Would the constant movement of cars in the tunnel induce a wind, reducing the pressure differential?

I guess that depends a lot on the density of cars and topology of the tunnels (a simple loop would be ideal, I think).