For each transmission line included in a power flow study, the line charging is accommodated as terminal shunt branches of the equivalent-PI model.
The terminal shunt branches thus created across the network get accommodated in the Y-bus matrix, by summation to diagonal terms corresponding to appropriate buses.
The rest is straightforward: Once included in the Y-bus, the Y-elements are included as admittance factors in the active and reactive (P and Q) relations at different buses.
As a follow-up of #3, Jacobian terms are defined in the usual manner, and play the usual role in Newton-Raphson definition and thereafter the power flow convergence.
Tried to make this as precise as I could - hope that helped !!
There are some good examples in the book:
J. Arrillaga, N. R. Watson, "Computer Modelling of Electrical Power Systems", John Wiley.
And by the way, line charging can be significant in distribution systems as well, if the distribution routes involve significant length of cables.
Actually, voltage magnitude is dependent on reactive power level on a bus. If a bus is having more reactive power injection then the voltage magnitude of bus will be higher. Similarly, insufficient reactive power can cause a voltage deep on a bus. This indicates that the line charging VAR can affects load flow results.