Air Compressors for PET Bottle Blow Molding at Up to 40 Bar

Why 40 Bar

PET preforms have an axial stretch ratio of 2.5 to 3.5 times and a hoop stretch ratio of 3 to 5 times. Completing this deformation within under a second at 95 to 115°C, pressing the material flush against the mold wall, takes 25 to 40 bar of air pressure. Simple thin-wall bottles under 500 ml blow well at around 28 bar. Bottles above 1.5 L, irregular shapes, and bottles with fine surface textures tend to need 35 bar or more.

Process

PET Blow Molding Air Supply

The preform IV value (intrinsic viscosity) affects compressor sizing and gets overlooked constantly. Between 0.80 and 0.86 dl/g, the preform has good strain hardening, the wall self-equalizes in thickness during stretching, and the pressure window is forgiving. Below 0.76 dl/g, self-equalizing drops off and the same bottle shape might require another 3 to 5 bar. Batch-to-batch IV fluctuation in raw PET can hit ±0.03 dl/g, enough to push a machine sized for 30 bar right to its limit.

Air Volume Matching

Rotary blow molding machines consume air in pulses. Each station blows for 0.5 to 1.2 seconds, then exhaust and mold clamping. The compressor sees spikes, not steady draw. Sizing to the blow molder's nameplate average consumption produces whitened bases and thin shoulders when pressure sags during peaks.

Temperature, Cooling, and How Displacement Disappears

Theoretical adiabatic discharge temperature from atmosphere to 40 bar exceeds 400°C. Three-stage compression with intercooling gets final stage discharge to between 150 and 180°C.

Intercooler fouling bleeds displacement faster than most people think. Heat exchanger fins in environments with cotton fiber dust, PET flake dust, or general industrial particulate build up an insulating layer on the air side within weeks. How fast depends entirely on the specific plant. A blow molding operation co-located with a preform injection line running reground material fouls intercoolers at a completely different rate than a standalone blow molder in a clean food-grade hall. In a relatively dirty plant, second stage inlet temperature can drift from design 40°C to past 55°C in two to three months. That temperature rise costs around 5% volumetric efficiency, which is worse than what piston rings at end of life cost. But piston rings have OEM-published replacement intervals measured in hours, those intervals go into maintenance software, and work orders generate automatically. Intercooler fouling rate depends on the environment, so the interval in the maintenance manual says something like "inspect periodically" or "clean as needed". Plants that have added intercooler cleaning to a fixed bimonthly schedule can see the displacement recovery on their flow meters. Plants that leave it ad-hoc lose capacity gradually, and the production team compensates with longer shifts or overtime, usually without connecting the output shortfall to the intercoolers at all.

Intake air temperature is where compressor rooms in PET plants cause trouble. Manufacturers rate nameplate displacement at 20 or 25°C. Compressor rooms run 45°C in summer routinely, more in some facilities. A mezzanine compressor room above an injection molding hall, with the injection machines dumping heat upward and the roof absorbing solar load, can see 50°C without difficulty. Going from 25 to 45°C intake reduces displacement by roughly 6.5%. At 50°C the loss approaches 8%. A 100 Nm³/min rated machine might deliver 92 Nm³/min in July and 98 in January. If the line was sized for 96 Nm³/min demand, the summer shortfall shows up as inability to hold cycle time on the fastest blow molder stations, which the production team usually blames on the blow molder.

Oil-Free and Oil-Lubricated

Oil content below ISO 8573-1 Class 1 (≤0.01 mg/m³).

Oil-Free

PTFE Ring Technology

Oil-Lubricated

Cascaded Filtration

For PET blow molding, oil-free is the better choice. The 40% price premium (down from 60%+ a decade ago) gets recouped within a few years on most lines through lower filtration cost, lower filter element waste, and elimination of the oil contamination risk that comes with activated carbon filter timing. The argument for oil-lubricated machines with cascaded filtration is that they can achieve 0.003 mg/m³ at the outlet and that metal piston rings last 8,000 to 12,000 hours versus 3,000 to 5,000 for PTFE rings on oil-free machines. Both points are accurate. But activated carbon adsorption capacity holds through about 80% of element life and then drops off a cliff. Calendar-based replacement either catches it too early (wasted money) or too late (oily air reaching the blow molder). Plants that run oil-lubricated compressors and manage the filtration well can match or beat oil-free on total cost. Plants that do not manage it well, and that is a lot of plants, end up with periodic contamination events that are expensive to trace and expensive to resolve. For a new installation where the operating team's discipline level is unknown, oil-free eliminates a variable.

Number of Compression Stages

Total ratio of 40. Three stages gives about 3.4 per stage, four gives about 2.5.

Below 50 Nm³/h, three-stage. Above that, four-stage. The reasoning: four-stage machines run cooler per stage with higher volumetric efficiency and more even crankshaft loading. Three-stage machines cost less and have fewer parts, but per-stage temperature is higher and any cooling system weakness gets exposed. The crossover point shifts with electricity price and annual running hours. A plant running 4,000 hours a year with cheap power reaches a different breakeven than one running 8,000 hours at 0.9 RMB/kWh.

Interstage pressure distribution has a larger effect on machine performance than whether the machine has three or four stages. Equal ratios across all stages is not optimal. A three-stage arrangement that works well: first stage outlet 3.5 bar, second stage 13 bar, third stage 40 bar, giving ratios of 3.5, 3.7, and 3.1. The second stage runs the hottest discharge temperature under this arrangement. On most frame layouts the second stage cylinder sits in the middle of the machine where cooling connections are closest. The final stage ratio drops to 3.1, keeping discharge within about 160°C, which eases thermal conditions for the last stage's valves and rings. Equal ratio (3.4 across all three) pushes final stage discharge up 15 to 20°C, which shortens PTFE ring life by roughly 30%, give or take depending on ring compound. Over ten years, one to two extra ring changes per year.

Specific Power and FAD Labeling

Specific power in kW/(Nm³/min). For 40 bar blow molding machines the range runs from about 5.5 to over 9. Compressor electricity is 25 to 35% of total blow molding line energy consumption. At 100 Nm³/min, 8,000 hours a year, and 0.7 RMB/kWh, each 1 kW/(Nm³/min) difference means 560,000 RMB per year.

Comparing across brands requires knowing how each vendor labels FAD. Three methods exist. Actual Inlet Conditions: temperature, pressure, humidity at the intake. Standard Reference Conditions per ISO 1217 Annex C: 1 bar, 20°C, 0% RH. Site Conditions: whatever the buyer specifies. Same compressor, FAD at actual inlet conditions reads 8 to 12% higher than standard. Specific power calculated from the bigger number looks better.

Some quotation packages put inlet condition FAD on the spec page and standard condition FAD in the contract appendix. Both are correct. If a buyer does not notice the switch and compares one brand's inlet condition number against another's standard condition number, the ranking can invert. On any quotation, look for the reference condition note. If missing, request it. Convert everything to ISO 1217 Annex C before comparing.

Valves

Valves account for more unplanned downtime and more maintenance cost than everything else on a 40 bar compressor combined. This section goes deep because no other component decision has as much long-term financial impact.

Valves at 40 bar see 500 to 1,500 impacts per minute depending on stage and speed. Working temperature 80 to 170°C. Pressure differential across sealing faces from a few bar to over ten.

Components

Valve Engineering at 40 Bar

Ring valves dominate at 40 bar. Several concentric metal rings moving independently give large flow area and fast response. Plate valves (single disc or sector-shaped plates) have less flow area, limited lift, and above 600 rpm seat late, losing efficiency to backflow.

Among ring valves, go concentric. Eccentric types offset ring centers to squeeze more flow area from the same valve diameter, but sealing contact uniformity suffers. At 40 bar, uneven contact produces leak paths and hot spots. Some manufacturers have been pushing improved eccentric designs, and some of those designs test well on the manufacturer's bench at controlled conditions. Field data from plants running them 24/7 for multiple years is thinner, and what exists is mixed. Concentric ring valves are the conservative, proven specification at this pressure level, and the flow area penalty relative to eccentric types is not large enough to matter on most machine designs.

Inconel 718 springs hold their force below 200°C without measurable relaxation. Cost increment is 15 to 20% of the valve price. On the final stage, where discharge temperature is highest and pressure differential is greatest, Inconel springs extend valve overhaul intervals by a factor of 3 to 5 compared to stainless at the same conditions. On first and second stage valves running below 120°C, stainless springs do not relax much, and the higher spring cost may not pay back before the valve is overhauled for other reasons like ring wear or seat damage from foreign object ingestion.

Recommendation: specify Inconel 718 springs on the final stage always. On other stages, stainless is adequate for most operating profiles.

For maintenance planning and spare parts, the valve manufacturer and series and material grades tell you more than the compressor brand name. Valve supply channels run separate from the compressor OEM's distribution network. After the warranty period, plants can often source replacement valve plates and springs from the valve manufacturer's aftermarket channel at 20 to 30% below what the compressor OEM's parts desk charges. Whether this is practical depends on the plant's procurement rules and how comfortable the maintenance team is specifying parts by valve manufacturer part number rather than compressor OEM part number.

First stage valves fail less often and the parts are sometimes available from regional distributors or fabricated locally if the design is simple enough. Stocking priorities should follow failure frequency and procurement lead time rather than unit price.

System-Side Matters

Pressure dew point for blow molding air: -20 to -40°C. At 40 bar, -40°C pressure dew point corresponds to about -58°C atmospheric. If a dryer supplier reports atmospheric dew point and the buyer reads it as pressure dew point, the sizing error is severe. Adsorption dryer regeneration at 40 bar uses 10 to 18% of total air supply, which must be deducted from effective delivery volume.

AIS (Air Recovery System) captures exhaust air from high-blow and routes it to pre-blow. Vendor specs say 30 to 40% savings. On high-speed blow molders running above 2,000 bottles per hour per cavity, exhaust valve response speed limits recovery and field savings come in lower, often 25 to 32% rather than the 38 or 40% in the brochure. Sizing the compressor at 80% of the quoted AIS rate avoids undersizing.

Piping pressure drop costs about 2.5% additional compressor power per bar. Cumulative drop from discharge to blow molder inlet (through dryer, filters, receiver, headers, branches) commonly adds up to 2 to 4 bar. Piping design usually finalizes after the compressor purchase order is placed. On more than a few projects, the compressor spec assumed 2 bar total piping loss and the as-built piping delivered 3.5 because pipe diameter was reduced during value engineering and extra elbows were added to route around structural steel.

Machine Layout

W-type and L-type cylinder arrangements give better vibration balance than V-type on large displacement machines, with lower foundation loads.

Forged steel crankshafts outlast cast iron by roughly an order of magnitude in fatigue life. The difference shows up as cracking or bearing journal wear in cast iron after three to five years of 24/7 running while the forged steel crankshaft is still within service limits. Requesting material certificates during procurement is the only way to verify. Siad Macchine Impianti provides these as standard documentation; other manufacturers may require a specific request.

Water cooling only at 40 bar. Air cooling cannot remove enough heat at this pressure level.

Water quality matters more than water temperature for heat exchanger life. Hardness, pH, corrosion inhibitor concentration. The failure mode is internal scaling. Thermal conductivity of calcium carbonate scale is roughly 2.2 W/m·K versus copper at 385 or stainless steel at 16. Even half a millimeter of scale degrades heat transfer by 15 to 20%, and thicker buildup (1.5 to 2 mm, which accumulates over two to three years in hard water regions running without treatment) pushes degradation past 30%. Chemical descaling can restore most of the performance if done before the scale mineralizes into a hard crystalline layer, but once it reaches that stage (usually after three or more years), mechanical cleaning or tube replacement is needed. A basic water softening system or dosing pump for corrosion inhibitor at commissioning prevents the problem at negligible cost relative to the compressor purchase price.