For those who don't know, I'm an HVAC engineer. From time to time, the issue of heat pump efficiency comes up. When picking a system, obviously efficiency (cost) is important, so we need to decide between, usually, heat pumps (air source) and gas for a small split system. The government has a rating system for seasonal energy efficiency, and a heat pump can be rated according to coefficient of performance, but as COP changes with temperature, that doesn't really tell you very much. So my boss and I decided to do a real-world test of a heat pump and attempt to get it published. And so the next time someone asks us "heat pump or gas furnace?" we'll have a meaningful answer. There are a few potential issues with heat pumps: -As said, COP drops with outside temperature. -Capacity drops with outside temperature - so heat pumps require another heat source as a backup. -Depending on the weather, a heat pump may collect frost. A lot of frost. This has to be melted by switching the heat pump back to air conditioning mode. Yeah, that's right, heat pumps will start air conditioning your house in the middle of winter. That, of course, needs to be offset with supplimental heat. So, what is the real-world performance? So we selected the lowest capacity Carrier window heat pump and installed it in the window of my walk-out basement (ZQA101RB: http://www.docs.hvacpartners.com/idc/groups/public/documents/marketing/02-racbr07-25.pdf ). It has a listed COP of 2.8 and the marketing literature says "ZQ Heat Pumps use 1/2 to 1/3 less electricity than an electric heater (to 45F). That implies to me that a communications major who doesn't understand engineering didn't understand what a 2-3 COP means (actually, maybe it was written by an engineer with bad grammar, but whatever). But anyway, 2.8 is still pretty good. Sounds promising. Now as it turns out, my basement is big, but doesn't have a lot of load. It's a walk-out with two walls exposed to the air, that I insulated pretty well. I tried to match the heat pump size to the load so we'd get a high usage factor (and keep costs down). We attached temperature, voltage, and amperage sensors, and started-it up. As it turns out, the marketing info is actually trying to tell you something and hoping you aren't listening. That 45F isn't just a rating point, it's the point at which the unit cuts from heat pump mode and switches to electric resistance heating mode. It's just a coincidence, but the night I installed it was an extremely humid, somewhat warm winter night, of about 45F, with 100% humidity. And the unit didn't frost-over. We had a sensor attached to the reversing valve, and it never de-energized to switch to a/c mode. I can only conclude that the unit was designed with frost avoidance in mind. This makes sense, because despite still having a good COP of 2.5 at 45F, the low capacity and constraints of the electrical service mean you can't afford to switch to air conditioning mode to defrost it. The next few days were that record warmpth of a few weeks ago, so I got to see pretty well the performance profile (I haven't finished reducing the data yet). But anyway, in Pennsylvania, running as a heat pump only down to 45F just doesn't cut it. And we wanted to know how a heat pump really performs when it is cold out. So we're terminating the study of this unit (anyone wanna buy a heat pump?). I'm still going to do the full analysis, though, and we did get about all we need for it. Fortunately, one of my coworkers has a heat pump at his house that we're going to study. Hopefully, we have enough winter left to get good data. I'll keep you guys aprised of how it's going.