I noticed there has not been one response to your post after several weeks. I do not profess to be an expert in such design of natural gas pipelines, but I wonder if installing the pipeline e.g. "on supports" (or piles etc.) in this area, perhaps also with some "redundancy" (e.g. in the way of heavier beam strength pipes and/or with more frequent axial spacing of supports?) might be one approach. If one or more supports e.g. were to succumb to a sinkhole, is it possible the line might still survive based on its beam strength and adjacent supports? [You might want to view some of the responses/linked references on the "seismic" thread that immediately precedes this thread.] There might also be some be some ball joint and/or specialty expansion devices (manufacturers) available for gas pipelines that might be used to provide for some local ground/line movement, perhaps also without unreasonably stressing the pipeline.