>Here's my two cents worth:
>
>    I'm doing a 3V app with a 16C54A that sleeps 99.9% of the time.  WDT off.
>The databook says average drain should be about 1 microamp.  Using 2 alkaline
>button cells @ 190 mah.
>    190 milliamp-hours is 190,000 microamp-hours, so at 1 microamp they should
>give me 190,000 hours, or about 20 years, eh?
Remember shelf-life - you need lithium for this!
Battery mfr's often give curves for life at given loads. There are
some issues at very low powers - maximum battery efficiency can depend
on drain, and I'm told it can decrease at very low drains!. 
Adding a capacitor to sink impulse loads when the thing wakes can also
help. 
>    I'm in the process of measuring this, but don't fully trust my
>instrumentation.
A 1K or 10K resistor inline with the supply and the mV range on a DMM
helps here!
>How do I a test for ten-year life in less than ten years?
Measure the AVARAGE current, including wake-up current - a while ago I
designed an integrating microammeter for this purpose (basically a
current shunt, precision opamp, integrator, and sample/hold) - it's
otherwise hard to measure avarage power for devices which wake up
occasionally. I'd expect an asleep PIC to take well under 1 uA.

Looks like I'll be using a NEC 4-bit uC for my product!
     ____                                                           ____
   _/ L_/  Mike Harrison / White Wing Logic / wwl@netcomuk.co.uk  _/ L_/
 _/ W_/  Hardware & Software design / PCB Design / Consultancy  _/ W_/
/_W_/  Industrial / Computer Peripherals / Hazardous Area      /_W_/