When calculating fast_level, fast/slow time weighting is to be used. For this, as the standard points out [Fig 1, Sec 3.5, pp. 3], the filter should have a real pole at -1/tau. This filter is created in line 160 of iec_61672_1_2013.py as:

b, a = zpk2tf([1.0], [1.0, integration_time], [1.0])

The pole here is 'integration_time' which is tau. Shouldn't it be 1/integration_time?