Thanks Paul for your answer.
I understand SAM uses iterative algorithms and convergence error are part of them, that is natural. However, my query still remains given the relative error magnitude, at time step level, I can see on the SF HTF mass balance calculated using the variables presented in my initial query. I would like to take it forward differentiating different operation scenarios if you allow me.
(1) SF Off + TES Discharging + Cycle On: min rel. error = 0% / max rel. error = 0%
(2) SF On + TES Discharging + Cycle On: min rel. error = -0.10% / max rel. error = 0.09%
(3) SF On + TES Idle + Cycle On: min rel. error = -0.08% / max rel. error = 35.54%
(4) SF On + TES Charging + Cycle Off: min rel. error = 0% / max rel. error = 94.24% (single time step)
(5) SF On + TES Charging + Cycle On: min rel. error = -4.60% / max rel. error = 33.35%
The relative error size makes me think we are facing something else than convergence errors. I could be missing or misinterpretating the problem.
Warm regards.
Javier Pastoriza.