GBA Logo horizontal Facebook LinkedIn Email Pinterest Twitter 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

Efficiency gains from upgrading a steam boiler

johns3km | Posted in General Questions on

With recent talk on steam boilers I was wondering what types of gains can be made from upgrading a 35+ year old Penco 192k/152k one pipe steam residential system for my home. The Boston area home has 450 EDR of attached radiation. I’ve gotten it down to 305 EDR after losing a couple radiators in a kitchen remodel and turning off the unused attic and basement rads. I could lose another 25 EDR as the two bathrooms have heated floors. If I downsized a couple of the large radiators after doing a heatload analysis I imagine I could get it close to 250EDR. With a standard 1.33 pickup, I’d be looking at an 80k boiler.

Would that in theory reduce my nat gas usage by 60%? Even more with decreased jacket losses/dampers etc?

I used 1120 therms 2016 calendar year/5300HDD. I’m trying to see how much I could save by downsizing and making the system as tight and efficient as possible. I have some minisplit heads as well but would like to retain steam system for deep winter comfort.


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. BrianPontolilo | | #1


    This is not my field of expertise at all, but there are some very knowledgeable GBA members when it comes mechanicals. I'm sure some of them will chime in soon.

  2. tommay | | #2

    Efficiency is in the user. Keeping your thermostat set at one temperature, rather than up and down, especially on a steam system, keeps your boiler warm and reduces run time. Air vent adjustment and operation is also critical to send steam to selected radiators first and to ensure low resistance allowing the radiators to heat up faster. If any vents are leaking or pissing, change them or remove, shake and blow through them, if the float inside is stuck it won't shut off properly once the radiator is full (a little WD40). If you downsize your boiler you may not have the volume of water necessary depending on how many radiators you have, how big they are and the amount of pipe run you have. Monitoring the water level so as to keep the minimum amount of water necessary to provide the heat you need will reduce boiler run time also. Make sure radiators are pitched back towards the valve, as well as pipes back to the boiler, to ensure water returns properly, if water builds up you will get banging. If you find yourself constantly adding water to the boiler, then water is leaving the system somehow, usually through the radiator vents.

    1. Expert Member
      Dana Dorsett | | #5

      >" Keeping your thermostat set at one temperature, rather than up and down, especially on a steam system, keeps your boiler warm and reduces run time."

      Warmer boiler = greater standby loss.

      Warmer house = greater heat loss from the house.

      Keeping it at a single temperature is more comfortable (and often lower maintenance with steam), but does not reduce the total burner-on time or fuel use. It increases fuel use.

      In a well insulated house with the boiler in an insulated basement the difference in fuel use is small, but that doesn't describe the house most steam boilers are located in. But even in the mythical well insulated house using setback strategies saves rather than increase fuel use.

    2. johns3km | | #16

      I don't use setbacks- just keep it at 71 unless I'm away for the weekend or longer. Too much high pressure on recovery.

      1. tommay | | #18

        Exactly, the time it takes for the steam to purge the colder, denser air from the radiators via the vents and for the heat to dissipate from the radiators to satisfy the t-stat will cause the boiler to run harder raising internal boiler pressure which will cause the pressure switch to turn off the boiler until the pressure drops. (be sure to clean out the pigtail pipe going to the pressure switch occasionally from any type of blockage.)

  3. rhl_ | | #3

    Has the house had a conservative manual J calculation done? This can give you a sense for how your existing convectors match the load on a room by room basis.

    Do the rads all feed to/from a basement? If you convert to a two pipe direct/reverse return system, or a manifold back near the boiler room, you can achieve better balancing, and more importantly has better zoning.

    Also ECM pumps are very energy efficient. The grundfos alpha 2 uses less than 30 watts, I have one for some floor heat zones which runs on a constant delta pressure model, it uses 11 watts

  4. Expert Member
    Dana Dorsett | | #4

    Reducing the radiation by 60% doesn't reduce the fuel use by 60%. But right-sizing the radiation & boiler to the load reduces standby losses, bringing the as-used efficiency close to the raw combustion efficiency, which for new steam boilers can be as high as 82%.

    If the boiler is in a basement with air sealed R15 continuous insulated walls the standby losses of the oversized idling boiler aren't really lost, so even fairly substantial oversizing doesn't affect the fuel use all that much. When the boiler is in an un-insulated basement and the boiler room is always the warmest room in the house those standby and distribution losses are substantially lost. Spending the money on marginally more efficient mechanicals is almost always a lousier investment than upgrading the building envelope, until/unless there has already been substantial insulating & air sealing work on the house.

    Start with the room by room Manual-J on the house in it's "after planned upgrades" version before diving into the mechanicals.

    1. rhl_ | | #7

      “Reducing the radiation by 60% doesn't reduce the fuel use by 60%.”

      Can you expand on what you mean by that? Does it reduce fuel usage at all? Aside from going up the smoke stack, where is the heat energy going ?

      1. Jon_R | | #8

        Perhaps clearer to say "reducing the radiator capacity...".

        An overized boiler cycles off and the energy stays in the gas lines. I think Kevin is chasing the savings from less cycling with a smaller boiler - which isn't much.

        Kevin, check that your thermostat has as much dead-band/hysteresis as you find comfortable (a low cost optimization). Have you had an energy audit and air sealing done?

  5. Jon_R | | #6

    How well balanced is the system? If you have rooms that are overheated, there is some potential savings available by reducing it.

    1. johns3km | | #15

      Rooms are all balanced and system runs nice and quiet.

  6. tommay | | #9

    Dana, realize steam is not an instant heat. It takes time to bring the water up to boiling. If you use setback strategies anything more than a few degrees, the boiler gets cold and may take up to 20 minutes or more, especially with an older, larger boiler, just to get the water to temp. If you fluctuate say 10 degrees the boiler will run longer than necessary because of the time it takes for the heat to leave the radiators and satisfy the thermostat. Then, since they are radiators, they will continue to radiate heat making the house hotter and will contribute to greater heat loss as you mention. Remember fancy programmable t-stats were not available when steam dominated, so adding modern tech to old tech may not work so well. If you leave the t-stat set at one temperature or 1 or 2 deg differential, the boiler will stay warm and will bring the temperature of the water up faster as well as keeping the pipe runs warm. As far as standby losses, if the boiler is warm it still contributes heat within the house and piping. You will find that the boiler will run for a shorter time and may only need to heat up half the radiator in order to maintain the homes temperature.

    1. Expert Member
      Dana Dorsett | | #10

      I'm not a newbie to steam. I've lived with and maintained a number of steam systems.

      The laws of physics are self enforcing.

      >"If you use setback strategies anything more than a few degrees, the boiler gets cold and may take up to 20 minutes or more, especially with an older, larger boiler, just to get the water to temp. If you fluctuate say 10 degrees the boiler will run longer than necessary because of the time it takes for the heat to leave the radiators and satisfy the thermostat. Then, since they are radiators, they will continue to radiate heat making the house hotter and will contribute to greater heat loss as you mention."


      Running longer to bring the thermal mass of the water & plumbing up to temp does not make the boiler " longer than necessary because of the time it takes for the heat to leave the radiators and satisfy the thermostat...".

      The boiler runs just as long as necessary, and doesn't cause a temperature overshoot. The cooler boiler and pipes during the setback are losing less heat.

      The longer burn time during a recovery from setback is in fact more efficient than shorter burns just to maintain temperature- every ignition cycle takes a loss.

      > "As far as standby losses, if the boiler is warm it still contributes heat within the house and piping. "

      Unless the boiler & plumbing are inside of insulated fully insulated space a good fraction of the standby & distribution losses are truly lost.

      1. tommay | | #11

        Well, let's compare what you say to other types of heating equipment. Hot water boilers as well as things like ovens work off a differential temperature so as not to completely cool during run times or use. Otherwise your oven would turn off completely after reaching temperature rather than turning on again after a small temperature loss in order to maintain the correct temp, same as with a hot water boiler, it cycles between two set point temperatures rather than completely shutting down and cooling off completely, especially with larger systems. Why is this ? Because it would take more energy to raise the temperature from a cold start. How about the heat in your car? Do you stop the engine once the interior reaches temp and let it cool down completely, then have to restart your engine and let it run, wasting fuel just to get back up to temp or is it better to leave it idling while in use?

        1. Jon_R | | #13

          My oven cycles within a narrow temperature band because otherwise my food wouldn't cook fast/properly. Nothing to do with taking less energy. Similar for my car - I don't let it cool back down completely because that would be uncomfortable (and interfere with me getting anywhere).

          1. tommay | | #17

            Jon, exactly. Same principle.

          2. Expert Member
            Dana Dorsett | | #21

            But you don't leave the oven on all night so you can warm up the bread in the morning, eh? I don't leave my car running with the heat on all day & night either. :-)

            Steam boilers don't overshoot the setpoint on an overnight setback. And it doesn't "... take more energy to raise the temperature from a cold start..." than it would have expended in shorter burn cycles maintaining temperature. Recovery from setback takes time, but that doesn't use more energy.

            Yes, it's more comfortable to "set and forget" the thermostat on a steam system, but it doesn't save energy- quite the opposite.

  7. user-723121 | | #12

    I would consider improving the building envelope before making changes to the heating system. If your house is 2,500 SF your Btu/sf/hdd is 8.45. That is high and there should room for improvement in reducing the heating load. Once the efficiency of the house is improved then any changes to the heating system will be properly considered and sized.

  8. johns3km | | #14

    Thanks for everyones comments. As far as the steam system is concerned, I have it dialed in pretty well. Good main venting and all the radiators get heat at the same time. It's quiet and requires minimal maintenance each year. Guess I'm surprised that replacing a 35 year old boiler would yield little cost benefits at even half the size. I've done a rough Manual J and came up with 43K including uninsulated basement, but not accounting for infiltration. Take that # with a huge grain of salt.

    As far as the envelope here are some rough numbers:

    1940s home, Boston
    1000sf uninsulated walkout basement: ~3/4 above grade, 4 leaky hopper doors. Large block/stone foundation walls probably 18" thick
    1000sf 1st floor: combination of MassSave blown in and open cell in kitchen/bath
    800sf 2nd floor: No in wall insulation as we had maxed out the MassSave credit- on the list to do. Just plaster on lath
    230sf attic playroom on 3rd floor: Attic has R-24 open cell in rafters

    Windows are all upgraded Harvey energy star replacement.

    I mean is $1700-1900 per year in natural gas usage (includes hot water) in line with ~2000sf conditioned space? Is the best I can do to get my bills down is blown in insulation on second floor, and sprayfoam R-15 all my basement walls and sills? I was hoping downsizing the boiler would help but it seems like there isn't as big of gains as I originally thought.


    1. tommay | | #20

      Yes, especially with the price of steam boilers today. The old supply and demand thing. Plus a lot of newbies in the field have no idea or much experience with steam and the old timers are dying out fast. I recently serviced a newer steam boiler after a newbie looked at. He told the homeowner they possibly needed a new boiler due to water on the floor. Ended up being a clogged pressure tube so the boiler wasn't shutting off and the relief valve was opening.

  9. Expert Member
    Dana Dorsett | | #19

    That's headed toward 2x what we're paying Eversource for gas every year in a 1920s vintage 2400' house + 1600' of basement for a 3 person household (some months 5+) after dense packing cellulose in the walls, judicious use of open cell foam for air sealing and insulating behind the kneewalls in the half-story upstairs, and adding 3" of reclaimed polyiso to the basement walls.

    Just because MassSave isn't going to subsidize it doesn't mean further upgrades aren't worthwhile. If I had to guess (which I sort of have to, with insufficient information) insulating the walls would be the first order of business, but with 3/4 of the foundation walls above grade it could easily be the biggest heat loss. It wouldn't surprise me if the basement losses accounted for more than 25% of the fuel use. (In my case it was ~15%, but my average exposure is less than 2 feet, and the above grade walls had been insulated the year prior, so mine was a much lower number overall than yours would be. A poured concrete foundation isn't as air leaky as quarried stone either.)

    At $2.50-3 per square foot, 2" of HFO blown closed cell foam isn't cheap, but it would bring it to code-min (on a U-factor basis, not R-value.) With the uneven surface of the quarried stone you might consider building a steel-stud studwall with 1.5-2" steel furring/studs set 2" from he wall, mounting the fire-protective gypsum board on the steel furring. This is actually quicker & easier than wood studs, since there is an inch or so of vertical slop in steel top/bottom plate channels, so the studs can all be rough cut to length and still accommodated some unevenness in height.

    Assuming something like a 130' of perimeter foundation wall and 9' from slab to subfloor you're looking at about 1000 square feet of basement wall after subtracting off the doors & windows. Order of magnitude budget $3K for the foam, maybe another $2K for the studs & wallboard (installed), $5K for the project which is about the same amount of money it takes to swap in a new steam boiler, but with a much bigger "payback." Even if it "only" saves you $500/year, that's an IRR of 10% even without looking at the fact that the gas bill is paid in after-taxes dollars, etc.

  10. johns3km | | #22

    Apologies Dana- never got the email on your response.

    I will take advantage of the MassSave blown in on second floor- I was referring to about 5 years ago where the incentives were capped each year, so we did the first floor only and planned to finish it the following year.

    It's roughly ~160' linear feet of block stone foundation, front of house is pretty much at grade and slopes to backyard - one corner is 4.5' above grade with the windows, the other corner is 5-6' above grade. Basement ceilings are 7'. Hopefully pictures help a little bit.

    I like your thinking with steel studs. Are fire-resistant boards necessary all around? I thought that was boiler room only and closed cell foam had some fire protection in it. What did you mean it's code-min for ufactor and not R value? Should I try and achieve more?

    Slab and basement temps are currently 60 degrees. I would be happy saving $500/yr for sure for almost 3,000sf of conditioned space.


    1. Expert Member
      Dana Dorsett | | #28

      Either an intumescent paint or a timed thermal barrier such has half-inch wall board would be necessary. Not all inspectors are keen on the intumescent paint approach- call your building department.

      Using the steel furring allows for the foam to be continuous, not thermally bridged by the steel studs, taking up less room area. I first encountered that approach a bit over year ago when a co-worker finally broke down and insulated his basement (after years of prompting from me.) The idea came from his general contractor.

      1. johns3km | | #29

        I did a quick heat loss on my basement and I came up with ~11,000BTU/hr @ 9F, not including infiltration losses; ~7,000 cubic feet with 4 leaky hopper windows, a terrible door, and 6 new Harvey replacement windows...another +8500 BTUs sound realistic just for the basement? Seems high.

        I backed in to my total heatloss for my home at 45,000 BTU @ 9F, including basement losses with 80% boiler (7.6 therms/day @ 36 HDD, base 65).

        A couple exterior pics,

        Any idea on best case scenario for basement losses with R-15 on the walls and sills?


        1. Expert Member
          Dana Dorsett | | #30

          R15 on the walls is going to cut the wall losses by about an order of magnitude. The walls are about R1.5-R2, counting air films. Add R15 to that and it'll be R16.5-R17, about 10x more resistant to heat transfer. So if the wall losses alone were 10,000 BTU/hr before, insulating, the wall losses after insulating will be about 1000 BTU/hr, give or take.

          It won't affect your window or slab losses though.

          With the lower heat loss the indoor temperature will rise by about 5F or more, so the temperature difference will be a bit higher, taking a tiny amount of those gains back, since the loss to the slab will nudge up a bit due to the temperature change. But these are secondary affects of the change. If you're updating the windows the new windows will likely be sufficiently high performance to overcome the affect of the higher indoor temperature with a large amount of margin to spare.

          1. johns3km | | #31

            Thanks Dana, really appreciate the detailed info. Looks like I have some work to do before next winter.

            Last question- once the envelope is in good shape would you consider boiler replacement or are there really no gains to be had there from a 35 year old boiler to today’s boilers? Steam to radiator can definitely be improved as the near boiler piping could use a better header, improving pipe insulation etc. Just ride out the boiler vs the $8k boiler upgrades?

            Thanks again.

  11. Deleted | | #23


  12. Deleted | | #24


  13. Deleted | | #25


  14. Deleted | | #26


  15. johns3km | | #27
  16. Expert Member
    Dana Dorsett | | #32

    >"Last question- once the envelope is in good shape would you consider boiler replacement or are there really no gains to be had there from a 35 year old boiler to today’s boilers? "

    In most cases there won't be a significant enough improvement in efficiency to ever "pay off" the $8K for a new boiler. It's worth keeping up with the system maintenance, keeping vents operating properly, blowing down the boiler sludge periodically, etc.

    If your radiators are all painted silver, painting them any non-metallic color will improve output, shortening the amount of time it takes to satisfy the thermostats by 15-20%, reducing some of the distribution losses to the basement, etc. If any of the radiators are COVERED, that is reducing the heat transfer efficiency too.

    During the shoulder seasons the mini-splits may have a lower operating cost than keeping the steam boiler hot to serve a tiny load. The raw combustion efficiency of 35 year old gas fired steam boiler might still be 75-80% (a brand new one is required to be 80%, some are in the 83-84% range), but the distribution and standby losses add up to a large fraction of the fuel use when operating at a low duty cycle, whereas during the deeper part of the heating season that fraction is much less, and mini-split efficiency is quite a bit lower when it's in the teens & 20s rather than the 30s & 40s.

    FWIW: A handful of years ago a friend of mine was forced to replace her ~1920 vintage gas fired steam boiler after a catastrophic failure (cracked cast iron, copious leakage), and replaced it with an 82% AFUE Burnham sized appropriately to the radiation. Subsequent analysis showed about a 15% reduction in fuel use per heating degree-day, not more. Given the antiquity of the failed beast we were hoping for more. The "problem" with steam boilers is that the don't quit, lose efficiency fairly slowly over time, and are usually so oversized that they'll still heat the place even at 50% combustion efficiency.

    Eventually you may want to DIS-place rather than RE-place the Pennco using much higher efficiency or non-fossil energy sources, but it's hard to make the financial case for that when gas prices are low, and there is still comparatively low hanging fruit on the building efficiency end.

  17. johns3km | | #33

    I learned from this past month's electric bill that my Fujitsu multisplits are pretty poor as a heating source and not $$ competitive to the boiler at 28HDD/day average month...shoulder season they are great. I let them run all month in conjunction with the boiler as a setpoint backup. Even with a 5 day vacation with the house at 60F, I used 935kwh for heating costs (probably less with Christmas lights, bathroom heated floor etc). Just tallied up my ecobee runtimes for the same 33 days and I still burned 5.2 therms/day heating only. $275 for natural gas and ~$205 for electric is crazy for a mild December. The last time I had a $400+ gas bill was with a 41HDD/day average month in Feb '15.

    This month I turned off the second floor splits/compressor and will just run the first floor unit. If it consistently gets below 30F I'll probably turn it off for good until the Spring. Pains me as I was trying to free myself from the National Grid empire.


    1. Expert Member
      Dana Dorsett | | #34

      You realize that National Grid is a UK based corporation that operates both gas and electric utilities, all over the world, including the US right? (They happen to be my electric utility...) National Grid is one of the largest electric utilities in MA, and one of the largest gas utilities in that state.

      Utilities in the US are regulated monopolies. Short of going off-grid, only if your electric utility is a tiny locally owned utility can you escape some of the empire-ness of the energy grids. (In Littelton/Boxborough MA the residential retail rate from the tiny local utility is about half the state average price for electricity.)

      Both gas and electricity rates vary widely, and the crossover point on operational cost will vary.

      By using both the boiler and mini-splits the boiler ran at a lower duty cycle, but had about the same standby losses. It's so oversized for the actual load that any further reduction in duty cycle reduces it's efficiency as a system. Once the basement is insulated much of that standby loss will still be retained.

      But in general it's tough for an HSPF 9-ish multi-split running on 20-22 cent electricity to break even with buck a therm gas (or even buck-fifty.) It becomes more competitive with HSPF 13-14-ish equipment. Even at 70% net efficiency in a steam boiler $1/therm gas is $14/MMBTU, $21/MMBTU at $1.50.

      At an as-used HSPF of 9 BTU/watt-hour at 20 cents/kwh you're looking at $22/MMBTU, but at an HSPF of 13 it's about $15/MMBTU.

      Move it to Littleton where they're paying about 12 cents/kwh an HSPF 13 heat pump comes in around $9/MMBTU and even the HSPF 9 multi-split becomes competitive with buck a therm gas.

      BTW: Was the $205 the total electric bill, or the additional amount above what you normally use in December?

      1. johns3km | | #35

        December ‘16 pre-minisplits I was at 630 kWh $128. I average 530 kwh in the non summer months presplits.

        Total bill was $360 this month @ $0.23 rate. Ouch.

        I don’t know too much about national grid other than their contractors striking, a pretty lousy website, and they raise my rates close to double digits each year.

        1. Expert Member
          Dana Dorsett | | #36

          Massachusetts is a decoupled utility rate state, with a competitive market for the energy portion of the electric bills via retail brokers.

          You can beat 23 cents even with 100% green electricity going through electricity brokers if you lock in the energy portion cost before the wintertime bump up. I'm currently paying about 11 cents + delivery charges for 100% wind on a 2 year fixed rate contract. The standard Nat'l Grid standard mix in my area is 13.7 cents on the energy portion right now- it bumps down in the spring, but I expect even then I'll be paying less than the standard mix.

          The state has a comparative shopping website here:

          If I were shopping right now the cheapest 100% renewable option would be 11.852 cents on a 21 month contract, but that too will readjust downward in spring when the standard mix rate is reduced. My contract is up in October 2020, so I'm not paying close attention right now.

          Watch out for the "automatic renewal" (which is usually at some exorbitant rate from brokers), or cancellation fees, and pay attention to when the contract term ends to avoid getting gouged. One of my prior contracts with an electricity broker was so egregious in the details the State Attorney General forced them to rebate all ratepayers some fraction, even those who managed to avoid the aut0matic renewal & cancellation fee traps.

          The green energy purchased isn't necessarily generated in MA or even New England (mine's from a wind farm in Texas, purchased from a broker in Florida.)

          There may be a community solar option cheaper than the 23 cents you're paying right now. Some of those operators a about as snake-oily as some of the electricity brokers too- they don't even post the structure of the deals available on their websites, so analyze carefully. Similarly the solar lease and even direct sales often come with bizarre terms to be weeded out. (What, they don't have enough car sales positions open? :-) ) But most well considered solar deals in MA are substantially cheaper than 23 cents/kwh when all incentives are factored in.

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |