Solar concentrators
 
  Focusing the sun

Solar concentrators have two basic types. Rectangles and dishes. Both use a parabolic mirror. The Rectangles make the best use of the available space and focus the sunlight along a straight line while dishes focus the sunlight to a single point. Both can produce very high temperatures although the dish design has the clear edge, producing up to 3,000K. Both track the sun's movement across the sky. Both produce a lot more energy than solar electric panels per square meter but niether work in cloudy conditions. There is also such a thing as a solar furnace. These tend to be large complex structures in which a large number of (usually flat) mirrors, track the sun such that the sunlight focuses on a central point, usually at the top of the tower. Solar furnaces are used to produce megawatts of electricity while the rectangles and dishes typically produce a few kilowatts.

Operation of solar concentrators

In general solar concentrators use the high temperatures generated at the focal point or line, to boil water into steam which is then fed to a steam turbine and made to turn a generator to produce electricity. They don't have to be used this way though and there are examples of solar concentrators used to drive high temperature chemical processes such as the thermal dissassociation of methane or even water, producing hydrogen without electrolysis. The thermal dissasociation of water shows considerable promise because if the objective is to use the electricity normally produced by steam turbines to electolyse water, then this approach has much greater yeilds. Howevert a lot more R&D is needed to perfect the technique so for now we will only discuss the solar concentrators that produce electricity.

Efficiency: Dishes vursus rectangles

Of the sun's energy that falls onto a solar reflector, about 20% is in the form of useless scattered light. Of the remaining 80%, some 3% is absorbed by the mirror, a very small amount by the air between the mirror and the absorber and a significant amount is radiated back by the absorber. The extent to which re-radiation occurs is mostly due to the temperature of the absorber. The blackened surface of the absorber is a very good radiator of energy as well and the extent to which selective coatings can mitigate against this is limited. The maximum temperature the absorber can reach depends on how concentrated the sunlight is. A very high quality dish reflector could concentrate sunlight up to 10,000 times. If ordinary sunlight can heat things up to say 60oC (330K) and the radiated energy rises as a power of 4, such as reflector could produce temperatures 10 times this, 3300K or 3000oC. At the maximum temperature the absorber is radiating as much energy as it recives leaving no energy available to be converted into anything useful like electricity! To actually get anything out of the concentrator we need the temperature of the absorber to be lower. For steam we only want 450oC or 825K, one quarter of what a high quality dish can produce. At that temperature the re-radiation would be a mere 1/4096th of that absorbed leaded to the question os why even bother to mention it!

The reason why it matters is because high quality dishes are very expensive. They need an awful lot of polishing. To be viable we want something of reasonable quality we can mass produce. Such dishes are not going to get anywhere near a concentration ratio of 10,000 and a more typical value would be around 500. This would still produce a maximum temperature of some 1650K but as this is only double our hoped for operating temperature, one 16th of the energy is re-radiated and that is a significant loss.

And it is much worse with the rectanglar design as these have a focal line rather than a focal point. With the quality of reflectors that are viable to mass produce, concentration ratios are going to be more like 60. This corresponds to a maximum temperature of about 1300K and if we still want the same operating temperature we would lose about one 6th of the energy.

Design constraints

The larger the dish or rectangle the more sunlight it can capture but this relationship is broardly linear. What mitigates against size are structural considerations. The larger the mirror the more support it needs and this is not linear. It is more costly to build 10 small concentrators than 1 concentrator 10 times the area. Conversely, if the area of the mirror is too small it will not generate enough steam to efficiently run even the smallest of turbines and using a piston type steam engine is at best only half as efficient as a turbine although it does operate at lower temperatures mitigating the re-radiation losses. Also in favor of size is the tracking device. The dish needs the same complexity to track the sun regardless of how big it is.

So ideally, we need to miniturize the steam turbine with the minimum impact on thermal efficiency and reduce the cost as far as possible and simplify the tracking as far as possible without compromizing tracking accuracy. Here's what we came up with ...

The steam turbine component has two options. The first of these is the standard steam piston but with the cylinder surounded by a partial vacuum. The stroke is arranged such that at the end, the pressure is much lower than one atmosphere and yet is still doing work. Instead of being expelled from the cylinder and into a condenser at over 100oC it comes out at about 70oC. This increases the efficiency of the cycle because thermal efficiency is a function of (Tin -Tout)/Tin. The second option is a steam concentric rotary engine, again with a reduced temperature output. We still have a little R&D to do on this so that the sealing problems rotary engines are notorious for, can be solved the way we think they can. Even with the sealing problems the performance of the prototype rotary is outstanding. It produces more energy per unit weight and per unit volume and at RPM's more akin to a turbine than a steam piston and there is no vibration and no crankshaft so fewer moving parts. And it can be cheaply scaled down to the size of a cotton reel whilst still retaining the bulk of it's thermal efficincy of >35%. More will be published on this in due course.

The prototypes to date

So far we have two prototypes, a 3 meter dish and a 6 meter dish generating 5KW and 20KW of thermal energy with output temperatures of between 450oC and 480oC, corresponding to 1.5KW and 6KW of electrical output respectively. We believe we can boost the electrical output by fine tuning the heat engine but note that the output we do get in the trials is easily enough to justify mass production. Both dishes are identical except in size, both use an aluminium reflector coated with clear plastic to prevent oxidization. The 3 meter dish has an aluminium thickness of 2.5 mm and no support while the 6m dish has the same but is supported with a steel frame, demonstrating the earlier point that larger concentrators cost more per unit area.

The tracking system on the prototypes consist of thermocouple embedded in the absorber, a free standing photocell and a ring of photcells surrounding the absober. These provide data to a microcontroller which drives two small stepping motors, one for setting the latitude, the other for longtitude. The transition with time, of the sun accross the sky is something that can be programed into the microcontroller but this presents complications. If we were to rely on this alone, the time would have to be syncroized with the actual time as left to itself, will either gain or lose time. Worse still, we would have to know exactly where each dish was to be located as the exact transition path depends on location. While this could be done it would complicate the operations of both suppliers and users of the concentrators. It is therefore of some importance that concentrators are self calibrating. This is where the therocouple and the photocells come in. The role of the free standing cell is to determine if there is sufficient light to operate the dish. This is not just a matter of it being daytime but being sunny as well. This is important because it tells the microcontroller how to interpret data from the thermocouple and the ring of cells. If all the cells in the ring are registering low levels of light and the thermocouple is cold or cooling, the output of the free standing cell indicates if this is due to the sun having been obsured by cloud or the alignment being out by enough to miss the absorber altogether. If the sun has gone in nothing needs to be done until it comes out again but if it is sunny, the concentrator will waggle around until one or more cells in the ring are hit with the focussed sunlight. The concentrator will then have a fix on the sun and will know it only has to move a little to the East or West or up or down and the focussed sunlight will be all on the absorber. Once properly aligned the thermocouple will start warming up and confirm this.

Plans for mass production and expected prices

We plan to begin shipping dishes in March 2008 at prices of around $1,500 for a 3 meter dish, $3,000 for a 4.5 meter dish and $6,000 for a 6 meter dish. The dishes will at this point be supplied as self build kits with the parabolic reflectors divided into sections that clip together. The price does not include the power management unit that interfaces to the grid or the cabling to connect the dish to the power management unit. This is because no assumptions can be made about where the dishes will fit in to your overall energy solution. The reason it will be 2008 rather than sooner is because there is still outstanding R&D to fine tune the design, much detail concerning who will manufacture what and where and of course, more extensive field tests. It is anticipated however that some dishes will be shipped as part of bespoke energy solutions as early as June 2007. The first major bespoke solution will be at our own site.

Back to top of page