I just read something by Matt Walker about my thoughts on a bypass. So I wish to add some clarity on my thoughts and then, I hope, all the folks with bypass experience can add their thoughts.
All of my experience with a bypass is in the steam punk testing contraption in my office. We are now a couple of months into the second year year of use.
I thought it would be something that could be used so that a fire could be started with less smoke back. But we really haven't suffered from much smoke back problem on this. I also thought if there was ever a time that wanted a rocket-ier burn, I could activate the bypass. This never was the result.
HOWEVER! The bypass did turn out to have a HUGE value in something I did not expect: it can double the efficiency of the mass heating the room. When the inside temp is, say, 70, and the outside temp is, say, 30, and fire is down to a few embers, I activate the bypass. Now, as the vertical exhaust does the thermo-siphon thing, it is not pulling any gasses through the mass. Air is pulled in through the woodfeed and out the roof. Without the bypass, air from the room would be pulled in through the wood feed and then through the mass and out the roof. So all that warm stuff I built up is just being taken outside.
Here is one amazing thing I learned with the bypass: I was routing the exhaust through the wall - low. The fire wanted to smoke back very much. The bypass was activated from the beginning of the burn. I thought I would close the bypass for a few seconds and force the exhaust through the mass. I was sure that this would definitely cause smokeback because now the exhaust would have to be pushed through a much longer path. The bizarre thing: no real difference. Maybe a little. What I concluded was that a person could, if they want, have 100 feet, or maybe 200 feet of run and the system would work about the same.
Good stuff Paul. I am a big fan of them for the reason you state, that you can bypass a warm mass when the thing is on it's way out to better conserve the heat in the mass. I also see them as a valuable tool that can allow savvy operators to build systems that either are too large for reliable operation, or for systems that see intermittent use and are started from cold often.
IIRC, your reservations regarding a bypass were mostly based on the goal of creating a system that was dead simple and fool proof for an operator of any level of skill or care. I think that's a dang good design goal, and as always my thoughts with regards to most of these techniques/design features is that they are all tools in the kit. Whether or not they are appropriate will change with every build according to the design goals.