[RE-wrenches] Inverter with two strings of different orientation

[boB]

> Date: Mon, 29 Dec 2008 11:32:42 -0700
> From: "Starlight Solar, Yuma, AZ" <larry at starlightsolar.com>
>
>
> I have pondered this problem often. Here's my thoughts: If you could
> disconnect the two orientations and monitor them simultaneously, in
> the morning the easterly array would have a higher Vmp while the
> westerly array would necessarily be lower. This relationship would
> change throughout the day with an equilibrium being when the sun is
> equidistant between the two arrays. (I'm just stating the obvious here)
>
>   
It may be that in the morning, the west pointing array is at a lower 
temperature than the east pointing array
and therefore the Vmp is higher for that cooler array.
If the Vmp doesn't change ("much") with irradiance, then this shouldn't 
make a difference **IF** the array
that is not getting full sun is getting an *EVEN* irradiance across the 
whole array. i.e. No PARTIAL shading.

If you darken just one module of an array, or even part of one module 
THEN the Vmp will go down by approximately
one panel because that panel is now out of the string, basically, and 
being bypassed by its bypass diode.

If you look at the curve in one of those Sandia papers, you will see 
that Vmp changes much more with irradiance when
the irradiance is nearing darkness. At that point, it is bordering on 
shade I would say, and then that module is not really working
as part of the string anymore. That's where "much" now means something 
and is getting into that lower portion of the Vmp-irradiance
graph.


> My question is will the lower Vmp array pull down the voltage of the
> higher? 
An array can pull down the voltage of another parallel array only if its 
Voc is lower than the Voc of the
other parallel array. Usually it's the inverter or controller that's 
doing the pulling down of voltage to get
some power out of the array. Remember, the array is a string of diodes 
that happen to make voltage
as well as being able to be forward biased by a voltage impressed across 
them.

>  I notice this effect with dissimilar solar panels and MPPT
> controllers that many DIY RV customers have.  It was stated that you
> may have a 2-5% loss. I can't imagine that the losses would be that
> low averaged for the day.
>
>   
Production will be lost compared to the production if all modules are 
un-shaded and receiving the
most possible amount of sun.

> If the strings were split equally accross the two orientations, wont
> the Vmp track throughout the day and give the most production
> possible?
>   
If the MPPT algorithm any is good, it should track this solar jungle as 
best as can be done by finding
the one "sweet spot". Sometimes there are 2 very close sweet spots in a 
case like this, power wise.
That's when it gets harder.

Personally, I think wind is easier to track because of all this..

boB


Of course the inverter threshold voltage will not be met for as many 
hours daily but how much will this arrangement loose?
> Kindest Regards,
>
> Larry Crutcher
> Starlight Solar
> 11279 S. Glenwood Ave #11114
> Yuma, AZ 85367
> (928) 941-1660
>
> larry at starlightsolar.com
> www.starlightsolar.com
>
> Retail Store: 2998 Shari, Yuma, AZ
>
> Renewable Energy Products, Service and Installation
>
>
>
> On Dec 24, 2008, at 12:41 PM, Peter Parrish wrote:
>
>   
>> I don?t think this topic has been dealt with in this forum.
>> Ocassionally we want to use a single inverter with two or more
>> strings, each string being loaded by a single sub-array. A sub-array
>> being defined as having a single uniform tilt angle and azimuth. All
>> modules are identical. And each string/sub-array contains the same
>> number of modules.
>>
>> The question is: ?How different can the orientation be for the
>> different sub-arrays?
>>
>> Let me pose a concrete example: Xantrex GT5.0, two strings of BP
>> Solar SX3190s. One string has an orientation of 15 deg tilt and 188
>> degrees azimuth; the other has 23 degrees tilt and 139 degrees
>> azimuth. For reasons that are difficult to explain, none of the
>> orientation angles can be changed, at least for the purposes of this
>> discussion. To visualize what I am talking about: the tilts are
>> fairly close (8 degrees) and the azimuths differ by about 45 degrees.
>>
>> How do I calculate the combined performance? How much power does one
>> give up compared to using separate inverters for each sub-array/
>> string? Are there rules of thumb as to what works and doesn?t? Do
>> conventional inverters have algorithms that can navigate the non-
>> ideal I-V characteristic under load? I know there was a tech
>> bulletin from Fronius a while ago: I read it but didn?t come away
>> with any definitive answer. I also had first hand knowledge with
>> MPPT algorithm problems with the Fronius inverters during that time
>> period, which caused me to discount the tech bulletin.
>>
>> I am quoting a system as we speak, so my inquiry is not hypothetical
>> in nature. Any help would be appreciated.
>>
>> - Peter
>> Peter T. Parrish, Ph.D., President
>> California Solar Engineering, Inc.
>> 820 Cynthia Ave., Los Angeles, CA 90065
>> Ph 323-258-8883, Mobile 323-839-6108, Fax 323-258-8885
>> CA Lic. 854779, NABCEP Cert. 031806-26
>> peter.parrish at calsolareng.com
>>
>> _________
>>