Describe the bug
DC connected batteries can cause the "inverter total power loss" metric to display a negative number when grid charging

To Reproduce
Steps to reproduce the behavior:

1. Set up a DC connected battery with grid charging
2. Run the simulation
3. Look at the data table output for inverter total power loss during a step with grid charging
4. Observe that the number is negative

Expected behavior
It looks like a calculation needs to use fabs that isn't. This number should always be non-negative, like many of the other loss categories.

Screenshots

Observed on develop, likely older than that.