backchannel: solar harvest, battery charging

from this post:

Although solar is sold to the public as watt/hr rated panels, to achieve their sales rating the panels must be oriented perfectly facing the sun, on a cloudless day, with no obscurations or diffusing humidity to achieve their rated power.

I’d say it’s an apples-apples objective rating which salescritters might [mis]use to make a sale. For onlookers, lab ratings are done at 1000W of light per square meter with cell temps of 25C (ambient approx 0C.)

Cell temp is specified because panel maximum output decreases as cell temperature increases. Luckily it can be calculated easily. Note to new folks: losses due to heat are the reason we leave airspace under panels. That way air can get under the panels and wick away heat from the bottom of the cells, which is typically 50-60F higher than ambient.

Example: today’s 100F temp in El Paso means my panels will lose 21.24% of otherwise-available power to temperature derating alone.

Explanation:  Pmax (max power output) is depressed at higher temperatures because Vmp (voltage at which the panel makes max power) goes down.  And power (watts) comes is voltage x current.

Imagine a panel that for reasons of clarity has a Vmp of 20 and Imp (current at max power, very stable across the power curve) of 5A under lab conditions/temps.  20v x 5A = 100w.

Now let’s subject that panel to a 100F day.  Vmp drops to 18v but Imp stays stable as it is prone to do.  18v x 5A = 90w.  <– temperature derating  of 10%

Interesting ramification:  PWM controllers don’t run the panels at Vmp, they run the panels at Vbatt (battery voltage).  Since this is generally much lower that Vmp they are not affected by temp derating until ambient gets very high.  This isn’t because of any kind of superiority of PWM;  the controller left that excess voltage on the table and hasn’t been making power from it anyhow.  So it’s not that PWM wins, it’s that PWM lost earlier in the game and so is not additionally affected now.

In some edge cases (12v bank + 12v panels + high ambient temps) PWM really can make more power than MPPT.  This is possible because MPPT typically has a ~5% loss due to the DC-DC magic it uses to trade excess voltage for more current.  You’d see this a bit below 100F ambient for poly panels and a bit above 100F for mono, owing to the difference in their original Vmp.

This overtaking by PWM can be addressed by running those 12v panels in series, or using a higher voltage panel to start with.

We have two 290 watt Solar City panels mounted flat on top of the RV. They do not produce 580 watts of power. In fact I’ve never seen them produce more than 510 watts of power.

I see similar results here. I have 570W on the campervan and the highest I’ve seen is in the low 500s. 200-300w is more common with my setup. It can be a bit misleading though on overpaneled systems. They can get through Bulk so early in the day that they are already on the downslope of Absorption by solar noon when max power might be generated. The high output observation I mentioned above was due to being in a forest; there was minimal direct sun until it was almost solar noon.  Full speed ahead!  If you have a breaker or switch between your panels and controller you can simulate this setting by keeping the panels offline until solar noon.  Then flip the switch and see what she can really do!

Warning:  one might see power (and controller-killing voltage) greater than the panel’s rating in very cold temps + high altitudes.  The Vmp rises in the cold and insolation increases with less atmosphere between you and the sun.

If you have 1,200 watts of solar panels, and you are operating them in a setting similar to ours (here in the SE United States with diffused sunlight, high humidity and high temperatures in the summer) you should expect 33% of your rated power on average.

This was very nicely qualified.  The rule of thumb does work quite well for the area and time he describes. With a couple calculations we can get predict average daily solar harvest from the panels

Insolation averages are published by location and often by season/month. They are given for flat-mounted panels, and factor in latitude, average atmospheric conditions, etc.

Example: In September, Miami [the stand-in I chose for the SE he mentions] averages 5.47 hours of Full Sun Equivalent (FSE) a day.   “Full sun” in this context means the 1000w/m2 lab rating spec. Average high in Sept is 89F, or 31.67C.  Under those conditions my 570w system would make, on average:

570W x 5.47h FSE = 3117.9 3117.9 - (3117.9 x 18.168% derating = 566.46) 2551.44W every solar day (2.551kW/day) assuming the system was fully loaded. Note how close that is to the 1/3rd rule of thumb this time of year (2551/12hrs = 212.58W/hr) at that location.

On the other end of the scale, Anchorage averages 1.98hrs of FSE daily in September but will normally have less temp derating (only ~9%) for about 1.02kW/day. Our 1/3rd rule of thumb breaks down here and becomes something like 1/6.7 (1020 / 12 = 85w).

The reason I bring this up is it’s pretty easy to derive actual daily power generation output for any area/season using insolation averages and temperature derating. We don’t have to make local rules of thumb.  The rules would be more helpful for folks who stay in one area moreso than for nomads.

The sweet spot: In the morning before your solar panels actually start producing any significant amount of power (6am – 8am), start your quiet generator and let it put two hours of energy into your batteries, or 1,620 x 2 = 3,240 watts towards the 11,178 depletion. This brings the number down to (11,178 – 3,240 =) 7,938 watts remaining. Then start your engine driven alternator for an hour (8am – 9am) and it will bring things down much faster by inputting 3,510 watts into the recharge leaving (7938 – 3510 =) 4428 watts remaining. At 9am let the Solar system take over and during the course of the day it will erode away another 3,600 watts of the remaining 4,428. At lunch time again start your genset to use your microwave or air conditioner, and as long as you are not trying to operate multiple air conditioners your Inverter/Charger should still sense a bulk charge need and dump in 100% or 1,620 watts of power for that lunch hour. Between the genset lunch hour (1620 watts) and the 12 hours of solar charging (3600 watts) the overall remaining 4,428 watts needed to return the battery bank to 100% SOC should be achieved before sunset. If you do not want to start your engine and use the alternator, you can achieve the same by operating the genset again for two hours during dinner, where it will again produce almost as much recharging power during those two hours of operation as the engine driven alternator does in 1 hour.The sweet spot...

Whew!  A simpler way to describe the process would be “add generator/alternator charging for lead chemistries any time the system is in Bulk stage”.  Once current acceptance starts dropping off in Absorption solar can take over the long duration, lower current, higher voltage charging Pb requires.  I agree it is optimal to save fuel and  wear & tear on the genny/alt once we are clear of Bulk.

I also agree that LFP would greatly simplify the dance that Pb demands.   Charge when you want, for however long you want, through whatever charger you want.

Hopefully everyone realizes that a 50 amp shoreline connection negates the concern all together because a 50 amp shoreline connection has a 12,000 watt capacity and could recharge your entire bank back at whatever rate your Inverter/Charger is capable of - typically 2-3 hours if your charger is mean enough.

Deeply cycled Pb batteries require 2-3 hours of Absorption stage charging alone.  Plus however long Bulk takes.  2-3hrs to get a full charge could be possible with lightly-cycled batteries.  Lithium can charge much faster naturally but for longevity reasons the BMS typically caps charging at 1C, sometimes 0.5C.  For a ~800Ah bank specified in the thread it  would take up to 8hrs to charge a depleted Li bank even if it were being charged by a nuclear powered aircraft carrier.

Obviously tripling the size of your solar array from 600 watts to 1,800 watts means you can recharge the eight 207 Ah AGM batteries in 1 Day. However, to achieve this feat, one must NOT use any power from the batteries or from the charging solar array during the daylight hours or it will lengthen the charging time necessary to achieve 100% SOC. You’ll need to operate your generator or engine driven alternator occasionally to achieve this.

Well, one must not use any power that would reduce the amount of charging current the batteries want.  As current acceptance drops off during Absorption one can use excess power to run loads with zero impact on Pb charging. And of course during Float you have access to almost the full power of the system, as only a tiny bit of current is required to hold Vfloat.