If you are using this website:

http://www.solarelectricityhandbook.com/solar-irradiance.html

Note that it uses the angle measurement from vertical. So 50 degrees from vertical is equivalent to 40 degrees from horizontal (90-50=40 degrees).

If you are at 35 degrees latitude, that is pretty close to the typical latitude=best overall yearly harvest (i.e., 90-35l at=45 degrees from vertical). I would not expect a lot of change in harvest with a 5 degree "change" (see below for PV Watts test).

If you use something like PV Watts (and most other programs) seem to pretty much use tilt from horizontal measurements.

PV Watts does not give a suggested best summer (typically +15 degrees flatter), yearly best harvest (typically = Latitude), or winter (typically +15 more tilt).

Using PV Watts for Slow Low AZ, playing with the numbers for a fixed south facing array):
https://pvwatts.nrel.gov/pvwatts.php

20 degrees from horizontal (35-15=20): 6.26 hours of sun yearly average (daily average over one year)
35 degrees from horizontal (35 = Latitude): 6.39 hours of sun yearly average
45 degrees from horizontal (35+10 degrees): 6.28 hours of sun yearly average
50 degrees from horizontal (35+15=50 summer): 6.17 hours of sun yearly average

Of course, the above does not include the best/worst of summer/winter month harvest--Just total yearly average (may be best for GT solar systems; but for an off grid system, you may want more summer or winter harvest based on A/C and Refrigeration/irrigation/etc. loads/needs).

So, from a "rule of thumb" design, the +/- and zero 15 to latitude seems to be pretty consistent with PV Watts.

Of course, PV Watts is based on actual data (include weather, dust storms and panel temperature--As I recall)... So if you have, for example, afternoon thunderstorms during the wet season, changing tilt and/or direct facing (say more South East) may give you better harvest during stormy weather.

One of our retired members here lived in Florida, and found facing more south east gave him better overall harvest (both because of weather and because panels were colder in the morning vs afternoons). However, the utility folks "pushing" GT solar wanted afternoon harvests (when folks turned on AC systems in the offices and after they headed home)--And they were surprised to see the variance in harvest results.

-Bill

This a "simple" solar calculator--It gives you several options/configurations:

Note this calculator uses the "angle from vertical". Pretty much everything else uses the "angle from horizontal"--Which makes things easier for "setting angle to lattitude" (52 degrees from Vert = ( 90-52= ) 38 degrees from horizontal.

http://www.solarelectricityhandbook.com/solar-irradiance.html

Pueblo
Average Solar Insolation figures

Measured in kWh/m2/day onto a solar panel set at a 52° angle from vertical:
(For best year-round performance)

Jan	Feb	Mar	Apr	May	Jun
4.49	4.93	5.61	5.74	5.88	6.10
Jul	Aug	Sep	Oct	Nov	Dec
5.79	5.62	5.95	5.72	4.68	4.30

You can play with the numbers and see how things are affected. It is "good enough" for our planning (within ~10% of "real" numbers, IMHO). And trying for more accuracy is not really helpful--Weather, clouds, humidity, temperature, dust on panels, etc. all affect numbers by 10% or more over days/weeks/months/years).

A more sophisticated program that makes things a bit more complicated:

https://pvwatts.nrel.gov/

38 degrees from horizontal, facing south, 52% overall efficiency (for off grid AC power systems).

RESULTS

System output may range from 1,027 to 1,090 kWh per year near this location.
Click HERE for more information.

Month	Solar Radiation ( kWh / m2 / day )	AC Energy ( kWh )	Value ( $ )
January	5.03	79	13
February	5.65	79	13
March	6.63	101	16
April	6.68	98	16
May	6.60	96	16
June	6.87	93	15
July	6.63	92	15
August	6.76	94	15
September	6.88	95	15
October	6.08	90	15
November	5.29	78	13
December	4.92	78	13
Annual	6.17	1,073	$ 175

User Comments

* Caution: The PVWatts® energy estimate is based on an hourly performance simulation using a typical-year weather file that represents a multi-year historical period for Pueblo, CO for a Fixed (open rack) photovoltaic system. The kWh range is based on analysis of a nearby data site described here.

The estimate for the value of this energy is the product of the AC energy and the average retail electricity rate. This value is useful for basic comparisons but does not account for financial considerations in a cash flow-based analysis. All of these results are based on assumptions described in Help that may not accurately represent technical or economic characteristics of the project you are modeling.

Location and Station Identification
Requested Location	Pueblo colorado
Weather Data Source	Lat, Lon: 38.25, -104.62 1.2 mi
Latitude	38.25° N
Longitude	104.62° W
PV System Specifications (Residential)
DC System Size	1 kW
Module Type	Standard
Array Type	Fixed (open rack)
Array Tilt	38°
Array Azimuth	180°
System Losses	48%
Inverter Efficiency	99.5%
DC to AC Size Ratio	1
Economics
Average Retail Electricity Rate	0.163 $/kWh
Performance Metrics
Capacity Factor	12.2%

I have not played with all the numbers/options... The above is setup for Grid Tied utility interactive systems which are much more "efficient" vs DC off grid power systems.

You can see the PVWatts program seems to give about 10% higher "hours of sun" per day (roughly 1,000 WH of sun per sq meter = "1 hour" of noon time equivalent sun per day.

One of the neat things about PVWATTS is you can download a spread sheet of the hourly power productions for a "generic/average" years worth of data.

The above assumes a 1,000 Watt Array (smallest number PVWATTS accepts). And to compare with our manual calculations using January as an example:

• 5.03 "hours of sun" per day (PVWATTS) * 0.52 off grid AC system efficiency * 1,000 Watt array = 2,615 WH per day (Jan average)
• 2,615.5 WH per day (Jan) * 31 days = 81,085 WH per Jan = 81 kWH per Jan (vs 79 kWH per day from PVWATTS using "my" deratings for PVWATTS).
• 4.49 hours of sun per day (Jan from SolarElectric link) * 0.52 off grid AC system eff * 1,000 Watt array = 2,335 WH per day (2.3 kWH per day)

"How you run your system" (day/night power usage, type of batteries, depth of cycling, do you clean dust of panels, etc.) are all ~10% variation too.

A couple of items to think about... With Lead Acid batteries, they need "hours of sun" "on charge" per day... You cannot dump 3 hours of "fast charge" into a discharged Lead Acid battery bank and force a recharge to 100% without slowly damaging the battery (high current, high gassing, high temperatures). So--For folks that are off grid and need more energy in the winter--Using a mechanical tracking array to follow the sun across the sky gives "more time on charge" during the day (great for lead acid batteries--lower peak charge current, longer time on charge). An option to the mechanical tracking (since solar panels are so "cheap" these days)--Create two arrays, 1/2 facing south east, the other 1/2 facing south west... You lose a bit of overall harvest, but increase hours on charge (cheap panels=>Buy some more panels).

Something to think about is using LiFePO4 (or similar) Lithium Ion batteries... They simply need Amps*Hours to fill (nearly 100% efficient). So they can take a large array with winter sun (fewer hours on charge). Li Ion batteries have other issues (don't work well below ~40F, need battery management system to reduce risk of damage, and are still very expensive vs lead acid batteries--Although longer Li Ion life does reduce the cost "spread").

And sizing for variable weather conditions... Never plan on using 100% of a solar array's predicted output every day... The numbers are an "average"--Some days more, some days less... Some days much less (stormy weather). To minimize run time of your genset, and stresses on your family--Another suggestion... Break your loads into "base loads" (loads that need to be powered every day for living/work) and "optional loads" (washing machine, water pumping for irrigation, movie night, etc.). Typically suggest that base loads run at 50% to 65% of predicted output, and optional loads run on sunny days.

In the end--Sizing the array should take many factors into account. Make sure you "know your loads" (Watt*Hours per day, seasonal usage, base vs optional loads, etc.)... And do a full "paper design" first, before you go out and start buying hardware. There are "configuration" issues to make sure everything "plays well together"... It is very easy to buy hardware "on sale" that ends up not working well/or at all/for your needs.

There are no "right answers"--Understanding the math and assumptions under the calculations will help you decide your actual system design. And especially for off grid power--Yes, you will most likely end up with an "over sized" solar array--Solar is highly variable, and batteries need to be kept "happy" (operating within voltage/current/state of charge parameters). Batteries being some of the most expensive components, and the easiest to "murder" if operated incorrectly (typically under charging / over discharging is what kills batteries for most folks).

-Bill