A plant manager I work with in Selangor opened his utility bill one Monday and found that 41% of his monthly cost had landed in a six-hour window every afternoon. His duct shop ran one shift, 8 to 5, which is precisely that window. Nothing was wrong with his machines. His schedule was just pointed at the most expensive electrons the utility sells. Over the following months we rebuilt his production calendar around the tariff, moved the heavy cutting into the night block, and cut the electricity line on his P&L by roughly a fifth without touching a machine. Time-of-use tariffs are spreading fast — most of the Gulf, Southeast Asia, and increasingly Latin America — and a duct shop is unusually well suited to exploit them because our work is batch-shaped. Here is how to work out whether it pays for yours.
You cannot schedule around a tariff you have not measured. Start with the nameplate, then correct it with a clamp meter, because nameplates lie by omission — they state rated draw, not duty.
On a typical line like our auto duct line, the decoiler drive pulls 3 to 4 kW but runs intermittently. The leveler adds 2 to 3 kW when it is working. The shear motor is 5.5 kW for a two-second stroke every so often. The plasma unit is the heavyweight: 15 to 25 kW while arc-on, but arc-on time on duct blanks is maybe 30% of cutting-station cycle. The flange former is 4 kW continuous while forming, the beading station 3 kW, the folding station hydraulic pack 7.5 to 11 kW in short bursts. Installed total looks like 45 to 60 kW; measured average draw across a productive hour is usually 18 to 25 kW. That distinction is the whole game — the line does not burn its nameplate, it burns its duty cycle, and duty cycles can be moved.
Two different numbers appear on industrial tariffs and they punish different sins.
This is the per-unit price that varies by hour. In our Selangor example the off-peak block, 22:00 to 08:00, cost about a third of the mid-afternoon peak rate. A kWh is a kWh — a blank cut at 2 a.m. is identical to one cut at 2 p.m. — so every movable kWh shifts cleanly.
Many utilities bill the highest 15- or 30-minute average demand in the month, and some ratchet it against the previous eleven months. This one is sneakier. Two lines starting their plasma units simultaneously at 08:00 can set a demand peak that bills all month regardless of what happens afterwards. Staggering large motor starts across a 15-minute window, and keeping plasma arc-on out of the metered peak window, often saves more than the energy shift itself. One factory I know cut its demand charge 30% purely by interlocking the two plasma units so they could not arc at the same time. No new equipment, just a relay and ten lines of logic.
Not everything should move. The trick is sorting production into energy-heavy-but-pausable and labour-bound-now.
| Operation | Typical draw | Duty | Movable off-peak? |
|---|---|---|---|
| Coil loading and decoiler setup | 3–4 kW | Intermittent | Partly — prep at shift end for the night run |
| Plasma cutting / blanking | 15–25 kW arc-on | Heavy bursts | Yes — the single biggest movable block |
| Flange forming and beading | 3–7 kW | Continuous while running | Yes, but labour-bound; often stays day shift |
| Folding station hydraulics | 7.5–11 kW | Bursts | Follows the forming schedule |
| Welding (spiral / reinforcement) | 8–12 kW | Batches | Yes — batch it into the night block |
| Compressed air | 11–22 kW | Background leak floor | Pressure-band it; fix leaks first, always |
On most shops the plasma work plus welding is 60 to 70% of the shiftable load. Forming and folding follow the people, and that is fine — the objective is not zero day-shift draw, it is moving the big arc-heavy blocks.
Take a shop running 22 productive days, averaging 20 kW on one shift. Roughly 55% of that energy is plasma and welding, so about 11 kW average is shiftable. Moving it into a block priced at a 60% discount saves 11 kW × 8 h × 22 days × 0.6 × tariff. At 0.15 currency units per peak kWh, that is about 174 units a month. Modest but real, and it scales with a second shift.
Now add demand. If the afternoon plasma arc-on coincides with the folding station's hydraulic start, a 40 kW spike for 15 minutes can set the monthly demand figure. Shaving that peak saves the demand charge on every following day of the month. In tariff structures with ratchets, one bad Monday in March bills through next February. Measure your own demand history from the utility portal before you assume anything — most utilities expose 15-minute interval data for free, and it tells you exactly when your money is leaving.
Night shifts for cutting-only blocks work better than full night shifts for most duct shops, and here is why. Cutting and welding are machine-paced; forming and folding are operator-paced and benefit most from experienced hands. A lean night crew of two running plasma blanks and spiral welding batches needs less supervision than a full forming line and pays for the tariff delta on its own. But: QC does not automatically transfer. If your only inspector is a day person, the night blocks need either a written first-off procedure with photos, or the first hour of day shift reserved for verifying the night's output before it progresses to forming. Skipping this is how you win on electricity and lose on rework. The math on rework is brutal — a mis-cut blank that survives to assembly costs more than the electricity saving on ten good ones.
Two practical notes from shops that run this way. First, plasma consumable life is unaffected by running at night, but operators left alone at 3 a.m. will stretch electrode change intervals — build the check into the night log. Second, keep the CNC plasma table dust collector filter checks on a fixed weekly slot regardless of shift pattern; night crews deprioritise what nobody schedules.
Blank staging closes the loop between the shifts. The night crew cuts and stacks blanks against the morning's forming schedule, and the day crew arrives to a full rack rather than a queue at the plasma station. That handoff — cutting finished off-peak, forming fed continuously on-peak — is the shape most successful conversions settle into. It also cleans up the material flow: one operator I respect calls it "cutting tonight what we fold tomorrow", and his shop's shearing station idle time dropped visibly once blanks stopped chasing the forming line.
Three cases where I tell people to skip the exercise. If your tariff's peak-to-off-peak ratio is under about 1.5:1, the saving rarely covers the supervision cost of a night block. If your bottleneck is labour rather than machine capacity — common in markets where skilled fitters are scarce — then idle forming stations at night generate no offsetting benefit and night premiums can eat the delta. And if the plant is on a legacy flat tariff, obviously there is nothing to arbitrage yet, though it is worth asking the utility what a time-of-use meter would change; several utilities in the region now offer it below a surprisingly small connected load.
The sequence that worked in Selangor: pull 15-minute interval data from the utility for a month and find your actual peak window; clamp-meter the line for one representative day and split the load into shiftable and fixed; move plasma and welding batches into the off-peak block with a two-person crew; interlock the big motor starts; review the next full bill against the same production volume. Nothing in that list needs capital. It needs a printout, a clamp meter, and someone willing to move the cutting to when the electricity is cheap — which, given how batch-shaped duct production is, is easier than almost any other energy project you will be offered this year.
How much of a duct line's load can realistically move off-peak?
On a typical straight line, plasma cutting plus welding is 60 to 70% of the shiftable energy — roughly half the line's average draw. Forming and folding usually stay with the day shift because they are operator-paced. Plan the project around moving the arc-heavy blocks, not the whole line.
Are demand charges or energy rates the bigger saving?
Check your bill. Where the utility bills a monthly peak demand with a ratchet, staggering motor starts and keeping plasma arc-on out of the peak window often saves more than the kWh shift itself. Interval data from the utility portal tells you within a month which lever is bigger for your site.
Does running plasma at night affect cut quality or consumables?
Not inherently — arc voltage stability depends on the machine and gas supply, not the hour. The real risk is thinner supervision at night: electrode change intervals get stretched and first-off checks get skipped. Put both into a written night log and quality holds.
What is the first step if I want to try this?
Pull a month of 15-minute interval data from your utility and clamp-meter the line for one representative day. Between those two documents you will know your peak window, your shiftable load and your tariff ratio — enough to do the payback math on a single sheet of paper before committing anyone to a night shift.