First Dog socks now in stock at the Crikey Emporium!
A cautionary tale…
What is solar power?
|
The Carbon Pollution Reduction Scheme is on the table, with concessions flying thick and fast to some of Australia’s biggest polluters. But to many, the debate should be looking towards the future — and renewable energy. Solar power is often the most talked-about possibility. After all, we do have all this sun. But could Australia actually be 100% powered by it? Evan Beaver takes a look. So first things first, what is solar power? Broadly speaking, it is any energy derived directly from the sun. This includes direct-heat uses such as the hot-water systems people are familiar with, but also includes drying applications such as salt production and sludge drying in sewage treatment plants. This little robot has what I consider the worst job in the world, and will be the first to go if the robots ever form a union. In the developing world, solar power, particularly the UV radiation associated with it, is widely used for sterilising water. Just fill a PET bottle with water and leave it in the sun for eight hours. Voila! Sterile water. But the big-ticket item is electricity production. What are the technologies competing to produce electricity from solar power? There are two distinct categories, with a few different technologies within each. Photovoltaic (PV); literally electricity from light. This is becoming more common on roofs and street lights around Australia, and has been popular for small-scale RAPS (remote area power supplies) for many years. Solar thermal; concentrating the heat of the sun to make steam, then using this steam to drive a turbine as in coal-fired power generation. Which is the best? That’s a tough question. They each have advantages and disadvantages and the “best” will depend on regional variations, resource availability and relative costs. PV has the advantage of very simple operation. Put it in the sun and electricity comes out. That’s it. No need for moving parts, maintenance crews or sophisticated tracking mechanisms. There’s more efficiency to be gained by adding tracking, but it’s not really necessary. On the other hand, they don’t generate much electricity per unit area compared to the other technologies, and at grid scale, tracking becomes mandatory. I’ve somewhat unkindly heard them described as “23% efficient; 77% dicking around”. But, efficiencies are improving in labs across the world. One company is claiming that it is on track for cost parity with coal-fired power by 2010. The Institute of Electrical and Electronics Engineers think they’re being optimistic. Also, because they are a direct-light electricity conversion, once the sun stops hitting it, say if a cloud or very large bird passes over, the electricity production stops instantly as well. Further, within the PV category, there are some more sophisticated technologies. Such as concentrating solar PV, so called ‘thin film’ (like Origin Energy’s Sliver technology) and flexible or organic PV. The concentrating PV uses a magnifying glass or curved mirror to focus the sunlight onto a very sophisticated and small PV cell. The goal here is twofold; minimise the amount of silicon (which is expensive) and concentrate the sun, which brings efficiency benefits. This last point is true of all technologies. Thin-film technology pushes the boundaries of the “minimise the silicon” aspect, by making slivers of photovoltaic material to minimise cost. Organic solar is cutting edge; again, they try to minimise production costs, but instead of silicon they use organic chemistry. The drawback is embarrassingly poor comparative efficiency numbers (~5% vs. ~25% for silicon PV). However, research continues because the costs per unit electricity are getting close to silicon PV. Organic PV is also useful because they can make fancy flexible films that allow some interesting military applications. Solar thermal
Thermal in general makes use of higher temperatures and better energy reclamation over a unit area. It is also much easier to incorporate storage into a thermal plant. There are a bunch of different methods of doing this, broken down broadly into linear (troughs, Fresnel), arrays (tower systems, including Lloyd Energy’s excellent carbon block storage) and parabolic dishes (such as this Stirling engine, one of which has the highest efficiency measured at ~40%, or Wizard Power) Most of the bigger installations worldwide are solar thermal, usually using trough-based reflectors. Sounds expensive. It is, but only compared to coal power, which is really cheap. Solar thermal exploits subtle energy gradients by concentrating solar energy. Coal power comes from digging a hole and burning what you find in the hole. As with all of these sorts of financial analyses, they vary greatly, depending on the starting assumptions. Here and here, you can see what I’m talking about. As a coarse summary, solar is somewhere in the range of 50-100% more expensive than coal (without capture and storage) and pretty close to nuclear projections. Translating to a carbon cost, solar would need at least $20 per tonne to be competitive. OK, there’s lots of technology. But does it actually work in any installations outside quirky labs, research organisations and hippy communes?
Currently, the US, Spain, Portugal and Germany are leading the solar charge. In the US, the SEGS 1 plant has been operating, and generating useful electricity since 1984. This was the first stage of the massive solar energy generating system in the Mojave Desert. This collection of plants (there are nine) now produce about 350MW; typical coal plants are 500-1000MW. In recent years, favourable government policies in Spain have led to an explosion in solar power projects. In 2008 alone, more than 3000MW were installed, with another 1300MW announced. Government support, through ‘feed-in-tariffs’, has changed the economics considerably for solar, so much so the government has created a maximum project size (50MW) to prevent a budget blow-out. What about baseload? Won’t we need electricity at night and on cloudy days as well? Leaving aside objections to the term ‘baseload’, there are a couple of ways of skinning this particular cat. What’s needed is energy storage; either as heat or as electricity. Then, during the times when electricity demand is less than electricity supply, we can store some for use later. This technology already exists and is part of a normal electricity grid. But, the energy storage needs are likely to increase dramatically with the introduction of more renewables. Energy can be stored as electricity in a number of ways. For grid scale storage they’re mostly different types of advanced batteries. These can be deployed anywhere in the grid. There is also ‘pumped-hydro’, good for large amounts of energy, but geographically difficult to build. Storing energy directly as heat is a little easier and well suited to solar thermal projects. Even a “standard” solar thermal plant has about an hour of storage, due to the mass of hot material (steam)in the system. This is enough to ride through cloud cover, which can be a big problem for PV. There are plants in Spain with 7.5 hours of storage, (essentially just steam in a boiler) which is just about enough to get them all the way through the night. This amount of storage increases project costs by about a third, putting them in the upper ranges of solar costs. But this has benefits for the generating company because more reliable electricity production is more valuable to electricity retailing companies. Heat energy storage has been an area with a lot of research directed at it recently and the results are starting to improve. Lloyd Energy has an innovative storage system that stores heat directly in 10-tonne blocks of graphite; this provides “on-demand” solar. Wizard Power stores heat by dissociating ammonia in a reverse of the industrial classic Haber Bosch process. I’m a big fan of this project for two reasons; it’s being built in Canberra and the dish they use as a test bed could easily be modified into a death ray (I’ve seen it burn a hole in a ceramic heat shrouds like it was butter). Can we power the whole country with the sun? The answer is probably yes, but there are some risks and some serious costs associated. Professor Keith Lovegrove, a solar researcher from ANU, is fond of saying that an area about 150 x 150km (from this presentation), covered in a solar thermal power plant could power the whole country, even including storage to get through the night and average conversion efficiencies. This is approximately the same size as greater Sydney. But, if it’s raining there for a few days, we’re stuffed. Most reports discuss the advantages of geographically spaced plants, to spread the weather chances, and a mix of generation technologies. These would include natural gas turbines that can peak as demand ebbs and flows; wind turbines and a smart grid. A mix of wind and solar offers synergies when Mother Nature is playing games with us; also high pressure over Australia means no wind, but lots of sun. Low pressure systems bring wind and rain, but no sun. Perfect! Enough with the tech nonsense and caveats. Is it the final solution or not? Using more evasive language, hardly anyone would advocate one generation technology and one only. Smart energy investors will realise there are little regional variations that push the favourability of a project back and forth depending on the resources available. With that in mind, Australia has a lot going for it with solar thermal. Plenty of sun, bugger-all rain and sunny places where the grid already exists and people are living. If it’s not viable here, it’s probably not viable anywhere. Apologies for the Wikipedia links, but for this sort of general information it’s perfect, albeit a little out of date on some specifics. Evan Beaver is an engineer, recent chook owner and convicted public servant. |
|
|
53 Comments
Nice summary there Evan.
Feeling the sun’s burning rays on my head at lunchtime and retreating back into the sanctuary of the “heaps of power needed to remove the sun’s energy” air-conditioning in the office you can see how much energy there is to be had from the sun.
Hi Evan. Solar power will almost certainly have to be a significant component of our future energy supply.
Do you have any figures on the ground area required to generate power in the American / Spanish plants? What sort of protection do those plants place around their capture devices, e.g. do they need to guard them from terrorists, or chase away birds from nesting on top of them, etc?
Do you have any figures on the ground area required to generate power in the American / Spanish plants?
Nope, haven’t seen any, though their mostly parabolic troughs, so using averages per GW is probably not unreasonable. The whole thing is so ad hoc in Spain that I wouldn’t place much credit in the numbers anyway. Apparently the market really went crazy with the introduction of the tarrif and projects sprung up all over the place. Particularly PV in the 20-50kW range, where every farmer and their dog wants a piece of the action .
What sort of protection do those plants place around their capture devices, e.g. do they need to guard them from terrorists, or chase away birds from nesting on top of them, etc?
Birds aren’t much of a problem, the mirrors are pretty unpleasant for most birds, but dust is a problem. One of the solar tower installations (Abengoa I think?) has a tractor with a big rotary brush on it that is constantly cleaning the mirrors. Reflector efficiency drops off fast with dust.
Lots of this info is based on colleague’s photos from a recent trip. Most of the installations looked like any other industrial area; 6ft cyclone wire fencing with a strand of barbed wire along the top. I guess it is an advantage of this sort of distributed generation; even if someone can take out a whole plant, you’re only going to lose 50-100MW.
I believe the wikipedia entry for the American SEGS plants has a lot of detail on the areas and technology being used in each of the stages of the plant.
Good summary. I couldn’t have resisted pointing out that world-leading technology for both solar-PV and solar-thermal were developed in Australia but were forced offshore (to China and California, respectively) by indifference of both the Howard government and the investment community (who are always more interested in holes in the ground). Expat-Australian Dr. Zhengrong Shi, CEO of the world’s largest solar-PV company, Suntech, is reputedly one of China’s richest men.
Also I believe that solar-HW is underestimated in the contribution it can make. It could take a huge load off the grid, is already economic and deserves to be made mandatory for all Australian houses.
Another solar technology that deserves a mention is solar air-conditioning. Not using electricity from rooftop solar-PV which is way too inefficient, but where solar thermal collectors provide thermal energy to drive thermally-driven chillers. Here is the wiki reference.
en.wikipedia.org/wiki/Solar_air_conditioning
en.wikipedia.org/wiki/Absorption_heat_pump
Currently it is being prototyped on large installations and I believe there was a competition last year in Australia for an appropriate test site (schools, hospitals etc). These systems can also provide HW as a side benefit. Obviously cloud or intermittency is irrelevant since it is perfectly matched to the need, and also the geography. It seems like it is about a decade away from commercial rollout and maybe longer for domestic scale. Mostly German, Japanese and American but I believe CSIRO also are researching it.
Yeah, I’ve been wondering about solar cooling. Saw a few interesting waste heat chillers back at Sydney Water that were deferring huge loads. Interesting stuff.
For solar air con/refrigeration, I’m on the cusp of an ill advised project at home. Remember the ammonia cycle refrigerators from the 70s and 80s? I’ve got an old one lying around at home. All it needs is a heat source at one point, and it makes the other end cold. No moving parts, only needs a heat input. If I can figure out how to install a solar heat supply I’m building a huge beer fridge.
What? Ignoring comparisons with the cheapest options provided by your on links?
I do appreciate the links, which appear to be close to the mark in most respects. However, your anti-nuclear stance is showing, to the point of irrationality.
It is not reasonable to ignore nuclear without even a comment.
Goodness, there is a wide range of possible comments to make, including:
* Legally prohibited by federal legislation;
* Has poor public image which could cripple its chances;
* Cheapest but with unresolved longer term site and wast management options;
and so forth.
But to not comment at all!!!
I have reluctantly come to realise that the probable best path will include a mix of nuclear for the foreseeable future, because this is possible as a replacement option where existin infrastructure exists but relies on black or (worse) brown coal. A coal fired power station will already have access to cooling water, transmission connection, land appropriately zoned, and so forth. It is attractive as an alternative site utilisation option for coal fired generators. For example, Wallerawang in NSW or Hazelwood in Victoria. Appropraitely sized packages in these two locations alone could replace 15GWH of coal (7 plus million tonnes of coal) within a decade if the political willand the finance were available.
Yet the author has allowed his prejudice to cloud reality, again.
Nuclear
@JohnBennetts - This thread is specifically about solar power. It’s not about ‘options’ or anything else. You’ll note that it is also not about wind, geothermal, coal, gas, or magic muffin power. It is entirely reasonable to ignore nuclear in this thread.
Evan,
First try a troughed parabolic mirror with a black steel heating tube at the focal point. No need for dynamic aiming - just prop it at the correct gradient on the roof.
I suspect that a reflector consisting of strips of ordinary glass mirror, say 100mm wide and 1000 long glued to curved galvanised backing sheet and mounted in timber troughs (plywood?) will be reasonably cheap. The idea is to run the water/steam in a closed circuit incorporating a tank to avoid needing to top-up and blowdown all the time, operating on the thermocycle principle.
Alternatively, a small PV collector to run a 12V hot water pump to circulate the heated water. Simple and elegant. Unused power could be fed to an electric fence energiser to ward off would be beer thieves. Beware! 10kV Fridge!
There will be variation on cloudy and rainy days, but it should fly on a hot summer afternoon!
The indoor fridge (negotiate with the domestic goddess) can be pressed into service for extended poor weather. Or, just get used to warmer beer. Dark brews are often much better only a few degrees below room temp.
Kirk, if this thread is not aboutanything except the three solar power options mentioned, then why bother with comparisons, including links? And why exclude the cheapest option in those comparisons, which are all about cost?
Evan a solar powered beer fridge is perfect since the hotter the sun, the colder the beer.
Solar PV is still unbelievably expensive for the energy output. Despite massive government subsidies and feed-in tariffs it is an upper middle class indulgence. However from a technology point of view, eliminating the need to have large boilers and turbines and moving parts is crucial for a very small scale technology. The advantage for solar PV is that it is distributed and therefore doesn’t impose large grid requirements. The problem is that people install a 1.5 kW panel at a cost of over $10K, closer to $13K usually I think. Large air conditioners use over 5 kW. So to power that plus a fridge and any other appliances during peak demand you need more like 6 kW of panels or an investment of more than $40K.
Solar thermal looks much more promising. But remember that winter peak demand occurs at about 7-8 a.m. in many states and the 6pm. For the morning peak solar thermal is not going to make a real contribution since its been dark for 12+ hours.
Please excuse my pedantry, but I believe that coal is technically a form of solar power as it is created by the compaction of organic material generated by photosynthesis. Irrelevant to the discussion here, I just had to get it off my chest.
John Bennetts, is there anyone you haven’t called “irrational” yet?
Excellent guide, Evan. Thanks.
Chris, coal is just a long life-cycle renewable. Nuclear probably is too if you hang around long enough.
Absolutely, I just enjoy the irony when people say that solar could never replace coal. The opposite is true. And solar is produced from nuclear reactions in the sun, so I guess it is all nuclear.
Regards pumped storage: there are some in Australia e.g. this 500MW plant at Wivenhoe QLD which is described at URL http://www.seqwater.com.au/public/catch-store-treat/dams/wivenhoe-dam. This example doesn’t have much vertical displacement for the water so I understand this limited the practical maximum power output (not sure what the ideal vertical distance would be for uphill pumping AND downhill motor/generator sets).
FYI modular nuclear (allows for cheaper production of plant): see this IEEE Spectrum magazine article for some info on 125MW proposals, at URL http://spectrum.ieee.org/blog/energy/renewables/energywise/at-last-something-new-under-the-nuclear-sun.
EnergyPedant - “Solar PV is still unbelievably expensive for the energy output. Despite massive government subsidies and feed-in tariffs it is an upper middle class indulgence.” Most new products start as “upper middle class” indulgences. As demand grows and production increases, costs reduce due to economies of scale and the technology moves to the mass market.
My partner and I here in the ACT were looking at solar hot water (we currently use off-peak), evacuated tube as it works better in Canberra’s, ahem, challenging winters. Vs off-peak electricity pay-back time looked around 15 years. To my surprise PV was about the same payback period, ignoring feed-in tariffs. Add the ACT’s gross feed-in tariff and it came down to eight years. *PV ended up more justifiable as a pure investment to us than solar hot water*.
The point of gross feed-in tariffs (as per Germany) I understand it to kick-start the widespread adoption of domestic PV which will push it along the production/cost curve quicker than it otherwise would.
Evan - excellent overview; particularly interested in the solar-thermal stuff. “Nuclear probably is [renewable] too if you hang around long enough.” You’d have to hang around for a *really* long time! AFAIK all natural radioactive material on Earth is hanging around from the creation of the solar system, and just steadily running down its half-life. There’s no way new say U235 can be created during the existence of the current solar system. You’d have to wait for a convenient supernova to create some new stuff, like in about 5 billion years’ time
Why is the cost of Solar PV still UNBELIEVABLY expensive. You can buy them now so it is not hard to believe the current cost. Cars were very expensive once in terms of the average weekly wage, but they have come down a bit since them. Solar is expensive compared to candles, but so are plasma TV’s swimming pools Porches and yachts. Plenty of people still buy them though. Perhaps value is the issue here not cost.
The issue of cloud and rain is a red herring. If you put all your solar systems in the same place of course it’s an issue. Diversity of location can overcome this problem and not only that it gives you the opportunity to place the sytems closer to where the power is used. As an example the coal and gas fires power stations that provide power to Perth are in Western Australia. The ones supplying Queensland are mostly in Queensland. Surely this isn’t to difficult an intellectual bridge to cross.
@James McDonald, 5:29pm today:
James, I try to remain logical. I really do. If you have noticed me use the term “illogical”, it is not without reason.
Just out of interest I thought I would check Keith Lovegrove’s claim that it is possible to supply all of Australia’s electricity needs with a solar farm 150 km by 150 km.
Just to put that into context that is an area equal to 22,500 square kilometres, approximately the size of Wales. My back of an envelope calculation shows that he is outby about a factor of 6. Please feel free to point out any mistakes:
Australian electricity consumption per capita = 11,332 kWh (2006)
Population = 21,374,000 (2008)
Irradiance of sunlight at the equator in March or September at midday = 1000 W/m^2.
The latitude of Alice Springs is 23 degrees. Therefore in July the irradiance is 1000 W/m^2 but in January it is 620 W/m^2 (cosine 50). The average of this is 810 W/m^2.
If we assume that the efficiency of conversion from sunlight to electricity is 50% (realistic for thermal solar, photovoltaic are more like 20%) then that is 405 W/m^2.
Assuming that there are an average of 12 hours of sunlight per day 365 days a year (no cloudy days) therefore 6.4 m^2 would be required to supply the needs of each person.
If we multiply this by the population it turns out we would need a space equal to 136,794 square kilometres (370 x 370 km). That is 1.7% of the country or an area half the size of the UK.
Let’s remember this covers only the electricity needs of the country and we have assumed there are no clouds, no transport losses, no outage for maintenance and note that I have assumed it is midday for the entire time that the sun is shining not just 12 pm, so I would suggest this is still an optimistic estimate.
Thomas, there is also a need for maintenance roads - probably 15% or so of the site, because mirrors must be dust free to work OK.
Check out one site which discusses area at desertec-australia.org/content/concentratingsolarpower.html. dd the httpwww stuff at the front of the address.
Even if the area for Australia is doubled, the rough idea stays the same. Area used is not the problem, however cost, construction time, transmission and reliability are significant, potentially limiting issues.
Thanks Thomas, great contribution.
As you point out, of course you aren’t getting 100% sunlight all day.
Another mitigating factor is the impossibility of 100% use of the land. You would, of course, need plenty of walkways or other spacing for maintenance. Very difficult to estimate, but would be a significant percentage.
I question the final calculation. 6.4m^2 per person, ~21m people calculates to ~136m^2 total area required. This is 136km^2 - there are one million square metres to a square kilometre (1000m x 1000m = 1km x 1km). This is a much smaller area than the 150km x 150km estimated, so I assume that the estimation does take into account the above factors, and perhaps some more we haven’t considered.
This is a huge area if we park it into one massive facility in the desert, but have plenty of effectively ‘useless’ space sitting above our heads. The average household roof must be at least 100m^2, and if each person only needs 6.4m^2 then a household roof will supply enough power for almost 15 people. Maybe this is a bad idea as it doesn’t optimise the location for maximum sun, but if each person was somehow responsible for their own rooftop power generation it would be a beautiful thing.
@Kirk B: 9:55pm.
“…If each person was somehow responsible for their own rooftop power generation it would be a beautiful thing.”
Beautiful?
A couple of million rooftops cluttered with mechanical crud? Ugly, perhaps? Laborious? Grossly inefficient? Need something else at home that can go wrong and wreck your day?
There is a role for centralised, efficient power generation, regardless of the technologies employed.
Sorry, John. I didn’t realise you lived in that Colorbond world where you gaze lovingly at your roof anytime you set foot outside your house. I don’t spend too much time looking at my roof. I don’t care if it is ugly, and I don’t see why it needs to be ugly anyway - no worse than the solar hot water systems, tv antennae, horrible cables in the air all down my street, etc.
Installing solar power for the entire country is always going to be laborious.
It also seems somewhat efficient to generate power where it is used. A significant percentage of electricity is lost during transmission over distance. I don’t see how this idea is grossly inefficient. In terms of cost efficiency, instead of having a number of people maintaining panels in some remote yet centralised plant there would be roaming technicians on call to visit individual homes / power sites.
The point is that with some government subsidy, this could be taken up by individual homeowners. As cost decreases and takeup increases, the dependency on existing power stations is slowly reduced. The alternative? Find some huge area of lad, build a multi-billion dollar power station, wait a few years for it to come online. Maintain existing power stations while this happens, and then as soon as it is finished, turn them off. Does that sound efficient?
Of course, we can always look forward to privatising the power station in a decade - that would be great, wouldn’t it?
ps John - Consider why we have centralised power at the moment - we use dirty great plants that do not work on a micro scale, and no one wants near their house due to their toxicity.
The solar ‘farm’ that we are talking about is simply a collection of millions of ‘units’ that can be deployed individually. The reasons for centralisation do not apply.
Great article, thanks.
Kirk is getting off topic a bit, but I will follow him again.
There always seems to be a cry for government subsidy when the maths don’t work out, and solar PV or solar thermal, or a mix of thetwo will require enormous subsidies and cfross-linkage (sounds like a transmission system) if they are to provide reliability
The dream of obtaining adequate power for the nation from individual cottage roofs is just that - a dream, and a not very “beautiful” picture (your word, not mine).
I do agree up to a point, though - and that is in relation to incremental additions to the total system capacity. Managed correctly, the ADDITIONAL generating capacity may very well come from roofs, etc and small community installations. Hopefully, the existing power station ites will provide the basic sites for the larger installations and make use of the existing high and low voltage transmission assets.
Additional loads, especially industrial loads, may also be partnered with new small to medium renewable power generating capacity. That would indeed be beautiful, in many ways.
So, KB, we do probably share some objectives. Probably because I am an engineer, I keep a stronger eye on the costs and the physical task. You may not share my conviction that centralised power generation in some form will be needed for a long time to come.
I do hope that you find my contributions, in total, positive.
Thanks for the positive feedback everyone. Also note I don’t mind negative feedback either. Like something I’ve left out or not adequately explained.
Something I did leave out, partly because it’s a bit complex, but also because there’s not much info about, is the CSIRO syngas project.
Here’s some pics of the infrastructure
http://www.csiro.au/science/Solar-Thermal-Energy-Research.html
The idea is this. Use solar energy to split the hydrogen off methane (nat gas) to form H2 and (I think) ethyne (ethylene, or C2H2). This creates a ‘synthetic gas’ of higher calorific value than normal methane. This high energy gas can be combusted later through a normal gas turbine.
I like this one because of the problems it solves and options it leaves open. If we build gas peaking turbines in the coming years, which is likely and will add flexibility to the grid, this project feeds into it. It becomes a solar battery, which can be used very quickly. A kooky end point of the project could see some BIG towers at the Moombah gas fields, collecting solar power, juicing up the gas as it comes from the ground, then piping it into Adelaide using the existing pipeline. It’s probably years off, like maybe 10, but I can see some awesome possibilities.
RE Domestic solar/rooftop solar
I’m generally opposed to it, particularly from the viewpoint of efficient use of Government funds. There are some technical advantages to point of use generation, but also some disadvantages. I’m no network engineer, but imagine the polarity of a suburb essentially flipping over on a hot sunny day as a cloud passes over, flipping from an energy positive suburb to energy negative and back again.
But my principle objection is just that it’s an inefficient way to deliver the MW, and difficult to quantify in the long term. Under the old scheme, the government was getting 1kW/$8000. That’s a MW for $8,000,000, where ideally it should be closer to $5M/MW. Decentralised generation means an inverter for everyone, and they strongly benefit from economies of scale. Also, as the systems age, maintenance will vary greatly and inverters will break. If you get a system for free, I would argue you are less likely to maintain it properly.
There are some intangibles with rooftop solar though, particularly the ‘I’m part of the process now, so more involved’ aspect. If someone can see electricity generation in their life, they might take more care of it. However, this is negated a little by gross feed in tariffs. Because a household is paid regardless of how much they use there is no incentive to use less power.
All of that said though, I’m spending today organising my free PV install. At least Canberra has decent insolation.
Rooftop solar is great for direct heating of water, taking the load off mains. I think it’s amazing that Malcolm Street’s cost-benefit comparison of PV and direct optical heating came out even. Direct optical heating is simple and elegant, but has recently been perfected so the materials are still expensive and the economy of scale is not there. PV, on the other hand, has been around so long and in such quantities that if it’s not economical now it never will be.
Malcolm, I’m really surprised that solar PV would work out better than solar hot water. Although I normally look at CO2 mitigation per $. Off-peak electric hot water is super cheap (if environmentally appalling since it is all coal generated).
I tend to agree with Evan that solar PV is still currently just a plain expensive way to generate the MWs. From a pure efficiency point of view it doesn’t matter what the government subsidy is (since it is my tax money), solar PV is $8-13 Million per MW. Usually the capacity factor is maybe 25%. Compare that to $2-3 Million per MW of wind with a capacity factor of 35-40%.
Evan,
Nice summary. There is another aspect to large scale solar thermal. In Spain where there are a number of plants there has been a turnaround for local employment. its not just permanent mirror cleaners ( now off welfare benefits) but other staff who would not otherwise be employed.
A proposal put up by Desertec is to provide the EU with energy from across the “sun belt” of N.Africa. This hs some big players putting their names to it although no plant has been built yet.
Evan, thank you for your clarifier, it briefs readers on the current state of solar energy.
However you really must surrender that whimsy about ammonia! The reference to “Wizard Power” needs a link to an authoritative webpage that shows that dissociating and reforming ammonia is easy. I couldn’t find one, nor could I find a reputable reference to Wizard Power.
Instead the link you gave us shows that reforming ammonia is difficult and that the successful processes gave rise to two Nobel Prizes in Chemistry.
Please check your links, it would save your readers a wild goose chase, or worse, give your more credulous readers false hopes.
Wizard are using this technology, developed at ANU.
I didn’t say it was easy, and don’t think it will be. To my knowledge this is a world first (using the reverse of the Haber-Bosch process) and one of a couple of energy storage options being researched by ANU.
The Wizard Whyalla project will prove 2 things; mass production of the Big Dish and heat storage using ammonia. The Big Dish is built off a jig that was built originally at ANU in Canberra. This will test their on-site mass production ability and the accuracy of their jig. Part of the advantage of this parabolic dish is that they have been able to mass produce the individual mirrors identically, then achieve focussing accuracy through the design of the dish. It’s a nice, efficient design-for-manufacture project in my opinion.
I am still waiting, after all the discussion of more expensive options, for Evan’s analysis re nuclear, the cheapest option and based on well proven technology.
Even his own references indicate that this is so. Why do we spend time and space considering fermented pigeon poo and aardvark feather combustion, when there is another, proven, costed, option staring us in the face.
Come on, Evan. Display your biases right up front and centre.
Defend your determination to discuss everything except the elephant in the room.
The best commercially available solar thermal plants are Molten Salt Power Towers. These are from US Company Solar Reserve and Spanish Company SENER.
Youtube of Solar Reserve, Youtube of SENER Torresol Gemasolar
This technology was first proven in a prototype in the French Pyrenees at the thermis tower in 1978 and a commercial scale demonstration Solar II was proven by the US DOE at the Sandia Laboratories which are run in a Joint Venture with Lockheed Martin.
Solar Reserve is owned by UTC which owns Pratt and Whitney and Hamilton Sunstrand — this company built the Apollo space rockets. Those same engineers developed the technology behind the Solar Reserve towers.
Spain dabbled with Nuclear and lost 6 billion euro’s on power plants that were never commssioned. Now SENER one of the companies that built nuclear plants that did actually work is building molten salt power towers. Unlike the nuclear plants they built in the 1980’s they actually want to own the solar thermal plants.
Also with power towers 50% - 66% less molten salt (potassium nitrate / sodium nitrate) is used than trough power plants. And a molten salt power tower is cheaper with storage than without.
These plants run 75% capacity factor - the same as a NSW black coal plant.
Matthew, I think it’s physically impossible for a solar facility to have a CF over half.
My understanding of CF is that it is actual generation / maximum possible generation. There’s no way that can be greater than the daylight hours / total hours.
It might provide power over night, but not it’s maximum rated power. Might have availability that high, but surely not capacity factors.
John, if you’ve been concentrating you should have noticed 2 things. 1, this is an article about the current status of solar power world wide. It is a comparison to nothing. 2. I made my objections to nuclear pretty clear on the other thread. What do you hope to achieve by baiting me into listing them again?
Evan the capacity factor of Molten salt power towers is 75% please check the Sargent and Lundy due diligence like document on Sandia (Lockheed Martin) Sunlab’s Molten Salt Power Towers and I highly recommend the youtubes above.
Depending on the sun resource at a given location the plants can run 90% of the time throttling the turbine from 60%-100% output. They also have the ability for fast start (dispatchable) because they bleed off a small amount of the cold salt 290C to keep the seals on the turbine hot.
PAGE 126
Assessment of Parabolic
Trough and Power Tower Solar Technology Cost and Performance Forecasts
PAGE126
If you’d like to know more check out the Beyond Zero Emissions website http://beyondzeroemissions.org/
I still don’t get it. They must be using a different definition of capacity factor. For a solar thermal plant I would always compare actual output with nameplate size x the hours in a year. So, a 1MW plant should be able to make 8760MWh per year. But, the sun is only up for half of those hours, so maximum CF should be 0.5. It doesn’t matter if the plant has storage, because that just means that the sunlight wasn’t used straight away and was stored for use later.
Unless their mirror field is massively oversized for their turbine. In which case I still think it’s dishonest.
Evan,
The capacity factor is measured against the nameplate (or more precisely the net output minus plant parasites) of the plants turbine.
So the sun resource for a good location will be giving DNI output for 25% of the year. So in order to recieve a 75% capacity factor you basically triple your mirror field and add a bit more. So you basically spec for 2190+ hours of annual sun, run 1/3 of the mirrorfield for on sun operation. And store 2/3 of the field’s output in latent heat for operating 2/3 of the time when the sun has set and also in the vvery early morning and very late afternoon or early evening when the output of the mirror field falls below 1/3 of the recievers spec.
These plants also have the advantage that they can cycle the cold storage tank ~290C and increase the temperature of that early in the morning / late at night when it is too difficult to achiever the full 565C (or in future versions 650C).
Also Trough plants lose 7% of their collected heat in heat exchange losses from moving the field oil collected heat to salt storage then back to oil and then flashing to steam. Towers actually use molten salt as the working fluid and storage media so there are no heat exchange losses. In fact the battery is greater than 99% efficient. Losing less than 1% in thermal losses over a year of operation.
All these things trough plants cannot do — molten salt tower plants are far superior to trough plants.
Evan,
Solar thermal (including salt) can couple heat storage with a smaller turbine. Thus the nameplate rating is not the equivalent of the maximum insolation on the best, hottest, clearest day of the year.
Thus, the strategy is to harvest heat during the sunny bits of the day and to generate electrical power, via a slightly smaller and cheaper turbine, during as many hours as possible, based on system load and stored heat.
For example, the solar linear array in the Hunter Valley is not able to run flat out - it is limited by the steam capacity of the system, including the receiving boiler. This occasionally means that a bit of energy goes unused for a couple of hours. If thermal storage was installed on this system, it could theoretically save the excess and use it later.
Hence, the nameplate rating essentially reflects max output rather than max input.
Re nuclear, I note that you have again sidestepped on the basis that this thread is discussing solar. Then why your references to about everything including solar, coal and CCGT? And nuclear? Your references, not mine. You brought them into the debate and then try to claim that they weren’t invited. And not a word about the fact that they are realistic, available and cheapest - again, your own references show this to be so. I reluctantly have come to see that all forms of low-CO2 energy production must be considered if our poor world is to be saved from its inhabitants.
I hope that my on-topic comment above makes good sense to you re nameplate ratings. Installation of generating and despatch capacity to match the max conceivable input for the best hour of the year is uneconomical. Perhaps 80% of peak nett input is closer to the mark, with the rest either time-shifted via storage or wasted by turning a mirror or two away from the sun or by venting surplus steam.
Beware the elephant.
Oh for Christ’s sake John. Nuclear is not ‘the elephant in the room’. My references only included a cost comparison of the various technologies available. Nuclear happened to be on the list, along with hydro, coal, gas and solar.
I have made my position very clear on why I will not support nuclear power in Australia. Do you want me to repeat it in summary?
1. Too slow to implement. The UMPNER (Switkowski) report lists 10 years minimum, more likely 15 years to build a plant. If they manage 25 by 2050, this will reduce emissions by 18%. it is not the silver bullet everyone thinks it is.
2. Waste storage is not solved and contains unknown future costs. Ditto decommissioning.
3. Australia essentially has no nuclear expertise. So, the design, materials and people to complete the project will ALL need to be imported. In my experience, the more contractors used on a project, the less control one has over the final price. Australia could well be held ransom to contract variations and protracted legal proceedings for another 10 years past the proposed commissioning date. I note with interest the number of nuclear projects world wide that caught in similar loops. Solar thermal in contrast, uses Australia’s natural advantages in sun, steel and manufacturing capability.
4. The political upheaval to actually get a plant built puts it well outside the scope of useful technologies. We need to reduce emissions sharply in the next 25 years. How long will it take to get Australian law changed to allow nukes? Why bother?
Back to Capacity Factors; I think that system is odd. Seems to me that the mirrors are greatly oversized compared to the turbine. I don’t think a capacity factor that just reports turbine generation is of any use at all. That means a grossly inefficient system, with hundreds of square somethings of mirrors could supply to a 1kW turbine and achieve near perfect capacity factor. I don’t see how this is useful or factual.
The mirrors have to be oversized in order to make effective use of the turbine, otherwise the turbine would not be able to reach its nameplate rating, except on the maximum hour of the hottest sunny day. This is grossly uneconomic and inefficient.
There is thus a balance between collector size and turbine size. I think that, in the “real world” that lies at 75 or 80% of the max input of sun’s heat. To improve the reliability of the overall system, this spare capacity can be directed to heat storage. If there was no such spare heat, there would be none left over to divert to storage and power could only be produced ramping up from after dawn and running down after dusk, and none at all on cloudy or rainy days or when thunderstorms are about or winds are so strong that the collectors are structurally endangered.
If storage was free and 100% efficient (we wish!) then the ideal turbine might be about 15 or 20% of the best insolation, running 24/7, with storage for several rainy days as backup. Hey presto! reliable baseload solar thermal with no need for CCGT as backup (1). Thermal storage costs big money and has losses, so each installation must be optimised(2).
(1) Most solar thermal scenarios include a large additional cost for CCGT backup, because of the unreliable nature of ST with only limited thermal storage - eg Spain’s 7 hours, which doesn’t even get through the night. The CCGT (or OCGT) are useful also for peaking power.
(2) A a conceptual exercise, consider a large field of mirrors backed up by a very large infinite capacity thermal battery and having a large number of turbines, which are only turned on when needed. There would be no need for GT’s because the solar turbine capacity is equal to peak demand. The downside is the cost of the infinite size and infinitely efficient (no losses!) heat battery. That heat store is the equivalent, in engineering terms, of a stockpile full of coal.
Now, why not make that heat store out of a number of underground holes full of superheated pressurised water with a steam blanket on top? The water and earth hold the heat, the overlying soil insulates the heat store and the result is a series of cells of super heat, all waiting to be discharged on demand via the turbines, with condensed spent steam returned to a header tank and, eventually, fed back to recharge the underground heat store.
I am not the first to think of this, but I don’t know of it being trialled.
To me, it is very much do-able, if there is such a word. Probably much better than hot rocks in the long run.
Okay, thanks JohnB, that makes a lot more sense.
I guess then for solar thermal CF only matters WRT the nameplate capacity of the plant, which is fine, but it makes the numbers for working out averages for a huge area much less dependable.
I have heard of someone in Australia talking about test holes for doing just that. I thought it was Ausra at Liddell, but I got a severe dressing down over being wrong on that. So, all I can tell you i that it’s being tested.
While you’re on that bent, have a look at Compressed air energy storage
I’ve often thought from working at Sydney Water about energy efficiency opportunities by avoiding an energy conversion step. I don’t think hydrogen will get far, mostly because there is so many steps, each with inherent losses. So rather than firing a biogas engine to generate electricity and offset imported electricity, why not use the engine to pump water directly? I think CAES has some potential applications with this in mind.
John Bennets… That’s just not true, Torresol is a 75% capacity factor plant. This is the standard size. Because it’s cheaper to add mirrors and storage than it is to add more turbines. 3 turbines each running 25% of the time costs 3x as much as 1 turbine running 75% of the time.
So if your turbine costs say 200million and your storage costs $50 million to scale and your tripling the size of the mirror field costs $150 million well you are $200 million ahead.
And with 75% capacity factor power towers the actual fluid running through tower is molten salt, and it is always going from salt to steam - so scaling this is cheap.
@Matthew W:
I suggest that you re-read my last contribution. We are not far apart. Molten salt Vs steam storage will come up with different costings, but the basics are the same.
The relative costs of additional storage and additional turbines and the value of additional storage will make each installation different, but the principles are the same.
I have no hands on experience with molten salt, but I have worked with solar thermal trial plants and I am sure that the principles are as I stated.
Whether Torresol has a “standard size”, by which I take it you mean “standard configuration”, I do not know. My main point is in relation to the ability to collect solar energy/heat durng the 12 hour day and store some for the hours of night, thus reducing the need or value of 100% capacity turbines. No plant can operate at 75% of mirror capacity 24/7, because this implies 150% or better efficiency… not possible. I guess that we are on different wavelengths.
If you are a designer or have actual experience of the Torresol plant, I would like to analyse the thermodynamics of the plant you are describing.
John,
It is the turbine that runs at 75% capacity factor. The mirror field runs at a lower efficiency that 75% However they are different things.
By trying to measure the capacity factor of the mirror field, you are effectively doing the equivalent of measuring the conversion efficiency of the energy in coal to what is released at the turbine then mmultiplying that by the capacity factor.
So for comparison a NSW coal plant runs at say 75% capacity factor, however the amount of energy in the coal is 4x what is actually delivered. So what you are asking would describe a coal plant as having an annual capacity factor (this is not the right terminology but I’m going along with it for sake of the discussion) of .75 x .25 or 18% No body ever quotes the figure like this.
. However the energy payback ratio or the embeded CO2 per kwh is what counts.
To illustrate this point again say a wind turbine operates at 55% efficiency for the wind passing through the swept area. We do not quote this as .5 x .3 and then claim the annual average capacity factor of this wind turbine will be 15% It is standard to claim that the annual capacity factor of the wind turbine is 30%
And both are very low for CSP. Energy Payback Ratio on one of these CSP plants is in the order of 4 -6 months. That means that of the 30 years the plant should run it will take 1/2 a year to pay off the total cost of building including making steel, concrete, glass etc, and dismantling.
There are some advantages with dismantling though as they generally use concrete steel and glass and it all can be recycled so the energy payback ratio on the subsequent concentrating solar thermal plant that replaces the first one built in 30 years time would be much faster in the order of 1-2 months.
I basically agree with you. The energy sent out from NSW coal fired units is more like 38% than 25%, but that does not alter the principles you quote.
This discussion has a positive side, though. It indicates the need for input power (sunlight) to be greater than output power, and that each step along the way is a balancing act, where construction costs (land, materials, labour and energy) may limit the capacity of the overall system.
The figure you quote for energy payback is interesting. It is much shorter than I expected. Nice work.
John,
Phew — these are always difficult things to get across.
The 38% of a NSW coal plant is thermal efficiency of the turbine.
However the plant uses a lot of parasites and embodied energy in order to get coal out of the ground and into the coal power plant.
This is where by 25% comes from — however depending on the plant maybe upto 30%. Victorian Latrobe Valley Brown coal plants are lower as a lot of water is used in drying the brown coal which comes out at 62% moisture content.
Other parasites including running scrubbers, demineralisation and cooling pumping etc.
Matthew, the actual turbine is much better than 38% efficient in typical NSW installations.
I speak from certain knowledge, that the ESO, after all losses such as condenser and auxiliaries is 38% or better for a typical NSW 660MW generator.
The 500 MW generators are better than 30% - probably close to 35%.
Thus, any difference between my quoted figure and the one you are using must relate to the energy cost of mining and part of the transport cost of coal from the mine. 13% differential for upstream energy uses seems steep to me, but the story is correct:
(1) If we are comparing energy generation technologies, the whole picture should be assessed - upstream energy use during construction and mining, losses during generation and the energy recovered or expended during site restoration at the end of life.
(2) We need to recognise that these figures may be elusive and must be available for audit to ensure that they are truly comparable.
Neither 1 nor 2 above is easy. Gathering reliable figures needs skill, knowledge and professionalism in an area where a lot of money is on the line and some data is held close to the chest and may be commercial-in-confidence.
I will leave brown coal to the Mexicans south of the border.