Sun hours

[DanKegel]

Do any utilities incentivize morning and evening generation yet?

[sensij]

Here are some curves made on recent sunny days. Note the highest curve is the
summation of the (1)W and (3)E facing panels, elevated 58 degrees on 13 June.
The implication is that enough of these will produce excess power at noon without
any help from the (2)S panel. The 17 June curve attempts to level things with an
elevation of 61 degrees; the E-W sum is quite different for a small change. I'm
considering just using E & W facing strings, but some winter curves are still
needed. Bruce Roe

I hope that if you have the terrain / space for it, you aren't overthinking yourself out of simply putting as many panels facing south as possible. 3 east + 1 west might make sense off-grid, or *severely* inverter limited, but otherwise, putting them south will surely generate the most over the course of the year (especially with adjustable tilt).

[bcroe]

MODELING

The array here started as a classic SW desert design. In 3 years of studying
performance, I have a list of things it doesn't deal with very well. Like, CLOUDS,
SHADING, and SNOW. The thought here is to improve these things by changing
the placement of panels. At this stage some model panels could be placed in the
field to compare actual performance to ideas.

3 panels have been set up to model any new design over an operating day, starting
in summer. Maybe they will still be there in winter. In fact about as much energy
can currently be collected on clear day in April, May, or June; this seems to be
because shadowing here increases as the sun moves north. The placement of these 3
test panels will check for any remaining shading. Notice the locating stakes, measured
out within inches. The angle of elevation is adjustable.

Output of a panel recorded over a day will give a curve of power vs time.
Adding together these curves on an hour by hour basis over a clear day, will
show how well total power output is (or is not) maintained at a fairly constant level.
Some fine tuning of the orientation will be involved.

I didn't really want to bother with an MPPT control for each of the 3 test panels,
representing the 3 tilts and locations of a complete array. I decided to just wire
a shunt to each panel (0.1 ohm 10 watt), putting them in the shorted output mode.
This will give a current nearly proportional to the MPPT value. Here is one plugged
into its panel, with meter jacks. Power of a full sized array is just a matter of
estimating the Vmp & Imp, multiplying them together and by the number of panels.

Here are some curves made on recent sunny days. Note the highest curve is the
summation of the (1)W and (3)E facing panels, elevated 58 degrees on 13 June.
The implication is that enough of these will produce excess power at noon without
any help from the (2)S panel. The 17 June curve attempts to level things with an
elevation of 61 degrees; the E-W sum is quite different for a small change. I'm
considering just using E & W facing strings, but some winter curves are still
needed. Bruce Roe

[sensij]

I think you would also need to know how the current would change with voltage under those specific conditions; it isn't constant. Not an easy problem to solve.

[bcroe]

Here is a measurement question. Lets say a string (720 cells total in this case) has the inverter in clipping. The inverter will tell me
the power delivered. I would like to look at the voltage, which will rise with clipping limiting, and estimate how much power is
clipped instead of delivered. I could have a conversion chart of voltage vs % delivered/not delivered. To make this actually work
the panel temperature would be needed to make a temperature correction. Bruce Roe

[J.P.M.]

Then, would you say the reflectance loss will change as panel electrical loading changes? Bruce

No. The reflectance losses for glass used as panel glazing are f(glazing characteristics, incidence angle, nature of the irradiance). See Duffie & Beckman for details. Electrical loading has no influence on those things.

[bcroe]

Respectfully, somewhat, but not entirely. Under MPP conditions, there is still the reflectance loss so
everything not converted to electricity does not get rejected as heat.

Then, would you say the reflectance loss will change as panel electrical loading changes? Bruce

[J.P.M.]

In other words, you agree.

I am doing some model array testing, and rather set up MPPT equipment, think I'll just measure the short circuit
current which is pretty much proportional to Imp. I'll be suffering about a quarter more heating. Bruce

Respectfully, somewhat, but not entirely. Under MPP conditions, there is still the reflectance loss so everything not converted to electricity does not get rejected as heat.

Assuming steady state conditions, whatever energy does not leave as electricity leaves either as reflected irradiance or gets dissipated as heat. A panel that turns 20% of the POA irradiance into electricity and loses, say, 5% off the glass from reflectance will reject (100-20-5) = 75% of the POA irradiance as heat. What goes in must go out for steady state conditions to prevail. If not steady state, the panel temp. will chance in response to the changing conditions until a new steady state is reached.

As Sensij notes under Isc, or Voc all the POA irradiance will be turned into heat or lost via reflectance (assuming steady state conditions).

Panels often perform less well under low irradiance levels so that will change the numbers under low light conditions from high irradiance levels.

Also, high irradiance levels, low wind and high(er) ambient temps. will also change the heat balance because their effect on panel temps., through the negative coefficient of power with respect to panel temps. That will lower electrical output and increase the portion of the POA irradiance that gets rejected to maintain the energy balance. The percentage of POA irradiance converted to electricity will never be the same from min. to min. for a variety of reasons.

FWIW, the amount and type of dirt on a panel will also have a slight effect on the reflectance, probably lowering it slightly, but not enough to make up for the irradiance loss from the dirt. That's beginning to separate fly crap from pepper however, and the stuff of papers for the journals.

[bcroe]

As a 1st approx., all the irradiance hitting a panel that is projected into the plane of the array is either reflected or absorbed with the absorbed portion then either turned into electricity or rejected as excess heat to the environment via conduction through the frame, convective loss to the ambient air or radiation loss to the surroundings. All the energy is accounted for via an energy balance on the cell, panel or array. The amounts/%ages will vary as the panel and environment warrant to maintain an energy balance.

In other words, you agree.

I am doing some model array testing, and rather than set up MPPT equipment, think I'll just measure the short circuit
current which is pretty much proportional to Imp. I'll be suffering about a quarter more heating. Bruce

[sensij]

More specifically, yes, in the Isc or Voc operating condition, zero electrical power is produced, so any light that is not reflected is absorbed and becomes heat. However, neither of those are real world operating states for a functional array. Under normal conditions, for the reasons J.P.M. described, panel temperature is positively correlated to power output.

[J.P.M.]

I'm wondering how much of the energy hitting a panel turns into pure heat for different conditions? For open circuit, short
circuit, and delivering into an MPPT load. Seems to me, for MPPT about 20% of the incident sun leaves, but for the
other conditions, all of the sunlight is converted to heat. Another way to suggest this, is a panel heats the least when
delivering the maximum power? Bruce Roe

As a 1st approx., all the irradiance hitting a panel that is projected into the plane of the array is either reflected or absorbed with the absorbed portion then either turned into electricity or rejected as excess heat to the environment via conduction through the frame, convective loss to the ambient air or radiation loss to the surroundings. All the energy is accounted for via an energy balance on the cell, panel or array. The amounts/%ages will vary as the panel and environment warrant to maintain an energy balance.

[bcroe]

I'm wondering how much of the energy hitting a panel turns into pure heat for different conditions? For open circuit, short
circuit, and delivering into an MPPT load. Seems to me, for MPPT about 20% of the incident sun leaves, but for the
other conditions, all of the sunlight is converted to heat. Another way to suggest this, is a panel heats the least when
delivering the maximum power? Bruce Roe

[DanKegel]

Does sound like shifting some load to electric is in your future. I'm in a similar situation, throwing away power until I get an A/C or something. At least I got a Leaf, that soaks up a bit.

[bcroe]

1 April 2016 anniversary date for my net metering has passed, and after 3 weeks the POCO got
around to mailing me a statement. The 2 questions were:

Can I make it through April without buying any KWH, starting with no reserve, and;

How did the KWH reserve for 2015-2016 play out.

April started cold and cloudy, but then it got warm and sunny. At 100 KWH a day average
generation and the outside temp now 81 deg F, its obvious I'll finish the month in the black.

A year ago, after a pretty cold winter, remaining reserve at anniversary true up was about
700 KWH or 2.5% of generation for the year. Pretty close, but in the black.

This year was warmer, and with the heat pump, more generation went into reserve instead
of being immediately consumed. Generation for very cloudy 2015/16 was less at 26,800 KWH,
down 3.1% from 27,670 KWH in 2014/15. Perhaps some will say that is due mostly to panel
aging, but I don't think so. Any time the panels are unshaded the inverters are clipping as
before; aging could effect cloudy generation, but its reduced to begin with.

Anyway, the graph shows that reduced consumption left a huge anniversary date surplus of
8348 KWH. The POCO gets to keep it for free, but I didn't pay anything extra for it. What
I read, is better energy management is needed. The amount generated and the amount
consumed for heating can't be controlled. But water heating and heat in the shop building
could be shifted over to electric as reserve is available. If some mini splits get set up, the
consumption could be even more efficient. I would need to better keep track of day to day
reserve, waiting a month for a POCO statement won't do.

BUT, its been years since I bought a KWH, and a minimum amount of propane is purchased
annually at the rock bottom July price. Bruce Roe

[bcroe]

The electric bill came Wed. The sun came out seriously for the first time in
Feb and gave me 101 KWH. That kept my 15 KW inverters pegged out for
6.73 sun hours. Today I didn't see the sun for the clouds, but got 65 KWH
anyway.

Cloudy Jan averaged only 38.1 KWH per day or 1200 for the month, but
still a lot better than 1000 a year ago. This might be weather, but reduced
shading probably helps esp when the sun is low. 18 Jan at 80 KWH
showed what is possible when the sun comes out. I think the 2 dozen trees
that disappeared here in 2015 have something to do with getting power out
earlier and later in the day. Half were killed by the severe drought (severe
by IL standards, not CA) a few years back. The others were trying to take
over the south fence line, so now they are gone. There are 7 more trees
back there that will not survive 2016, if I do.

Net metering reserve 1 Feb 16 was 9713 KWH, vs 3200 KWH 1 Feb 2015,
3 times a year ago. Some credit goes to the heat pump in earlier months,
but I think my generation is just up. I made it into positive reserve before
bottoming out a year ago, but this year looks like a big surplus at the true
up this April first.

Jan had some serious cold and I used 127.4 KWH a day, that's 88.7 KWH
out of my summer reserves after 38.1 KWH I generated.

14 Feb it snowed all day; 13 KWH. 15 Feb was solid overcast, with 3" on
the south facing panels it was running at 1/6 capacity. Those near vertical,
snow free east and west facing panels really work. I cleared the snow and
it doubled.

I am not done improving the array. If this surplus is repeatable, I will make
arrangements for some to go to my shop building. Generally I have kept it
around freezing with propane, and blasted it up to 60 F to do car repairs.
Maybe a mini split (capable of operating down to zero F) can take over
maintaining a minimum temp. Bruce Roe