Skip to content
Apex PreciCost Consultant

Where the missing 5% of your steel actually goes

8 min readApex PreciCost Consultant
  • Reconciliation
  • Wastage
  • Cost optimization

On most RCC projects, the tonnage of steel that arrives on site is not the tonnage the drawings called for. A gap of three to five percent is normal. A gap of eight to twelve percent is common enough that nobody blinks at it.

What is unusual is a project that can say, line by line, where that gap went. Most write it off as "wastage" at final account and move on. On a 500 tonne job at ₹62,000 a tonne, five unexplained percentage points is about ₹1.5 crore of steel that nobody has accounted for.

The five places it actually goes

Wastage is not one thing. It is at least five separate things with five different fixes, and lumping them together is exactly why the problem never gets solved.

1. Cutting loss

Steel is delivered in fixed stock lengths — commonly 12 m, sometimes 9 m or 6 m. Your bars are whatever the drawing needs. Every time a cut leaves a remainder too short to use, that remainder is loss.

This is the single largest component on most jobs, and it is also the most fixable, because it is decided at the scheduling desk rather than on site. A schedule written without reference to available stock lengths will generate offcuts no amount of site discipline can recover.

2. Over-ordering

Ordering per pour with a safety margin, then ordering again because the margin was consumed. The margin compounds: 5% added to each of twenty pours is not 5% overall, because the surplus from each pour rarely gets used on the next one.

3. Unrecorded variation

Design changes after the steel was ordered. The extra steel is real and legitimate, but if the variation was never recorded against the reinforcement quantity, it turns up in the reconciliation looking exactly like waste. This is often the largest misattributed component — a genuine cost that was recoverable and got written off because nobody documented it.

4. Offcuts that never come back

Usable offcuts are supposed to be returned to the yard and reissued for shorter bars, stirrups and starter bars. In practice they get buried, mixed across diameters, or scrapped. The steel was paid for at full rate and recovered at scrap rate, if at all.

5. Theft

It happens, it is a small fraction of the total on a well-run site, and it is almost always the first explanation reached for. Chasing it before the other four are quantified is a good way to spend a lot of effort recovering the smallest number on the list.

How to reconcile properly

Reconciliation is arithmetic. The difficulty is entirely in having the records.

Theoretical consumption comes from the approved drawings and the BBS, diameter by diameter. This is what the structure should have taken.

Actual consumption comes from delivery challans, material receipt notes and the site issue register. This is what was bought and issued.

The variance is the difference, and the whole exercise is attributing it:

Theoretical (from BBS)          412.0 t
+ Laps and splices               18.5 t
+ Approved variations             9.2 t
= Adjusted theoretical          439.7 t

Actual delivered                472.4 t
Variance                         32.7 t   (7.4%)

  of which:
  cutting loss (measured)        19.1 t   (4.3%)
  offcuts recovered              -4.2 t
  unreturned offcuts              8.8 t
  unexplained                     9.0 t   (2.0%)

That last line is the one that matters. It is the number you can actually act on, and you cannot even see it until the four lines above it have been separated out.

Do it monthly, not at handover

This is the part that changes the outcome more than any technique.

Reconciliation at the end of a project is an accounting exercise. It tells you what happened, after the point where anything can be done about it.

Reconciliation at every slab cycle, or at minimum monthly, is a cost control. If the cutting loss on floors 1 to 6 is running at 9%, that finding can be applied to floors 7 to 22 — where it is still worth something. On a twenty-storey tower, catching it at floor six recovers three quarters of the available saving. Catching it at handover recovers none.

What a realistic target looks like

For well-managed RCC work with schedules optimised against stock lengths, three to five percent total wastage is achievable. Not zero — offcuts are a physical consequence of cutting fixed lengths into varied ones.

If a project is running consistently above that, the cause is far more often the schedule than the site. Optimising cutting patterns against available stock lengths, rationalising bar diameters across repeating members, and reviewing lap strategy will usually move the number more than any amount of supervision on the deck.

You can put rough numbers on that gap with the savings estimator, or work the real figure out from your own drawings.

The short version

  • Wastage is five separate problems; measure them separately or you will fix the wrong one
  • Unrecorded variation is a recoverable cost that routinely gets written off as waste
  • Reconcile monthly, because findings are only worth something on pours that have not happened yet
  • Three to five percent is achievable; above that, look at the schedule before you look at the site