Problems importing India meteo data with ASCII tool

[spelland74]

Hello,

We had been importing our own (3TIER) weather data for India successfully with past versions of PVsyst (6.42 is the one that we used for the longest time) using the ASCII import tool. Now, with version 6.47, we are having several issues. One seems to be a bug, the other is simply quite time-consuming.

To start with the time-consuming issue:

-We use ASCII import often, especially for India, where the time convention is UTC+5.5 (tm2 format doesn't correctly handle the half-integer UTC difference).

-Now in version 6.47 (and possibly some earlier ones too), we are asked to import with hour shift =1 if PVsyst calculates a time shift greater than 30 minutes, and with hour shift = 0 if PVsyst calculates a time shift less than 30 minutes. This is problematic for our India data, where the "normal" time shift is 30 minutes (i.e. data runs from 0:30 to 1:30, 1:30 to 2:30, etc). That means that for any non-zero deviation from this that PVsyst calculates, we need to import one way or the other. That's time-consuming, as it means potentially re-doing all our file imports twice (sometimes we import up to 25 or so weather files)

Now with the bug:

-I will send by a separate email one file that we have not been able to import properly in PVsyst 6.47 at all. When we put hour shift=0, time shift=30, PVsyst calculates an average time shift of -37 minutes and asks us to import with hour shift =1. When we import with hour shift=1 (time shift=-30), then PVsyst calculates an average time shift of 30 minutes (inconsistent) and asks that we import with hour shift of 0.

-Note: I tried importing with PVsyst 6.42, and there is no bug in that case, and the .MET file is judged to be ok. With PVsyst 6.47, we cannot even run it!

-If possible, it would be helpful if "hour shift=1" would be reserved for cases where the average time shift on clear days is one hour or more...