Hi Mara,
I guess your are measuring your in-house glass against another similar glass containing Br and I? Am I right? So, say, if you're using NIST 610 (provided it has Br and I, I don't know), simply reduce your data using the trace elements DRS. Do you see a signal for Br and I in your glass? That's the first thing I'd check. Also (I'm not a glass standard expert, be advised), you should probably check whether those elements are distributed homogeneously in the in-house glass.
Cheers
Luigi