Though I haven't done much technical development since early this year, the nine installed septic system monitors are all working well, collecting and analyzing the activity of their owners' systems. Reviewing the data that is being collected I noticed something unexpected and pretty cool: When the septic pumps turn on, the amount of time they stay on is very consistent. I added a new charting option to the web site to show this fact. First, here is the 'classic' chart seen on the web site: It shows the cumulative time a septic pump has run in terms of "minutes per day", in this case broken down weekly:





As you can see, this system goes on from as little as 1m:15s per day up to nearly 3m per day. Now consider how long the pump turns on each time it does so:


This pump always runs for about 2m:20s. It's extremely consistent! The yellow bars show the median run time, and the dark error bars delimit the inter-quartile range (the range of values within which half the run times lie). This is just a statistically fancy way of saying that the pump is extremely consistent in its run time.

What does this mean and why is it cool? We have to consider what affects the cumulative time (the first chart above) and the run time (the second chart above). The cumulative time reflects how much you use your septic system. If you go on vacation it drops to zero. If relatives stay for the holidays and bathe a lot, the cumulative time will spike up.

However the pump run time is insensitive to all of that. The pump is switched on and off by a float switch down in the pump tank. The diagram at the top of this page shows the idea

http://www.plumbingsupply.com/alarm.html

(though the specifics of your system may differ). The run time is how long it takes for the pump to lower the water level from the 'on' position to the 'off' position, and this only depends on (1) how strong the pump is and (2) how much resistance it encounters. Since the run time is so consistent, we should be able to detect the beginnings of mechanical failure in the pump and/or minute changes in downstream resistance due to a developing clog.

And that, to me, is pretty cool.