tag:blogger.com,1999:blog-13900197.post3493131384073301445..comments2023-10-15T05:20:00.675-06:00Comments on Entropy Production: The Glittering Future of Solar Power:Prognostication of Photovoltaic Capacity Extrapolated from Historical TrendsRobert McLeodhttp://www.blogger.com/profile/05270962906437456350noreply@blogger.comBlogger40125tag:blogger.com,1999:blog-13900197.post-45676888158566555242012-06-27T00:26:33.217-06:002012-06-27T00:26:33.217-06:00This blog site has lots of really useful info on i...This blog site has lots of really useful info on it! Cheers for informing me!<br /><br /><a href="http://www.polarracking.com/h/news" rel="nofollow">Solar Racking Manufacturers</a>Stevehttps://www.blogger.com/profile/08746314915189529019noreply@blogger.comtag:blogger.com,1999:blog-13900197.post-8954514247433313622009-05-18T01:53:00.000-06:002009-05-18T01:53:00.000-06:00Thanks for this wonderful post.......Thanks for this wonderful post.......Solar Power Businesshttp://www.solarenergy-solarpower.comnoreply@blogger.comtag:blogger.com,1999:blog-13900197.post-36930130155300884422009-04-05T07:20:00.000-06:002009-04-05T07:20:00.000-06:00This comment has been removed by a blog administrator.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-13900197.post-37409238415349310452008-10-03T10:09:00.000-06:002008-10-03T10:09:00.000-06:00Lynch report:http://rapidshare.com/files/150589716...Lynch report:<BR/><BR/><A HREF="http://rapidshare.com/files/150589716/9_04802_.pdf.html" REL="nofollow">http://rapidshare.com/files/150589716/9_04802_.pdf.html</A>Robert McLeodhttps://www.blogger.com/profile/05270962906437456350noreply@blogger.comtag:blogger.com,1999:blog-13900197.post-84613044693884139412008-09-29T08:09:00.000-06:002008-09-29T08:09:00.000-06:00Would you happen to have a copy of report you refe...Would you happen to have a copy of report you referenced "Solar Wave - Apr-07 Edition" that you could share. The link is broken and I have been unable to find any reference to it elsewhere on the net.<BR/><BR/>ThanksAlexander Forstnerhttps://www.blogger.com/profile/05482200633214504759noreply@blogger.comtag:blogger.com,1999:blog-13900197.post-11548461510927971132008-07-07T05:25:00.000-06:002008-07-07T05:25:00.000-06:00Whoops, 15 and 30 square meters, respectively.Whoops, 15 and 30 square meters, respectively.Cyril Rhttps://www.blogger.com/profile/17667288494374310919noreply@blogger.comtag:blogger.com,1999:blog-13900197.post-55308621231624572102008-07-07T05:21:00.000-06:002008-07-07T05:21:00.000-06:00Oh, and that's 90 billion divided by 6 billion peo...Oh, and that's 90 billion divided by 6 billion people (it's probably closer to 7 billion actually) means 10 square meters per person. For the low insolation case there's half the energy available, so double the area becomes 20 square meters per person. <BR/><BR/>This doens't include storage losses. But even then the area is just not big. My house is pretty big at more than 200 square meters, and much more still if you count the backyard. And the agricultural land I use (by consuming food) has to be many times that figure. So solar land area won't be the bottleneck at all.Cyril Rhttps://www.blogger.com/profile/17667288494374310919noreply@blogger.comtag:blogger.com,1999:blog-13900197.post-41874589256218442602008-07-07T05:09:00.000-06:002008-07-07T05:09:00.000-06:00Perhaps I can clarify a few points that KD made:Th...Perhaps I can clarify a few points that KD made:<BR/><BR/><I>The cost of processing a sq. cm. of silicon has been surprisingly constant. You can't expect that to change in such drastic measures...</I><BR/><BR/>That's for traditional flat plates. You mean ingots, right?<BR/><BR/>First, as you said their efficiency kept going up. Double the efficiency at the same area costs means twice the energy. Given the fact that electricty cost is inversely proportional to the energy harvested, that means half the cost per unit of energy.<BR/><BR/>Of course, you have a valid point in implying that the limits of efficiency of traditional flat plates silicon PV are in sight, so novel approaches are needed to further reduce cost. You are wrong to think this isn't happening: Sliver cells and String Ribbon are just two of the approaches that could plausibly reduce the cost per square meter quite drastically.<BR/><BR/>Second, there's manufacturing improvements. Non cell related costs have gone down a lot due to improved manufacturing and technology, and larger scale production. Don't underestimate those cost, they are big and subject to continuous improvement.<BR/><BR/>Here is where thinfilms often have an advantage. I am skeptical about rare earth thinfilms (CIGS, CdTe etc) being able to keep on growing exponentially long enough. As you mentioned also, there are huge uncertainties with the availability of rare earths. Some people argue that the percentage of rare earth cost of the total cost is so low that prices could go up a lot, and in that way stimulate new exploration and exploitation of previously uneconomical deposits. But this is very speculative, and even if true it will take a long time to get the capacity up and running. Will the mining capacity be able to grow exponentially?<BR/><BR/>But silicon thinfilms show a lot of promise. Manufacturing investment is much lower, especially with new manufacturing processes like roll-to-roll production, printing technologies etc. <BR/><BR/>Installation can be made very easy, for example one could buy rolls of thinfilm "carpet" with an adhesive underneath and just stick it on your roof. I think United Ovonics has such a product. Other approaches are to make the modules very large in order to reduce overhead costs. For new build, or roofs that require heavy refurbishment, Building Integrated PhotoVoltaics (BIPV) is promising.<BR/><BR/>As for inverters, mass production really does wonders. Car inverters cost 40 cents per Watt according to the engineer-poet. That's less than half of the cost of solar inverters.<BR/><BR/>A bit further in the future, there's a lot more intersting stuff coming up. Infrared nano-antennas potentially up to 80% efficiency. Amorphous diamond as a semiconductor in stead of silicon promises up to 50% efficiency at lower production costs per square meter. Silicon nano-crystals of up to 30% effiency while using a tiny fraction of the raw silicon.<BR/><BR/><I>Vapour Deposition is not an answer, I don’t know where you got that from… It is quite expensive actually.</I><BR/><BR/>Depends on the type of CVD and the type of material that has to be deposited and onto what material. Some processes are already very cheap. Just look at how cheap low e coatings are for insulated windows. Ten years ago they were really expensive. I think that saying CVD is expensive is a blanket statement that has no value without a good reference. This is an industry that is developing fast, with innovations in manufacturing following quickly one after another (just look up low e coatings development for an example).<BR/><BR/><I>With respect to your PV electricity cost simulation: if the discount rrate was 1% over the central bank rate, that’s too low, it won’t excite too many investors with such returns. Also, I don’t see cleaning costs in your models.</I><BR/><BR/>I could think of a model in which the government stimulated PV installations by allowing PV owners to get most of their loan from them. The government is able to borrow money very cheaply from private investors. However, the extra demand for loans means the govt has to provide sligthly higher interest for private entities in order to attract the extra money. And the banks aren't going to be happy about all this of course; they'll cry that it's unfair competition, and they may have a good point. But it's definately doable; it's a strategic policy which may very well be enough justification. <BR/><BR/>Also, cleaning costs are very low for large systems. The Springerville generation station has a few mills or something like that in cleaning costs. In fact the total variable costs are under one cent per kWh. Of course, a smaller system is going to have higher costs per kWh. A 5 kW system could require 15 minutes of cleaning every month to keep it close to optimal in output, according to a friend of mine which has a similar installation. Let's take 50 bucks an hour including materials. (all you need is a bucket of water and a long cleaning pole so you don't have to get up the roof). That's 150 bucks of cleaning per year. In a good solar location you'd get 10000 kWh per year from this installation which is 1.5 cents per kWh. If the solar system gets cheaper, it will actually make sense to devote less time to cleaning. Cleaning only 10 minutes six times a year means just 0.5 cents per kWh. That would have to be compared to the loss in output, if any. Of course, a location with less sunshine is going to be more expensive, but even at half the above production, still not exactly a showstopper. For now, I don't suggest solar panels in areas that get less than 20 percent capacity factor, unless for off-grid and remote applications of course.<BR/><BR/><I>You reference 18TW for human use, divided by 6B people, this equals 3kW per person. 3.5 sq.m with 300W constant insolation and 10% efficiency gives only 105W per person. And that’s for peak output. You have capacity factor, which for baseload (which solar is) is very low.</I><BR/><BR/>You are confusing peak figures with average figures. You already calculated constant insolation, so you already have average output - not peak.<BR/><BR/>However, the way to do these macro scale calculations is to look at energy, not power. The world uses about 18000 billion kWhs per year. If your figure of 18 TW is correct then that implies a capacity factor of 11 or 12 percent, very reasonable I think. <BR/><BR/>But let's do our own calculations. Using a reasonable solar insolation figure of 2000 kWh/square meter per year and 10 percent efficiency that's 200 kWh net. That's 90 billion square meters or 90000 square kilometers. The world is about 149 million square kilometers of land area; about 0.06 percent of the world's land area. Even at half the insolation I used that's still just 0.12 percent of the world's land area.<BR/><BR/>Whoa. This post has gotten way too long!Cyril Rhttps://www.blogger.com/profile/17667288494374310919noreply@blogger.comtag:blogger.com,1999:blog-13900197.post-68972160414094664002008-06-04T16:41:00.000-06:002008-06-04T16:41:00.000-06:00I'm a long-time reader, first-time poster. Could R...I'm a long-time reader, first-time poster. Could Robert respond to "dr. Krassen Dimitrov"? I'm just really, really curious...and pretty damn unfamiliar with science and technology. I try to keep up as best I can, though. :)Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-13900197.post-25513272307871248952007-09-21T01:17:00.000-06:002007-09-21T01:17:00.000-06:00Interesting... but flawed:1. Even though PV techno...Interesting... but flawed:<BR/><BR/>1. Even though PV technology is akin to semiconductor, the same scaling effects do not apply. When people talk about the scaling in semiconductors it is in the context of an Integrated Circuit: size of a transistor transistor density. The cost of processing a sq. cm. of silicon has been surprisingly constant. You can't expect that to change in such drastic measures...<BR/>2. The historic data on the learning curve has nothing to do with decreasing production costs per sq.cm. due to technological innovation as you imply. The two major drivers have been: (i) improved efficiency lower price per Wp (not per sq. cm), and (ii) the growth in semiconductor fab capacity has created slack in older generation foundries, which cannot process at the current node. PVs have been a good way for them to extend their economic life by utilizing the already sunken cost. None of these two factors will keep up with more significant instalment.<BR/>3. Vapour Deposition is not an answer, I don’t know where you got that from… It is quite expensive actually. The big rage now is with baking thin films of nanoparticles (NanoSolar), however this is not just silicon, it is CGS (copper-gallium-silicon), which of course is not scalable (worldwide gallium production is only a few tons).<BR/>4. With respect to your PV electricity cost simulation: if the discount rrate was 1% over the central bank rate, that’s too low, it won’t excite too many investors with such returns. Also, I don’t see cleaning costs in your models.<BR/>5. I am totally confused by your estimates of 3.5 sq.m. per person of solar panels.<BR/><BR/>You reference 18TW for human use, divided by 6B people, this equals 3kW per person. 3.5 sq.m with 300W constant insolation and 10% efficiency gives only 105W per person. And that’s for peak output. You have capacity factor, which for baseload (which solar is) is very low.Krassen Dimitrovhttps://www.blogger.com/profile/06462795325843234990noreply@blogger.comtag:blogger.com,1999:blog-13900197.post-2840502672593119632007-09-01T10:40:00.000-06:002007-09-01T10:40:00.000-06:00Here's some solar ready home info. http://www.foc...Here's some solar ready home info. <BR/><BR/>http://www.focusonenergy.com/data/common/dmsFiles/R_EH_MKFS_RenewableReadyNewHomes.pdfAnonymousnoreply@blogger.comtag:blogger.com,1999:blog-13900197.post-4493372268097864402007-08-24T22:34:00.000-06:002007-08-24T22:34:00.000-06:00I too am concerned with CdTe panels. I can almost ...I too am concerned with CdTe panels. I can almost see the contaminated groundwater headlines now.<BR/><BR/>I appreciate your manufacturing explanations for thin-film and your conviction. Thin-film definately wins out in the flexibility of application department. <BR/><BR/>Your numbers make a little more sense now. But you did state that all energy and not just electricity could be replaced. Did you mean just electricity? Would it not be more reasonable to extrapolate where electricity use per capita is headed and use that to figure for solar area requirements?Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-13900197.post-52045836957645291442007-08-24T13:55:00.000-06:002007-08-24T13:55:00.000-06:00mrshiba said:But here's what bothers me. Suntech w...mrshiba said:<I><BR/>But here's what bothers me. Suntech was selling conventional panels for $3/Wp in 2004. Their panels go for $4/Wp currently. The price hike has been driven mainly by inflated raw silicon and wafer costs. When I ponder what the Wp costs would be without the inflation I'm left thinking the 50-75 cents in ancillary cost savings is very significant. Does this reasoning make sense? When I apply the learning curve idea to conventional panels it appears as though grid parity will be reached with or without thin film.</I><BR/><BR/>Well you've discovered the difference between price and cost, yes. From what I understand, the Chinese manufacturers can produce amorphous Silicon panels for about $1.0/Wp, with perhaps an efficiency of 7.5 % after the initial photodegradation. Due to worldwide demand, they are selling them for several times that.<BR/><BR/>For the PV industry to continuously grow at 33 % a year, it will need to realize very healthy profit margins. That will, of course, slow down the rate at which the market price drops, assuming demand doesn't sag. <BR/><BR/>Thin film does have a couple major steps to take that will greatly reduce its ancillary costs, and those are: application of a thin hardcoat (instead of thick glass glazing), and application onto roofing material as the substrate.<BR/><BR/>Thin-film PV is currently made backwards: you use the glass as the substrate, apply the transparent conductive oxide, apply your active layers to form junctions, and then finally the back electrode (typically aluminium). However, there's no real need for a ~2.5 mm sheet of glass if you can develop something harder and tougher and apply it via an evaporation process. If you use a roofing material as a substrate so that your panels are now 'dual-use' (absorb sunlight, shed rain) you've now negated the bulk of those efficiency dependent ancillary costs.<BR/><BR/>I personally do not especially like the toxic panels (CdTe) from a market acceptance perspective. Some of the formation gases in thin-film silicon are just as toxic (e.g. phosphine gas) but not present in the final product.<BR/><BR/><I>You mentioned it would take 3.5-7 m2 of 10% efficient panels per person to supply all our energy needs. 1.25-5 kWh per person per day isn't high enough. What did you mean with this quote?</I><BR/><BR/>The main difference being that I was giving a world average, and all the countries at the bottom of the list balance the OEDC countries at the top. I am also making the assumption that we could replace our energy consumption with electricity at a 3:1 ratio due to the lower entropy of electricity compared to to fossil fuels. <BR/><BR/>I consume about 5 kWh a day at home, BTW. About 2/3rds of that is my refrigerator (I rent).Robert McLeodhttps://www.blogger.com/profile/05270962906437456350noreply@blogger.comtag:blogger.com,1999:blog-13900197.post-21270858858202898592007-08-23T17:12:00.000-06:002007-08-23T17:12:00.000-06:00Thank you for your response. I'll check out the pa...Thank you for your response. I'll check out the paper and references. My point with efficiency is that it seems to be improving surprisingly quickly. I recognize this translates into ancillary cost reductions but not necessarily the lowest cost per watt. <BR/><BR/>You mentioned ancillary costs of $2.50/Wp. This would go to $1.50 if you take out the inverter. This leaves roughly 50-75 cents in ancillary costs that can be cut by installing 20% vs 10% efficient panels. This isn't a whole lot... I'll give you that.<BR/><BR/>But here's what bothers me. Suntech was selling conventional panels for $3/Wp in 2004. Their panels go for $4/Wp currently. The price hike has been driven mainly by inflated raw silicon and wafer costs. When I ponder what the Wp costs would be without the inflation I'm left thinking the 50-75 cents in ancillary cost savings is very significant. Does this reasoning make sense? When I apply the learning curve idea to conventional panels it appears as though grid parity will be reached with or without thin film.<BR/><BR/>But how long will it take before the inflation goes away? And how will thin film develop? These things I don't know. <BR/><BR/>This report projects 13% efficiency CdTe modules in 5 years: http://www.nrel.gov/pv/thin_film/docs/first_solar_update2003.pdf<BR/><BR/>There are clearly applications where size or efficiency matter but 10% efficient panels are plenty good for roofs.<BR/><BR/>You mentioned it would take 3.5-7 m2 of 10% efficient panels per person to supply all our energy needs. 1.25-5 kWh per person per day isn't high enough. What did you mean with this quote?<BR/><BR/>http://www.nationmaster.com/graph/ene_ele_con_percap-energy-electricity-consumption-per-capitamrshabahttps://www.blogger.com/profile/14148254862303866175noreply@blogger.comtag:blogger.com,1999:blog-13900197.post-57774516544897505772007-08-23T12:57:00.000-06:002007-08-23T12:57:00.000-06:00anonymous said:The Hoffman citation for 33% per ye...anonymous said:<I><BR/>The Hoffman citation for 33% per year is somewhat academic - I wonder if Merrill, GS, or any of the research departments in the banks have commented in an increase in capacity over time.</I><BR/><BR/>I'm not sure that the investment banks have been tracking PV that long. The first reference is to a Merrill-Lynch report references SolarBuzz, which is a rather expensive report.<BR/><BR/>mrshaba said:<I><BR/>Could you name a few of these silly PV concepts please? What do you think of wide acceptance angle concentrators? </I><BR/><BR/>I'm not going to 'name names' without backing up my opinions (i.e. with a full post). Wide-angle concentrators are better than tracking optics (on a cost basis) but still suffer in the marginal times (overcast, mornings). As such, caveat emptor applies because if someone is relying on concentrators to boost their cell conversion efficiency, it won't perform as well in the real world as against the Air Mass 1.5 test conditions.<BR/><BR/><I>I’ve always found thin film technologies inferior because of the comparatively low efficiencies. In the next 5 years average silicon modules could well get to 20% efficiency. SunPower is already above 19% module efficiency and Suntech is advertising a similar technology being available next year. The triple junction Sharp module you mention comes in at 10% efficiency. Do you have a guess where thin film performance will be in 5 years?</I><BR/><BR/>Efficiency isn't the best metric for determining the cost per watt. It has an impact, but it's mostly in terms of the ancillary costs, i.e. the glazing, the frames, etc. It will become more important in time, but I do not see the ancillary costs dominating before PV becomes an important energy source. Thin-films have the potential to greatly reduce both the amount of raw material required and the amount of material processing required.<BR/><BR/>On the Sunpower monocrystalline cells, well, I am working on electron microscopy techniques for my Ph.D. I do a lot of specimen prep work, which involves polishing, cutting, etc. of semiconductors which is practically identical to what's required for monocrystalline silicon. It's very labour intensive.<BR/><BR/>Growing thin films via chemical vapour deposition is a lot more straight-forward and allows for vastly higher throughputs. Amorphous silicon has some serious issues but I do think you'll see better and better microcrystalline silicon. I recently saw a paper showing an electron mobility of 450 cm^-2 V^-1 s^-1 for a type of laser annealed microcrystalline silicon. Semiconductor grade Silicon is only about 3x better, so the material continues to improve. See:<BR/><BR/>http://dx.doi.org/10.1016/j.ultramic.2007.04.014<BR/><BR/>and referenes [27] and [28] therein.<BR/><BR/>Your suggestion about building roofs ready for PV is an interesting one, and maybe something Southwest states in the USA should consider. Corrosion, is probably the biggest potential drawback, since that is one of the primary limiting factors in the lifetime of PV system.Robert McLeodhttps://www.blogger.com/profile/05270962906437456350noreply@blogger.comtag:blogger.com,1999:blog-13900197.post-78346118747843398582007-08-22T10:35:00.000-06:002007-08-22T10:35:00.000-06:00Could you name a few of these silly PV concepts pl...Could you name a few of these silly PV concepts please? What do you think of wide acceptance angle concentrators? <BR/><BR/>I’ve always found thin film technologies inferior because of the comparatively low efficiencies. In the next 5 years average silicon modules could well get to 20% efficiency. SunPower is already above 19% module efficiency and Suntech is advertising a similar technology being available next year. The triple junction Sharp module you mention comes in at 10% efficiency. Do you have a guess where thin film performance will be in 5 years? <BR/><BR/><BR/>Nuclear engineers aren’t much different than MEs btw. Building and fixing nuclear plants has recently been more limited by welders than engineers. Your personnel shortfall comments talk about the top of the food chain. I’ve always seen more of a personnel shortfall on the installation end of the chain compared to design or manufacturing. The personnel shortfall issue is interesting though. You’ve got to figure that solar energy systems will become easier to install. Once solar subsidies go away there won’t be much stopping a do-it-yourselfer from installing the lion’s share of a system over a few weekends and calling up an electrician to hook up all the hot stuff. We might even see homes pre-prepared for solar with roof anchors and wiring paths etc. The unfinished roof version of the unfinished basement. This sort of development would tend to ease the personnel issue I’m worried about. <BR/><BR/>But who knows. Everything is growing so fast it's hard to tell. It's fun to extrapolate though.mrshabahttps://www.blogger.com/profile/14148254862303866175noreply@blogger.comtag:blogger.com,1999:blog-13900197.post-62322393773489271522007-08-22T09:40:00.000-06:002007-08-22T09:40:00.000-06:00Great post - very informative. Perhaps for all in...Great post - very informative. Perhaps for all involved here can we agree on some sort of highly defensible expansion curve for solar. The Hoffman citation for 33% per year is somewhat academic - I wonder if Merrill, GS, or any of the research departments in the banks have commented in an increase in capacity over time.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-13900197.post-51316232768255862632007-08-15T10:18:00.000-06:002007-08-15T10:18:00.000-06:00I've always thought of solar being the eventual wi...I've always thought of solar being the eventual winner in the energy lottery, but I am pretty dubious of some of the silly PV concepts designed to liberate technophobic investors from their money. KISS (keep it simple stupid) rules.Robert McLeodhttps://www.blogger.com/profile/05270962906437456350noreply@blogger.comtag:blogger.com,1999:blog-13900197.post-51321142786164359892007-08-15T09:06:00.000-06:002007-08-15T09:06:00.000-06:00Great article Robert. Amazingly good really. It ...Great article Robert. Amazingly good really. It seems as though some of your earlier posts were rather down on solar. Did you convert to solar in the last few years or have I misinterpreted your earlier posts?mrshabahttps://www.blogger.com/profile/14148254862303866175noreply@blogger.comtag:blogger.com,1999:blog-13900197.post-79524299524365908252007-07-14T08:31:00.000-06:002007-07-14T08:31:00.000-06:00Graphing wind installed capacity against solar PV ...Graphing wind installed capacity against solar PV installed capacity over the sort of time horizon and making the same assumptions as you have made would result in the wind power far exceeding the installed capacity of PV for the foreseeable future, in spite of it's slightly lower growth rate.<BR/>The relatively large installed based of wind compared to solar at the present time insures this.<BR/>Of course, if your statement about the lower geographical availability of wind as against solar is correct than things would change.<BR/>However, things change dramatically if the requirement to generate wind power from towers is relaxed.<BR/>Various tether-based proposals would mean that the winds of the jet stream could be utilised, which although it wanders around somewhat would provide much less intermitancy than either ground-based wind or solar power.<BR/>It therefore seems entirely possible to me that the present explosive growth of wind-power will continue, with higher altitude winds being utilised as the easier low-level resources are exploited.<BR/>So you end up with a tussle for dominance, with the areas of low wind and high solar activity such as the tropics favouring solar, but a much more closely fought battle at more northerly latitudes will better high altitude winds.<BR/>I'd be very interested in seeing the graphs of wind vs solar projected into the future, using the methodologies you employed in the original article.<BR/>Regards,<BR/>DaveMartAnonymousnoreply@blogger.comtag:blogger.com,1999:blog-13900197.post-90610689218477152562007-07-14T08:06:00.000-06:002007-07-14T08:06:00.000-06:00Only the one weak point in your argument that I ca...Only the one weak point in your argument that I can see.<BR/>That is the assumption that ancillary costs will drop at the same rate as the module costs.<BR/>In comparable industries eg computer production this just does not happen.<BR/>The price of the electronic gizmos follows Moore's Law, and drops rapidly, but the boxes to put it in, assembly and distribution costs don't.<BR/>So I can see the transition being a bit slower and stickier than you project, with ancillary costs making up a more substantial proportion of total costs as module costs drop.<BR/>It doesn't invalidate your basic thesis, that PV costs will drop to below grid prices in fairly short order, but it does affect the time scale, and perhaps indicates that technologies such as PV incorporation into roof shingles and so on may be the way to go, as the installation costs are much reduced as you need a roof anyway, although not of course the electronics and so on.<BR/>Very interesting blog - thanks.<BR/>DaveMartAnonymousnoreply@blogger.comtag:blogger.com,1999:blog-13900197.post-58549279730004150982007-07-12T07:34:00.000-06:002007-07-12T07:34:00.000-06:00In my opinion in the developing world, LEDs paired...In my opinion in the developing world, LEDs paired with solar panels could provide a cheap, sustainable light source that doesn't need a traditional power grid. In reality, solar power may not provide enough energy for ALL of our energy needs. But it could provide be a large percentage energy source. Coupled with nuclear power, wind power and water power, we could stop global warming. And as an added bonus, the U.S. with lose its dependency on foreign oil.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-13900197.post-10116375137152364452007-06-14T08:22:00.000-06:002007-06-14T08:22:00.000-06:00I found your piece absolutely fascinating and thou...I found your piece absolutely fascinating and thought-provoking. But I am very worried about your assumptions, particularly, as others have commented, on the 33% growth rate. One always has to be concerned about extrapolations where the timescales of the prediction and the data are significantly different. In the case of solar energy capture, much of the growth has been underpinned by massive subsidies by governments which are not growing to keep up with demand. In the UK, for example, the growth was fuelled by a 50% subsidy which was very recently cut by about a half - not surprisingly growth has slowed substantially. Price elasticity is a bugger, isn't it?<BR/>Let us hope that the combination of learning curve and improvements in efficiency are able bring the price down faster than the losses in subsidy. <BR/>So I think I'm pretty sceptical about solar making really big inroads since I suspect that as coverage increases, the exponential will start to flatten out.<BR/>And that's a shame 'cos I'm committed to the extent of having replaced half of my roof with PVs...<BR/>Thanks for giving me so much to think about, thoughAnonymousnoreply@blogger.comtag:blogger.com,1999:blog-13900197.post-55106382926607781252007-06-08T16:26:00.000-06:002007-06-08T16:26:00.000-06:00Thanks for the response and the links. Just to cla...Thanks for the response and the links. Just to clarify, I certainly support solar and think there should be aggressive government incentives to help develop the industry. Guess I'm just quibbling a bit with the core assumption of your otherwise excellent analysis.<BR/><BR/>My state (Oregon) will be doing its part to help encourage solar. The governor <A HREF="http://www.bendweekly.com/Statewide-News/6896.html" REL="nofollow">signed a law this week</A> that requires utilities to produce 25% of their power from "new" renewable sources by 2025 ("new" being anything built after 1995). That should give a nice little boost to the solar industry.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-13900197.post-75790917791990475142007-06-05T10:36:00.000-06:002007-06-05T10:36:00.000-06:00Orygunner:I'll try to address your points one at a...Orygunner:<BR/><BR/>I'll try to address your points one at a time.<BR/><BR/>In terms of sustaining the growth rate, I can see a few of means for continued strong demand. <BR/><BR/>One would be any sort of greenhouse gas legislation in the major economies of the world. A carbon trading system or simple tax will certainly favour solar and other renewables over coal and natural gas. Even if some nations do not get on-board the CO2 wagon, we will likely eventually see the developed nations applying a carbon tariff to imported goods. Since CO2 sequestration is largely vapourware, and it will definitely be expensive, nations, their industry and citizens, will look to solar and wind as being most competitive supply-side method of curtailing greenhouse gas emissions to meet mandates.<BR/><BR/>Another thing that could push solar is local depletion of coal and natural gas. We've already seen depletion of the coal resources in the UK and Germany, and if it continues to expand at its current rate in China we'll see it there too within my lifetime. Coal depletion is more an issue of declining quality than tonnes but it will still give solar a competitive advantage. Low-heat coal is expensive to ship over long distances. Of the major coal producers, only Australia exports a major fraction of their production.<BR/><BR/>Natural gas depletion will all have local effects even if there's large quantities available elsewhere in the world. Cryogenic liquefaction will work, and reasonably well, but it will be a lot more expensive than our current oil tanker method. <BR/><BR/>Lastly, we should not discount the impact on the developing world. Most developing nations are resource-poor, while the OECD countries are all resource rich (with the notable exception of Japan). For a lot of these countries solar will be the affordable means of electricity generation now and in the future since it won't require the massive capital investment of a distribution grid. A lot of the equatorial nations also receive the most insolation. A country like Kenya, for example, isn't going to buy anywhere near as many panels per capita as Germany but they will realize greater economic benefits from things such as indoor lighting or refrigeration.<BR/><BR/>On the environmental cost of PV, I have addressed the energy inputs previously <A HREF="http://entropyproduction.blogspot.com/2006/05/solar-payback-period.html" REL="nofollow">here</A>. <BR/>A reference in that post leads to a Dutch study on the <A HREF="http://www.ecn.nl/publications/default.aspx?nr=c06002" REL="nofollow">various inputs for solar cells</A>. You can check out the Excel spreadsheet they provide.<BR/><BR/>The land issue is addressed in the original post. I suggested 3.5-7.0 m2 of panels per person, which is relatively limited compared to the resources required to build a car or house.Robert McLeodhttps://www.blogger.com/profile/05270962906437456350noreply@blogger.com