
Two quotes leave the factory for the same 500 g jar of honey. One is glass, one is PET. The honey in them is identical, the freight invoice is not. The reason sits in a number most buyers never ask for: the empty weight of the container. This article works through that number, converts it into pallets and 40 ft containers, and puts it next to breakage, cushioning and barrier data so that glass vs PET honey jars freight cost becomes a calculation rather than a preference.
It is written for packers, importers and distributors buying jars by the container, not by the case. Every figure below is either published by a jar maker, a packer or a carrier index, or it is a calculation whose inputs are printed in the same row so you can substitute your own. Where no published figure exists, the row says so instead of guessing.

What one empty jar weighs, in glass and in PET
Container weight is where the two materials separate. Published jar weights for a 500 g honey jar run from 195 g to 240 g in glass against 26 g to 60 g in PET, a spread of roughly four to nine times depending on which mould and which lightweighting programme the glass comes from. Two independent sources put the lightest 500 g glass honey jar at 195 g and 196 g, so the floor of the glass range is well established.
| Honey fill (g) | Glass jar, empty (g) | Glass, filled + jar (g) | PET jar, empty (g) | PET, filled + jar (g) | Closure weight, both |
|---|---|---|---|---|---|
| 250 | 140–147 (published 250 ml honey jars) | 390–397 | 45 (published 250 g PET square jar) | 295 | Add separately; a basic plastic screw cap is published at about 2 g, pumps at over 10 g |
| 500 | 195–240 (lightest published 500 g jar 196 g; a 450 ml round jar 240 g) | 695–740 | 26–60 (26 g square, 33 g hexagonal, 45 g hexagonal, 60 g square) | 526–560 | Same basis; lug and twist-off caps vary by diameter and liner |
| 1,000 | 315–330 (published 700 ml jars) | 1,315–1,330 | 45 (published 1,000 g / 750 ml PET round jar) | 1,045 | Same basis |
Read the filled columns, because that is what the carrier charges you for. A 40 ft box of 500 g glass carries between 695 g and 740 g per unit before the carton. The same format in PET carries 526 g to 560 g. On a container of 24,000 jars that is 3.3 tonnes to 4.3 tonnes of difference, and it is the whole argument.
Two caveats belong with this table. Jar weight is mould-specific: hexagonal and heavy-based decorative jars sit at the top of each range and lightweight rounds at the bottom, so the spread inside glass is wider than the spread between the cheapest glass and the heaviest PET. Closures are not included in any of these figures, and a heavy metal lug cap on glass adds grams that a light plastic PET cap does not.
The millilitres behind 250 g, 500 g and 1 kg
Net weight sells the honey; volume sizes the jar. Honey is denser than water, so a 500 g label does not need a 500 ml jar. Working from the widely used 1.42 kg per litre density, 250 g occupies about 176 ml, 500 g about 352 ml and 1 kg about 704 ml. That density is moisture-dependent: published conversion tables built on reference data give specific gravity around 1.4237 at 17% water falling to about 1.4027 at 20% water, and the EU compositional standard caps moisture at 20% with defined exceptions.
The practical consequence for a buying order is headspace. A packer filling 500 g into a 500 ml jar is paying to ship air and is carrying a bigger, heavier mould than the fill needs. A 500 g fill in a 375 ml jar saves glass weight, carton volume and pallet height at the same time, which is why the 375–405 ml band is the common commercial choice for a half-kilo.
Name the jar by its nominal volume in the purchase order, not by the honey weight it is intended to hold. A “500 g jar” is not a specification. “405 ml, brimful, 63 mm twist-off finish, target empty weight 195 g” is a specification, and it is the only form in which a supplier can be held to a freight number.
Pallets and containers: the arithmetic that decides the freight rate
A 40 ft standard container is fixed by ISO 668: external 12,192 × 2,438 × 2,591 mm, internal about 12,032 × 2,352 × 2,385 mm, internal volume 67.7 m³, maximum gross weight 30,480 kg. Container makers publish tare weights from about 3,630 kg to 3,780 kg for the same box, so the usable payload lands between roughly 26,700 kg and 26,850 kg depending on whose box you load. Note the internal height: published figures range from 2,385 mm to 2,393 mm, and one pallet layer can turn on that.
Pallet counts follow from the pattern. Published loading references put 25 Euro pallets (1,200 × 800 mm) or 21 industrial pallets (1,200 × 1,000 mm) into a 40 ft box in a single tier. A 1,200 × 1,200 mm pallet does not tile two across a 2,352 mm internal width, so it loads 10 per tier and no more.
| Configuration | Jars per pallet | Pallets per 40 ft | Jars per 40 ft | Cargo weight per 40 ft (kg) | Basis and what limits the load |
|---|---|---|---|---|---|
| 500 g glass, 12-jar cartons, Euro pallet, 6 layers | 864 | 15 | 12,960 | 9,588 | Published packer loading plan; limited by the plan, not by payload |
| 500 g glass, 12-jar cartons, industrial pallet, 6 layers | 1,152 | 21 | 24,192 | 17,758 | Published packer loading plan; limited by the plan |
| 500 g glass, loose in trays | 1,960 | 21 by floor, 19 by weight | 37,240 | 25,882 | Pallet build published; count capped by the 26,730 kg payload at 695 g per filled jar |
| 1 kg PET, 1.2 × 1.2 m pallets | 960 | 10 | 9,600 | 10,080 (filled at 1,045 g) | Pallet build published at 960 units; count capped by pallet footprint, not weight |
| 12-jar cartons, any size, packer’s reported container counts | Varies | Varies | 3,200 cartons PET vs 2,250 cartons glass | Not published | Published by a honey packer; PET loads about 1.4 times the cartons of glass |
The third row is the one to argue about internally. Loose-tray loading of 500 g glass reaches about 37,000 jars before the payload stops it, and at that point a third of the container’s volume is still empty air. Glass weighs out. PET in the fourth row is the opposite case: 9,600 jars of a 1 kg format fill the floor pattern at 10 tonnes and leave 16 tonnes of unused payload.
That inversion is the whole freight story. PET containers are limited by geometry and glass containers are limited by the scale. Once you know which side of that line your format sits on, the optimisation is different: PET buyers should be arguing about pallet footprint, carton height and stacking; glass buyers should be arguing about grams per jar.

Where glass hits the weight ceiling before the space runs out
Run the division for your own specification. Take the payload for the box you are actually loading, somewhere between 26,700 kg and 26,850 kg for a 40 ft standard, and divide by the filled weight including carton, dividers and pallet. At 695 g filled and 12 per carton with carton and divider at 250 g, one carton is 8.6 kg, which matches a published packer figure of 8.6 kg per 12-jar box of 500 g glass. The payload then allows about 3,100 cartons and 37,000 jars.
Volume, meanwhile, would allow more. A 12-jar carton at roughly 340 × 255 × 95 mm occupies about 0.0082 m³, so 67.7 m³ holds over 8,000 of them before pallet gaps, dunnage and load securing take their share. The gap between the two answers is the air you are paying to ship.
Three levers close it, and they are the ones worth putting in a supplier brief. First, empty jar weight: moving from a 240 g mould to a 196 g mould on a 500 g jar removes 44 g per unit, which is 1,100 kg on a 25,000-jar container and about 1,600 more jars inside the same payload. Second, carton height: a jar 2 cm shorter can win an extra layer on every pallet, and the saving multiplies across 21 pallets. Third, pallet pattern: a format that tiles a 1,200 × 800 mm pallet without a turned pattern loads more pallet positions per container.
Volume-per-jar published by a bottle maker for a 40 ft high cube shows the same mechanics in another product: a 76 mm diameter, 95 mm tall 250 ml honey jar stacks 24 layers and reaches 113,400 pieces by volume across 21 pallets, but at 500 g each that would be 56 tonnes, so the payload caps the real order far below the volumetric number. Published calculators of this kind switch to the weight answer automatically, and that is the behaviour to demand from your supplier’s loading plan.
Breakage, cushions and the cost of the carton you did not specify
Glass breakage in transit is a packaging-design problem more often than a glass-quality problem. Published guidance for imported glass jars puts the industry-accepted range at 0.5% to 2% for well-packaged shipments and over 5% where packaging is poor, and states that rates above 1% are almost always caused by carton design, dividers, pallet build or the loading plan rather than by the glass itself. Direct-to-consumer food packing targets below 0.5% breakage across all orders, on the argument that above 1% the combined replacement, reshipment and service cost exceeds the cost of upgrading the insert.
The published fixes are specific and cheap compared with a claim. Full-height dividers in every carton cell stop the glass-on-glass contact that vibration turns into failure; double-wall cartons are specified for jars over 500 g filled and single-wall for lighter ones; dunnage airbags between pallets stop lateral shift on long voyages. A carrier that publishes drop-test data for 1-gallon jars shows break rates from 18.3% per 1,000 units at 2.8–3.2 mm wall thickness down to 0.7% at 4.0–4.5 mm, and reports that in a majority of high-breakage investigations the root cause was void fill or edge loading rather than glass quality. Published insert costs for glass protection run about 0.12–0.32 USD per unit for moulded pulp and 0.18–0.45 USD for die-cut EVA foam, before any per-parcel handling.
PET does not carry these lines at all. That is its structural advantage, and for any channel where a parcel, not a pallet, is the shipping unit, it is usually the deciding one. The cost to compare is not the jar price, it is the jar price plus the insert, plus the freight on the extra grams, plus the claims.
What the barrier difference does to shelf life
Glass is a total barrier; PET is a good one. A peer-reviewed comparison of 30 cm³ bottles at 40 °C and 75% relative humidity measured water vapour transmission at 0.07 mg per day for glass against 9.9 mg per day for PET and 0.8 mg for HDPE, and ranked oxygen barrier in the same order: glass, then PET, then HDPE. Published packaging comparisons put PET oxygen transmission around 2–5 cm³ per m² per day against effectively zero for soda-lime glass, and put water loss from PET at 0.1% to 0.5% per year.
What that means for honey is narrower than the marketing suggests. Honey is a low-water-activity, high-sugar, acidic product; the failure modes that bother a bottler are moisture pickup at the surface, crystallisation habit and aroma loss, not oxidation in the way a juice or an oil would suffer it. A published syrup-vessel comparison claims 24+ months of shelf-life stability in glass against 6–12 months in PET, but that figure is for a high-brix syrup and should not be transferred to honey without your own stability data.
The honest position for a specification sheet: for ambient retail honey with a stated best-before of 18–24 months, glass removes a variable; PET at standard wall thickness will need a moisture-barrier closure and, if you want a long date, either a co-extruded barrier layer or a shorter declared date. Either way, ask the supplier for migration and transmission test reports on the exact jar and closure combination, and date-stamp them.
Food contact rules both materials still have to pass
Material choice does not remove the compliance file. In the United States, indirect food additives used in food contact articles are cleared under the FDA’s food additive regulations, with the general provisions in 21 CFR Part 174 and the polymer-specific clearances in Part 177 of the same subchapter. In the European Union, food contact materials are governed by the framework regulation described on the European Commission’s food contact materials pages, supported by national provisions in several member states.
For a container arriving at an EU or US port, the practical checklist is short. Declaration of compliance naming the regulation and the specific migration limits; overall migration test conditions matched to your filling temperature and contact time; a statement on the closure liner, which is a separate material from the jar; and, for the wooden pallets the jars arrive on, heat-treatment or fumigation marking under the IPPC standard. A jar that passes and a pallet that fails costs the same delay.
Order timing across two hemispheres
Jars are bought against a harvest that arrives once. Northern hemisphere packers extracting from roughly May to August are packing from September to December, so the jar container has to be booked and landed before the honey is in the tank; a jar order placed in August misses the year it was bought for. Southern hemisphere packers extracting from roughly November to March pack from January to April, which moves the jar order into the middle of the year.
That offset matters because both hemispheres compete for the same factory slots and the same vessels. Published market reporting for October 2026 put the Drewry World Container Index at 4,485 USD per 40 ft, with Shanghai–Rotterdam at 3,522 USD per 40 ft and the Freightos Baltic Index at 3,344 USD, with weekly swings of several per cent. Treat those as a reference point with a date on it, not as a quotation: the rate that matters is the one your forwarder quotes on your lane in your booking week, and the northern pre-Christmas peak and the Chinese pre-Golden-Week peak are the two weeks it is most expensive to need a box.
Unit landed cost per jar, and the format each channel should buy
The table below is deliberately incomplete in one column and complete in the others. Unit prices move with resin, cullet, energy and order size, so any number printed here would be wrong next month; ask the supplier for a current quotation and put it in the first row. Freight uses the published index figure for Shanghai–Rotterdam in October 2026 and the container counts from the rows above, so it is a worked result rather than a quotation.
| Cost line | 500 g glass | 500 g PET | How to use the row |
|---|---|---|---|
| Jar unit price, FOB | Ask the supplier for a current quotation | Ask the supplier for a current quotation | Quote against a named mould, empty weight and finish, not against a size alone |
| Carton, dividers and inserts per 1,000 jars | Dividers and double-wall cartons required; insert costs published at 0.18–0.45 USD per unit for EVA foam and 0.12–0.32 USD for moulded pulp in parcel packing | Single-wall cartons normally sufficient; no parcel insert needed | This row is where the two materials converge or separate depending on the sales channel |
| Jars per 40 ft container (published plans) | 24,192 in 12-jar cartons on industrial pallets; up to 37,240 loose in trays before payload caps it | About 1.4 times the glass carton count on a packer’s published comparison | Use your own loading plan; a supplier’s number should arrive as a drawing |
| Ocean freight per jar at 3,522 USD per 40 ft | About 0.15 USD at 24,192 jars; about 0.09 USD at 37,240 jars | About 0.10 USD at 33,888 jars | Worked from the index figure dated October 2026 and the counts in the row above |
| Breakage allowance | 0.5%–2% published for well-packed glass; over 5% reported where packaging is poor | Not material in palletised freight | Put the number in the cost model before it appears in a claim |
| Channel recommendation | E-commerce: only with a tested insert. Supermarket: yes, ambient shelf. Gift: yes, the barrier and the weight read as quality. | E-commerce: the default. Supermarket: yes, where the shelf allows. Gift: only with a heavy-wall mould and a metalised closure. | Choose by shipping unit first, then by shelf |
Read across the last row and the recommendation follows without sentiment. A business that ships single jars through a parcel network should buy PET, because its breakage, insert and reshipment costs are the ones the glass format cannot recover. A business that ships pallets to a supermarket and sells across a two-year date should buy glass, because the barrier removes a shelf-life variable and the freight penalty is absorbed across 24,000 units. Gift and premium channels can carry either, but the PET version has to be specified as a heavy-wall mould with a good closure, or it will be judged against glass on the shelf and lose.
The two formats can also be bought together. Enzher supplies both sides of this comparison and loads them in the same box: the glass range sits under glass jar honey containers, small formats for hotel, airline and gift programmes are covered by mini custom glass honey jars, and the PET side, including a 250 g jar published at 45 g empty and a 500 g jar at 60 g empty, is listed as the black square PET honey container and as versatile PET square bottles from 180 ml to 500 g. A mixed box of glass for the retail shelf and PET for the parcel channel is usually the cheapest answer to a question that gets asked as if it had one answer.
Production context for the volumes you are planning against is published in the FAO FAOSTAT crops and livestock products dataset, and the pollination dependence behind those volumes is set out by the FAO pollination programme.