“It was the best of decks, it was the worst of decks. It was a deck built wisely, it was a deck built foolishly.” ¹
Recently, I had the opportunity to examine a deck damaged by a tree. Well, “damaged” is not the right word, more correctly the deck was demolished by a tree. Here it is, on the ground:
Wow. This deck is toast; it’s just a pile of lumber. Luckily, no one was on or under the deck when a wind and snow storm crashed a tree on it. Did this have to happen? Could the deck have survived? We cannot know for sure, but I’d say: Yes.
Really? “Yes” is a bold statement. Can you back that up?
Let’s look at the facts.
The deck was constructed a few years earlier in Wilmington, MA. First, the good news. The frame consisted of 2×8 PT joists, 16″ on/center over a triple 2×10 beam. OK. The joist span was 10 ft., which is proper. Two 4×6 columns supported the beam and created an 8’6″ span – which is appropriate for that size beam with a design load of 390 pounds per linear foot or plf.² Good. It had galvanized joist hangers that fastened each joist to a 2×8 ledger at the house. And the ledger was attached to the house with some nails and some ledger screws. Partially good.
Before we get to the bad news about this frame, let’s consider another deck.
Several months earlier, a client hired us to repair a deck we had built 24 years ago – which, ironically, was also damaged by a tree. We’ll call this deck the Westboro deck. The situation in Westboro was very similar to Wilmington’s: a wind storm toppled a tree that fell onto the left side of the deck, just over the end of a main support beam.
The trees were surprisingly similar: Westboro’s was an oak tree, 16″ in diameter, about 70 feet away. The Wilmington tree was actually a big branch, about 14″ in diameter, from a pine tree about 80 feet away.
We cannot know the exact force of either impact. But the trees were similarly sized and each tree smashed through the rail and hit the side of the deck over the beam.
The Westboro deck that we built survived. Indeed, the oak tree broke over the edge of the deck — like a branch breaks over your knee. And the deck remained proudly erect. It needed some repairs: We replaced two joists, a brace, one beam member (the outer 2ft was broken), and of course the rail section and some floor boards. Here it is with repairs nearly completed:
But the Wilmington deck was demolished beyond repair. Why?
Let’s look closer. Wilmington’s structural members were adequately sized and spaced. But they were poorly connected. The collapsed frame members tell the story:
The structural members did not break. The joists did not break. The beam did not break. The columns did not break. What broke was the beam to column connections. The beam balanced on top of the columns and was merely toe-nailed to each column with 3 nails.
Do 3 nails seem proper?
No way, you say.
You’re right. I expect Mother Nature laughed at those 3 nails. Here’s a drawing of that post and beam connection before it collapsed.
And here’s what the current Building Code requires:
The Code allows either the “Notched Post” or the “Post Cap” method to “resist lateral displacement.” Yeah, in Wilmington Mother displaced the deck beam right onto the ground.
Does that “notched post” look familiar? Recall the Westboro deck:
In 1993 we notched the 6×6 support column, set the double 2×10 beam in that notch and bolted them together with two 1/2″ galvanized through bolts. That was two decades before the Code required it. And then we further braced that connection with a pair of 2×6 Y braces, screwed to the beam and column. It’s good we did — this combination withstood the direct attack of that angry oak tree.
Without such a solid connection, the Wilmington deck collapsed. Literally, like a house of cards.
Another factor may have contributed to this collapse: the deck ledger’s connection to the house. Its 2×8 ledger was fastened directly over the copper flashing. This connection is responsible for supporting almost half of the deck’s design load — the dead load of the structure plus the live load of people and furniture — but no visiting trees. Still, that design load on the house would be almost 3,500#. It was fastened to the house with nails (a building code violation), augmented with Fasten Master ledger screws (proper). But there were only 8 ledger screws, where at least 13 were required,³ and half of the nails were rusted.
If the deck ledger was properly attached to the house, would the deck have survived the tree’s attack? No. That toe-nailed connection doomed the beam. Might the deck have remained attached to the house? I suspect not, but we cannot know.
The Building Code in effect when Wilmington was built did not address beam/ column connection. So technically, that balanced, toe-nailed beam did not violate the Code at that time. It does now.
The moral of this tale of two decks? Comply with the Building Code. Indeed, exceed the code — for a safer deck and a longer lasting deck.
- I stole and twisted this quote from the opening lines of Charles Dickens’ A Tale of Two Cities, an extraordinarily popular novel set in 19th century Paris and London.
Design Load: Snow Load of 50 pounds per square foot, psf, plus Dead Load of 10 psf. Load on beam thus = ((10’/2) + 1.5′)) * (50+10) psf = 390 pounds per linear foot, plf. Here’s the framing plan of the deck before collapse:
We at Archadeck of Suburban Boston offer professional design and build services for clients west and north of Boston. Over the past 25 years we have designed and built over 950 projects. We have enhanced the depth of our expertise by limiting our work to decks, porches, sunrooms, and patios. To view some of these projects, visit our website. To learn how we treat our clients, check our Super Service Awards on Angie’s List or read a recent article about us in Remodeling Magazine. For a free design consultation and a relaxed and rewarding design and construction experience, contact us via e-mail, email@example.com or by phone, 781-273-3500.
© 2018 Advantage Design & Construction, Inc.