Open Thread 23

Here’s another of Solar Reserve’s projects – this time in Chile.

http://cleantechnica.com/2015/08/25/cheap-baseload-solar-copiapo-gets-ok-chile-exclusive-info/

[Solar Reserve’s Copiapó] 260 MW project will comprise 150 MW of PV panels for daytime generation, and two 130 MW CSP towers utilizing the company’s molten salt storage, for an installation of 410 MW between the two solar technologies.

By oversizing the CSP, SolarReserve can guarantee a round-the-clock baseload supply for a firm 260 MW at more than 90% capacity.

“Total installed capacity is 410MW, but different parts of the plant run for different hours to provide 260 MW 24 hours a day,” he said.

With a price expected to be well under 10 cents per kilowatt-hour, the pioneering 24-hour solar project in Chile’s Atacama desert can compete on price against other baseload generation. Much of Chile has been dependent on pricy fossil fuel imports from its neighbors.

260MW capacity at 90% capacity factor from a hybrid solar system is probably the nuclear industry’s worst nightmare. Fortunately for them it can only be installed where there really is a lot of sun such as Chile’s Atacama desert – or maybe parts of Australia.

LikeLike

Chris Harries said :

All the authenticated ERoEI figures I’ve seen for solar thermal power come in underneath that of solar photovoltaic panels. That, and the fact that solar thermal opportunity exists in listed locations around the world, would seem to render it a very doubtful technology.

Chris,

You would expect the EROEI for CSP / STP to be about half that for solar PV (assuming the same lifetime) to reflect the cost ratio – PV is half the price of CSP and will continue to be cheaper even as the costs of both come down.

But that’s not the whole story. Solar PV is destined to be so cheap that you should be using it when you can, but it can never generate when the sun is not shining. Whatever you do outside daylight hours is always going to be more expensive than solar PV.

So the EROEI comparison is not a reason for ditching CSP / STP.

The locations suitable for installing CSP + storage are generally the sub-tropics and tropics.

That might sound like it disqualifies CSP, but in the long-term, HVDC lines with very low losses (3.5% per 1000km) can be justified to pipe this power up to 2000km away to quite a few of the large human population centres.

Saharan CSP is within reach of Europe up to UK and Berlin.

Western China CSP will probably be connected to the population centres in the East of China. It is not scared of 50 year investments in UHVDC lines across the width of the country – http://www.ee.co.za/wp-content/uploads/legacy/energize_2012/07_TT_02_A%20review-of-HVDC.pdf.

Australia has reasonable CSP available in the interior but zero political will to exploit it.

LikeLike

@singletonengineer

Fresnel trough + steam systems cannot readily be used with storage, so solar PV probably overtook them some time ago in terms of economics.

Solar tower with hot salt and storage seems to have advantages over Fresnel reflecting trough systems. The distributed equipment is simpler, apart from the requirement for two axis tracking instead of one. The salt can be heated in the solar collector tower to a higher temperature than oil (or steam) in the trough pipes, and there are no losses from transferring the heat from oil to salt for storage. It is straightforward to overconfigure the heliostats to provide full generation power in winter as well as summer. In summer or at noon you just don’t point all of them at the collector (or you might break it)

The higher steam input temperatures equate to a bigger efficiency improvement than you might think because most of these systems are in deserts and for daytime operation the thermodynamic exit temperature of the steam is high than normal. The storage capacity is improved when CSP generates at night, so a hybrid daytime solar PV + nighttime CSP system is good. In the Sahara I spent a very cold couple of hours viewing the spectacular star field through binoculars before spending a night freezing in the coldest bedroom with no heating that I can remember. The much lower steam exit temperature at night again increases the thermodynamic efficiency compared with a daytime-only system, improving the storage time over what you might expect.

Solar Reserve is very proud of its solar collector tower which lets them work at a high temperature. The US is subsidising Solar Reserve’s research to find materials for even higher collector temperatures which might improve thermodynamic effiency and reduce the cost further. There’s a thermodynamic limit. The temperature of the collector depends on the thermodynamic temperature of the sunlight filtering through the (clear) atmosphere and the ratio of the solid angle of all the reflectors at the collector divided by the possible 2 pi solid angle. But current systems are way off this.

Solar Reserve uses heliostats with separate IP addresses on a common network which has to point them all in different directions to reflect sunlight onto the tower, although many competent engineers working with programmers could solve this particular problem.

These improvements over Fresnel trough are not individually big deals, but as an engineer, surely you have to wonder whether together they constitute a step change in the practicality and pricing of solar thermal. Solar Reserve’s projects look like the real deal to me.

LikeLike

@singletonengineer

The question was whether there was anything to stop the Solar Reserve technology being configured to meet the customer requirement, which was expressed as baseload generation of a fixed output with a capacity factor of

(direct sunlight hours / year) x 2 / 8760.

Instead of taking the trouble to understand the technology and analyse how such a requirement might be met with it, singletonengineer comes up with :

If 24/7 means all day, every day, for at least a week, then an overcast week in a winter’s month of 9-hour days will require arrays that are of the order of 10 times the size of an array sized to just meet 100% capacity during a sunny summer day of 24 hours.
…..
There are a few opinions regarding 24/7 availability (at what capacity factor?) and an assertion regarding complete lack of emissions or waste which is far from a cradle to grave analysis.

This response is from someone unwilling or unable to analyse the problem, despite having worked on other types of CSP system before.

Further the comment :

Perhaps there are diesel heaters to top up the oil temp on bad insolation days and months. Remember, 25% of energy sent out from one large 12/7 solar thermal operation was found to be diesel powered discussed above

betrays the fact singletonengineer either hasn’t read or hasn’t understood that the particular installations we are discussing use permanently molten salt, not hot oil, and use no gas whatsoever.

As a service to the rest of the readers here, singletonengineer should go back and read the technology documents on the Solar Reserve site properly before posting irrelevant FUD on the Crescent Dunes, Redstone or Copiapó CSP installations.

If the required capacity factor is specified as “direct sunlight hours per year x 2 / 8760” then that allows the system to be configured for the shortest day of the year on the assumption that the sun shines all that day. The inclusion of the capacity factor in the requirements and contract then takes care of the fact that the sun has a known probability of not shining at any particular daytime time.

By varying the capacity of the following high-level components :

x solar PV panels
x heliostats
x solar collectors
x hot and cold salt tanks
x hot salt to steam boilers
x steam turbines and generators

you can configure a system which, with no interuption to daytime sunlight on a particular day, will allow 24 hours generation of the required output subject to no equipment failures. And if it will do that for the shortest day each year then it will do it for any other day.

In addition there needs to be a financial optimisation of the ratio of solar PV generation to CSP generation to minimise the cost to the customer. Only some components are affected.

Thus the specification of a capacity factor related to the number of direct sunlight hours per year allows the design of a baseload hybrid solar system.

About the only valid point singletonengineer makes is that some allowance must be made for component failures, either in the contractual requirements or by providing redundancy in various components.

Accepting sales blurb at face value isn’t engineering. It is intellectually lazy consumerism.

However, contracted PPA’s for a system are an entirely different matter as the supplier is then on the hook for the performance of the system. singeltonengineer confuses the two.

Let’s have some logical analysis around here instead of a refusal to properly understand renewable technologies competing with nuclear. The mantra “nuclear good, renewables bad” is not an acceptable substitute for an open mind.

LikeLike

Peter Davies — Do not accuse of cherry picking until you are at least half as good a statiscian as Grant Foster, who runs his Open Mind blog as Tamino. In particular, 2015 was simply the most recent full year for the USA capacity average. Indeed, looking at the graphic you posted from EIA, one sees that the US nuclear power fleet has obtained around 90% capacity since the year 2000.

I assume that this growth to maturity is partly a learning curve for these LWRs and partly a matter of preferring nuclear over coal more recently.

As for the British capacity averages I am not well informed of the British regulatory environment nor how British utilities choose to run their equipment. I have enough difficulty with just the USA. Possibly you would care to learn more about how it is done in Britain and, thoughtfully, inform us. Every country is different.

LikeLike

From the Engineering ToolBox here are some higher calorific values. The three grades of coal are highly variable by source and I only give a central value.
Lignite, i.e, brown coal: 16300
Bituminous coal: 20000
Anthracite: 33000

Methane: 55530

with units of kJ/kg.

This corrects the wrong explanation of why natural gas, almost all methane, is a superior fuel, as found in a linked article by a polymath in a comment several back.

While correct, these values fail to take into account the fugitive methane which escapes into the atmosphere in the course of mining, refining and distribution. Some hold this is enough to make natgas use worse for global warming than even burning lignite.

LikeLike

Peter Davies, as an engineer, how can you say “give the irrelevant history a rest” when discussing mass deployment of expensive, unproven technology (CSP in this case)?

Surely your training focussed on delivering practical systems at economically sensible costs so that resources (including money/finance) are not wasted.

Your line of thinking would have had Edison force people to buy light bulbs that burnt out every few hours of operation rather than test the 6000 odd materials and configurations before putting a durable and economically viable light bulb on sale (and letting market forces determine its success).

Or should people have been forced to abandon gas, oil and kerosene lamps and returned to candles when Edison first started his light bulb experiments in expectation that he would come up with a solution in short order?

Engineering is about solutions. If the problem to be solved is for a critical requirement, then the solution should be proven, not experimental. .

LikeLike