ioprog

I finally achieved a target run time (estimated) of a year on a couple of AAA batteries for the NRF24L01 and the STM32F030. The current flow is mostly zero but once a second the MCU wakes up and transmits a message leading to a current waveform like this:

The best estimate I can come up with for the average current flow puts it below 100uA which should allow very long run times off a pair of AAA batteries.
Code, circuits, results and further discussion is here