DEAR GLENN: I’ve read many of your articles in JLC and I’ve watched videos on your website, too. You seem to be the go-to guy when it comes to deck construction. Maybe you can help me with my headache! I apologize if my rant is long (see below), but I’m hoping it’ll straighten out the issues for other builders, too. — Scott Wells – Scott Wells Construction Co., Medford, OR
A Note from the Publisher:
This article was originally written by Scott Wells, and we submitted it to Glenn Mathewson for his thoughts and feedback. For those of you who don’t know Glenn, he is a long-time tradesman, contractor, consultant and building inspector with significant experience in the field of code development and education. We decided to publish both Scott’s original article as well as Glenn’s responses as a two-part series; this first segment pertains to ledger attachments. Look for the next article on lateral load requirements soon!
It doesn’t seem to me like I’ve been in the construction industry that long. But when I see all of the continually changing code requirements I start to feel like I have. Many of the code changes are good and force the industry to build healthier, safer, and more efficient homes. But once in a while a code requirement comes along that I just don’t get. The lateral bracing requirements for decks is a perfect example. Those ever-changing requirements have become unnecessarily ridiculous. This is one of those times when too much of a good thing becomes a problem.
We’ve all seen the photos, YouTube videos, or read the stories: Overloaded deck. Improper flashing. Excessively rotten ledger board. Ledger attached with nails. Ledger attached to 3/4-in. OSB rim board. It seemed like every week or so, you heard a new story about a deck collapsing. And the main culprit: the ledger attachment. No doubt, something had to be done. It was time for the code makers to look at the way decks were being attached.
Here in southern Oregon, I’ve torn out or performed work on countless decks and seen several different ways that carpenters attached ledgers. Most are installed the same way: ledger nailed to the wall, sometimes right on the siding, or nailed right on the sheathing, with no flashing anywhere. What were they thinking? Usually the sheathing and siding are rotten and sometimes so is the rim board. Occasionally, I remodel a home and find a ledger was installed properly, with metal flashing redirecting water to the outside of the ledger, attached with lag bolts to a solid lumber rim joist, and maybe even with a planned air gap between the ledger and the home. But that’s the exception, not the rule. Here’s what I’ve been doing:
The contracting industry often looks at building codes as a thorn in their side, a government hurdle that must be cleared on each job, and throughout each job. While at times this is an understandable perspective, decks are a perfect example of what can happen to an industry, and a nationwide stock of structures, when there are no codes. The deck industry has been the Wild West since its inception—a place where lawmen are charged to administer laws that don’t exist, or are established only locally and partially. When decks have no codes, anything goes, which means that, from job to job and jurisdiction to jurisdiction, there was (and still often is) no consistency in what’s expected by the building authority. Codes provide consistency to contractors and clear and uniform rules for both builder and inspector. They minimize surprise. They keep all players in the game of construction accountable to the same set of rules. As we’re seeing in the news feeds, decks built without codes are now showing signs of age and poor integrity, and they’re often dying in catastrophic display.
I remember in the mid 2000s when the first wave of deck ledger code requirements surfaced. Suddenly building departments in my area required bolts for attaching ledgers to walls, and they specified metal “Z” flashings above the ledger and behind the siding. Clearly a sensible and reasonable improvement, even though I heard some carpenters complain about unnecessary and expensive code requirements. Those carpenters were obviously lazy and couldn’t have been more wrong. Hey, we’re in Oregon, man! It’s wet here and sometimes stays wet for months! This is not a good place to trap moisture! Rot will happen! Sure, it may take longer to remove or cut back some siding to slip flashing in. Sure, it may cost more to purchase lag screws. But this is the bare minimum for best practice. And “best practice” has become a key phrase for deck builders, or as the American Wood Council (AWC) puts it—”good practice.”
The history of the AWC is pretty interesting, especially if you’re in the deck business. Evolved from the American Forest and Paper Association, the AWC represents over 75% of wood products manufacturers. Their staff develops and promotes standards for wood products to ensure their best use. The AWC also “advocates for balanced government policies that affect wood products,” which is one reason they produce the Design For Code Acceptance manual (DCA-6), a graphic guide that really helps builders understand and incorporate often confusing code changes.
Rather than creating a new code, the AWC augments the IRC code; they pull together the best resources available—like the Fairfax County Typical Deck Details document, that enable builders to improve their practices. The DCA-6 is an all-inclusive guide, and many of their recommendations (and graphics) are used to develop the IRC code. However, the AWC reminds everyone that the IRC code always takes precedence, though DCA-6 includes provisions and details that are not found in the IRC.
Personally, I’d be lost without DCA-6, which helps me navigate the prickly IRC code and better understand the installation requirements for deck hardware.
When the three prominent publishing building code organizations across the US merged into the International Code Council and published the first IRC in 2000 (a 1998-IRC was published, but it was never mainstream), they included a paragraph prohibiting the use of nails “subject to withdrawal” for ledger connections. While good intentioned, a mere prohibition of one fastener orientation helped the decking industry about as much as the advice not to eat wild mushrooms would lead someone to a healthy meal. It prompted the end of ledger attachments with nails, but did not provide any guidance for a correct way to attach them. While it has its place in standards, prohibitive language alone does not suffice. Little changed in the first few editions of the IRC, and little changed in the potential hazard of decks built across the country.
The AWC recognized this problem and the wide assortment of locally developed deck standards throughout the US, and they decided that their sixth installment of their “Design for Code Acceptance” line of free publications would pertain to deck codes. Beginning with the 2006 IRC, it proved rather quickly to be their most successful DCA to date.
DCA documents provide ways to meet building code requirements. Provisions contained in DCA documents that are not included in the building code are considered good practice recommendations. Many building officials look to the DCA-6 as a means to approve deck construction, and the state of Georgia even legally adopted an amended version as the state deck code. Many provisions from the DCA-6 were later incorporated into the 2015 IRC, where decks are now much more comprehensively addressed than in any previous model code in history.
A few years ago, deck ledger fastening requirements, introduced in the 2009 IRC, were enforced in my region. We were given a “Fastener Spacing for a Deck Ledger” schedule that made a lot of sense and gave us some guides for ledger fastening, which was a big help and a boost to our industry.
Since 2009, new timber-type screws, like Simpson Strong-Tie’s SDS and FastenMaster’s LedgerLOK have also became available options (when approved by local building officials). Pre-drilling separate pilot holes for the rim board and ledger in order to install lag screws has become a thing of the past. These new types of fasteners can be simply installed with an impact driver following the manufacturer’s specified layout.
|Ledgers can also now be spaced off the wall with stacked washers or Deck-to-Wall spacers, providing an air gap so ledgers can dry out and prevent rot.|
In 2012, the IRC made major changes to the bolting requirements.
In fact, the new guidelines for ledger fastening are so complicated, I can’t understand how anyone in their right mind thought contractors would be able to meet them. I sure hope the folks in southern Oregon never adopt that part of the code!
In 2005 a wood research team out of Virginia Tech acknowledged the need for a recognized method of attaching a deck ledger to a wood-framed floor and they set out to test it. With limited research funding, they tested the best variety of attachments they could and then they published an article on their findings. By 2009, the IRC was eager for a code solution and, during their hearings, they reviewed a proposal based on the Virginia Tech research that involved a simplified ledger-fastening table. While this was legendary research for the deck industry and a much-needed addition to the IRC, there is an inherent challenge in the nature of codes based on testing: the codes will be limited to the exact method of testing, and any deviation will not yet be proven—it’ll be much more risky for an authority to approve. For ledgers, decks went from Wild West to dictatorship, and it’ll still take more time to iron out wrinkles and create a well-represented democracy when it comes to ledger codes.
After publishing the 2009 edition, the AWC noticed many of the ledger connection stipulations were not in line with the engineering standards in the well-recognized and IRC-referenced document, the National Design Specifications (NDS). So the AWC submitted a code modification that was approved for the 2012 edition, but ultimately it created a bit of a mess. The NDS is a document for engineered design, not tested design. Without testing, one must make conservative generalizations of past experiences and gained knowledge, and this is represented in engineering standards, such as edge distance for fasteners. Merging the tested results from Virginia Tech with the edge distance requirements of the NDS design document quite frankly resulted in code that is unusable to both builder and inspector, as seen in this video:
But that’s not the only ridiculous code requirement….
In the next article, we’ll look at lateral load requirements, and not on handrail but on the ledger and joist diaphragm.