Short answer: specific energy is motor power divided by throughput, in kWh per kg. On a compounding line it typically sits between 0.15 and 0.35 kWh/kg. Most plants can take 8–15% out of that number without buying anything, by fixing tip clearance, screw configuration and fill level — in that order.
Specific energy is the energy you put into the polymer per kilogram of product. It is written kWh/kg, or more often kWh/t on a plant report. Two versions matter, and confusing them is the most common mistake in energy discussions.
Gross specific energy is what the electricity meter sees: total motor input power divided by throughput. It includes the energy needed just to spin the shafts, which produces no melting at all.
Net specific energy, also called specific mechanical energy, subtracts the no-load power at the same screw speed. It is the energy that actually went into the material. Net is the number that tells you about dispersion and melt quality; gross is the number that tells you about your electricity bill.
For an energy project you need both, because they point at different fixes. A high gross with a normal net means you are paying to spin empty screws — a throughput or fill-level problem. A high net at normal gross means the screws are doing too much work per kilogram — a configuration or clearance problem.
The HMI load percentage is not a measurement. It is a derived display, usually scaled to a nominal current, and it will happily disagree with the meter by 10%. Put a power analyser on the VFD input and log three things at steady state:
Then run the machine at production screw speed with the feed stopped and the die open, and record the no-load power. That single reading is worth more than a week of HMI screenshots, because it separates the fixed cost from the variable one.
| Reading | Value | What it means |
|---|---|---|
| Drive input power at production | 240 kW | What you pay for |
| Throughput | 800 kg/h | Weighed, not setpoint |
| Gross specific energy | 0.300 kWh/kg | 240 / 800 |
| No-load power at 600 rpm | 60 kW | Empty screws, die open |
| Net specific energy | 0.225 kWh/kg | (240 − 60) / 800 |
| No-load share of the bill | 25% | 60 / 240 |
A quarter of that machine's energy is going into bearings, seals and gearbox losses. That is not unusual, and it is the reason throughput levers usually beat hardware levers.
The order below is deliberate. The first three cost nothing but a shutdown and a decision; the last two cost money. Work down the list, not up it.
| # | Lever | Realistic saving | Cost |
|---|---|---|---|
| 1 | Tip clearance and element wear | 3–8% | Shutdown + parts |
| 2 | Fill level and throughput | 3–10% | None |
| 3 | Screw configuration | 2–6% | Shutdown + parts |
| 4 | Barrel temperature profile | 2–5% | None |
| 5 | Die, screen pack and downstream restriction | 1–4% | Low |
| 6 | Element surface finish | 1–3% | Parts |
| 7 | Drive, belts and VFD tuning | 1–3% | Low |
This is the single biggest mechanical contributor and the one most often ignored, because a worn screw still runs and still makes product. As the flight tip wears, the gap between tip and barrel grows. Material that should be dragged forward slips backwards over the tip instead, so the same throughput needs more screw speed, and more screw speed means more of the energy goes into shear heating you then have to remove with barrel cooling.
The effect compounds: worn elements raise specific energy, the higher energy raises melt temperature, the higher melt temperature needs more cooling water, and the cooling water is another electricity bill nobody attributes to the screws. See how to measure screw element wear and clearance for the measurement itself. A clearance check costs an afternoon; a set of elements costs money but pays back in months on a machine running three shifts.
Specific energy falls as the screw fills, because the no-load power is spread over more kilograms. Running a machine at 60% of its capability is the most expensive habit in compounding, and it is extremely common: lines are set up for the hardest product they will ever run, then left there for everything else.
Push throughput up in steps and watch melt temperature, die pressure and dispersion together. Most lines have 10–20% headroom on easy products that nobody has claimed, because the recipe was fixed years ago at a cautious number.
Kneading blocks are where the energy goes. A 90° stagger disc set is a high-shear, high-energy element; a 45° set does similar distributive work for less energy; a 30° set less again. If a block was specified to solve a dispersion problem that no longer exists — a different filler, a finer grade, a changed carrier — it is now simply a heater.
Count the blocks and ask what each one is for. A typical over-configured screw has one kneading block too many and one too many 90° discs in the block that remains. Removing energy here also lowers melt temperature, which usually improves product quality rather than hurting it.
Heating the feed end reduces the mechanical work needed later, because the polymer is softer when it reaches the kneading zone. A flatter, warmer feed profile often cuts net specific energy noticeably. The limit is feed-zone bridging: too much heat at the throat and the granules stick and the feed screws starve. Move the profile in 5 °C steps and watch the feed current, not just the melt.
Preheating the feed itself, if you have waste heat available from the die or from a dryer, is free energy in the most literal sense.
Everything downstream of the screw tip is a pressure load on the screw. A dirty screen pack, an undersized die, a long transfer pipe or a partially closed valve all raise head pressure, and the screws pay for it. Log the pressure differential across the screen changer and change screens on pressure, not on the calendar.
Ground and polished flanks let material slide rather than drag. The effect is smaller than clearance but it is permanent, and it also reduces wear on the elements themselves. Article 13 covers the mechanism and the roughness figures: screw element surface finish and why it matters.
The least glamorous item, and occasionally the one that finds a real fault. Check belt tension or coupling alignment, confirm the VFD is not running in a fixed-torque mode that wastes energy at part load, and compare the motor's no-load current with its nameplate. A misaligned coupling can quietly cost 2% forever.
Take the 800 kg/h line above, running 6,000 hours a year. That is 4,800 tonnes a year.
| Scenario | Gross kWh/kg | Annual kWh | Annual cost at 0.65/kWh |
|---|---|---|---|
| Before | 0.300 | 1,440,000 | 936,000 |
| After a 10% improvement | 0.270 | 1,296,000 | 842,400 |
| Saving | 0.030 | 144,000 | 93,600 |
Put the numbers in your own currency and the argument makes itself. A set of screw elements is usually recovered inside a year on a line of this size, and the melt-temperature reduction is a bonus that shows up in product quality rather than on the bill.
Specific energy is a cost metric, not a quality metric. Three cases where driving it down will hurt you:
The honest target is not the lowest number. It is the lowest number at which every quality test still passes with margin, and that number is specific to your product, not to a benchmark table.
None of that requires a capital request, and together it is usually worth more than the hardware options on the list. If you would rather start from the parts side, our screw elements and kneading blocks page lists the geometries and steel grades, and we will match a configuration to your current process rather than to a catalogue.
For compounding, gross specific energy usually falls between 0.15 and 0.35 kWh per kg, with most glass-filled or heavily filled products at the upper end and unfilled or easy-flowing products at the lower end. Net specific energy is typically 70 to 80 per cent of gross. A number outside that band is not automatically wrong, but it is worth explaining.
Both, for different reasons. Gross is what you pay for and is the number to use for an energy-saving project. Net, which subtracts no-load power at the same screw speed, tells you how much work actually went into the polymer and is the better indicator of dispersion and melt quality. If gross is high but net is normal, the problem is fill level, not screw geometry.
Almost always, up to the point where the screws are full and the die pressure starts to surge. The no-load power is fixed, so spreading it over more kilograms lowers the number. The practical ceiling is set by melt temperature, die pressure stability and the quality tests, not by the motor.
A set of elements at the wear limit typically adds 3 to 8 per cent to gross specific energy, and the effect is worse than the number suggests because the extra energy arrives as shear heat that you then pay to remove with barrel cooling. On a three-shift line the electricity saving alone often covers a replacement set within a year.
Partly. Lower screw speed reduces total power but also reduces throughput, so specific energy per kilogram often stays roughly flat while melt temperature falls. The real gain from lower speed is quality, not cost. The gains that reduce cost come from fill level, clearance and configuration, which raise kilograms per kWh rather than lowering kW.