Heat, Thermostats and Serious Content (photo: Dan Zeng)
Studies show that people can be lazy, intimidated, etc. But I want to discuss a significant reason pointed out in the study: people have an incorrect “mental model” of how programmable thermostats work (PDF).
A mental model is just how you picture something working — how you understand stuff in order to get through a complicated world, right or wrong. A classic mismatch of mental model and reality is that “the computer” is the the screen, rather than the part that has the CPU, Memory and Disk in it (leave it to Apple to make a computer that matches peoples’ mental models!)
Apparently a certain Alaskan Senator had the mental model of the Internet as a “series of tubes.” But I won’t go there.
My hope is that where it matters, we can get a proper mental model that helps us make good decisions. Here are some that caused people to not use programmable thermostats.
Incorrect Mental Models of Thermostats
(photo daveelf/fourworlds)
Assumptions
I’ll assume the thermostat is used for heating — if you live in Hotlanta, everything is the same for cooling.
I’ll also assume that it’s colder outside than inside, and its a given temperature, like 30 degrees.
It’s also important to know that a thermostat is just a “furnace switch”. It turns the furnace on when the temperature in the room is colder than the set point. It turns the furnace off when it warms up to the set point. (There are some clever thermostats, but in the end they switch the furnace on or off.)
So, what did those studies find?
Thermostat Is A Volume Control
One incorrect mental model is that the house will warm up faster if you turn up the thermostat higher. (The mental model may be of a knob on a stove, or a volume button on a radio.) Generally a furnace turns on and heats water or air which is then circulated around the house making the air warmer. In reality, the longer the furnace runs, the warmer the house gets.
Same Energy To Heat Up
Another incorrect mental model is that it takes the same amount of energy to warm a house from 68 to 70 as it does from 70 to 72. Nope: it requires more energy as the difference between outdoor and indoor temperatures increase.
And neither of these misconceptions have anything to do with programmable thermostats. Things get more fun when you add this twist.
Heat Is Stored
There appears also to be a notion that heat is stored in the walls, so warming up your house is really a matter of filling the walls with heat — kind of like warm bricks on a fireplace. But the main thing that makes you feel warm is the temperature of the air in the room; all walls do is slow down the loss of heat that makes the air feel warm.
Heat Momentum
Some believe that there’s a sort of “momentum” so that if your house temperature falls really low, it takes more energy to heat, than if you just kept it warm. Yes, it takes longer to heat a house from 60 to 70 than it does to heat it from 68 to 70, those last few degrees are the hardest. The key fact here is that during the time the house was gradually cooling, you were not running the furnace and therefore not using extra energy that would just heat up the great outdoors.
The Cost Of Incorrect Mental Models
These misconceptions, and others have lead people to use programmable thermostats incorrectly. The study showed that because of peoples’ misconceptions, the possible savings of programmable thermostats were lost.
(I am an engineer, and I think I know why this happens. Engineers often tend to believe that other people think the way we do. If something is obvious to them, it’s obvious to everyone. Actually, this may be true of all people, but it’s the engineers who design thermostats. And I blame them for designing things so poorly, they are used in a way that does the opposite of what is intended.)
But, until we engineer types start getting more socially adjusted, here’s a crack at explaining how heat works — a new mental model.
My Analogy for Heating Your House
Pump Me Up! (photo: 24thcentury)
At first, it’s easy. As the ball gets fuller, the air leaks out faster, so it’s harder to get to “full”. Once it’s full, you still have to keep pumping it up since air leaks out. If you stop pumping the air still leaks out — quickly at first, and more slowly as the ball gets flatter.
So in this silly analogy, a full beach ball is a warm house. Pumping uses energy, like the furnace. The leak represents the way your walls actually work, some are better or worse at slowing down the rate of heat loss, but they all leak.
So what strategies balance our desire for a full beach ball against the amount of energy needed to keep it pumped up?
You could run the pump just before you used the ball, but then you have to wait (manual control). If you really wanted the ball full whenever you wanted to use it, you could run a motorized pump that ran whenever the ball got a little flat (normal thermostat). If you were clever, you could put the pump on a timer so that the ball was kept full only at times you might need it, but let it go flat other times (programmable thermostat).
Using this analogy, it makes more sense (to me) that a programmable thermostat makes the furnace run for a shorter time overall than a regular thermostat.
Yes? Not so much? If you don’t like mine, what’s your analogy? You can certainly do better than a leaky beach ball! All comments accepted. The best response will get a free beach ball.
(Credit for getting the ball rolling, so to speak, to Chris at MapAWatt.com who had a question from one of his readers, and reached out to me and Allison Bailes of EnergyVanguard.com to validate his explanation. So I asked someone to make sure I had it right.)
So enough with the analogies. What about some actual facts?
The Science of Heat Transfer
Scientist, Not Andrew (photo: aeter)
So I asked for help from my friend and almost-classmate, Andrew Robbins, who took the same physics class I did. He aced it. He then went on to get a Masters in Mechanical Engineering from MIT, and if that weren’t enough an MBA from Harvard. I didn’t share my beach ball analogy (for fear of being ridiculed) and just said I needed some help on how to explain how heating a house actually works.
He responded:
It is not like heating a pot of water, which would be enthalpy. The air in your house doesn’t store a lot of energy. This is about maintaining the heat in your house against an external drain – energy transfer out of the house. For the most part, the energy drain will go with the C*(Tin-Tout).
[A brief interjection: Andrew is a very funny guy, a great dad, and beyond brilliant. But he is an engineer, and assumes the rest of us know what “C” is, and what “Tin-Tout” means. Here, C measures how poorly the wall keeps heat inside. Tin-Tout is the difference between inside temperature and outside temperature, or temperature differential, or Delta T.]
You can lower the heat transfer by lowering C; that is, adding insulation, sealing air leaks, etc… Or you can lower the temperature differential. If the ambient temperature outside is 30 degrees in January, the difference between heating your house to 68 vs. 70 should be (38/40) or a 5% reduction in energy. But you have to put on a sweater and who wants to do that!
[See -- I told you he was funny!]
There are three types of heat transfer that might matter and one that doesn’t:
- Conduction (e.g. transfer through studs in the wall)
- Convection (warm air embracing you)
- Thermal Radiation (think of a fire)
- Direct exchange (cold air leaking in and hot air leaking out).
Thermal radiation goes with the temperature differential to the fourth power. I don’t think that it is a significant factor because with delta T to the fourth power, a small temperature differential is not a big deal. It is also the difference of the outer wall’s temperature NOT the inside of your house.
Conduction and Convection both go linearly with delta temperature.
Direct air exchange will also go with delta temperature. This is like heating a pot of water where you keep putting more cold water in the pot. You need more heat if the water is colder.
So, just as I said: it’s like pumping up a leaking beach ball. Get it? Right? Yeah, me neither.
I suddenly feel less bad about washing out of physics.
And just because you made it all the way, here’s my first attempt at an Xtranormal video. I don’t think MS would have liked this.
Tom,
I don’t believe science anyway!
And I love Delta Tea!
Sarah Palin is pretty.
Great article.
Comment by Chris — October 7, 2010 @ 2:05 pm
Inspired by you, Chris!
Comment by Tom Harrison — October 7, 2010 @ 2:08 pm
Nice post, Tom. (Thanks for the mention!) I like your discussion of mental models – and their frequent misalignment with reality. It’s important to understand how people construct knowledge if we want to show them how things really work.
I like the beach ball analogy, too! It works, at least for me. Keeping your house at a high Delta T, analogous to keeping the beach ball full, clearly takes more energy. When it’s flat, there’s no Delta, and thus no loss.
Your friend Andrew’s explanations are good, too, except he didn’t get radiation quite right. First, it’s proportional to T to the 4th power (T^4), not Delta-T to the 4th power. Since it’s based on the absolute temperature scale, with room temperature approximately 300 K, the net radiation radiation would be T2^4 – T1^4, not (T2 – T1)^4. Put 300 and 299 in for T2 and T1, and you’ll see the difference.
Anyway, radiation can be the dominant effect on comfort when you have poorly performing parts of the building envelope, like uninsulated walls. The air temperature may be fine, but the radiant temperature difference between your body and those uninsulated walls can make a room uncomfortable. (See my article on the 4 factors of comfort: http://hub.am/9JTNRl.)
This topic has been on my list for a while, so look for my article on thermostat setbacks in the next month or two. I’ll link to yours and Chris’s, too.
Comment by Allison A. Bailes III — October 7, 2010 @ 2:36 pm
Allison –
So I tried the calculation and came up with a big number, but I don’t really know what it means other than it’s big!
I wonder if we’re conflating comfort and heat loss? No doubt, radiation affects comfort. But is it a major factor in heat loss, and subsequent indoor air temperature falling? (That’s not a rhetorical question, it’s a real one — I am a software guy, like I said, I washed out of my Physics major :-).
Off to EnergyVanguard to learn some real facts.
Tom
Comment by Tom Harrison — October 7, 2010 @ 2:52 pm
Ah, yes, I did start talking about comfort and forget to come back to heat loss. And yes, my attempt to explain the scale of the difference between Delta-T to the 4th and the difference between two numbers to the 4th power failed, as it seems this explanation of my previous explanation is failing also. Sorry about that. Get out your old physics book and look up the Stefan-Boltzmann law if you want to get into the details.
As for heat loss, most of it would be through conduction and convection, as Andrew said, as long as we’re talking about insulated houses. Uninsulated homes can lose significant amounts of heat through radiation, however, because their outer surfaces are at a higher temperature. That’s what we see in those pretty pictures taken with all the infrared cameras out there these days. If those uninsulated homes are on the prairie and have lots of exposure to the night sky, radiative losses can really hurt.
Comment by Allison A. Bailes III — October 7, 2010 @ 3:44 pm
I don’t think you’re going to get very far trying to change consumers’ mental models, because the UI in part creates the mental model. It looks like a volume control. So the consumer thinks they’re “turning up the volume.” The end. So I would say, change the UI. The mental model by itself is too hard to change. It involves math, as you say. I can barely solve the spam protection sum on your site!
So, just to state the obvious question: What is the appropriate UI for a thermostat? I.e., one that “forces” an appropriate mental model.
What UI would give the visual impression that:
(1) The house will NOT warm up faster if you turn up the thermostat.
(2) it does NOT take the same amount of energy to warm a house from 68 to 70 as it does from 70 to 72.
I would guess that the answer is in the monitoring devices you’ve described elsewhere on your blog, that represent energy usage as dollars spent. If, when I turned the thermostat to 72, it told me immediately that I was now spending money at a rate of $100/month (instead of $70/month), that would be a great help.
Of course, it would be anti-capitalist to introduce such a device cheaply and universally, because now the consumer can’t be as easily swindled to just want it to be “warmer” with no consequences.
Comment by John — October 8, 2010 @ 11:17 am
John –
Totally with you on the “can’t change mental models” idea.
Logitech makes a remote control called “Harmony”. It has buttons “Watch TV”, “Watch a Movie”, “Listen To Music”, “Off” and “Help”.
When one of those buttons is pressed, all the magic happens (TV on, AVR on, input set to X, DVD off, and so on). If something fails (because IR is spotty), hitting Help automatically tries to get you back to a known state, and if that doesn’t work, it prompts you through questions like “Did that solve the problem?” and “Is the TV still on?” and so on. Initial definition of the various components is done through an Internet site that has a huge library of devices and their IR codes. A USB connection downloads the config to the remote.
It is brilliant.
Why couldn’t UI for a thermostat be similar? There are now some internet-enabled thermostats, as well as ones having predictive capabilities and Zigbee integration. A motion detector would be cool.
And it could be part of the display for electricity monitors.
Now imagine an integrated site that would ask you about your house, your heating system, e.g. BTU/hr, your zipcode and so on. Then it might ask about your usual schedule, your current heating/cooling bill and your objectives (save money, save planet, etc.). Then it could run for a few days and send you and email saying something like “It looks like you could save x dollars by setting your thermostat down further during work days” or “You saved X dollars this week — was the house comfortable? Want to try for more?” or whatever.
Periodic emails with progress reports. Comparison with similar houses in your area — “Your house seems to use a lot more energy than others — have you considered getting an energy audit to find ways to improve the efficiency of the house?” This kind of stuff already exists in Microsoft Hohm. The bits and pieces are mostly there.
I could go on and on.
Gee: a business idea is born!
(I won’t dignify your anti-capitalist remark with a response :-)
Tom
Comment by Tom Harrison — October 8, 2010 @ 12:16 pm
[...] Harrison, over at Five Percent, unpacks these myths and gives you the breakdown on how you should REALLY think of your thermostat [...]
Pingback by Everything I needed to know about home energy I learned from a beach ball. | Firefly Ecometrics — October 8, 2010 @ 2:49 pm
Hi Tom,
I like the mental model of a leaky ball, that does it perfectly for me. I think the idea of hitting the right mental-model is more important than people generally understand, it makes such a difference.
Allison has already commented on the difference between (T2-T1)^4 and (T2^4-T1^4), but you can break that down further into something more understandable. Apply the binomial theorem twice, and you get:
(T2^4-T1^4) = (T2^2+T1^2)(T2+T1)(T2-T1)
and if T1 is fairly close to T2, that reduces to (something big)*(T2-T1). In other words, radiation loss is proportional to delta-T, just like convection and conduction.
Of course, that doesn’t say anything about which contributes more to any specific case, and I suspect you’re right that radiation matters less, for the reasons you give.
Comment by Tony Wildish — October 9, 2010 @ 8:53 am
Tony –
I
blamegive credit to Andrew for all scientific content :-)Tom
Comment by Tom Harrison — October 9, 2010 @ 10:29 am
Oh, I’ll repost my analogy here since you actually asked for it here ;)
Imagine you have a bucket. It leaks, because there is a line of small holes drilled up the side. So the fuller it is, the faster it leaks.
Sometimes you need the bucket full; other times you can live with it half empty. Clearly you let it be half empty when you can be, because it leaks less that way. And filling it to the top only when necessary doesn’t take any “more” water….
BTW did you get your heater cycle measuring device going? I am sorely tempted to get a wifi thermostat and hack it to measure & graph this…
Comment by Eric — February 9, 2011 @ 12:16 am
Eric –
Here’s the link to my prototype datalogger using the components from Lego Mindstorm NXT kit: http://fivepercent.us/2009/02/26/heat-and-hot-water-energy-usage-for-my-house/
Comment by Tom Harrison — February 9, 2011 @ 9:43 pm
[...] this can be a confusing concept to grasp, this is a really great site that explains common mental misconceptions of how heating your home works, and the scientific [...]
Pingback by Energy Saving Myths: Part 2! « The Minnesota Energy Challenge Blog — February 8, 2012 @ 10:09 am