Almost every gravity load takedown starts with the same question: of all the floor up there, how much of it does this support carry? The answer is the support’s tributary area, and once you have it the rest is multiplication.
This post is the mental model I wish every junior engineer started with, plus the two or three places where the tidy textbook version stops being tidy.
The rain test
Here is the only picture you really need. Imagine rain falling evenly on the floor slab. Every drop runs downhill to the nearest support. The patch of floor that drains to a given column or beam is that support’s tributary area.
On a flat slab there is no real slope, so “nearest” is geometric: a point belongs to whichever support it is closest to. Draw the boundary halfway between each pair of neighbouring supports and you have carved the floor into one region per support. Structural engineers call these lines the midlines; a mathematician would call the result a Voronoi diagram. They are the same thing. The equivalence is used well beyond floor slabs: a Journal of Structural Engineering study by Gierson, Phillips, Duthinh and Ayyub assigns tributary areas to irregularly spaced wind-tunnel pressure taps the same way, concluding that Voronoi diagrams are optimal for tributary area assignment because they give a unique answer independent of whoever draws them.
That is the whole idea. The load on a support is:
load = floor pressure × tributary area
Floor pressure is your dead and live load in kPa or psf. Tributary area is geometry. Keep them separate in your head and the takedown stays clean.
The regular-grid shortcut
When columns sit on a rectangular grid, the midlines are just horizontal and vertical lines at the half-bays, so the tributary area becomes a rectangle you can read straight off the plan.
| Column position | Tributary width (per direction) | Share of a full bay |
|---|---|---|
| Interior | half-bay on both sides | full bay |
| Edge | half-bay one side only | about one half |
| Corner | quarter of the surrounding bay | about one quarter |
So for a 6 m by 5 m grid, an interior column carries 6 × 5 = 30 m², an edge column about 15 m², and a corner column about 7.5 m². Those ratios of roughly 1, 1/2 and 1/4 are worth memorising. They are the fastest sanity check you have: if your interior column reaction is not in the right ballpark relative to the corners, something upstream is wrong.
This is the case where hand methods and software agree to the last decimal, because the geometry is genuinely simple.
Where hand sketches quietly go wrong
The trouble starts the moment the plan stops being a clean grid. Three situations cause most of the errors I see:
1. Irregular spacing
If the bays are not equal, the midline still sits halfway between two supports, but now every column gets a different rectangle and it is easy to mis-measure. A column flanked by a 4 m bay on one side and a 7 m bay on the other carries 2 m plus 3.5 m, which is 5.5 m of width, not the 5.5 you would get by averaging only if you remember to halve each side independently. People average the wrong things under time pressure.
2. Skewed or non-orthogonal grids
When the column lines are not at right angles, the tributary regions become slanted polygons, not rectangles. The midline construction still works perfectly, but you can no longer read the area off as width times depth. You have to actually compute the polygon area, and a hand sketch turns into a coordinate-geometry exercise.
3. L-shapes, re-entrant corners and openings
Around a stair core, an atrium, or the inside corner of an L-shaped floor, the nearest support for a given patch of slab can be surprisingly far away, and the tributary polygon wraps around the opening in a way that is genuinely hard to draw by eye. This is the classic place where two competent engineers sketch two different answers, and neither notices until the column sizes disagree.
The honest summary: the midline method is always correct, but doing it by hand is only easy when the geometry is easy. The further the plan is from a clean grid, the more the error budget moves from the physics into your sketching.
A worked sanity check
Say you have an interior column on a 6 m by 5 m grid, carrying three identical suspended floors plus a roof. Each floor has a dead load of 4.5 kPa and a live load of 3.0 kPa, and the roof contributes 1.5 kPa.
The tributary area is 30 m² at every level. Per floor:
- dead: 4.5 × 30 = 135 kN
- live: 3.0 × 30 = 90 kN
Across three floors that is 405 kN dead and 270 kN live, plus a roof of 1.5 × 30 = 45 kN. Before any load combination, the column is already gathering roughly 720 kN of service load. Apply a gravity combination such as 1.2 dead plus 1.6 live and you are comfortably into the hundreds of kilonewtons that decide the column size.
Notice how little of that calculation was hard once the tributary area was settled. That is the point: the geometry is the part worth getting exactly right.
Why we compute it instead of sketching it
StructLoads draws the midline partition for you, including the awkward skewed and L-shaped cases, by clipping each support’s region to the actual floor outline. You place the supports, and the tributary polygons resolve in real time, with an area-balance check that confirms every square metre of floor has been assigned to exactly one support. No square metre counted twice, none lost.
If you want to see it on your own plan, open the studio and drop four columns on a blank floor. The fastest way to understand tributary areas is to drag a column and watch its region grow while its neighbours shrink.
From here, two natural next steps: how those per-floor loads stack down through a building in a floor-by-floor worked example, and when the simple rectangular split needs rethinking in one-way versus two-way slabs.
Quick answers
What is a tributary area in simple terms?
It is the portion of a floor whose weight is carried by one particular support. Imagine rain falling on the slab: the tributary area of a column is the region where the water would run to that column rather than to any other support.
How do you calculate tributary area for an interior column?
On a regular grid, take half the distance to the column on each side. For a column on a grid with bays of 6 m by 5 m in the two directions, the tributary area is (6/2 + 6/2) by (5/2 + 5/2), which is 6 m by 5 m, or 30 square metres.
Why do edge and corner columns carry less load?
Because there is no floor on one or both sides to contribute. An edge column draws from a half-bay on the missing side, and a corner column draws from a quarter of the surrounding bay, so they pick up roughly half and a quarter of an interior column’s area.
Related guides
More on tributary areas: the 45-degree rule for tributary area, tributary areas on an L-shaped floor and tributary area vs influence area. It also covers tributary area for a wall.