GBA Logo horizontal Facebook LinkedIn Email Pinterest Twitter X Instagram YouTube Icon Navigation Search Icon Main Search Icon Video Play Icon Plus Icon Minus Icon Picture icon Hamburger Icon Close Icon Sorted

Community and Q&A

Doors: blowing agents and thermal bridging of multi-point locks

charlie_sullivan | Posted in Green Products and Materials on

I’ve been shopping for doors, and mostly looking at foam-filled fiberglass in order to get a decent R-value. Most are polyurethane (PU) foam, and I’m guessing that they use HFC if not HCFC blowing agents. I’m not thrilled by the global warming impact of those. But it seems very difficult to find any information on what blowing agents they use. Therma-Tru replied to my query and said they “do not disclose the formulation of the blowing agents used.” The only line of doors I could confirm uses low-GWP blowing agents is Jeld-Wen, who use Neopor, which is a graphite enhanced EPS foam, so you get similar R-value to PU, with a relatively benign hydrocarbon blowing agent.

Question1: does anyone know of other door manufacturers who are willing to tell you what is in their foam?

Question 2: I also need a garage door. R-value there won’t affect energy use as it’s not a conditioned space, but it will affect the temperature the garage floats at so I’d prefer some decent R-value there. I found that Clopay gives you the option of PU or EPS, though they don’t tell you what the PU is blown with. But there are quite a few with 2″ thick EPS, which is thicker than most. Anyone know of another garage door manufacturer who discloses their blowing agent?

Question 3: It seems that multi-point locking systems provide an improved air seal both by preventing the door from warping and being loose at the top and bottom, and by drawing it in to the jamb as it latches. But they are installed by putting a metal strip along the door edge. That seems to me like the opposite of installing a thermal break. My intuition is that it’s OK. The fiberglass skin won’t collect heat and conduct it to that thermal weak point the way glass or a metal skin would. But does anyone know of any data on the effect on U value of adding a multi-point lock to a door?

Overall, I’m feeling good about Jeld-Wen doors with Neopor insulation and a multipoint lock, and Clopay garage doors with 2″ of EPS, but I’m surprised how little information there is on this, so I’m curious to hear whether others have considered these issues.

GBA Prime

Join the leading community of building science experts

Become a GBA Prime member and get instant access to the latest developments in green building, research, and reports from the field.


  1. Expert Member
    Dana Dorsett | | #1

    #1: We don't know- neither PU SIP manufacturers or door manufacturers specify whose polyurethane they use (and they may use multiple vendors), which would be one way to determine the blowing agents.

    #2 The temperature/insulation level of an attached garage affects the load on the house, and thus energy use, just as the temperature/insulation of unconditioned attics and basements affect energy use. But it's the same answer as #1 regarding blowing agents.

    #3 You'd have to calculated the U-factor from first principles based on the conductivity and cross sectional area of the steel. The actual U-factors of garage doors are well above 1/R using any of their published R-values, which is a center-panel number and does not include the hardware or the thermally bridging plastic at the panel edges.

  2. GBA Editor
    Martin Holladay | | #2

    For more information on the routine lying that plagues the garage-door industry, see Energy-Efficient Garage Doors.

  3. charlie_sullivan | | #3

    Thanks to Dana and Martin. I did see that informative though depressing article on garage door R values. Instead of looking at those numbers, I'm trying to just go by the construction. The challenge is that if I want to stick to EPS, that tends to put in in the low-end of a given manufacturer's line, which then means they drop some of the features like thermal breaks and weatherstripping between panels. So the challenge is to find one with thick EPS, good weatherstripping, and a thermal break.

    Clopay publishes a chart comparing their garage doors to other manufacturers', including tested whole-door U-values for some of theirs! But don't get too excited--it's only for commercial doors. But you can get commercial doors 3" thick, and give your garage that loading-dock look for less than $3k.

    On entry doors, it sounds like the Jeld-Wen Neopor doors are a good find and I'll stick with them.

  4. kevin_in_denver | | #4

    After recently ordering a few "triple point locking" doors, I discovered that they can be of two different designs. The better design is what you described, with a rod going up and down and into the jamb above and below the door.

    What I got instead on most of my doors was a cheaper design that locks in three places on the jamb's vertical surface. It has the normal deadbolt, doorknob, and two additional "hooks" that are a foot away from the top and bottom of the door. The design of the hooks doesn't improve the air seal by drawing the door against the jamb as you describe.

    Is there a lock style I should have specified?

  5. charlie_sullivan | | #5

    Wow, thanks Stephen, for addressing the thermal bridging effect of the lock so thoroughly!
    I'm curious about the difference between the "thermally broken" aluminum sill and the wood sill. Presumably that wood sill is better, but how big is that difference?

  6. Stephen Thwaites | | #6


    Not sure that you should be concerned about the thermal effect of the multipt lock. Other factors have greater effects.
    Quickly added a continuous lockbar to an existing fglass skinned door model. It reduced door R-value by about 0.1. This makes sense when you consider;
    - Door frame overlaps door edge, so the edge of the door panel is a bit buried
    - Lockbar is steel (less conductive than aluminum)
    - At the edge, door panel is wood (~R-1/inch) not foam (~R-4/inch), so the door edge is not the best insulating part of the door panel
    Of greater importance is the amount of wood in the door panel. Premium fglass skinned doors tend to have larger internal Lock Stiles to reduce warping when installed with single point locks. Typically this larger Lock Stile is 3-4” wide. In round numbers making it smaller, to say a similar size to the Hinge Stile, increases the door insulating value by about R 0.5 in a typical foam filled door.
    You can see the bridging effect of the Stiles in the attached IR photo of a multi-pt locked fglass door (Note door is installed a ‘reefer’ trailer so the wall looks unusually warm).
    In theory these internal effects should be accounted for in the NFRC U-value for the door. However, most NFRC whole door U-values are for doors with wood sills. Most people buy doors with ‘thermally broken’ aluminum sills. So many NFRC U-values for doors are not as accurate as they could be.
    Stephen Thwaites
    Thermotech Fiberglass Fenestration

  7. charlie_sullivan | | #7


    From what I've seen, Hoppe is the dominant supplier of multi-point locks in the US. Their HLS7 seems to be the standard, and they claim in their HLS7 catalog that both the "hook" and "tongue" versions of the extra bolts on the side of the door "pull the door into the frame for added protection against
    extreme weather conditions." There is an option, which perhaps can be added to your system as a retrofit, to also have a bolt the extends vertically out the top and bottom of the door. They call this a "shootbolt".

    They also have an HLS9000 system that has a "roller" system that they say "is easily adjusted with the simple turn of an hex wrench for improved weather sealing." So it might be that the HLS9000 is the way to go. It comes in a "manual" version and an "automatic" version. It seems that the automatic version engages the multiple points when you close it regardless of whether you lock it. The mechanism (without the decorative handles/trim) costs about $120 for the HLS7, and about $220 for the HLS9000, from what I can see online.

  8. wjrobinson | | #8

    Garages should be insulated, drywalled and properly taped and painted. IMO it should not be on the penny pinching list of shorting. Just like the horrible electric jobs done in tract housing. I am adding electric now to a 30 day old new home. They need two more circuits in the kitchen should have had 4 more and they need another in the garage and basement.

    Back to the garage, 85 degrees out, 76 degrees inside with AC on, open the door to go out in the garage to take a few trips during the day, and the garage is like 95 degrees! This winter it will be freezing cold in that garage, Insulation would buffer that.

    As to multi lock doors, I would not use such in the main use door. Its a lot of hardware to move easily and to keep maintained for 40 years. Try the brand you want to buy and make sure you want to operate all that hardware a thousands of times. I have had good experience with 3 brands of doors and not so good with one. Jeld Wen is not big in my area and have not tried them.

  9. Stephen Thwaites | | #9


    That's actually a tougher question than the effect of the multi-pt lock, because there are lots of thermally broken aluminum sills.
    Here's my best attempt at a short answer;
    Most non-embossed (flush) 1 3/4" door panels with frame and wood sill are typically ~R-5.
    For these doors the sill could affect the overall R-value of the assembly by ~R-0.2
    Perhaps more importantly, the least thermally broken sills can be a source of condensation or frost in cold climates - especially if they are installed onto a thermally weak sub-floor - like say concrete or a metal sill pan.
    If it's a better than average insulating door, say panel and frame together is ~R-8, the choice of sill has a bigger effect; ~R-0.5 --- which, in my view, is alot for something that is maybe 2% of the area of the door.

  10. charlie_sullivan | | #10

    Stephen, very interesting. Conventional wisdom would be that a wood sill would be a bad idea because it would eventually rot, but given the condensation issue, it might be a better choice for durability, vs. the Al sill leading to rot in adjacent wood. But the Al sill units have nice adjusting features built in, whereas I don't see that with wood. So would I be asking for trouble in getting everything lined up right and sealing properly with a simple wood sill?

  11. Stephen Thwaites | | #11

    In theory if a door is made well and installed well, weatherstripping should not need to be adjusted.
    However, even if no provision is made for adjusting, there are always a few options, should adjusting be required. For example, kerfed in sweeps can be shimmed down with foam tape. And perhaps while not desirable, wood sills can be modified to make doors fit tighter (or looser) if needed.
    So even if there is "trouble" with a non-adjustable wood sill, there are options...

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |