Programmable thermostats are sometimes touted as a good way to save money. They are used to turn the heat down in the night when you are sleeping or during the day when you are at work out of the house.

But in our green show house in Brooklyn we are not using programmable thermostats. We plan on insulating so well that the house will not be affected that much by outside temperature. This means that if we did turn the heat down on a cold night it would take a lot more than one night to cool down. This means that the boiler wouldn’t have to come on anyway.

So having a thermostat that turns the heat down at night when we sleep is pointless since the boiler won’t have to come on until we are awake again anyway. I’m not doing a good job at explaining the logic.

That is why I have quoted a fellow listmember who did a better job of saying what I mean:

I was thinking about all these rocks people are throwing at Programmable Thermostats, and usually when I think it involves a model of some kind. In the 70’s, programmable stats were touted as a cost effective way to turn down energy use, and I’d never questioned that idea.

A UADT spreadsheet calculation will say that a programmable thermostat saves a lot of money. But it doesn’t account for mass.

Today’s model was a TRACE run simulating several buildings. While simple UADT spreadsheet calculations just consider insulation and temperature difference, TRACE is an energy simulation package that can take into account mass and thermal lag.

Building 1: 1024 square feet, R-13 walls, R-19 roof, lightweight frame construction, wood floor over unconditioned basement, leaky double pane windows on every wall, electric heat and AC, fixed thermostat. (basically the 50’s tract home I grew up in. )

Building 2: Same lightweight house with a programmable thermostat, set back at night and during unoccupied times in the day. 72F/65F setback in the winter, 78F/85F setup in the summer.

Building 3 – same size, but with heavyweight concrete walls, externally insulated at R-22, R60 roof, concrete slab with perimeter insulation, tight construction, same heat and AC, fixed thermostat.

Building 4 – same as building 3 (both would make nice F-5 tornado shelters) but with programmable thermostat.

Results were : the lightweight frame house saved about $60 a year, enough money to pay for a programmable thermostat in one year, but not any more than that. Whereas in the heavy building the programmable stat didn’t save a watt nor a dime.

The argument is the same as the argument against turning a water heater up and down. In a massive body, turning the stat down doesn’t change the temperature in the building. The building loses the same amount of heat, because its mass holds it at a nearly constant temperature, and you have a constant delta T across the insulation all night. Same heat loss whether you turn the stat down or not. You replace the lost heat the next day by charging up the mass.

In a lightweight building, the building really does cool off and delta T is less, therefore energy loss is less. You still replace some heat when you charge up the mass the next day, but it is less than you saved so it is a winner. But not by as much as people used to expect. I imagine that the lighter the building, the better these things would work. They might be good for trailer homes.

I can’t think of a way that a programmable stat would use MORE energy, as has been asserted, but if someone will propose a scenario I might put it into these models and see how it comes out.

So, Gennarro, if you are remodeling a Brick brownstone, don’t bother with a programmable stat, it probably won’t help you much.

–Lawrence Lile