Renewables: "talking 'bout my generation"

Oscargrouch

jackieblack wrote: »

Interesting

My understanding was/is that a smaller inverter will run more efficiently during periods of less than maximum generation, (which are likely to far exceed periods of maximum generation) thus compensating for the slight reduction in output during the (far fewer) periods of maximum generation.
Is this not correct?

Fronius tech data

I am not really bothered, my install forecast was for 2,046 kWh a year and the system achieved 2,458 kWh; so no complaints from me...:D

spgsc531

Martyn1981 wrote: »

I was so concerned at those posts that I sought clarification from SMA themselves, but the 'twist' Graham had used made it tricky for them to act, here is what they said to me:-

"I have read the various posts and decided its worth following up but as Graham states what is written in the manual is correct there seems we cannot do a lot about it, however if he wants to continue SMA may write him a nice legal letter."

grahamc2003 wrote: »

Holy Exploding Undersized Inverters Batman!

I think you're confusing the voltage that the whole set up is designed to deliver, and the wattage that the inverter can cope with. courtesy of our resident solar guru (and no, it makes no sense whatsoever, but that never stopped him).

The following comes from Chapter 2 - Safety Information Page 9 of the Sunnyboy sb1700 Installation manual.....


'Overvoltage on the DC side can even cause dangerous explosions of the Sunnyboy input capacitors and explosions of the electrolytes coming from the exploded capacitors'
courtesy of Sunnyboy.

Read all about it starting at 1693 in the following thread

https://forums.moneysavingexpert.com/discussion/comment/51280983#Comment_51280983

Oops! Looks like SMA Legal Department might be getting in touch. :doh:

grahamc2003

jackieblack wrote: »

My understanding was/is that a smaller inverter will run more efficiently during periods of less than maximum generation, (which are likely to far exceed periods of maximum generation) thus compensating for the slight reduction in output during the (far fewer) periods of maximum generation.
Is this not correct?

Fronius tech data

Up to a point, yes.

There are a couple of levels of overvoltage protection, and obviously the protection they give is limited - you can't simply stick whatever DC voltage you like and expect it it handle the situation, and you are in the realm of exploding inverters if anyone tried to stick a 100kW array on a 4kW inverter.

The first level of protection is the tracker itself. While it usually tries to set the panel voltage to that which produces the maximum power at any instant, it will moderate the voltage away form the max power point at higher voltages.

When the ability of the tracker to hold the voltage down is exceeded, the power is absorbed by varistors, which are simply voltage sensitive resistors wired to the dc input. Usually their resistance is very high, but as the voltage rises the resistance decreases, and heat is dumped by them, and the voltage fed to the rest of the inverter is lowered. If the voltage rises further, then more heat is dumped, protecting the electronics of the inverter in one respect (from overvoltages), but introducing heat, which lowers the life of electronics longer term. If the voltage rises further beyond the spec of the varistors, then the ability of them to protect the inverter is breached, and they, or the electronics following, are likely to explode/melt or othrwise be ruined. Obviously, the dc voltage protecting isn't limitless as the dynamic duo thought, and only offer protection to volatges perhaps double the spec of the inverter. (Not to be confused with transient high voltages lasting for milliseconds such as lightening effects, which may well be handled).

So stick a 100kW of panels on a 4kw inverter and when the sun comes out - pretty obviously - the inverter would be wrecked and possibly explode, just as the warnings on inverters say they will.

If you have say a 4kw of panels feeding a 3.5kW inverter, then the overvoltages there are likely to be handled invisibly, and you're likely to get more generation overall, with the earlier and later synching of the system likely to exceed the loss due to the voltage protection losses above 3.5kW. But, the downside are heat is being dumped into the inverter, which has long term reliability effects, and the varistors finite life and efficiency is decreased too. With time and use, varistors charcteritics change such that the low voltage high resistance lowers, meaning they are dumping small amounts of heat even at low voltages. So the lower life and efficiency of the varistors and the inverter itself have to be netted off from the generation gained by underspeccing.

KevinG

Pretty good day yesterday despite quite a lot of cloud later on. 8.56 kWh, 4.28 O's

Dave_Fowler

Grahamc correctly identified the problem which may arise if an array is not matched correctly to the inverter.

If the number of panels in a string is too great then there is a danger of over-volting the input to the inverter. (By the way, the voltage sensitive resistors are connected across the DC input and are of very high resistance until the maximum input voltage is exceeded).

However, all panels have a specified maximum open circuit output voltage. Provided that the total open circuit voltage of the panels connected in series in a string does not exceed the input voltage of the inverter then no damage will be done by having any number of panels connected to the input of the inverter as a number of parallel short strings. A 100kW array could be connected as 40 parallel strings of 10 (250W) panels.

Most panels produce 80% of their final voltage at quite low light levels provided the current is kept low. The effect of having a lot of panels connected in short strings and the strings connected in parallel would be that the inverter will have a greater output at lower light levels. The inverter would take only a small current from each string.

Under bright conditions, the maximum power output of each panel would not be taken as the inverter would limit its DC input current. The output voltage of each panel would rise towards its open circuit voltage but there would be no danger to the inverter due to an over-voltage as the number of panels in each string has been limited to prevent this from happening. PV design software checks that the open circuit voltage of a string can not damage the inverter and then recommends a size of inverter to match the number of panels.

It is probably not cost-effective to have a grossly oversized array of panels connected to an inverter due to the cost of the panels and the limited output from the inverter. In this country it is rare that the sun is shining brightly whilst the temperature is low so it is reasonable to design an array with panels rated 5% - 10% percent greater than the inverter. Perhaps as the cost of panels reduces, this ratio could rise to a higher percentage. Just think, if the cost of the panels were low enough, the output from a solar farm could be constant during all daylight hours if the number of panels grossly exceeded the inverter output (That's just a tongue-in-cheek joke, but who knows!)

Dave F

zeupater

Dave_Fowler wrote: »

Grahamc correctly identified the problem which may arise if an array is not matched correctly to the inverter.

If the number of panels in a string is too great then there is a danger of over-volting the input to the inverter. (By the way, the voltage sensitive resistors are connected across the DC input and are of very high resistance until the maximum input voltage is exceeded).

However, all panels have a specified maximum open circuit output voltage. Provided that the total open circuit voltage of the panels connected in series in a string does not exceed the input voltage of the inverter then no damage will be done by having any number of panels connected to the input of the inverter as a number of parallel short strings. A 100kW array could be connected as 40 parallel strings of 10 (250W) panels.

Most panels produce 80% of their final voltage at quite low light levels provided the current is kept low. The effect of having a lot of panels connected in short strings and the strings connected in parallel would be that the inverter will have a greater output at lower light levels. The inverter would take only a small current from each string.

Under bright conditions, the maximum power output of each panel would not be taken as the inverter would limit its DC input current. The output voltage of each panel would rise towards its open circuit voltage but there would be no danger to the inverter due to an over-voltage as the number of panels in each string has been limited to prevent this from happening. PV design software checks that the open circuit voltage of a string can not damage the inverter and then recommends a size of inverter to match the number of panels.

It is probably not cost-effective to have a grossly oversized array of panels connected to an inverter due to the cost of the panels and the limited output from the inverter. In this country it is rare that the sun is shining brightly whilst the temperature is low so it is reasonable to design an array with panels rated 5% - 10% percent greater than the inverter. Perhaps as the cost of panels reduces, this ratio could rise to a higher percentage. Just think, if the cost of the panels were low enough, the output from a solar farm could be constant during all daylight hours if the number of panels grossly exceeded the inverter output (That's just a tongue-in-cheek joke, but who knows!)

Dave F

Hi

Excellent post Dave, it's just a shame that it was this very point which was overlooked in the discussion from 12months ago which grahamc2003 recently referenced, both at the time and again a few posts ago ....

Ref post : 23-02-2012, 1:31 PM

zeupater wrote: »

Hi Graham

This is where people become a little confused between Volts & Watts and why most inverters have the ability to connect multiple parallel strings in order to limit the voltage entering on the DC side ....

HTH
Z

... Our particular model of inverter has two MPPTs with two strings per tracker, so four string inputs in total. However, there's absolutely no way that I would consider attaching a considerably larger array than the manufacturer guideline, even if the overall voltage was within specification, this both being due to the effect on the inverter lifespan and the loss of generation potential through 'clipping' in bright conditions.

HTH
Z

Martyn1981

Dave_Fowler wrote: »

However, all panels have a specified maximum open circuit output voltage. Provided that the total open circuit voltage of the panels connected in series in a string does not exceed the input voltage of the inverter then no damage will be done by having any number of panels connected to the input of the inverter as a number of parallel short strings. A 100kW array could be connected as 40 parallel strings of 10 (250W) panels.

Dave F

Cheers Dave, the voice of reason. Hopefully this will be the end of automatically leaping to 'over-voltage' claims every time mention is made of undersizing an inverter - within recommended tolerances of course (as described in the recent SMA and Fronious examples).

Mart.

Edit: Zeup, I didn't overlook your post, in fact I mentioned you today. M.

legoman62

4.7kWh for me.

Another accurate prediction from the weatherman.

Hope he's wrong for the next week....predicted rain, sleet and cold spells:(

Oscargrouch

legoman62 wrote: »

4.7kWh for me.

Another accurate prediction from the weatherman.

Hope he's wrong for the next week....predicted rain, sleet and cold spells:(

Ahh, at last, back to a post of sanity......:D

1.72 UK O's for me..... :doh: C'mon back Shafeeq, all is forgiven....

shafeeq

2.49 kwh today