AWE operation is for a long time. So the process of endosmosis (see texts and links above) is likely crippling. As indicated in the quotes above when 3.5 volumes of hydrogen leave the balloon, 1 volume of air enters, which in the long run gives an explosive mixture.
Perhaps a kind of valve located at the bottom of the balloon would allow the air which is heavier to be evacuated. However, the instability of hydrogen makes any transfer and manipulation perilous, as for slowly filling by using an electrolyser on board as described on AWES using LTA with PV for hydrogen production on board - #16 by PierreB.
Besides it (not burned) hydrogen “acts as an indirect greenhouse gas” (Other Greenhouse gases - Hydrogen).
I remember a way to make clean hydrogen I do strongly not recommend because it is very dangerous, the process giving off a lot of heat. Put aluminum, caustic soda (unblocker), and water in a plastic bottle or a jerry can, fix the opening of the balloon and … move away. I inflated like this several hydrogen balloons up to 1 m³ with 1 kg of aluminum.