When we turn on the lights or TV we do not worry about how the needed electricity is produced. It just is. Somehow, backstage, supply matches demand. And when everyone uses electricity at the same time, consumption peaks, triggering a production peak. This peak production, rather than the average production, is what sets the scale of the production capacity and of the grid. And not all electricity sources are equal in this respect:
With this third category, one may ask: is production completely random or is there some correlation between the weather (and thus production) and electricity consumption? In the case of solar power the link is clear: the production is highest in summer and during the day, with little production in winter and none at night. In some countries this is most welcome: electricity peaks are due to air-conditioning, and thus occur in summer. And the sunnier and hotter the summer, the greater the consumption… and the solar generation. In very sunny and very hot countries, air-conditioning consumes a lot of electricity and solar panels produce a lot of electricity, and they do it on the very same day.
In Germany, Britain and especially France, on the other hand, the consumption peaks occur in winter and summer consumption is lower. Solar production is thus completely off: it zigs when consumption zags. Solar panels contribute mostly to summer overproduction. And while the reports of negative prices of electricity in summer are somewhat misleading, it is a fact that in Europe electricity is far more valuable in winter than in summer. (Wind on the other hand blows harder in winter — I know this is true in France, not sure about other countries.)
This does not mean that most European countries cannot have any solar power at all. For instance, the projected large solar installations in Maghreb to export electricity to Europe would not have the same flaw given that their winter production would be substantial and the negative correlation between production and value weaker.
In order to compare various types of electricity production (coal, gas, nuclear, wind, solar, etc.) one looks at the production of a power plant and at the cost of building and running it: the ratio gives the cost per kilowatt-hour (kWh). But this implicitly assumes that all electricity is equally valuable, that a kWh produced in winter is worth as much as a kWh produced in summer.
The negative correlation between solar production and power consumption in Europe (i.e. one is high when the other is low) means that solar cells produce a lot of electricity of little use (and low price) in summer, and little that is very useful (and getting a high price) in winter. Consequently, the true cost of solar energy is underestimated. (It would be overestimated in air-conditioned countries, where the correlation is positive.)
Let us imagine an extreme country with two clear-cut seasons: very sunny summer followed by very dark and cold winter, with nothing in between. Half the year solar production would be phenomenal and electricity consumption derisory, and the other half of the year there would be no solar production and massive consumption. What would be the value of the solar production? Exactly zero: solar panels would produce electricity only when it is worthless and would not produce when it is very valuable. Yet, the ratio of total cost over total energy produced would give a seemingly attractive price. Real countries will of course exhibit a smaller difference, but it is clear that such a calculation is very misleading.
Claims that solar electricity is or soon will be at "grid parity" (i.e. competitive without subsidies) are based on such a mistaken calculation. Paying solar producers the same price for their summer production as others are paid in winter (or making such an assumption in calculations) is itself a subsidy, since it pays solar power more than it is worth.