big_m mentioned here: when adding the constraints to ensure that the head loss and the flow direction are aligned
During the development of the head loss constraints of a gas network, the big_m value had to be reduced to big_m = 1.1 max_head_loss instead of using a value similar to the heat physics head loss constraints where big_m = 2.0 2.0 * max_head_loss