GBA Logo horizontal Facebook LinkedIn Email Pinterest Twitter X Instagram YouTube Icon Navigation Search Icon Main Search Icon Video Play Icon Plus Icon Minus Icon Picture icon Hamburger Icon Close Icon Sorted

Community and Q&A

Calculating Savings

bdrfab | Posted in Energy Efficiency and Durability on

Is there a way to figure out an exact cost savings made by upgrading equiptment/insulation etc? Anecdotally, my parents upgraded to a 95% furnace last year, and are seeing the savings, but it would have to be compared to the HDD for each year, correct? And even at that its something of a crap shoot since they have a gas water heater as well? One would need to moniter volume of gas at the furnace, correct? is there any sensor for this?

GBA Prime

Join the leading community of building science experts

Become a GBA Prime member and get instant access to the latest developments in green building, research, and reports from the field.


  1. KHWillets | | #1

    My utility offers a usage history feature where we can plot gas, etc., usage vs. degree days. It's a very useful option. We saw a sizeable drop in gas consumption after insulating a couple of years ago, and we've continued to improve with various smaller projects.

    I'm not aware of how to get HDD per billing period, but if that data is available somewhere you could do your own comparison.

  2. user-943732 | | #2

    "Is there a way to figure out an exact cost savings made by upgrading equipment/insulation etc?"

    simple answer - no.

    There is no way to figure out the exact savings from a retrofit because we don't have a perfect parallel version of the world handy where everything else was the same except for the retrofit. You may get a fairly good idea of the savings by analyzing energy use data using a weather normalization approach (e.g., PRISM, degree day regression, etc.) but you can't control for everything in any given home -- differences in occupancy, behavior, non-temperature weather (wind,solar gain), and other changes in the building and equipment can all affect the observed savings. Directly measuring the heating system and measuring with a finer time resolution can help some but not enough to provide a firm answer

    By analyzing the energy use of large groups of homes you can learn a lot more about retrofit impacts, but the findings in any one home are always be suspect.

  3. bdrfab | | #3

    Thanks Michael, you raised some other variables I hadn't thought of.
    will calculate degree days going back 3 years.

  4. jklingel | | #4

    I'd argue that it is an emphatic "yes". Not down to the gnat's butt, as you will never get "exact" for anything, anywhere, anyhow. But if you can assume that whoever will live there for the next few years will behave like they did in the prior few years, and the weather will generally be what it has been, then all the variables are gone. Is that a fair assumption? Well, what other options do you have? Does it matter if you are off 5%? That's your call. Just do a heat analysis w/ insulation A and insulation B, or whatever A and B you want to compare, and you've got a pretty good handle on it. I don't see any issues with that at all.

  5. user-943732 | | #5


    You are more than welcome to give it a try -- but it becomes a sort of self-fulfilling prophesy. If the savings are far off from your expectations, then you can think up some excuse for how things are different. If they are consistent with what you think, then you feel sure that you nailed it.

    In general (i.e., most of the time) the weather will not be the same and the occupancy patterns will not be the same. The differences won't usually matter that much if you are expecting energy savings of 40% or so, but they will matter quite a bit if you are expecting savings of 20% or less. Energy use tends to change from year to year with a standard deviation of perhaps 10%. Since you are only looking at one house, it limits how strong a conclusion you can make regardless of the results.

    I base my beliefs on having analyzed the energy use patterns of millions of homes and having looked at the savings from retrofits in many tens of thousands in a wide range of climates and undergoing a wide range of interventions. If you don't see any issues at all then you aren't looking very closely.

  6. davidmeiland | | #6

    So, should someone doing an audit for an individual homeowner make projections regarding energy use before and after improvements? Give them payback periods for improvements?

  7. user-659915 | | #7

    "So, should someone doing an audit for an individual homeowner make projections regarding energy use before and after improvements? Give them payback periods for improvements?"
    Consulting professionals of many stripes do this all the time, as do contractors and installers wanting to make a sale. Whether they SHOULD do is more a question of ethics than technology, given that nearly all have motivation to err on the optimistic side, sometimes wildly. I have known energy consultants go so far as to guarantee performance but (given the multiple variables Michael mentions) this did not prove a sustainable business model.

  8. bdrfab | | #8

    Slightly off tangent, but, is it fair to say that in a 95% furnace, 95% of each gas unit is used for heat?

  9. jklingel | | #9

    Michael: Ok, I see your point, but I think we're looking at this from different angles. Sure, just because you double your insulation doesn't mean that your fuel bill will be cut in half the next year. Certainly, Nature is going to vary and confound/camouflage your results ; that's Nature. But, all that aside, you've still got your savings tucked into that variability. It's still there, whether it shows its face or not. That was my point: The savings will exist, but not necessarily be recognizable year to year. Heat loss is heat loss. Fuel usage, and subsequent variability, is another issue.

    Aaron: There are a couple of rating systems used for boilers, and I don't study them enough to recall their differences. If I recall, some refer to what % of your fuel goes into real heat in your house, and some refer to what % of the fuel is burned (some may go up the chimney). You may have to Google that one and learn the various systems.

  10. user-943732 | | #10


    Yes -- I see no problem with providing an estimate of energy savings from retrofits -- as long as you use a reasonable method for making that estimate (a surprisingly big "if") . That's not the same thing as telling people what their bills will be next year. The savings are how much less energy they use compared to what they would have used if they didn't do the retrofit.

    For example, if a household's heating energy use was $1000 last year and you install retrofits to reduce that by about $200, that doesn't mean their energy use will be $800 next year. It could be that they just had a baby and their heating use would have gone up to $1100 but instead it's now $900 -- they actually saved $200 but their bills only went down by $100. In any given home we don't really know how the energy use would have changed without the retrofit, but across large groups of homes we can generally confirm the impacts of retrofits using evaluation methods.

    John - I'm glad you see my point but now I''m still a little confused. The question we are answering is whether one can accurately see the savings from a retrofit by measuring the energy use in a given home. You said it wouldn't be a problem. Now it seems you were answering a different question -- are there actual savings even though they may be masked. I agree with you that "heat loss is heat loss" but I'm not so sure we really have a very good handle on what those potentially camouflaged savings really are for many common retrofits. Empirical data seem to indicate that commonly used modeling tools significantly over-estimate the energy savings from common retrofits of poorly insulated and leaky homes. So it doesn't appear to be so straightforward.

    Aaron -- a 95% efficiency rating for a furnace can be interpreted as meaning that 95% of the energy content of the fuel ends up going into the distribution system. The distribution system losses may lower the net amount that actually heats the living space. If you want to estimate the % heating savings from replacing a furnace, you can get a pretty decent estimate from the simple calculation of 1 - AFUE_old/AFUE_new. In retrofits, the AFUE_old can be the slightly tricky number to get.

  11. jklingel | | #11

    Michael: I think we're on the same page, but an interpretation of his words is in the way. The OP asked "Is there a way to FIGURE out an exact cost savings..." My position is, yes, you can FIGURE it, reasonably well, but due to a variety of variables you may not be able to see it, specifically. Your angle is, as you said here: "The question we are answering is whether one can accurately SEE..." I did not interpret the OP as asking SEE, but rather FIGURE. You apparently interpreted that as SEE. So I think we are saying the same thing. IMO, closed case.

  12. bdrfab | | #12

    I appreciate your responses. I wish there was a better way, too many people can't justify spending a little more up front to save on the back end. Thanks for the responses guys.

  13. v5PHTcyPPU | | #13


    I'm joining this discussion late, but I wanted to encourage you not to give up on finding "a better way" to estimate energy savings from your retrofits. What Michael Blasnik says about it being hard is right, but that doesn't mean it's impossible. No, we don't have a parallel control world, but we do have methods of controlling variation statistically. If we didn't, epidemiology, astronomy, economics and other non-experimental fields would be a waste of time.

    So how could you measure the savings from your retrofit? Here are some things to think about. You could adjust for temperature differences before and after the retrofit by calculating energy usage per degree-day, as already mentioned. But the same is true for wind speed, wind direction, and solar gain. Information is available for them from the National Weather Service and elsewhere. But first you have to learn a bit about modeling energy use when you have multiple sources of variation (temperature, wind speed, wind direction, sunshine, etc.)

    You also have to take household behavior into account. Hopefully, the homeowners haven't changed their thermostat settings, but if they have, you would need to adjust for that by finding degree-days based not on an assumed constant thermostat setting (like 65) but on the actual setting, on a day-by-day or even an hour-by-hour basis. A change in the number of occupants is trickier, but there are techniques to address this also.

    The alternative to controlling things statistically is to measure a large sample of buildings, comparing them before and after the changes, and hope that all the unwanted changes between buildings over time cancel each other out. There's nothing wrong with this approach, but it's not the only one.

    Energy Metrics

  14. user-943732 | | #14

    reply to David-

    I will tell you how epidemiology and other fields analyze their data -- they use large groups and perform careful statistical analysis -- they don't work with a sample size of 1. Of course, when working with observational data, they often still get things very wrong. What they don't do is look at one person who lived near a garbage dump and got cancer and say -- aha -- the dump caused their cancer. That would be ridiculous. But you are claiming nearly the same thing -- basing conclusions on an N of 1.

    Yes, we can try to adjust for weather -- outdoor temperatures are the most amenable to that (I've done that for millions of homes).. But I guess you don't know that the best weather normalization approaches -- employing variable-base degree day regression models -- still don't correct for weather properly. For heating, they over-adjust for outdoor temperature with errors of a few percent due to seasonality of baseload. For cooling it is much worse because much of the cooling load is driven by solar gain and not dT and so the dT adjustment overshoots by much more.

    You say you can adjust for wind speed and solar -- how would actually do that? Yes, I do know a thing or two about modeling buildings and statistical analysis -- it's what I've been doing for the past 20+ years)? We don't have very accurate models for how each of these things affect energy use and we don't have good sources for the inputs for any given home. Each adjustment will have considerable uncertainty and there are many factors that can vary that you may never find out about.

    Sure, in many cases the changes in these other factors aren't very large and you can still clearly see large savings. But for more modest/typical retrofit impacts, the potential errors for any one home can be large compared to the savings. People who think they can reliably make these adjustments usually have little experience actually doing it or are living in a world of self-fulfilling prophecies.

  15. tgray | | #15

    For what its worth we replaced a circa 1990 80% furnace this fall w/ a 96% unit. I tracked my MCF use for past 3 winter months and saw no savings over same 3 2009-2010 heating months. The degree days as posted on the gas bill were within 1 degree. Our habits and setpoints were unchanged. I also insulated a cold garage wall and 2 basement walls fall 2010 so we are dissapointed. It makes it harder for me to advocate for envelope improvements and equipment upgrades when the #s do not support the cause. I am hopeful that next winter will show better results, maybe just a fluke or due to all the variables discussed above?

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |