I first wrote this article in December 2020, when the duck curve was still a forecast for Cyprus. In 2026 it is the daily reality of our grid. Below is the original explanation, tightened, followed by a dated 2026 section on what has actually happened.
What the duck curve is
The duck curve is the shape the daily net load takes when a lot of solar is connected to a power system. Net load is demand minus solar generation, in other words what the conventional power plants still have to supply. As solar grows, net load is pushed down in the middle of the day and then rises sharply in the late afternoon when the sun sets. Plotted over 24 hours, the curve looks like a duck: a low belly at midday and a steep neck into the evening.
The term comes from California, which saw it first. Any sunny grid with high solar penetration gets the same shape.
Why it is a problem: the ramp-up
The duck curve creates two problems at once.
The first is the midday belly. When net load falls too low, the conventional units that must stay online for technical reasons cannot go any lower, and the surplus solar has nowhere to go. The result is curtailment: clean, cheap energy that is simply thrown away.
The second is the evening ramp-up. Demand in Cyprus peaks in the evening, at the same time that solar output is collapsing. The conventional fleet has to increase its output very quickly to fill that gap. Steam turbine plants like the ones we have in Cyprus need time to go from low output to full output because of heating-up times. Ramping fast is hard on them, it raises the cost of electricity and, in extreme cases, it puts security of supply at risk.
What I projected in 2020
In 2020 I modelled the Cyprus load curve with the solar capacity the government expected by 2030 under the National Energy and Climate Plan. My conclusion then was that the first 300 MW of solar would not change the shape of the load much. Beyond that, the evening ramp would grow quickly: with 500 MW of solar the ramp from 4 pm to 6 pm reached about 100 MW per hour in my model, and with 750 MW about 140 MW per hour. Cyprus had never seen ramp rates like that.
I expected the problem to become serious around 2029. I also wrote that the answers already existed: energy storage, demand response and anything else that adds flexibility to the grid.
2026 update: Cyprus is on the duck curve now
Section added October 2026.
I was wrong about the timing. The duck curve did not wait for 2029. The clearest evidence is curtailment, which is exactly what the midday belly produces in an isolated system that cannot export its surplus.
Curtailment is now the norm, not the exception. At Renewable Labs we went through the DSO’s (ΔΣΔ) daily curtailment reports for October 2025 to August 2026, a period of 335 days. Photovoltaic systems of 120 to 400 kW controlled via SCADA were curtailed on 236 of those 335 days. Systems above 400 kW were curtailed on 325 of the 335 days. The curtailments fall mostly between 10:00 and 16:00, the best solar hours of the day, which is precisely the belly of the duck. (Source: DSO daily curtailment reports, Renewable Labs analysis.)
Large parks lose a big share of their energy. As I wrote in CBN in June 2026, residential photovoltaic systems were curtailed up to 17 times within one month, and in large photovoltaic parks curtailments reached 66% of their energy in April (The solution is right in front of us: energy storage).
Must-run units make the belly worse. The conventional units that have to stay online for system security set a technical minimum. The more of them that must run, the less room is left for solar at midday. I covered this in my July 2026 MoneyMatters piece on central storage (Central storage: an important step, not the only road).
Storage is arriving, slowly. The central storage contract was signed in 2026, for 120 MW expected in 2027. A grid-forming battery of that kind can, if it performs as designed, reduce the need for must-run units and so leave more room for solar. But one central project is not enough, and as I wrote in June, the DSO had issued connection terms for only 16 applications for storage at existing photovoltaic parks, 29 MW in total, before it stopped issuing terms.
What this means for businesses with solar in Cyprus
If you own a commercial or industrial PV system in Cyprus, the duck curve is no longer an abstract system issue. It shows up as lost production on your own meter. Your inverter app shows what the system produced, not what it lost when it was curtailed, so most owners do not know the real cost.
The practical steps are the same ones I named in 2020, only now they are urgent: measure the curtailment loss on your own system, then look at behind-the-meter storage and at shifting consumption into the midday hours. A battery that charges from your own curtailed solar at noon and discharges into the evening peak works directly against the duck curve, for you and for the grid. If you want help sizing that case, see my battery storage advisory for businesses in Cyprus.