Please note: This older article by our former faculty member remains available on our site for archival purposes. Some information contained in it may be outdated.
A low-cost alternative to poured concrete foundation walls.
Most light-frame construction projects follow a similar sequence of events: the perimeter of the planned structure is marked with stakes, soil within the defined area is excavated to a depth that is safely below the frost line (4’0″ minimum in my area), foundation walls are formed, concrete is poured, and then a wood-frame structure is erected on top of the poured foundation. Excavations live a short life. They are dug only to be filled – with thousands of dollars worth of concrete. This building practice is at times a necessary. But often, filling an excavated trench with concrete is nothing more than a bad habit. There are less costly alternatives. Using a wooden grade beam is one option that saves time, money, labor and resources.
The design goals for a recent project I worked on were clear from the start: build a detached 2-car garage that includes ample room for storage. Budget was an important consideration. So we designed a 3-car garage with one bay closed to the outside forming a storage bay. The future conversion of this storage space to a third stall would be simple if it was planned in advance. After considering several preliminary sketches, we decided to build a simple 32′ x 24′ structure with a hip roof .
A pragmatic solution to the budget constraint hatched while developing a cost estimate for the project. The estimate for materials(no labor) weighed in at $6,500, but the first 3 items on the initial checklist tipped the balance: excavation/backfill = $450; poured perimeter footing and 8″ thick frost wall(4’height) = $2,100 and 24′ x 32′ concrete slab = $750. More than half the cost of all materials required for the garage was buried in a hole. The question followed: “Do we really need to bury 25 yards of concrete and $3,300 before we even get started?”
A second option seemed attractive. Pouring a series of concrete piers to grade, bolting a triple nail-laminated pressure-treated beam (2×8’s) onto the tops of the piers and laying a 6-inch thick dense-grade stone-dust pad as a garage floor trimmed $2,100 from the cost. The estimate for the alternative plan:
excavation = $200
twelve-inch diameter poured piers(13) = $120
builder’s tubes, 2×8 pressure treated lumber, 1/2″ threaded rod = $300
15 yards of dense-grade stone dust = $180
shovel labor not required in the first estimate = $400.
I reasoned that a grade-beam system would force us to concede lower quality. The grade-beam system would allow vermin and weather to penetrate the garage; the alternate plan would be less durable and unfamiliar; stone dust wouldn’t supply a hard durable surface to drive on.
But one by one the trade-offs fell as insignificant. I’ve never seen a vermin-proof garage. If the grade-beam was laid directly “on grade” and the inside was filled with stone dust it would be reasonably weathertight. Pressure-treated lumber is guaranteed to last more than 40 years in contact with soil. And if the owners tired of the stone dust pad they could pour a slab. However, I’ve been impressed with the dense-grade stone dust I’ve used as a driveway top coat.
System #2 seemed like a good bet.
It is critical to determine a pier-and-beam system’s ability to redistribute the structural load to the soil. And equally important is determining the soil’s ability to support the load.
We lucked out. The soil is course-grained gravel with plenty of large rocks mixed in. It’s the best soil when it comes to foundation systems (the worst when it comes to driving in batter-board stakes!). It supports 5 tons per square foot, resists frost action and provides excellent drainage. This type of soil forms a stable base because it is not likely to change volume as its moisture content changes. It also resists settling.
Knowing what load the soil will carry, logically leads to the all-important question: What total load are we asking the soil to carry? Here we are concerned with dead load, live load and in the Northeast region, snow load.
Live loads, the weight of transitory loads like people or lawn tractors, is not factored into this calculation because the weight of people, automobiles and stored garden equipment does not affect the bearing of this structure. These loads will bear directly on the stone-dust pad. However, dead loads, the weight of the building materials themselves, and snow loads must be considered.
The total dead load of the roofing material, sheathing, ceiling joists, rafters, walls, grade beam and concrete piers are additive and ultimately bear on the soil through the foundation system. Their weight must be factored into the design. It’s easiest to calculate total load in two parts: first roof loads and then loads for the remainder of structure.
Many handbooks list loading allowances for various structural assemblies. But Western Wood Products Association (WWPA) sells a span computer slide-rule for $2 that is especially handy when calculating loads of beams, joists and rafters. For instance, WWPA lists high-slope roofs (over 3/12 pitch) with no finished ceiling as having a dead load of 7 psf when light-weight roofing is applied to the roof deck. That is the appropriate allowance for this project, since I planned conventional asphalt shingles. I also carry an additional 10 psf as the roof-system’s dead load for ceiling joists with limited attic storage, yielding a total dead load of 17 psf for the roof assembly. In other words: the entire roof/ceiling assembly weighs 17 pounds per square foot of floor area (horizontal projection of roof).
Snow load is determined by the local building code. The amount and weight of snow varies dramatically from region to region. In my area snow load is listed as 35 psf. All loads are based on the square footage of the structure’s horizontal projection. That is the footprint of the structure, not the actual square footage of the roof deck surface. WWPA treats snow loads as live loads on its slide rule. The roof load for this garage (combined dead and snow loads) equals 52 psf —– OR —– 52 psf x 24′ x 32′ = 39,936 pounds (total load).
Several elements contribute to the remainder of the structural load: wall sheathing, siding, wall frame, grade beam and concrete piers. Weights of various building materials are listed in several handbooks. The Architectural Graphic Standards is a good source of material weights. It’s easiest to calculate the weight of 1 square foot of wall and translate that figure to the weight per running-foot of wall. The dead load of the garage walls, beams and piers totaled 13,650 pounds, bringing the total load of the garage to 53,586 (roof load 39,936 + remaining load 13,650 = 53,586 pounds).
So, how many square feet of soil is required to support 53,586 pounds if every square foot of soil can support 10,000 pounds? (remember that the soil on this site could support 5 tons per square foot.) About 5 1/2 square feet of bearing surface is needed to hold the weight of the garage.
Taking inventory of the facts – soil bearing capability, total loading and size of footing required – it becomes apparent that we need one more piece of information to finish the design calculations: the allowable span between concrete piers. I made some assumptions before determining the allowable strength or stiffness of the beam. I decided to use a triple 2″x8″ CCA-treated southern yellow pine grade-beam for two reasons. The top of a triple 2×8 beam would stick up above the 6-inch thick pad and it would provide the lateral rigidity needed at the base of the wall.
The load transferred to the grade-beam is not uniformly distributed because hip roofs deliver more weight to the center portion of their supporting walls (along its length) where the peak is highest and roof area is greatest. There is less roof load transferred to the wall as you approach its ends. Considering the heaviest loading point, the WWPA span computer indicates that a triple-laminated No. 2 grade southern yellow pine beam could easily span 7′ FOR THE LOADS IMPOSED BY THIS STRUCTURE. WWPA does not write the grade rules for southern yellow pine, the Southern Forest Products Association (SFPA) does. However, SFPA design values can be used with the span computer to calculate the allowable maximum spans.
Pier spacing is determined by the beam’s ability to span a given distance: in this case 7’0″. A total of 13 piers are required. Five piers are needed along the 32′-long back wall, but only four piers are used along the front since the design calls for 3 garage stalls. Nail-laminated 2″x 10″ headers carry the load over the 9-foot wide garage openings. Two piers were placed on each side between the front and back corners.
Footing size is based on anticipated load and soil bearing capability. Since 5 1/2 square feet of bearing area was required, it was determined that the base of each pier must cover at least 0.43 square feet of soil. Nine-inch diameter builder’s tubes would do the trick — even without footings. Instead, I used 12-inch diameter tubes (0.79 square foot area) imparting my own factor of safety.
Putting It All Together
Fabrication is the acid test for any design. And as expected this design worked neatly.
Location and orientation were arbitrary, since the garage sits 50 feet from the house. Nevertheless, the front wall of the garage was laid parallel to the front wall of the house. The four corners of the garage were conventionally squared using diagonal measurements and the sides of the structure were laid out using batter boards and line. The line marking the location of the front wall was strung first and used as a reference to orient the remaining walls. Stakes were driven at appropriate points along the lines marking the centers of each pier. The height of the perimeter lines was not adjusted to establish the finished height of the piers. They were only used to establish the direction and squareness of the rectangular shape.
The lines were temporarily removed while a backhoe excavated the 13 holes to a depth below the frost line. Once the holes were excavated, lines were restrung to define the exact shape and size of the garage.
Builder’s tubes were carefully leveled and positioned at the four corners first. The tubes were left long enough to be leveled down from the intersecting lines that formed the footprint of the garage. The disadvantage of using builder’s tubes instead of formed walls is that you have to backfill tubes by hand. This took 16 man-hours to backfill. Its easiest to set the tubes if you place and tamp several inches of fine soil around the bottom of the tubes. This gives you a chance to accurately secure the bottom of the tubes before backfilling. Intermediate tubes were located and set after backfilling the corner tubes. I use a spacer stick to gauge the location of intermediate piers. Once all the tubes were backfilled, the grade-beam height was established and the tubes were cut to the appropriate height.
The fastest, easiest and most accurate way to establish the grade beam’s elevation is by shooting a transit or dumpy level. Fortunately, this building site was fairly level. The elevation of three corners was within a couple of inches. The grade at one of the back corners dropped off about a foot. I cut the first builder’s tube flush with the level of soil at the grade’s highest point. The top of the severed tube became my benchmark. Based on transit readings, I cut the four corner tubes and stretched lines between the corners to mark the heights of the intermediate tubes.
The rest of the project was straightforward: the tubes were filled by the ready-mix truck; anchor bolts(that extended 8″ above the pour) were set; pressure-treated 2″x8″ ‘s were spiked together; the grade beam was bolted down onto the piers; and the garage frame was erected on top of the beam. A 6-inch deep layer of stone dust was spread within the perimeter of the beam, sealing the inside space. A water table trim detail made of pressure-treated stock was applied to the base around the outside of the wall. And the exterior soil was graded to meet the bottom edge of the water table trim detail.
All-in-all the grade-beam system works well and is cost efficient. It certainly doesn’t lend itself to all designs and applications. But the system provides an alternative for low-cost construction. I’m currently designing an insulated version for a northwoods camp where access to heavy equipment is limited.