Skip to content
Get a Quote
Insight

When the process changes, the air system is now oversized: a steel plant’s fix

Compressed air systems are usually designed once, for the plant that existed at the time. When the process changes — a new product mix, an upgraded line, a different operating rhythm — the air system does not adapt, and the mismatch shows up as air being blown off to atmosphere while the compressors run at their minimum.

A steel plant’s continuous casting operation went through exactly that. Five compressors at 400 m³/min each supplied four casting machines. A product-mix change led to a casting process upgrade, which lowered the air demand below the original design. Even with the compressors turned down to their minimum output, surplus air was being released, and energy was being spent to compress air that nobody used.

Why the obvious fix was wrong

The station had been built around large machines because the original load was large. Once the load fell, the problem was not capacity but granularity: five 400 m³/min units have a very coarse step size, and their minimum stable output was already above the plant’s demand.

The instinctive response — reduce pressure, run fewer machines — was tried first and did not solve it. A step change in demand can only be matched by a step change in capacity, and the station had no unit small enough to provide one.

How the right size was established

Rather than size a new machine from the process documentation, the operating team spent several days recording pressure, flow and the casting rhythm, then compared that record against what the casting department actually needed. The result was a data set of more than a thousand points, cross-checked with the user department.

That exercise produced an answer — 250 m³/min — that no nameplate or design document would have given. It is less than two-thirds of one existing machine, which is precisely why the station could not trim its output with what it had.

The operating change mattered as much as the machine

Once the smaller unit was installed, the gain came from how the station was run:

  • The new unit carries the trim while the large machines carry the base, so surplus no longer has to be blown off.
  • Instrument air is fed into the atomising air system where the two can be combined, using one system’s spare capacity to cover the other’s peaks instead of treating them as separate supplies.
  • Operating curves per casting machine were established, so the station can anticipate demand from the casting schedule rather than react to header pressure.

The reported outcome: specific consumption fell to 0.118 kWh/m³, described as the best the site had recorded, with annual electricity savings of 4.464 million kWh and a cost benefit of RMB 2.58 million.

The general pattern

This is one of the more common and least discussed inefficiencies in mature plants. It has a recognisable signature:

  • Air is being vented or blown off while compressors run at minimum.
  • Header pressure is held high because the station cannot modulate further down.
  • The station’s step size exceeds the demand variation, so demand tracking is impossible.
  • The process has changed since the station was designed — new product, new line, new shift pattern.

The fix is rarely a bigger investment. In this case it was one machine of the right size, plus the discipline to measure the demand before choosing it.

What to do before buying the trim unit

  1. Log pressure and flow for at least a week, covering the full production cycle including the quiet periods. The trim machine is sized for the difference between the base machines’ minimum and the actual demand.
  2. Check whether surplus is really surplus. Vented air can be a symptom of a control problem — a sequencing error, a stuck valve, a set point that no longer matches the process.
  3. Look for a second use for spare capacity. Feeding instrument air into process air, or vice versa, can absorb peaks that would otherwise need another machine.
  4. Restate the control logic afterwards. A station with a new unit and the old sequencing will simply run the new unit flat out.

The savings here came from a plant that had measured its own demand and matched capacity to it — not from a more efficient compressor.

—

Industry practice note. Figures as reported by industry press; the plant described is not a SIND project.

Planning a compressed-air upgrade? Talk to SIND’s engineers.

Facts as reported by industry press; the project described is not a SIND project.

Planning a compressed air project?

Talk to an engineer
SIND Applications Engineering

The SIND applications engineering team reviews compressed air duty points, sizes systems and writes the technical articles published on this site.

Published November 16, 2026.

SIND
Jafor Nayan Sales Engineer Biruk Argaw Sales Engineer