sdz,
I have always used ASCE 7 loading coefficients for open signs and lattice frameworks. I did some research on the topic several years ago and I read (I have forgotten where) that the normal force coeficients range from 1.6 to 2.0 for parrallel trusses placed somewhat close together. This seems to hold up in conveyor design also. If the conveyor has hoods I use the arched roof coeficients from ASCE 7. (This will be your worst case). If the belt and idlers are not covered, I use the open signs coefficients on the projected area, measured from the top chord. This seems to give good results. Back in the old day's people around here used 20 - 30psf depending on the height and we have never had a failure in 26 years. I think if you look into this methodology you will find it satisfactory. If you are nervous, don't consider the low importance factor. The trussed tower force coeficients are another method, but my experience, including hurricanes, seems to agree with the open signs and lattice frameworks method.