I am a geothermal heating contractor and we performed a Manual J on the home. Home has roughly 36,000 cubic feet of volume when empty and 3,540 sq feet. The Manual J, based on average leakage for a two story home shows a CFM50 of around 2100-2200.

The problem lies in the fact that the home was tested recently and has a 4,400 CFM50, the builder and insulator sealed some leaky can lights and brought the CFM50 down to 4,020 which gives me 23K BTU of infiltration for heating and my manual J has 11,505 figured in. We sized the system at a 3 ton(actual capacity is 42,500 BTU's of unit). Homeowner is mad that the HVAC system can't heat the home, the builder is not fully up to speed on air infiltration so he assumes the HVAC is the problem as well instead of the home.

For you home performance guys, what should the leakage be on a house of this volume, built to code(actually a bit higher than code since it is a custom home) and is my Manual J number low or is the CFM50 number high?

In zone 4, under 2009 IECC, code was 7.0 ACH50, which it looks like he just squeaked under after the rework.

Now, it's 3.0 ACH50 under 2012 IECC. at 6.7 ACH50, it looks like he's way over that.

For reference, Passive House is 0.6 ACH50, I believe.

Sky,
 If I assume a 2 story home located in Portland OR 36,000 cubic ft tested when outdoor temps was 35 and indoor temps were 65(guess) I get ~ ach50 = 6.3, if I don’t correct for temp, I get 6.7 ach50. This shows a winter natural leakage of around 330 cfm.
I don’t look at many new homes, but the ones I have tested have been ~ 5 and 7 achn50 here in Connecticut.
 Depending on what code you have to deal with ACH50< 7  under IECC 2009 it would be OK.  It looks like OR is IECC 2009 still? 
 I looked up Washington St.  Energy code 2012, it specifies ACH50 < 5 so the structure clearly does not meet the code. It would have to be below ~3000 cfm50 to qualify. The one caveat is that it looks like Washington has an alternative compliance checklist which might be why the builder got away with it.

Did you also  measure duct leakage? If you have ducts in unconditioned space that can be a huge contributor to leakage AND energy loss. I would also want to have the number for total duct leakage and leakage to the outside. If you have a sensitive manometer, you might also want to measure what the relative pressures in the different rooms are with the doors closed. When the room pressures are more than 3-4 pa that can drive air leakage quite a bit.
 
If you want a very crude ballpark, I  figure that every 100 cfm reduction would take 2 man hours and 12$ in materials for guys that know what they are doing. Since they got a 10% reduction just sealing can lights, there may be some other big leaks somewhere that are easily correctable. My guess is that you are going to need a crew of 2 to work for a couple of days with a blower door and a thermal imager working hard to reduce air leakage to get it under 5 achn
I don’t know what is the best solution, but I would start with that Holes that are very high in the structure and or very low in the structure drive the most leakage through the stack effect. If the house is exposed  on a hill top it can exacerbate that also.
Good Luck,

Cheers,
Eric
Edited because all those western states look alike!

Perhaps a good reminder that a Manual J only works if the house is actually built to the assumptions. Even worse, ACH50 doesn't say much about ACH-normal (what actually matters; a tracer gas test is better that a blower door). As Eric says, leaky ducts or pressurized rooms (a supply duct in a room with no return) will change the numbers significantly. And it's worse on the coldest days (when the ducts are used the most).

IMO there is something to be said for testing vs estimating heat loss. Many contractors seem to add a big fudge factor (more money for them) and then swear that their Manual J is perfectly accurate.

The ducts were actually performing amazing, they allowed 400 CFM leakage and were 200 or less.
Sadly I don't know nearly as much as I would like to about the air leakage tests, but I do know quiet a bit. Being an HVAC contractor we only need to know to an extent since we don't fix any air leaks and we don't do any sealing I just need to know so that my calculations can be correct.

Since its a new home the leakage is the only thing we have to guess since the house can't be tested until its complete.
My BPI certified guy is meeting the homeowner and the builder and the insulator/air sealing guy next week since I will be out of town. My BPI guy knows a little about manual J and a lot about air sealing and testing and I know a lot about manual J and D but little about air testing and sealing. Im just wanting to make sure we don't throw another contractor under the bus, but I already know the builder is blaming our system and wanting us to put in a 4 ton because the 3 ton won't keep the home comfortable and I am telling them that we size based on industry standards and if the home is leaking more than industry standards that is not our fault or the systems fault because its the only number we have to make an assumption on(except ductwork but that is half of what we assumed).

For our meeting i just want to have some ammunition to show the builder in the codes stating Air changes per hour. The builder also told us that "he needs to know extra requirements that geo needs and propane does not" I keep telling him that propane and geo have the same requirements and we would have put in a 60K propane furnace and had the same issues of the 60K furnace being almost undersized due to air leakage. I just want to know who is in the wrong here because we are all spending money on the issue and we feel our equipment is sized properly given the information we were given and the information we asked for(we ask on every home for an average sealed home, window U-vales and SHGC, wall insulation values, floor insulation, attic insulation and floor plans so we can find out if anything is needed for the manual J). The insulator is saying he insulated to code and that his leakage is not high based on builder spec and the builder says the heating system is not the right size since it has troubles heating the home without using backup heat.

Sounds like you sized for an average new house and this one is a code minimum house (at least air infiltration wise). So responsibility depends on contracts and wording.

I guess I need to clarify more with this builder because they told me "this is a custom house and we want a custom system and we are putting in above code windows and higher R-value walls and spray foaming basement walls" the only thing we did not fully discuss was air infiltration.

On my end of things I would assume that Average would be code built and not that loose is considered code built today. I would expect loose to be 1980's homes and earlier. To me a manual J is an industry standard and using "average" for air leakage would be building an industry standard home. Of all the classes I have been to on the manual J and manual D this has been the one thing that every class has skipped over or only covered briefly because they all recomend using "average" for all new homes and semi-tight for spray foam/sealed homes and tight should only be used on verified tight homes. Semi loose from 1990's til 2009ish and loose anything before 1980's

That's an insane amount leakage for a new-construction house that size in my neighborhood, and if they'd spent even 2% of what they spent in better-than-code windows and foaming the foundation on paying attention to air-sealing as it came together they'd be under 1000cfm/50.

A 2500 cfm/50 number would probably be a typical pre-retrofit air sealing for QUALITY new construction 3500' houses with simpler framing in my neighborhood. A tract home McMansion that size might break 4000cfm/50 unless the contractor had been beat up by the inspectors about it too many times and started to at least pretend to care.

On my machine here at work I have a plot emailed to me by a who blower-door tests homes for friends as a retirement hobby, and tracks air sealing progress for them. The leakage levels are plotted against the age (not size) of the houses. While there are many pre-1975 homes with initially tested leakage over 4000cfm in his sample set, only one circa 1975 house came anywhere near that number, and was just shy of 4000cfm/50. (After a deep energy retrofit project that house came in at about 600cfm/50 according to his notes.) The sample size is small (only 8 post 1975 houses, the newest of which was built in 2001- the others date back as far as 1765). The range of leakage of the 8 non-antique homes is from ~1800cfm/50 to about 3500cfm/50, with the exception of the nearly 4000cfm/50 outlier already noted. The median of the set of 8 about 2200cfm/50. That jives with my less-than formal non-data-collecting experience as well.

The IRC 2009 requirement for <7ACH/50 is more like a stripe on the floor than a hurdle. Homes that are sheathed with plywood/OSB and wrapped in housewrap regularly come under 4ACH/50 without even trying. The ones that fail are those where the contractor skipped the interior side wallboard behind bathroom tub surrounds or failed to put air-barriers on soffited ceilings, or built in thermal bypasses in the joists for the bonus-room etc., some of which are readily retrofit-fixable, others a lot messier. The IRC 2012 spec of 3ACH/50 isn't really much of a challenge to retrofit in most homes, but big houses with lots of complicated framing bump-outs, kneewalled dormer rooms, etc. can sometimes be challenging.

The builder has to be at least somewhat versed in the discipline of air sealing to come in under 500cfm/50 on a house that size, but it's been done without resorting to PassiveHouse builder levels of air-sealing obsession.

I agree with Eric- this house probably has at least a few glaringly large air leaks accounting for half of the total. It will likely take both blower door and infra-red imaging to find it all, but that's the daily bread of folks in the air-sealing biz. If you're lucky it can be fixed by one sheet of wallboard or something (not likely) but don't hold your breath. If the house is leaking due to super-complex framing that only a junior-architect could love with a dozen bump-outs, 17 outside corners, at 7 hipped roof dormers it could be a "death by 1000 cuts" scenario, and take forever to fix.

Sky,
 I am not trying to throw you or anyone under the bus. We are at a place where the existing system is not doing what it is advertised to do. Manual J tends to overestimate heat loss by ~20% so it should provide a bit of a cushion from the start.
If you have a 3 ton system- assuming you have around 400 cfm/ton of air, 200 cfm of leakage is 200/ (400*3) or ~ 17% of the air in the system not coming from or going where you want. If it is leaking inside the heated portion of the house, not so bad for energy usage, if it is going outside the heated portion or coming from outside the heated portion, this is heat loss. If you are starting with lower airflow rates, the % of leakage is magnified.

Here would be my thought process on tracking down what is going on.
First, I would want to make damn sure that the system is actually producing the amount of heat we think it is producing. To do that you need to confirm what is the actual airflow at the air handler and that it is appropriate for the system. You also need to measure the temp rise across the heat exchanger. Now you know how many btu’s you are producing.
Next you need to measure air flow at the registers as well as the delivered air temperature at the register. Sum all these up and from this you can calculate what percentage of air is being delivered and how many btu’s are in that air. With these 2 measurements you can figure out how many btu’s are being delivered to the house and calculate supply distribution system losses.
 Do the same thing with the return system.
If you measure the return air temp of the duct system a lot lower than the average temp of the house you know you have a problem. Next figure out if the house has mechanical ventilation, and account for any heat loss associated with that.

Now you have the heating system part of the equation.
The second ½ of the puzzle is the shell, which we already covered. Make sure that the windows that are speced are actually installed and done correctly . A quick run around the house with a thermal imager may turn up a few surprises and is worth doing. You might be able to have the bpi auditor generate the heat loss calculations independently and compare notes with him. That may show up something you missed.
Before I put in a larger system, I would work hard to reduce duct losses and tighten up the shell- I bet it would be more cost effective.
Good Luck,
 Eric

Eric, the leakage of air is at the CFM50 rate so more than double the standard pressure through a duct system, in the end is equates to roughly 3% of the ductwork airflow is leaking and on a 1,200 CFM system thats only 36 CFM. As I said I am not very good at how the numbers are concluded just what I am told they mean and my BPI guy said it equated to a 3% loss and 6% here is code. Even the insulator/home performance guy said they wrere performing exceptionaly.

We have verified system performance, this is a geo so we need to perform a heat of extraction calculation with waterflow etc. We are getting 35 Degree EWT's and a total heating capacity of around 42,000 BTU's on a 3 ton design heat pump system. We have also done airflow testing and I would have to look up the numbers but we were getting OVER 375 CFM per ton of verified airflow out of the system and were probably in the 400 CFM per ton range or higher but I can't remember.

Based on what Dana1 is saying this house is super leaky and I just wanted to get some home performance guys opinions that deal with air leakage more than me to know if this was normal for a "custom home" or not and to make sure I correct an errors in the Manual J in the future. Its a frustrating situation all the way around and ultimately the customer is who we want to make happy but im not going to put in a 4 ton system and loop when the house is not performing nor am I going to spend money fixing my duct work and system when its not the problem and has been third party verified that it is working to specification and when a manual J has been performed to industry standard and the best of our knowledge with specifications from the builder.

im not going to put in a 4 ton system and loop when the house is not performing

We need more contractors that will hold the line on things like this.

Sky,
Just for your knowledge, the duct leakage rates for compliance in iecc 2009 are a maximum of 6 cfm @ 25 pa of duct leakage PER 100 square feet of heated space so for a 3500 sf house, you are allowed 6* 35 or 210 cfm of leakage.
This does not mean that only 6% of the air is leaking out of the duct system.
These tests are usually performed at 25 PA of pressure (0.1” WC), not 50 pa . Most of the forced air systems I test have external static pressure of 0.5” WC or more (125 pa) so the average pressure in the ducts are 0.2+" WC (50 pa) or greater. This indicates that the duct leakage value you get from the duct blaster UNDER estimates how much leakage you are getting. Again, it matters a good bit where the leakage is going, inside or outside.
I have never tested a geo system though so I don't know what kind of pressures those systems run at.

On the whole, I agree with you that the shell is likely the majority of the problem, but I might want to appear to be contributing to the solution by reducing duct leakage as much as possible.
Cheers,
Eric

This is a good post in that it shows what happens when a typical builder, who does not care a whit about "air sealing" or any of this weirdo "sustainable" stuff meets a homeowner who demands a house that works as it should. It's time for the majority of builders to start paying attention. All the references to "code" compliance and "normal" construction ignores the fact that the codes and "normal" are decades out of whack. As tighter houses come into the public eye, and people finally realize they can be comfortable without spending a fortune on oil or gas, builders will be forced to adopt modern building methods.
This house can and should work a lot better; I'm hoping the builder is made to repair it and realizes that he can easily change his practices. For the record, my houses (generally smaller than this one) are coming in around 400 CFM50, so the 4400 number is, to me, absurd. This builder thinks that upgrading materials will solve his problem; it won't; he has to change the way he builds.

The other thing that seems to be happening is that the leakage testers care more about getting a check than they do about the results. I've been told more than once over the phone "we can make it come out the way you want".

there are dishonest people in every field.

Posted By SkyHeating on 26 Mar 2014 07:52 PM
Eric, the leakage of air is at the CFM50 rate so more than double the standard pressure through a duct system, in the end is equates to roughly 3% of the ductwork airflow is leaking and on a 1,200 CFM system thats only 36 CFM. As I said I am not very good at how the numbers are concluded just what I am told they mean and my BPI guy said it equated to a 3% loss and 6% here is code. Even the insulator/home performance guy said they wrere performing exceptionaly.

We have verified system performance, this is a geo so we need to perform a heat of extraction calculation with waterflow etc. We are getting 35 Degree EWT's and a total heating capacity of around 42,000 BTU's on a 3 ton design heat pump system. We have also done airflow testing and I would have to look up the numbers but we were getting OVER 375 CFM per ton of verified airflow out of the system and were probably in the 400 CFM per ton range or higher but I can't remember.

Based on what Dana1 is saying this house is super leaky and I just wanted to get some home performance guys opinions that deal with air leakage more than me to know if this was normal for a "custom home" or not and to make sure I correct an errors in the Manual J in the future. Its a frustrating situation all the way around and ultimately the customer is who we want to make happy but im not going to put in a 4 ton system and loop when the house is not performing nor am I going to spend money fixing my duct work and system when its not the problem and has been third party verified that it is working to specification and when a manual J has been performed to industry standard and the best of our knowledge with specifications from the builder.

I dunno about "super leaky"-  one of circa 1850 antiques my friend tested in Georgetown, ME came in at a whopping 9000cfm/50 before squirting some cellulose into the wall cavities, and was still ~4200 cfm after insulating.  It's a work in progress.

As it happens I spent a weekend in that house last year. It has a rough-quarried granite foundation (grouted seams but not mortared) and a rough hewn beam for a foundation sill (grouted in place), with wide-plank heavy pine flooring between the living space and the dirt-floored cellar, and most of the original large original 8/8 light sash single pane double hungs (in various condition).  I expect about half the air leakage is the foundation, and maybe 1/3 is the windows, but since I wasn't there when it was tested that's just a WAG.  With judicious use of spray foam at the foundation sill and maybe some housewrap on the underside of the floor joists sealed to the foundation sill it would probably duck under 3000cfm/50 (tbd).

He also lists a 1780s house in Harvard MA listed that started out at ~4100cfm/50 , and dropped to ~3500cfm/50 by tightening up the windows. There are clearly more holes left to deal with there.  I don't know any of the construction details of that house, but typical homes of that vintage in that town have fieldstone or quarried foundations, and clapboard siding with or without plank sheathing. I'm not sure what is/isn't reasonable for a retrofit on that one.

But these are the kinds of houses that hit 4000cfm/50. The only way to get there in a house sheathed with 4x8' OSB/ply on one side and sheet gypsum on the other, and brand new better than code windows is to have sections of sheathing/gypsum missing &/or a raft of other intermediate-large-ish holes (like the recessed lighting they addressed), or unsealed plumbing/flue/electric/duct chases that go from the leaky basement to a vented attic, etc. Air sealing the house to sub-2000cfm/50 will do a lot more for comfort (and air quality!) than up-sizing the HVAC by a ton or two to make up for the infiltration factor.

This has been an interesting discussion of the technical aspects of how Manual J calcs and air-sealing interact. We have different subs pointing the finger at each other. The consensus seems to boil down to the builder has built a leaky shell. Well,the builder apparently hasn't been able to offer support on why his shell may not be leaky. Ultimately the builder has to convince the homeowner as to the validity of his techniques. Just for the edification of myself and other Newbies, what are the appropriate techniques used to achieve a solid ACH50 under 2? I chose 2 since the 2013 standard code is for under 3; PassivHaus is 0.6, and 2 is about halfway. There must be a few basic items to check off; rim joists, window caulking, locations around fireplace, etc. What else? And how does one establish that a good job has been done? I called for a ACH50 test under 2 in my contract; after the builder signed he asked me how to obtain it. I told him to educate himself, offered several building science sites, and said; you signed the contract, figure it out. So as green builders, what are the 5 or 10 most basic steps to take?

The single most important thing to do is decide where and what the air barrier is going to be. It can be on the interior side ie air tight drywall approach- it can be at the sheathing layer. Ideally it is both for a bit of redundancy.
When you switch from wall to ceiling- the air barrier must remain intact across the boundary. Same when you go from foundation wall to sill plate. Everyplace the barrier is breached- there needs to be a strategy to seal the breach

Big leaks usually include Attic Hatch, gaps around chimney (especially masonry), fireplace dampers, between sill plate and concrete, plumbing chases, electrical penetrations, recessed lights, where interior walls intersect the ceiling plane, vents for the bathroom and range with poor dampers, ductwork in the unconditioned attic, bilco door in the basement, kneewalls and vented cathedral ceiling, oil fillup pipe.

I think the KEY to being sure is using a method where the air barrier testing can be done prior to insulation, or at least the drywall. Once the insulation is in, it is very hard to track down the small leaks that all add up.
Zip sheathing+ zip tape makes it easy to establish an air barrier at the sheathing layer.
 My airsealing box contains: tapes- siga rissan, foilmastic tape, zip tape, aluminum flex duct tape, aluminum duct tape; sealants- silicone caulk, fireblock caulk, mastic; one part foams- standard, low expanding(windows and doors), and fireblock. I use ridged materials- foam and coil stock to cover big holes and foam gaskets for things that open and close.

Cheers,
Eric

Eric's list is excellent. Probably the most important thing for the builder to do is recognize that he has to pay attention; air sealing has traditionally never been important to framers; that needs to change. And he needs to spend $40 for a foam gun.

Yep. Good quality caulk gun. Foam gun.
Buy your 6 or 9 cases of caulk and case of foam cans up front or it will kill you to buy them one at a time.
Verify the sealing yourself. Over and over again.
Don't count on the contractor. Even if he wants to, his guys simply can't do it.