SP1ML & SPSGRF Spirit modules inconsistent current with STM32L151 in Stop mode

Goodday,

After a few months on the  SP1ML (STM32L151+SPIRIT combined) then switching to SPSGRF (STM32L151 seperate from the SPIRIT module) I am experiencing the same phenomena on both designs. My software toggles the system in and out of STOP mode using timer 7 (to go to shutdown) and RTC to wake up. Eight out of ten times current consumption will be 50uA which seems correct with external pull-ups etc, but two out of ten times the current jumps to  220uA.

I have removed all additional semiconductors from my design and reverted back to the original firmware from ST. I am now toggling the unit in and out of shutdown mode with an external wire connected to ground and still every now and then I get 220-300uA in shut down mode instead of 19uA.

My project is current sensitive and will be asleep 99% for months at a time so I cant afford to take a chance where it goes into stop mode drawing 220uA.

I have narrowed it down to the SPIRIT Module if I remove it and only toggle the STM32L151 I get 0.4uA consistently.

To date I have build around 40 different units (SP1ML & SPSGRF ) and they all do the same.  I also looked for eratta on the Spirit module.

I suspect the problem lies here 

Anyone ever came across something similar? Any ideas on how to narrow it down, I looked at clocks, GPIO states, even the ST StDPeriph_Driver_PWR.

Thanks in Advance

Melt