Energy Sub-Metering: Find the Hidden Problems in Your Plant

Compressed air is the most expensive utility per unit of useful work in most Malaysian factories, and the one almost nobody meters. A mid-size plant running a single 75 kW screw compressor around the clock burns on the order of 500,000 kWh a year, enough electricity to add a six-figure ringgit line to the annual bill before you count the maximum demand it locks in every month. The uncomfortable part: the US Department of Energy estimates that 20–30% of that air leaks straight out of the pipework, doing no work at all.

That gap is where compressed air energy efficiency lives. It is rarely a matter of buying a better compressor. It is a matter of seeing where the energy goes and stopping the leaks, the over-pressure, and the misuse that a plant runs blind to.

Why compressed air is so expensive to make

Compressing air is thermodynamically wasteful. As little as 10–15% of the electrical energy fed into a compressor reaches the tool as useful work; most of the rest leaves as heat (much of it recoverable, but rarely recovered). The Carbon Trust and DOE both put compressed air among the least efficient conversions in a plant.

The practical consequence: every cubic metre of air you make and then waste costs several times what the same energy would cost delivered as electricity or steam. So the cheapest compressed air is the air you never had to compress.

The three places the energy actually goes

Across most factories the waste concentrates in three areas:

- Leaks. The single biggest loss. DOE field studies put untreated leak rates at 20–30% of total output. A 3 mm hole at 7 bar leaks continuously, 24/7, whether or not the line is producing anything. Leaks are invisible on a P&L and audible only if you go looking with an ultrasonic detector.
- Over-pressurisation. Every 2 psi (≈0.14 bar) you run above what the tools actually need adds roughly 1% to compressor energy. Plenty of plants sit at 7–8 bar because "that's how it was set," while the process only needs 6.
- Artificial demand and misuse. Open-blow cleaning, unregulated cooling, and drops left cracked open create demand that has nothing to do with production. It reads as load; it delivers nothing.

None of these show up on the TNB meter as "compressed air." They show up as a base load that never drops, which is exactly why they survive for years.

What it costs on a Malaysian tariff

Under TNB's RP4 tariff structure, a compressor room hurts you twice. First on energy (kWh): a leaky system runs the compressors longer and harder. Second on maximum demand: because compressors run continuously and cycle under load, they prop up your monthly peak, and the RP4 Capacity and Network charges bill that peak in RM per kW every month regardless of how little useful work it produced. (If you're still working out how that peak is set, see our guide on cutting TNB maximum demand charges.)

Cut compressed air waste and you move both numbers at once: fewer kWh and a lower, flatter demand profile.

You can't fix what you can't see: sub-meter the compressor room

The reason compressed air waste persists is measurement. A factory-level TNB meter cannot tell you the compressors drew more last night than during production, but a sub-meter can, and that pattern is the fingerprint of a leak.

This is where CobiNeural earns its place. Sub-metering the compressor feed under Insights → Equipment gives you continuous kW and kWh at the source. Anomaly detection then does the watching for you: when night-time or weekend demand fails to fall while the plant is idle, that steady draw is the leak load, quantified in kW and ringgit rather than guessed at during an annual audit. The Max Demand KPI under Insights → Energy shows exactly how much the compressor room contributes to your billed peak, and Actions can sequence or stage compressors so you're not running a 75 kW unit to serve a 20 kW night load.

The point isn't a dashboard for its own sake. It's that leaks and pressure creep come back: a survey fixes them once, monitoring keeps them fixed.

A practical sequence that pays back fast

You don't need a capital project to start. Work in this order:

1. Baseline it. Sub-meter the compressor room so you have real kW/kWh and a clear night-vs-day load curve before touching anything.
2. Hunt the leaks. An ultrasonic survey plus tag-and-repair typically recovers a large share of that 20–30% within weeks, and it's mostly labour.
3. Right-size the pressure. Drop the setpoint to the lowest the process tolerates. Each 2 psi is about 1% off the bill, every hour, forever.
4. Kill artificial demand. Replace open-blow with regulated nozzles, close abandoned drops, fix the obvious misuse.
5. Sequence multiple compressors so part-load is met efficiently instead of a big machine idling.
6. Verify the saving. Use the same sub-meter to confirm the new load curve. Proper measurement and verification turns "we fixed some leaks" into a number you can put in an EECA report or an ISO 50001 review, and defend.

Manufacturers we work with, from textiles to precision production, find the compressor room is where the fastest energy ROI hides. See how that plays out in real plants on our case studies.

The takeaway

Compressed air energy efficiency isn't won at the compressor. It's won in the 20–30% you're currently leaking, the pressure you don't need, and the base load that never sleeps, all of which stay invisible until you meter them. Fix the leaks once, then let continuous monitoring make sure they don't quietly come back.

If you want to see what your compressor room is actually costing you, book a demo and we'll walk through it on your own data.