I think I got it right now. As I said in my first comment, the temperature drops because the species is generated at its reference temperature, which is lower than the current temperature of the medium. The equilibrium temperature is necessarily lower.

Even though I was already defining an equivalent heat load to compensate that phenomenon, I was not obtaining constant temperature in the volume because I was not defining the heat capacity accurately.

I want to remark that, even though a generation term is similar to a generation because of chemical reaction, this is not treated in the same way. The heat associated to the generation from a reaction is the formation enthalpy as described in the equation (5-11) of the section

5.2.1. Heat Transfer Theory (ansys.com)

However, the heat associated to a generation term defined by the user is the sensible heat required to bring the species from its reference temperature to the equilibrium temperature. And this is, as far as I know, not described in the manual.