The SMOKE manual says the SCCXREF file “provides mapping between reference SCC (i.e., SCC8) and full SCC (i.e., SCC10) for Smkinven to convert reference SCC-level inventory to full SCC inventory.” There is then an example given:

Each Reference SCC points to more than one Full SCC. So, how does smkinven split the reference SCC inputs across the full SCC outputs that GRDMAT will then use?

We are running SMOKE-MOVES in v4.5.

Hi Farren, Once Smkinven converts reference SCCs to full SCCs for onroad mobile sources, the rest of program like Grdmat, Temporal, and Spcmat will expect full SCCs. So you need to update/create your GREF (Grdmat cross-reference input file) with full SCCs.

Thanks BH. Yes I got that much figured out… but what is still unclear is how smkinven splits the reference SCCs across full SCCs when there is more than one match in the XREF file (e.g. in the example).

This reference SCCs are used for onroad VMT and Vehicle activity dataset which does not vary by emission processes (exhaust, evaporative, refueling, break, tire,). By using reference SCCs for these activity inventory, we were able to reduce the size of inventory files. However, SMOKE does need them in full SCCs during another step like chemical speciation since the speciation varies by emission processes. So, SMOKE internally converts reference SCC-based activity into full SCCs ones so then we can apply properly profiles by full SCCs level. It means that the same amount of reference SCC VMT will be assigned to the full SCCs internally. So then same amount of activity for all full SCCs will be used for the rest of SMOKE processing.

Hi, The idea here is, all processes (exhaust, evaporative, refueling, etc) use the same activity data in its emissions calculations. So, the SCCXREF takes each non-process-specific SCC from the activity data (last two digits 00), and maps to one or more process-specific
SCCs (last two digits 62, 72, 40, …) according to the SCCXREF. The activity data is copied, rather than split, to each full SCC. For example, if you run a Smkreport, you will see the VMT for 2201210200 repeated for 2201210240, 2201210262, and 2201210272. Grdmat,
Temporal, and Movesmrg only use the full 10-digit SCCs.

Thank you, that is helpful.

To follow up, I am surprised that the SMOKE-MOVES tool “” create a SCCXREF for rateperdistance that includes refueling. I have always considered refueling to be “off-network” (e.g. occurs at gas stations). Though, running MOVES standalone does show refueling emissions in both rateperdistance and ratepervehicle.

What do raterperdistance (“on-network”) refueling emissions represent? And, should they be allocated to gas stations? It seems like it would be double counting to put refueling in both rateperdistance and ratepervehicle portion of SMOKE-MOVES.

Hi, refueling is allocated to RPD because the amount of refueling needed is proportional to the amount of driving done. During spatial allocation it can be allocated to gas stations if an appropriate surrogate is available, but this is
not currently where it is being allocated due to the difficulty in obtaining an accurate spatial surrogate for gas stations nationwide.