Thursday, August 17, 2017

Cost of solar power (69)

As reported widely in RenewEconomy and elsewhere, there is big news from South Australia. A 150 MW solar thermal power station with molten salt storage for 8 hours is to be built by Solar Reserve 30 km north of Port Augusta at the top end of Spencer Gulf.

The politics of power generation in South Australia is extremely vexed, with proponents and opponents of renewable energy fighting bitterly for years.  In recent times, the State government has taken decisive steps to assure the local electricity supply at a reasonable price.  Earlier this year there was an announcement of a large battery system to be built by Tesla, and now we have the announced Aurora concentrated solar thermal (CST) facility.

As a keen student of CST, I'm keen to assess the Levelised Cost of Electricity for the Aurora project, the first of its kind in Australia.

First of all, some reported facts.  RenewEconomy says the cost of the 150 MW project is AUD 650 million.  According to RenewEconomy, the South Australian government will pay between AUD 75 and AUD 78 per MWh, with the deal contingent on financing of AUD 110 million coming from the federal government.

Solar Reserve says the project will have thermal storage good for 1,100 MWh of electricity and will produce 495 GWh of electricity per year (around 5% of the State's needs).  Solar Reserve further says that the normal output will be 135 MW, but there will be the capability to increase that to 150 MW at times of peak demand.  The Capacity Factor based on peak capacity is therefore 495,000 / (150 x 24 x 365) = 0.377.

Let me now estimate the LCOE for the Aurora project using my standard assumptions:
  • there is no inflation, 
  • taxation implications are neglected, 
  • projects are entirely funded by debt,
  • all projects have the same interest rate (8%) and payback period (25 years), which means that the required rate of capital return is 9.4%, 
  • all projects have the same annual maintenance and operating costs (2% of the total project cost), and 
  • government subsidies are neglected.
For further commentary on my LCOE methodology, see Yet more on LEC.

The results are as follows:

Cost per peak Watt          AUD 4.33/Wp
LCOE                              AUD 149/MWh

The components of the LCOE are:
Capital     {0.094 x 650 x 10^6}/{495,000 MWh} = AUD 123/MWh
O&M       {0.02 x 650 x 10^6}/{495,000 MWh} = AUD 26/MWh

Conclusion

I have been waiting for a long time to be able to present such an analysis for an Australian CST project.  At AUD 149/MWh, the LCOE is significantly above that for top-tier Australian PV projects such as Sun Metals (AUD 72/MWh) and Ross River (AUD 77/MWh).  Nevertheless, the benefits of despatchability are significant, especially in the context of bitterly disputed power shortages in South Australia in recent years.  Note again that the maximum price that the State government will pay for the electricity is AUD 78/MWh, so we can conclude (a) that Solar Reserve has performed some fancy financial engineering in arranging this project and (b) that this is a really good deal for the citizens of South Australia.


The graphic shows my LCOE results in USD/MWh over eight years at current exchange rates (AUD = USD 0.7928, EUR = USD 1.1799, JPY = 0.00903, GBP = USD 1.3044) and with the value of currency depreciated at 1.75% per year.  Red indicates solar thermal projects; blue indicates PV projects.  Filled-in circles are for projects that were completed when I made my LCOE assessment; non-filled-in circles are for projects as announced, even if not completed.


Monday, August 7, 2017

Cost of solar power (68)

I've been busy on other projects lately, without time to blog on developments with the cost of solar power.  In the meantime, there have been various announcements of new PV systems in the state of Queensland, culminating yesterday in the announcement of a 1 GW facility.

Now 1 GW is a big round number, surely warranting that I add the project to my set of case studies.

As ever, RenewEconomy has the story.  The project is the Wandoan South Solar Project, to be built near Wandoan in the Surat Basin, up to now better known as a coal and gas province.  The project developer is Singapore-based Equis Energy and the project is described in this press release.  In brief, the 1 GW project will involve a capital investment of AUD 1.5 billion and construction will begin in 2018 and be complete in 20 19.  The annual output from the 3 million solar panels will be 1,800,000 MWh.

The Capacity Factor for the Wandoan project is 1,800,000 / (1000 X 365 X 24) = 0.205.  That's not particularly high, so I conclude the panels will be fixed, although that's not stated in the press release.

Let me now estimate the LCOE for the Wandoan South Solar Project using my standard assumptions:
  • there is no inflation,
  • taxation implications are neglected,
  • projects are entirely funded by debt,
  • all projects have the same interest rate (8%) and payback period (25 years), which means that the required rate of capital return is 9.4%,
  • all projects have the same annual maintenance and operating costs (2% of the total project cost), and
  • government subsidies are neglected.
For further commentary on my LCOE methodology, see Yet more on LEC.

Note that I am now using annual maintenance costs at 2% of capital cost.

The results are as follows:

Cost per peak Watt   AUD 1.50/Wp
LCOE                        AUD 95.00 / MWh

The components of the LCOE are:

Capital  {0.094 X 1.5 X 10^9} / 1,800,000 = AUD 78.33 / MWh
O&M      {0.02 X 1.5 X 10^9} / 1,800,000 = AUD 16.66 / MWh

Conclusion

The LCOE (AUD 95/MWh) is not as good as the last two projects that I have analysed from Queensland, namely Sun Metals (AUD 72/MWh) and Ross River (AUD 77/MWh).  I wonder if Equis Energy has just been cautious in their early announcement.  I also think the Capacity Factor (0.205) is on the low side for fixed panels in an area where the solar resource is excellent.


The graphic shows my LCOE results in USD/MWh over eight years at current exchange rates (AUD = USD 0.7928, EUR = USD 1.1779, JPY = USD 0.00903, GBP = USD 1.3044) and with the value of currency depreciated at 1.75% per year.  Red indicates solar thermal projects; blue indicates PV projects.  Filled-in circles are for projects that were completed when I made my LCOE assessment; non-filled-in circles are for projects as announced, even if not completed.





Thursday, February 9, 2017

Cost of solar power (67)


Three weeks ago, I analysed the Levelised Cost of Electricity (LCOE) for the Sun Metals utility-scale PV project near Townsville, Queensland.  The LCOE was excellent, AUD 72 per MWh according to my standard assumptions.

Today, I’ll analyse another recently-announced PV installation in the same region, namely the Ross River Solar Farm.  As with Sun Metals, the Ross River project will have single-axis horizontal tracking, but now the panels will be crystalline PV as opposed to thin-film Cd-Te.  According to the project developers, the site is a 202 Ha disused mango farm, the peak capacity will be 135 MW and the cost is AUD 225 million.  Construction will commence in the first quarter of 2017 and is expected to take 12 months.

The output from Ross River Solar Farm is the subject of a power purchase agreement with EnergyAustralia as part of its obligations under the federally mandated Renewable Energy Target.  I’ll estimate the annual output using the same Capacity Factor as for the Sun Metals project, namely 0.28.  A defence of that CF is given here.  The annual production is therefore estimated as 0.28 × 365 × 24 × 135 = 331,128 MWh.

 Let me now estimate the LCOE for the Ross River Solar Farm using my standard assumptions:
  • there is no inflation
  • taxation implications are neglected,
  • projects are funded entirely by debt,
  • all projects have the same interest rate (8%) and payback period (25 years), which means that the required rate of capital return is 9.4%,
  • all projects have the same annual maintenance and operating costs (2% of the total project cost), and
  • government subsidies are neglected.
For further commentary on my LCOE methodology, see posts on Real cost of coal-fired power, LEC – the accountant’s view, Cost of solar power (10) and (especially) Yet more on LEC.

Note that I am now using annual maintenance costs of 2% of capital cost.

The results are as follows:

Cost per peak Watt              AUD 1.67/Wp
LCOE                                     AUD 77.46/MWh

The components of the LCOE are:

Capital           {0.094 × 225 × 106}/{331,128 MWh} = AUD 63.87/MWh
O&M              {0.020 × 225 × 106}/{331,128 MWh} = AUD 13.59/MWh

Conclusion

Well, the LCOE (AUD 77/MWh) isn’t quite as good as that for the Sun Metals project (AUD 72/MWh), but it’s still very good, and the second best that I’ve analysed in Australia.

This confirms the continuing decrease in the cost of solar power, as shown in the LCOE graphic below.  The graphic shows my LCOE results in USD/MWh over eight years at current exchange rates (AUD = USD 0.7541, EUR = USD 1.069, JPY = USD 0.00868, GBP = USD 1.2544) and with the value of currency depreciated at 1.75% per year.  Red indicates solar thermal projects; blue indicates PV projects.  Filled-in circles are for projects that were completed when I made my LCOE assessment; non-filled-in circles are for projects as announced, even if not completed.

My concluding observation is that whilst Australian politicians continue to bicker about the development of renewal energy in the local market, our enterprises and our citizens are just getting on with the transformation.  One day, in a short amount of time, the politicians will wake up and realise that the energy landscape has changed before their very eyes, at great disruption to incumbent large players, and in a way that helps lower our greenhouse gas emissions in line with the Paris Agreement.  What will the politicians squabble about then?


Monday, January 30, 2017

Cost of solar power (66)


This will be a first for me – I’m going to analyse the Levelised Cost of Electricity (LCOE) for a large PV installation in Cixi, China, 150 km south of Shanghai.  Because of China’s rapid industrialisation and cheap labour, I’m expecting the LCOE will be somewhere near the best in the world to date, but let’s see.

PV Magazine has the story, also reprinted in RenewEconomy.

The PV installation is 200 MW, presumably AC to grid.  The panels are mounted over a 299.5 Ha fish farm, “deliberately spaced far apart for enough sunlight to penetrate the water, which is critical for the growth of the fish beneath the surface”.  From the picture accompanying the article, it looks like the panels are fixed.

The project is expected to generate 220 GWh of electricity per annum (roughly enough for 100,000 homes) and cost 1.8 billion yuan (or USD 262.6 million at today’s exchange rate).  Construction started in late June 2016 and finished in December 2016.

The Capacity Factor for the Cixi project is 220,000 / (200 * 365 * 24) = 0.126, which is surprisingly low, given that the latitude is about 30°N and the solar resource should be good.  The CF value is consistent with fixed panels.

Let me now estimate the LCOE for the Cixi project using my standard assumptions:
  • there is no inflation,
  • taxation implications are neglected,
  • projects are funded entirely by debt,
  • all projects have the same interest rate (8%) and payback period (25 years), which means that the required rate of capital return is 9.4%,
  • all projects have the same annual maintenance and operating costs (2% of the total project cost), and
  • government subsidies are neglected.
For further commentary on my LCOE methodology, see posts on Real cost of coal-fired power, LEC – the accountant’s view, Cost of solar power (10) and (especially) Yet more on LEC.

Note that I am now using annual maintenance costs of 2% of capital cost. 

The results are as follows:

Cost per peak Watt              CNY 9/Wp
LCOE                                     CNY 933/MWh

The components of the LCOE are:
Capital           {0.094 × 1.8 × 109}/{220,000 MWh} = CNY 769/MWh
O&M              {0.020 × 1.8 × 109}/{220,000 MWh} = CNY 164/MWh

Conclusion

At the exchange rate of CNY 6.85 to the USD, the LCOE for the project is USD 136/MWh.  That’s quite a lot more expensive than recent big projects I’ve analysed as you can see from the graphic below.  The Capacity factor for the Cixi project is poor, hence the annual output is not as large as anticipated, and hence the LCOE is not up to international best practice.  I’m surprised.

The graphic shows my LCOE results in USD/MWh over eight years at today’s exchange rates (AUD = USD 0.75431, EUR = USD 1.06928, JPY = USD 0.00868, GBP = USD 1.25437) and with the value of currency depreciated at 1.75% per year.  Red indicates solar thermal projects; blue indicates PV projects.  Filled-in circles are for projects that were completed when I made my LCOE assessment; non-filled-in circles are for projects as announced, even if not completed.

Sunday, January 15, 2017

Cost of solar power (65)


It’s been six months since I blogged about the cost of PV projects, and I suspect that the cost of solar power has continued to fall rapidly.  Let’s see whether that’s true by estimating the Levelised Cost of Electricity (LCOE) for the Sun Metals solar PV plant, 15 km south of Townsville in Queensland, Australia.

PV magazine has the story.  The 100 MW installation will be complete in Q1 2018 and will exploit a connection to Sun Metals’ existing substation.  The project features one-axis tracking, thin film CdTe panels from First Solar and is part of a major upgrade of Sun Metals zinc operations. 

PV magazine states the cost of the project is AUD 155 million, but doesn’t explicitly mention the Capacity Factor.  So, let me use data from the Australian Renewable Energy Agency (ARENA), which says that the average Capacity Factor for one-axis tracking installations in Queensland is 0.28.  (By way of defence of this CF, it should be noted that the solar resource near Townsville is excellent, even if it doesn’t quite match that of the best locations in the world, such as Chile.)

Under that CF assumption, the annual output of the Sun Metals project would be 0.28 × 365 × 24 × 100 = 245,280 MWh per year.

Let me now estimate the LCOE for the Sun Metals project using my standard assumptions:
  • there is no inflation,
  • taxation implications are neglected,
  • projects are funded entirely by debt,
  • all projects have the same interest rate (8%) and payback period (25 years), which means that the required rate of capital return is 9.4%,
  • all projects have the same annual maintenance and operating costs (2% of the total project cost), and
  • government subsidies are neglected.
For further commentary on my LCOE methodology, see posts on Real cost of coal-fired power, LEC – the accountant’s view, Cost of solar power (10) and (especially) Yet more on LEC.

Note that I am now using annual maintenance costs of 2% of capital cost. 

The results are as follows:

Cost per peak Watt              AUD 1.55/Wp
LCOE                                     AUD 72.04/MWh

The components of the LCOE are:

Capital           {0.094 × 155 × 106}/{245,280 MWh} = AUD 59.40/MWh
O&M              {0.020 × 155 × 106}/{245,280 MWh} = AUD 12.64/MWh

Conclusion

Wow, the LCOE of AUD 72/MWh is a stunning figure for Australia and not far behind global best practice.  At today’s foreign exchange rate of AUD 1.00 = USD 0.75, my LCOE estimate is USD 54/MWh, which is way below my previous best result for Australia. 

As shown in the LCOE graphic below, the cost of solar power continues to fall rapidly.  The graphic shows my LCOE results in USD/MWh over eight years at today's exchange rates (AUD = USD 0.7495, EUR = USD 1.064, JPY = USD 0.00873, GBP = USD 1.2179) and with the value of currency depreciated at 1.75% per year.  Red indicates solar thermal projects; blue indicates PV projects.  Filled-in circles are for projects that were completed when I made the LCOE assessment; non-filled-in circles are for projects as announced, even if not completed.

It’s clear we are in the midst of an amazing technological/industrial revolution, even if this is not widely appreciated by the media in this country.  One might well ask how much longer this decline in the LCOE will continue!