Looking at the circuits build for a client during a Tor process, I have noticed the following: Four circuits were built.

On the first circuit the guard node was a1.

On the second circuit the exit node was a1.

On the third circuit the guard node was a1.

On the fourth circuit the guard node was a1.

What are the risk of simultaneously having circuits that use same node for guard relay and exit relay? P.S. This is not a hypothetical question. I have documented the actual circuits built by Tor process for reference.