Leaving the "lights" on - add $800 to our gas bill

I've collected a few more data points this winter altering how Crista and I use the thermostat.  The short version is that by being lazy with the thermostat, our estimated annual gas bill has increased by over 800 therms compared to using a six degree setback.  That is an increase in consumption of 179%.

From late January to early February, Crista and I did not touch our thermostat.  It stayed at 68 degrees, 24 hours a day.  After taking three meter readings over that time period, PRISM estimated our annual usage to be 1,376 (+/- 73) therms.  Just as a refresher, with the house under the exact same conditions using a six degree setback, the house's annual consumption was 494 therms.

So by being lazy, or comfortable, or whatever one's reasons for not setting their thermostat back, the additional running of the furnace added over 800 therms of energy usage.  However, more important is the inverse.  By setting the thermostat back six degrees during work and while sleeping, we were able to save over 800 therms and lower our consumption by 64%.

As is evident from the PRISM plots, only three data points were used to estimate the annual savings without using a setback.  Which means while the normalized annual consumption estimate is reliable, it would be poor form to put too much emphasis on that estimate.

To further explore the thermostat's effect on our house's energy consumption we're now using a three degree setback.  Six degrees may be a bit extreme for many people.  Three degrees might better represent a more acceptable level of setback for most people.  It is also more affordable for us to run our house at for longer periods of time.  That way more data points can be collected and PRISM will have more data to model.

Un-greenwashing

Some Truth in Advertising.

Really Good

Those of you who haven't already found Radiolab, this might be a good introduction to the show.  It is just in time for the holidays, without hitting you over the head about the holidays.  Good for getting one thinking about being Good.

Retrofit, phase 1

The next spring our First time home buyer 0% interest loan (we got in before it was a full blown tax credit) was put to some use.  Dirk & Trevor Fyffe installed a 40,000 btu American Standard furnace with an AFUE of 95.  They also installed the new thermostat.  I also did a little air sealing in the attic.  Mostly along the bathroom wet wall and the top plate going down the middle of the house.  Dirk & Trevor had also re-wired the old knob & tube, so those holes were also sealed.  All told, the attic went from a 45 Pa pressure difference with reference to the house, to about 49 Pa pressure difference.  I also had the box sills and crawl foundation walls foamed.  About 3" in the box sills and about 1.5" on the crawl walls.  The house was also used for building technician certification events.  Meaning the uninsulated sidewalls had dense pack cellulose installed.  After all that, the house's blower door reading is about 2300 cfm @ 50 Pa.

Notably missing from the efforts are the following:

• No new windows were installed.
• No new doors were installed.
• No additional attic insulation was installed

So I still have the windows from the 1920s when the house was built (albeit with leaky aluminum storms).  The two doors are metal with foam insulation & storm doors.  And I only have an average of about 2" of old rockwool insulation in the attic.  Which resulted in the following savings the next winter:

The annual usage went from 772 therms down to 494 (+/- 12) therms.  A savings of 278 therms (36%).  The energy intensity dropped to 6 btus/sqft*HDD.  

During this time, there was a 6 degree setback in the t-stat.  So that is the baseline for my next experiment.  Data is currently being collected using a 3 degree setback to see how that changes the house's energy consumption.

Garfield, circa mid 80's

Crista and I bought this house late in the summer of 2008.

• There were a wall or two without insulation.
• No insulation below the floor.
• A leaky attic access and open-ish crawlspace vents.
• A 100,000 btu draft hood furnace (installed in the early 1980's).
• A blower door of about 3200 cfm@50 Pa.
• A thermostat that probably came with the furnace, mercury switch and all.
• About 1000 sqft of above grade living space with an additional 400 sqft of basement space.

If I had wanted to get the full effect of how the house would "normally" perform I would have left the thermostat set at 68 degrees all winter long. But Crista and I did manual set backs at night and during the day when the last person left the house (if they remembered).

As a result, the weather normalized annual usage was 772 therms (+/- 41 therms). Below is the consumption run through PRISM:

The house also has a gas water heater and gas cook stove.  Despite those gas baseload appliances, 95% of the gas consumption went to space heating.  The house's energy intensity was about 10 btus/sqft*HDD.  Not bad, but not too good either.

Even though we bought the house in 2008, it had the same energy characteristics as it did back in the mid 1980s.  Well, maybe it has a bit more wall insulation.  At first glance, the annual consumption is quite a bit below average (average is about 1000 therms/yr) and the EI of 10 indicates that, while there are some savings to be had, I would be surprised to see big reductions from a whole house retrofit.

A year after this, I was a little surprised.

Meena helps spread the EE word in Chicagoland

Here

and

Here

My light switch thermostat

If there is going to be a focus of this bit of internet clutter, I think it will be on doing little energy experiments and posting my amateur findings. The first experiment or project I'm curious about is the issue of thermostat (t-stat) setbacks. Weatherization tends to gives lots of credit to the energy saving capabilities of programmable thermostats. Generally speaking, not using energy in the first place saves more energy than using energy more efficiently. And setting the the t-stat back doesn't cost anything (compared to the cost of adding insulation or installing a more efficient furnace).

The problem I have is the same thing that Energy Star had several years ago. Many weatherization (wx) clients don't use the programmability of a programmable t-stat. There are many obstacles preventing a complicated programmable t-stat from being used to the fullest.

• Many elderly Wx clients find the programmable t-stats "too complicated" and would rather have the old round one. Keep it simple.
• Fewer people have routine schedules that fit into a regular program. Many programmable t-stats have a 5-2 configuration with separate programs for week days and weekends. That is fine if you happen to work a typical 9-5 work week with weekends off. But if you're working a service sector job or have a random work schedule, there is no point in programming anything in.
• And there is the path of least resistance. It is just easier to keep the t-stat set the same all the time. The house is warm when you want it warm.

I was turned onto this non-programmable t-stat from one of the heating trainers at work. Being non-programmable, there is no clock to set or times to worry about. What it does have is a nifty little button with a dollar sign. I think that button is the sweet spot between a plain, non-programmable t-stat and the sophisticated programmable t-stats.

I think it is fantastic because it works with odd schedules. Basically, it turns the t-stat into a modified light switch for the furnace. When talking to wx clients, an easy energy saving tip to stress is turning lights off when not in the room. With this t-stat, the same principle applies. About the leave for work at some odd time? Press the setback button (turn off the switch). Back home? Press the button again & the t-stat temperature rises back.

I've been using it for over a year now so there is plenty of data to look at what a 6 degree setback looks like. Next will be some posts describing the house as it was when Crista and I moved in, what changes we made to bring the house to its current configuration. And by then some time will have passed to see what different set back settings do to the energy consumption.