Expert opinions wanted

System

That is great! I have a new installation of 20 Evergreen 190 panels (with SB4000 inverter) in Massachusetts and on a sunny day (like yesterday, ambient temperature 90 degrees), I can get only 7 KWH. I noticed that the peak watt at around 1 pm (EST) was only 1,155 Watts. I was expecting around 3,400 Watts. Is anything wrong with the panels or the system? I would appreciate your expert insights.
Nihar Mohanty

BB.

Re: Suprass 1 MWh yesterday

So you have two strings of 10 panels each?

If so--it sounds like one of the strings is not generating. Either partial shade, open electrical connection, or a problem with a panel.

What is the voltage and current readings from the inverter (DC input)?

Did you ever get more power--or is this it.

According to this link, you should be getting around 2,400 watts around 12 or 1 pm (full sun, default installation for 3.8 kW worth of solar panels...

You can either try an diagnose it yourself--or just call the installer.

-Bill

mike95490

Re: Suprass 1 MWh yesterday

nmohanty wrote: »

That is great! I have a new installation of 20 Evergreen 190 panels (with SB4000 inverter) in Massachusetts and on a sunny day (like yesterday, ambient temperature 90 degrees), I can get only 7 KWH. I noticed that the peak watt at around 1 pm (EST) was only 1,155 Watts. I was expecting around 3,400 Watts. Is anything wrong with the panels or the system? I would appreciate your expert insights.
Nihar Mohanty

Sounds a little "light" on output. First questions I'd ask:
1) ventilation. Is there good airflow under the panels to let them stay as cool as possible ?
2) how are they oriented - South roof, North roof ??
3) angle of install , flat on roof, or on a bracket to get optimum angle ?
4) any shade on them ? Even a little bit on a corner will drastically reduce power for the whole array. (I can see a drop as the shadows from the nearby power lines (wires, not the pole) crawl across the array)

My array on a west facing roof, peaks about 2 - 3PM pacific, when the sun is perpendicular to them. In the winter, when the sun is further south, and I don't get perpendicular, my output is about 50% of summer rate.
22ea, 208W panels Los angeles.
January 2008 11685w daily, 2188w peak
June 2007 27805w daily, 3558w peak I have about 5 inches under my panels, and usually a good breeze to help keep them cool. Today, they have warmed up so much, there is about a 20V loss of voltage, from when they were cold this morning

I've got several sites linked to displays on my page at http://www.mike-burgess.org/PVinfo_2.html where you can compare what different systems produce daily.
In Mass, look at the Riverland Farms site
http://www.riverlandfarm.com/solar (give them a call and tell them to kick their server!)

so, heat & angle are critical for the PV's to work right

BB.

Re: Suprass 1 MWh yesterday

I should add--although I find the linked site from my post above, I just compared with my own unit today (3.5kW of panels, 85F, 1-2pm)--I would guess that your 2,400 Watt reading (based on the predicted output) could easily be 20-25% higher or near 2,900-3,100 Watts expected for a clear/warm day (the higher numbers based on my system vs predicted output--mine is over 2.5 years old now--and the panels are pretty dusty, but still generating nicely).

-Bill

System2

Re: Suprass 1 MWh yesterday

Thanks for the comments. Yes, I have two strings of 10 panels each. The panels are at 21 degrees tilt (not the optimal 42) and 16 degrees SE (The panels are pointed to magnetic south). At the time of my observation, there was no shading on the panels (between 12 and 1 pm). There is about 5 inches of spacing underneath the panels so there is probably some ventilation. The voltage was approx. 250V and I did not check on the current display (I will tomorrow) but estimated 1,150W/250V= 4.6 amps? Today was the second day so I do not have much information.

BB.

Re: Suprass 1 MWh yesterday

250 volts DC for panel voltage sounds OK...

You should be getting around 7 amps DC per string--or 14.x with two strings (you probably want to see at least 12-13 amps in full sun).

With the flatter install--you should even get a hundred or so more Watts this time of year than I predicted before.

Peak wattage still will be around 3,160--probably on cooler/windy days (this is after solar panel deratings and inverter losses of 5% or so). On hot days, anywhere above ~2,600 Watts is good.

Remember that current is pretty much proportional to sun energy... Solar panel output voltage will drop on hot days (my guess, for your system on a 95F no-wind day--panel voltage around 219 volts)--leading to less power generated (Power=Voltage*Current -- so lower voltage gives lower power).

Hmmm... 250 VDC actually sounds a little high for a hot day--could be an issue with the inverter too (if panels, wiring, current, voltage all look good to the inverter). Don't know for sure--just comparing with my 3kW installation.

Need to call your installer on Monday.

-Bill

crewzer

Re: Suprass 1 MWh yesterday

The SB4000's (240 VAC model) peak power tracking voltage range is 250 - 480 V DC. If the SB has raised the array's operational V (adjusted for high ambient temp and mounting style) from ~220 Vmp to ~250 V, then the array's output current would be dramatically reduced. This in turn would reduce output power from the inverter.

What's the indicated ~mid-day total array current when the inverter indicates 250 V and ~1,200 W -- ~5 A or so?

The SB sizing tool suggests to me that 10 Evergreen 190's per series-string will operate sub-optimally in a hot ambient environment and when mounted close to the roof surface.

HTH,
Jim / crewzer

Solar Guppy

Re: Suprass 1 MWh yesterday

Having evergreens, I know for sure your problem is you have to few panels in series on your strings, I have 15 in series with My Xantrex GT5.0, similar input requirements as the Sunnyboy.

The Evergreens are 18V nomimial panels, the vmp is 26.7V ( at 25C ) adjusted for your roof mounted ambient tempature I'd guess the true vmp will be ~21V or for a string of 10, 210V and your sunnyboy needs a minimum of 250V BIG PROBLEM!!!!

Also 22deg for Mass is WAY TO LOW ... you will only get decent harvest near June. Your Latitute is near 43 degree's and in the winter, your panels will be off almost 45 degrees as the Sun is at 65 degrees in december. Further, you may have serious snow/ice damming problem in the winter, 22Deg is unusual for the area and snow will not shed well on a shallow roof, do a search and you will find panels get damaged from snow-ice build up from under the panels expanding.

At a minimum, you need to add two panels per string ... I'd be very suspect of your installers as what you have told us point so them lacking very basic PV knowledge

newenergy

Re: Suprass 1 MWh yesterday

Solar Guppy wrote: »

At a minimum, you need to add two panels per string ... I'd be very suspect of your installers as what you have told us point so them lacking very basic PV knowledge

Looks like 2 more per string would be ideal. He starts to hit inefficiency at low temperatures with more than 12 per string since it's cold in MA.

1 more per string is pretty good most of the time. The average high temp in Boston is only 73-82 throughout the summer.

It must be a real pain to string modules in an area that gets very hot in the summer and very cold in the winter.

Solar Guppy

Re: Suprass 1 MWh yesterday

One panel isn't enough. Roof mount regardless of where it is added close to 45C in panel tempatures, all that asphalt just radiates back at the panels.

My Panels with ambients in the high 80's low 90's are getting over 65C based on the measured voltage and IR measurements

For the Evergreens, the delta is -0.120V/ degree C over 25C ...

So panels at 65C would have a drop of 45 * 0.120 = 5.4V per panel or 54 volts in a ten panel string. vmp of 26.7 - 5.4 = 210.3V

Add one panel and you only get ~231V , still a full 20V below the minimum operational voltage of the Sunnyboy

I have a hunch the installer is clueless about this stuff and just used the published vmp @25C values.

BB.

Re: Suprass 1 MWh yesterday

There is another slight possibility--220 VDC is MPPT minimum for a 208 VAC service... Typically, 120/240 VAC is standard for home service and 208 VAC (in California) for commercial.

Besides adding panels, changing to another inverter (Xantrex GT 3.3)--Or can an autotransformer be used to step down the voltage if more panels/different inverter are not an option? (bit of a hack though--I think the Sunnyboy is 240/208 auto detect and Vmp goes to 220 VDC--still pretty close with only 10 panels)...

-Bill

newenergy

Re: Suprass 1 MWh yesterday

Installer needs to sign up here I guess.

Xantrex GT 3.8 has V mppt min at only 195 Volts. It seems to string well with two strings of 10 ES 190s. Right?

Depending on roof space he might not be able to add modules, but could possibly switch inverters.

Also, since this is a hijack of the thread I'll add "Contgrats autoxsteve!"

edit: woah, I guess it's not a hijack anymore.

niel

Re: Expert opinions wanted

newenergy,
i agree it was derailed from the original thread, but not with malice so i split up the thread and you'll have to give steve the congrats on the original thread.
i don't know what i did wrong as all posts are not addressing the new title i had given the thread except for my reply, but i'm not going to modify every individual post to read correctly. if crewzer or windsun knows what i did wrong or fixed it please pm me on it.

Solar Guppy

Re: Expert opinions wanted

Looks like there is no more GT3.8, only GT4.0 & GT5.0 , there range is 235-550Vdc on the GT's Mppt, a bit better than the SunnyBoy units thought. Interesting, I beleive thats how SMA gets to claim best in class efficiency, reduce the operational range :roll:



Xantrex 235-550Vdc
http://www.xantrex.com/web/id/1416/DocServe.aspx

SMA 250-480Vdc
http://download.sma.de/smaprosa/dateien/4776/SB3000US-4000US-101907.pdf

For the OP, use this to check configurations:
http://www.xantrex.com/support/gtsizing/index.asp?lang=eng#calculator

newenergy

Re: Expert opinions wanted

Yeah, I don't see the 3.8 for sale, but the GT 3.3 is available and it's range is 195-550V. PTC for the 20 ES 190s is 3344W. You are hitting the upper power range for the 3.3, and should lose a little efficiency doing that. Maybe you won't even be that close to the limit in MA.

mike95490

Re: Expert opinions wanted

newenergy wrote: »

Yeah, I don't see the 3.8 for sale, but the GT 3.3 is available and it's range is 195-550V.

How would a 200V array come on line to feed a 220V circuit (or a 240V north america ckt)?

Solar Guppy

Re: Expert opinions wanted

Its the ratio of the transformer between the DC and AC sections internal to the unit.

BTW, it takes 400Vdc to make the 240Vac as 240 is RMS not peak, the 190 or 235 or 250vdc is just a design choice to limit the current, which is the root cause of all loses in an inverter design.

newenergy

Re: Expert opinions wanted

Solar Guppy wrote: »

Its the ratio of the transformer between the DC and AC sections internal to the unit.

BTW, it takes 400Vdc to make the 240Vac as 240 is RMS not peak, the 190 or 235 or 250vdc is just a design choice to limit the current, which is the root cause of all loses in an inverter design.

So basically, when you can you should design towards the upper end of the voltage limits for a given power output?

Solar Guppy

Re: Expert opinions wanted

Well, there are other losses, the higher the voltage, the more switching losses so its a balance. My other post was more to the point of why as the wattage goes up, the operational range shrinks.

From a design for installation view point, one should stay at the lower end of the operational ranges as that is the sweet spot for most inverters, but were talking 1-2% improvement over running at the highend of the operational range.

System2

Re: Expert opinions wanted

First,
I am impressed with your advice and insight. I am new to this so do not fully understand all the nuances. All comments are not at the same place so hard for me to quote you guys. I checked the voltages again, it is not 250 but around 247V around noon time with the peak wattage of 1150 as I mentioned earlier. I called the installer (Borrego Solar) and they came in to take a look. They will be back on Friday to pull a few panels and check them again. Before they suggest ideas to me let me run a few ideas past you:

(1) I can add two more panels and it would increase the voltage somewhat and increase the efficiency. I am thinking it is not my fault with the design so should I be paying for this?

(2) to increase the voltage, is it possible to replace a few Evergreens with other makes that generate higher voltages? Say, replace two panels of each string with these higher voltage panels. Based on what I have read, it may not be a good idea...

(3) Replace the inverter with one as some of you said, with another Xantrex model 3.8 that has a lower efficiency...
(4) I can not increase the shallow angle of the panels so that idea can't be implemented
Any other ideas to resolve the problem?
Thanks in advance,
Nihar

mike95490

Re: Expert opinions wanted

nmohanty wrote: »

First,
I am impressed with your advice and insight. I am new to this so do not fully understand all the nuances. All comments are not at the same place so hard for me to quote you guys. I checked the voltages again, it is not 250 but around 247V around noon time with the peak wattage of 1150 as I mentioned earlier. I called the installer (Borrego Solar) and they came in to take a look. They will be back on Friday to pull a few panels and check them again. Before they suggest ideas to me let me run a few ideas past you:

(1) I can add two more panels and it would increase the voltage somewhat and increase the efficiency. I am thinking it is not my fault with the design so should I be paying for this?

(2) to increase the voltage, is it possible to replace a few Evergreens with other makes that generate higher voltages? Say, replace two panels of each string with these higher voltage panels. Based on what I have read, it may not be a good idea...

(3) Replace the inverter with one as some of you said, with another Xantrex model 3.8 that has a lower efficiency...
(4) I can not increase the shallow angle of the panels so that idea can't be implemented
Any other ideas to resolve the problem?
Thanks in advance,
Nihar

If there is space on the roof, they should install the 2 extra panels, free labor, you buy the parts @ their cost. They made a mistake, and need to step in to correct it, but you can't expect them to eat the price for panels. They should think of the good reports you will have for them when it's working right.

OR they can swap out the inverter for one that matches the panels better.

BUT the ROOF will be the killer. When you get an ICE dam, and it floods your house, who ya gonna call ?? The mounting needs to be fixed, so you don't get an ice dam.

Solar Guppy

Re: Expert opinions wanted

There is nothing to "pull" or "look at"

The installer simply didn't do proper string sizing ... have your install look at the Xantrex string sizing tool,

http://www.xantrex.com/support/gtsizing/index.asp?lang=eng

its black and white, you setup is short TWO PANELS PER STRING, for a total need of 4 panels need to be added. There is nothing wrong with the panels, its a configuration mismatch, the panels, ten in series is TO LOW for the Sunnyboy inverter. If your installer doesn't get that, pull up the sizing tool on your computer and sit him down and show him what he did wrong ( select a GT5.0 for the inverter type as its the closest to the Sunnyboys requirements )

The most straight forward change would to be to change out the Sunnyboy with a Xantrex GT3.8, as far as efficiency, they are the same, its a spec game the Sunnyboy is using to "claim" they are better. The GT, for its wide voltage range is number one, no if and or butts about this.

Since the GT3.8 isn't manufactured anymore ( it is not UL1741-2005 ), the next step would be to add ONE PANEL PER STRING and change the Sunnyboy to a Xantrex GT4.0, that would work since the GT4.0 is good down to 235V where the Sunnyboy needs 250V minimum

As for having the installer pay for extra watts, thats not his responsibility, it would be fair as Mike suggests to have you pay actual cost for the 2 or 4 panels and they install for free.

If they do an inverter swap, they should pay YOU, the Xantrex is about 500 dollars less street price, has built in disconnects and you get free communications to boot

Its really sad to see an installer have done this, but I'm not surprised. Re-read this post until you understand what it means .. then sit your installer down and tell him what your expectations are

SG

System2

Re: Expert opinions wanted

Thanks SG,
I played with the string sizing of Xantrex and I understand what you are saying. Although I have not heard much about Fronius inverter's reliability or efficiency, what do you think about the model Fronius 4000 that has an input DC voltage range from 150V to 450V? Then I may not have to add any more panels?

newenergy

Re: Expert opinions wanted

SG,

I'm not doubting anything you're saying, but if you use the SMA string sizer it suggests 2 strings of 10 ES 190s as an acceptable arrangement for the SB4000US.

If you look at the predicted outputs you will see a bunch of yellow numbers, but it is offered as a viable configuration. I suppose since the highest average high temperature for any month is 82 deg, they are willing to perform poorly on 90 degree days.

It seems installers need to be careful using the SMA sizer.

Solar Guppy

Re: Expert opinions wanted

nmohanty wrote: »

Thanks SG,
I played with the string sizing of Xantrex and I understand what you are saying. Although I have not heard much about Fronius inverter's reliability or efficiency, what do you think about the model Fronius 4000 that has an input DC voltage range from 150V to 450V? Then I may not have to add any more panels?

I personally don't like Fronius, They are exposed ( internal electronics ) fan cooled. Their efficiency was one of the lowest when I tested them 3 years ago.

You could also use a pair of Xantrex GT2.5's, they are low cost enough to be similar to a SB4000 if you don't like the other options.

I'm thinking add two panels and goto a GT4.0 would be your best option, followed by a adding just a GT3.8 followed by adding a pair of GT2.5's

Solar Guppy

Re: Expert opinions wanted

newenergy wrote: »

SG,

I'm not doubting anything you're saying, but if you use the SMA string sizer it suggests 2 strings of 10 ES 190s as an acceptable arrangement for the SB4000US.

If you look at the predicted outputs you will see a bunch of yellow numbers, but it is offered as a viable configuration. I suppose since the highest average high temperature for any month is 82 deg, they are willing to perform poorly on 90 degree days.

It seems installers need to be careful using the SMA sizer.

Looking at all the charts, the one that clearly is in error is the "Estimated PV Array Maximum Output Power vs High Temp" ... The error it is just applying the Vmp drop ( ~0.12V/C per ES-190 ) and not taking into concideration the Mppt minimumum of the inverter.

Any value below 250V and your losing potentional power with the SB4000 and 10 ES-190's, in this case big time. I'm also not sure what they are using for the roof mounted array delta, seems low to me. Something as simple as a dark shingle vs. light shingle can make a 20% difference in tempatures.

Even using the SMA sizer, it only showes ambient tempatures below 41F as working, so its pretty clear its an issue.

I can see how an installer would think its not a big deal based on the power chart ( which is wrong ). I've sent and email to SMA-America alerting them to this concern

System2

Re: Expert opinions wanted

Thanks SG and Newenergy for your comments. I sopke to the installer and they suggested as you said... to put in two extra panels and I need to pay them at their cost price. My shingles are "weather wood" color so between light and dark I guess. I also got a call from a public TV in Boston for discussing the system and I am going to forward the installer's name to the interviewer. I will report when I see more progress.

Solar Guppy

Re: Expert opinions wanted

2 extra panels aren't enough ... you need 4 panels to have it perform above the SB4000 inverter's 250Vdc minimum requirement in all weather/seasons

System2

Re: Expert opinions wanted

SG,
They found out that they had a loose connection in one string to the system and fixing that that increased the peak wattage to about 2300W. They will change the inverter to suit the voltage. There is a new inverter model by Solectria (PVI 4000) that can accept voltages in the range of 200-550V and have 96% efficiency so may work out but does anyone know about their reliability?