Concentrating thermal solar power – CSP or CST – has been the perpetual bridesmaid to solar PV. Just as itÂ´s about to go commercial, the price of solar PV drops again and the projects are shelved. CSP is growing, but at a fraction of PVÂ´s headlong pace. Brightsource, the leading commercial player in the US, only predicts the global market to reach 30 GW by 2020. (One prediction for PV at the same date is 330 GW.) CSP has one big thing going for it: storage. As you start with a hot fluid, itÂ´s fairly easy to add a thermal heat reservoir at the point of production, The Spanish company Gemasolar has demonstrated 24-hour electricity generation with hot salt storage coupled to a solar tower. The Australian grid operator AEMO reckons that this isnÂ´t optimal, as thereÂ´s very little demand for electricity at 3 am, and 9 hours of storage will get you comfortably through the dark evening (pdf report, Appendix 2). Still, a 9-hour Gemasolar clone remains expensive.
The likeliest way to bring costs down is to go hotter. Currently operational CSP plants use water/steam as the working fluid, and standard Rankine-cycle steam generators. The PS10 solar power tower near Seville, the largest currently operational, has a steam working temperature of only 275 deg C at the receiver outlet. BrightsourceÂ´s Ivanpah tower in California, the largest under construction, will run hotter at 560 deg C.
The Rankine cycle isnÂ´t very efficient. Most current steam generators are at 35-42%. This is a very mature technology and itÂ´s very hard to improve on it, especially with relatively small 100MW rigs. To push up efficiency, the favoured way is to go hotter. Partly because of the basic physics of heat engines; partly because above 800 deg C or so you can switch to or add a Brayton cycle, aka a gas turbine. Feed the still hot gas turbine exhaust into a conventional steam generator, and you have a combined-cycle generator plant. These are routinely used in natural gas plants and regularly achieve efficiencies over 60%.
A solar tower can easily reach such temperatures at the focal point. Large research parabolic solar furnacesÂ get over 3,500 deg C. Long before the theoretical limit of 6,000 deg – the temperature of the SunÂ´s surface – you would run out of materials to hold targets in, and you are left with a fun but completely useless fixed death ray. The problem is capturing the heat with a non-magical material and transferring it to your heat engine.
One candidate is an alveolated ceramic block fixed in the receiver, with blown air as the working fluid. The EU went into this in the last decade but the scheme hasnÂ´t made it into precommercial pilots. I suspect the problem isnÂ´t the ceramic but the air: perhaps itÂ´s not dense enough to support high rates of heat transfer.
Enter Sandia with a different idea: grind the (EUÂ´s?) refractory ceramic into sand and use a sand-air mixture as the working fluid. Capturing the heat is beautifully simple: the sand just falls in a curtain past the receiver window and is heated up to 1000 deg C. It is collected at the bottom of the tower in a heat exchanger which feeds air or steam into the turbine. The sand has a lot of thermal inertia so it could be its own storage, depending on cost. The working fluidÂ´s higher temperature and density translates to a more compact and efficient plant with – they hope – a much lower cost of generation.
This photo shows a demonstration receiver. Clearly this is still an early-day lab rig. The aim is a complete operational design by 2017, and they still have to overcome a host of problems, including machinery (pumps, bucket elevators) that will survive a diet of red-hot abrasives. More sketches here. Via CleanTechnica.
Still, itÂ´s very pretty idea. Gravity is as cheap and reliable a way of moving stuff around as you can possibly get. Give them more money!
Is it frivolous to judge technological innovations on aesthetic grounds, as IÂ´m doing here? It would be to make it the only criterion. But form follows function. The human aesthetic response is surely part of Daniel KahnemanÂ´s automatic, instinctive, efficient mental System 1, even if itÂ´s educated by conscious learning and reflection. The ev-psych argument that itÂ´s partly based on utility is very strong. Beautiful people have the characteristics of biologically good reproductive partners. Beautiful landscapes of the Arcadian type correspond to good environments for hunter-gatherers. ItÂ´s not a stretch that this extends to artefacts as well as images: a beautiful house is one you would like to live in, an iPhone is a tool you enjoy using. So our aesthetic reaponse to engineering reflects an instinctive judgement that itÂ´s fit on a range of practical dimensions: transparent, controllable, economical, compact, stable, reliable.
Of course the aesthetic reponse has other triggers than fitness for use. We are also attracted to environments – mountains, storms,Â deserts of sand and snowÂ – that are extremely hostile; animals like tigers and dinosaurs that are (or would be) very dangerous; and representations of violence and terror in tragedies and action movies, embedded and distanced in a moral narrative. Why this should be so is a puzzle. AristotleÂ´s theory of catharsis about the last category look fishy, but I donÂ´t know of a better one. The Romantics distinguished between the Sublime (the Alps, Gothic cathedrals, King Lear) and the Beautiful (girls, flowers, songs). ThereÂ´s something in this, but as far as I can tell the subjective sensation is the same. Maybe thereÂ´s a connection with the paradoxes of sexual selection: features like the peacockÂ´s tail or behaviour like the pronking of gazelles that are worse than useless, but demonstrate to potential mates a maleÂ´s superior fitness as obstacles he has successfully overcome. (Sexual selection is not gender-neutral.) Perhaps we admire Gothic cathedrals because of the sheer crazy prowess of building such unnecessarily tall and airy structures out of stone and weak mortar. In any case it doesnÂ´t look as if my aesthetic response to the sand CSP concept is driven by any of these factors.
Early-stage R&D is necessarily based on inadequate information. ItÂ´s likely to fail. The decision to fund or not is a poker bet based on the few cards in your hand. ItÂ´s sound to pay attention to your instincts as well as your reason – though as information accumulates during the game or project, reason should exert greater control. The aesthetic response is one such input from instinct. So donÂ´t believe in the beautiful idea as Dirac advised – but do give it a chance.
There is a second, unrelated argument for making and publishing an aesthetic judgement. Most R&D projects fail. Indeed most human projects of any ambition do too: most scientific papers and novels are unread, most films are scarcely watched, most competitors in a sporting or electoral competition lose. But while generally failing, we can all still occasionally achieve good work. The relegated football team can score a fine goal in a match it loses. I suggest that itÂ´s therefore a social and cultural duty to praise aretÃ© wherever we see it. Success, which is often due to luck, is a very crude and misleading proxy for excellence.