Power-limit loss when it would have clipped

Method

On every timestep on which the array operates at the MPPT minimum, its maximum power is converted through the efficiency curves at the floor voltage and compared with the AC limit. When the result exceeds the limit, the array would have clipped, and the whole power lost at the floor is counted in the power-limit row, Inverter Loss over nominal inv. power. Otherwise it is counted in the voltage-threshold row, Inverter Loss due to voltage threshold. The loss is never split between the two rows. When the floor point itself still exceeds the AC limit, the AC-limit rule raises the voltage from the floor and the whole loss is a power-limit loss.

The math

1. Power-limit row
2. Voltage-threshold row

$$L_P = \begin{cases} \Delta P & P_{out}(P_{mpp},\ V_{min}) > P_{lim} \\ 0 & \text{otherwise} \end{cases} \tag{1}$$ $$L_V = \Delta P - L_P \tag{2}$$

where

Symbol	Meaning	Unit
$L_P$	the loss counted in the power-limit row	W
$\Delta P$	the power lost at the MPPT minimum	W
$P_{out}(P, U)$	the AC output of the efficiency curves at DC power $P$ and voltage $U$	W
$P_{mpp}$	the array power at its maximum-power point after the DC losses	W
$V_{min}$	the minimum MPP voltage, the bottom of the inverter's MPPT window	V
$P_{lim}$	the AC limit of one inverter at the timestep: its nameplate, derated with the ambient temperature and lowered by a grid power limitation	W
$L_V$	the loss counted in the voltage-threshold row	W

References

1. Analysis of PVsyst 8.1.5 outputs shows this convention in its results.

API

inverter.vmin_loss_bucket = would_clip

Symbols follow the symbol registry.