How to calculate solar panel hourly power output?

[gocool767]

I'm trying to think of a way to calculate the power output of a solar panel by taking in information from hourly weather data from weather.com (e.g. sunny or cloudy?, temperature, etc.) and basic characteristics of the location and panels, but i cannot find or think of a way to do so. Basically, my final output would be a list of power outputs from hour 0 to 23. Hour 0 will have a power output, hour 1 will have a certain output, etc. Does anyone have any ideas as to how I can accomplish this task.

 

[johnbbahm]

There should be published tables for your area with the average number of sun hours.
From there just use the equations.
http://www.recgroup.com/en/aboutsolar/basic-solar-maths/
I think there are tables for both fixed and moving panels.
Always plan for a little more than you need, as the panels degrade over time.

 

[gocool767]

But how will i be able to get hourly solar panel power output from this? My method must also take in environment data (sunny vs partly cloudy vs cloudy), which i don't see how this method takes into account, unless my understanding of your method is mistaken.

 

[johnbbahm]

You won't, Because of the margin of error, the best you can hope for
is an average (maybe) monthly output.
If you could tell how much sun would hit a panel during a given hour of a given
day, you could do someones weather forecast.
http://www.longtermsolar.com/solar-sunlight-hours/
On any given day, it could be raining, or cloudy, but average over
a month, can be fairly accurate.

 

[Emeline]

Hi,
I am also looking for the calculation of an hourly solar PV output curve. I understand it would not reflect the energy produced in a month or a year, but I want to optimize my hourly production for a specific power demand. Basically I want to play around with the azimuth and tilt angle to maximize my PV output in the afternoon when my load is higher. Any idea on a software/excel sheet that could help me do that ?
Thanks,
Emeline

 

[russ_watters]

Welcome to PF!

Here's the published data on clearness and solar irradiance, hourly, for creating such a model:
http://rredc.nrel.gov/solar/old_data/nsrdb/1991-2005/tmy3/

 

[Emeline]

Thanks Russ. Now how can I translate this into a PV system output ? I need to take in consideration the plane orientation of the panel...

 

[anorlunda]

Hi,
I am also looking for the calculation of an hourly solar PV output curve. I understand it would not reflect the energy produced in a month or a year, but I want to optimize my hourly production for a specific power demand. Basically I want to play around with the azimuth and tilt angle to maximize my PV output in the afternoon when my load is higher. Any idea on a software/excel sheet that could help me do that ?
Thanks,
Emeline

If you really want to optimize it, don't use any tables or predetermined calibrations, build a feedback control system instead.

In overly simplified terms, change the direction of the panels, if the change makes more power, keep changing in the same direction, otherwise change in the opposite direction. Repeat for all axes of freedom. That way the system "seeks" the optimum at any time and any weather. Of course a real life system is not quite that simple, but I hope you get the idea.

I would expect that the forums on Rasberry PI or Ardiuno computers might turn up some ready-made open-source software to do exactly that.

 

[Emeline]

Hi Anorlunda, thanks for the methodology. It's exactly what I am trying to do. My only problem now is that I am not trying to maximize the output of the plant (for that I can just use any PV software, compute the yearly PV output, change a bit the design, recompute,...). What I am trying now is to produce more in the afternoon than in the morning. Basically I currently through away PV power by capping my inverter output in the morning due to very low loads (no storage in the system), whereas the power demand is very high in the afternoon-evening. I won't be able to shift all my PV output to the evening time, I know the yearly production will be lower than positioning the plant straight to the South, however, maybe the loss is lower than what I dump every morning...
I really need to model my solar output hourly to do so... HomerEnergy provides the hourly curves but they are not easy to export. I was wondering if another software could help me compute that at a technical level only, I don't need to assess the financials of the project for this exercise.

 

[johnbbahm]

Perhaps you could attack the problem from the other way.
What time of day does you peak load start, and where is the sun at that time.
Set your panels so they are most normal to the sun around the time the usage starts to peak.

 

[anorlunda]

Based on your comments in #9, I think we are talking about two different things. It seems that you are trying to calculate the best fixed orientation for your panels. I was talking about an active system with motors that changes the orientation minute by minute. That even has a name, a heliostat https://en.wikipedia.org/wiki/Heliostat
An active system will optimize the orientation in the afternoon hours (and all other hours) independent of what happens in the morning. Almost by definition, any fixed orientation is not as energy efficient as a heliostat.

But let me also ask about your efforts to use the morning power. Hot water heating is typically something that can easily be scheduled to happen only in a specified time window. If you heat/cool with a heat exchanger, you could also add a water reservoir and a heat exchanger to improve the efficiency of your heat pump. Solar energy could heat or cool water in the reservoir during the morning. Then, of course there are batteries. The recent announcement by Tesla got a lot of publicity. Green Mountain Power in Vermont, just started selling Tesla batteries to their customers.

 

[CWatters]

What I am trying now is to produce more in the afternoon than in the morning.

I asked a question about this elsewhere. I was told that some programs allow you to enter details of when the panels are in shadow and optimise annual output accordingly. So to find the angles needed to maximise output in the afternoon just tell it the panels are in shadow before midday (they suggested using a "horizon file" to block the sun totally before mid day?). A program called PVGIS was suggested but I've not used it.

 

[russ_watters]

Thanks Russ. Now how can I translate this into a PV system output ? I need to take in consideration the plane orientation of the panel...

I use the data for a different purpose and exactly how you would, and I'm not sure exactly what you want, so you may need to take a swing at it yourself. But for an idea:

One of the provided points is irradiance perpendicular to the sun. If, separately, you get a model of the sun's position in the sky, you can then use trigonometry to calculate the solar panel output for each hour of the year.