This page is supplementary material to the post “Electricity in Texas part II – the cost of a 100% renewable grid” on Climate etc.
The PTC is no longer available for new solar PV projects. Current Federal legislation provides an ongoing 10% ITC (investment tax credit) for solar PV generation projects. The ITC (at higher levels) was an alternative to the PTC when that was still available for solar PV.
The future estimates below assume no subsidy.
Solar – current unsubsidised Texas LCOE – $47 / MWh
The Lawrence Berkeley National Laboratory Utility Scale Solar 2015 report August 2016 shows two recent Texas solar PV PPA’s for $31 and $34 / MWh after deduction of the PTC subsidy. These are for relatively small projects around 150 MW. Taking the lower value of $31 / MWh, and, as for wind, adding $15 / MWh for the value (not cost) of the PTC subsidy and $1 / MWh for the market value of the Texas RECs (Renewable Energy Certificates) gives an estimate for the current lowest total unsubsidised solar PV price of $47 / MWh.
Solar – current lowest unsubsidised global LCOE – around $30 / MWh
Reverse auctions in ideally sunny locations have yielded the following unsubsidised PPA prices for systems to be delivered within a couple of years:-
- $29.1 / MWh, Atacama Desert, Chile, 120MW, August 2016 (40% cheaper than the previous auction)
- $29.2 / MWh (unweighted) $24.2 / MWh (weighted), Dubai, Abu Dhabi, 1.2GW. Construction started May 2017 so the project is real. 3 of the 5 bidders bid under $30 / MWh.
- Around $40 / MWh for many large projects in India, Chile, UAE and elsewhere
Similarly to wind prices in Morocco, the cost of capital in these bids is unknown, and Texas labor rates will be higher than for these bids.
The Texas CREZ zones receive 20-25% fewer kWh per square metre annually than Abu Dhabi, and 30% fewer than Chile’s Atacama desert which has the best sunlight in the world. The LCOE is inversely proportional to this irradiance. If capital and labor costs were the same, the unsubsidised LCOE for Texas PV would thus be 25-33% higher than Dubai ($29.9 becomes $38-$40 / MWh), and 45-50% higher than Chile ($29.1 becomes $42 to $44 / MWh). Thus $47 / MWh for current Texas costs seems to be the right ball park after making some allowance for increased labour charges and possible cost of capital differences in Texas.
Solar PV 2030-2040 unsubsidised Texas high LCOE estimate – $30 / MWh
The learning rate (reduction in cost with each doubling of installed capacity) for solar PV systems varies between the different system components. At present the cost of panels and inverters forms the majority of the costs and the learning rates for these components are of the order of 21% per doubling. (See the Agora Energiewende document Current and Future Cost of Photovoltaics document).
The conservative, high-end estimate for 2030-2040 assumes two-third of this learning rate, or 14% per doubling. A 10% compound growth rate for global solar PV capacity would give 3-4 doublings for a total of eight to sixteen times the current solar PV global installed capacity by 2035. On the conservative assumptions, 3 doublings at 14% per doubling would bring the cost of solar PV down by 1/3rd to around $30 / MWh for utility-scale solar PV in Texas CREZ areas. By that time the balance of system costs (including inverters) would predominate over solar panel costs.
The required CREZ network transmission network is included separately in the cost model.
Solar PV 2030-2040 unsubsidised Texas low LCOE estimate – $21 / MWh
Current silicon PV cells (single junction) are in the low 20%s efficiency. Current multi-junction technologies can double the efficiency to 40-45% at high concentration (typically a few hundred “suns” which independently benefits efficiency of any solar PV cells). Multi-junction technology will always be much more expensive for a given active area. Using concentrators made from cheap materials such as plastics, the expensive active PV area required can be reduced by a factor of one hundred or more (e.g. for 300 suns concentration). However, concentrators require two-axis tracking to keep direct sunlight focused on the active PV areas, which adds cost. This combination of technologies is not yet mature, and development of it has to be financed while the market price of standard silicon solar PV panels is still dropping like a stone.
Balance of system costs, excluding inverters, come down with increased module efficiency. In a 15-20 year time frame doubling efficiency would halve the panel interconnection and labour costs and not quite halve the physical solar farm area as two-axis tracking takes a little more space. The inverter and other AC equipment costs are unchanged. A crude estimate might be that costs could be reduced by 1/3rd compared with the high LCOE estimate, for a $20 / MWh low LCOE in Texas CREZ areas.
An alternative low 2030s LCOE estimate assumes the full 21% historical learning rate for LCOE reductions in the document above for three and a half doublings, giving a 56% reduction on $47 / MWh for $21 / MWh, which is comparable and is the figure used in the cost model spreadsheet.
CREZ network transmission costs are a separate item in the cost model spreadsheet.