boiler therms per degree-day increases during heating season

Nom_Deplume

Hi...I am a homeowner with an HTP Elite FT condensing boiler. I have been monitoring gas consumption per degree-day for the last three years. Every season, it starts most efficient in October (about 0.25 therms/degree-day) but then steadily gets worse, ending up at about 0.31 therms/degree-day in May. That's almost 25% worse. Next year, it's back to 0.25 in October, etc.

I realize that therms/degree-day is not a standard measure of efficiency, and depends on the house; it's just something that is easy for me to calculate. But my question is about the trend. If the efficiency were worst in Jan/Feb I would attribute it to less condensation, but that's not what I am seeing.

PS the therms data comes from the gas bill, and the degree-day data from www.degree-days.net. Summer gas consumption is about 0.5 therms/day, so I subtract that baseline from the winter consumption before calculating the ratio. I clean the boiler myself each summer; there are some coffee grounds fireside and in the trap, but not that much, i.e. nothing to indicate that this cleaning will somehow improve the efficiency.

Is this normal, and what could it be due to? Thanks in advance!

Zman

I wonder if your net efficiency is taking a hit due to short cycling on the shoulder season days. How is the boiler sized as compared to the load.

Jamie Hall

Interesting. I've never noticed a pattern like that -- but I have steam.

EdTheHeaterMan

Have you ever let the boiler operate for some time thru the summer (before maintenance) to see if the 0.5 therms per day happens in June or July? What other gas appliances are connected to the gas meter? Cooking, Fireplace, Clothes dryer? If so, have you also accounted for different usage of those appliances from one season to the next?

All things being equal... You do have a interesting query! To really know the answer you could place an hour meter on the gas valves of each appliance that uses gas to see what the usage of each appliance actually is.

Hot_water_fan

In addition to the above advice: It’s correct to try to remove domestic hot water/cooking/grilling/etc. by subtracting out summer usage. However, DHW gas usage increases as incoming water temps decrease, which would happen starting in Oct and continue through May. Look at weatherspark for your city - they might have a water temperature graph like this (an approximation of what your incoming water temps might be, depending on your water source):

fentonc

I noticed something very similar, that I had initially been attributing to a more-than-linear increase in gas usage vs heating-degree-days, but I'm now thinking is more related to the ground cooling (similar to the effect on DHW energy usage as the water temp cools). I'm both directly monitoring boiler gas usage and getting heating degree data from a temperature sensor in my backyard. I have a partially finished basement on its own small hydronic zone that I keep at ~62F, but I was heating the rest of the house for several weeks before I finally turned on the thermostat in the basement. The basement average temperature was probably dropping around 0.5F per week as the ground cooled, sapping increasing amounts of heat from the rest of the house, and now it runs for a few minutes every hour or so pretty much continuously to keep it at 62F. I would bet the ground (along with the incoming water) cooling down are the main culprits.

Nom_Deplume

Thanks for the comments!

Neither shoulder season short-cycling, nor the possibility that non-space-heating usage is higher in winter, explains why efficiency should be better in fall but worse in spring. November is about the same as mid-March-to-mid-April in our area in terms of outside temperature, but the efficiency is not the same.

Now the ground temperature IS definitely lower in March-April than in November, so that could certainly be an explanation. We do have a basement, and the foundation is not insulated.

If that's it, there's nothing that can realistically be done. Lot of things have been attached to or installed up against the foundation walls, and moving them would be a huge project in this >100-year-old brick house.

hot_rod

Hot goes to cold, so if the temperature on the outside if the foundation are dropping the load is increasing

In addition the wider that temperature difference the higher the transfer rate

Robert O'Brien

All degree days are not created equal. How much time was spent at the high and low of the day isn't calculated. Yesterday in NYC area is a perfect example, low of 14F shortly after midnight rising steadily to 40's 24 hours later. On paper it's 39 degree days but is it really?

fentonc

I was under the impression heating degree days are calculated using average temperature, which does account for how much time was spent at any given temperature (in my home monitoring setup I sample the temperature once an hour and use that to calculate the average for any given day), although certainly things like like wind, cloud cover, etc. can have an effect above and beyond just the average temperature. Anyway, I would put my money on cooling ground temperature.

Nom_Deplume

There is no standard definition of degree-days, but they are customarily calculated by averaging the max and min temps and subtracting from 65. In the old days, max and min data were more easily available than minute-by-minute readings.

As a result, it is not particularly accurate for any given day--it's not supposed to be. However, over longer periods such as a month, the variations go both ways and are smoothed out (averaged).

MikeGordon

Hello Nom;

In my experience, what you are seeing with the increased fuel usage on a degree day basis is very typical.

Looking at it, causes to consider are decreased solar gain from fall to winter, and then decreasing ground temperature as a result until summer. There is a big lag between increased solar activity in spring, and when the ground temperature heats up. Also, average surface wind speed and direction can make a real difference in heat loss and therefore fuel consumption for heating.

Air temperature is the major but not the only factor in determining the heat loss of a structure. Depending on the design and orientation of the structure, solar gain, wind, internal gains and ground temperature can have a big effect.

Hope this helps

The Steam Whisperer

One other thing to consider is that heat loss does not have a linear relationship with outdoor air temp. Air leakage increases at a steadily increasing rate as the air temperature drops. For the differences between fall and spring in a high mass home, the ground temperature is probably the main cause of differences. With high mass walls, the ground temperature probably has a greater effect than in frame homes, since that mass also pulls heat up out of the ground into the exposed walls, reducing heating loads. I have never seen anything in the ASHRAE calcs to model this impact, but it probably is a contributing factor for why old brick homes often have much lower heating fuel usage than expected when compared to newer insulated frame homes.

fentonc

I plotted BTU/HDD-Hour for the first ~2 months (mid November to mid January) of the heating season, and I think you can pretty clearly see the effect of the ground cooling down and then stabilizing, with BTU/HDD-Hour starting in the 300-400 range and then settling into the 500-600 range.

LS123

do you have a thermostat that tell you how many hours a day it runs? I calculate my heating oil consumption based on nozzle size, PSI for oil use by the burner. Don't have much knowledge with gas heating....

LS123

I mean
Heating oil use = number of hours a day burner runs (from thermostat data) + nozzle size ( based on PSI oil settings for burner) ... I find that to be almost on pint with the fuel consumption.

fentonc

I built my own monitoring system - it estimates fuel usage based on the “burn” status LED of the boiler and the nominal input power (140k btu/hr), although I think my estimates are probably too high by a few percent based on recent utility bills. The boiler is only firing for a third or so of the time any of the thermostats are calling for heat, since it’s so oversized.

LS123

@fentonc ... I am impressed! I was in IT over 30 years and wondering how you have built a system that calculate fuel use specifically for the burner? did you use assembly language and other programing codes (software) to communicate with hardware?
By the way, you may already done this.... I found out that insulating the basement and steam pipes in the basement have made a significant difference how fast the rads get heated.... if you have not insulated the main pipes in the basement that might be a good way to save some fuel... there was a thread I had in the past and many members provided me with how to calculate heat loss and or savings by insulating main pipes in the basement... will try to find that thread and post it some other time.

LS123

@fentonc , I misunderstood when you said system.... I was thinking about hardware and s/w.
Correction to my oil use formula

Heating oil use = number of hours a day burner runs (from thermostat data) * nozzle size ( based on PSI oil settings for burner) ... I find that to be almost on pint with the fuel consumption.

fentonc

I built something using a little microcontroller that is mounted next to my boiler, and it has 6 temperature probes to monitor the supply/return pipes for all 3 zones, as well as 2 light-sensors to monitor the state of the 'circulate' and 'burn' LEDs on the boiler. It also has a little screen that displays the current temperature values and graphs the last 10 minutes or so of operation. That board communicates the state of the temperature sensors and LEDs back to another computer via bluetooth every 1 second. That computer is running a script that tracks boiler heating/burn cycles, what the current outdoor temperature is (via an outdoor weather station), and, based on the heat output vs temperature charts provided for the baseboards in my house, it estimates how many BTUs are being delivered to each zone during a heating cycle (as well as residual heat delivered once the circulator stops running). It saves a daily summary of the heating system performance vs heating-degree-days (gas used, burn cycles, heating cycles, estimated BTUs delivered to each zone and overall estimated efficiency). I got this up and running in mid-November, so I have nearly the entire heating season logged in excessive detail (and it did involve writing a lot of software and doing a lot of soldering).

LS123

@fentonc , This sounds AWESOME!! sound like a good system!. Undoubtedly I can assume your codes are accurate to do all the calculations. What did you use for script? Perl, python? I am also guessing you are using as DB or log files collect data from all the sensors, then do the calcs..
by the way as mentioned if your basement is and steam mains are well insulated you would loose less heat and more savings on fuel use.
I didn't know steam can be zoned, I think you mentioned your system is steam... how do you zone steam heat have have zones? I have single pipe system.... is your steam system also single pipe?
I would love to learn more about it...

fentonc

I just have a regular hydronic (hot water) baseboard system with one zone per floor (basement + 1st flr + 2nd flr).

LS123

ok thank you for clarifying about the zoning.

fentonc

The boiler-side is pretty 'dumb', and it uses an Arduino-compatible board, so it's written in C++. It just reads the sensors and broadcasts them via bluetooth every second. A python script on my computer listens for the broadcasts, tracks the daily data and then logs heating cycles and daily summaries to a log file, as well as spitting them out in a terminal for debugging purposes. It reports log messages like this:

2022-01-23 09:45:25.8 HEATING_CYCLE_COMPLETE BTUS Fuel 11897.4 Heat 9942.83 B 1665.31 1st 1995.62 2nd 0.0 Eff (est) 0.31 Total_cycle_time 671.03 Burn_time 305.93 Cycle_hist 203.96 101.97

So both the basement and 1st floor were calling for heat over an ~11 minute period (671 seconds), with a 204 second initial burn followed by a 102 second burn later, with an estimated efficiency of 31%. There are definitely some inaccuracies in how the estimated delivered heat is tracked, but during long heating cycles (like recovering from the night time setback), the estimated efficiency does top out around ~82% on a nominally 83% efficient non-condensing cast iron boiler, so it should be reasonable.

Here is a photo of it mounted on the wall near my boiler:

LS123

This is one of the best thread on heating help... I am more than impressed having a member in the forum, taking actions to create a hardware and software to determine fuel consumption... I will reach out to you via heatinghelp emails... I would love to learn more about the circuits, sensors etc... you utilized in your project to create a system....... although I am comfortable with C++, python, I can understand perl better.... My preference is ADA object oriented programing.... which is used by commercial air liners, ICBCs, subway systems, and some defensive and attack helicopters....Thank you for sharing the pics and info. Best regards

LS123

* I mean old ICBMs

Nom_Deplume

Our boiler (the one in the original post) is a modulating boiler with outdoor reset, so our run time isn't a measure of anything. Our curve is designed to make stat call for heat as much as possible.

I suppose one could put a secondary meter in the gas line at the boiler---I'm surprised at how cheaply one can buy one with an output signal from Amazon, but I don't know if it is reliable:

https://www.amazon.com/Pulse-Output-Gas-Meter/dp/B00998ULE4 .

Nom_Deplume

fentonc said:

I plotted BTU/HDD-Hour for the first ~2 months (mid November to mid January) of the heating season, and I think you can pretty clearly see the effect of the ground cooling down and then stabilizing, with BTU/HDD-Hour starting in the 300-400 range and then settling into the 500-600 range.

I would suggest smoothing that curve with 7-day averaging, but in any case, that's a huge change in energy consumption over time.

Our trend is much weaker, no significant trend in the Nov-Feb period, but lower BTU/HDD in October and higher in March-May.

LS123

depending where your home is at, how well the crib is insulated, wind factors etc plays a role in energy consumption....I can tell you for sure, that I can keep run less than 10 gallons of a day when it is single digits, or occasionally even minus degrees where I live at. prior to basement insulation etc... we used a lot of heating fuel.... I would thing at some point more than 15 gallons a day...as of yesterday to today burner runtime is about 6.5 hours with in 24 hours, that is thermostat set at 70 F, if i drop to 65F, it is much less fuel consumption....

fentonc

I didn't turn on the basement zone until day 32 or so, when the average temp down there finally fell below 60F, and I think we've had it at 62F (no setback) since then. The basement is not well air sealed, although it does seem to have fiberglass insulation on all but the utility area (roughly half of one wall). It's divided into a finished playroom / storage area, with a door separating it from half-finished workshop/utility/laundry area. The thermostat is in the playroom, and both of the rooms have 11 feet of ~800 BTU@180F/ft baseboard.

About once an hour or so, I see the basement zone call for heat, with a really typical pattern like this (picking an example from my log):

1. Thermostat calls for heat
2. Boiler aquastat temp has fallen back to 70F, and burns for 245 seconds before it hits 160F aquastat high limit.
3. Circulator keeps pumping for another 140 seconds before thermostat stops calling for heat, total cycle has taken 380 seconds (never fell below the aquastat low limit of 140F).
4. Boiler slowly bleeds heat into surrounding uninsulated concrete block walls until next call for heat.

So I think the basement heating is just very inefficient, and it significantly increases the fuel usage vs heating the rest of the house (the upper 2 floors use setbacks during the day and night, and also get a lot of solar gain during the day time, whereas the basement really consistently calls for heat once every 50-70 minutes or so throughout the day).

LS123

if basement is not used frequently, I would keep the stats to be no more than 50F or (60F) to use less heat. Assuming that utility area is insulated, and all water lines are insulated and will not freeze your water pipes by dropping the temp in the basement. Insulation makes a huge difference in my case. Try it for a day or two and see if that makes a noticeable difference in fuel consumption. not sure if the heaters in the basement are correct size base on sq footage it has to heat.

Nom_Deplume

Yes, I would leave the basement at 50F or even less (just enough to keep pipes from freezing). However, you may find that this just increases the heat demand upstairs.

fentonc

I might experiment with some setbacks for the basement thermostat, as I have a 7-day programmable one down there, but half of it gets used as a kids playroom (mostly on the weekends) and the other half gets used as a workshop area (also mostly on the weekends), and below 60 it's pretty chilly feeling down there. I really wish there was an easy way to implement something like "Whenever the upper 2 floors are done calling for heat, open the basement zone and run the circulator until the water temp drops down to 70F", to dump the residual heat into the basement air rather than letting it bleed into the uninsulated walls around the boiler.

LS123

not sure where you are at, where I am it drops, to single digits and sometimes few degrees in to minus side. I dont know much about hydo heating but as far as I know, for hydo boiler should be min somewhere 140F and max 180F ( these settings were there are real cast iron rads) you can program the tstat in the basement to run on weekend at your preference, there is always manual override to increase as needed. I think dropping the boiler temp to run until 70F may end up using more fuel when heating is called for other sections (including basement) of the house.... just a thought and I am not a pro.
There are programmable tstats at very affordable prices at suppluhouse.com , other plumbing shops and home improvement places. These Tstats also give burner run time etc.
I think your original post was about how much fuel is being used. Try putting one in the basement. and calculate the run time by what your boiler use for fuels (based on its own settings)