Not for positioning, but when I had to take measurements of rocks.

I did some work taking oriented samples for a paleomagnetic study. In order to do this, one has to create orientation marks on the sample before removing it. Now the rock I was working with had a high magnetite content, was in an area that experiences higher than normal lightening strikes, and had been on the surface for tens of millions of years. As soon as you brought the compass next to the outcrop the needle would point directly to the rock. I had to get the samples so I borrowed a sun-compass from a geologist that used to work in the iron hills of Minnesota (where that's the only thing that worked). After studying the equation of time and other solar time keeping techniques (fascinating, really), I headed back to the field to get my samples. Mixed blessing when it comes to using a sun compass: the sun is often blocked by clouds. Bad because you can't get a reading and good 'cause you get paid to sit on an outcrop and wait for the sun to come back out. But, after all of that, the samples were worthless. The lighting that had magnetized the rocks had totally obliterated the p-mag signal.

Then an opportunity opened to take the samples in a tunnel under construction and presented a new problem. With the train tracks and the 30Kv powerline's induction field, I couldn't use a magnetic compass there either! My workaround involved protractors, levels, and PVC pipe to create a device that would sight down the tunnel (surveyed by laser) and give me a fairly good bearing.

I have encountered other highly magnetic rocks in my mapping that confused the Brunton. Sometimes just sitting on these rocks has demagnetized my credit cards.

The lesson, I suppose, is redundant systems and skill sets.