Different Peak Sun Hours data

**LETitROLL** · 11-06-2015, 10:10 AM

I'm sure you know the H in KWH is hours, hours by itself without the fixed unit of production (kw) married to it wouldnt help anyone design a system. If you bought 1,000W worth of solar panels and knew how much sun you get at your location in december but did not know the relative strength of those sun hours you could not convert it to expected gross output from your 1,000W array. As for why the different websites show different results, i have not looked at them close enough to help with that, i assume at least part of it may be different assumptions about panel angle, temps, and all the other data that goes into calculating the answer (each exact install will vary a little with those particulars)

**J.P.M.** · 11-06-2015, 12:04 PM

What you are seeing is, for the most part, some of the technically and commonly accepted ways of measuring the solar resource.

Because I'm not as well versed in off grid or smaller PV systems as some other subjects, I'm not sure of the origin or practicality of the concept of "sun hours". Not knocking it - I just don't see the use, probably due to my ignorance. I think the term "sun hours" may have an assumed value of 1,000 W/m^ buried in it somewhere (similar to the way pipe schedules (pipe wall thicknesses) have an allowable stress buried in them), so that, for example, "4 sun hours" may mean 4kWh/m^2, or 4kW/day without a surface area reference, or something else. I've seen "sun hours" have the units of hours, kWh/m^2, kWh/day, and others, often used by folks who are as (or more), confused than you, but don't know it.

I think the concept of sun hours causes a lot of unnecessary confusion for little benefit as exemplified by your post.

As for responses to your confusions:

The unit of kWh/((m^2)*(day)) is the amount of energy (and usually/often referred to as "insolation", and for solar work usually, but not always W/m^2 per time period) received from the sun by a surface over the course of one day. The term actually refers to the time integrated sum of the instantaneous solar flux (called irradiance, with units of W/m^2, or kW/m^2 ) received by a surface of 1m^2 in area. That (time summed) insolation term can be for a day, an hour, a minute, week, month, year, or any time period.

BTW, the insolation or irradiance incident on a surface is different for different surface orientations. For example, a gimbel mounted solar panel always pointed toward the sun will usually have more incident irradiance and insolation than the same panel mounted in a fixed position. Converting the amount of energy incident at different incidence angles is an interesting subject, but sort of off topic at this juncture.

The insolation term (W or kW/m^2 per time period) is the total solar energy received by a surface over that time period and is the quantity of solar "fuel" available.

As for trust: All those sources are probably using the same units, but perhaps wearing different clothes. They probably don't use the term "sun hours", but they're probably all correct.

If I were you, I'd check out two sources: PVWatts, as you already have - only this time read the info help screens and chase down/root out some of the reference material to be found in the info/help screens. The second source is something called the "TMY manual", also found on line. The manual explains the logic behind how the (estimated) solar resource (as insolation) is derived for a bunch of U.S. and some other locations.

If you want a correct and thorough discussion of the subject, also see Duffie and Beckman: "Solar Engineering of Thermal Processes", Wiley, ISBN #:0-4714-51056-4.

**Sunking** · 11-06-2015, 12:44 PM

"1: Why the units are in kWh/m² day and not in terms of hours?"

Kwh/m2 and Sun Hours are the same number and interchangeable. 4.1 Kwh/m2 = 4.1 Sun Hours.

**Esi** · 11-10-2015, 07:56 AM

Thank you all for your comments,

Ok, so I understand that PSH, is the number of hours in a day during which the radiation is at its maximum of 1000W/m².
I think I should just accept that there is a mismatch between the terminology used and the units (and kick the guy who came up with the name).
Where I come from, the units are just as important as the formula and they do tell a lot about what is being quantified.
There is noone that can ever convince me that hours and kWh/m² are the same thing. The first is a measure of time and the second is a measure of energy.
(Kilometers an hour has also hours buried in it; - can you relate that to energy as well?

)

Thank you for your recommendation, LetitRoll. I will investigate the plane of the insolation for those data.
Thank you JPM for all your clarifications. I don't know what TMY manual is. I will check it out immediately.

**Esi** · 11-10-2015, 08:26 AM

Ok, just to clarify: the above data is for various module tilts.
Some sites optimize the tilt or let you change the tilt and recalculate PSH.
Even with the same tilt, the data are not exactly the same, but similar enough to make some assumptions when designing an off-grid system.

**LETitROLL** · 11-10-2015, 10:40 AM

Maybe this will help, since you bring up the kilometers per hour example, of course you cannot tell energy from that, but you can tell distance (but not everything about total distance) if someone says "Jim traveled 10 KM per hour yesterday" you wouldn't know how far his trip was at all , when you multiply a time component (6 hrs) "jim went on a 6 hour trip at 10 KM per hour yesterday" now you have something. It is kind of the same with the solar, once you have your exact sun hours for a given location you can multiply that by your panel(s) capabilities (watts) and get the total information about what you can expect for production. But you have to know the relative strength of the sun, just that it is up and shining on your panels does not tell you how much energy is being collected, the elevation of the sun in the sky relative to your place on earth makes a world of difference to the amount of energy striking your panels, official sun hours numbers adjust the hours component of the number while the energy per given area stays fixed, when the sun is low it just takes longer to collect a certain amount of energy so they reduce or dilute the hours component accordingly, thus 8 hours of weak sunshine may only = 2.3 sun hours. Without the KWH/m2 you could not calculate total energy to be collected, it seems to be equal to the speed component in the KM example "jim traveled 8 hours yesterday" how many days will it take him to get to dallas? You have to know how efficient his movement is (KMH) then you can figure all the rest.

**J.P.M.** · 11-10-2015, 12:34 PM

You're welcome. The TMY manual is the NREL blurb that explains a lot about the concept of Typical Meteorological Year. The TMY, is one, and at this time the most common tool used in many solar modeling tools as sort of a standard for available solar radiation at over a thousand U.S. locations and many foreign ones as well.

(Dirty little open secret: What most folks don't know, or care about is that, for the most part, the TMY hourly irradiance values are themselves estimates and are often and for the most part, synthetic, being generated from other site and location variables, some of which are also synthetic. The TMY irradiance values are probably not too bad, but they are usually not actual recorded data. This is not an exact science.)

Relative to your comment about possible sun hours being "...the number of hours in a day during which the radiation is at its maximum of 1,000W/m^2" : Not correct,. or at least a good example, to me anyway, of why the term "sun hours" or "equivalent sun hours" can be confusing, not to mention unnecessary.

For one thing, a level of solar irradiance of 1,000 W/m^2 is pretty high and rarely seen in most locations, except perhaps @ elevations > ~ 2,500 m or so, and then only on very clear days.

FWIW, that level of irradiance is used as part of the PV panel rating criteria, partly because it is sort of a maximum level that might commonly be seen and partly because it's a nice round # - that makes things easier to calculate.

30 or so years ago, there was nowhere near the irradiance data that's available today. That availability makes the concept of "sun hours" something of an anachronism. These days, most anyone can find information on estimated, and in many cases actual, irradiance data for many locations, and for just about any orientation, fixed or moving for the sake of a few mouse clicks. 30 years ago, I'd have killed for the type of data I now get on a min. by min. basis from the instrument on my roof. I'd also have saved a lot of time not writing computer code to transform GHI raw data to other useful forms as that information is much more available today and better than ever.

Knowing what all the data means and how best to use it still seems a problem however.

With so much solid information readily available, the concept of sun hours seems unnecessary, incomplete and, as your quoted statement seems to show, more than a bit misleading, and more often than not sending folks off in the wrong direction and into more confusion.

Maybe the best use of the terms "sun hours" or "equivalent sun hours" may have is as some relative gauge as to the "sunniness" of one location, orientation or time, to another, or several. Even then, I'd think for the sake of clarity, or at least to avoid muddying the waters, instead of saying, for example, "4 equivalent sun hours", saying instead, " 4,000 Wh/m^2". The latter, while still of very limited usefulness, at least does not imply power (by having the "hour" term) and would be more technically correct while perhaps not introducing any more confusion from the use of the term " sun hours".

Still, the "4,000 Wh/m^2", while it says something about energy, says nothing about power, and it says nothing about whether that energy is incident (and if so, at what angle, and/or the distribution of beam, diffuse, or reflected components), or turned into something useful.

Maybe my ignorance is causing me to miss something, but I don't see any advantages worth the price of the probability of confusion introduced by the concept of "sun hours".

**Esi** · 11-11-2015, 07:09 AM

Dear Solar Fanatic,

I am amazed at how involved you are with this field and how much you know about it. I, on the contrary have just started learning the basics.
I did check TMY yesterday, but it wasn't of any use, since it does not contain any other data outside of the US. And I know about how the data is gathered and estimated. You are right, there is no way of knowing any radiation data unless you measure it; even then, there are a lot of errors and other factors to consider. And I know that all the weather data for my country is all interpolated from weather data taken at two weather stations in two neighboring countries-imagine that!

In Europe PVGIS of JRC/EU is the equivalent of NREL/PVWATTs and, as far as I know is the only free and reliable database one can use.

Back my PSH question, which until now, I have only seen it being used to design off-grid PV/hybrid systems, I have decided to do the following:
1. Find the month of the lowest average irradiation from the whole operation period, at that specific location (which is December if the operation of the system is year-around)
2. Find the tilt (optimize the system) for the entire operation period
3. Find the PSH (from 2-3 databases online) for the given month at the given tilt.
4. Hope that I won't get any calls around Christmas time and hear someone yell at me, if their critical load is off!

Cheers!

**J.P.M.** · 11-11-2015, 10:54 AM

Esi: For a semi technical but still understandable discussion and explanation of the availability of the solar resource (and many other things solar for that matter) see : Duffie & Beckman, "Solar Engineering of Thermal Processes", ISBN #: 0-471-51056-4. It covers most of the basics. The subject is more broad than deep. Lots of simple concepts put together in understandable ways. Eat the elephant one bite at a time. Fortunately, most of the bites are small. Somewhat unfortunately, there seems to be a lot of them.

Good luck.

**sensij** · 11-11-2015, 12:44 PM

Originally posted by J.P.M.

Still, the "4,000 Wh/m^2", while it says something about energy, says nothing about power, and it says nothing about whether that energy is incident (and if so, at what angle, and/or the distribution of beam, diffuse, or reflected components), or turned into something useful.

To this discussion I'd like to contribute that not all sources may be reporting the same value, even when the units are the same. For example, PVWatts reports the insolation in the plane of the array in (kWh/m2/day), while many other sources report the global horizontal insolation in the same units. Without a look at the documentation (among other things), it is impossible to know just from the units what you are looking at.

From what is sounds like the OP is trying to accomplish, I would suggest the following:
1) Don't trust monthly averages. If you are designing an off-grid system, drill down into the actual hourly data and see what a typical worst case week will look like. This may give you some idea of the generator supplement required for your PV system in real-world conditions.

2) Why would you design the system to maximize overall output, and not tilt the array to maximize the winter output (if the system is to be in service in winter). Even better would be to look at the seasonal load and adjust the tilt to minimize the generator run time annually, which may skew steeper or flatter depending on what that profile looks like (lots of AC? lots of electric heat?).

3) To achieve accuracy at the level of precision it appears you want, free data sets may not be the best choice. Purchasing the data (and modeling) from a company like SolarGIS may be a better way to reliably achieve your goals.

4) If you want to stick with free, I would suggest that you learn how to use NREL's System Advisor Model. The engine is very similar to PVWatts, and it will accept a variety of weather file formats (epw, tmy, etc), which are frequently available for many more locations than just the data set included in PVWatts. Those files can be manipulated as you see fit to test model sensitivity to changes in irradiance, temperature, or other environmental factors. A benefit of the NREL stuff is that it is extensively documented... it is possible to recreate the results yourself by applying the same published, peer reviewed models. Not every modeling source is so open about how they generate their numbers.

**Esi** · 12-01-2015, 09:24 AM

I somehow missed the last message from sensij, sorry.

Unfortunately SAM and NREL provide weather data for US, not SE European countries.
We have purchased PVSOL, which uses weather data from Meteonorm, which again are interpolated for my place and, I am sure I can extract the GHI or global irradiation at the module from the simulation runs.

Yes, I could optimize the tilt of the off-grid system for the worse month of the year, but do I need to recalculate that that tilt is going to provide enough PSH for each of the 11 months? Or should I just assume that? See, I don't have enough experience to know that optimizing for the worse month will provide enough PSH for the other months as well...

**inetdog** · 12-01-2015, 10:34 PM

Originally posted by Esi

I somehow missed the last message from sensij, sorry.

Unfortunately SAM and NREL provide weather data for US, not SE European countries.
We have purchased PVSOL, which uses weather data from Meteonorm, which again are interpolated for my place and, I am sure I can extract the GHI or global irradiation at the module from the simulation runs.

Yes, I could optimize the tilt of the off-grid system for the worse month of the year, but do I need to recalculate that that tilt is going to provide enough PSH for each of the 11 months? Or should I just assume that? See, I don't have enough experience to know that optimizing for the worse month will provide enough PSH for the other months as well...

You can, except in the wildly divergent situation of being next to an overhanging cliff, count on the output for the non-winter months being better than the winter month output for that same tilt, just not as good as it would be with the optimal tilt for Summer.
If your usage patterns require more energy per day in the summer, then you will need to look more deeply.

Although it does not have a full range of closely spaced reporting points, pvoutput.nrel.gov will do the panel geometry calculation for a lot of sites outside North America.

Different Peak Sun Hours data

Different Peak Sun Hours data

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment