Extrusion Energy Consumption: Where the Kilowatt Hours Actually Go

Where the energy actually goes

The chart on the left is a measured breakdown for a 90 mm single screw line running a polyolefin profile. The main drive takes 52 percent of the electricity, the barrel heaters 26 percent, cooling pumps and fans 9 percent, the haul-off and winder 8 percent, and vacuum and compressed air 5 percent.

The drive dominates because it does two jobs at once. It conveys the material and it generates most of the heat that melts it, through viscous shear in the melt film. On a well designed screw, 70 to 80 percent of the energy needed to melt the polymer comes from the drive rather than from the barrel heaters, which is why the heaters are only a quarter of the total.

That has a direct consequence for efficiency work. Anything that improves the drive efficiency, such as a variable speed drive or a correctly sized motor, affects half the bill. Anything that improves the heaters affects a quarter, and some of that saving is offset by the drive having to work harder.

The breakdown also shows where the small numbers hide. Cooling pumps and fans at 9 percent are often left running at full speed when the line is idle, and compressed air at 5 percent is usually the most expensive energy on site per kilowatt hour because of the losses in generation.

Specific energy and the shoulder of the curve

Specific energy is the electricity consumed per kilogram of product, and it is the number that matters for both cost and carbon. On the chart on the right it falls steeply from 0.420 kWh/kg at 40 percent of rated output to 0.248 at 80 percent, then flattens and begins to rise again, reaching 0.258 at 100 percent.

The shape has two causes. At low output the machine is consuming a large fixed load, the heaters and the drive idling, spread over a small mass. As output rises the fixed load is diluted, so the specific energy falls. At high output the melting limit takes over: the screw does more work per kilogram, the melt temperature rises, and the specific energy turns back up.

The minimum is around 80 to 90 percent of rated output, which is where the line should be run for energy reasons. It is also close to the knee of the output curve, which is where it should be run for wear and melt quality reasons. The two optima coincide, which is convenient.

Running at 40 percent of capacity doubles the specific energy. That is the single largest energy waste in most extrusion plants, and it is caused by scheduling rather than by equipment. A line that is run at part load for a full shift uses more energy per kilogram than the same line run flat out for half a shift.

The drive: the largest and least understood load

The drive is an induction motor through a gearbox, and its efficiency at full load is typically 92 to 95 percent. At half load it falls to 88 to 91 percent, and at a quarter load to 80 percent or less. A line that runs at part load therefore pays twice: the fixed losses are spread over less product, and the motor itself is less efficient.

A variable speed drive does not improve the motor efficiency much, but it removes the gearbox losses at low speed and it allows the screw speed to be set to the actual requirement rather than to a fixed ratio. On a line with a wide product range, that is worth 5 to 10 percent of the drive energy.

The gearbox is the part that gets ignored. A worm gearbox is 60 to 80 percent efficient and a helical gearbox is 95 to 97 percent. A line with a worm gearbox is losing a fifth of its drive energy as heat in the oil, and the oil temperature is the evidence. Replacing the gearbox is rarely justified on energy alone, but it is worth knowing when the numbers do not add up.

Screw design affects the drive load directly. A barrier screw with a long melting section requires more torque per kilogram than a simple screw, because it is doing more of the melting work mechanically. That is a good trade, because it removes the need for heater energy, but it shows up as a higher drive load and it should be counted in the comparison.

Barrel heaters: set them for the melt, not the gauge

Barrel heaters are the second largest load at 26 percent, and they are the one that is most often set by habit. A profile that was established for one material is left in place for another, and the heaters then work against the screw, adding heat that the shear is already providing.

The energy penalty is not just the heater power. Every degree of excess barrel temperature has to be removed downstream by the cooling system, so the same kilowatt hour is paid for twice, once to add the heat and once to take it out. A profile that is 20 C too hot can add 8 percent to the total line energy.

The correct setting is the one that produces the right melt temperature at the die, not the one that matches a data sheet. A flat profile, with all zones at similar temperatures, is often more efficient than a rising profile, because it relies on shear rather than on conduction and leaves the cooling system less work.

Insulating the barrel is cheap and effective. A bare barrel radiates 100 to 200 W per metre of length, which on a 25 D barrel is 2.5 to 5 kW of continuous loss. A removable insulation blanket costs a few hundred euro and pays back in months, and it also stabilises the zone temperatures.

Downstream loads that are easy to forget

Cooling pumps and fans are 9 percent of the total, and they are almost always run at fixed speed. A cooling circuit that is sized for the hottest day runs at full flow on a mild one. A variable speed pump on the cooling circuit, controlled by the return water temperature, typically saves half of that 9 percent.

Vacuum and compressed air are 5 percent of the line total but a much larger share of plant energy, because compressed air is an inefficient medium. A vacuum pump running continuously for a profile that needs vacuum only at the calibrator is a common waste; a solenoid valve on the vacuum line costs almost nothing.

The haul-off and winder are 8 percent and are usually driven by servos that are already efficient. The waste there is in oversizing: a haul-off rated for 40 m/min running at 6 m/min has poor efficiency at that point, and a smaller unit would use less.

Lighting and space heating are outside the line but often attributed to it. A plant that measures energy per line should separate the two, because otherwise a change in the weather looks like a change in process efficiency.

Measuring what you actually consume

The measurement has to be at the line, not at the building meter. A three phase energy meter on the line supply, logging at one minute intervals, gives the profile of a shift and shows the idle consumption, the start-up peak and the steady state. That data is the basis for every decision that follows.

The second measurement is the mass. A line that does not weigh its output cannot calculate specific energy, and most lines do not weigh output continuously. A simple check, weighing the output over ten minutes at three speeds, gives enough to plot the curve in the chart and to find where the line actually runs.

The third is the melt temperature at the die, which links the energy to the quality. A line that is running at a high specific energy and a high melt temperature is wasting energy and risking degradation at the same time. A line at a low specific energy and a correct melt temperature is at its optimum.

Table 1 — Five measurements that cost almost nothing

What to change first

The first measure is scheduling, and it costs nothing. Running lines at 80 percent of capacity instead of 40 halves the specific energy, and it requires no capital at all. It is the single largest available saving in most plants, and it is the one that is most often resisted because it constrains the production plan.

The second is barrel insulation and heater zone control, which together are worth 5 to 10 percent of the total and pay back in under a year. Both are retrofits that can be done on a shutdown weekend.

The third is variable speed on the cooling circuit and on the vacuum, worth another 4 to 6 percent, and the fourth is a variable speed drive on the main motor if the line does not already have one. Those two are capital projects with paybacks of two to four years at current energy prices.

What is not on the list is a new machine. A modern line is more efficient than a 1990s one, but the improvement is 10 to 20 percent, and it comes with a capital cost that no energy saving can justify on its own. Energy is a reason to run the existing line properly, not a reason to replace it.

Table 2 — Measures in order of payback, not of saving

The order in the table is the order of payback, and it is deliberately not the order of saving. The largest saving is free and the smallest saving is expensive, which is usually the case and is usually ignored because the free measure requires a change in how the plant is scheduled rather than a purchase.