Fix PvModel output calculation

[danielfeismann]

The electric output calculation from PvModel seems to be partly wrong when assuming the elevationAngle as the angle between the horizontal plane / ground and the panel.

Assesment so far:

• Output of east-directed PVs are higher than expected values
• Output of west-directed PVs are lower than expected values
• elevationAngle takes all values, documentation gives just the interval from 0 to 90°
• Output of west-directed PVs with elevationAngle between 360° to 270° seems to be approx what would be expected for 0 - 90°

To do's:

• Do further investigations
• Find solution
• Extend test methods

[danielfeismann]

Result of further investigations:

• Weather data that is not in ZonedDateTime will be handled as LocalDateTime even if it is originaly in UTC.
• Given that the first daylight beams that reach a east-oriented PV panel based on the simulation time (e.g. 09:00 UTC) will be matched with weather data at 10:00 (LocalDateTime) since this is 09:00 UTC in case of daylight saving time (e.g. January) or matched with weather data at 11:00 (LocalDateTime) in e.g. June.
• In that case the production of east-oriented PV panels will be overestimated since the beam intensity is wrong.
• For west-oriented PV panels on the other hand, the last beams that could reach the panel from geometrical perspective are weak or even zero since the false weather data (one or two hours later) contains a lower or zero beam.

Solution:

• Make ZonedDateTime mandatory in PSDM.