Maybe OT: Save fuel with a setback thermostat?

If the fact that my furnace is metal qualifies for "metal content", then let it be so. If not, then this post is off topic. It could, however, relate to heating one's shop so----

Anyway, here's the question:

Does a setback thermostat really save you any fuel cost?

For our house and the way we live, the answer is "hardly none". I hooked up a simple thermistor to the DAU (Data Acquisition Unit) that's hanging off the back of my computer and have been graphing "furnace-on" events for several days. We stay up till about midnight or so and then the thermostat goes from 74° F down to 65° F until morning. The thing is set to go back to 74° at 8AM, but it anticipates for us, so on real cold days it starts in as early as about 5:30AM. The amount of time it takes to get house back to 74°F is almost equal to the amount of time that the furnace would have been running if there was no setback. During the night, the house hasn't ever gotten down to 65°F before the furnace came out of setback. Some googling on the subject sorta confirms what I found. The general answer is: "it depends". They say that the worse your insulation, the more a setback thermostat can cut energy bills.

My measurement process: I use a Radio Shack "Sensor Lab" to make up simple circuits. I bought a 9volt battery eliminator for it. I put the thermistor right on top of the closest wall register. The thermistor is simply in series with a 10K ohm pot across the 9 volt supply and I took the signal across the wiper arm of the pot and ground. These two leads connect to one of the inputs on my Dataq DI-194RS $25 (now $50) DAU, which connects to the computer through a serial port. The software that came with it makes the graphs. Only the top trace is active on the graphs referenced below. The units are volts. The range I chose is simply to give me a fairly full chart excursion. The peak voltage corresponds to about 108°F and the overnight min is about 63°F.

Here's a link to a graph of our home furnace cycling rate for a typical

24 hour period with a cloudy day. The graph starts at about 08:30.

formatting link

Here's a link to a graph of our home furnace cycling rate for about 2 24 hour periods with a sunny day then a cloudy day:

formatting link

What have you guys found?

Pete Stanaitis

---------------------------------------

Reply to
spaco
Loading thread data ...

The actual input range of the A/D is 0 to +4V or +5V. You can change the full scale range to increase the sensitivity by removing the offset and attenuating resistors.

formatting link
jsw

Reply to
Jim Wilkins

Hello, Jim.

The input range of this DAU is 0-10 volts on the outside of the box. I added the 10k pot so I could reduce the current a lot more than just applying the whole 9 v across the the thermistor, since I was 1) originally using a 9 v battery and 2) to minimize the heating of the thermistor. I saw no need to mess with the DAU circuitry, as long as I could get a useful excursion. For determining the actual on and off points, I import the raw readings into an Excel spreadsheet and mess with them there. Of course, all of this is really overkill, because I could easily enough just take the signal from the gas solenoid, or something in that area. I just wanted to do it this way for the heck of it. It is kinda interesting, though, to see the rate of change of the temp as it all happens. The graphs I show have been considerable compressed. There are a little over 900 samples taken per day.

I really have been delighted with this inexpensive Dataq DAU. I used it a couple of years ago to run a 1 1/4 year solar energy test with one of those HF 45 watt PV kits.

Pete Stanaitis

-------------------

Jim Wilk>

Reply to
spaco

An inexpensive thermocouple amplifier for it:

formatting link
The high temperature limit is a bit low for some of my projects but I may buy a few to play with. The Small Outline package can be spider- webbed to a DIP component platform.

jsw

Reply to
Jim Wilkins

yes

Would be interesting to know the outside temperatures.

First question is, "why do you care?". If you care about cost, you should measure what you pay for. Eyeballing a temperature graph has insufficient resolution. While the "loss" is relative to temperature difference, you don't know the temperatures where the loss happens.

You should always set the thermostat as low as you can tolerate, but no lower. It costs what it costs. If your measurement leads you to lower the thermostat to save cost, you weren't honest about setting it as low as you could tolerate.

Assuming you're comfortable and you care about what it's costing you...beyond what it says on the gas bill...measure that.

Faced with a similar question, I put a flapper on a microswitch, set it on a register and stuffed the signal into the serial port of an old Palm IIIc. After some quality time with a stopwatch, the furnace and the gas meter, I know the relationship between fan on-time and gas consumed. I graph gas consumed.

I've been struggling to understand why my gas consumption doesn't track outside temperature. After doing the math, I discovered a possible explanation. The graph resolution is equivalent to 17 watts of electricity converted to heat. That's WELL below the noise level of the measurement, but does seem to explain the anomalies.

Gas consumption tracks "inversely" the added 200W of heat while I watch TV. Go figger. Outside temperature is around 40F this time of year.

So, what have I learned from all this?

1) I turned off the setback and run the house at 65 degrees 24/7 in winter. I'm home 24/7 and sleep when I feel like it. YMMV.

2) Air movement is EXPENSIVE. My furnace fan runs 1.5 minutes longer than the typical 6 minutes of heat. You get some of that back in winter, but it's still a significant electricity consumption and 3x the cost of gas. That's the dirty little secret of high efficiency furnaces. Fans run longer and that doesn't figger into the efficiency numbers. Summers in Oregon are cool enough to open the windows at night. I'd been running a big circulation fan all night. I learned that I could run the air conditioner all night for not much more $$.

3)Floor insulation may not be all it's cracked up to be. Temperature in my crawl space was 55F when it was 20F outside. After adding floor insulation and heating duct insulation, it's still 55F under there. Heat loss thru the floor insulation should be proportionally better, but the deltaT isn't much.

4) The better your insulation, the less "anything you do inside" matters.

5)I managed to consume a bunch of time working on this, but in the end, the answer is simple. Use as little energy as you can tolerate. I really didn't need any measurements to determine that.

Are we having fun yet?

Reply to
mike

-------------------------------------------------------------- Home Page:

formatting link
If you're not the lead dog, the view never changes Doubt yourself, and the real world will eat you alive The world doesn't revolve around you, it revolves around me No skeletons in the closet; just decomposing corpses

-------------------------------------------------------------- Dependence is Vulnerability:

-------------------------------------------------------------- "Open the Pod Bay Doors please, Hal" "I'm sorry, Dave, I'm afraid I can't do that.."

Reply to
JR North

Thanks for your insight. The graph is simply the visaul representation of the data. I represents about 900 samples per day. I calculate the "percent on" for each day from that. I can also get the on length on each cycle that way to determine the difference between setback and no-setback.

It'll be below zero every night for the next 4 or 5 here and it isn't even "winter" yet. If I had your warmer climate, I wouldn't worry much, either. Our propane bill went from $2400 to $4400 in one year, so it's a real issue for us.

Pete Stanaitis

mike wrote:

Reply to
spaco

Yikes! $4400 would be hard to swallow. We have been heating our modest 1100 SF home in Northern Minnesota with off peak electric for about $90.00 a month. We used to keep our day and evening temp at 70 but this year I lowered the temp from 70 to 65 in the late afternoon and evening with a short bump up to 65 at 6:30 in the morning. I keep it at 60 at night and during the day when my wife and I are out in the shop. On sunny days the south facing windows help immensely. We often see temps up to 80 in the living room because of the sun. I should try a steady 65 to see if it would affect the electric bill. Steve

Reply to
Up North

Interesting. What do you do for cold junction compensation with a device such as this?

Are you aware of the Arduino? The woodgas guys at allpowerlabs have fitted one out with LOTS of stuff like this. Looks like a really neat almost general purpose process controller.

formatting link

Reply to
spaco

Holy Shit, Maynard! I'd say it's time to buy a more efficient furnace, triple up on the insulation, install triple-glazed winders, put dampers in the kitchen & bath fans, and caulk, caulk, caulk!

-- You can't wait for inspiration. You have to go after it with a club. --Jack London

Reply to
Larry Jaques

That info gives an upper limit of about 1000C for a 5V supply but it has a much higher supply capability so it implies the temperature limit can go higher. Some of my interests include glass blowing so being able to monitor temperatures upto 1300C would be good but that also typically requires a R or S type thermocouple (platinum + platinum rhodium) for longevity. As they have a lower output/ degree C then maybe they will work OK with a 5V supply which would be easy. Do you have any idea if they can be used directly with non K or J thermocouples?. Maybe a software map to do any correction required.

Reply to
David Billington

After posting I read the full data sheet. The device reads its own temperature to compensate, or if you short the t/c inputs it outputs its temperature, at 5mV/C.

I've considered learning the PIC because I've seen them embedded in quite a few products, like Lithium battery pack monitors. But for logging data an old ~15 Watt laptop on AC makes more sense to me because I can write substantial programs in QBASIC and have access to the Windows file system, at least for 8:3 filenames.

An engineer I worked for encouraged me to use the parallel port for general purpose I/O on his projects and it worked very well.

The data format from the DI-194 is messy but not THAT difficult to capture and untangle. I also wrote a driver for the Radio Shack serial data output multimeter, which is opto-isolated and useful to capture AC voltage or current. With QBASIC I can read data from both.

The Dell Diagnostics partition on some older machines is FAT32 and runs DOS7 from Win98. I edited the autoexec.bat with Knoppix to remove all the reboots so it works as a DOS + Basic partition, and they are incredibly fast on a >2GHz PC. An integer FOR loop takes one nanosecond. Unlike Windows, DOS has almost no overhead, just the clock tick every 55mS.

jsw

Reply to
Jim Wilkins

I can suggest a couple of tweaks that will save you some money. We also heat by propane which is about as expensive as electric resistive heating ( at least it is here ). So we have the thermostat set to switch to a much lower temperature several hours before we go to bed. And use a space heater to keep the room where we are warm if needed. So my guess is that you could set the thermostat to setback at about 9 pm instead of midnight.

I also do not set the temperature as high as you do. Just wear warm clothes and have the thermostat set to about 68 degrees.

In the mornings I wait as long as possible before cranking up the heat. If the weather is sunny, I will shut the furnace off while the house is still pretty cool. We get a lot of solar gain.

Dan

Reply to
dcaster

Since you don't need extremely fast response times for glass, why not just go with a stainless steel sheathed C/A TC? IIRC, you can get them with the TC junction welded to the end of the sheath or insulated from it.

Pete Stanaitis

-------------------

David Bill>

formatting link

Reply to
spaco

Yes, but how much depends on your usage. Houses lose heat at a rate roughly proportional to the temperature difference between inside and outside. If the inside temp is reduced, then heat loss will be reduced during that time compared to what would have been lost if indoor temp remained constant.

Let's say the outdoor temp is 20F, indoor temp is 70 so temp diff is

50degF. If we reduce indoor temp to 60 then temp diff becomes 40degF and heat loss will be reduced by 20%. But if it ramps down slowly to 60 and then immediately starts ramping back up, the temp diff averaged over the period is 45 degF and loss is reduced by only 10% over that period. If that period is 8 hours and the house is kept at 70 the rest of the day (16 hours) then you've saved only 1/3 of 10% or about 3.3%.

Better insulation reduces total heat loss so the savings becomes 3.3% of a smaller total bill.

A setback stat doesn't save me any money because I don't use one -- and I used to work for Honeywell! I have at least a couple of 'em NIB somewhere, not motivated to install them. The stat on the wall is an original Chronotherm with the electromechanical clock that quit working a couple of decades ago. The new ones are smaller, so if I installed one of them I'd have to paint the wall. I hate painting.

Reply to
Don Foreman

On Sat, 4 Dec 2010 17:30:10 -0800 (PST), Jim Wilkins

Spaco is using a thermistor, not a thermocouple.

What does "spiderwebbed" mean? Is that when you stick the SMT part on a board upside down with glue and then solder #30 wires to its leads?

Reply to
Don Foreman

Just make a nice brass or aluminum trim plate to cover the old paint. My dad preferred oak, to metal. ;-)

Reply to
Michael A. Terrell

By C/A do you mean chromel alumel, if so that would be a K type IIRC. The trouble is they don't live long at temps in the range 1300C and neither do stainless sheaths, even the high grade ones especially with certain glasses such as leaded glass. The normal thermocouple used would be R or S and in a ceramic sheath as they can stand the temperature for long periods.

formatting link

Reply to
David Billington

Right side up, on one of these:

formatting link
jsw

Reply to
Jim Wilkins

In Europe, IME, type R or S (Platinum-Rhodium) thermocouples are common in this kind of application. In the US it's common just to use a big fat (eg. AWG 8) cheap open thermocouple and replace it every now and then.

Best regards, Spehro Pefhany

Reply to
Spehro Pefhany

PolyTech Forum website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.