CompareToEnergyPlus

Comparison to EnergyPlus

The unit tests are doing a good job of ensuring specific outputs of the functions. They will continually be improved. In addition, the demos on the wiki are demonstrating the functions over time with explanations. As a final check before these are considered validated, we will run EnergyPlus and compare the outputs.

EnergyPlus Model Setup

A shoebox model has been created and the east and west walls will be used as the demonstrative surfaces. The relevant solar angles will be reported from EnergyPlus, which has been validated extensively, and used as a verification (near-validation) source. The model will include the following characteristics:

• Design day located in Golden, CO
• Site:Location:Latitude: 39.57 degrees NORTH
• Site:Location:Longitude: 104.85 degrees WEST
• Winter Date: 12/21
• Summer Date: 7/21

Case 1: Site-specific angles

Three site-related angles are reported from EnergyPlus

• Site Solar Azimuth Angle
• Site Solar Altitude Angle
• Site Solar Hour Angle

Here is the comparison for the Winter day:

Note that I had to negate the hour angle for the comparison because EnergyPlus uses the convention of positive hour angle in the morning and negative in the afternoon. Overall the results match up perfectly.

Here is the comparison for the Summer day:

Note that I left daylight savings time OFF because in EnergyPlus, design days do not have daylight savings time enabled (at least not by default). Again, the results line up wonderfully.

Case 2: Solar Incidence Angle

EnergyPlus is capable of reporting the cosine of the solar angle of incidence on a particular surface. For this case, we will add that variable to the east and west facing surfaces of the shoebox model. The resulting comparison is here:

The results are a great match. The only real deviation is that EnergyPlus reports a positive value at the 5th hour shown. This is because EnergyPlus is actually calculating this value at each time step (15 minutes), and then averaging to report the hourly value. The library just samples once, in this case at the 30 minute mark of the hour. So this is showing a difference because the sun must be slightly up at the :45 mark of the hour, so EnergyPlus shows a positive value. This could be remedied by calling the library on a 15-minute interval.

On that same note, during initial comparison, I found a deviation of approximately 30 minutes between the library results and EnergyPlus. Investigation revealed this was because of how EnergyPlus was sampling over the entirety of the hour and averaging, while my original library calls were at the top of each hour. Once the sampling was changed to the 30 minute mark of the hour, the results got much better.

This essentially validates the current capabilities of this solar calculation library.