Tanked hot water heater to heat 2 new radiators

[Dan Westerfield]

I have an original 1914 coal fired (converted to gas) boiler that runs my gravity hot water heating via cast iron radiators. I have a pantry, and and a small breakfast room off of the kitchen that are currently unheated by the main boiler. The pantry is roughly 4’x9’, and the breakfast room is roughly 9’x12’. After doing the calculations the pantry requires about 3,500 BTU, and the pantry requires about 7,500 BTU for heating requirements. I will be replacing my old boiler at some point, and converting the entire system to pump fed, with zones. But I don’t have that kind of change just yet. I am wondering how hard it would be to go ahead and buy the proper sized rads for these two unheated rooms, plumb them in as a “zone”. But since I can’t tie them into my old system, I thought I could run it though a heat exchanger fed and run it off of my existing 50gal, 40,000BTU DHW heater on a closed system, separate from my DHW. This way, I can get functional, period correct heat in these room, with “minimal” investment, and when I do replace the old boiler I can easily tie it into the new system, no need to “redo” anything. Am I crazy?

 

[Dana]

The only crazy part of it is the heat load numbers adding up to 11,000 BTU/hr which seem crazy high for less than 150 square feet of space if there is glass in the windows and any exterior doors close. How did you calculate those load numbers?

Try running this freebie online Manual J load tool. Set the air infiltration and ventilation assumptions to 0 cfm or it will overshoot reality by 30% or more, and be aggressive rather than conservative on insulation R-values. (eg: If it's a 2x4 framed house with fiberglass in the walls, assume R13 rather than R11.)

The actual heat load of that space is probably less than 4000 BTU/hr, and could easily be less than 3000 BTU/hr if the windows are double paned (or have storm windows) and reasonbly tight. Even with single pane windows, zero wall insulation and only R10 in the attic it wouldn't be much more than 5000, BTU/hr. But even a 150 square foot uninsulated tent would have a heat load less than 11,000 BTU/hr at 0F outdoors, +70F indoors, with the tent flaps closed.

At 120F average water temp (AWT) a cast iron radiator delivers about 50-60 BTU/hr per square feet "equivalent direct radiation". A typical fin-tube baseboard or 10" tall cast iron baseboard delivers about 200 BTU/hr per linear foot of baseboard. So if we assume a design load of 4000 BTU/hr that only takes ~20' of baseboard, or 80 inches of Sunrad or similar 5" x 20" thin profile radiator (often available for cheap on the used market- check the local craigslist for used cast iron radiators.)

With a counterflow plate type heat exchanger the output temperature of the exchanger will be only a couple degrees below the storate temperature, so designing the radiation for 120F AWT is conservative, and would also be appropriate for running with a condensing boiler when it's time to replace the gravity-feed monster.

The potable water side of the heat exchanger needs to be an appropriate potable-supply compatible material. Fortunately there are a number of stainless steel pumps used for water recirculation systems that will work here. The pumping head is low- go with the smallest one you can find, and install a ball valve to be able to throttle back the flow a bit. On the heating loop side any cheap circulation pump can work. The pumps can be energized together by just one zone relay with a thermostat input. It will also need a small expansion tank, and a connection to the potable side for filling it. Set it up to run 8-12psi, to avoid oxygen migration into the system. This is approximately what you'd be looking at (assume "buffer tank" and "heat source" combined is the same as the water heater :

 

[Sponsor]

 

[Dan Westerfield]

The only crazy part of it is the heat load numbers adding up to 11,000 BTU/hr which seem crazy high for less than 150 square feet of space if there is glass in the windows and any exterior doors close. How did you calculate those load numbers?

Try running this freebie online Manual J load tool. Set the air infiltration and ventilation assumptions to 0 cfm or it will overshoot reality by 30% or more, and be aggressive rather than conservative on insulation R-values. (eg: If it's a 2x4 framed house with fiberglass in the walls, assume R13 rather than R11.)

The actual heat load of that space is probably less than 4000 BTU/hr, and could easily be less than 3000 BTU/hr if the windows are double paned (or have storm windows) and reasonbly tight. Even with single pane windows, zero wall insulation and only R10 in the attic it wouldn't be much more than 5000, BTU/hr. But even a 150 square foot uninsulated tent would have a heat load less than 11,000 BTU/hr at 0F outdoors, +70F indoors, with the tent flaps closed.

At 120F average water temp (AWT) a cast iron radiator delivers about 50-60 BTU/hr per square feet "equivalent direct radiation". A typical fin-tube baseboard or 10" tall cast iron baseboard delivers about 200 BTU/hr per linear foot of baseboard. So if we assume a design load of 4000 BTU/hr that only takes ~20' of baseboard, or 80 inches of Sunrad or similar 5" x 20" thin profile radiator (often available for cheap on the used market- check the local craigslist for used cast iron radiators.)

With a counterflow plate type heat exchanger the output temperature of the exchanger will be only a couple degrees below the storate temperature, so designing the radiation for 120F AWT is conservative, and would also be appropriate for running with a condensing boiler when it's time to replace the gravity-feed monster.

The potable water side of the heat exchanger needs to be an appropriate potable-supply compatible material. Fortunately there are a number of stainless steel pumps used for water recirculation systems that will work here. The pumping head is low- go with the smallest one you can find, and install a ball valve to be able to throttle back the flow a bit. On the heating loop side any cheap circulation pump can work. The pumps can be energized together by just one zone relay with a thermostat input. It will also need a small expansion tank, and a connection to the potable side for filling it. Set it up to run 8-12psi, to avoid oxygen migration into the system. This is approximately what you'd be looking at (assume "buffer tank" and "heat source" combined is the same as the water heater :

I appreciate the reply! I used the “slant fin” heat loss calculator app for my iPad to do the calculation. Is that not accurate? It was room by room. It showed my entire home to need about 180k btu. Does that sound crazy? My house in total is about 3400 sqft, without the basement. I also made a spreadsheet for all my current radiators to compare thier output to what the slant fin calculation said. On average it shows the rads that I have are 37% oversized for the needs to heat the house. Except for 2 bedrooms that are fairly “undersized”, which does explain why my daughter complain so much about thier rooms being cold! Lol

So if you’re saying my calculations for those 2 small rooms is SUPER oversized, then the rest of my house sure is too. All calculations were done for a water temp of 180. There are two small rooms I’m talking about on this post:

Pantry - 5’x9’, 45sqft, 1’ thick solid brick walls with plaster on the inside. Crawl space underneath that has access from the basement. R-30 fiberglass insulation in the ceiling. One short wall is exterior, no windows or doors, and one long wall is exterior with one single pane wood framed window with storms about 2’x4’. The other short wall has a door into my kitchen, and the other long wall is shared with the pantry. The other room I’m trying to heat.

Landry - 9’x12’, 108sqft. 1’ thick solid brick walls, with no plaster on the inside. This room is on an uninsulated slab. One long wall has 2 single pane windows (2’x4’ each) with storms, and an exterior half glass (single pane) door with a storm door. One short wall has 2 single pane windows (2’x4’ each) with storms. The other long wall is shared with my kitchen, and the other short wall is shared with the pantry.

 

[Dan Westerfield]

I appreciate the reply! I used the “slant fin” heat loss calculator app for my iPad to do the calculation. Is that not accurate? It was room by room. It showed my entire home to need about 180k btu. Does that sound crazy? My house in total is about 3400 sqft, without the basement. I also made a spreadsheet for all my current radiators to compare thier output to what the slant fin calculation said. On average it shows the rads that I have are 37% oversized for the needs to heat the house. Except for 2 bedrooms that are fairly “undersized”, which does explain why my daughter complain so much about thier rooms being cold! Lol

So if you’re saying my calculations for those 2 small rooms is SUPER oversized, then the rest of my house sure is too. All calculations were done for a water temp of 180. There are two small rooms I’m talking about on this post:

Pantry - 5’x9’, 45sqft, 1’ thick solid brick walls with plaster on the inside. Crawl space underneath that has access from the basement. R-30 fiberglass insulation in the ceiling. One short wall is exterior, no windows or doors, and one long wall is exterior with one single pane wood framed window with storms about 2’x4’. The other short wall has a door into my kitchen, and the other long wall is shared with the pantry. The other room I’m trying to heat.

Landry - 9’x12’, 108sqft. 1’ thick solid brick walls, with no plaster on the inside. This room is on an uninsulated slab. One long wall has 2 single pane windows (2’x4’ each) with storms, and an exterior half glass (single pane) door with a storm door. One short wall has 2 single pane windows (2’x4’ each) with storms. The other long wall is shared with my kitchen, and the other short wall is shared with the pantry.

I ran the manual J you suggested, a bit confusing, but I did my best, it’s saying my entire house should take 64544 btu’s To heat. That is quite the swing!!!

For real world data - I currently have a the smallest wall mounted ventless gas heater (I’m guessing 10,000 Btu) I could find in the laundry room and keep it on the lowest setting and it keeps that room at 70 degrees.

 

[Dana]

The SlantFin load calculation app is a bad joke, especially when assessing solid masonry buildings. Even on framed insulated buildings it reliably overshoots reality by 30%, often more.

A heat load of 64544 BTU/hr at Kentucky style 99% outside design temperatures on a reasonably tight 3400' foot-thick masonry house is credible, at a bit under 20 BTU/hr per square foot of conditioned space. A heat load of 180K (over 50 BTU/hr per square foot!) is not. What is your 99% outside design temperature, as indicated by either the SlantFin app or loadcalc.net app? (I think SlantFin uses the 99.6 outside design temp, but I haven't looked at any recent versions of it.)

You can sanity-check those numbers by running a fuel-use load calculation. It may be hard to determine the combustion efficiency and distribution losses of the gravity system and that can't be separated out of the fuel use calc, but if you (optimistically) assumed 75% combustion efficiency on the conversion burner in the antique boiler it would put a firm upper bound on it. The actual efficiency might be as low as 50%, so don't be surprised if the fuel use calculation comes up with something like 90-100,000 BTU/hr for the load instead of 65,000 BTU/hr. It would be truly astounding if a fuel use load calc came up with 180K. Comparing the fuel use numbers to the Manual-J numbers vs. the SlantFin app numbers you'll have a basis for deciding which tool is more credible, but you don't have to guess where I'd put my money, eh?

For a third opinion, give coolcalc a shot.

 

[Dan Westerfield]

The SlantFin load calculation app is a bad joke, especially when assessing solid masonry buildings. Even on framed insulated buildings it reliably overshoots reality by 30%, often more.

A heat load of 64544 BTU/hr at Kentucky style 99% outside design temperatures on a reasonably tight 3400' foot-thick masonry house is credible, at a bit under 20 BTU/hr per square foot of conditioned space. A heat load of 180K (over 50 BTU/hr per square foot!) is not. What is your 99% outside design temperature, as indicated by either the SlantFin app or loadcalc.net app? (I think SlantFin uses the 99.6 outside design temp, but I haven't looked at any recent versions of it.)

You can sanity-check those numbers by running a fuel-use load calculation. It may be hard to determine the combustion efficiency and distribution losses of the gravity system and that can't be separated out of the fuel use calc, but if you (optimistically) assumed 75% combustion efficiency on the conversion burner in the antique boiler it would put a firm upper bound on it. The actual efficiency might be as low as 50%, so don't be surprised if the fuel use calculation comes up with something like 90-100,000 BTU/hr for the load instead of 65,000 BTU/hr. It would be truly astounding if a fuel use load calc came up with 180K. Comparing the fuel use numbers to the Manual-J numbers vs. the SlantFin app numbers you'll have a basis for deciding which tool is more credible, but you don't have to guess where I'd put my money, eh?

For a third opinion, give coolcalc a shot.

Coolcalc is showing 190,000btu for heating?!?! Lol

I have all the data and have tried to do the fuel usage calculation, but I don’t have the days that the meter was read, so I can’t be perfectly accurate. I average about 3000 cuft per month in the warmer months, and in the coldest month I used 31,500 CuFt. So if I take off 3,000 (water heater), then my highest usage is about 28,500 CuFt in the coldest month for heating alone. If I take that and convert it to BTU’s By multiplying my CuFt used by 1020 it gives me 29,070,000 BTU in the coldest month, divided by 31 days (?), that gives me 937,742 BTUS/day, divided by 24 hours gives me 39,073 BTUs/hr on average for the coldest month of the year in past 3 years. I know that a 40k BTU boiler can’t cut it, since there are ups and downs, but a 90k should have ZERO problem. One caveat, I didn’t account for the inefficiently of my current boiler in these calculations. Am I crazy? Or just plain wrong?

I just can’t understand why every calculation is SO much different.

 

[Dana]

The average BTUs/hr over a month is meaningless unless it's accompanied with the average TEMPERATURE. Averaging it over the month just tells you the average load, but you need to size the heating system for the PEAK load, not the average.

So let's say your coldest month the boiler used 30,000,000 BTU for the month. At 75% efficiency that means the house used 0.75 x 30,000,000= 22,500,000 BTU, or (/31=) 725,806 BTU/day, or (/24=) 30,241 BTU/hr.

But that's your average load, not the 99% peak load.

Let's assume you were in Lexington where the 99% outside design temperature is +10F, and the average temperature in January is about 35F. Assuming you keep the place between 65-70F the heating/cooling balance point is about 65F. If the average indoor temp is cooler than that we can adjust it, but this is just an example.

The average temp of 35F is 30F colder than the balance point. So for a linear approximation (which is good enough for boiler sizing purposes) you are looking at 30,241/30F= 1008 BTU per hour of heat load for every degree below the balance point.

At +10F the house is (65F -10F)= 55F below the heating/cooling balance point, for an implied load of 1008 BTU/F-hr x 55F = 55,444 BTU/hr.

Even your input BTUs if used at 100% efficiency (instead of 75% efficiency) would only imply a load of 73,925 BTU/hr @ 10F.

So, which load calculator do YOU think is more accurate, based on the fuel use?

I'm still going with loadcalc, since the others would imply your antique boiler is operating at 250-300% efficiency (which is not very likely.)

 

[Dan Westerfield]

The average BTUs/hr over a month is meaningless unless it's accompanied with the average TEMPERATURE. Averaging it over the month just tells you the average load, but you need to size the heating system for the PEAK load, not the average.

So let's say your coldest month the boiler used 30,000,000 BTU for the month. At 75% efficiency that means the house used 0.75 x 30,000,000= 22,500,000 BTU, or (/31=) 725,806 BTU/day, or (/24=) 30,241 BTU/hr.

But that's your average load, not the 99% peak load.

Let's assume you were in Lexington where the 99% outside design temperature is +10F, and the average temperature in January is about 35F. Assuming you keep the place between 65-70F the heating/cooling balance point is about 65F. If the average indoor temp is cooler than that we can adjust it, but this is just an example.

The average temp of 35F is 30F colder than the balance point. So for a linear approximation (which is good enough for boiler sizing purposes) you are looking at 30,241/30F= 1008 BTU per hour of heat load for every degree below the balance point.

At +10F the house is (65F -10F)= 55F below the heating/cooling balance point, for an implied load of 1008 BTU/F-hr x 55F = 55,444 BTU/hr.

Even your input BTUs if used at 100% efficiency (instead of 75% efficiency) would only imply a load of 73,925 BTU/hr @ 10F.

So, which load calculator do YOU think is more accurate, based on the fuel use?

I'm still going with loadcalc, since the others would imply your antique boiler is operating at 250-300% efficiency (which is not very likely.)

You have explained that calculation MUCH better than I have been able to find anywhere else! Thank you! We keep our house at about 70 degrees. I actually am about 30 minutes from Lexington. We do get the occasional cold snap like the one 2 weeks ago where we get 4 or 5 degrees below 0. So worst case scenario, it’s -5f outside and I want my indoor temp to be 70f, that is a difference of 75 degrees x 1008btu /hr/degree = 75,600 BTU per hour to heat my house to 70 degrees. If I go with a 85% conventional boiler I would need a unit capable of 86,940 BTU. If I go with a 95% high efficiency boiler I would need one capable of 79,380 BTU’s. Am I right?

 

[Dana]

ASHRAE recommends 1.4x the 99% load. So with more careful calculation and better weather data you come up with 55,000 BTU/hr (as in the example), 1.4 x 55K= 77,000 BTU/hr for output. That would be for either cast iron or conventional boilers and it would be the DOE output not the IBR net or similar derating. Those deratings are to account for the standby and distribution losses, but in a fuel use calculation those are already included.

Going a bit higher or lower than that isn't a deal breaker. In a high mass masonry house you'll ride through a few hours of even -10F without the temperature dropping more than a degree or two if you're only a few thousand BTU/hr shy on boiler output.

If you run the system at a lower temperature (as you would naturally with a condensing boiler) you'll find the actual heat loss goes down, since the boiler room would run cooler for a smaller inside to outside temperature difference in that part of the house. With a cast iron boiler it may be worth setting it up with an "outdoor reset" control to achieve a similar effect, but all condensing boilers come with that feature.

 

[Dan Westerfield]

ASHRAE recommends 1.4x the 99% load. So with more careful calculation and better weather data you come up with 55,000 BTU/hr (as in the example), 1.4 x 55K= 77,000 BTU/hr for output. That would be for either cast iron or conventional boilers and it would be the DOE output not the IBR net or similar derating. Those deratings are to account for the standby and distribution losses, but in a fuel use calculation those are already included.

Going a bit higher or lower than that isn't a deal breaker. In a high mass masonry house you'll ride through a few hours of even -10F without the temperature dropping more than a degree or two if you're only a few thousand BTU/hr shy on boiler output.

If you run the system at a lower temperature (as you would naturally with a condensing boiler) you'll find the actual heat loss goes down, since the boiler room would run cooler for a smaller inside to outside temperature difference in that part of the house. With a cast iron boiler it may be worth setting it up with an "outdoor reset" control to achieve a similar effect, but all condensing boilers come with that feature.

I really appreciate your help!

 

[Jadnashua]

The temperature usually doesn't stay at the coldest point for really long times. So, say the heating can keep things at your set point at 0-degrees, if it got to -1, over time, the house would cool off about 1-degree. Having a little excess isn't bad as it will let you recover from a setback. Way excessive, and you start to get short cycling, which is both a life and comfort issue.