How can we compare the cost, performance and value-for-money of alternative large-scale clean energy projects? Actually, it’s pretty tough to try and avoid apples-and-oranges comparisons. Still, some adjustments can be made, such as for capacity factor, to partially levelise comparisons.

Below is a simplified comparison of four recent real-world projects. All can be considered first-of-a-kind installations, except for the wind farm.

1. A large proposed wind farm in South Australia (600 MWe peak)

The wind project will use 180 of the 3.4 MWe Suzlon turbines and “generate enough electricity to power 225,000 homes“. It includes a biomass plant that could produce up to 120 MWe of backup power to cover low-wind periods, and might offset up to 2.5 million tonnes of CO2 per year. At average 8m/s winds the capacity factor is estimated to be about 35%. A 60 km undersea high voltage direct current cable will connect it with Adelaide. Cost is $1.3 billion for the generating infrastructure and $0.2 billion for the cable.

2. A large Generation III+ nuclear power plant in Finland (1600 MWe peak)

The in(famous) Olkiluoto 3 NP unit, a European Pressurised Reactor (EPR) being built by the French (AREVA). The project has seen significant delays (first electricity now expected in 2014), and a cost blowout from the original € 3.7 billion to a new figure of € 6.4 billion. Despite this, the Fins have ordered two more EPR units. Assume it runs at the average Finnish capacity factor of 86%.

3. A large solar PV plant under construction in New South Wales (150 MWe peak)

To be built in Moree, this will cover 3.4 km squared with 645,000 multi-crystalline PV panels, and is forecast to output 404 GWh per year (enough for 45,000 households). Part of the “Solar Flagships” programme, the cost is $A 923 million. Estimated to abate 364,000 tonnes of CO2 per year (based on NSW emission factor 0f 0.9 tCO2/MWh). Estimated capacity factor is 30.7% (based on peak power and GWh forecasts) — this seems high compared to typical PV performance.



4. A state-of-the-art solar concentrating power plant with some energy storage, in Spain (20 MWe peak)

The Gemasolar CSP plant in Andalucía started electricity production in late 2010. There is a detailed summary of projected performance here. Peak output is 19.9 MWe, anticipated output of 110,000 MWh/yr from a mirror field covering 190 ha. Cost of € 230 million (much higher estimate of £ 260 million here). It is a Power Tower facility with 2,650 heliostat units (120 m2 each) and 15 hours of thermal storage (not sure of total amount of thermal energy stored – I presume enough for peak turbine output, which would be just under 300 MWh of final electrical energy). The capacity factor as a result of the thermal storage is anticipated to be 63.1% (the site has 270 productive days per year thanks to the excellent desert siting).

——————-

Okay, let’s try a few ways of squaring these options off — and then explore details and alternative assumptions/calculations in the comments. (I know some have already been reported by various commenters in the BNC Open Threads, but this is a good place to centralise and reiterate/update them.)

First, let’s consider the capital cost after standardisation for capacity factor, and bringing output up/down to an equal power rating of a large commercial facility (1 GWe electric average, the size of a big coal-fired power station or the average output of the AP1000 reactor) and then equalise this to $USD. Average exchange rates for 2011 are 1 USD = 0.73 Euro = 1.03 AUD = 0.63 UKPS.

Wind (biomass backup): $US 6.9 billion/GWe

Gen III+ Nuclear: $US 6.0 billion/GWe

Solar PV (no storage/backup): $US 19.6 billion/GWe

Solar CSP (thermal storage): $US 25.1 billion/GWe (or $32.9 billion if the higher cost estimate is correct)

The rank is nuclear, wind, PV, CSP. The first two are one third the cost of the solar options, but the wind has only 20% of its peak output backed by the biomass.

What about the levelised cost of electricity (LCOE)? Such calculation involves many assumptions. Here is just one set, entered into the handy NREL calculator. All discounts set to 8.5%. The nuclear plant has an estimated operational lifespan of 60 years, whereas we can generously assume a 25 year lifespan for the three renewable installations (Q: does anyone know wind/solar farms that have run for longer?). For financing purposes, however, I will set the nuclear option to be a 30-year term, after which the LCOE is lowered for the remaining 30 years of operation.

Wind [ignoring biomass fuel] (Cap $2,427; Fixed O&M 50; Var O&M 0.002; Heat Rate 0; Fuel 0), LCOE = 10.0 c/kWh for 25 years, then infrastructure replaced

Nuclear (Cap $5,137; Fixed O&M 150; Var O&M 0.005; Heat Rate 10,000; Fuel 1), LCOE = 9.9 c/kWh for 30 years, then 3.6 c/kWh for the next 30 years (average = 6.75 c/kWh)

Solar PV (Cap $6,032; rest same as wind), LCOE = 24.4 c/kWh for 25 years, then infrastructure replaced

Solar CSP (Cap $16,833; rest same as wind), LCOE = 29.6 c/kWh for 25 years, then infrastructure replaced

It is clear again which option is most cost effective (especially given the baseload output and load-following capacity of the EPR reactor), based on current or recently proposed costs and performance figures. If people argue strongly below for me to modify any of the LCOE assumptions, I’ll consider it and may edit the above accordingly — I don’t claim these are final or definitive.