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

Crawlspace Beam Layout

alexallo | Posted in Energy Efficiency and Durability on

Hi Friends,

Long time reader, first time poster. I’m looking for some feedback on my plans on fixing some floor bounce/sag.

House layout is pretty simple, 62×32 with one corner consisting of a 20×20 slab garage. Floor Joists are 2x10s, spaced 16″ on center. They span almost 16′. The house has long been neglected, and the crawl was filled with moisture for years. Poor landscape drainage, washing machine leak, and a dryer vented to the crawlspace led to sagging floor joists and a main beam with rotten ends. Plus plenty of mold.

The current design meets code, and I live in a rural area with almost no building code requirements. Rather than trying to sister damaged joists and jack all of the sagging areas, I’ve decided to cut the span in half by adding two more beams. Where I could really use some input is on beam sizing, and the number of piers. I’ve got about 30″ of head room, I’d like to do the least digging possible for footings. Right now I’m leaning towards using steel I-beams. I think I can manage to get a 20′ long section in through the current door, but might have to cut back to 15′. With this being an above and beyond repair over code, and a project that won’t support a high resale value, I’ve decided not to seek an engineered solution from a PE. The main goal in going with the steel I-beam is to limit the number of posts necessary for support. Any feedback or suggestions on my plan? Is the steel worth it, or should I just stick with a double 2×10? Pier spacing suggestions?

Layout is attached. Thanks!

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. seabornman | | #1

    A free program that lets you size joists and beams is Forte by Weyerhauser. Takes a little learning, but very valuable. Of course it's all wood based products, so no steel. There are some free steel beam calculators on-line also. You can figure out footing sizes from building code. The International Residential Code is available on-line.

  2. Yupster | | #2

    Forte is excellent, if you know how to use it properly. You could size your beam from the tables in the Ontario Building Code. Scroll all the way to the bottom of Part 9 of Division 2 and you will find tables for built-up wood beams and steel beams. The spans are in meters, so if you are a Yankee then better bring a conversion calculator along. Tried to link to a copy of the OBC but I got blocked as spam.

  3. Expert Member
    Michael Maines | | #3

    Sizing steel beams is trickier than sizing wood beams--for each nominal dimension of an "I beam" (technically called a W-section, the kind most often used because it has the strongest overall strength to weight ratio) there are a wide range of weights per foot, which mean the strength or stiffness varies a lot. I know how to size steel beams and sometimes do for my own projects, but I don't know any rules of thumb or books with span tables as there are for wood framing. There are a lot of variables. An undersized steel beam is no better than an undersized wood beam. The longer the span, the bigger each footing needs to be. If the footing is too small, the soil will compress and the beam will lose contact with the joists. There are several ways to attach a steel beam to a post (and to the joists) but it depends on the situation, the type of post, etc. An engineer will be able to run through these calculations quickly (likely using software, though your situation is pretty straightforward), will have standard details and ideas of costs in your area. Some steel fabricators have engineers right on staff.

  4. bennettg | | #4

    At the risk of under thinking this, I'll suggest an alternate approach.

    I'm thinking the issue is that your floors are "bouncy" and sag some in the middle of the span. They likely are at the span limit for 2x10x16'@16". I'm assuming you're addressing the moisture issues - bulk water getting in and ground moisture coming up.

    _If_ you're comfortable that the floor is structurally sound and the goal is to reduce the bounce and level up the floor, I'd use wood beams (either 3-2xnn as MM suggests or 2-2xnn+1/2" as with a header, maybe breaking the 40' length under your partition walls, jackposts at a spacing that worked and a 2x2 stack of 4x8x16 solid cinder blocks or precast post footers.

    You're going to have to tweak the jacks to level out the floor over some time anyway as the floor structure adjusts.

    I did exactly this on a smaller scale for a kitchen floor in an old house that was a little saggy and bouncy for my liking, using a single 2+1/2" 2x10x~10' "header" supported by one jackpost in the middle.

  5. Expert Member
    Michael Maines | | #5

    Alexallo, since you're a long-time reader (and I'm procrastinating) I ran some calcs for you. I'm not a licensed engineer or architect so this is conceptual only. If it was my house, and I was going to follow your strategy, I would consider spacing piers about 16' apart, using 24" x 24" footings, and choosing from the following beams, to get L/480 deflection (or better):
    (3) 2x12 (a bit undersized for that deflection but over L/360)
    (2) 1 3/4" x 14" LVL (MOE=2.0)
    (3) 1 3/4" x 11 1/4" LVL
    W8x24 A992 (or A36) steel beam

    If you change the pier spacing, obviously these calcs are wrong. If there are structural loads on the floor, these calcs are wrong. Basically just assume that these calcs are wrong, but maybe enough to point you in the right direction.

  6. alexallo | | #6

    Thank you all for the answers and feedback. I'm working on mastering the Forte software, I really like it and appreciate the reference.

    Michael, I appreciate all of the leg work. I was just about right on my footing plan, but undersized on the beam. A 20ft W8x24 is a 500lb beam, no way I'm getting that set in place without a huge team. I think I'll have to compromise by more piers so the span is shorter.

    Bennett, using a couple standard blocks was my first plan. But the previous owner tried it and they settled/slanted and bent a jack post. That was fun to remove. The joists meet design under 10lb dead, 30lb live at L/240, so a stretch but technically acceptable. We've got granite in the kitchen, but it falls under the ~11ft span behind the garage. The wife and I love stone and are planning on expanding the master bath and putting in marble. The current joists definitely won't hold up to that, so I'm thinking the replacement beam with poured footings is my real solution.

    When these repairs are completed, I'll be encapsulating the crawl. Then haven't decided on using hvac for humidity control or a dehumidifier. Leaning toward the dehu, but that's a small issue compared to my current repairs.

  7. Expert Member
    Michael Maines | | #7

    Alexallo, I made a mistake--you can span up to 20' with W8x10s, a much more manageable weight. That's a good example of why you shouldn't trust internet engineering.

    At 20' your footings should be bigger. At that span there is 8,000 lbs on each footing so it's not a trivial load. Again, if you want to pursue this route, I would strongly suggest investing a few dollars in a design professional who can make sure that what you are doing is safe and not a waste of time. I would not want to install marble (or other expensive, fragile materials) without knowing that my structure was safe. You have to start with a good foundation.

    By the way, the IRC code minimum for floors other than bedrooms is 40 psf and L/360 deflection. All but the worst grades of framing lumber meet those specs at your spans--again, assuming there are no additional loads on your joists.

  8. CMObuilds | | #8

    Michael, where are you getting your steel beam load tables from? Mine says for a 20' span 40/15 you need a w8x18, w8x10 will go 16', 36 ksi steel. 8' tributary width.

    Alexallo, I'm assuming you abated the mold, step one.

    Step 2, address your rotting main beam ends. Even if you installed 2 midspan beams to carry the entire floor load they wouldn't support an 8' cantilever, so you can't ignore your existing main beam.

    40 lb live load and 15lb dead with an 8' tributary ignoring the end walls is 496 square feet at 55lbs per square foot or 27,280 lbs or 440 pounds per linear foot your beam needs to pick up. Size beam and desired spacing to that if you must. A 2 ply 11 7/8" 2.0 LP LVL passes total load at a support span of 16'6" L/360. You want better deflection shorten the span (I would, L/480 min, pref. L/720, I think tile TCNA calls for min. L/720 for most installations). There's some other details to it I wont get into but you already have a supported floor so go crazy if you want.

    Michaels advice regarding footings isn't real accurate, longer spans dont necessarily mean bigger footings, footing design depends on your bearing soil type and obviously total load on the footing.

    Since you're in a crawl I would make it easy on yourself and put in more footings for shorter spans and handle LIGHTER smaller beam material. You can form and pour right on top of the dirt, you dont necessarily need to dig it in unless the soil has been disturbed.

    For sizing steel I think Steel Alliance? has tables you can download for steel beam capacities, all LVL manufacturers publish span charts and material properties. Honestly, you could do some way hokey stuff and it will never give you a problem that I've seen in houses I go into that amazingly have worked just fine for many many years, but I wont divulge.

  9. Expert Member
    Michael Maines | | #9

    T Carlson, from the AISC Manual of Steel Construction, a textbook from college a long time ago, for allowable uniform load tables for 36 ksi steel. (Of course also ok for 50 ksi steel.) I also calculated deflection based on the design load; I don't recall what it was but I'm sure it was below L/480--although L/360 will meet code, at that span it may feel bouncy--if you're going to that effort, you might as well end up with a solid-feeling floor. I used a 10lb dead load because that's almost always enough.

    I did not check Fb, but a bit of interpolation on the maximum allowable moment chart for unbraced spans made it look like it should be fine there too. I know that's not the correct way to do it but it's accurate enough for this situation. I don't have or know where to get standard span tables for steel beams. I was trying to get the OP into the ballpark and made it clear that my calculations should not be relied upon. I also tend to do my own calculations for LVLs and dimensional lumber, as for me it's faster than looking at the code book, but whenever I compare results it's always fine.

    Regarding footings, I suppose I could have added, "With all other conditions remaining the same, expanding the span increases the load on the footings" but that seems like common sense. I had done rough calculations at a 16' span and he mentioned increasing that to 20'. At 16' on 1500psi soil (the worst common soil, but commonly used as a boundary condition) he would need roughly 24" square footings. At 20', he would need just under 30" square footings. And again I made it clear that he should talk to a licensed engineer in person.

  10. alexallo | | #10

    T Carlson, all of the tables I've seen are based off of end walls with one main beam. I didn't think those applied, so I've tried to do the real math. I'm not an engineer, but I do have an engineering background and enjoy learning how to find the real answer, not just a rule of thumb.

    I've attached the calculations I did on Forte. Assuming I've entered all the data correctly, it calculates 634 pounds to support on each joist contact. For beam sizing I'm thinking I can convert that to linear lb/ft by dividing by 1.33 (16" on center to 1ft). That gives me 476 linear lb/ft. Note on my calculation, roof loads aren't included. There are no internal load bearing walls, and I didn't feel it necessary to calculate the point load of the exterior walls for this.

    I moved my pier spacing to 10' to minimize the beam size. The Forte beam sizing gives me such a massive requirement that I feel like I must have made a mistake in my data input somewhere. I also noticed that it gives me deflections at L/999+ which makes me think I did something wrong. If it's L/999+, it's overbuilt. Especially on this price level of house. The current main beam is only a triple 2x10, and it spans in 8ft sections.

    For the footings, they need to be deep. Very loose dirt. I've been replacing the top plate along the rim due to termite damage, and I can sink a standard block 4" down with a jack lifting up 4 or 5 joists. Of all the things that I think could stand to be overbuilt, footings top the list in my book. Every house on my street has foundation issues/repairs. I think Michael was about right on his rough sizing. With spacing 10ft, I'm thinking I'll go back to my original plan of an 18" round form 24" deep.

    I can't say thank you enough for all the input and help. I enjoy learning all about these solutions. I'm sure I won't enjoy implementing them, but this part is fun!

  11. user-6623302 | | #11

    Have you checked to see if the joists are bowed? Jacking up the middle of a bowed joist will not make it flat. You are planning a lot of work in a small space. I had a bad floor system over a crawlspace in my kitchen. I pulled out the whole floor system and reframed it. If I were you, I would rethink the plan. What will take you all day crawling around in the dirt will take you an hour if you can stand up. A new floor frame would allow for proper insulation and air sealing as well. A all round better job.


  12. Jon_R | | #12

    > For the footings, they need to be deep.

    Seems to me that you need to keep digging until you get to undisturbed soil and then measure the bearing capacity. You can't pre-determine a depth or optimal footing size. On the other hand, you can re-shim if the footing sinks a little as disturbed soil compresses.

    In your case, any added beam will make the floor stiffer/flatter and be safe. So calculations are useful, but not critical.

  13. alexallo | | #13

    Jon R, good points. Though if I get down to 24" and still have soft soil, I'll likely reach the end of my patience and pour a footing anyway. I replaced a bunch of subfloor when I remodeled the kitchen, it sure did make repairs in that area easier. I've thought about pulling up a single sheet over each spot that a footing needs to be poured. Even thought about trying to fit an auger in over that area, but honestly the crawl space isn't as miserable as many I've been in. Plus I'm looking at 10 footings at this point, that's a lot of subfloor. I've got some teenagers at church looking for work, it'll teach work ethic. And that $10/hr isn't as exciting as it sounds.

    All, some pictures for reference if interested.

    Jon, they are both bowed and cupped in most areas, but not significantly. Except for the one the dryer vent blew on for a decade. I'll have to sister that one to level. And several that I've worked with have bent back with jack posts without lifting up too much on the ends. I'm only worried about perfectly level over the bathroom with stone going in. The other two will have 12"x12" tile, and these are close enough to level that with a healthy backer board and a bit of leveling compound, they'll be fine. The rest of the house will be #2 quality solid oak hardwood. It'll be thicker than the subfloor, should go down just fine.

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |