C J

I believe that heaters of that nature are in fairly common use for water heating purposes, but as far as I know nobody is using them for power generation. That's probably because they don't generate sufficient temperature differentials to run a conventional engine, and explains why companies like Stirling Energy are using parabolic mirrors to generate a 'hot spot' for their engines to operate in. The Stirling Energy generators are an impressive piece of engineering, but you can see what I mean about needing to keep the mirrors clean! They also need clear skies, which means we're unlikely to see any of them deployed in the UK. C J

I'm sure part of it stems from the cost of keeping all those parabolic mirrors clean and all the hassle involved in keeping them pointing at the sun, plus the fact that they're no use when it's overcast. To be honest, I'm not sure why everyone insists on using mirrors, since apparently unconcentrated collectors are only 20% less productive than ideal concentrators. FWIW I just found this figure in 'Advanced Engineering Thermodynamics' by Adrian Bejan, but haven't had time to read his analysis in detail so am not sure what the practical implications are. C J

The point I was trying to make was that I think conventional bulk substrate PV cells are a dead end, and we need the kind of techologies you describe to reach the market ASAP. I should probably have qualified my earlier post by saying I'm not a fan of _current_ PV technology! It's over ten years since I did any device physics, so I'll admit to being a bit vague there. The fact remains, though, that the currently available technology leaves a lot to be desired in terms of efficiency. Hopefully your colleagues will come up with a solution. Like the good Doctor said, it suffers from being somewhat boring. Unfortunately, the VCs and funding agencies like their technology new, shiny, high risk and exciting (and preferably either nano-this or quantum-that). Leaves no place in the portfolio for ideas which rely on good old thermodynamics and control theory (and plumbing!). C J

I can't see there being any cost advantages from moving to conventional PV cells with bulk substrates other than Silicon, since the manufacturing process will be pretty much the same as for bulk Silicon but with something slightly more esoteric (and expensive) than sand at the input. My expectation is that the cost breakthrough will occur once a suitable thin film technology is in place. The interesting thing about this is that the people best placed to benefit are unlikely to be the incumbent PV manufacturers - it will be the people who're currently experts in manufacturing LCD panels! The technical reason I still prefer thermal solar is that in principle a thermal solar solution can be designed which converts all incident light (regardless of wavelength) into energy. With PV cells, you're only ever going to convert the incident light whose wavelength matches the energy bandgap of the material you're using. I know that research is ongoing to develop cells with multiple bandgaps, but I can't see how anyone is going to be able to develop a cell which can absorb all wavelengths from infra-red to ultra-violet the way a black box thermal collector can. The catch with the black box collector thermal solar solution is that you then need an ultra-efficient (ie, approaching Carnot cycle efficiency) low temperature differential heat engine to make use of the collected heat... and these don't exist (yet). C J

I suspect this is a factor... Historically PV input costs have been artificially depressed because they've been able to make use of the semiconductor industry castoffs... in this situation, increasing demand for PV devices should not have had a feedthrough to front end silicon prices and would not have had an effect on the price of the reject silicon (since the PV manufacturers are effectively bidding against the scrapheap). If the PV industry demand for non-IC grade silicon now exceeds supply, good old price elasticity will start to assert itself on the non-IC grade stuff, and the PV manufacturers will also have to compete with the IC manufacturers for their excess requirements - a double whammy. I'm not a fan of PV - a lot of the assumptions about its cost effectiveness have been based on the artifically depressed prices of the input costs (junk silicon). This has severely biassed cost comparisons against competing thermal solar solutions in favour of PV, when in reality thermal solar should be the most cost effective and carbon neutral solution. C J