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

[Kirk Herander, VSE]

Can't you use an inverter with two separate MPPT circuits - ie. Aurora?

-----Original Message-----
From: re-wrenches-bounces at lists.re-wrenches.org
[mailto:re-wrenches-bounces at lists.re-wrenches.org] On Behalf Of Peter
Parrish
Sent: Friday, December 26, 2008 2:30 PM
To: 'RE-wrenches'
Subject: Re: [RE-wrenches] Inverter with two strings of different
orientation

I was afraid that the ultimate resolution would be another software cost
item. I'll look into PVSYST, and consider purchasing it if it is a
professionally-developed and maintained product, with pre-and post-sales
technical support, a policy of regular upgrades, runs on PCs and MACs, and
has a model library containing current commercial PV modules. Am I asking
for too much?

As for Bill's comments, I have a problem with his reasoning. I agree that
Voc is predominantly a function of temperature. Secondarily, temperature is
a function of the normalized irradiance (my term, which is the irradiance
times the cosine of the angle of incidence), ambient temperature and wind
speed (the latter two being constant between sub-arrays). However, Voc is
not the important factor; Vmp, or something close to Vmp, is. Vmp is a
function of irradiance (otherwise MPPT algorithms wouldn't work).

Here is the problem as I see it: Vmp for two otherwise identical sub-arrays
with significantly different tilts/azimuths will have different I-V
characteristics and different MPPT points -- so long as they are not
connected in parallel. In specific, they will have different Vmp and
different Imp. 

It is actually worse than that: looking at the two I-V characteristics,
there are no points in common between the two sub-arrays (with the exception
of very close to Voc, which is not a practical operating point, so it can be
ignored). There must be some sort of dynamic interaction between the two
arrays to establish a common voltage. This has nothing to do with inverters,
because we could examine this problem by loading up the strings with
resistors.

When two strings are wired in parallel, the voltage across both strings must
be equal, by definition. From a practical point of view, when wired in
parallel, no operating point can possibly maximize the power from each
string, there has to be some reduction in output power.

If PVSYST can produce a correct I-V characteristic of two strings in
parallel with different tilts/azimuths, and if commercial inverters have the
capability to deliver a MPPT capability under this condition, great,
otherwise I remain skeptical.

Anyone have a citation for the articles purporting to treat this subject? 

- Peter

Peter T. Parrish, 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