Imagine a computer game in which a planet is struggling to reduce CO2 levels and failure to do so will result in total annihilation of human life on said planet. You are in charge of the planet's eco system and are given 1 billion tons of aluminum to use in a manner which reduces the most CO2. The game is set-up such that if you use it well you will save the planet, but if you squander it, you will doom the inhabitants.
Where would you use it? You are given lots of choices. You could use it as a substitute for steel in cars to reduce weight and increase gas mileage. You could put it in electric transmission lines to reduce transmission losses. You can also use it to make frames for mounting solar panels. The game also allows use of 1 billion tons of copper, which could be put in electric motors and generators, reducing their internal electrical resistance and increasing the output of hydro, wind, or geothermal generators. Likewise labor is available and can be used to install insulation; capital can be used to finance energy efficiency projects with multi-year paybacks. In real life, all of these valuable resources can be and are used to save energy every day.
So to make the decision about where to use the resources in this game you will need to consider the relative potential of each resource to save CO2. If you decide to use these resources in an application which has a relatively low return for reducing CO2 emissions, you will lose the game. If you use 1 lb of aluminum to save 1 lb of CO2 when you could have saved 10, you wasted the resource. And that resource is now not available for the wiser use. This is the concept of the energy opportunity cost of non-energy inputs. Aluminum is not energy, but could have been used to make or save energy (and therefore reduce CO2).
I can tell you with certainty, that if you use these resources to make on-grid PV systems, you’ll lose the game and doom the planet. This is known because one can easily calculate the CO2 savings from PV and the number is $150/ton, ignoring any embedded CO2 in the productions and installation. We also know that CO2 trades on the EU market for $20/ton, meaning that someone, somewhere in the EU can reduce CO2 for a cost of $20/ton. So there are much better (less expensive) ways to reduce CO2 than installing solar panels on homes already connected to the grid.
PV is a very expensive way to address CO2 and might even have the effect of increasing CO2 when the energy opportunity cost of PV is considered. Using resources for PV crowds out the use of those resources in better applications. Using copper for PV will reduce the efficiency of electric motors. This may seem strange or non-connected, but PV systems consume copper and this will drive up the price of copper and result in less copper being used in the next generation of electric generators. The market is warning us not to install on-grid PV, but we are ignoring price signals. Sometimes ignoring price signals is the correct action, but one must be very careful when charting a course which ignores prices. Price signals are by far the best way to determine were resources are optimally used. When price signals tell you that using aluminum in transmission lines is more valuable than using it in PV, then that is a reliable answer. Today the market is screaming to stop using resources for on-grid PV, but we can’t hear the screams because solar subsidies have almost silenced the cries. Installing on-Grid PV systems is a terrible misuse of these resources and may cause us to lose the real game.