Abdelhalim, I was reading an article that mentioned that, although the research community had a real hard time achieving efficiencies beyond 20%, almost 2x efficiencies were achieved at the CIRCUIT LEVEL by understanding how to properly control the solar panels.

This was the motivation behind my answer above about WEAR LEVELING. These circuit control techniques include things like, 1) using a smaller portion of the cells, when the load demand is low, 2) the wear leveling I talked about above, 3) short-circuit, overload and other types of protections, current limiting, etc ...

This is an amazing result showing that, while the materials people will do their thing, and try to improve the cells themselves, we can achieve much better results NOW by simply knowing how to use these panels through proper CIRCUIT and CONTROL techniques. You can get 100 panels and use NAIVE circuits and naive control mechanisms, vs. ones incorporating all of these fancy control techniques along with their proper circuits. The difference is going to be 10 year lifetime, vs. 20 years ! The cost difference will be $2000 vs. $2050 !!! Because, the additional circuitry is minimal, it is all control software that makes the difference ...

Efficiency

apart from efficiency .... I think the solar panel manufacturing cost is also a key element, so you could think of it as a "Power to Cost" ratio.

improvements in Semi conductor materials and better manufecturing techniqus will increase efficiency and life span of the Solar pannels(However costs must be taken into account )

edf tells us that "in a typical UK location the sun only shines for 34% of a typical day". That affects the reliability of the energy transfer. But I guess you are more interested in 'reliability' in the sense of how long it stays close to its peak condition, and that is a function of production quality.

Dear Colleagues,

Thank you for trying to answer my question about the reliability of the solar panels.

The reliability of any product is determined by its life span in which it remains full functioning.Longer life is very required for the solar panels as their initial cost is relatively high and longer life grantee their pay back and increase their profit. Indeed , when designing , manufacturing and testing solar panels one wish they sustain very long life time. Great efforts have been spent by scientists and manufacturer to make the panels more reliable by making them more durable to the hard environmental conditions. In their use in the field solar panels are subjected to hard weather and environmental conditions. In spite of stringent design and production conditions during manicuring, the environmental cyclic stresses cause irreversible changes in the solar cells that partially or totally malfunction them normally gradually with time.

The solar panels are subjected to the following environmental effects:

-Wind stresses

-Cyclic thermal stresses from day to night and from summer to winter.

These two major cyclic stresses may lead to fractures of the cells and or their demetallization. These are two main observed defects caused by the repetitive environmental effects. Advancements in the panel design and making led to greatly relief of the stresses themselves on the solar cells inside the panels. The solar cells are embedded in relatively soft plastic potant in the panel.

Thanks to these advancements, the reliability of the solar panel reached to about 25 guaranteed lifetime.

I think the use of the concept of repairable panels will share even more in enhancing the reliability of the solar panels.

Please see the link:

https://www.researchgate.net/publication/232709358_A.A._Elshazly_A._Zekry__M.A._El-Koosy_and_M.s._Farid_Three_flat_plate_terrestrial_module_fabrication_and_testing_Renewable_Energy_23_2001?ev=prf_pub

Article A.A. Elshazly, A. Zekry , M.A. El-Koosy and M.s. Farid, “Thr...

Nice question. I am not in the field of solar panels but the answers of you Prof. Zekry and the colleagues are enlightening on the issue. Nevertheless, it came to my mind the impact of environment and the nice argument of Vasilo in the in the previous comment (that Solar panels might be degraded when they are mostly useful). I like to discuss the issue differently. From mathematical point of view, the reliability of any device, which can be defined as the probability of working time, is dependent on the number of components/parts and the reliability of each component. The larger number of components the less the total reliability of the device and the reliability of the whole device will be less than the smallest reliability of any one part. Nowadays, reliability of electronic devices are higher due to improvements in the reliability of the components themselves. There are less or no reliance on determinate crucial parts that can easily fail. Examples, no more hot parts like the thermionic valves, less and less moving/ mechanical parts or switches, less high voltage pats like the high voltage components needed to drive a CRT monitor and less reliance on electrolytic capacitors for filtering in power supplies due to the use of high frequency SMPSs instead of 50Hz power supplies. The second source for increased reliability is the reliance on highly reliable integrated parts. Hence improvements in VLSI technology and bread boards manufacturing produced highly reliable sub-boards and devices. I believe solar panels are electronic components that can enjoy the benefits of improved reliability of all other electronics and a highly reliable solar panels can be produced. It is the job of those in this industry to make use of these advances to attain the required reliability. From the early times of electronics there was military standards of components more reliable components and electronics that are designed to last are there, let be one pioneer example the "Voyager" space probe of NASA.

As I said I am not in this field but wonder why long lasting solar panels can't be done (except of course that they don't want them to last for commercial purposes). Thanks. @AlDmour.

Abdelhalim, I know that, there are circuit-level measures you can take , such as OVER-PROVISIONING the number of panels. In other words, if you need 100 panels, build the circuit with 105 to correct for issues in the following 25 years. The circuit (though simple), is built to automatically activate the inactive cells in case of failures, since it is very hard to know which ones are GOING TO fail in the following 25 years.

So, one natural question is : WHY NOT ACTIVATE THE EXTRA 5 CELLS AT THE BEGINNING ? It is because, the AGING of a cell is proportional to temperature IF THEY ARE BEING USED. The temperature has pretty much zero effect if they are not being used. So, those five cells could be switched over for the first ones that fail (or, gradually fail).

This WEAR LEVELING is a very similar concept to how SSD's work with additional cells. The additional cells are used to replace the first failing ones. For example, in a 240GB SSD, you really have 256GB, but, the 16GB (about 6%) is left for redundancy to address the WEAR LEVELING issue.

This is a circuit-level technique that won't really increase the lifetime of solar panels, but, will allow you to have 25 years from 105 over-provisioned panels instead of 20 years from 100 panels without over-provisioning ...

I would like to thank my colleagues Vasile, Ismat and Tolga for their answers that enriched the discussion. It is the system reliability that has been increased by switching additional panels instead of the panels that have the defective cell or cells.According to Tolga example one may extend the useful life time the array.

On the other side, according to vasile , it may be very hard to increase the life time of the high quality panels appreciably beyond twenty years because of cell heating.

Ismat is the opinion that the reliability of the solar panel could be increased by importing the high technology of the integrated circuit.

Thank you again, let us wait an other answer from a colleague who can have some more excessive insight into the problem Or we may wait the future. It certainly has an answer.

Abdelhalim, I was reading an article that mentioned that, although the research community had a real hard time achieving efficiencies beyond 20%, almost 2x efficiencies were achieved at the CIRCUIT LEVEL by understanding how to properly control the solar panels.

This was the motivation behind my answer above about WEAR LEVELING. These circuit control techniques include things like, 1) using a smaller portion of the cells, when the load demand is low, 2) the wear leveling I talked about above, 3) short-circuit, overload and other types of protections, current limiting, etc ...

This is an amazing result showing that, while the materials people will do their thing, and try to improve the cells themselves, we can achieve much better results NOW by simply knowing how to use these panels through proper CIRCUIT and CONTROL techniques. You can get 100 panels and use NAIVE circuits and naive control mechanisms, vs. ones incorporating all of these fancy control techniques along with their proper circuits. The difference is going to be 10 year lifetime, vs. 20 years ! The cost difference will be $2000 vs. $2050 !!! Because, the additional circuitry is minimal, it is all control software that makes the difference ...

Thank you Tolga for clarifying the role of sophisticated control methods to increase the photovoltaic system reliability. This may serve in the direction of Ismat idea to use the advanced control devices produced by the integrated circuit technoloy to control thr operation of the solar panels.The dc to dc converters and the inverters used in the photovoltaic power generators are controlled to operate the photovoltaic array at its maximum power point. One may also control the operation of the solar cells to relief their electrical stress and hence they can live longer.

On the other side, if the solar cells operate at one sun, the maximum power dissipated in the the cell is equal to the output power of about .5V* 35mA per cm^2=.175 watts. So, the electrical heating is smaller than the direct solar heating.Additionally, in case of solar cell , the solar cell works as a constant current source and so the short circuit current is limited by the incident solar radiation. It is a safe working operating condition for the solar cells. It is believed that the severe environmental conditions are the main cause for the failure of the the solar cells and their interconnects. Circuit level solutions to the reliability issues over come the defective cells by adding redundancy cells and modifying the inter connect configuration to substitute the defected cells. One of the simple solution is to connect a backward diode to every cell to by pass it in case of getting failed.

Tolga your comment was useful for me and i think for the other colleagues.

Thank you again.

Vasile, although I have a bunch of papers, I am a very hands-on guy, and I agree with your ideas very much. I completely believe in nicely combining theory and practice.

And, the saying that you introduced : "simple things works better" is my RESEARCH PRINCIPLE :) Whenever you complicate things, some weird thing pops up somewhere that messes something up :)

The failures you talked about can be managed with good control software (even very simple ones) :) because, they can be detected. However, when the detection circuitry itself fails, you are toast :) So, there is a lot of research out there for making the detection circuitry itself redundant, by making redundant copies of it etc ... But, in the end, exactly like you said, simple things work better.

So, make the circuitry out of better components, use the most reliable solar panels, and if the failures are small, you can handle them. If 50% of the panels are going to fail in the first 5 years, then, possibly, your problem is not the circuitry, get new RELIABLE PANELS :)

This discussion is getting richer and richer everyday ^_^

I think some research work is going on in the field of failure detection for solar panels. Prof. Tolga, "Redundancy" is a key element in the field of fault detection and tolerance control. Instead of using several detection circuits, you can implement redundancy in the software level, such as using state estimators or observers.

Mohammed, software-based redundancy measures are great, since they are FREE :)

However, if the hardware fails, software is DEAD ! If the software is not working, since the underlying hardware is DEAD, software can't do anything :)

On the other hand, hardware based redundancy is good, since FAILURE is generally on the hardware, and the redundant pieces of hardware can detect each others' issues ...

So, the perfect solution is to employ BOTH !

In other words, have redundant hardware, with its own control software ... However, each redundant piece is REDUNDANCY-AWARE, i.e., its gotta communicate with the other redundant pieces and know that, they have to take over, if the other parts are dead.

BTW, this is very similar to how Google File System (GFS) works :)

The trick is to assume that, ANYTHING CAN DIE. and, design a system, where there is no SINGLE POINT OF FAILURE. No different with solar panels.

Ironically, since a microcontroller costs $2 these days, far cheaper than a solar panel, it is not terribly difficult to add a microcontroller-based "heartbeat checker" into each panel, and have it measure some vitals. Even if these little microcontrollers do not do anything, they can ALERT a bigger authority, such as the HARVESTER.

Clearly, it costs a lot to repeat the harvester, with the large MOSFETs and POWER DIODEs etc ... but, it costs a negligible amount to add teeny little heartbeat checkers . It is amazing what a $2 PIC microcontroller can do, with 14 channels of ADC, temperature sensor, and the ability to shut itself down in case of a brownout ... you name it ... and, the communication among these microcontrollers can be either wireless (like Zigbee, which is pretty immune to failures), or, I2C, SPI etc ... which are fairly immune to failures, since they were designed for very harsh environments ...

Vasile, very interesting practical feedback from the field. Now you are handing over the culprit to the court. It seems that failures are caused by some very basic mechanical / soldering links,..etc according to you. Again, I can't discuss that but I can say that If technology can/ will provide more reliable links/ lead soldering we expect higher reliability solar panels. This can be at cost rendering the total renewable energy "more story than reality" as per your saying above. Mohammad and Tolga provided redundancy as a solution which is generally an approach for increased reliability at the cost of complexity (control complexity) while jeopardizing reliability (as per Tolga who at the same time fears control failure). As you like simple things (like us as well), a simple approach to higher reliability is the provisioning for replaceable parts hence getting a higher percentage of working time over the 25 years lap. Of course, this might depend on lots of factors and you tell whether this is possible /economical or not. Thanks. @AlDmour.

George, BETTER THAN WORST CASE DESIGN is possibly an interesting research topic in many fields. My question is this : Consider two different approaches:

OPTION 1) Design solar panels that has a 0.1% long-term defect rate over 25 years (meaning, 0.1% of the panels fail in 25 years), and costs $300 per panel. I am completely using made-up numbers to make the point ...

OPTION 2) Design solar panels that have a 10% long-term defect rate over 25 years and cost $30 per panel.

Which option is better ? Clearly, depends on what your goal is. OPTION 1 almost eliminates a sophisticated CONTROL-CIRCUITRY that checks the vitals like you mentioned ...

But, OPTION 2 can be deployed with a 15% over-provisioned setup, bringing the REAL COST to $34.5 per panel. Well, I now, have to have a sophisticated circuitry to do all sorts of cool tricks to disengage the dead panels etc ... Ok, throw another $5.5 for that. Total is $40. That's still far better than $300 !

My point is, if the additional reliability is going to come at the expense of a disproportionately increased cost, it's not worth it, since I can make up for the reliability issues by over-provisioning and taking care of it on the BACK END (i.e., by using sophisticated control circuitry).

This concept is very similar to Error Correction Codes in memory chips. OPTION 1) Do you use extremely expensive memory chips ? or, OPTION 2) use ordinary chips bundled with Error correction codes that can not only DETECT, but CORRECT errors (like SECDED - Single Error Detection - Double Error Correction).

On the other hand, if you can give me the OPTION 1 solar panels at the same price as the OPTION 2 solar panels, well, OPTION 2 will become obsolete, and you possibly have a disruptive technology anyway ...

Do you agree ?

I have not completely studied the above replies. However, every one in the SPV industry knows that for the SPV modules two IEC standards are used, these are IEC 61215 for monoand multi crysatlline silicon modules, and IEC 61646 for thin film modules. For concentrator modules 62108. However, these standards can be viewed as general guidelines. here a set of 8 modules tested as per a specific sequence defined and specified level of harshness given in the standard, if the level of degradation is after these tests is

Dear Colleagues,

After this rich discussion i would like to invite you to read the book chapter in the link: https://www.researchgate.net/publication/323309527_Solar_cells_and_arrays_Principles_analysis_and_design

It contains a section about the reliability of the solar cell modules.

Best wishes