Account for distribution loss impact of DER

[elainethale]

From looking at customer_group.jl and regulator.jl, customers.d is consistently treated as inclusive of distribution losses, and is scaled by (1 - utility.loss_dist) when used to calculate quantities at the customer-level.

In contrast, DER energy data (rho_DG) is always used as-is, at both the customer and the system level. Given that the DER capacity is nameplate, I think rho_DG * capacity should be interpreted as meter-level, and thus at times inflated with the factor (1 + utility.loss_dist) when used at the system level (sometimes in regulator.jl, and whenever used as part of a VPP in utility.jl or ipp_group.jl) under the assumption that changes in net load will mostly not travel very far through the distribution system.

[Nongchao]

Ah, do you mean if BTM load is d_BTM, BTM gen is g_BTM, the load with transmission loss would be (d_BTM-g_BTM)*(1+loss_dist), that's why we need to scale BTM generation up by (1+loss_dist)? That makes sense. To be consistent, I guess when we get the load without loss, we currently do d*(1-loss_dist) (where d is the load with loss), maybe it is more accurate to do d/(1+loss_dist)?

[elainethale]

Yes, all that is correct.