Different distances to toxic endpoints with different versions of RMP*Comp

I've noticed that for certain chemicals, RMP*Comp gives substantially different distances to the toxic endpoint than previous versions. Why?

In the current version of RMP*Comp, we have incorporated new chemical-specific distance tables for ammonia, chlorine, and sulfur dioxide. The generic tables are still used for other chemicals (you can see the generic tables in the Offsite Consequence Analysis Guidance (OCA Guidance).