Profitability of Dutch Utility-Scale PV Parks under Volatile and Negative Wholesale Electricity Prices

Abstract

Solar power has become a pillar of the Dutch electricity system, with 29.4 GWp installed by the end of 2025. That success undermines its own economics: because all solar parks produce in the same midday hours, they depress the price exactly when they have the most to sell. Negative-price hours on the Dutch day-ahead market reached 584 in 2025, and the solar capture factor, the ratio of the average price solar earns to the average market price, fell to 0.63–0.66. This thesis quantifies which design and operational strategies defend the lifetime profitability of a utility-scale solar park against that trend.

A 50 MWp reference plant on a 25 MW grid connection is simulated hour by hour on eleven years of measured weather and prices (2015–2025), with the SDE++ negative-price rule and full-load-hour cap inside the dispatch decision. Six scenarios add one change each to the same plant: naïve export (S0), smart curtailment (S1: no export at negative prices), a 10 MW/20 MWh battery on the firm connection (S2), the same battery trading the balancing market on a non-firm connection, curtailable by the grid operator in exchange for a lower import tariff (S3), the cap-and-floor contract for difference that replaces the SDE++ premium from 2027 (S4), and power-to-heat through an industrial heat pump (S5). The battery is priced at the end-2025 European turnkey price of €164/kWh; the higher NREL benchmark of about €395/kWh is carried as a sensitivity.

Design alone cannot repair the capture factor: at a two-to-one ratio of module to grid capacity the connection, not the array or the inverter, is the binding constraint. Smart curtailment adds €0.17–0.19M a year at no capital cost, lifts the plant's return from 6.9 % to 8.6 % on 2025 prices and is the only configuration that clears the 5 % cost of capital in all eleven market years. The battery lifts the 2025 capture factor from 0.57 to 0.93, repays its capital on day-ahead trading alone (break-even turnkey price €215–307/kWh with grid charging) and is the most valuable firm-connection configuration in both anchor years (NPV +€9.4M, IRR 9.4 % on 2025 prices); 27 of 36 battery sizes are value-positive, with a three-hour 20 MW/60 MWh optimum. The balancing market is the upside: a simple strategy without future price knowledge captures 15–40 % of its perfect-foresight pool against fee-inclusive break-evens of 18–28 %. At the higher benchmark cost no battery size repays on day-ahead trading and the case becomes a balancing decision. The non-firm connection is net-negative in every modelled case (−0.30 to −0.94 M€ a year in 2025): the discount applies only to charging while the curtailment risk falls on the exports. The cap-and-floor contract, settled against an annual reference price, cuts the year-to-year revenue standard deviation from €2.55M to €0.87M while leaving the battery's day-ahead cycling intact; a contract settled every period at a single strike would halve it. Power-to-heat is a conditional add-on carried by the heat off-taker, and a seasonal heat store cannot repay the charging capacity it requires.

On the harsher 2024 prices, and in a synthetic 2030 stress year built from the measured cannibalisation trend, the naïve plant falls below the cost of capital while the flexible configurations stay above it: flexibility becomes a condition for viability rather than an optimisation. A Dutch park should size its connection deliberately, curtail negative-price hours by default, add storage with balancing access as the upside, and treat power-to-heat as an add-on. For policy, the per-period suspension of support is incentive-compatible and belongs in the new contract, the annual reference price preserves the storage signal, and grid tariffs rather than subsidy decide the storage case at the margin.