Original radiators, want to replace boiler and add radiators

[Dan Westerfield]

I have a 1914 home with the original coal fired boiler that was converted to gas at some point. There are 4 floors in total. Basement, 1st (ground), 2nd, and a finished attic. My current (original) system supplies the 1st and second floor radiators with hot water through a gravity system. I have 11 radiators in total now, and the house is about 4500 sqft. The exterior walls are solid brick, 2 brick thick with plaster on the inside. Interior walls are plaster and lathe. R-30 insulation has been installed in the attic. I do have storm windows on all windows, and doors except for 2 set of French doors that lead out to my side porch. The heat in the basement comes from the wildly inefficient boiler and pipes. The attic heat currently comes from a stand alone ventless gas heater. On the first floor, there is a pantry (over a small crawl space, and back room (on a slab) that we converted to our laundry room that were unheated. I added a small wall mount ventless gas heater to the newly converted laundry room.

What I would like to do is replace my old boiler with a properly sized, ultra high efficiency boiler, with a recirc pump, and add radiators to my attic, pantry, laundry room, and basement. I would like to make the first floor and basement one Zone, and the second and attic a second zone. It would be an added bonus for my boiler to also provide DHW for my home.

I have read about installing 2 smaller boilers, that way the majority of the time I’m only running one small one, and not firing up the big one everytime, and only kicking on both when it gets really cold. Pros - Cons? Or should I go with 2 small boilers and put 2 floors on each boiler instead of doing zones on a bigger boiler?

Any help would be appreciated. I want to make sure I get this done right, so it will be comfortable, dependable, balanced, and efficient.

 

[Dana]

Is it steam, or pumped hot water? A 1-pipe steam system isn't easy to convert, but a 2-pipe steam usually is.

A right sized modulating boiler with a 10:1 turn down ratio makes more sense than multiple boilers. This is possible if it's pumped hot water, but steam needs to be sized for the radiation, not the heat load.

Start by running a room-by-room Manual-J type heat load analysis, or at least an IBR type heat load calculation for each room separately, and add them up for the entire house load. Don't pad the numbers- be aggressive rather than conservative on R-value & U-factor estimates or you'll oversize it. If you need help figuring out U-factors for the walls we can do that here. The thickness of the brick (in inches) as well as the plaster & lath layer would be necessary to ball-park it.

Then radiator by radiator, come up with the equivalent direct radiation (EDR) square feet, and keep tabs on the ratio of the room's heat load to the EDR'. It will usually fall within some range, and the middle range if load/EDR is how you'd want to size the new radiators.

If you have a heating history on the place, run a fuel-use heat load calculation (wintertime gas bills only). For an old converted coal boiler assume 60-70% combustion efficiency, unless it's been tested & tagged at some other number by a burner tech.

More on sizing modulating condensing boilers here. With high volume heat emitter such as cast iron radiators there is a lot of thermal mass forgiveness already built-in. Zoning by floor is going to be fine, but micro-zoning it might still have short-cycling potential.

Without knowing the wall thickness, window sizes or much else about it, I'd take a WAG that the 99% heat load of a 4500' house like that will come in between 85-135,000 BTU/hr @ 68F indoors, +10F outdoors- a typical KY 99% outside design temperature. (That's a big range, I know, which is why it's important to run load calculations.) A 150K condensing boiler with a 1o:1 turn down ratio would almost surely heat the place, but a 100K or 120K boiler would be better, assuming it covers the load.

HTP's UFT series are pretty cheap and easy to deal with (can usually be pumped direct) and will have something that fills the bill. Even the 140K version is only a couple grand at internet pricing- the others are less.

Lochinvar's KHB or WHB series have more bells & whistles and are more expensive, but there'll be something there that works too.

There are others. Let's figure out your load & radiation numbers first.

 

[Sponsor]

 

[Dan Westerfield]

Is it steam, or pumped hot water? A 1-pipe steam system isn't easy to convert, but a 2-pipe steam usually is.

A right sized modulating boiler with a 10:1 turn down ratio makes more sense than multiple boilers. This is possible if it's pumped hot water, but steam needs to be sized for the radiation, not the heat load.

Start by running a room-by-room Manual-J type heat load analysis, or at least an IBR type heat load calculation for each room separately, and add them up for the entire house load. Don't pad the numbers- be aggressive rather than conservative on R-value & U-factor estimates or you'll oversize it. If you need help figuring out U-factors for the walls we can do that here. The thickness of the brick (in inches) as well as the plaster & lath layer would be necessary to ball-park it.

Then radiator by radiator, come up with the equivalent direct radiation (EDR) square feet, and keep tabs on the ratio of the room's heat load to the EDR'. It will usually fall within some range, and the middle range if load/EDR is how you'd want to size the new radiators.

If you have a heating history on the place, run a fuel-use heat load calculation (wintertime gas bills only). For an old converted coal boiler assume 60-70% combustion efficiency, unless it's been tested & tagged at some other number by a burner tech.

More on sizing modulating condensing boilers here. With high volume heat emitter such as cast iron radiators there is a lot of thermal mass forgiveness already built-in. Zoning by floor is going to be fine, but micro-zoning it might still have short-cycling potential.

Without knowing the wall thickness, window sizes or much else about it, I'd take a WAG that the 99% heat load of a 4500' house like that will come in between 85-135,000 BTU/hr @ 68F indoors, +10F outdoors- a typical KY 99% outside design temperature. (That's a big range, I know, which is why it's important to run load calculations.) A 150K condensing boiler with a 1o:1 turn down ratio would almost surely heat the place, but a 100K or 120K boiler would be better, assuming it covers the load.

HTP's UFT series are pretty cheap and easy to deal with (can usually be pumped direct) and will have something that fills the bill. Even the 140K version is only a couple grand at internet pricing- the others are less.

Lochinvar's KHB or WHB series have more bells & whistles and are more expensive, but there'll be something there that works too.

There are others. Let's figure out your load & radiation numbers first.

Thank you very much! I’m traveling on a business trip right now, but will be home tomorrow. I will get to work on the calculations. I have owned this home for 3 winters now, so I should be able get a good idea from my gas usage. I will post again once I get more information!

 

[Dana]

Thank you very much! I’m traveling on a business trip right now, but will be home tomorrow. I will get to work on the calculations. I have owned this home for 3 winters now, so I should be able get a good idea from my gas usage. I will post again once I get more information!

If you were using deep overnight setbacks or keeping the house at something other than 68-72F that will affect the fuel-use load numbers, so the thermostat setting would help fine tune it. The error bars on the ancient boiler's actual steady state efficiency are pretty big too, but with fuel use numbers it'll be possible to put brackets around the actual load, and sanity check the IBR &/or Manual-J load calculations.

 

[Dan Westerfield]

If you were using deep overnight setbacks or keeping the house at something other than 68-72F that will affect the fuel-use load numbers, so the thermostat setting would help fine tune it. The error bars on the ancient boiler's actual steady state efficiency are pretty big too, but with fuel use numbers it'll be possible to put brackets around the actual load, and sanity check the IBR &/or Manual-J load calculations.

Right now it just stays set on 70. I would like to drop the heat upstairs, and bring the heat up down stairs during the day, and the opposite at night. This is why I want 2 zones. I would like to be able turn off the heat in the finished attic since it is not used often. Not sure if I can do that though without short cycling the boiler. The attic is only 700 sqft though. . I plan to home run all radiators to supply and return manifolds so I can easily balance each room. Maybe I can turn off the attic there with out causing short cycling?

 

[Dana]

With most homes having at least one thermostat/zone per floor delivers better comfort than a 4 story house all running as a single zone. Short cycling won't be problem if the boiler isn't ridiculously oversized and has an adequate turn-down ratio. The availability of models with 10:1 turn-down has gone up considerably in recent years, making this a lot easier than it was 20 years ago when 3:1 or 5:1 ratios were the norm. The thermal mass of the water in radiators offers a lot of design forgiveness too.

If you get the sizing right, with a big turn-down boiler you can completely turn off one or more floors without taking an efficiency hit or short cycling.

It wouldn't surprise me if installers who don't run any analysis on it would propose a 250K or larger boiler for this house, using a rule of thumb approach: "Lessee, 4500 feet of uninsulated house times 50 BTU per foot comes to 225K, so bump it 15-25% just to be sure, call it 250-300K" type of rule of thumb. Those houses will never be cold, but it's far from optimal, and could run into cycling issues if zoned by floor. If the real load is 100K, and the smallest zone's load is only 15K at design condition, when just that zone is calling for heat during lower-load periods even with a 10:1 turn down a 300K boiler may short-cycle even with high-mass radiators, but a 120K boiler with a 10:1 ratio probably won't.

In my neighborhood (with design temps in the low to mid single digits) the rule of thumb hacks seem to use 50 BTU per square foot for uninsulated brick, 35 BTU per square foot for insulated 2x4 framed houses, 25 BTU/hr for 2x6 houses, and the oversize factors are more than 2x. (In other words, the real loads are less than half that.)

 

[Dan Westerfield]

I had a local HVAC company that does boilers come to give me a quote. I told him what I wanted to do, I.e. add radiators to the finished attic, add radiators to the basement, and home run each radiator to manifolds. I have an elevator shaft in the center of my home to be able to easily pull the PEX to the attic. This guy suggested to have each floor on a Zone and install thermostatic valves on each radiator to fine tune/balance each zone, and not do the homerun setup, since the running new lines to the 2nd floor would be pretty intensive, and the end effect would be the same. I am interested to see what he comes back with. He has me a bit concerned as to how he will calculate the BTU requirement, as he only measured the foot print of entire house.

I am going to download my gas usage for the last three winters to calculate it on my own. He told me that current boiler (coal fired converted to gas) is likely 20-30 efficient. Could it be that bad?

 

[Dana]

I had a local HVAC company that does boilers come to give me a quote. I told him what I wanted to do, I.e. add radiators to the finished attic, add radiators to the basement, and home run each radiator to manifolds. I have an elevator shaft in the center of my home to be able to easily pull the PEX to the attic. This guy suggested to have each floor on a Zone and install thermostatic valves on each radiator to fine tune/balance each zone, and not do the homerun setup, since the running new lines to the 2nd floor would be pretty intensive, and the end effect would be the same. I am interested to see what he comes back with. He has me a bit concerned as to how he will calculate the BTU requirement, as he only measured the foot print of entire house.

I am going to download my gas usage for the last three winters to calculate it on my own. He told me that current boiler (coal fired converted to gas) is likely 20-30 efficient. Could it be that bad?

Depending on the age, condition and type of coal boiler it could be under 50% efficiency, but efficiency as low as 20% wouldn't be very likely.

The elevator penetration to the attic is likely to be a huge heat leak that should be addressed- it's effectively a chimney sucking heat out of your house. Ideally it would be entirely within the insulation & pressure boundary of the house.

 

[Dan Westerfield]

Depending on the age, condition and type of coal boiler it could be under 50% efficiency, but efficiency as low as 20% wouldn't be very likely.

The elevator penetration to the attic is likely to be a huge heat leak that should be addressed- it's effectively a chimney sucking heat out of your house. Ideally it would be entirely within the insulation & pressure boundary of the house.

The elevator shaft only goes from the first floor to the second. It doesn’t go to the basement or the attic, so it is captured completely inside of insulated space.

The smallest Combi boiler I can find is a 155. If I only need a 100 is that going to be short cycling?

 

[Dana]

Combi boilers have fairly limited turn-down ratios, and usually only a good fit for houses with modest or moderate hot water needs combined with a high heat load. That might be you, but might not be.

Short cycling potential a function of the MINIMUM modulated output, not it's maximum. A 150K boiler with a 10:1 turn down can throttle back to 15,000 BTU/hr, whereas a 100K boiler with a 5:1 ratio can only dial it back to 20,000 BTU/hr. Read the blog article on sizing modulating condensing boilers linked to in my first response for the napkin-math on that.

It's important to figure out both the heat load and the radiation sizing for making these determinations. If the old boiler is too decrepit to be useful as a measuring instrument, an I=B=R or Manual-J load calc based on the construction materials & type is your only real option. But since you'll need to run those numbers for sizing the new radiation anyway it's not time wasted.

 

[Dan Westerfield]

I did a manual J calculation on coolcalc.com. It says I need 139,934BTU to heat my house. So maybe a 150K BTU boiler isn’t crazytown?

 

[Dana]

The ~140K load number feels a bit on the high side for a 4500' house, but not insanely high, especially when it's the output of a less nuanced online tool, which tends to oversize relative to where it would be using a pro tool and aggressive assumptions (often by 25% or more.) When using load calculation tools it's important to be aggressive rather than conservative on all the input assumptions. How much of the load did coolcalc attribute to air infiltration?

Run the same calculation using loadcalc.net or Slantfin's tool as a "second opinion", but yes it's quite possible that a ~140-150K boiler is in your future (but maybe not a combi boiler.)

The room by room load numbers and the ratio of load to radiator square feet EDR can still work for sizing the radiators. Even if the tool is trending to oversizing by 1.5x the ratio of load to radiator size will still fall within a range allowing you to specify the new rads, even if the actual load is lower.

Also be sure to run the total zone load and total zone radiation EDR numbers for each zone/floor separately. With the zone's load & radition numbers it'll be possible to figure out the water temperature at which a particular boiler would start cycling rather than modulating on single-zone calls.

 

[Dan Westerfield]

Loadcalc = 101,377 Total BTU’s heating
Slantfin’s tool = 141,574 Total BTU’s heating.

I’m waiting to get the quote back from the contractor, but after looking at the house he ballparked it at about 150k. He said he was going to go back and actually do the calculations. He did take lots of picture inside and out, so hopefully he has enough detail.

By my calculations, my smallest zone is just over 10,000 BTU’s. This would be the basement. The next smallest would be the finished attic at around 18,000 BTU’s. The 1st and second floors are neck and neck at about 55,000 BTU’s each. So if I can get a 150,000 btu mod/con boiler, with a 10:1 turn down ratio, I’m still a bit over sized for my smallest zone. Although I would guess that it would be pretty rare for the basement to call for heat and not any of the other 3 zones, but I guess it is possible. This would cause a short cycle in my boiler, right? Or is using 2/3rd’s of my lowest capacity ok?

I would love to have a Combi boiler, but if I can’t get one with a large enough turn down ratio, then I will have to buy a separate tankless DHW heater. I will be running manifolds and homerun PEX to each fixture.

 

[Dana]

If even loadcalc is coming up with ~100K you probably don't want to go more than 140K (1.4x) on the boiler, and 120K would probably be even better. Most of the time loadcalc is high by a double-digit percentage. Slantfin's tool is always at least 25% to the high side, so don't think of this as a 2:1 vote just because coolcalc came up with the same number.

The radiation sizing for the basement and other zones is the next critical piece. That 10,000 BTU basement zone only needs 10,000 BTU/hr when it's 10F outside (or whatever your local design temp is.) The water needs to be hot enough when it's that cold out to emit that much heat, and it will likely be above the temperature at which condensing efficiency can be had. Programming in the outdoor reset curve to automatically raise & lower the boiler temp in response to the outdoors is critical for getting the efficiency out of them. At condensing temperatures that hit the mid-90s for efficiency you don't get more than ~50 BTU/hr per square foot EDR out of the radiators, so we really need to know how much radiation there is on each zone. Ideally you'd have enough radiator on each zone to emit the full min-fire output of the boiler at condensing termpertures or at least well over half. That's easier to do with a 120K boiler with a 10:1 turn down ratio that can cruise along at 12K than a 150K combi boiler with a 5:1 ratio. The thermal mass of high volume radiators offers quite a bit of forgiveness on that though.

HTP has a 199K combi boiler with a 10:1 turn down ratio that might be worth considering, since it has better domestic hot water delivery than most but also a decently low (~19K) firing rate for a boiler that size. The EFTC-199W is similar to their fire-tube boilers, and runs ~USD$2K. The 140K version of that combi boiler has only a 5:1 turn down, with a ~28K minimum firing rate, which can become a temperature control problem during the summer months at low flow on the domestic hot water. But let's see where your zone radiation numbers come out first. This series is made by a first-tier Korean manufacturer (Kiturami), and imported to North America by a few other vendors. There are Westinghouse labeled versions but I believe Westinghouse buys them through HTP, not direct from Korea. Even though it's the same equipment under the paint, there will be differences in the level of support available.

The Navien NCB-240E has a similar minimum firing rate, but is more expensive and can't be pumped-direct, must be plumbed primary/secondary. In heating mode it's only good for 120K, but 199K for the domestic hot water. Given the loadcalc numbers it's going to be enough. The local support for Navien may or may not be better than HTP in your area.

 

[Dan Westerfield]

I’m working on my zone radiation numbers now. It seems the basement zone is the tough part to get around. Maybe it would be easier to just tie it into my first floor zone, then turn down the flow meter on the manifold?

 

[Dana]

Heat loss characteristics of basements differ pretty widely with changes in outdoor temperature from those of fully above-grade rooms, since the ground's temperature changes much more slowly- seasonally rather than hourly. Combining it with a first-floor zone rarely works out well.

Is it important to fully heat & control the temperature in the basement?

How much radiator is there in the basement?

If you can find specs for similar radiators to yours with published water volumes and weights per section keep track of those numbers too. A pound of water has about 9x the thermal mass of a pound of cast iron, but even the iron does adds up. With the total zone thermal mass and the potential shortfall in output it's possible to ballpark the potential cycling behavior with low-temp water.

 

[Dan Westerfield]

Right now there is no radiators in my basement. I have a gravity fed system, so the only heat in the basement is what radiates from the wildly inefficient boiler, and mostly uninsulated large black iron pipes associated with the boiler. It actually stays pretty nice down there, but I’m assuming when I replace the boiler with a high efficiency one, and redo over all of the big iron pipes, I won’t be so cozy down there. The basement is unfinished, but I do workout down there, and work on projects occasionally. This is mostly due to our detached garage being unheated.

 

[Dana]

In that case, stub out for a zone on the manifold, but defer the decision to actively heat the basement. It might be cheaper to insulate the basement walls to code min than it is to add enough radiation to not short-cycle a 100K+ boiler.

Current IRC code minimum basement insulation for zone 4A (includes all of Kentucky) us R10 continuous insulation. At 1.5" most foil faced polyiso is rated R9-R10, but most roofing insulation would be R8.5 @ 1.5", R11+ @ 2". So 1.5" of fire-rated foil faced polyiso, or 2" of reclaimed roofing polyiso strapped in place with 1x4s through screwed to the walls on to which half inch wallboard gets installed as a timed thermal barrier to meet fire codes, in which case it would likely stay above 60F even in the dead of winter, maybe even above 65F.

R10-ish wall insulation will probably lower the heat load of the basement from about 10,000 BTU/hr to less than 2000 BTU/hr, but you can test that with one of the load tools. If a particular tool doesn't have sufficient options for below grade walls, pretend the height of the room is just the above grade portion plus 2' in height-it'll be close enough. In much colder places such as Winnepeg or Montreal insulating brick walls from the interior side risks freeze/thaw spalling of the exterior brick, but most of KY should be fairly low risk. If that would concern you, 2.5" of EPS on the exterior dug down to 2' below grade up to the level of the first floor joist bottoms, protected with a purpose-made cemeticious faux stucco such as Quikrete Foam Coating would keep the brick warmer & drier and would lower the heat load almost as much.

In my somewhat cooler location 3" of reclaimed roofing polyiso (~R16-R17) keeps it a bit above 65F and very stable in the basement when the first floor is maintained at ~70F. It's a reasonable temp for a workout, not too hot, not too cold. My binned hourly mean temperature for the month of January runs about 25F outdoors. Unless your house is on a mountaintop in the Smokies it's probably something like 8-10 degrees warmer in winter on average than my house, even in cooler locations such as Covington (which has a 99% design temp of +7F).

 

[Dan Westerfield]

I did find charts online for BTU output per hour for cast iron radiators. I measured all of mine, made a fancy spreadsheet to do all the calculation for me. I can change the water temp, it then changes all the BTU calculations for each radiator. So after all the calculations, I have found that with all my radiators put together, I can raidiate 130,352.6 BTH/hr with a water temp of 180. Although, I’m pretty sure that I don’t run that hot now.

 

[Dana]

That surely puts a firm upper bound on it, eh?

If the radiation is pretty much keeping up with the heat load and the water temp is averaging less than 180F (190F out, 170F return) your load is under 130K.