AC or DC Solar Panels in Extreme Heat?

[russ]

Changing the microinverters on the roof might get tricky - Naptown or KRenn can probably give a comment based on experience for that point.

If there are any shade problems - even just that of wires - microinverters have an edge and probably on warranty (that we will know for sure in 20+ years.

[Taylori]

Might make sense to you but it is meaningless - by that definition any panel using a micro inverter would be an AC panel

You obviously know a lot more about solar panels and systems than I do. I just want to know which panels, if any, are likely to be more durable under the extreme heat of the desert: (1) DC panels with the inverter inside the garage or (2) panels with a microinverter built into each panel. Which would you chose, cost and warranties being equal? The exposure is ideal, directly south with no shade.

[russ]

You obviously know a lot more about solar panels and systems than I do. I just want to know which panels, if any, are likely to be more durable under the extreme heat of the desert: (1) DC panels with the inverter inside the garage or (2) panels with a microinverter built into each panel. Which would you chose, cost and warranties being equal? The exposure is ideal, directly south with no shade.

A couple of guys in the business are from Arizona - can't get too much more extreme in the US - they should provide the best answer.

[Naptown]

This is actually easier than you would think and in reality safer.

Try to avoid the Westinghouse panels. If a panel or inverter goes bad in the middle you may end up taking the majority of the array apart to get to the problem because of the way they go together.

If you use a typical top down racking system from say Unirac or any of the others you simply remove 4 bolts and the panel will come out. (Remember if there are weeb clips in there you must replace any that were loosened)

The inverter will remove from the racking with one bolt then simply unplug the panel MC4 connectors and the inverter cable from the engage cable.
If you turn off the AC where it ties into the service panel the engage cable and inverter will be inert. Since these use 60 cell panels the highest DC voltage is about 37 volts.
A lot less than the 300-500 on a string inverter which cannot be turned off unless you cover the entire string with no light getting in to them.

Lastly if there ever is a problem with a panel or inverter finding the offending part is very easy with the enphase system. The monitoring system will tell you exactly where and what the problem is. If a panel goes bad on a string inverter you are on what could be a very long easter egg hunt tracking down the problem.

Should you decide to go the string inverter route I would suggest you look at the Fronius The modular design makes replacing the inverter portion very easy as this simply unplugs from the power and cable connection portion. I also happen to like their 3 stage arrangement to keep the inverter on the top end of the efficiency spectrum.

[Sunking]

Now that we can boost and drop DC voltage, it wins for a different reason. To transmit an average of 1000 amps in an AC system you have to transmit a peak of 1400 amps, and to transmit an average of 12,000 volts you have to have peak voltages of 16,000 volts. However, to transmit 1000 amp/12,000 volts DC you need to handle only 1000 amps/12,000 volts. That means that towers have to be higher, wire spacing has to be greater and conductors have to be thicker to handle that power level at AC.



That's a good argument for a HVDC array. Les stuff on the roof.

That is kind of half truth, but is not how it is done with HVDC. I have built many of them starting in TX as TX does not allow any AC lines in or out of the state. TX is not part of the national grid. The AC high voltage is just simply rectified using very ancient selenium rectifiers as it is the only semi-conductor that can handle the very high voltages of 1 million volts. At the receiving ends in converted back to AC at the same high voltage. So if it starts out as 750 KV it is converted to 1 MVDC, and converted back to 750 KV. Al the HVDC transmission lines are point to point transmission and bipolar. Very expensive to implement but saves the utilities huge line losses so they pay for themselves very quickly.

For and the foreseeable future secondary transmission and distribution will always be AC. DC is just to expensive and difficult to regulate and control.

FWIW Manhattan Island was totally DC up until I think 2007 when Con Edison turned off the last DC fedders. Many of the sky scrapers in NYC still have DC light, heat, air, and mechanical like elevators. Those building had to either convert to AC, or most bought rectifiers. NYC is one of a kind.

[inetdog]

The AC high voltage is just simply rectified using very ancient selenium rectifiers....

You are bringing back memories of the smell of melted selenium rectifiers on my HO gauge model railroad!

Is the fact that their reverse losses are so high actually a help in balancing them in a high voltage series connection? That and the fact that they do not break down catastrophically when their maximum voltage is exceded?

[Sunking]

You are bringing back memories of the smell of melted selenium rectifiers on my HO gauge model railroad!

Is the fact that their reverse losses are so high actually a help in balancing them in a high voltage series connection? That and the fact that they do not break down catastrophically when their maximum voltage is exceded?

You are getting into th emanufacturing side which is not my area of expetise. What I can tell you is they use many rectifies in series to build up the voltage and have improved reverse losses.. Most of what i worked with 69, 138, and 345 KV on various points on the TX border where they interchange with out of state energy providers.

The out of state comes into the substation at 345 KV, changes to DC, connects to TXU, and then converts right back to 345 KV to go into or out of state. They are bi-directional interconnects. TX rarely ever buys power, we export power mostly to NM, LA, OK, AR, and AZ.

I have not personally worked with the long line 750 KV and 1 MV systems. Not a lot of those around. None in TX that I can think of except maybe at some of the Mega Wind Farms out in the Panhandle and West TX, don't get in on that action, just a smaller 200 MW wind farm around Muenster TX.

[inetdog]

You are getting into th emanufacturing side which is not my area of expetise. What I can tell you is they use many rectifies in series to build up the voltage and have improved reverse losses.. Most of what i worked with 69, 138, and 345 KV on various points on the TX border where they interchange with out of state energy providers.

Along the lines of future improvements (As Mayor Richard J. Daley of Chicago once said "I see a bright future ahead of us once the technologists have learned to harass [sic] the atom.")* when superconducting transmission lines become practical, they will be limited to DC.

* Quote from the Little Green Book, Quotations from Chairman Daley.

[Sunking]

when superconducting transmission lines become practical, they will be limited to DC.

Wished I had a patent on room temperature super conductors. I'd buy me a little island in the South Pacific somewhere and throw the most of the inhabitants off except a few to run stuff. An Island like Australia or something.

[SoCalsolar]

No consensus? Or no opinion?

Doesn't seem like there is any consensus among the club here on which setup would work better. I think however if you posted the pricing for each system, AC SunPower or the string CS panels you would get more of a consensus. I'm curious to the responses you will get when you reveal the difference in price.

[inetdog]

Wished I had a patent on room temperature super conductors. I'd buy me a little island in the South Pacific somewhere and throw the most of the inhabitants off except a few to run stuff. An Island like Australia or something.

A group of grad students at Stanford, working in the low temperature physics group, once did a skit on a proposed new grant to build a 12' x 15' x 8' foot box, insulate it and cool the entire assembly to near liquid helium temperature. The idea was that if they could not raise the transition temperature of the superconductor, they would lower the temperature of the room. The punch line was "...at that temperature, even peanut butter is a superconductor."

More seriously, the need is not for a room temperature superconductor, it is for one which does not lose its superconductivity when placed in a strong magnetic field (like when the wire is carrying current. )

[russ]

This is a good point. I was also interested in how much space they take on your roof, as SunPower claims that their panels take up less space, so you can add more/put more.

There is a big difference between thin film and mono or poly - there you would see the difference but the difference between 16% and 17% won't amount to much.

Mostly salesman's chatter.

[SoCalsolar]

my guess is 6-10 panel difference

My guess is a 6-10 panel difference ( 8-12k system size) they seem to spin their meters pretty hard out in Palm Springs and the "AC" panels are a touch over 19% efficient compared to Canadian S. which are between 14.6%-almost 15.8%. If space is the major concern SP has a panel that is a smidge over 20% efficient but those are not "AC" just regular "DC" but they are half the cost of the "AC" panels. Still waiting for the cost of the "AC" panels I want to see the talking heads explode.

[billvon]

The AC high voltage is just simply rectified using very ancient selenium rectifiers as it is the only semi-conductor that can handle the very high voltages of 1 million volts. At the receiving ends in converted back to AC at the same high voltage.

Yep, converting to DC is the easy part. Going back to AC is the hard part.

In the olden timey days you could do both with a motor/generator set (this is still how some older equipment, like elevators, are powered.) Induction motor drives DC generator and vice versa. For power transmission, though, the cost is prohibitive.

First mercury arc rectifiers came along which allowed rectification to DC without the MG set. This allowed conversion of AC to DC for things like electric trains. With great difficulty, a version of these (mercury arc valves) could also be used to convert DC to AC, and some early HVDC systems used these. Next came semiconductors - first diodes, then thyristors. Thyristors were critical because they allowed relatively easy switching of high voltage DC power, which allowed conversion back into AC. That's the first time that DC transmission really became practical.

For and the foreseeable future secondary transmission and distribution will always be AC. DC is just to expensive and difficult to regulate and control.

Buildings are already starting to convert to DC power to reduce costs and increase efficiency. Data centers especially benefit from easily-paralleled, redundant HV DC supplies. ETSI now has a 400 volt DC standard that's starting to be used in-building.

For external local distribution AC will probably reign supreme for at least the next 10 years. Once DC/DC converters become cheaper than transformers (and within 10 years they will likely be) then you'll see a move away from AC to DC to save money; every gauge you can drop in wire size saves a utility hundreds of thousands of dollars.

[inetdog]

Buildings are already starting to convert to DC power to reduce costs and increase efficiency. Data centers especially benefit from easily-paralleled, redundant HV DC supplies. ETSI now has a 400 volt DC standard that's starting to be used in-building.
.

APC has put out a white paper on this subject which concludes that low voltage DC (48 volt telco style) does not make sense economically, while high voltage DC (300-500 volt) is competitive but still does not have a clear advantage.

Switching remains an issue among other things.