Solutions for Combustion Safety in Existing Homes


  • Larry Brand, Research and Development Manager at the Gas Technology Institute and the Partnership for Advanced Residential Retrofit (PARR)
  • David Bohac, Director of Research with the Center for Energy and Environment

Nicole: Hello everyone! I am Nicole Harrison with the National Renewable Energy Laboratory, and I’d like to welcome you to today’s webinar hosted by the Building America program. We are excited to have Larry Brand and David Bohac here today to discuss both current guidelines and a field study supporting revised protocols for combustion safety testing in low-rise residential buildings. 

Before we begin, I’ll quickly go over some of the webinar features. For audio, you have two options. You may either listen through your computer or telephone. If you choose to listen through your computer, please select the “mic and speakers” option in the audio pane. By doing so, we will eliminate the possibility of feedback and echo. If you select the telephone option, a box on the right side will display the telephone number and audio PIN you should use to dial in. Panelists, we ask that you please mute your audio device while you are not presenting. If you have technical difficulties with the webinar, you may contact the GoToWebinar Help Desk for assistance.

If you would like to ask a question, please use the “Questions” pane to type in your question. Please indicate which speaker you are addressing your question to.

If you are having difficulty viewing the materials through the webinar portal, you may find PDF copies of the presentation at the website listed here and you may follow along as our speakers present. Today’s webinar is being recorded and the recording will be available on the DOE YouTube channel within a few weeks.

We have an exciting program prepared for you today. It is often challenging for residential contractors to conduct appropriate combustion safety test procedures in existing homes and use effective actions to address any failures. Commonly used testing methods can be complicated to follow and may be overly conservative, producing false positive results without addressing the fundamental need for inspection of the venting system itself. Additionally, there are situations where comprehensive combustion safety tests are not necessary and prescriptive procedures can be used to address safety concerns. Methods to easily identify those situations and conduct the required actions may reduce test costs and provide a higher level of confidence that contractors are following the proper procedures to protect against combustion safety failures. 

To assist in addressing these concerns and opportunities, the Building America Program is hosting this webinar that will discuss both current guidelines and a field study supporting revised protocols for combustion safety testing in low-rise residential buildings. 

Before our speakers begin, I will provide a short overview of the Building America program. Following the presentations, we will have a Question and Answer session, closing remarks, and a brief survey.

The U.S. Department of Energy’s Building America program has been a source of innovations in residential building energy performance, durability, quality, affordability, and comfort for 20 years. This world-class research program partners with industry to bring cutting-edge innovations and resources to market. Visit our website at to find more information about the program and the Building America Solution Center, which provides expert information on hundreds of high-performance construction topics, including air sealing and insulation, HVAC components, windows, indoor air quality, and much more.

Building America is supported by five industry research teams and four national labs. Each of these teams and labs partner with dozens of industry professionals, including builders, remodelers, manufacturers, and utilities. The best and the brightest in the residential buildings industry can be found here.

And now, on to today’s presentation!

Our webinar today is Solutions for Combustion Safety in Existing Homes.

If you would like more detailed information about this effort, please feel free to contact our speakers.

Our first speaker today is Larry Brand. Larry is a research and development manager at the Gas Technology Institute and the Partnership for Advanced Residential Retrofit (PARR) working in the areas of building energy efficiency, product development, codes and standards, and technology deployment. He has 30 years of experience managing building systems research and product development programs for private and public companies, and has launched over 20 products into the marketplace. Larry is currently the lead investigator for the PARR Building America team and principal investigator for the California Energy Commission Measured Home Performance project.

Also presenting today is David Bohac, the Director of Research with the Center for Energy and Environment. He has over 30 years of experience in energy efficiency and indoor air quality in commercial and residential buildings. He was the manager of the CEE’s Indoor Air Quality group for the Minneapolis-St. Paul Residential Sound Insulation Program that performed combustion safety tests, air leakage tests, insulation/air sealing quality control inspections, and remediation of over 7,500 single family homes and hundreds of multifamily units. His energy research includes measurements of commercial building envelope and duct sealing, aerosol envelope sealing and ventilation improvements for multifamily units, the efficiency of tankless water heaters and combination systems in houses, and effectiveness of combustion safety protocols. His indoor air quality research includes assessments of secondhand smoke exposure in cars, apartments, outdoor patios, and bars & restaurants, the transfer of secondhand smoke in multifamily buildings, the transfer of contaminants from residential garages, and strategies to reduce the transfer of airborne infectious diseases in hospitals.

With that, I’d like to hand it over to Larry and welcome him to start the presentation.

Larry: Thanks.  Let me get you a full screen view of this, there we go.  So Dave and I are excited and glad to be a part of the webinar today, before we kick it off, we are going to talk about three main topics, and before we get into that, I wanted to acknowledge some of the great work being done by several other organizations and their work in combustion safety, so, Paul Francisco for Illinois Sustainable Technology Center (ISTC) at the university, Dan Cautley at Seventhwave is also a part of the work we are talking about today.  So Paul and Dan as well, Brett Singer and his team at LBNL are doing some great work in this area and we have presented with them in the past, of CEE – Jim Fitzgerald, the Center for Energy Environment.  We may hear from Jim today as well as some of our partners with the American Gas Association, AHRI, Building Performances and RESNET, so you know this is a very important topic for the entire industry, and there are some pretty good minds trying to address and resolve some of the issues with combustion safety. 

So, today this is what we are going to do, I’ll give you a little bit of an introduction.  There were two measure guidelines that were developed for Building America, one that uses indoor air and one that uses outdoor air.  I was encouraged not to use the term confined space, because I think that means something else now in the building inspection business.  So, I’ll use those for now.  Then we are going to talk about recent research, so Combustion Safety Simplified Test Procedure that Dave and I and others have been working on for Building America, some survey work that was done by Paul that I mentioned, some short term testing and monitoring, then we’ll look forward a little bit and talk about harmonization and the various codes that are out there, then we’ll do Q&A.  Dave and I will pass it back and forth as we go through our own topics here. 

So what’s combustion safety? There’s a lot of disconnect about what is it and what isn’t it.  And in general, it refers to natural draft appliances creating a draft in the vent, within a short period after ignition, so it relates to excessive spillage in general in the community. This is my definition.  Things tend to go back to spillage, measurement of spillage, rather than the more serious issues related to back drafting, then you know there is a time associated with allowable spillage, so that’s what we’re focusing on here.  It does apply to common vented and fan assisted, both category one appliances, natural draft venting with V vents. Only in the perspective that there should be no flow through gasses or gasses from one appliances to the other when they are common vented.  That’s a critical part of it.  Good practice for combustion safety is that the appliances are properly installed, and they are operating correctly.  There is sufficient air for combustion and dilution, so we will get into some of that today.  The vent is properly sized and installed.  I can’t underline that enough.  We see a lot of cases where that is not the case.  And then appliances that are operating at CO levels that are within safety certification limits.  I’ll talk about that a little bit as well.

We start with the code.  The combustion air requirements and vent sizing tables are all the same in the model fuel gas code and sizing code.  So here’s the national fuel gas code, 2015 edition.  There is also an international fuel gas code and there are other model codes.  And they all kind of reflect the same set of requirements for combustion air requirements and vents and vent sizing, so I would encourage you to take a look at those.  The fuel gas code here has been an ANSI standard since 1974, all the category 1 appliances have been installed according to the tables in the fuel gas code.  All the other categories I believe have their own installation instructions, so fuel gas code talks about category 1 only or natural draft appliances.  There was a 2015 update to Annex G, if you are familiar with the code, that was I believe a major change since the last version which was I believe was 2012. At any rate, the code is enforceable when it is adopted by the authority having jurisdiction, so the AHJ references the code, it becomes sort of the law of the land wherever that jurisdiction is. So I always start with the fuel gas code and I encourage you to take a look at it.

Getting into that a bit, here is category 1 natural gas appliances, fan assisted draft hood equipped, negative vent pressure, not condensing.  The label on the appliance will show its category, so if you are not sure if this is a category 1 or a category 2 or something, look for a cat 1 on the label and that tells you what you are working with.  Water heater with draft hoods are not categorized, maybe (inaudible 11:15) is on the line, he can bring us up to speed on that, but I believe that’s the case.  They are considered category 1 for combustion and safety, so they do follow the tables that you’ll see in the national fuel and gas code.

Proper installation of the equipment is important as well, so clearances to combustible materials, I want to make sure that we cover this just briefly, that’s important for safety. The combustion air requirements are met and that the appliances are tested according to the test procedures.  I think today we are going to focus on 2 and 3 in that.  These are the two building America measure guidelines that are available through the building America website, and you can also contact us and we can share these with you. Combustion safety for natural appliances using indoor air, and combustion safety through client zone ventilation, that we will be summarizing these two today. 

Let me go into the indoor air measure guideline and Dave will talk about the zone isolation.  Indoor air as I mentioned, this is section 9.1.3, and it is air combustion and fuel ventilation in the gas code, so the air is for combustion air ventilation and dilution, in accordance with the manufacturer’s instructions.  It is very important to the combustion safety in the venting process to get the air correct for the appliance.

So, there are some requirements for indoor combustion air.  They are using the required volume method.  SO there’s two parts to this.  Part one is if you know the air exchange rate and the air tightness of the home to be .4 ACH, that’s based on natural, that’s not ACH 50 number, then you can use the standard method here, and the standard says that the volume of the space – this is for indoor air combustion so – the volume of the space has to be 50 cubic foot per 1000 Btu/hr of the appliance.  So for example here if you have 120,000 Btu per hour appliance, category 1, so it could be fan assisted or a natural draft, you need 50 times 120 or 6000 cubic feet of volume that’s communicating freely with the appliance.  And that’s about a 750sq ft area with if you have an 8 ft tall ceiling in that home.  If you have more than .4 ACH and .4 ACH can relate to – in an ACH 50 term, I believe it’s around 8 ACH 50 so that’s a fairly leaky house by today’s standards. 

So this result part 1 is 750 square feet.  Part 2, if you know that you’re air infiltration rate is less than .4 ACH natural you can use this known air infiltration rate method.  (Inaudible 14:34) code, other than fan assisted you use 21 cubic foot per ACH if you have a fan assisted furnace for example, you use 15 cubic feet.  So here’s my example again.  Let’s use .3 Air Changes per Hour on a natural basis, 120,000 natural draft appliance, you would need 8,400 cubic feet for 1050 for a natural draft, with a fan assisted you would need 6,000 cubic feet of volume freely communicating with the appliance, or 750 square feet with 8 foot ceilings, so that’s back to sort of the same number we were getting in part 1. So these are the two main parts of the code on air infiltration, or air, so that’s used by the appliances. 

Then I want to mention CO’s within limits.  This is a table from the 2012 standard, but I believe the numbers are the same in the 2016.  There’s a more exhaustive list. And these are the CO air free numbers in the appliance.  With a central furnace I’ll categorize about ANSI certification standard requires 400 ppm air free of CO or better.  So, that’s what these are.  These are following the appliance certification standards.  You can look down the list here to wall furnaces and gravity furnaces and water heaters. And this also shows you how to do the calculation for air free CO, so that dilution is taken out of the calculation. So, make sure the CO is within the limits for the appliance.

And then vent sizing is the other part of this that I wanted to mention.  It applies to both zone isolation and using indoor air for combustion.  And this is my opinion on the top.  Most combustion safety related incidents are related to poor sizing of the vent. And you see a lot of that on there. So the venting tables on the code have been around since the 50’s, for natural draft appliances, and they were updated in the 80’s, for fan system appliances, so all of us were a part of that.  And they are good.  The vent sizing tables have been time tested and they will address most problems so I kind of encourage people who are having combustion safety question or see a house that has an issue to start with looking at the venting tables in the code.

Here is an example of how to use the venting tables in the code, I won’t go through this too much, so you have a furnace and a water heater with various capacities, they are common vented, you go to a table like this that says, this is the vent connector capacity for type B double wall connector with two or more appliances. You start on the left with the vent height, you work your way across and it shows the capacity for a natural draft appliance, which is a maximum capacity, you look up your appliance for a fan assisted furnace in this case, you look for that range.  Then this tells you that you have a 4” double wall connector for the water heater and a 5” double wall connector for the furnace in this particular case.  It’s fairly simple, but I encourage people to take a look at these connector sizing tables and some of the other information available on the lateral sets on the code when you get a situation and you’re not sure why it’s not working correctly.  I have some pictures later on that I’ll show you of some installations that aren’t maybe according to the code.

So that’s my set on indoor air for combustion, and Dave is going to talk about the zone isolation option. 

David: Ok, thanks Larry.  My understanding is that the reason this came about was that there was an expert meeting um for Building America, quite a bit of discussion about the concerns with the ability of contractors to properly do the combustion safety testing under worst case conditions and being able to set up worst case, so I tried to identify situations that would be pretty straightforward for contractors and something that they could implement, and maybe without uncertainties of testing.  So the one situation that was identified was where the furnace water heater were in an isolated zone all by themselves away from the house so this measure guideline focuses on that. Just wanted to mention too, as Larry mentioned, Jim Fitzgerald who is the co-author on this is here with me today and for those of you who know Jim, it is very likely that he will be chiming in along the way. 

So Larry if you want to go ahead on to the next slide.  These are natural draft appliances located in an enclosed space, typical locations, mechanical closets, off of the living spaces, attached garages, attics.  And what we’re looking at doing is providing a proper air barrier to make sure that those appliances are truly isolated from house and house depressurization, and really providing combustion air from the outside to make sure that they have sufficient air.  What the expectation is that this would then eliminate the need for doing any depressurization combustion spillage testing.  So if you can make sure that these kind of prescriptive measures are complied with then you don’t have to worry about depressurization in the house.

So, the five steps that are involved with this are: number 1, you know Larry finished off his discussion with talks about his combustion venting system. Make sure that the combustion venting system is sized properly according to codes, I’ll just echo that.  We tested thousands of houses with our sound insulation program and what we found was of the houses that were spilling under worst case conditions that, even if you took away worst case conditions entirely, still half of those were failing in spillage or under what we would call natural conditions.  And that if we just had those venting systems modified so that they met code, that spillage went away. Number 1 make sure that the combustion venting system complies with code. Number 2, make sure that you have a good air seal for that physical boundary, you need to make sure everything is sealed well.  Number 3, ducts and cabinets – air handler cabinets – are in that isolated zone, you want to make sure that they are sealed.  Any leakage to the return duct and the return side of the blower is going to pull in air from that space, depressurizing that space.  And then also if there is any spillage, any combustion spillage, its then going to get pulled in to that air handler and distributed to the house.  You need to make sure that ducts and cabinets are sealed properly. And then make sure that we have number 4, properly sized combustion air openings, and then number 5, although it’s not really recommended, you really don’t want to have exhaust appliances in that space, if they do, make sure that you have adequate makeup air to deal with those. 

So Larry already covered the issue about looking up or using the code tables for combustion venting so I won’t talk any more about that. Sealing the enclosure, you really need to make sure things are airtight.  First of all, if there’s any large openings, use sheetrock to seal those up, and then any other small gaps, seams, penetrations, between the living space and the zone, kinda use your normal sealants.  Joint tape, sealant foam, make sure it’s listed as a firestop and for use on uncovered use or sealant caulk.  And then very often, especially mechanical closets will have a door to the house, you want to make sure that that’s weather-stripped, has a nice quality seal on it, and really you’d like that to be self-closing and latching, so that there’s not an opportunity for people to be able to leave that closet door open.  Sometimes we’ve seen those doors with louvered doors, and certainly that’s not something you want to do to try to isolate that space so either replace that or that you block off all the louvers. 

Number 3, the duct and cabinet sealing. Again, you need to seal all the gaps and so forth.  The straightforward thing is that you want to make sure that you have a continuous sealed duct from the living space to the blower housing, in general what you’d like to see is that that duct area is the same as the area of the return inlet, or at least 2 square inches per 1000 Btu of output, and that’s recommendations that have come up from Proctor and Wilcox in their work.  Finally you also need to have that mechanically fastened at the duct to the blower housing.  You want to tape all the cabinet service openings and joints of that air handler to the furnace.  And you are going to be doing that with Mastic and mesh tape.  UL181 compliance or foil tapes UL181 A-P or B-FX compliance and then of course you don’t want any return air from that space that’s gonna depressurize that or cause you to bring air in from that space and distribute it to the house. 

So as far as combustion air from the outside, again follow the code.  It works.  If you have two openings, you want the one at the top to be within 12” from the top and the one at the bottom to be within 12” of the bottom, the diagram on the left shows when you have either direct or vertical ducts, to the outside from the space, in that case you need an opening of 1 sq inches per 4000 Btu/hr input for the appliances that are in the space.  If you have those going horizontal, then you need more so you need 1 sq inch per 2000 btu/hr input.

And the other option there, you can go ahead and get by with one opening, that should be within 12” of the top, and now you need a sq inch per 3,000 btu/hr input, but it also needs to be not less than sum of the area of all the vent connectors.  When you’re looking at this you need to make sure that you consider the effects of any louvers, screens or grills, use the manufacturer guidelines for how to de-rate that opening.  If there isn’t any label, then guidance is assume 75% for metal louvers, and 25% for wood louvers. 

So makeup air and exhaust fans, again, the best thing to do is to make sure you don’t have any exhaust fans or dryers in that space if you do have a dryer, the code recommendation is for 100 sq inches of opening.  It is allowable to have a damper on that opening but it must be interlocked with the dryer.  If you got a situation where that equipment is up in the attic, you need to make sure that you either remove or disconnect any powered attic fans, so that they’re not going to be causing any sort of depressurization, or you need to follow the manufacturer recommendations for the amount of air inlet openings that you would need to properly compensate for the amount of air that would be needed for that fan.

So, that’s kind of the way of coming up with your work scope for an isolated space you’ll get that work completed and these are the measures that you would need to do to confirm that everything has been done appropriately.  So first a visual inspection of again, those gaps and leaks in the air handler and ductwork, make sure that those have been properly sealed.  As a follow on to that you know to really ensure that that’s been done properly you’d want to smoke test the duct and cabinet leakage, and that involves taking a fan and pressurizing that cabinet and duct work that are in that space, up to 25 pascals. Bring theatrical fog into the system and then look at that ductwork and cabinets and see if you have any fogs that are leaving any of those spaces.  If you see anything leaving there, you need to seal it up.  And then finally, you want to make sure that your air sealing of that enclosure has been done properly.  Our recommendation is that when you have the house pressurized to 50 pascals, you’d like to see that isolated space be depressurized by no more than 5 pascals, so 1/10th of what you would get in the house.  So it’s not 100% isolation, but we feel that this is a good guideline or a good rule of thumb that will provide adequate isolation in almost all cases.  You also want to take a look at the HVAC operation and make sure that it doesn’t depressurize that space by more than 1 pascal when it operates. If you do have a dryer in the space, you want to make sure that doesn’t depressurize the space by more than 2 and a half pascals. And same thing if you are in a garage, and there is an exhaust fan in the garage for ventilation, you want to make sure that that doesn’t depressurize by more than 2 and a half pascals. 

So I just want to note, that in the measure guideline there’s case studies, to provide a little bit more guidance in a few situations: 1 is the attic furnace, 2 is the attached garage, and 3 is mechanical closet with particular attention if you have a door to a bedroom or a bathroom and when you have a building cavity that’s being used as a return, so you can look at the measure guideline for more information on those situations. 

And I just wanted to point out too, that one of the key details that is often overlook, or is difficult to deal with, is that return platforms underneath the air handler, and there is a nice guidance on how to do that air sealing that was provided by Ray David from NREL.

So just to sum up, this is meant to be a very straightforward, very prescriptive measures to isolate that space, use inspections and isolation tests to confirm that you have the appropriate amount of isolation that’s gonna protect the natural draft appliances from house depressurization so now you don’t care how tight the house gets, or how much exhaust you have in the house, that’s gonna be isolated from your combustion appliances.  And then, it eliminates the need for depressurization combustion spillage testing.

Back to you, Larry.

Larry: Great. Thanks Dave.  We also wanted to cover some of the more recent building America research we have been doing over the past year or two in addition to the measure guidelines, so we have a section on that here. We are looking at field test procedures here and the question is setting up the worst case depressurization that you can create necessary to create combustion safety failures.  Is there a way to do more simple test procedure, and still catch the cases that you want to and get your false positives so you don’t start out an upgrade for a home by replacing appliances that may or may not be replaced. So in the past couple of years there was some activity sponsored by building America to do some research on field performance and the opportunity for doing a simplified de-pressure safety test procedure. 

Here is the scope, we started with a survey working with weatherization teams around the country to get their experience on how they approach this issue in the field, what kind of testing do they do, do they have guidelines that they work with, and there was quite a bit of great information that came out of that. And then there was a field evaluations, starting with simplified test procedure that the major components are fixed or positions, air handler, if it reduces air pressure and clothes dryer on we will have a little more simplified procedure on this later on.  And then sites were selected based on – for the field testing where there was significant data collection – the house had to fail among the must fail criteria and had to pass a must pass criteria. We needed to identify houses that were in this range that could potentially have a combustion safety failure.  The must fail had the kitchen fan operating on high, the next largest fan on, and look for continuous spillage after a set time on the time dependent on – is it heating season or not heating season.  And then there was a must pass criteria where the kitchen fan could either be on high or low. So we made it a little bit less severe.  The next largest fan could be either on or off and would require no spillage after a set time.  And that’s a time range that we’ll talk about.  That’s the overall scope of the whole program.

Here is the simplified test procedure that we all identified.  Paul, Francisco and I, Dave and Pat were the folks that were working on this together.  This is a set up the house, keep the doors open, if there is a return or exhaust in the system, otherwise we would close the doors. And that is opposed to the way BPI used to do it looking for the worst case door position.  A little simpler for the field tech. Turn on all the exhausts including the dryer.  I have a star under the exhausts, because I think in some cases you have to identify the speed of the fan for the exhaust. And then check for if the house is more negative with or without an air handler operating, and you take the most negative case, then you check with and without the combustion appliance zone door open then you take the most negative case there as well. So here are the spillage limits, beyond 2 minutes for water heaters and furnaces in heating mode, if that was a failure if it was billed past that amount of time, and 5 minutes for furnaces not in heating mode. The assumption here is that in heating mode, those appliances wouldn’t be cold, so there would likely be working with a warm water heater or a warm exchange furnace. And then of course checking the CO against the ANSI certification standards. These are the three major elements of the simplified test procedure that we were trying to collect data around. 

I’ll take you through the survey, then Dave will talk about the field data collection. On the survey, the work was done by the Illinois Sustainable Technology Center, so we were trying to identify how common the combustion and safety failures are out there in the field. There’s a lot of anecdotal information, I think there was a study in Canada done many years ago that said about 10% of homes had failures for combustion safety reasons and needed remediation.  I think it was back in the 80’s. So we are trying to figure out what does that really look like with real data. National Association for State Community Service Programs, were our partner on this and they disseminated this survey vehicle, and you can see here which states were surveyed, not all of them provided responses to the survey information, but we did get quite a few.

So there were two parts to the survey, and I don’t have a comprehensive list here, but we did ask questions about housing and the work done in the last program, how many homes were treated, how many homes used fossil fuels and how many homes had natural draft appliances within the pressure boundary to identify where depressurization could cause a concern.  Outside of the depressurization boundary, if you follow the zone isolation rules then it should be immune to depressurization related safety failures. We asked them what test procedures they were using, was it BPI or some other procedure. We were looking for granularity on the houses that failed. The test procedure they were using, did they know how many were due to air handler operation, or exhaust related including dryers.  How many cases was the appliance taken out and replaced with a new appliance.  Or, what was the use of power vent kits for remediation, and how many houses had deferred weatherization because of the issues.

Here’s the home sample size, 30,000+ houses, 94% of those were WAP: weatherization assistance low income homes, well there is a low income requirement for WAP, so a good population of those.  78% were oil gas and propane, so many were using fuels like that. One state said they didn’t track the – they had some data – but they didn’t track the program with the granularity we were looking for.  So in the end I think we had 28,000 WAP homes 30,000 homes in the program, pretty good sample. 

Six states provided estimates.  They report 41% of the houses had natural gas appliances within the pressure boundary so that’s pretty good.  There were 4 states that had precise values and better record keeping. And they said that 81% of those homes had natural gas appliances within the pressure boundary, and seven states didn’t track it at all.  The difference we believe between the six that did provide estimates and the four that didn’t provide precise values was more related to geography more than it was to whether or not they were keeping good track of their data. Anyway, so, a pretty good proportion of the homes had appliances within the pressure boundary. 

So the states that did have good data, this is survey results on failure rates. The states that did have good data, there were four of them, and 4% of the houses that were remediated because they had some expected failure in the home, and about 5% were remediated because they did do the test and observed the spillage beyond the spillage limits.  In the states that didn’t have precise data but had estimates, about 6% were remediated because of expected failures, and about 16% were remediated due to observed spillage, but that included Alaska, which estimated about 40 -50% of their homes were remediated because of that.  That’s a little bit of a tough data point, so if we take that data out, we had about 8%, so if we had good data, you were in the 4-5% range, and the states that were estimating, gave us from 6 – 8 % range of remediation.  So compared to the Canada study from the 80’s, it looks like there were fewer cases of remediation that we would have expected.  The pie chart just shows the number of data for the states that had estimates and didn’t have data. I left the no data case out of here, a lot of states didn’t track.

Also, so failure causes.  We asked to go a little deeper, so the states that did have good data, about 4% of the homes did fail due to air handler operation, which was a little bit of a surprise to us, and 4% of the homes failed because of exhaust and that exhaust could include dryers. So we were in the – we can identify what percentage could be associated with these different things.  That’s a very high number for air handler operation, it tends to take us down a different direction.  Then performance of the dryers, I know the industry has been talking about the contribution of dryers to depressurization and this would tend to support that. The states with estimates, 6 and a half failed because of air handler operation and about 18% were on the exhaust including dryers, so quite a bit higher where good data wasn’t tracked quite as well.  I guess you can make your own decisions about that.

Also compared to states providing precise data this is in general states providing estimates said that homes failed tests at a higher rate and that exhausts were more frequently the cause, so this kind of gets back to the anecdotal information that we’ve had in the past that it tends to overestimate what the failures are. The failure rate tends to point to exhaust more frequently when we saw some data here that showed it could be distribution systems issues, deck systems. Could be correct because of the small sample and geographic differences, tough to say.

Some states volunteered that most failures are due to, many failures, in addition to those major items are improper flue sizing, roof caps that are crushed for one reason or another and the air handler operation, dryer operation. Pretty interesting stuff.

So, remediation states with data 31 or 3.2 % of the homes got new appliances. In the states that had estimates, if you take out Alaska, 2.3% received new appliances and Alaska estimated about 60% so they are running their program a little differently. So those are pretty comparable. So you know we are into the 2-3% neighborhood. The power vent kits which we thought would be a really good solution in most case weren’t that commonly used out there. Only less than 2 tenths of a percent - not sure why that’s not used as frequently. It’s a less expensive option.

And 6 homes in one agency deferred the upgrade of the insulation, the weatherization, so it’s very infrequent that the weatherization doesn’t happen due to a combustion safety issue, which was a question we had at one point.

So conclusions, failure is not as common as expected. Maybe half of what was expected in 3-5% range. Not often due to exhaust fans, we did find a lot of air handler or dryer or vent failures that could be more likely root cause. Air handlers because of the leaky duct systems. Very little tracking of the info so we see a research opportunity here for the future, betting better granularity and actual working with the weatherization agencies to get good data on failure modes.

Ok and that’s the survey part and I’m going to turn it back to Dave to talk about the field study.


Ok, thanks Larry. So in addition to the survey there was also a field competent to this study. That is we want to take a look at houses that failed the simplified protocol for spillage and yet passed when they were put under less extreme depressurization conditions to look to see how often those appliances actually spilled so the question is that are we, is the simplified procedure properly identifying appliances that are prone to spillage or is it kind of overestimating the amount of spillage that might a=occur, or underestimating.

We used 11 homes, 10 in Minnesota, and one in Wisconsin. We are looking for homes in order parts of the county and reached out to other organizations to help identify those homes but it turned out it was quite difficulty to find houses that had a narrow window of opportunity that it failed under deep depressurization conditions and yet when you turned the kitchen fam to low and other exhaust fans to off it just passed. So there’s a smaller number of houses where that occurred so instead of seeing a national study, this was centered on Minnesota and Wisconsin. So for all of those houses we did a series of short terms tests and we also went in and put monitoring equipment in, we looked at the burner operation and so when and how often the burner is operating. We used CO2 monitored near the draft hood as an  indicator or spillage, I should say that Dan Cautley from Energy Center of Wisconsin for compiling the data and doing all of that analysis and so what he was able to see that if you monitored CO2 near the draft hood that when there was spillage you would see a sharp rise in CO2 concentration and when there wasn’t it might be more gradual or there’d be none at all so that seemed to be a very reliable indicator of spillage and the duration of spillage.


In addition we are measuring the pressure in the combustion air zone around the appliances relative to outside and we were looking at the status of all the exhaust fans, dryer (when is it running) kitchen fan, bath fans and also for some of the houses we were able to take a look at the door closest to the appliances was open or closed.

We collected that data from 3 to 6 months and we ended up between those 11 sites and we had 1500 days of data. This monitoring stated in February and ended in the summertime.

Go ahead Larry.

Let’s give you an indication of the amount of spillage or number of spillage events by minute of operation and by site. So and then 01 is the first house, and all the way through and then 10 and then WI01 is the house in Wisconsin.

The bars that are kind of maroon or red indicate that those are the events where there was spillage and the ones in gray indicate that those are well all of the events for that water heater.

And from left on the 0 means this data was binned by number of minutes into the operation of that burner so on the left means that you know, there was one minute of operation and over toward the right 15 minutes. So typically two to five burner events per day for these houses and five minutes or more was the typical duration of the burner cycle. So you can see right off that and then 04 and WI01 had significant amounts of spillage so it turns out all of the operation of that burner for those two sights (MN04) 47% of the time there was spillage and for WI01 there was 91% of the time there was spillage.

But otherwise if you look at the rest and you look at all of the operation the fraction of time that there was spillage varied from a low of .3 % for one of the houses up to 5%. And the other 9 houses that didn’t have significant spillage and then 01 had some significant spillage but really only in the first minute of operation. And pretty much the same is true for MN03.

And if you take out those two houses you know what you would find is that of all of the rest of the houses there was spillage for 1% of less of all of the burner operation for those water heaters.

I should reiterate here too that these are water heaters that when tested under a simplified protocol for depressurization that they failed that combustion spillage test, that there was spillage beyond 2 min of operation. So this is real world operationable water heater compared to what you get from a short term test. So in all of these cases of regular depressurization test would say that it’s failed, and that it needs to be replaced and yet in certainly 9 of the 11 when you look at the frequency of spillage it’s around 1$ or less if you ignore the first minute of spillage operation. And kind of going back to what is causing them, and then 04 and WI01 these are vent system problems. Larry if you want to go ahead?

So both of them had vent defects and then 04 they have a picture there on the right so it has three inch vent connector off the water heater with a size foot run and 4 elbows. And when you look in the national fuel gas code sizing tables and the vent capacity was actually 75% of the burner input and you would like to see that to b21 100% or greater. W01 there was actually that was previously common vented with a furnace and the furnace was removed but the vent had stayed in place and was still partially opened on the end of that vent so in both of the cases where we saw significant spillage it turned out it was an entire system issue.

If you want to go ahead and move forward. So this is Dan and the people at Seventh Wave did a logistic regression model to look at the factors that [inaudible] to spillage and you know what they were finding in there is that again, the model says that 04 and 01 were the only ones that had significant amount of spillage beyond the first minute, that’s those red bars.

And 01 and 03 there was significant spillage but only for those first minutes, that’s the blue bars. And they also looked at the effect of outside temperature and you can see in these cases that there is an increasing incident of spillage with warmer outside temperatures. And we’ve seen the same kind of things with our son insulation program and testing that water heaters when we did a sharp turn test we saw higher failure rates in warm weather conditions than in cold conditions.

Can you go ahead and move on Larry?

So we are also doing short term testing there and we are measuring the depressurization of the combustion air zone relative to the outside and we are looking at that that depressurization level was for our simplified protocol compared to in this case it is a comprehensive test so if you use pressure to indicate whether the interior door should be open and closed and that includes bedroom doors and doors between floors and the combustion air zone. And if you use pressure to use whether air handler should be on or off and you turn on all of the fans these are the 11 houses and how the simplified depressurization on the X axis how that compares to the comprehensive test to pressurization on the Y axis so for the most part there’s not very much difference between the simplified dep and the comprehensive except in one case and that turned out to be MN 06 and when our simplified protocol we weren’t turning on, so we turned on the kitchen fan to high, and the next largest exhaust fan, if there was a second bath fan that was not turned on for the simplified protocol, and then o6 there was a 68 cfm exhaust fan in the second bathroom and in addition in our simplified proto we were always closing the door closes to the CAS and in this case it turned out the depressurization was worse with that door open.

And then also there were two main floor bedroom doors where using pressure as an indicator had those be closed instead of open.

So that’s the reason for that one point. The difference between the simplified and comprehensive depressurization measurements.

Go ahead Larry.

So we also looked at using the BPI proto for setting up the house in depressurization conditions and how that compared to the comprehensive and so now you are looking at mainly the difference between using pressure to indicate whether interior doors should be open or closed or not, and so you can see for R10 or 11 houses, very little difference in using pressure as an indicator or whether there was an exhaust fan or return in that room.

So some conclusions again this was turned out to be a 100 page final report that included information on the survey and the short term test. All of the monitoring data. An awful lot of info that came out of this study I’m hitting really just the high points. Overall though we are finding that overall normal systems that had properly venting systems weren’t spilling excessively and they weren’t producing much CM and these again were ones that failed depressurization spillage tests.

Again the vent defects were important cause of the problem. And the worst case conditions were about equal for the simplified and comprehensive methods except for the one house of our 11. I should also note too we are doing another test a cold vent establishment pressure test looking at what’s the level of depressurization that this water heater can vent against or when you back off on the depressurization at what level depressurization will the water heater stop spilling and vent properly.

And for the ones in our study we saw a great variation anywhere from minus 1.7 up to 1.6 and that would be probably similar to a standard that you would use as level of depressurization that when you go about that you would expect to see signify spillage problems. So most people might use minus 5 as a guideline but what we are seeing is that a number of the water heaters that less than the minus 5 and somewhere above that minus 5.

And I think Larry wraps things up for the field part.

Larry: Great, thanks Dave. As Dave mentioned there’s a hundred pages of really good stuff there. I’ll give you the report info in a second. There’s some really good work and there’s some good data to be analyzed and evaluated again I think.

Harmonization - I just wanted to make a comment about harmonization part of the work that we were doing was to take a look at all the differences amongst the various codes and practices. And the top part of this box above the and in the middle is the older codes, 2012 and 2011. The fuel gas code ACCA and BPI and NREL standard specification and the bottom part is updated version so of those codes or recommendations.

You can see where in the top there is quite a variety of the CAS door open or closed, or do you evaluate about other doors. Was the air handler considered, what’s the spillage limit, so there is quite a variety and the bottom you are seeing more commonality I think between the recommendations from the various organizations.  We were a part of the BPI procedure, redevelopment of the BPI procedure. So that actually is a tech transfer vehicle for much of the simplified comb safety work that we did with this project.

Tips, just a couple of things that I wanted to put in here at the end.

Check the vent connector, I think we got the message across. Too long, too narrow, corroded, consider Type B. There’s a photograph from one of the reports and you can start to tell where that connector is going.

So another option if you look at the tables and the tables will allow a larger diameter vent connector to get more within the appliance capacity then you can sometimes has a beneficial impact.

Chimney liner: when downsizing appliances, I know that’s all in the code now. But if you are taking out an appliance or significantly reducing the capacity then the chimneys need to be considered and need to be relined. Unblock the combustion air openings - one of my favorite pictures is someone who left a warning and then it’s totally blocked. Look for them and unblock them. Look for. Use the latest test procedures, look for 2015 versions. And then consider a power vent kit which I think it’s a pretty good solution.

Here’s the report that we discussed today. I think these slides will be available so that you can go to the NREL/DOE website and pick up these reports.

Simplified test particle field study is what we’ve bene talking about for the last 15-20 minutes or so.

So I think at this point I’m turning it back and we are more than willing to answer questions.


Great, thank you so much. We have a number or questions. I’m just going to start with the ones that came in at the beginning of the presentation and then we’ll work through.

Larry, from your part of the presentation at the beginning we had the question whether you had any idea what the impact has been regarding comb in energy efficient homes being built.

Larry: In new construction, you know is definitely tighter because of a variety of things. State building codes or other regulations. It will also tend to have fan assisted furnaces instead of traditional but there will be atmospheric water heaters in many cases. So in some cases we are seeing that it works quite well in new construction. Some markets they use zone isolation exclusively and that’s a good solution. And in some cases the builders prefer to use fan assisted appliances altogether to avoid any kind of issue with combustion safety. So it can be very much a mixed bag.

Nicole: Ok great, thank you. We had a question from someone that it was there understanding that power vented appliances and high efficiency furnaces do not require fresh air inlets but do they?

Larry: Well if it’s a direct vent appliance then it will come with piping to provide its own combustion air so those will generally be category 4 appliances. There are some category 1 appliances, REAM had a furnace that I’m aware of that did have a pipe that used outdoor air for combustion So if you are in category 4 not as much of an issue, air supply – if you are in category 1 then you need to look into this.

Nicole: Ok, great. Thank you.

And there was also a question about which states provided the precise valuable data to the survey and I’m not sure if you can reveal that or not but someone is curious.

Larry: I don’t’ have that here, but if you send me an email I will forward your question along to the people that have all the data.

Nicole: Ok, great. Did the field data include measurement of Carbon Monoxide levels in the “failed” homes?

Larry: Dave? Do you know?

Dave: Um, I know when we did the short term test that it did. I’m looking very quickly right now.

No, it was just CO2 measurements but we did put a Carbon Monoxide detector in the space itself but nothing in the flue.

Nicole: Ok, great. Then we had another question about whether there was Carbon Dioxide or Monoxide being measured in the draft hood. So I’m assuming from your answer that would be Dioxide.

Dave: Correct.

Nicole: Ok, great. And then from the part with the graphs, we had someone say that it seems like the graphs are saying even when there’s spillage it rarely lasted longer than 15 minutes across all the measurements is that correct?

Dave: Yes, the graphs go out to 15 minutes, I’d have to check with Dan but I believe that was because there weren’t that many cycles that lasted beyond 15 minutes. But if maybe you want to go on to the next question I’ll take a look at that real quickly and see if I can provide an answer.

Nicole: Yeah, sure. Do Carbon Dioxide measurements also include effects from fuel fired ovens and ranges?

Dave: Well, in this case, see I don’t have a diagram but there is a nice diagram in the report that shows the location of the CO2 sensor. Actually these were pumped sensors so the inlet for that was placed right at the edge of the draft hood. So we would have been far away from any other combustion appliances.

Nicole: Ok, great. Someone was wondering if we could just repeat was the differences were between the simplified and a comprehensive testing approach and give us an idea of how much time was saved by employing the simplified test approach.

Larry: I think at a high level in the past it was the worst case depressurization testing so the tech would cycle through the house and open and close doors till they identified the door position that provided the worst case depressurization for the appliances, and then exhaust fans I believe were all operating so was the general approach in the past. Identify the worst case and then evaluation based on that. The more simplified approach that I summarized in here was you pick a door position depending on if there was a return air or a jump duct I guess. In this space itself so you identify, you took a look at the room and identified the door position requirement. And then what we did evaluate a couple of options for exhaust appliances so the vent system was operating, the range hood was operating, and then the dryer was operating. So, it’s slightly different than all of the exhaust fans operating at the same time at maximum speed.

There’s quite a bit of that detail in the report.

Nicole: Ok, great, thank you.

Dave: One thing I’ll add to that too is that, maybe, part of it might be a little bit of a time savings. But it’s just more straightforward to describe to a contractor how to do the simplified test if you are only looking at whether there’s an exhaust fan or return that’s a little bit simpler than having to do a pressure measurement under a door or use smoke under all the doors and figure out is it positive pressure, negative pressure, and for people that do one or two of these every day it’s becomes very straightforward or routine. But for a contractor that might do it only once a week or once every few weeks it becomes more confusing.

Nicole: Ok, great thank you. Do you have any advice, data or anecdotes about CSAS testing in multifamily buildings. And are there any other tests that need to be considered for multifamily testing?

Larry: Could you give me the first part of that question again, I didn’t catch it.

Nicole: Yes, it was advice, data or anecdotes about testing in multifamily buildings.

Larry: Wow, there’s a good research topic for you.

Jim: Yeah, we have some experience by the airport with six or eight buildings and we went through opening all the doors to all the units and finding all the fans and all the everything else in one building. In 12 unites, maybe more. And found that in that building and what looked like in all the other ones, the only door that mattered was the mechanical room door. And we found most of those wedged open and dryers present in all the mechanical rooms. So we didn’t have to any of the complicated testing normally done in houses and concentrated just on the building mechanical room.

Dave: We will say that’s a situation where there’s a central mechanical, central boiler. And not individual units in each unit, that’s what Jim was describing.

Jim: Central water heater, central boiler, with the laundry in the same room.  Stupid design, but that’s what we saw.


Jim: And that’s all we had to test.

Nicole: Great, thank you. We had a couple questions about whether this study informs BPI to simplify their comp test or if they’ve chosen to focus on a worst case scenario to be conservative? Can you speak a little bit to whether the testing represents a worst case scenario or whether it should be modified.

Larry: Actually, that’s a good question. Paul Francisco and I and Dave and Pat have all been working out the BPI committee that’s revising their test procedures and actually we’ve got quite a bit of the simplified method in that test procedure now – it’s not identical but it’s a major step in the right direction so if you pick up the newest BPI test procedure you are using a lot of the simplified method here.

Nicole: Ok, great. Dave, can you mute yourself while you are not talking. I think we are getting some feedback.

Alright, thank you.

Someone asked what you recommend for wood stoves and fireplaces since these can’t be isolated.

Larry: I, myself, haven’t studied wood stoves and fireplaces and their effect on combustion safety. I think the recommendation is that you just close the damper to a fireplace. Dave and Pat, have you done anything there?

Dave: Well our rec was always that wood fireplaces aren’t, well, you need to open up a window to make sure they start drafting properly and that in tight houses you very often will have problems with wood fireplaces and woodstoves. You know, what often is a problem is that maybe you might get a good draft at the very start when you have a roaring fire but as time goes on that kind of dwindles down. You don’t have such a hot flue anymore and that’s also a time when you can be producing higher levels of CO. So that was our recommendation to people.

Jim: People have to have a CO alarm in the space. And I was tasked with following up on the complaints of wood smell and we found in houses with basements two story chimneys that air flow in standby condition was from the fireplace into the house. Cold air would fall from the fireplace into the house all the time.

Even in the summer with all the windows open. It’s nuts. This is with old masonry fireplaces in the house. Some are using wood for heat then the advice we gave was to use an EPA listed insert or direct vent gas logs.  But following up with existing fireplaces, I’ve had have a mechanical inspector that taught me that why are you looking at that junk.  Very difficult to make into more than a hobby. So. People want them, and they are an imp part of real estate transactions. So. If that’s helpful.

Dave; One thing I want to point out is that my email address, I missed this, is misspelled. It should be first initial Sorry for not catching that.

Nicole: Thank you for bringing that up, we actually had someone ask already.

Larry: Not going to let me change it. There we go. Is that right Dave?

Nicole: Great, thank you very much. Just a couple more questions. Someone asked what is there a supply in the bathroom that overpowers the exhaust fan would you still leave the door open?

Dave: I think by the simplified protocol yes,

Larry: Yes, yeah.

Dave: If you were going and using pressure as a determination then possibly not, so that would be kind of one of those nuances that you potentially could catch by looking at pressure across a door that you wouldn’t with the simplified protocol.

So there’s a trade-off there.

Nicole: Ok, great. Thank you. And then someone said the presentation mentioned power vent kits a lot and that they looked into this recently and found that the code said the appliance must be listed and labeled for such use. Natural draft appliances do not appear to meet this requirement making after market power vent kits not allowable by code. Is this consistent with your understanding?

Larry: The power vent kits are available and it’s not a modification to the appliance so you have to look at the manufacture’s installation instructions for the power vent kit. That’s my understanding of how they work.

Have you used them at all Dave?

Dave: We used them for a period of time in our installation program and had an interesting interpretation from our code officials that when we were putting them on water heaters with standing pilot they were uncomfortable that if you were side venting those that the pilot could potentially be spilling into the space and because you were blocking the vent with the motor and for that reason they wouldn’t let us use them. Which seems kind of counterintuitive because how many stoves do you have out there with the standing pilot and ranges and that’s fine but a water heater isn’t.

Jim: We did have repeated complaints and calls to the fire department of gas odors in the house from pilot spillage on a boiler. After the boiler gas fell was updated to a modern one with an electric ignition there was no issue. But we ended up having to pull out 24 draft inducing fans and kits that either homeowners had installed or we had installed as remediation.

So that’s more than all of the kits reported in the 1700 or the 28,000 in this survey.

Larry: So it looks like that’s a good point, you have to check the listing to see if there’s any kind of violation to the listing of the code.

Jim:  It’s really hard to tell somebody that’s just the right amount of gas odor in your house.


We can’t even go there.  And if you think that’s funny..

Larry: Oh boy.

Nicole: Alright great. Thank you very much to our presenters. I think we are going to move on to our short survey. If we weren’t able to get to your question we will try to follow up with you offline, so thank you for all of the questions that were submitted.


Thank you for answering our survey.

You can subscribe to receive notices about upcoming webinars and other news at the URL on the screen.

On behalf of the Building America program, I’d like to thank our expert speakers for their time today, and to our attendees for participating in today’s webinar. We’ve had a terrific audience, and we very much appreciate your time.

Please visit the Building America website to download a copy of the slides and learn about the program. We also invite you to inform your colleagues about Building America resources and services.

Have a great rest of your day and we hope to see you again at future Building America events. This concludes our webinar.