Oil-containing metal shavings oil removal and briquetting line: Transforming "Waste disposal fee" into "Second output value"

In most CNC shops, the fate of the chips is always the same: swept into a bin, piled up in a corner, and when the bin is full, hauled away by the scrap dealer and priced as "light material" or "wet material" — at whatever the day's rate happens to be.

Very few people treat this stream as a production line. But once you open the books, you find the chips are leaking money in three places at once — and a de-oiling and briquetting line recovers two and a half of them in one go.

Start with the physical situation of the chip itself. A 10 kg aluminium ingot has a surface area of a few tenths of a square metre. The same 10 kg as chips 0.1 mm thick reaches several hundred square metres. Three orders of magnitude of difference.

The chemistry has not changed. The shape has multiplied its weaknesses a thousandfold.

Bulk density: 0.10–0.35 t/m³ for aluminium, 0.5–1.0 t/m³ for steel. A 10-tonne truck loaded with loose aluminium swarf actually carries only 1–2 tonnes.

Specific surface area: several hundred to several thousand times that of solid metal. Extremely fast oxidation, heavy dross formation once it hits the furnace.

Entrained liquid: commonly 8–12%. The cutting fluid is paid for twice and thrown away once, plus a disposal fee on top.

Chip shape: long chips tangle together. They bridge, they jam the feed, and they spring back after compression.

The three losses below come directly from these four weaknesses.

Loss #1: The layer floating on top of the melt is burning away for nothing

This is the first reason a briquetting press exists, and the most repeated — and most often mis-stated — claim in the industry.

"Briquetting reduces melting loss" — that statement is only half true.

Loose chips are far less dense than the molten metal, so once charged they float on the surface of the bath. They meet air and flame first, not liquid metal. Oxidation and then melting happen at the surface, and a large share of the metal turns into dross that has to be skimmed off. Industrial measurements are unambiguous: 20–30% melting loss for loose aluminium chips, versus 2–8% after briquetting. In published remelting trials of cold-pressed aluminium briquettes, dross fraction was 2.25%, relative melting loss 3.55%, and metallurgical yield 94.2% — at the top of the 90–96% range reported in the literature.

But the decisive point lies elsewhere: briquetting itself is not a sufficient condition. Crossing the density threshold of the melt is.

Liquid aluminium has a density of about 2.3 g/cm³. Put another way, a briquette pressed to only 2.0 t/m³ still floats — and a briquette that floats oxidises, pressed or not.

This is precisely why the scientific literature keeps producing contradictory results:

One study reports that within the 30–50 MPa compaction range, raising the compaction pressure actually lowers metal recovery.

Another, covering 0–1200 MPa, found that at 170–210 MPa — the typical range of commercially advertised presses — aluminium briquettes reach less than 60% of the density of bulk aluminium, with remelting recovery of around 50%. Even at 1200 MPa, density stayed below 94% and recovery below 72%.

Some authors even argue that briquetting is detrimental, for exactly this reason: the briquette is still lighter than the melt, so it floats and oxidises faster.

Put these conflicting findings side by side and the conclusion is sharp: the argument is not "should you briquet," it is "how dense." The published finding that briquettes above roughly 2.2 g/cm³ show markedly lower oxidation losses than loose chips matches exactly the industrial practice of specifying aluminium briquettes at 2.0–2.4 t/m³, with good equipment reaching 2.3–2.5 t/m³. Both are aiming at the same thing — making the briquette sink.

Worth noting too: Chinese industry standards quantify the requirement. Under JB/T 12778-2021 and GB/T 4223-2017, a briquette only counts as compliant furnace charge at 5.5 t/m³ or above for cast iron chips, 6.0 t/m³ for steel chips, and 2.0 t/m³ for aluminium chips, with a drop-disintegration rate of ≤10%. Note that for aluminium, 2.0 t/m³ is a compliance threshold, not the sinking threshold.

So when you select equipment, do not ask "how many tonnes?" Ask: for my alloy and my furnace temperature, will you put briquette density in writing in the contract? If the answer is only "around 2.0," you are buying a machine that makes floating pucks.

Loss #2: You paid for the cutting fluid twice, and you are paying again to dispose of it

This loss is often larger than the first, and it is the one most often filed away as a "side benefit."

Cutting fluid leaves the shop in two ways: as atomised mist that evaporates, and clinging to every chip surface and every capillary gap. At 8–12% oil content, a mid-size shop generating 3,000 tonnes of chips a year carries away roughly 240–360 tonnes of cutting fluid — already paid for.

High-pressure compaction wrings the chips out like a sponge. The expelled fluid runs through the die drainage channels into the collection tank, is filtered, and goes back to the machine tool sumps. Recovery rates reported across the industry run 85–98%, and a fully configured line brings residual oil down from 8–12% to below 3%.

Why 3%? Because that number governs environmental compliance and selling price at the same time:

The exemption list in the 2025 edition of the National Catalogue of Hazardous Wastes (China) states that oily metal swarf de-oiled to the point of no dripping at rest, and briquetted for metal smelting, need not be managed as hazardous waste during the utilisation stage. The cost gap between hazardous waste and ordinary solid waste can approach RMB 10,000 per tonne.

A second reference point is Annex B of GB 5085.6-2007: petroleum solvent content at or above 3% constitutes identification as hazardous waste. Some regions already accept de-oiled ferrous chips below 3% petroleum hydrocarbons as ordinary solid waste.

And on the selling side: above 5% oil content, most scrap buyers refuse the lot or discount it heavily.

The applicable criteria must be confirmed with your local environmental authority. Test methods and thresholds vary by market.

It also has to be said plainly: a briquetting press is not a dedicated de-oiling machine. The judgement used across the industry is:

Liquid content up to 5%: direct briquetting is usually fine, with part of the fluid recovered during compression.

Liquid content 5–10%: assess case by case, by chip form, material and fluid type.

Liquid content above 10%: centrifuge de-oiling or dewatering first, then briquet.

Expecting a briquetting press to handle heavily saturated material is forcing it to do a job that is not its own. What comes out is discounted "wet material."

Loss #3: Volume, freight, and the most valuable one — priceability

The third loss is the most visible and the easiest to calculate.

Volume: cast iron chips reach 5.2–5.5 t/m³, steel turnings 3.5–4.5 t/m³ or higher, aluminium 1.5–2.5 t/m³. Overall volume reduction of 80–98%, with a compression ratio typically around 10:1.

Freight: with the same truck, payload goes from 1–2 tonnes of loose aluminium swarf to 8–9 tonnes of briquettes — four to five times more per trip, with unit transport cost down 70–80%.

Priceability — the most underrated one. The buyer's reasons for discounting loose material are always the same: "mixed material," "awkward to load," "oil and moisture unknown." Once briquetted into a standard format, density, weight and composition all become verifiable, and the purchase price typically runs 10–30% above loose material.

What a briquetting line actually delivers is not compression. It is standardisation — turning a pile of waste that nobody wants to price carefully back into a raw material that trades by the tonne.

Why it has to be ONE line: crushing → de-oiling → briquetting

These three stages are not three machines you buy separately and bolt together later. They are three answers to one physical problem.

1. Crushing (pre-treatment)

Long, curled ribbon chips tangle into balls: physically, a heap of springs compressed together. The moment pressure is released, the stored elastic energy discharges and the briquette swells, cracks, or falls apart on the spot. At the same time, liquid trapped inside the tangled structure is very hard to expel.

So "crush first, briquet second" is not an option — it is a precondition. Shearing chips with too large a length-to-diameter ratio into short, uniform pieces (3 cm is a common reference value, but the real criterion is whether the material feeds stably and continuously) solves three problems at once: no bridging in the feed, higher briquette density, lower residual oil. Field comparisons show that after crushing, briquette density reaches 1.7–2.2 t/m³, with residual oil far below what you get pressing long curly chips directly.

2. De-oiling

Centrifugal separation brings heavily oiled chips back into a pressable range and noticeably improves final briquette quality.

3. Briquetting

The full cycle has five stages:

1. Feed — chips enter the die; on horizontal machines a pre-compression device may first push fluffy or tangled material into the main chamber.

2. Main compression — the main cylinder applies sustained pressure; voids close, particles deform plastically, surface oxide films rupture under the pressure, and fresh metal surfaces come into contact, producing cold welding and mechanical interlocking.

3. Solid-liquid separation and dwell — separated fluid is routed to the collection system; pressure is held so internal stresses relax fully.

4. Controlled decompression — pressure is released progressively. Sudden release causes hydraulic shock, vibration and noise and, more seriously, cracks the briquette. This is a function many low-cost machines leave out.

5. Ejection and reset.

Two parameters decide the outcome:

Dwell time. Its physical meaning is not "press harder" — it is giving the delayed elastic effect time to release inside the die. Common reference values: 3–8 s for metal chips, 5–12 s for light, fluffy material. Insufficient dwell produces the classic symptom: a briquette that looks perfect on ejection and falls apart when it hits the floor.

Specific pressure matched to the material. Around 150–200 MPa for aluminium alloys, above 300 MPa for steel and stainless. Pressure is not "the more the better": plot density against pressure and you see a clear inflection point. Beyond it, extra pressure buys almost no density, while energy consumption, die wear and radial stress cracking in the briquette all rise together. The correct way to tune is to sample continuously and read the cross-section — porous section, raise pressure slightly; radial cracks at the edge, lower peak pressure.

Briquetting needs no binder, and you should not accept a proposal that includes one. Sustained hydraulic pressure cold-welds metal chips on its own. A binder adds consumable cost and contaminates the melt.

Hydraulic vs mechanical: a hydraulic press applies slow, sustained pressure across the full stroke and produces dense, dimensionally consistent briquettes without binder — the right choice for metal chips. A mechanical press uses a flywheel and crank for fast repeated impacts: higher stroke rate, but cohesion generally requires a binder, which is why it belongs with coal fines and biomass rather than your swarf.

What a saleable briquette looks like, by material

Cast iron chips: bulk density 0.2–0.5 t/m³, target ≥5.2–5.5 t/m³ — chargeable directly to the furnace.

Steel turnings and borings: 0.5–1.0 t/m³, target 3.5–4.5 t/m³, premium furnace grade ≥6.0 t/m³.

Aluminium chips: 0.10–0.35 t/m³, target 2.3–2.5 t/m³ — must exceed about 2.3 to sink in liquid aluminium.

Copper and brass chips: target around 75% of solid metal density. Copper in an aluminium melt downgrades the whole heat.

Stainless steel, titanium: higher specific pressure required (>300 MPa); confirm the pressure class before ordering.

Never mix different materials — this is not fastidiousness, it is margin

Mixing triggers three losses at once:

1. Galvanic corrosion. Aluminium and iron chips in prolonged contact with residual electrolyte oxidise faster; briquettes can powder within three months of storage.

2. Density mismatch creates a "false compact." Cast iron compacts to around 6 t/m³, aluminium to 2.0–2.4 t/m³. Pressed in the same cavity, the hard material bridges over the soft: you get a hollow briquette that the buyer discounts as mixed scrap, by RMB 50–120 per tonne.

3. Melt contamination. Copper in an aluminium ingot and the entire heat is downgraded.

Recommended practice: four separate holding hoppers in the shop — steel, cast iron, aluminium, copper — each with its own briquetting line. If floor space is tight, apply at minimum the rule "same shift, same material," and never allow overnight mixed storage. Today's PLCs store more than 20 pressure curves: switching material is a program change, not a machine change.

Payback: a mid-size aluminium shop

Example figures — replace them with your own output, alloy, cutting fluid price and local aluminium price.

Basis: 60 tonnes of oily aluminium chips per week × 50 weeks = 3,000 t/year

Aluminium selling premium: +20% × 3,000 t × RMB 15,000/t = RMB 9,000,000

Cutting fluid recovered: 3,000 t × (8% − 3%) ≈ 150 t × RMB 8,000/t = RMB 1,200,000

Hazardous waste reclassified as ordinary solid waste: 3,000 t × local differential — varies widely by region, calculate case by case

Freight reduction: payload up four to five times, depending on distance and frequency

Labour saving: annual gap between a semi-automatic and a fully automatic line of roughly RMB 90,000–100,000

Payback periods reported across the industry for industrial briquetting presses run 6–18 months; high-volume aluminium die casting and CNC shops sit at the short end.

Semi-automatic or fully automatic? Do not take the sales pitch of "go straight to the top model." Multiply daily output × shifts × your local hourly labour cost and the answer comes out by itself: above 10 t/day, on two shifts, or in regions where labour is scarce, the labour saving usually covers the depreciation. Below 12 months payback, the automatic line is justified.

Five questions to ask the supplier before you sign

1. Can you test my material and send the briquettes back to me? Tonnage in a brochure means nothing. A serious manufacturer will ask you to ship 1–2 kg of sample and will answer with measured density and measured residual oil. Insist on this step.

2. Will you put density in the contract? Not "up to" — a specific number, a test method, and an acceptance clause.

3. Is the de-oiling stage sized for my oil content, or is it a token drain channel? Above 10% oil, a press with a collection tray is not a de-oiling line.

4. How is decompression handled? Progressive release protects both the briquette and the hydraulic circuit. Instantaneous dump means a cracked briquette.

5. What happens when I change material? Stored pressure curves, adjustable dwell time, quick-change dies: switching alloy should be a matter of minutes, not the purchase of another machine.

FAQ

Can a briquetting press handle chips wet with cutting fluid?

Yes, but with limits. Up to 5% liquid, direct briquetting is usually fine, with part of the fluid recovered during compression. Between 5% and 10%, judge by chip form and fluid type. Above 10%, centrifuge de-oil or dewater first. Material with more than about 10% water must be dewatered before pressing.

Is a binder required?

No. A hydraulic press cold-welds steel, aluminium, copper and cast iron chips under sustained pressure, with no additive at all. A binder costs money and contaminates the melt.

How much cutting fluid can be recovered?

Reported rates run 85–98%, depending on chip form, fluid type and configuration. The recovered fluid is filtered and returned to the machine tools.

What residual oil level counts as compliant?

Take 3% as the key reference: it governs the "no dripping at rest" criterion and matches the 3% petroleum solvent threshold in Annex B of GB 5085.6-2007. Above 5% oil, most buyers refuse the lot. Confirm the applicable criteria with your environmental authority.

What density do aluminium briquettes need?

Do not stop at the 2.0 t/m³ compliance threshold. What determines melting yield is exceeding the density of liquid aluminium (about 2.3 g/cm³), so the briquette sinks into the melt instead of oxidising on the surface. Recommended target: 2.3–2.5 t/m³.

The briquette falls apart at the slightest touch — what should I do?

Check three things: dwell time too short (the elastic effect releases after ejection — extend to 3–8 s, or 5–12 s for fluffy material); long curly chips not crushed (spring effect — crush first); pressure curve wrong (sample, read the cross-section, adjust in steps).

How long is one cycle? How do I calculate throughput?

A cycle typically runs 30–180 seconds depending on model and material. Daily capacity ranges from 3–5 tonnes to well over 100 tonnes. Theoretical throughput = weight per briquette × cycles per hour.

I generate 2–5 tonnes of steel turnings a day — what tonnage do I need?

Common ranges: 100–125 t gives 0.6–1.5 t/h; 160–200 t gives 2.2–5 t/h and is the standard in machining; 250–315 t and above serves centralised processing centres. Up to 5 t/day, a single-cylinder 125–160 t model is enough. For 5–15 t/day, go for a 200 t twin-cylinder model with pre-compression and drainage.

What are your chips worth? Send us a sample

Send 1–2 kg of chips straight off your machines, oil and all, stating the alloy grade, monthly output and current oil content (or water content). We will run them on our line and send back:

1. Measured briquette density — and whether it clears the sinking threshold for your alloy.

2. Measured residual oil — against the 3% line.

3. Configuration recommendation — of crushing, de-oiling and briquetting, what you must have and what you can skip.

4. Throughput and payback calculation — using your real numbers, not industry averages.

Request a sample test →

Data in this article is drawn from industry publications and published literature and does not constitute a performance commitment for any specific model. Actual configuration and benefit are determined by the test report on your own material and by a site survey.