Possible factor-of-two error in the metric_battery_cycle documentation example?

[mpi]

Hi everyone,

I’m trying to calculate metric_battery_cycle from battery cost and expected cycle life, and I think there may be a factor-of-two inconsistency in the documentation example.

The example uses a £4,000 battery with 9.5 kWh capacity and an assumed lifetime of 6,000 complete cycles. It calculates:

Combined charge/discharge throughput per cycle:
2 × 9.5 = 19 kWh

Cost per kWh of throughput:
£4,000 / (19 × 6,000) = £0.0351 = 3.51p/kWh

However, the next paragraph recommends setting metric_battery_cycle to 1.75, halving that result again because the penalty applies to both charging and discharging.

Doesn’t the 19 kWh denominator already account for both directions?

My understanding of the code is that battery_cycle accumulates abs(battery_draw), and the optimiser adds:

battery_cycle * metric_battery_cycle

With a setting of 1.75p, the accumulated penalty over the assumed lifetime would therefore be:

1.75p × 19 kWh × 6,000 / 100 = £1,995

A setting of approximately 3.51p would allocate the full £4,000.

Would the intended formula therefore be:

metric_battery_cycle =
    battery cost / (2 × usable capacity × expected cycles)

with the result converted to the minor currency unit per kWh?

Am I interpreting the throughput calculation correctly, or is there another reason for the additional halving?

Thanks!

[gcoan]

thanks for raising, well spotted, it is incorrect, the second divide by two is not needed. I will remove it in the documentation.

By the way, I personally wouldn't recommend you set this metric, it will hamper Predbat's ability to charge/discharge the battery. I work on the principle that since I have already paid for the battery I might as well use it.
But of course your view of cycles and warranty may differ

Cheers

[gcoan]

documentation changed and awaiting PR approval