TECH · Technology2D and 3D — the distance between quantities · 1/3BuilderHub-R · Quantity takeoff

Why reinforcement is hard to count

Take one column. 500 × 500, 2,800 tall. Concrete is 0.5 × 0.5 × 2.8 = 1 m³, formwork is 0.5 × 2.8 × 4 = 6 m². Done. But for reinforcement the drawing says only “longitudinal 10-HD19, ties D10@200”. How many kilograms is that?

Published August 21, 2026·Last checked · August 2026

Key points

Concrete and formwork come straight out of the dimensions on the drawing. Reinforcement does not — the drawing gives how many bars at what spacing, never how many metres each bar runs
To fix a length you need to know where it is spliced, where it ends, how far it sits inside the concrete, and how far the ends are bent — and none of that is on a 2D drawing
So every takeoff assumes. The assumption is not an error; it is a decision about what 2D can compute
The problem is that the error does not run one way. In the same building a column comes out about 4% short and an irregular mat foundation about 10% over

1Concrete is multiplication, reinforcement is judgement

Take one column: 500 × 500, 2,800 tall. Concrete is a volume and formwork is an area, so multiplying the dimensions on the drawing gives the answer. Anyone who runs the numbers gets the same value.

Column 500×500×2800 — concrete 1 m³ and formwork 6 m² fall out of multiplication, while reinforcement leaves a question mark — **Concrete and formwork end in multiplication; reinforcement leaves a question mark.** The drawing gives how many bars at what spacing, never how many metres each bar runs.

Reinforcement is different. What the drawing tells you stops at “place ten 19 mm bars” and “wrap 10 mm bars at 200 mm centres”. Without a length there is no weight, and without a weight there is no quantity.

A rebar quantity does not start with arithmetic. It starts with a judgement.

Three terms, and that is all. Longitudinal bars — the heavy bars running along a column or beam; they carry load directly. Ties and stirrups — bars wrapping the longitudinal steel at a set spacing. HD19 / D10@200 — HD19 is a 19 mm deformed bar; D10@200 means 10 mm bars at 200 mm centres.

2Seven things the drawing does not tell you

Fixing the length of a single bar takes the seven items below. And not one of the seven is on a 2D drawing.

2.1① Where to splice — laps

Bars are not butted end to end; they overlap. The code sets the lap length, but where to splice is not on the drawing.

Two bars overlapped in a lap splice — takeoff applies laps uniformly, 3D judges continuity against the ordered length — **Bars are not butted end to end; they overlap. The overlap is extra steel.**

Site reality differs. Order 8 m bars and you splice once every 8 m. If the member ends at 5 m there is no splice at all.

2.2② Where it ends — continuity and development

Uniform development length at both ends of every beam, versus judging continuity and placing laps only where needed — **A 2D set is drawn one floor at a time, so continuity above, below and across cannot be judged from a single sheet.** So the safe side is chosen — everything is treated as continuous.

Deducting only the slab thickness versus deducting down to the wall girder changes the height of the same wall — **How much of the floor above is deducted decides whether the same wall is 3.35 m or 2.75 m.**

2.3③ Steel is smaller than the concrete — cover · ④ The ends are bent — hooks

A bar sitting too close to the concrete face corrodes and loses fire resistance, so it is held back inside the section. That margin is cover, typically 40 mm. And a tie is bent inward at its ends and hooked. Without the bend it unwinds inside the concrete.

2D computes the full 2.0 m column perimeter, 3D deducts cover to 1.68 m and adds the hook extension — **A 500 column does not mean the bar runs on 500.** About 30 cm per tie. And cover makes the bar shorter while the hook makes it longer — **the two run in opposite directions.**

The Japanese public standard says the same: the length of hoops and stirrups is taken as the perimeter of the designed concrete section, and hooks are taken as absent. It asks for the dimension of the concrete, not the dimension of the steel.

2.4⑤ The bar count is arrived at differently

A wall 10,390 long with vertical bars at HD10@150. Same wall, different count.

	Takeoff (2D)	3D (BH)
Calculation	10,390 ÷ 150 = 69.26 → round up to 70	(10,390 − cover 40×2) ÷ 150 = 68.73 → 69
Two layers, each face	140	138
Intersection reinforcement	duplicates removed, then +4 at the ends	already counted in the vertical bars
Total	144	138

Six bars in one wall. On one floor of this project there were 68 such walls. The clause in the Japanese standard reads the same way — divide the length by the spacing, round up, and add one.

2.5⑥ The shape is not rectangular

An irregular mat foundation with a sloped edge — concrete and formwork still fall out of multiplication, reinforcement does not — **Concrete and formwork for this foundation still come out of multiplication. Reinforcement is a question mark again.** A schematic redrawn from the dimensions of the original example.

In the real geometry every bar along the slope differs; 2D replaces it with a rectangle of equal area — **It becomes computable, and the geometry disappears.**

Along the sloped edge the bars run from 9.31 m to 14.98 m — 64 distinct lengths — **Reinforced in 3D, the bars run from 9.31 m to 14.98 m — 64 distinct lengths.** Replace the shape with a rectangle and those 64 collapse into a single average.

3D leaves every bar as its own row; 2D leaves only a division formula based on one side of the substituted rectangle — **The same foundation.** Above, the real geometry is kept and every bar stays as a row in the schedule. Below, it is replaced by a square of equal area and survives as a division formula. The “29.31 ÷ 0.36 + 1 = 82” in the takeoff sheet comes from here. *The 3D view is the actual model; the table has been retypeset from the takeoff content.*

2.6⑦ Some reinforcement is not on the drawing at all

Structural reinforcement 92–95% and construction-support reinforcement 5–8%, which never appears on the structural drawings — **Support reinforcement is decided by the fabricator’s experience at the shop-drawing stage.** It is not on the structural drawings, so it is not in the takeoff either. **It is still ordered.**

3So how far apart do they actually land

Back to that one column. The takeoff adds a 0.5 m lap to ten longitudinal bars for 33 m, times 2.25 kg/m, giving 74.25 kg. Ties are 1.12 kg each on a 2.0 m perimeter × 14 = 15.68 kg. Total 89.68 kg.

What is actually fabricated differs. Longitudinal bars are split between laps and hooks, with cover applied. The tie end zone is taken as max(storey height / 6, largest column dimension, 450), so 500 governs; a longer end zone extends the @200 region and raises the tie count. Total 93.00 kg.

	Takeoff (2D)	Bar detailing	3D (BH)	Difference
Weight of one column	0.0897 t	0.0930 t	0.0930 t	+3.9%

2D comes out 3.9% low. But the irregular mat foundation runs the other way.

Method	Quantity
Takeoff (2D) — replaced by a rectangle	1.968 t
Bar detailing — grouped to follow the geometry	1.789 t
3D (BH)	1.797 t

Column — 2D

-3.9%

falls short

Irregular mat foundation — 2D

+9.5%

overshoots

Within one building

opposite

one falls short, the other overshoots

“A 2D takeoff comes out a few percent higher” does not hold.
It differs building by building.

4So how has it been reconciled until now — the waste factor

Whatever could not be computed has been covered by a single ratio. The Japanese public standard writes it into the text — in arriving at the required quantity of reinforcement, a 4% addition to the design quantity is taken as standard.

And the method has largely worked. Sites did not stop; work rarely halted for want of steel. So no one went back and asked what the number contained.

But once it works that way, things of quite different kinds sit inside that 4% — what was assumed because it could not be computed, the offcuts genuinely lost in fabrication, and quantities missed altogether. What comes out when you open it is the subject of Part 2.

2D assumed because it could not compute. 3D does not assume, because it can.

Sources

Public Building Quantity Takeoff Standard (2023 revision) — clauses on laps, hooks and required quantityMLIT, Japan
BuilderHub-R product overviewChangSoft Global

This article is a summary compiled from published material. It is not a legal interpretation and does not constitute advice. Regulation changes frequently, so please check the original documents and the latest notices from the responsible authority before relying on it in practice.