units.csv: two Dutch entries duplicate an English unit

In [`DataFiles/units.csv`](https://github.com/buildingSMART/bSDD/blob/master/DataFiles/units.csv):

| line | entry | duplicate of |
|---|---|---|
| 240 | `Jaar,Jaar,NULL,NULL` | `yr,year,NULL,YR` (line 613) |
| 195 | `GJ/jaar,GJ/jaar,NULL,NULL` | `GJ/yr,Gigajoule per year,NULL,NULL` (line 196) |

Jaar is Dutch for year. Anyone matching a Dutch label exactly lands on these instead of the English
code, and neither carries a QUDT reference while `yr` does. I work around it, but localised names as
codes will keep causing this.

Worth deprecating or pointing them at the English entries, after checking whether a published
dictionary already uses them. Related to #105.

Start in DataFiles/units.csv at lines 195, 196, 240, and 613, then review related issue #105 and check whether a published dictionary already uses the Dutch entries. Reconcile the duplicates by deprecating or pointing them to the English entries, preserving the appropriate code and QUDT reference, and verify both cases in the resulting CSV.

Written by the indexing model from the issue text.