Currently, the controls of the ThermalTank plugin will start to utilize the chiller as the charge goes to zero to alleviate potential issues with suddenly starting up the chiller as the tank is emptied of charge. This won't always be seen since the current sizing always oversizes:
- If the requested size is the equivalent of 22.1 tanks, 23 tanks are used, resulting in an oversize by nearly an entire tank. For 100% load coverage, the tank should still have nearly an entire tank of charge as the discharge window closes.
- If the requested size is the equivalent of 22.9 tanks, 23 tanks are still used, resulting in not much oversizing at all. For 100% load coverage, the tank may not have enough charge as the discharge window closes to prevent the chiller from coming on.
The second scenario could result in substantial energy use during the peak window, and that should be corrected.
The current calculation of size is (roughly):
$$ N_\text{tanks} = \text{ceiling}\left(\frac{\text{load}}{\text{capacity}}\right) $$
where capacity is the per-tank capacity and the ceiling function rounds up to the next integer. If a fraction $x < 1$ is added, the new calculation would be
$$ N_\text{tanks} = \text{ceiling}\left(\frac{\text{load}}{x\ \text{capacity}}\right) $$
and the system is at least oversized by $x$.
Currently, the controls of the ThermalTank plugin will start to utilize the chiller as the charge goes to zero to alleviate potential issues with suddenly starting up the chiller as the tank is emptied of charge. This won't always be seen since the current sizing always oversizes:
The second scenario could result in substantial energy use during the peak window, and that should be corrected.
The current calculation of size is (roughly):
where capacity is the per-tank capacity and the ceiling function rounds up to the next integer. If a fraction$x < 1$ is added, the new calculation would be
and the system is at least oversized by$x$ .