Hydrogen pick-up test failures

Each month I send 2 sample titanium (T6-4) coupons out for hydrogen pick-up testing. They are first scotchbrited then cleaned in Diversey 909 then one is etched in nitric hydrofluoric at 100 °F for 15 minutes. The coupons are then wrapped in acid free paper and sent for analysis. Our internal spec states the results must not exceed 15ppm. What reaction must I have when we exceed that limit? I have had some results back as high as 41ppm and all I do is send another sample until the results come back within tolerance. The material I am using is certified but almost 2 years old. I thought the scotchbrite treatment should remove any nitrogen picked up while the coupon is in the cupboard. Both the coupons come from the same sheet of Ti.
I hope I have given sufficient information, I do not have anyone in this building to turn to for advice.

I believe that we punch the blank and test holes right next to each other due to natural H variations in the alloy.

Do you guys do that?

Have you studied the variation in H across your test panel? Do you have a feel for the amount of "noisiness" in this test?