skip to Main Content

Climate Zones Are Getting Warmer (Mostly)

2021 IECC Climate Zone Map

Climate zones for energy code purposes are defined in the International Energy Conservation Code (IECC).  Every three years, that code gets updated.  The map that was in the last version of the IECC, which came out in 2018, had been the same since 2003.  The 2021 IECC, however, shows that climate zones are getting warmer in a bunch of counties.

The climes they are a-changin’

The map above shows a colorized version of the 2021 IECC climate zone map.  For comparison, the 2018 map is below.

The 2003 to 2018 IECC climate zone map
The 2003 to 2018 IECC climate zone map

It’s a bit hard to go back and forth between the two maps and catch all the changes, so the Pacific Northwest National Lab put together the version below showing which counties have changed and how they changed.  The white counties are still in the same climate zone.  The red ones have moved to warmer climate zones (lower numbers) and the green to cooler climate zones (higher numbers).

Climate zones are getting warmer
US climate zones are getting warmer, mostly.

As you can see, there’s a lot more warming than cooling.  It’s interesting that nearly all of the counties that moved to cooler climate zones are in Oklahoma.  The big warming changes are in the upper Midwest.  Wisconsin wins the award for the most change.  It no longer has any counties in climate zone 7.  The Upper Peninsula in Michigan also lost all of its CZ-7 counties but still has a toehold in that zone because one of the CZ-6 counties flipped to 7.  Minnesota, which was about half CZ-6 and half CZ-7 now has some CZ-5 in the south and about a third of the state in CZ-7.

What do these changes mean?

First, of course, these changes show that the climate really is changing.  The folks who update the IECC are builders, code officials, manufacturers, and other stakeholders in the construction industry.  Yes, there probably are some who talk about the “climate emergency,” but on the whole, I’d say this is a pretty conservative group.  The IECC climate zones are based on actual numbers for heating and cooling degree days.  This is a fact-based change.

Second, the changes mean that the counties moving to warmer climate zones may have to put in less insulation, do less air sealing, and generally have an easier time meeting their energy code.  Of course, this all depends on whether or not those counties have adopted the IECC, which is a model code.  What actually happens depends on what the states and smaller jurisdictions adopt (and enforce).  In Georgia, we just went on the 2015 IECC two years ago.  If past is prologue, we’ll think about adopting the 2021 IECC in another four years or so.

In short, the International Energy Conservation Code, using real data, has established that climate zones are getting warmer.  That has implications for homeowners, builders, manufacturers…everyone, really.  None of us can escape the Earth’s climate.


Note:  I found out about this change recently in an article by Clay Dekorne in the Journal of Light Construction.  Since then, Green Building Advisor published an article on the topic, and I found one from March by JR Babineau at Johns Manville.  See those articles for more details than I’ve provided here.


Allison Bailes of Atlanta, Georgia, is a speaker, writer, building science consultant, and the founder of Energy Vanguard. He has a PhD in physics and writes the Energy Vanguard Blog. He is also writing a book on building science. You can follow him on Twitter at @EnergyVanguard.


Related Articles

Do You Know Your Building Science Climate Zone?

The Fundamentals of Heating and Cooling Degree Days, Part 1

Could an Energy Code Really Be So Simple?


NOTE: Comments are moderated. Your comment will not appear below until approved.

This Post Has 27 Comments

  1. Residential retrofits responding to the IECC need accurate Manual J heat load calculations. When I needed to do that, I was disappointed to learn that my city had trashed all the construction information submitted by the builder when getting the building permit, so I had to reverse engineer it. Do you have any idea whether this is a widespread problem, or just an isolated incident? If widespread, I’m thinking that doesn’t bode well for us combatting climate change by improving our existing stock of homes. Or perhaps acceptable Manual Js are accomplished without the need for much precision?

    1. Bill, no, I don’t know how common that is, but I would imagine that a lot of building departments don’t hang onto paper documents longer than they’re required to because of the space needed. With electronic documents and the low cost of storage, that shouldn’t be an issue, though. Even without knowing what the builder told the building department he was going to do–which might be different from what actually got built–you can get enough info about the house to do a decent Manual J.

      1. Allison, I believe you saying that a decent Manual J can be done without having access to builder’s plans. Heck, I did one on my own house (circa 1997) without access to plans, using Wrightsoft, and I think I did it decently. I know I put a lot of time into it, because it was my first. But how can I know for certain that the parameters I used reflect reality? For instance, I could find nothing at all written on the windows, so I had to make an educated guess on that major component, and I did not break into the walls to see how much insulation exists in the 2 x 4 cavities, etc. I think the structure mostly yielded to my study, and consequently I do feel it was done adequately.

        I have no idea how you and Energy Vanguard do your Manual Js on existing residences, but expect that you have found ways (shortcuts, experience) that permit you to do them quickly and accurately, well enough maybe that you really do not have a need for original builder’s plans, or wouldn’t use them if they were available. After all, that’s a big part of your business, I suspect, so you cannot afford to spend a lot of time getting decent Manual Js done.

        It only struck me, a determined beginner, when I set out to do my Manual J, that the builder’s plans sure would have been helpful in saving ME lots of time, and I innocently extrapolated that large savings of time to beginners all across the country. What I was missing was the fact that the HVAC industry isn’t comprised of beginners, which indeed reduces the importance of inspections departments retaining those documents. That’s not to say that they shouldn’t keep them for the life of the residence, for other reasons, which, as you say, digitization helps immensely, it’s just that doing decent Manual Js isn’t the driving reason.

        I’m guessing you have a chapter in your upcoming book that describes a good way to do decent Manual Js on existing residences.

        1. Regardless of whether the HVAC industry consists mostly of beginners, I can say with reasonable certainty that most HVAC practitioners at the residential and light commercial level do not routinely perform Manual J calculations because in most circumstances it is not required.

          I can’t speak to Energy Vanguard’s methods, but I’ve done / supervised several hundred Man Js in the course of business in the past dozen years. Building age has ranged from the mid 1800s through homes not yet built.

          Plans are rarely available for existing homes. If they are available, their primary utility is not having to physically measure room dimensions onsite.

          Knowledge, skill and experience all contribute to insight as to the hidden elements of wall assemblies (principally insulation or lack thereof) and likely fenestration (window and door) energy performance based on age of the structure and observable characteristics.

          Since we own the equipment, it is not uncommon for us to demand / perform a blower door air infiltration test in the course of a Manual J calculation if circumstances suggest it would be useful.

          1. Curt, thank you for the information. When you said, “most HVAC practitioners at the residential and light commercial level do not routinely perform Manual J calculations because in most circumstances it is not required,” were you referring to requirements of local code authorities? If not could you explain how they confidently select system capacities?

      2. I’ve done a LOT of new construction J’s over the years and one thing I’ve learned is to never rely on construction drawings for critical details like insulation, window specs, internal loads, ventilation, etc, assuming these specs are even on the plans. The problem is this stuff often gets changed during construction when the bids come in.

        Some jurisdictions are getting strict about requiring energy related specs on permitted plans but that doesn’t mean they get updated if (I should say when) changes are made during construction. Bottom line, when doing a load calc, the bulk of my time is spent chasing down the stuff that’s not on the plans and/or has yet to be decided. In some cases, I have to issue a revised load report just prior to mechanical rough.

        As Curt said, the main benefit of plans for an existing home is not having to do field measurements. Aside from insulation R-value, which can usually be estimated, an infrared camera can help assess insulation quality. And you can usually tell the type of insulation by removing an outlet or light switch cover. Attics are easy, sub-slab insulation is likely to be impossible. A blower door test is one advantage you have in existing construction that you don’t have (soon enough) in new construction.

        If the existing HVAC equipment is still working and reasonably maintained, you can get a far more accurate estimate of the true load than any Manual J by timing the cycles when outside conditions are near design. That’s something I proactively help family and friends with, so when it comes time to replace their HVAC, they’ll already know what size they need.

        1. David, thanks. Your information is helpful. I definitely like your solution for doing a Manual J without doing one in existing buildings, by timing HVAC cycles at design temperatures. Must add that your family and friends are lucky that you exist.

          You said that you sometimes submit a revised load report just prior to mechanical rough-in. That report would be put into the permit package with the rest of the plans, I suppose, so, you are at least one person who believes in updating submitted plans, to document the structure “as built” for posterity. Yet you and others say that plans aren’t often updated according to changes that are made between design and implementation. Do you happen to know what authority has jurisdiction over what information is required by building permits to “document” each residence, and whether updated plans are normally required?

          1. It is interesting how two guys (David and I) that are highly committed to properly sizing HVAC systems deal with their respective marketplaces.

            90+% of my team’s new construction Man J calcs get built by us. Our calcs become part of the permitting process. In my 12 years at this gig, I can think of just 1-2 times where we have had to resubmit calculations and energy performance paperwork owing to changes made during construction…and then only if the changes were radical, such as switching between air source and geothermal heat pumps.

            I’m same as Dave with respect to duct designs – we rarely do them unless required for permitting, and we can be sure what we put on paper won’t get built owing to unforeseen and unforeseeable structural constraints…we almost never have anything approaching an “MEP” (Mechanical / Electrical / Plumbing) plans that incorporate those elements into the structure…those are typical for new commercial buildings, but not residential in my area.

            Our room-by-room Man J load calcs inform our installers what air flow each room and zone requires…how they get there is up to them…they redesign on the fly as dictated by structural reality, negotiating with the project super and other trades to work out how spaces in floor systems are allocated.

            I like the idea of cycle timing in the course of specifying retrofits / system replacements, but unfortunately by the time we are engaged to propose system replacement the existing system has either completely failed or is operating in a severely degraded state such that accurate cycle timing is no longer possible.

            Bill asked how contractors who do not perform Man J confidently size equipment…they either propose “like for like” or apply a “rule of dumb” (thumb) such as “1 ton per 600 SF” regardless of actual conditions.

          2. I love Curt’s comment because it shows what a dedicated, smart contractor looks like. If the typical HVAC contractor were like this, there’d be no need for 3rd party designers. That would be just fine by me! I’m getting too old for this!

            Curt’s wrote: “…how they get there is up to them…”

            Exactly! Manual D is nothing more than a cookbook design procedure that relies largely on utilizing various duct sizes and fittings to achieve room-to-room balance at a specified friction rate. It’s not the only duct design procedure, or even the best.

            As Curt points out, it doesn’t matter how one designs the duct system as long as it delivers the correct airflow and balance at an acceptable static pressure drop. In fact, I would argue that Manual D is getting less and less relevant as building enclosures get more efficient. After all, when the design load only calls for 0.25 cfm/ft2 or even less, there’s no way you can design a self-balancing duct system a la Manual D when it calls for pipes smaller than 4 inches for every branch. In building code parlance, Manual D is just a piece of paper that satisfies a checkbox. The real deal is post-install test & balance as standard QA practice.

          3. Anyone got a guide on cycle timing to measure load? I’d love to know how to set something like that up. I find the whole discussion on Manual J (or any modeling) a problem because there’s no feedback, you never get to know the actual number (and the equipment doesn’t provide any data on its operation!).

            So is cycle timing only possible when things are at design conditions? So it’s name plate hourly capacity of the equipment/average proportion of hour operating?

            Or can you plug in the outside conditions and do that anytime?

            And can one do this with VRF equipment or only single stage? You can’t use proportion of time operating because of the variability, can you use wattage draw measurements to know the output of the equipment?

            This is sort of the electrical version of fuel oil consumption estimation?

          4. @James, I’ve never seen a formal write-up of cycle timing. I developed my own procedure:

            Cycle timing, when possible, is a proactive test to reveal how much of an existing system’s capacity is required to offset the load. For example, if an air conditioner runs 60% of the time during a cycle timing test, the load at those conditions is roughly 60% of its capacity.

            Cycle timing should be done when outside temps are near or preferably above the design temperature (opposite in winter). The results can be extrapolated, but only by a few degrees (based on delta-T). Keep in mind that cycle timing indicates the real time load, which is different than the design load. We typically design to the 1st and 99th percentile temperatures because of thermal lag. OTOH, with cycle timing, you’re essentially on the receiving end of thermal lag so it’s best to wait for days that exceed the design temperature by a few degrees.

            In winter, cycle timing should be performed during the hour or two before sunrise (before sun’s radiant energy interferes), and during late afternoon in summer (cooling load typically peaks a couple of hours AFTER the daily high). You want to time at least 2 full on-off cycles. Three cycles is better but may not be feasible if cycles are long. Ideally, a cycle timing test lasts more than an hour but less than two. It’s important for the interior temperature to be stable for at least an hour before commencing the test.

            The procedure is simple: Note minutes and seconds (or use a stopwatch) at the start and stop of each cycle. It doesn’t matter if you begin at the start or end of a cycle as long as you end the test at the same point. Once you complete the test, divide the total “on time” by the total elapsed time for the test (i.e., total “on time” + total “off time”). The result is a fraction, which gets multiplied by the equipment’s nominal capacity. When extrapolating, I look deeper at the expanded performance data tables for more refinement, but that’s way beyond the scope of this already-off-topic diversion!

            As you noted, cycle timing won’t work on multi-stage or variable capacity equipment unless it’s locked into high stage. And it definitely won’t work if the existing system is impaired, or for an undersized system that runs continuously and can’t maintain the setpoint. Lastly, if major efficiency improvements are contemplated, hold off on replacing the HVAC until after the other improvements if possible, so cycle timing can be used to reveal the new load.

        2. @Bill, city or county building code officials determine what information is required for permitting purposes. This is sometimes regulated by city or county governance but the details are typically left up to the senior code official.

          In reality, my load reports rarely make it into the permit package. Here’s why:
          First, I never accept work for permitting purposes. Anyone who approaches me with permitting as their opening is unlikely to be interested in what I have to offer (the load calc itself is just one input in my design process). Second, in my experience, most jurisdictions still don’t require load calcs (my clientele is probably atypical as most build in rural areas, which tend to have lax building code enforcement). And third, in cases where J/S/D submittals are required, the mechanical contractor often ends up submitting the reports.

          In particular, I don’t do duct design. 99% of my projects are non-local so I can’t be on site during construction, requisite to getting good results. Instead, I’m often tasked to help the client vet prospective mechanical contractors. They must demonstrate good duct design-build skills.

          In addition, contractors under consideration must be willing to work with a 3rd party designer, relatively rare in residential. If the contractor uses Right-J, I’ll provide my project file that can be used as the starting point for duct design. It’s paramount that the contractor “own” the design since he’s the one that gets the call if there’s an issue. Understandably, this sometimes leads to friction. Let’s just say I’ve learned over the years how to negotiate these challenges!

          This is getting way off topic. Check out Allison’s article on 3rd party design.

          Just a clarification… cycle timing is not “doing a Manual J” (I know you didn’t mean it that way). Manual J is a specific procedure to establish the design cooling and/or heat load. After cycle timing, it’s necessary to do a Manual S or equivalent (equipment selection). Keep in mind that knowing the load is decidedly different from determining which equipment to use!

          1. Funny you specifically mention 4″ supply duct branches. Of the 52 supply branches set forth in our latest Man J / D, for a 7200 SF custom home, 12 of them are to be 3″

            3″ flex duct materials are not stocked locally, but can be special ordered, so we do via periodic bulk buys, and from what I gather, we are the only contractor in town doing that.

            Awhile back we worked a job where the homeowner had already engaged a 3rd party to run Man J and D calcs. I noted multiple 4″ branches slated to receive less than 24 CFM and wrote the author to ask if he objected to our substituting 3″ diameter in those instances…he was dumbfounded to learn that was even an option!

            I do admit we do not automatically perform a T&B, but we have the tools to do so and proceed as-needed basis using a baby Alnor flow hood and opposed blade damper registers. We’ve move away from hand dampers, even though they are much less expensive – every hand dumper introduces one or two duct leaks and insulation gaps.

          2. @David First, I feel bound to express my appreciation for your kind and thoughtful answer, as I found it quite helpful. You expanded my understanding on a number of related issues and helped me draw a clearer picture of how things work, in reality. I came here to learn, and people like yourself surely do help with that.

            Second, to make one point clear, last month I paid my second year’s lease payment to Wrightsoft for their simulation software: Right-Draw, -J, -D, -2Line, -ComDuct, -Comm2Line. That’s to let you know that I am (keenly) aware of the effort it takes to “do” a Manual J, what it means to “do” a Manual J, and how an HVAC design proceeds from J (load calc) to S (equipment selection) to D (duct layout) to T (space air distribution). That means I can definitely appreciate the value of having an easy way to verify how well the Manual J was done, the brilliant way you provided by suggesting the use of HVAC cycle measures at design temperatures. Your simple measure provides “ground truth.”

            Third, what I wanted to know you did answer. I was disappointed, but you told me what is real, and that’s what I most needed to know. You said that requirements on what “data” is collected during the building permitting process is decided locally, with responsibility normally given to the senior code official. You further said, “in my experience, most jurisdictions still don’t require load calcs.”

            Why I felt disappointed:
            – we have a huge stock of existing housing here in the US; HVACs at <=10 SEER
            – we are facing an existential threat due to climate change (aka global warming)
            – in partial response, we intend to improve residential energy efficiency (IECC)
            – Manual J is the way we calculate a home's thermal efficiency (total heat loss/gain)
            – Some jurisdictions don't even require Manual J
            – I had been thinking that the easiest way to quickly generate a Manual J would be to utilize builder's plans, but they may not exist (local decisions governing what is collected for the permit, lax follow-up when building changes are made, permit files tossed when filing systems overflow), and in my personal case, that information had been tossed by the city.

            That's disappointing, but Manual J isn't the only information needed to improve the energy efficiency of existing homes, and you showed an alternative means that could be used to do the equivalent thing as the Manual J does (determine the heat loss/gain of the residence). So, in reality, the situation isn't as bad as I had originally thought.

            Since your measurement technique needs to be made at "design temperatures," as a nation, we need to be figuring out how to best capture "equivalent Manual Js" for existing homes, so when the imperative comes down, saying that the country is finally serious, demanding we begin to actually make energy saving modifications to the many existing residences, we might have some basic information ready.

  2. Were the dry/humid designations examined as well? The 2018 IECC CZ map (and previous versions) shows rainfall patterns reflecting political boundaries other than in Texas and Oklahoma.

    1. Bryan, that’s not on the map this time, but they did designate the warm, humid counties with an asterisk in the IECC. Since the climate zones are by county, they’ve always reflected political boundaries.

  3. In my opinion, this is further evidence that the code differences based on climate should be ditched altogether and we should be building home everywhere that will work in any climate. We all know it’s changing, and since we can’t predict which zone will change in which direction we’d better prepare for all options.

    1. Joe, if we did that, we’d probably end up making things worse because we’d increase the carbon footprint of homes in warmer climates more than could be justified.

  4. I read recently that per recent “TMY” (typical meteorological year) data update that it isn’t getting hotter in Florida (dry bulb, anyway), but we are seeing more moisture, higher humidity, higher dewpoints, higher wet bulb temps.

    That dovetails with my 25+ years living here with 12 in the HVAC biz – our biggest challenges are managing moisture – dewpoint control failures, and I think it is getting worse.

    1. Curt, yeah, the climate is changing in a bunch of ways, and it’s different in different places. And another phenomenon is increasing the average temperatures without affecting what size air conditioner you might need, at least in some places. The heat waves get all the attention, but the nighttime low temperatures aren’t as low as they used to be. That means more runtime and energy use, but you don’t need a bigger air conditioner.

      1. Yes! If anything, the phenomenon of elevated nighttime low temps drives the need to pay ever more attention to building envelope details and optimizing HVAC system sizing and dehumidification capability since in my experience the overnight low temp often becomes the floor or minimum dewpoint for much of the rest of the following day.

        That the overnight low is 74-76 vs 72-74 likely has minimal impact on operating times and energy costs but more significantly stresses both the HVAC system and its host building’s ability to resist problems arising from elevated humidity – occupant comfort problems and uncontrolled moisture fails become much more likely whenever outdoor dewpoints and wetbulbs push into the upper 70s…a phenomenon we encounter more and more often.

  5. So … does that mean that vapor barriers need to be moved to the other side of the wall anywhere?

    More importantly, do we have a concept of “drying days” or “dehumidification days”? Drying days would be the days in which the wall can dry to the outside (outside RH lower than inside), dehumidification days would be the opposite (outside RH higher than outside). Hmmm, can dewpoint differences be used (pretty sure RH is what drives drying potential).

    And, are these decided based on average hdd or 1%/99% values? ie do these relate to the likelihood of extreme weather?

    References/links appreciated!

    (something like days in which the dewpoint outside

    1. James, I agree with what Curt said. Also, water vapor problems are rarely because of diffusion. The air barrier is still way more important than vapor retarders.

      Interesting idea about drying and dehumidification days. I don’t think that would ever be adopted beyond a few building science geeks. I don’t mean this in a condescending way, but a lot of homeowners don’t know how to use their thermostats properly. Curt has experience with Floridians “totaling” their homes because of setting the thermostat at 68F. And I’ve heard of HVAC techs telling homeowners to put their heat pump t-stat in emergency heat mode whenever the outdoor temperature drops int the 30s Fahrenheit.

  6. Those are interesting questions – caused me to think a bit about the construction assemblies I like best – closed cell spray foamed attics atop ICF walls. Those assemblies bury the dewpoint deep in a vapor impermeable volume rather than merely on the face of a plane.

    In a warm climate high mass walls store and slowly transfer sensible heat well after sunset, causing cooling systems to operate more often and longer at night, improving humidity control.

  7. ACCA Manual J’s table 1a has not changed in far too long.

    When will it get revised? Updated?

    ABC Morning News showed NYC is building sea walls for rising tide line.

    1. @Chris, table 1a comes from the ASHRAE Fundamentals Handbook, which is updated every four years, 2021 edition being the most recent. MJ software gets updated based on that.

Comments are closed.

Back To Top