and solar producers wait longer for new or heavier connections. https://www.tennet.eu/nl-en/news/major-expansion-projects-high-voltage-grid-utrecht-flevopolder-and-gelderland-are-taking-longer-planned
6. Commodity price. Eurostat, nrg_pc_205_c, electricity prices for non-household consumers, Netherlands, band ID (2-20 GWh), energy and supply component, 2025. https://ec.europa.eu/eurostat/databrowser/view/nrg_pc_205_c/default/table
7. Network tariff. Liander, Tarieven voor aansluiting en transport elektriciteit 2026, MS category >136-2,000 kW: EUR 0.0226/kWh and EUR 3.57/kW/month on the monthly peak, excluding VAT. https://www.liander.nl/-/media/files/tarieven/grootzakelijk/tarieven-2026/tarieven-voor-aansluiting-en-transport-elektriciteit-2026_2.pdf
8. Energy tax. Belastingdienst, tarieven energiebelasting 2026. Electricity, 50,001 kWh to 10 million kWh: EUR 0.03735/kWh excluding VAT. https://www.belastingdienst.nl/wps/wcm/connect/bldcontentnl/belastingdienst/zakelijk/overige_belastingen/belastingen_op_milieugrondslag/energiebelasting/energiebelasting
9. Day-ahead prices. Energy-Charts (Fraunhofer ISE), day-ahead auction prices, bidding zone NL, 2025, from Bundesnetzagentur/SMARD, CC BY 4.0. Hourly to 30 September, 15-minute from 1 October; both averaged to half-hours. https://api.energy-charts.info/
10. Irradiance and temperature. Open-Meteo Historical Weather API (ERA5-based), hourly global tilted irradiance and 2 m air temperature, Horsterparc Zeewolde, 2025, east-west array at 10 degrees. https://open-meteo.com/en/docs/historical-weather-api
11. Financing, imbalance cost. PBL, Eindadvies basisbedragen SDE++ 2026: 65/35 debt/equity, 4.25% debt, 9.0% equity, 25.8% corporate tax, 2.0% inflation; long-term imbalance cost EUR 7.0/MWh. https://www.pbl.nl/system/files/document/2026-02/pbl-2026-eindadvies-sde-plus-plus-2026-5698.pdf
12. Solar capex and O&M. PBL, Eindadvies basisbedragen SDE++ 2026, Table 8.13: building-mounted PV of 1 MWp and above, EUR 509/kWp investment, EUR 11.7/kWp/yr fixed O&M, EUR 0.0089/kWh variable O&M. https://www.pbl.nl/system/files/document/2026-02/pbl-2026-eindadvies-sde-plus-plus-2026-5698.pdf
13. Installed battery cost. PVB Energy Storage, How much does a C&I BESS really cost in 2026? Europe total installed cost guide (27 March 2026, updated 11 June 2026). A vendor guide citing no external data. Installed, including EPC, grid connection, fire safety, permits and commissioning: 241 kWh EUR 460-720/kWh; 422 kWh EUR 400-610; 1 MWh EUR 330-530; 5 MWh+ EUR 270-450. Hardware only EUR 140-200/kWh. https://www.pvb.com/nl/blog/how-much-does-a-ci-bess-really-cost-in-2026-europe-total-installed-cost-guide-for-cfos-updated-june-2026/
14. Time-dependent network tariffs. Netbeheer Nederland proposal, published by ACM for consultation on 19 March 2026: Voorstel codewijziging tijdsafhankelijke transporttarieven voor grootverbruikers. Planned start 2028 per Kennis van Energie, Tijdsafhankelijke transporttarieven voor grootverbruikers. https://www.acm.nl/nl/publicaties/voorstel-codewijziging-tijdsafhankelijke-transporttarieven-voor-grootverbruikers
15. Irradiance calibration. European Commission JRC, PVGIS 5.3, PVGIS-SARAH3 hourly series, Horsterparc Zeewolde, 2018-2023, east-west at 10 degrees, 14% loss: 871 kWh/kWp a year on average, against 969 from Open-Meteo irradiance through the same yield model. https://re.jrc.ec.europa.eu/pvg_tools/en/
16. Battery degradation. Figgener, J. et al., Multi-year field measurements of home storage systems and their use in capacity estimation, Nature Energy 9 (2024). 21 lithium-ion systems in Germany, up to 8 years of field data: usable capacity falls by about 2-3 percentage points a year. Mixed chemistries (NMC, LMO/NMC, LFP); a commercial LFP system in a conditioned enclosure is likely to fade more slowly. https://www.nature.com/articles/s41560-024-01620-9
17. Energy investment deduction. RVO, Energielijst 2026. EIA of 40%. Code 251118, storage of at least 5 kW and 15 kWh on the same connection as a renewable installation above 15 kW. Solar PV (code 251102) qualifies only from 15 to 100 kW per connection, and not with SDE++. https://www.rvo.nl/sites/default/files/2025-12/Brochure%20EIA%20-%20Energielijst%202026.pdf
1. Solar PV per capita. IEA PVPS, Snapshot of Global PV Markets 2026 (Report T1-49:2026, April 2026). PV per capita at the end of 2025: Australia 1,629 W, the Netherlands 1,601 W, Germany 1,404 W. https://iea-pvps.org/snapshot-reports/snapshot-2026/
2. Load profile. ELMAS dataset (Mines Paris for Enedis; Scientific Data, 2023), CC BY 4.0. Hourly smart-meter data from French business customers, 2018. Cluster 3: supermarkets, food wholesale and warehousing, including 719 medium-voltage warehousing sites. https://figshare.com/articles/dataset/ELMAS_dataset/23889780
3. Cold-store electricity use. Evans, J.A. et al., Specific energy consumption values for various refrigerated food cold stores, Energy and Buildings 74 (2014) 141-151. European mean of 65.9 kWh per refrigerated m3 per year for mixed chilled and frozen stores. https://www.sciencedirect.com/science/article/abs/pii/S0378778813007925
4. Feed-in congestion at the site. Gemeente Zeewolde, Beleidsvisie netcongestie Zeewolde (October 2025). Records TenneT’s declaration of feed-in congestion for Utrecht, Flevoland and Gelderland on 12 October 2021, now expected to last until 2033-2035. https://www.duurzaamzeewolde.nl/fileadmin/nieuwbouw-zeewolde/documenten/duurzaam/Beleidsvisie_netcongestie_Zeewolde_Def.pdf
5. Grid expansion delay. TenneT, Major expansion projects for the high-voltage grid in Utrecht, Flevopolder and Gelderland are taking longer than planned (12 June 2025). Completion moves from 2029 to 2033, or 2035 with further setbacks; splitting the regional grid into four sub-networks depends on a new high-voltage substation in Utrecht North. Large consumers and larger wind and solar producers wait longer for new or heavier connections. https://www.tennet.eu/nl-en/news/major-expansion-projects-high-voltage-grid-utrecht-flevopolder-and-gelderland-are-taking-longer-planned
6. Commodity price. Eurostat, nrg_pc_205_c, electricity prices for non-household consumers, Netherlands, band ID (2-20 GWh), energy and supply component, 2025. https://ec.europa.eu/eurostat/databrowser/view/nrg_pc_205_c/default/table
7. Network tariff. Liander, Tarieven voor aansluiting en transport elektriciteit 2026, MS category >136-2,000 kW: EUR 0.0226/kWh and EUR 3.57/kW/month on the monthly peak, excluding VAT. https://www.liander.nl/-/media/files/tarieven/grootzakelijk/tarieven-2026/tarieven-voor-aansluiting-en-transport-elektriciteit-2026_2.pdf
8. Energy tax. Belastingdienst, tarieven energiebelasting 2026. Electricity, 50,001 kWh to 10 million kWh: EUR 0.03735/kWh excluding VAT. https://www.belastingdienst.nl/wps/wcm/connect/bldcontentnl/belastingdienst/zakelijk/overige_belastingen/belastingen_op_milieugrondslag/energiebelasting/energiebelasting
9. Day-ahead prices. Energy-Charts (Fraunhofer ISE), day-ahead auction prices, bidding zone NL, 2025, from Bundesnetzagentur/SMARD, CC BY 4.0. Hourly to 30 September, 15-minute from 1 October; both averaged to half-hours. https://api.energy-charts.info/
10. Irradiance and temperature. Open-Meteo Historical Weather API (ERA5-based), hourly global tilted irradiance and 2 m air temperature, Horsterparc Zeewolde, 2025, east-west array at 10 degrees. https://open-meteo.com/en/docs/historical-weather-api
11. Financing, imbalance cost. PBL, Eindadvies basisbedragen SDE++ 2026: 65/35 debt/equity, 4.25% debt, 9.0% equity, 25.8% corporate tax, 2.0% inflation; long-term imbalance cost EUR 7.0/MWh. https://www.pbl.nl/system/files/document/2026-02/pbl-2026-eindadvies-sde-plus-plus-2026-5698.pdf
12. Solar capex and O&M. PBL, Eindadvies basisbedragen SDE++ 2026, Table 8.13: building-mounted PV of 1 MWp and above, EUR 509/kWp investment, EUR 11.7/kWp/yr fixed O&M, EUR 0.0089/kWh variable O&M. https://www.pbl.nl/system/files/document/2026-02/pbl-2026-eindadvies-sde-plus-plus-2026-5698.pdf
13. Installed battery cost. PVB Energy Storage, How much does a C&I BESS really cost in 2026? Europe total installed cost guide (27 March 2026, updated 11 June 2026). A vendor guide citing no external data. Installed, including EPC, grid connection, fire safety, permits and commissioning: 241 kWh EUR 460-720/kWh; 422 kWh EUR 400-610; 1 MWh EUR 330-530; 5 MWh+ EUR 270-450. Hardware only EUR 140-200/kWh. https://www.pvb.com/nl/blog/how-much-does-a-ci-bess-really-cost-in-2026-europe-total-installed-cost-guide-for-cfos-updated-june-2026/
14. Time-dependent network tariffs. Netbeheer Nederland proposal, published by ACM for consultation on 19 March 2026: Voorstel codewijziging tijdsafhankelijke transporttarieven voor grootverbruikers. Planned start 2028 per Kennis van Energie, Tijdsafhankelijke transporttarieven voor grootverbruikers. https://www.acm.nl/nl/publicaties/voorstel-codewijziging-tijdsafhankelijke-transporttarieven-voor-grootverbruikers
15. Irradiance calibration. European Commission JRC, PVGIS 5.3, PVGIS-SARAH3 hourly series, Horsterparc Zeewolde, 2018-2023, east-west at 10 degrees, 14% loss: 871 kWh/kWp a year on average, against 969 from Open-Meteo irradiance through the same yield model. https://re.jrc.ec.europa.eu/pvg_tools/en/
16. Battery degradation. Figgener, J. et al., Multi-year field measurements of home storage systems and their use in capacity estimation, Nature Energy 9 (2024). 21 lithium-ion systems in Germany, up to 8 years of field data: usable capacity falls by about 2-3 percentage points a year. Mixed chemistries (NMC, LMO/NMC, LFP); a commercial LFP system in a conditioned enclosure is likely to fade more slowly. https://www.nature.com/articles/s41560-024-01620-9
17. Energy investment deduction. RVO, Energielijst 2026. EIA of 40%. Code 251118, storage of at least 5 kW and 15 kWh on the same connection as a renewable installation above 15 kW. Solar PV (code 251102) qualifies only from 15 to 100 kW per connection, and not with SDE++. https://www.rvo.nl/sites/default/files/2025-12/Brochure%20EIA%20-%20Energielijst%202026.pdf