Thursday, June 30, 2011

Cost of solar power (16)

Beyond Zero Emissions (BZE), the Melbourne-based advocacy group, has attacked the decision of the Solar Flagships program to fund the Solar Dawn project.  According to the BZE report, the hybrid solar-gas concept of Solar Dawn is outdated and the Compact Linear Fresnel Reflectors (CLFR) don’t deliver high enough temperatures for best-practice renewable power generation.  BZE identifies the recently-opened Gemasolar tower plant with molten salt storage as the way of the future, so that installation will be the subject of today’s post.

First, let me mention that I analysed the Solar Dawn project in Cost of solar power (13).  My finding was that the Levelised Electricity Cost (LEC) was pretty much the same as several other large projects commissioned recently.  I don’t at all dispute BZE’s contention that molten salt storage is preferable to hybrid firing with gas.  The resources of natural gas are finite, and gas combustion only exacerbates anthropogenic global warming.  The sooner we move away from gas as an energy source the better.

I do however dispute BZE’s view that tower technology is superior to CLFR, and by inference to other solar thermal technologies.  What is important is to have cheap electricity available on a despatchable basis, and I consider the jury is still out regarding the best solar thermal technology.  Indeed I can highlight my own technology based on passive solar heat collection as a contender for cheap large-scale solar thermal power generation; for details see http://www.sunoba.com.au/.

The Gemasolar installation was commissioned in late May 2011 and is located near Seville in Southern Spain.  This is a very sunny location with 2,062 kWhr/(m^2.yr) solar resource.  Gemasolar is the outcome of a joint venture, Torresal Energy, between Masdar (Abu Dhabi energy company) and Sener (Spanish engineering firm).

In Gemasolar, the sun’s rays are reflected by 2,650 heliostats, each of 120 m^2, to a receiver at the top of a 150 m central tower.  The heliostats are dispersed over a land area of 185 Ha.  Sufficient heat energy is stored in molten salt for 15 hours operation, so the facility can almost be considered to provide base-load power.  Temperatures in the receivers reach 900°C and the temperature range of the molten salt storage is between 290°C and 565°C.  The peak output is 19.9 MW and the annual output is given as 110 GWhr.  Heat energy in the molten salt drives a wet-cooled Rankine-cycle steam turbine.

The UK Daily Mail reports that the cost of Gemasolar was GBP 260 million or EUR 289 million.  One further snippet I gleaned from various press reports is that the project will involve an Operations and Maintenance crew of 45.

I now evaluate the Levelised Electricity Cost (LEC) using my customary 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 (3% of the total project cost), and
          government subsidies are neglected.

For further commentary on my LEC methodology, see posts on Real cost of coal-fired power, LEC – the accountant’s view and Cost of solar power (10).

The results are:

Cost per peak Watt EUR 14.52/Wp
LEC                                        EUR 326/MWhr

The components of the LEC are:
Capital           {0.094× EUR 289 ×10^6}/{110,000 MWhr} = EUR 247/MWhr
O&M  {0.030× EUR 289 ×10^6}/{110,000 MWhr} = EUR 79/MWhr

By way of comparison, LEC figures (in appropriate currency per MWhr) for all projects I’ve investigated are given below.  The number in brackets is the reference to the blog post, all of which appear with the title “Cost of solar power ([number])”:

(2)        AUD 199 (Nyngan, Australia, PV)
(3)        EUR 547 (Olmedilla, Spain, PV)
(3)        EUR 205 (Andasol I, Spain, trough)
(4)        AUD 257 (Greenough, Australia, PV)
(5)        AUD 432 (Solar Oasis, Australia, dish)
(6)        USD 177 (Lazio, Italy, PV)
(7)        AUD 295 (Kogan Creek, Australia, CLFR pre-heat)
(8)        USD 248 (New Mexico, CdTe thin film PV)
(9)        EUR 218 (Ibersol, Spain, trough)
(10)      USD 251 (Ivanpah, California, tower)
(11)      CAD 445 (Stardale, Canada, PV)
(12)      USD 315 (Blythe, California, trough)
(13)      AUD 310 (Solar Dawn, Australia, CLFR)
(14)      AUD 286 (Moree Solar Farm, Australia, single-axis PV)
(15)      EUR 381 (Lieberose, Germany, thin-film PV)
(16)      EUR 326 (Gemasolar, Spain, tower)

The Capacity Factor for Gemasolar is 110,000 / (19.9 × 24 × 365) = 0.63, which I think is the world’s highest CF for a solar power station.

To summarise, the cost per peak Watt for Gemasolar is high (indeed the highest I have analysed), but that is not surprising in view of the fact that it has 15 hours of thermal storage.  The Capacity Factor is excellent, approaching base-load, and the LEC is actually quite high compared to trough installations (also with storage) at Andasol and Ibersol. 

Moreover, at today’s exchange rates, I calculate the LEC for Gemasolar is around AUD 447/MWhr, or 44% more expensive than my estimate for Solar Dawn.  To conclude, I don’t agree with BZE’s contention that tower technology has been proved superior to CLFR technology as installed at Solar Dawn.

Cost of solar power (15)

Today I’m going to analyse the Lieberose Solar Park in Germany. 

Firstly, however, I want to discuss yet another source of inexactitude in assessing the cost of solar power.  This issue is relevant to development of the solar power industry in general and to photovoltaics in particular.

To analyse the performance of a solar power station, three items of information are mandatory: the peak power output, the annual power output and the cost.  Now I can accept that the cost is not necessarily going to be publicly available.  Whether it is will depend on the way the financing has been put together and whether public money is involved.  If the cost is publicly known, then we can proceed to examine the annual power output.

Of course, what we are looking for here is a clear statement of the annual output to the grid measured in MWhr per year.  You’d think that would be simple, but not so.  In my experience, less than half of project press releases provide this information.  More usually the annual output is described in the form of tonnes of CO2 abatement per year.  But what does that mean?  In Australia where I am based, CO2 emissions from coal-fired electricity are typically estimated as 0.8 to 1.0 t CO2 per MWhr.  In Germany where there are many PV plants, I’m not sure of the basis for the estimate.  A detailed project report I have just obtained (see Lieberose below) gives 0.673 t CO2 per MWhr.  And in the USA, my analysis for a PV project in New Mexico gave 0.391 t CO2 per MWhr – see Cost of solar power (8).

Another issue relates to peak power.  Again you’d think this would be simple, but again it’s not, particularly for PV installations.  The key issue here is whether the output is measured in MW (DC) from the panels or in MW (AC) to the grid.  Tom Cheyney has written entertainingly and knowledgeably on this issue.  Here’s an extract from his article in which he comments on the Finsterwalde PV plant in Germany, rated at 82 MW:

“Using my usual quick-and-dirty formula for converting DC to AC for a PV system—MW (DC) divided by 1.2—the AC rating for the Finsterwalde trifecta comes to 68.3MW. Others use the following math: MW (DC) multiplied by 0.86, which in this case puts the German farm’s total at 70.52MW (AC).”

Cheyney goes on to explain that most European installations are measured in MW (DC), with the honourable exception of announcements from Sun Power.  On the other hand, most installations in North America are measured in MW (AC).  The above comments apply to utility-scale projects, not for domestic rooftop installations, in which DC ratings are generally quoted.

Having taken on board those cautionary comments, let’s look at the Lieberose project.  This installation, located in a former military training area approximately 100 km SE of Berlin in the German state of Brandenburg, went fully on-line in October 2009.  The area of the site is 162 Ha, the area of modules is approximately 500,000 m^2, the peak output is 52.8 MW (presumably DC as we have discussed above), the annual output is 52 GWhr (not sure whether this is AC or DC), the CO2 abated is 35,000 t per year, and the cost is approximately EUR 160 million.  In all, there are approximately 700,000 First Solar thin-film modules.  My understanding from the material I read is that the panels are fixed.

I now evaluate the Levelised Electricity Cost (LEC) using my customary 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 (3% of the total project cost), and
          government subsidies are neglected.

For further commentary on my LEC methodology, see posts on Real cost of coal-fired power, LEC – the accountant’s view and Cost of solar power (10).

The results are:

Cost per peak Watt EUR 3.03/Wp
LEC                            EUR 381/MWhr

The components of the LEC are:
Capital           {0.094× EUR 160 ×10^6}/{52,000 MWhr} = EUR 289/MWhr
O&M              {0.030× EUR 160 ×10^6}/{52,000 MWhr} = EUR 92/MWhr

By way of comparison, LEC figures (in appropriate currency per MWhr) for all projects I’ve investigated are given below.  The number in brackets is the reference to the blog post, all of which appear with the title “Cost of solar power ([number])”:

(2)        AUD 199 (Nyngan, Australia, PV)
(3)        EUR 547 (Olmedilla, Spain, PV)
(3)        EUR 205 (Andasol I, Spain, trough)
(4)        AUD 257 (Greenough, Australia, PV)
(5)        AUD 432 (Solar Oasis, Australia, dish)
(6)        USD 177 (Lazio, Italy, PV)
(7)        AUD 295 (Kogan Creek, Australia, CLFR pre-heat)
(8)        USD 248 (New Mexico, CdTe thin film PV)
(9)        EUR 218 (Ibersol, Spain, trough)
(10)      USD 251 (Ivanpah, California, tower)
(11)      CAD 445 (Stardale, Canada, PV)
(12)      USD 315 (Blythe, California, trough)
(13)      AUD 310 (Solar Dawn, Australia, CLFR)
(14)      AUD 286 (Moree Solar Farm, Australia, single-axis PV)
(15)      EUR 381 (Lieberose, Germany, thin-film PV)

I can also estimate the cost of CO2 abatement for the Lieberose project.  That is (0.094+0.030) × EUR 160 × 10^6 / 35,000 t CO2 = EUR 567 / t CO2.  Quite expensive abatement!

Another metric I can calculate is the Capacity Factor, which is 52,000 / (52.8 × 24 × 365) = 0.11.  That number seems not unreasonable for a plant with fixed PV modules in a northerly location.

I’ve learned a lot in preparing this post, particularly from the article by Tom Cheyney.   I’d also comment that the Lieberose output is expensive, which I put down to the facts that the modules are fixed and the location is quite northerly.  Finally, I note that the project information gives the LEC as EUR 319/MWhr, which is rather less than what is predicted using my standard assumptions.

Monday, June 20, 2011

Cost of solar power (14)

On 15 June 2011, the first two winners in the Australia federal governments Solar Flagships program were announced.  In my last post, I described the program briefly and analysed the winning solar thermal syndicate, Solar Dawn.  Today I’ll consider the winning PV syndicate – the Moree Solar Farm.

There are three members in the consortium behind this project:  Fotowatio Renewable Ventures (majority owner, Spanish-based developer), BP Solar (manufacturer of solar cells) and Pacific Hydro (Australian-based renewable energy investor).

The Moree Solar Farm will have an output of 150 MW peak from 645,000 multi-crystalline solar panels with single-axis tracking.  The whole site will occupy approximately 1,200 Ha.  The reported cost of the project is AUD 923 million, with funding of AUD 306.5 from the Australian federal government and AUD 120 million from the New South Wales state government.  Final planning approval is imminent and construction is expected to start in mid 2012.

As with the Solar Dawn project, the consortium seems shy about disclosing the annual output.  But an educated guess can be made in two ways as follows.

(1)  Press releases for the project claim a CO2 emissions saving of 400,000 tonnes.  Now most Australian electricity is generated from coal, and I’d expect that to be the basis for CO2 emissions savings estimates by the consortium.  The amount of CO2 emitted per MWhr varies according to the nature of the coal and the thermodynamic efficiency of the plant.  Typical emission rates are shown here.   For black coal, modern super-critical plants might be as low as 800 kg CO2 per MWhr, whereas a typical value for older large-scale plants would be 1,000 kg CO2 per MWhr.  I expect the consortium would choose the latter estimate since that is most favourable to them.  Therefore the estimated annual output would be 400,000 t CO2 / 1 t CO2 per MWhr = 400,000 MWhr.

(2)  If I use the estimate in my last post that the Capacity Factor for solar projects is in the range 22-24% (let’s make that 23%), then the annual output will be 0.23 × 150 × 365 × 24 = 302,220 MWhr.

Of those two estimates, I think the first is the more reliable since the 400,000 t CO2 figure was widely used in all press reports and the CO2 intensity of coal-fired power stations is well known.  So, let’s assume the annual output is 400,000 MWhr = 400 GWhr.

I now evaluate the Levelised Electricity Cost (LEC) using my customary 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 (3% of the total project cost), and
          government subsidies are neglected.
(For further commentary on my LEC methodology, see posts on 2011-04-23, 2011-04-27 and 2011-05-21.)

The results are:

Cost per peak Watt AUD 6.15/Wp
LEC                            AUD 286/MWhr

The components of the LEC are:
Capital           {0.094× AUD 923 ×10^6}/{400,000 MWhr} = AUD 217/MWhr
O&M              {0.030× AUD 923 ×10^6}/{400,000 MWhr} = AUD 69/MWhr

By way of comparison, LEC figures (in appropriate currency per MWhr) for all projects I’ve investigated are given below.  The number in brackets is the reference to the blog post, all of which appear with the title “Cost of solar power ([number])”:

(2)        AUD 199 (Nyngan, Australia, PV)
(3)        EUR 547 (Olmedilla, Spain, PV)
(3)        EUR 205 (Andasol I, Spain, trough)
(4)        AUD 257 (Greenough, Australia, PV)
(5)        AUD 432 (Whyalla, Australia, dish)
(6)        USD 177 (Lazio, Italy, PV)
(7)        AUD 295 (Kogan Creek, Australia, CLFR pre-heat)
(8)        USD 248 (New Mexico, CdTe thin film PV)
(9)        EUR 218 (Ibersol, Spain, trough)
(10)      USD 251 (Ivanpah, California, tower)
(11)      CAD 445 (Stardale, Canada, PV)
(12)      USD 315 (Blythe, California, trough)
(13)      AUD 310 (Solar Dawn, Australia, CLFR)
(14)      AUD 286 (Moree Solar Farm, Australia, single-axis PV)

I can also estimate the cost of CO2 abatement for the Moree Solar Farm.  That is (0.094+0.030) × AUD 923 × 10^6 / 400,000 t CO2 = AUD 286 / t CO2.

My general conclusion is that the LEC for Solar Dawn and the Moree Solar Farm are approximately the same (subject to due caution associated with all the assumptions I’ve made).  Moreover, at current exchange rates, the LEC for both projects is broadly comparable with other large projects I have analysed recently, with Ivanpah currently having the best figures.

Sunday, June 19, 2011

Cost of solar power (13)

Over the weekend, the Australian federal government announced two winners under its Solar Flagships funding program.  This post contains an analysis of the Levelised Electricity Cost for the Solar Dawn project; my next post will investigate the other winner (Moree Solar Farm).

The Solar Flagships program, as originally announced in 2009 with AUD 1.5 billion funding, had the intention to support more than 1 GW of installed solar generation capacity in four separate projects in Australia, two solar thermal and two PV.   In the 2011 federal budget, funds available over the forward estimates period were reduced by AUD 250 million.  Even in reduced form, the program represented a very big pot of money and there was vigorous competition with 44 original applicants and a subsequent shortlist of eight.

The Solar Dawn project is a hybrid between solar thermal and gas, with the gas component restricted to 15% of the total output.  The project uses the Compact Linear Fresnel Reflector (CLFR) technology devised by David Mills at the University of Sydney and now under the control of the French multinational company Areva.  This technology was helped to full commercial status by venture capitalists in California.  Wikipedia summarises the history of the technology.

It is of interest that no tower technology (such as Ivanpah in California and Abengoa in Spain) made the solar thermal shortlist.  The CLFR technology also triumphed over applicants based on parabolic trough technologies.  Climate Spectator today published a very interesting interview with Anthony Wiseman, Areva’s Regional Director and General Manager of Areva Renewables, who discusses technicalities associated with the CLFR concept.  Here are some quotations from that interview:

“... the attractiveness of CLFR technology is that it’s direct steam generation, so it doesn’t require that balance of plant or heat exchanger equipment, it doesn’t have the environmental hazards of dealing with a thermal oil circuit and consequently it has a capital cost advantage.

... our temperatures are in excess of what is achievable by a parabolic trough plant. By producing super heated steam at such a temperature and pressure, we increase the efficiency …, which improves the project economics.”

To achieve consistency of power output, the Solar Dawn project opted in favour of a hybrid solar-gas system instead of thermal storage.  According to Wiseman in the Climate Spectator interview:

... we have a research and development group that continues to pursue storage solutions, but at the current day it’s not a commercial proposition.

“The plant will be able to independently operate on solar, or it will be able to independently operate on gas. The 15 per cent restriction is from the guidelines under the Solar Flagships Program related to the annual energy that was dispatched from the plant. It’s not an instantaneous calculation. It’s a yearly calculation.

... Solar thermal in general has a sort of 22, 24 per cent capacity factor.”

Overall, the project has 250 MW capacity provided through two 125 MW steam turbines.  The site area will be 450 Ha.  The annual power output seems to be a secret, but it can be estimated from Wiseman’s comment above that the Capacity Factor due to solar would be in the range 22-24%.  If we take 23%, then the annual power output would be 0.23 × 250 × 365 × 24 = 503,700 MWhr = 503.7 GWhr.  The cost of the project has not been finalised, but the figure mentioned publicly is AUD 1.2 billion, of which AUD 464 million will come from the Solar Flagships program and AUD 75 million from the Queensland government. 

Note added: 23 July 2011

A month ago, I sent an e-mail enquiry to Solar Dawn asking for their estimate of power produced.  I received an answer yesterday.  The annual output is estimated to be 480 GWhr, and I have updated the calculations accordingly.

The Solar Dawn project is near Chinchilla in Queensland, and is adjacent to and will be operated by the same parties as the Kogan Creek project – see Cost of solar power (7).

I now evaluate the Levelised Electricity Cost (LEC) using my customary 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 (3% of the total project cost), and
          government subsidies are neglected.
(For further commentary on my LEC methodology, see posts on 2011-04-23, 2011-04-27 and 2011-05-21.)

The results are:

Cost per peak Watt AUD 4.80/Wp
LEC                            AUD 310/MWhr

The components of the LEC are:
Capital           {0.094× AUD 1.2 ×10^9}/{480000 MWhr} = AUD 235/MWhr
O&M              {0.030× AUD 1.2 ×10^9}/{480000 MWhr} = AUD 75/MWhr

By way of comparison, LEC figures for all projects I’ve investigated are given below.  The number in brackets is the reference to the blog post, all of which appear with the title “Cost of solar power ([number])”:

(2): AUD 199/MWhr (Nyngan, Australia, PV)
(3): EUR 547/MWhr (Olmedilla, Spain, PV)
(3): EUR 205/MWhr (Andasol I, Spain, trough)
(4): AUD 257/MWhr (Greenough, Australia, PV)
(5): AUD 432/MWhr (Whyalla, Australia, dish)
(6): USD 177/MWhr (Lazio, Italy, PV)
(7): AUD 295/MWhr (Kogan Creek, Australia, CLFR pre-heat)
(8): USD 248/MWhr (New Mexico, CdTe thin film PV)
(9): EUR 218/MWhr (Ibersol, Spain, trough)
(10): USD 251/MWhr (Ivanpah, California, tower)
(11): CAD 445/MWhr (Stardale, Canada, PV)
(12): USD 315/MWhr (Blythe, California, trough)
(13): AUD 310/MWhr (Solar Dawn, Queensland, CLFR)

At the current approximate conversion rate of AUD 1 = USD 0.95, the LEC for Solar Dawn is comparable to that for Blythe (Cost of solar power (12)) and around 25% more than that for Ivanpah (Cost of solar power (10)).   Note however that the Aussie dollar is volatile and was around USD 0.60 as little as three years ago. 

Also, the estimated LEC for Solar Dawn will be on the high side since the capital cost includes hardware for gas-fired steam generation, whereas the output is only for the solar component.

Tuesday, June 14, 2011

Cost of solar power (12)

Recently I’ve had a flurry of German visitors to my blog.  They come from a finance discussion group, and are particularly interested in activities of the German company Solar Millennium.  Solar Millennium AG, headquartered in Erlangen, builds solar thermal power stations such as Andasol and Ibersol in Spain, and I have previously estimated the Levelised Electricity Cost for these installations (see posts Cost of solar power (3) and Cost of solar power (9)).

Solar Millenium has a joint venture with another German company, Ferrostaal AG in the USA.  The JV, called the Solar Trust of America, is based in Oakland and plans to build a very large solar thermal plant in Blythe in the extreme south-east of California.

Wikipedia reports that the following approvals are in place for this project:
·         California Energy Commission (approved September 2010)
·         Bureau of Land Management (approved October 2010)
·         US Department of Energy USD 2.1 billion loan guarantee (approved April 2011)

The San Francisco Chronicle reported on 19 April 2011 that preliminary construction began last year.  In the same article is reported that

“Chevron Energy Solutions, a division of the San Ramon Oil Giant, helped develop the Blythe project but will not oversee construction”.

In the German discussion group, contributor ulm000 (post #8614) reports that the Chevron connection is as follows:

“Da Chevron ein kleines Grundstück hatte auf dem SM das Blythe CSP-Kraftwerk baut, wurde Chevron auch in allen Dokumenten zusammen mit SM erwähnt ("development agreement").   Das steht auch genau so in den offiziellen Dokumenten der BLM.  Der Entwickler von Blythe ist die Solar Trust of America und sonst niemand und die STA gehört SM und Ferrostaal.”

In English …

“Chevron had a small piece of land on which Blythe is being built, and that’s why Chevron appears together with SM in all the documents. … That’s exactly how it appears in the official documents.  The developer of Blythe is the Solar Trust of America and no-one else, and the STA belongs to SMA and Ferrostaal.”

From a technical standpoint, the Blythe project is conventional.  The sun’s energy is collected in parabolic troughs and used to raise steam to power Rankine-cycle steam turbines with air-cooled condensers.  According to the Wikipedia article, when complete the Blythe project will consist of four separate installations that will produce 968 MW peak and 2,200 GWhr per year.  The collectors will occupy 28.43 km^2, and the claimed CO2 emissions savings will be 884,000 short tons (approx 802,000 tonnes) per year.  There was no mention of thermal storage in any of the documents I saw.

The first half of the project is estimated to cost USD 2.8 billion.  The estimated cost for the full project is stated by Wikipedia to be USD 6 billion, but that would be an imprecise estimate since the full project is expected to take six years to develop.  I’ll make my cost estimates on the first half of the project, namely 484 MW, 1,100 GWhr/yr, USD 2.8 billion, 401,000 tonnes CO2 abatement per year.

I now evaluate the Levelised Electricity Cost (LEC) using my customary 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 (3% of the total project cost), and
          government subsidies are neglected.
(For further commentary on my LEC methodology, see posts on 2011-04-23, 2011-04-27 and 2011-05-21.)

The results are:

Cost per peak Watt USD 5.79/Wp
LEC                            USD 315/MWhr

The components of the LEC are:
Capital           {0.094× USD 2.8 ×10^9}/{1.1 ×10^6 MWhr} = USD 239/MWhr
O&M              {0.030× USD 2.8 ×10^9}/{1.1 ×10^6 MWhr} = USD 76/MWhr

The cost of CO2 abatement is {0.094+0.03} × USD 2.8 × 10^9/4.01× 10^5 = USD 866/t CO2.  That represents expensive CO2 abatement, although, to be fair, the attraction of building such solar power stations is more than just CO2 abatement in the short term; there is also the development of industrial capacity that will be needed in the post-Carbon economy of the future.

By way of comparison, here are LEC figures for all projects I’ve investigated:

Cost of solar power (2): AUD 199/MWhr (Nyngan, Australia, PV)
Cost of solar power (3): EUR 547/MWhr (Olmedilla, Spain, PV)
Cost of solar power (3): EUR 205/MWhr (Andasol I, Spain, trough)
Cost of solar power (4): AUD 257/MWhr (Greenough, Australia, PV)
Cost of solar power (5): AUD 432/MWhr (Whyalla, Australia, dish)
Cost of solar power (6): USD 177/MWhr (Lazio, Italy, PV)
Cost of solar power (7): AUD 295/MWhr (Kogan Creek, Australia, CLFR pre-heat)
Cost of solar power (8): USD 248/MWhr (New Mexico, CdTe thin film PV)
Cost of solar power (9): EUR 218/MWhr (Ibersol, Spain, trough)
Cost of solar power (10): USD 251/MWhr (Ivanpah, California, tower)
Cost of solar power (11): CAD 445/MWhr (Stardale, Canada, PV)
Cost of solar power (12): USD 315/MWhr (Blythe, California, trough)

On these numbers, the LEC for the Blythe project is 25% higher than for another big US project currently under construction – Ivanpah, see Cost of solar power (10).  After allowing for currency conversions at the current rate, USD 1 =EUR 0.691, the LEC for Blythe is very similar to that for Andasol and Ibersol.

Sunday, June 12, 2011

Savings in CO2 emissions (ECET)

Today I shall present an estimate of CO2 emissions savings that would be achievable if my ECET concept were to be introduced.  In marketing-speak, these savings would be described as low-hanging fruit.

ECET is the acronym for the Expansion-Cycle Evaporation Turbine, the continuous-flow version of a new thermodynamic cycle I invented in 2004.  In operation, the ECET expands hot air, evaporatively cools it at sub-atmospheric pressure and then re-compresses it to ambient pressure.  Several cooling and re-compression stages are advantageous.  Power is received in expansion and expended in re-compression (as well as in water purification and injection), and there is surplus work available in the cycle provided the air is hot enough and the adiabatic efficiencies of the turbo-expanders and compressors are good enough.

For further details, please visit www.sunoba.com.au and follow the link to “Expansion-Cycle Evaporation Turbine”.  At that web page, I describe a case study in which the ECET is used to boost the power of an Open-Cycle Gas Turbine (OCGT) by more than 20%, without using any extra fuel, and at a specific capital cost ($/MW) expected to be no more than that for the upstream OCGT that provides the hot exhaust.  The ECET boost is employed when the OCGT is operated, namely at peak demand in the electricity grid.

Let’s proceed with the estimate …

In 2008-09, the total electricity production in Australia was 266 TWhr or 9.6 ×10^17 J [1].  I don’t have data saying how much of this was produced by OCGTs, but let me estimate 5%, giving 4.8 × 10^16 J.  In the case study referred to above, the ECET boosts the power of an OCGT by more than 20%, so it is argued that 0.2/1.2 = 0.17 of the electricity currently produced by OCGTs could be produced by ECET boost.

The specific cost ($/MW) of ECET hardware required to produce that electricity is reckoned to be no more than the specific cost for the hardware for the upstream OCGT.  If the total amount of generation capacity need not be altered, neither would be the total capital cost.

That gives a plausible Australian ECET electricity production of 0.17 × 4.8 × 10^16 = 8.2 × 10^15 J.  That represents electricity that does not have to be produced by OCGTs, say at an average thermodynamic efficiency of 0.36.  Therefore the primary energy (as natural gas) saved is 8.2 × 10^15 J / 0.36 = 2.3 × 10^16 J.

Note added 20 July 2011.  In what follows, I now think I need to use the Higher Heating Value for the energy content of natural gas, rather than the Lower Heating Value.  Appropriate changes, which have a relatively minor effect, have been made and are indicated in red.

Now natural gas has energy content (Lower Heating Value) of 50.0 × 10^6 J /kg [2], so the amount of natural gas saved is 2.3 × 10^16 J / 55.5 × 10^6 J /kg = 4.1 × 10^8 kg.

Each kg of natural gas produces 2.75 kg of CO2 [3], so the emissions savings are 2.75 × 4.1 × 10^8 kg = 1.1 × 10^9 kg CO2 = 1.1 million tonnes CO2.

If the price of gas is AUD 4 / GJ [4], then the value of the natural gas saved is AUD 4 × 2.3 × 10^7 = AUD 92 million.

Each tonne of CO2 that is avoided is therefore associated with a saving, not a cost, of AUD 92 million / 1.1 million tonnes CO2 = AUD 81 / tonne CO2.


The following table gives a summary of the estimates for Australia:

Primary energy savings per year
2.3 × 10^16 J
Savings in natural gas per year
0.41 million tonnes
Savings in CO2 emissions per year
1.1 million tonnes [5]
Value of natural gas saved per year
AUD 92 million
Benefit of CO2 emissions reduction
AUD 81 / tonne CO2 [6]
To obtain corresponding figures for the entire world, a rule of thumb is that Australia represents around 1% of the world market.  Simply multiply the above estimates (except for the benefit) by 100 to get a global estimate.

I think these estimates justify my use of the term “low-hanging fruit”.  (Even if my unsubstantiated estimate of 5% for the amount of electricity produced by OCGTs is wrong by a factor of 5, the estimates would still be impressive.)

References

[5] this is 0.24 % of Australia’s annual CO2 emissions, see http://www.climatechange.gov.au/climate-change/emissions.aspx
[6] This is a benefit, not a cost; in an important recent publication, the Australian Productivity Commission gives implicit costs for CO2 abatement for various countries and technologies; see Figure 2 on p. XXXI , Productivity Commission 2011, Carbon Emission Policies in Key Economies, Research Report, Canberra.  Abatement costs for large-scale renewable such as wind are around AUD 40-60/t CO2.  Most solar projects have abatement costs above AUD 200/t CO2.