Chimney vs. Direct Vent vs Power Vent for oil boiler

[thomase00]

I looked more closely at the Manual-J report generated by HVAC-Calc and found some issues:

1.) Since I was using this primarily to calculate cooling load (heat gain) for my central AC, and my central AC has ducts (though insulated) in unconditioned attic space, the heat loss calculation ALSO included duct losses, assuming that the attic is at the outside temperature during heating season. These losses add up to 8K BTU, which I think is a bit much considering that I don't actually have air moving through the ducts during the heating season.

2.) For my brick fireplace, the software gave me 3 choices for "leakiness": poor, average, and good. In the winter, the bricks are cool to the touch and there can sometimes be a little draft coming from the hair-thin gaps between the hearth and the flooring. Not being sure what poor/average/good meant, I choose poor to be conservative. This had ZERO effect on the cooling load, but added 26K BTU to the heating load! I'm guessing the software assumes that "poor" means "wide open damper sucking heated air out of the house"! Also, when this whole project is done, the brick fireplace will be gone!

3.) Again, knowing that some of my windows are drafty, and hearing not so great things about the builder, I chose "average" tightness which amounts to 0.7 air exchanges per hour in the winter. This results in 32K of heat loss! Obviously, this is a HUGE factor, so it greatly skews the result if my guess is off. Unfortunately, I never got an accurate measure of ACH when MassSave came to do a blower door test. They just gave me some measure of the before/after difference after sealing my attic, and it WASN'T in units of air exchanges per hour.

So, out of the 91K BTU of estimated heat loss, 66K is due to what essentially amounts to a guess!

Edit: Also, the heat gain calculation (for cooling load) included about 5K BTU from occupants and appliances, and this was not automatically subtracted from the heating load.

 

[Dana]

The internal heat gains required for cooling load calculations are not the same as for heat load calculations. It's important to include all the reasonable worst-case for plug loads and occupants for peak cooling loads, which occur during the day when people are likely to be up & active, bathing & cooking, etc. For calculating heating loads you have to assume the lights & TV are going to be off a 5AM on the coldest night of the year, and only include sources that are going 24/7, such as refrigerators, water heater standby losses, etc.

Infiltration losses of 32,000 BTU/hr is almost like having a window or two open. At 0.018 BTU/cubic foot per degree-F and a presumed 67F temperature difference (+1F outside, +68F inside) every cubic foot of infiltration air is then worth 0.018 x 67F= 1.2 BTU, so to add up to 32K that's 32,000 /1.2= 26,667 cubic feet per hour, or (/60=) 444 cubic feet per minute. That's like having 5 bath fans running, or a larger than average kitchen range hood on high speed.

The fireplace flue and other infiltration assumptions are clearly what drove the large Manual-J error.

If you haven't already, it's worth caulking the AC register duct boots to the ceiling gypsum everywhere (with polyurethane caulk or can-foam if it's a big gap), since that can add up to a substantial air leak. Any penetrations of the upper floor ceiling, or from the outdoors into the basement are potentially high leak points disproportionate to their actual size, since the highest & lowest elevation leaks have the largest amount of stack-effect infiltration drive.

 

[Sponsor]

 

[thomase00]

I've been waiting for an estimate from a contractor for a propane boiler. It has been 2 weeks and still no estimate. This is from the company that delivers my oil, so they know how much oil I'm using. He "converted" my oil usage to hypothetical propane usage based on some assumptions about efficiency improvements, and used this to argue that using propane will cost a similar amount, or even less. He had previously commented that my house has more baseboard than he would have expected. I told him that I did a heat load calculation of 55K based on oil delivery records and degree days. He replied with the following:

"Your calculations are off. Oil delivery records and degree days have nothing to do with heat loss load calculations. What they are good for is determining consumptive use. What I propose for you will be the correct size boiler for this address and your existing baseboard. If you were to install a boiler with a max capacity of 55k BTU, your home would not heat well at all. Remember, your baseboard is existing."

 

[Dana]

He's way off base, on both fronts. He may be honest but ignorant, or he may be angling to hook you into propane micro-monopoly, but run the real numbers, and take a close look.

A gallon of oil has 138,000 BTU source energy. Burned at 85% efficiency it delivers 117,300 BTU /gallon into the heating system. Normalizing to gallons per million BTU (MMBTU) delivered, that's 1,000,000/117,300= 8.53 gallons/MMBU.

A gallon of propane has 91,600 BTU of source energy. Burned at 95% efficiency in a condensing boiler that delivers 87,020 BTU/gallon into the heating system. Normalizing to MMBTU, that's 1,000,000/87,020= 11.49 gallons/MMBTU.

MA propane & oil pricing averages can be found here.

At the recent season's ~$2.70 price point for residential oil, heating with oil costs (8.53 gallons/MMBTU x $2.70/gallon=) $22.92/MMBTU

For propane to break even it would have to cost $22.92/11.49 gallons= $2.00/gallon.

But for most people in MA it averaged over $3/gallon, which makes it over $34/MMBU.

That's more than just a little bit more expensive than oil, or about the same as oil, it's fully 50% more expensive than oil!

Propane vendors get in the door with a sucker-pricing of $2 or less for the initial tank, then market rate thereafter. Since they own the tank and exclusive right to fill it (unlike the oil model where you own the tank, and can buy from any vendor at will) it's tough to change vendors, there is usually an expensive tank removal fee (I've read of fees as high as $2000) for the tank if you want to change. It's a micro-monopoly, and you're kinda screwed.

Then there's the sizing to the radiation issue:

With 55,000 BTU/hr of boiler output at 180F into the 62 foot zone, that's 887 BTU/foot, more than the radiation can even deliver at 180F AVERAGE water temp (AWT), let alone entering water temp (EWT). Even so the delivery is well above the zone load, and zone calls will be relatively short.

With 55,000 BTU/hr of boiler 180F output into the 72' with the 8500BTU @ 180F EWT kickspace heater, that leaves 46,500 BTU for the baseboard, for a ratio of 645 BTU/foot, which is more than it will deliver with 180F EWT.

When both zones are calling for heat at the same time the boiler's temp will drop a bit, to where the radiation output matches the 55,000 BTU/hr output, but it'll still deliver the full 55,000 BTU/hr, and that's still well above your actual 99% design condition heat load, or even your 99.6% design condition load- it's your once or twice in a boiler's lifetime load. Even at -30F you probably have enough radiation to keep at least one of your zones fully up to temperature, even if you have to let the other one slip. (And even the smallest oil boilers all put out more than 55,000 BTU/hr, though there are some propane boilers that don't.)

Sizing the boiler to whatever the radiation will deliver at 180F (either EWT or AWT) is a common approach by hacks too lazy or busy to run any sort of heat load calculation, but it's also the WRONG way to size a boiler! That's even more so with a modulating condensing boiler. With a modulating condensing boiler you need the smallest-radiation zone to be able to operate at condensing temperatures without short-cycling. If you size it by the maximum output, to be able to reach 180F with the existing radiation you will almost always end up oversizing it for condensing purposes, since they do not have an infinite modulation range. The napkin math reasoning behind that lives here.

At condensing temperatures low enough to deliver 95% combustion efficiency most baseboard puts out between 150-200 BTU/hr per running foot, so your 62' zone really wants something that can dial back to 62' x 150= 9300 BTU/hr. But at 180F EWT your combined radiation will deliver about 85,000 BTU/hr. To be able to deliver 85K @180F out yet dial back to 9.3K @ 115F out you need a boiler with least a 10:1 turn down ratio, that is also magically sized exactly for 85,000 BTU/hr maximum. While there are one or two boilers out there that come close to filling that bill, none really do. Most modulating boilers have a 5: 1 turn down ratio many have less. But what you really need is one that can cover your design heat load with a bit of margin, that also modulates low enough to condense without short-cycling on zone calls, and a minimum modulation well below your average midwinter load so that it actually modulates rather than cycles on/off. There are many more options with those constraints.

Don't be surprised if your contractor comes back with a proposal for a 100-120K (input) boiler that has a minimum modulation over 20,000 BTU/hr (which is above your average wintertime load.) But don't buy it!

For your house a Navien NHB 80 would be fine, even though it "only" delivers ~70,000 BTU/hr at max fire when it's above condensing temperatures, but the NHB-110 may cycle enough on zone calls that you'd have to bump it's temp up, giving up a 2-3% in combustion efficiency for no good reason. The NTI Trinity TX51 would also be fine, even though it won't cover the full load at -15F. But even a right-sized propane mod-con won't be as cheap to operate as a 2x oversized oil-burner.

For the record- I live in a 2400' (+ 1500' insulated basement) 1.5 story 1920s antique, with antique single pane windows + clear glass storms, R19 in the attic, and known gaps in the wall insulation. This is no superinsulated hyper-efficiency house! I'm radiation-limited to less than 45,000 BTU/hr total out of the system at the water temperatures I'm running, and it sails through -10F weather without losing ground. At 180F the radiation could deliver about 120,000 BTU/hr, but that would be just plain silly, and not nearly as comfortable!

 

[thomase00]

I'm not sure it is valid to look at those average propane prices because of the way it is priced. Multiple suppliers that I have talked to have indicated that pricing is based on annual usage. There are different pricing tiers for <200, <400, <600, <800, <1000, and >=1000 gallons per year. Someone who only uses propane to cook may pay $4 per gallon while someone using over 1000 gallons annually for heating pays $1.89 from the same supplier. The price surveys include both types of customers. At the highest usage, propane is currently slightly cheaper per gallon than oil, but it needs to be about 75% the price of oil to break even, assuming 87% efficient oil combustion vs. 95% efficient propane combustion.

This particular suppler supposedly prices propane at 78% the price of oil, but their oil seems a bit higher than average and their propane is a bit lower than average. Another supplier that I've talked to has cheaper oil, but the propane is priced at 95% the price of oil.

It might close the gap a bit for propane if you consider that it can turn down the output in order to reduce combustion inefficiencies due to cycling. Then again, you can't count on it hitting 95% efficiency on average if it is not in condensing mode all of the time.

 

[thomase00]

Another thing to note about the Manual-J software...

When I selected "average" tightness, this equated to 0.7 ACH in winter, and 0.4 ACH in summer. This implies that the rate of air infiltration increases as the inside/outside temperature difference increases (i.e. much larger delta-T in winter). Doesn't this further imply that the BTU required to raise the inside temperature by 1F increases as the inside/outside delta-T increases? If so, what does this mean for our estimation method considering that it assumes a constant BTU per degree day?

 

[Dana]

The "...number of BTU required to raise the inside temperature by 1 degree..." is a constant, is not dependent on temperature difference, only the thermal mass of everything in the house (specific heat x mass).

BTU/hr is a rate, which is dependent upon the temperature difference and the thermal conductivity & emissivity, the air leakage, and the stack effect drive leakage. The leakage rates in winter are higher than summer in part due to the bigger temperature differences, but it's still fairly linear with temperature, unlike other aspects of heat loss.

There are many nonlinearities when zooming in on individual aspects. The U-factor of the windows very nonlinear with temperature, and are true only at a single specified temperature & temperature difference, since the radiative loss fraction is dependent on the 4th power of the absolute temperature( above absolute zero.) The R-values of fiber insulation increase at bigger temperature differences, foam insulation varies in a very non linear fashion with average temperature through the layer, but it's fairly linear in regions that matter, but with a much steeper slope than the changes with fiber insulation.

But a linear approximation of the BTU/hr per degree temperature difference is still good enough for HVAC estimation purposes, if not for the materials scientists. Being within 15% of the true numbers is fine for sizing heating or cooling equipment (if ONLY the real world installations were estimated as close as that, with only modest oversizing factors!)

Most homes in MA and other cool/cold regions are tighter than "average" assumptions built into most Manual-J software packages for the simple reason that when they are that leaky they are drafty and noticably uncomfortable. The cost retrofit air sealing a plywood sheathed 1970s house to bring it into current code (maximum of 3 ACH @ 50 pascals pressure) typically varies from $0 (meaning it's already that tight) to a couple thousand USD, though there are outliers to the high side.

Blower door directed air sealing finds the leaks quickly, but blower door leakage numbers on their own don't tell you where the leaks are, on the cumulative size of the hole. The air leaks in the upper floor ceiling and into the basement are quite a bit more important, since that's where the largest stack effect drive exists. The biggest unseen air leak in most of these houses is the seam between the foundation sill & the top of the foundation, and at the band joists, usually larger than all window & door crackage combined. Chimney, electrical & plumbing chases that extend from the basement to the attic can also be quite large. Fireplaces are generally leaky too, but depend a lot on the tightness of the masonry, and the tightness of the flue damper. Retrofit gasketed flue-top dampers can take a big dent out of it for some fireplaces, as can a tightly installed wood burning fireplace insert (which turns it into an effective heating appliance rather than than a big heat leak that operates with pathetically low efficiency when burning.)

If you think the place is pretty leaky it's worth a round of air sealing by a competent contractor. One symptom of excessive leakage is the dryness of the interior air. A 3ACH/50 house will usually stay above 30% relative humidity except during extended sub-zero weather, and needs to run kitchen & bath exhuast longer to keep humidity under 40%, and to get cooking odors to disappear in a reasonable time fraction.

 

[thomase00]

The "...number of BTU required to raise the inside temperature by 1 degree..." is a constant, is not dependent on temperature difference, only the thermal mass of everything in the house (specific heat x mass).

You said it yourself that BTU/hr depends on stack effect among other things, and that stack effect increases as the temperature difference increases. If stack effect is measured in ACH, and ACH increases with the indoor/outdoor temperature difference, doesn't this mean that the BTU/degree-hour also varies with the indoor/outdoor temperature difference? In other words, assuming an otherwise perfectly tight house, BTU/degree-hour will be higher with 2 skylights and 2 windows open vs. with 1 skylight and 1 window open (i.e. opening an extra skylight/window pair simulates an increased stack effect due to increased delta-T).

This is what I meant when I said its not a constant. In other words, the relationship between BTU/hr and delta-T is non-linear. I THINK we are saying the same thing.

I was just pointing out that according to your estimation model, we calculate an "average" BTU/degree-hour for a period and use that to extrapolate BTU/hour LINEARLY for our chosen design day delta-T. I guess you are arguing that the linear approximation is good enough.

I'm further pointing out that while there are factors that contribute to overestimation (the average boiler efficiency being somewhat less than the steady-state efficiency, the inability to factor out DHW fuel usage, etc.), the non-linearity of the delta-T vs. BTU/hour curve can contribute to underestimation, particularly if you want to extrapolate to a delta-T even greater than that for the design day (e.g. -15F). Therefore, can we ALWAYS say with confidence that this estimation yields an upper bound?

 

[Dana]

An exceptional case where wintertime fuel load calculations aren't an upper bound would be a super-insulated house with high solar gain, and where some other energy source (such as an electric water heater) is used for heating hot water.

Another exception would be if you set one of the zone thermostats to 50F and didn't use that space over the time period of that fuel use.

Another exception where fuel use calculations OVER estimates the heat load is when the boiler is outside of conditioned space, and the distribution & jacket losses are truly lost. When the heating designers are working from a Manual-J they have to estimate the distribution losses to size the boiler correctly, but with a fuel use calculation those losses are already included in the dumbed down simple load number. With a different boiler operating at a different temperature & duty cycle those distribution loss numbers move around a bit, but it doesn't tell you how much boiler you would need if you moved it all inside the insulation boundary of the house, which is going to be lower.

Yes, real heat loads are not perfectly linear, but different component nonlinearities aren't in the same direction, and will counteract. Don't let the perfect model become the enemy of "good enough model" here. All models will fail under reductio ad absurdum scrutiny. When Manual-J is done correctly (using aggressive rather than conservative assumptions) , and infiltration rates are estimated using blower-door data (which also has large error bars, since it doesn't tell you where the leaks are, only the size), careful fuel use calculations tend to hew pretty closely to Manual-J, at least closely enough for purposes of specifying a boiler. If you want a better model they do exist, and if you're really curious, download a copy of BeOpt and take the tutorial. It's not perfect either, but it's better than a Manual-J.

Unfortunately something like, 19 out of 20 residential heating contractors make no attempt to accurately estimate the heat load, and will either size it to the radiation (like your contractor), or use some idiot's BTU per square foot of conditioned space rule of thumb. The most popular rules of thumb I see in MA are "25 BTU per foot" and "35 BTU per foot". So for your 2200' house they do the "Lessee, 2200 feet times 25 BTU per foot comes ta 55K, so let's make it a 75K boiler just to be safe", or , " At 2200 feet times 35 BTU/foot we'll spec a minimum of 77,000 BTU boiler, or the next size up. Those methods reliably oversize the boiler usually by 2x or more. The client never gets cold, and the contractor never gets the mid-winter 5AM call from the irate shivering customer, but oversizing it leaves some efficiency and reliability on the table relative to a perfectly right-sized or 1.4x oversizing factor.

A more realistic rule of thumb (for the heat load, not the boiler size) for reasonably tight (not supertight) insulated 2x4 houses with double panes or storm windows is 15 BTU/hr @ 0F per square foot, which would put your house at 33,000 BTU/hr @ 0F. That may in fact be pretty close if you were to separate out the water heating fuel use and the distribution losses. But there will be numerous exceptions to that rule of thumb, some far enough off that it could leave you cold. It's fine to use that 15 BTU/ft rule of thumb for sanity-checking load estimates & calculations- there is usually a discernable reason when that rule of thumb is far off the mark, such an unusually large amount of window area, or large & obvious air leakage, such as un-dampered flues, etc.

The smaller the boiler, the fewer the burn cycles it will experience in a year. This is because it's smaller output makes it more likely that a larger zone can actually deliver the full output of the boiler without cycling, and even when the zone radiation can't take it all, the smaller size of the boiler leads to longer & few burns while cycling, and since it takes longer to satisfy the thermostat, it leads to overlapping zone calls that add enough radiation to take the full output. This is why with zoned systems you don't want to upsize the output of the boiler to match what the total radiation delivers at 180F. That is unless it's absolutely necessary for covering the design load, or (as in your case) when you don't have much of a choice, since oil burners smaller than 0.50-0.60 gph aren't generally available.

 

[thomase00]

Don't be surprised if your contractor comes back with a proposal for a 100-120K (input) boiler that has a minimum modulation over 20,000 BTU/hr (which is above your average wintertime load.) But don't buy it!

He wants to install a Bosch Greenstar ZBR35. He says he sized to the radiation, ADDED some margin for the indirect, and rounded up "in case I ever put on an addition".

He said SO many things in conflict with what you are arguing, but he didn't really back any of it up:

• The system "won't work right" if the boiler is undersized relative to the baseboard capacity.
• You never want to size the boiler exactly to your load, because then it would have to run all the time which will wear it out faster.
• When I pointed out that the indirect has priority and doesn't run at the same time as the other zones, he got frustrated and said it didn't matter and he doesn't know where I'm getting my info from.
• When I pointed out that if there is too much baseboard, the temperature will just settle below 180F, he again got frustrated and said I was wrong and that I'll get no heat, again wondering where I was getting my wrong information. I asked if I wouldn't just need a bypass to raise the return water temperature, and he said I had it backwards, but didn't elaborate
• I can't be sure that I did the Manual-J right (he's probably right about that).
• The ONLY thing that oil use records are good for is to price out how much the equivalent BTU of propane would be when comparing fuels.

He's definitely a nice guy overall and I don't think he is trying to deceive, but he also has no desire to debate this with me. I have a suspicion that he's in line with what most other contractors are going to want to do though. Is this a case of "perfect is the enemy of good enough"? After all, this is what he does for a living. I've only been researching this stuff for a week.

The price is good though. He foots the bill for the propane tank, and it sounds like their current pricing is $1.79 per gallon, but I suppose they could crank it up when I get hooked. The pricing is still based on usage though. The best tier is over 1000 gallons a year. I have a friend in NH who has moved around between different suppliers a few times. Each time, the new supplier buys out the rented tank from the previous supplier. I think he pays just a bit over $2 per gallon.

I'm a EE like you (ASICs, digital circuit design, etc.) and have a passion for optimization. The problem is that it becomes a big barrier to pulling the trigger on home improvement projects. We are not talking about life or death here.

I'm eagerly awaiting my appointment with the next contractor...

 

[Jadnashua]

One thing to consider is this...if your radiation and boiler were capable of keeping the house warm and supplying all of your hot water needs at say 0-degrees F, if it gets to -1, the house would (after time because you do have insulation and thermal mass), end up one degree cooler...not frigid. And, then consider that it is rare when it is super cold for the entire day to be below the (proper) 99% design temperature - it does get warmer generally once the sun comes up...then, you will almost always have at least some excess capacity. On/off cycles waste energy and tend to wear things out faster than continuous use. Last...the house is more comfortable when the heat output is constant rather than cycling up/down (which is where a modulating boiler can excel - matching load to output). Depending on the type of radiators...it tends to be quieter, too because you don't get any of those expansion/contraction noises when it cycles.

Having a lack of excess capacity, depending on the radiation available, can slow the recovery rate from a setback. SOme of the better thermostats learn that rate, and can be set to have the house the desired temp at a particular time, starting earlier on a cold day, and later on a warmer one, rather than you setting it to come on at 6am when you really want it warm at 7 - those thermostats might wait until 6:45 on a warmer day, saving 45-minutes of time at the higher house temp. How often are you needing 100% of the heat available, though? If it's that cold, you might just choose to not set back (in the worst case, it would drop because you need more than what's available), but it would recover once the sun came up.

Finding a tech that will properly size HVAC is unfortunately, sometimes pretty tough...for reasons Dana related, most err way on the high side. It's hard to know what the cost of fuel will be over the life of a boiler...generally, it goes up. We're in an abnormal situation right now, but something could quickly change that, and efficiency becomes a lot more important. Who doesn't really want to save some. AN oversized boiler is not a frill, it's a waste to both buy and in comfort and efficiency.

 

[thomase00]

The contractor was probably just speaking rhetorically when he mused "I don't know where you are getting your information from" but do you think I should share this blog link with him?

http://www.greenbuildingadvisor.com/blogs/dept/guest-blogs/sizing-modulating-condensing-boiler

I don't want to burn bridges or step on toes because I genuinely like the guy. Like I said earlier, this is what he does for a living.

 

[thomase00]

I looked more closely at the boiler, which raises some more questions.

It looks like its a Weil-McLain 566, which has an output of 173K BTU with a 1.5GPH oil firing rate. However, the service tags, of which there are many going back even to the previous owner, all indicate a 1.0GPH oil firing rate.

So, this has been down-fired to a 115K BTU output.

Why not get something a little bigger for "insurance" and down-fire it to more closely match the load? Also, with the same burner BTU output, doesn't the higher mass of water in the larger boiler act as a buffer and stretch out the cycles even more?.

 

[Dana]

The smallest nozzles that work with #2 oil in the US are 0.5 gph- there's not much down-firing room available if you buy a Biasi B10-3 with a 0.5gph nozzle or a smallest of the line Buderus or Burnham with a 0.6 gph nozzle.

You don't need any "insurance", any of these suckers are already more than 1.4 x oversized for your heat load at your 99th percentile temperature bin. There are no nozzles that will down fire enough to actually " match the load.

Down-firing a 1.5 gph boiler to 1.0 gph reduces the steady state combustion efficiency, but also reduces the standby losses- it's usually a gain. But vented into an oversized terra-cotta lined chimney (without a stainless steel liner) is a recipe for flue condensation destroying the mortar in the chimney. It can be done, but the stars have to align on the flue size to work with low risk. If your current flue does NOT have a stainless liner, this is a likely reason the masonry chimney is toast.

The thermal mass inside modern oil boilers is pretty small compared to 50-75 years ago. The Biasi B10-3 has about 30lbs of water in it and another 25lbs of water-equivalent cast iron- that's it. To double the thermal mass you'd have to bump up to a ridiculous oversizing level, and downfiring it would be LESS efficient. If you need/want to buffer the heat (say, for a micro-zoned system ) to establish longer minimum-burn times, an oil fired water heater has an order of magnitude more thermal mass than any of these boilers. Your best bet is to go as small as possible, and use heat purge control.

The Energy Kinetics System 2000 steel boilers have more sophisticated heat purge and thermal mass management controls (comes with an indirect HW heater as part of the system, used as a buffering thermal mass), but the smallest of the line there is in the 0.70 gph range. A 3-plate Biasi + indirect with a retrofit heat purge controller can probably match or beat it on as-used efficiency with your loads, as can the smallest MPO-IQ + indirect, etc.

[edited to add]

Note, the smallest System 2000 EK1 fitted with the 0.68 gph nozzle delivers about 82,600 BTU/hr. With your ~150' of baseboard + ~8500 BTU/hr kick heater that's matched almost EXACTLY to what the radiation can emit with a 180F EWT, and is the largest non-modulating boiler that makes any sense at all to install on your system. But even that is larger than optimal given the diminutive size of your actual loads.

 

[thomase00]

As an aside, I have an update on my heat loss calculation:

It's been bothering that the single largest factor contributing to heat loss in my Manual-J is infiltration, so I've been trying to come up with a better number.

The initial design day heat loss that I calculated from the Manual-J (for 1F outdoor and 68F indoor) was 91K. However, this includes a total guess at infiltration losses of 32K, fireplace losses of 26K (which I think is totally bogus, ESPECIALLY considering that I'm tearing it down), and duct losses of 8K (which I don't think apply because I don't use forced air for heat, and the calculation seems to assume losses due to 120F duct temp exposed to 1F ambient).

If I subtract out ALL of this from the 91K, I get to about 25K BTUH, but this is of course without air infiltration losses.

I was able to dig up the results of a blower door test done after air sealing last year, which measured 2735 CFM50. The following paper has a table of conversion factors (varying by region), by which you can divide ACH50 to get ACHnat (also applies to CFM50 vs CFMnat):

https://www.energystar.gov/ia/home_improvement/home_sealing/ES_HS_Spec_v1_0b.pdf

For my climate region, the factor seems to be 14.8, but just to be conservative, lets round that down to 10. If you do the math, you arrive at a natural infiltration rate of 273.5 CFM, or 16410 cubic feet per hour.

From here:

http://www.greenbuildingadvisor.com...manually-calculate-air-infiltration-heat-loss

I learned that "The heat capacity of air at sea level is, on average, 0.018 Btu per cu. ft. per degree F."

So doing the multiplication for a 68F design deltaT, I get 16410 * 68 * 0.018 = ~20K BTUH

Adding this to my baseline 25K BTUH without infiltration arrives at 45K BTUH, which agrees ALMOST EXACTLY with the load calculation I did based on fuel use and heating degree days.

I'm feeling more confident about my estimates now that 2 totally different methods are getting close to the same results! Granted, the Manual-J ignores the basement, but I can count on one hand the hours over the course of the entire winter where we EVER turned on that zone, and its only 320 square feet anyway.

 

[Dana]

The load of a mostly below grade 320' insulated basement is pretty tiny- within the error margin of the other parts of the heat load calculations.

At an EWT of 180F that 15' stick o' fin tube can only put out 7500 BTU/hr anyway, which is clearly overkill, since that zone almost never runs.

 

[thomase00]

I used the slant-fin program as well, converting my ACHnat estimate to slant-fin's "infiltration factor" according to instructions I found on greenbuildingadvisor.com, and I came up with 42K BTUH!

I'm feeling really confident now!

 

[Dana]

Mind you, SlantFin's load tool almost always delivers at least a 15% overshot, sometimes even 30% higher load numbers than a more aggressive & accurate professional Manual-J. I'm even more confident that your real heat load is under 40K, not that it makes a difference on your oil-boiler choices- it's still the smallest boiler in anybody's line:

Even if the 42K were real, using ASHRAE's recommended oversizing factor, 1.4 x 42K= 59K, the output of a smallest-available nozzle 0.50 gph triple-pass oil burner.

The 45K from the fuel use calc is an upper bound, but also likely to be pretty high, due to the 2x+ oversize factor and your reported affinity for taking long showers. Typical wintertime hot water fuel use is on the order of ~10% of the total (annually it's more like 25-30%). Yours is likely to be higher than typ if you tend to take 12-20 minute shower rather than 6-10 minutes.

 

[thomase00]

Maybe I didn't make it clear, but I have NOT yet installed a new central AC condenser and air handler. The current system is VERY old. I have been able to squeeze a few more years out of it, so I never pulled the trigger on a replacement. It seems like a replacement is going to cost me in the neighborhood of $10K (duct rework is needed). My air handler is in the attic and I have a single return on the ceiling of the 2nd floor (right below the air handler). I've had ice dams in the past, so I wouldn't want to add any sources of heat to my attic if I can avoid it. I know a fossil-fuel fired furnace in the attic wouldn't be great for that reason, but I don't know about a central heat pump air handler.

I hear a lot about "cold climate" mini-splits maintaining their efficiency down to sub-zero temperatures, but what about using a central heat pump (combined with AC) that takes advantage of my existing duct work? Does a more traditional style central heat pump differ from a "cold climate" mini-split in such a way that it does not work well during cold New England winters?

Keep in mind that we also want to install a propane fireplace. Originally, this was more for aesthetics/luxury, but would it be efficient enough (around 70%) to be a reasonable backup heat source, if a heat pump has problems at low temperatures? The fireplace units that we are looking at are capable of putting out up to ~30K BTUH.

Of course, going this way means abandoning in-place or removing my hot water baseboard heating system (along with all the associated "patching up" work), as well as swapping out my 6 year old indirect water heater for a propane-fired one.

Anecdotally, I have a local acquaintance in town whose similarly sized home has a central heat pump and solar panels. He claims that his electric heating costs for the ENTIRE past winter was only ~$400 dollars (unless I'm not remembering the conversation accurately).

 

[Dana]

There are ducted heat pumps that can use the existing ducts, but the duct losses are large if located in the attic, which would also be likely to cause ice damming. You would be probably be looking at a 4-5 ton Carrier Greenspeed or similar to cover the extra load of the duct losses to the attic, and the installed price tag would be in the ~$18-25K range if competitively bid, probably higher. Ducts in the attic add to the load, both from direct gains/losses, but also from air-handler driven infiltration that comes from punching the holes in the ceiling, magnifying the amount of infiltration losses due to induced room-to-room pressure differences. If it's all inside the pressure boundary of the house those effects diminish (a lot!)

I saw a bid for a 4 ton GreenSpeed + 1 ton mini-split come in at $40K on Marthas Vineyard a couple of years ago, but some of that was "Island Pricing" effects- it's cheaper elsewhere, but mini-splits & multi-splits are cheaper still and more effective at sub-zero temps. In MA the only realistic cold climate mini-split options that are well supported are Mitsubishi (who has more than half the total market) and Fujitsu.

A single 4-8 zone cold-climate ductless multisplit on a 2.5-4 tons of compressor such as a Fujitsu AOU30RLXEH, or Mitsubishi MXZ-8C48NAHZ can probably deliver enough heat for your whole house load, but whether that really works in your house depends on the floor plan.
At 8 zones it would be pushing GreenSpeed pricing, but at 3-5 zones it will be quite a bit cheaper.

I had the folks in Marthas Vineyard run an aggressive Manual-J and explore ductless & mini-ducted minisplit options. The eventually went for a pair of 3 zone 2 ton ductless Fujitsu multi-splits (6 zones total), which came in at $15K before MassSave rebates. After rebates it was on the order of $10-12K USD. This was for a ~3000' house, 99% outside design temp of +12F. A pair of those units would have enough capacity at 0F to cover your loads too.

Single zone mini-splits tend to be higher efficiency, and better low temp capacity and don't necessarily cost more to install than multi-splits, but at some number of zones you run out of space for the outdoor units. Some combination of mini-ducted and ductless mini-splits will often work for keeping it down to 3 zones in a house your size, but a lot depends on floor plan layout, and the room-by-room heat load numbers. In MA the only realistic cold climate mini-split options that are well supported are Mitsubishi (who has more than half the total market) and Fujitsu. For single zone ductless the Fujitsu xxRLS3H series have very good low-temp capacity & efficiency, as do Mitsubishi's -FHxxNA series. For ducted mini-splits it's hard to find competent installers who can do the duct designs, but it's easier to go with Fujitsu's xxRLFCD series than the competition, due to the more powerful mini-duct cassette blowers. The Marthas Vineyard people almost went with an 18RLFCD for half their house and a 3 zone multi for the other half, but couldn't convince the installer to put the ducts in the basement rather than the attic, where the duct losses would have been far more substantial. In the attic it didn't have any margin for the load after duct losses, but in the basement it would have been fine.

An electric water heater is cheaper to run than a propane water heater at MA prices, and cheaper to install too. A heat pump water heater is substantially cheaper to run than a dumb electric water heater, but for somebody interested in 20 minute showers it needs to be 80 gallons or bigger (for either a heat pump water heater or an electric tank.) A 300' basement is big enough to handle the air flow requirements of an 80 gallon heat pump water heater such as an ACCELERA 300. An 80 gallon dumb electric such as Westinghouse 2X045 plus a drainwater heat exchanger would be almost as efficient for long-showering families, and deliver longer shower times, longer lifespan, and might be cheaper to install overall.