Tuesday, August 07, 2012

Solar PV Cost per GW

By request, a quick post on the cost of new solar vs. new nuclear.

{N.B. This analysis is like comparing apples to oranges, since solar PV delivers "peak" or daytime power, while nuclear provides "baseload" or power around the clock .}

The cost of solar panels continues to fall...PV panels from China now cost ~$0.75/W +/- (kick that up to about $1/W in the US because of our recent tariffs).

Then you need to add the cost to install the panels, which has (before the last couple years) represented ~50% of the installed cost.  Because panels have fallen in price so quickly recently they now represent just 1/4-1/3 of the total installed cost (panel + balance of system cost). 

Using this range/ratio (i.e. installed cost = 3 to 4 times panel cost), the cost to install 1 Giga-watt of solar PV is $2.25B-$3B (Giga = 1 billion).  [The panels themselves cost just $0.75B/GW.] 

If we could get back to panels representing 50% of the installed cost (probably a limit) that would put the installed cost of each GW at $1.5B (or ~$2B in the US) based on current prices.

Germany is reportedly installing multiple GWs of solar for under $2.5B/GW.  Using even the top price of $3B/GW, PV capacity seems to compare favorably with the cost of a new nuclear. That begs the question, what exactly does 1GW of nuclear cost? Industry claims ~$8B/GW, but history teaches us to double or triple what the industry claims because of cost overruns to something in the $16-$24B/GW range). 

That said, 1GW of nuclear can produce around 5-6 times more kWh than 1 GW of solar in each 24 hour day (because the nuke runs constantly).  This means a "fair" comparison of cost (if all you care about is total number of electrons generated in a 24 hour period) requires 5-6 GW of PV installed per GW of nuclear.  In a bulk electron generating race, expensive solar ($3 x 6 = $18B) beats expensive nuclear ($24B) by 30%, although cheap nuclear ($8) beats cheap solar ($10-11B) by a similar margin.  {Cheap and expensive indicating extremes of the cost estimates given above.}

In truth, we often care as much or more (measured by price) about when we get our electrons as the total number--compare "peak" rates to "off-peak" rates.  Since (and as long as) peak rates occur during the day, and off-peak occur at night, the value of the (equal number of) electrons generated by 1GW of nuclear will be less than the value of the electrons generated by 5-6GW of PV.
For example if peak rates are 50% higher than off-peak rates, the value of all the PV electrons total 20% more than all the nuclear electrons. 

This difference in the value of the power generated is essentially why comparing nuclear to PV is an apples and oranges comparision.  PV provides electrons when they are most valued.