Published: May 15, 2021

Ten Green Bottles

The milkman, once an endangered species, is making a comeback. Deliveries of milk in reusable glass bottles have seen a surge in recent years in the UK (BBC 2018), and the COVID-19 pandemic has only strengthened demand for milk delivery services. Delivery services are convenient, but are also regarded as more sustainable: reusable glass bottles are seen as more environmentally friendly than plastic, and home delivery services can reduce CO2 emissions, compared to driving to the shops.

A bottling line. More than 1 million plastic bottles are made every minute

Glass does have its disadvantages, however. Glass bottles are much heavier than plastic, adding considerably to the environmental and financial cost of transport. Washing bottles for re-use also adds to the environmental impact. On the other hand, materials technology is continually improving, as manufacturers respond to environmental legislation and consumer demand for more sustainable packaging. So are reusable glass bottles really the most environmentally friendly packaging option? And what is the most sustainable solution for cafes?


Refill or Recycle?

Several life cycle analyses in different industries have found that refillable glass bottles are more environmentally friendly than plastic bottles. For example, a UK study on soft drink packaging found that refilling glass bottles more than three times would be enough to make them more sustainable than aluminum cans or plastic bottles (Amienyo et al 2012). However, the study showed that as the proportion of material that is recycled increases, glass’s advantage is eroded: if PET recycling rates were increased to 60%, then the bottles would need to be re-used at least 20 times to come out ahead.

Similarly, a major Life Cycle Analysis (LCA) carried out by the German Environment Agency (UBA) in 1995 found that re-usable glass bottles were slightly more environmentally friendly than other options available at the time (Kleene 2007). While re-usable glass was found to have the smallest environmental impact, paper-based cartons (such as Tetra Pak) came a close second, even though at the time most such cartons were disposed of in landfill.

Packaging options for milk (L-R): Carton, stand up pouch, and pillow pouch

However, since 1995, improvements in recycling rates and manufacturing technology meant that the gap narrowed, and a follow-up analysis was commissioned by the UBA five years later (Plinke et al 2020). This analysis found that while plastic bottles still had a higher environmental impact than other options, there was no longer a clear winner between re-usable glass and disposable cartons. This led the German Environment Minister, Jürgen Trittin, to say that “from an environmental point of view there is a stalemate between refillable bottles and disposable beverage cartons.”


The Cost of Washing

The UBA’s life cycle analyses factor in the environmental cost of washing and returning the glass bottles to be refilled — however, they assume that the bottles are washed at the bottling plant, which can be made very efficient. Some bottle return schemes, however, insist that the consumer must thoroughly wash the bottle before returning it. The instructions for one such scheme in the UK state that the bottle should be filled with hot water and detergent, scrubbed out, and then rinsed in hot water, before finally being put into the dishwasher to sanitise it.

The amount of hot water used to do this adds considerably to the carbon footprint: filling and then rinsing the bottle uses about 1.25 litres of hot water. If this water was heated from 15-55 °C in a gas boiler, it would release about 10g of CO2. Domestic dishwashers are also likely considerably less energy-efficient than the commercial washing process at a bottling plant, so the difference in emissions is probably even greater than this.

Considering that the total carbon emissions from a milk carton are approximately 25-35g (Fry et al 2010), then adding an extra 10 grams at minimum to the footprint of a milk bottle makes it a much less environmentally friendly option, from the point of view of carbon emissions.



There are even better options than re-usable glass bottles on the market. Since plastic bottles take less energy to manufacture than glass, re-usable plastic bottles could have an even lower environmental impact than glass. Simple polyethene sacks of milk (called ‘pillow pouches’) also take very little energy to manufacture because of the small amount of material required, resulting in emissions as low as 22g for a 1 litre container (Fry et al 2010). However, practical considerations mean these options make up only a very small part of the market (Keoleian & Spitzley 1999). Meanwhile, the more rigid ‘stand-up’ pouches, an increasingly popular option, result in roughly the same amount of emissions as paper-based cartons (Fry et al 2010). 

For cafes and other commercial settings, however, there is one option that is more practical than pouches, and may beat both cartons and re-usable glass: the pergal. Pergals are 3-gallon (13 litre) plastic sacks, usually housed in a cardboard box — similar to bag-in-box containers for wine. They are usually used in specially designed milk dispensers, ranging from a simple countertop fridge and tap, to more complex undercounter systems designed for cafes, such as the Milk Mate.

A milk dispenser fridge can save on packaging waste, but at the cost of running an extra fridge

Large containers result in less packaging being needed per litre of milk, potentially reducing the environmental cost. However, the disadvantage of pergals is that they typically need an extra fridge, built into the dispenser. As far as we know, there is no life cycle analysis comparing pergals to other milk packaging, so we made some calculations of our own.

One study estimates the total lifetime emissions of a fridge at 1670kg CO2, based on a lifespan of 10 years, or about 458g per day (Xiao et al 2015). This figure uses data on electricity generation in China in 2005, which at the time was very polluting, creating 1.1 kg per kWh of electricity (Bauer 2008). Since then, the emissions per unit of electricity have halved in China, and are lower still in Europe. Converting this to the current EU energy mix of 0.275 kg / kWh (European Environment Agency 2020) gives lifetime emissions of 165g CO2 per day. The fridge in the study uses 0.35 kWh per day, which is relatively efficient.

We don’t know of any figures for how much emissions a pergal can save, compared to milk cartons. However, a study of wine containers found that bag-in-box containers resulted in about 75% of the emissions of a paper-based carton (Le Guern & Tostivint 2010). Assuming the same ratio applies to milk cartons, a pergal would save about 9 grams of CO2 per litre.

Open Analyser app courtesy of Professor Steven Abbott, published under a CC-by licence. You can learn more about the app on Professor Abbott’s website.

We plugged these figures into Professor Abbot’s Life Cycle Analysis app, so that you can experiment with the results. Based on these assumptions, in order to offset the emissions of an extra fridge, a cafe would need to get through about 18 litres of milk every day of the year, so a pergal fridge will only reduce emissions for the busiest cafes.

If your fridge is more efficient than this, or you buy green electricity, your figures will be different. Perhaps we underestimate the amount of emissions a pergal can save? You can change the figures directly in the app by clicking on them and dragging the slider at the bottom to the new value.

Of course, if you can fit pergals into your existing fridge, then it’s even more efficient, since there are no extra emissions from the fridge to account for. For cafes where a pergal is not an option, however, then classic paper-based milk cartons can be just as environmentally friendly as returnable glass bottles — and are considerably more convenient to handle.


  1. Steven Abbott

    I often complain that debates about greener options are ill-informed and go more with what feels right than what is right. It’s great to see an intelligent discussion of the complex issues around milk containers. It was then amazing to find that you’d used Open Analyzer to help yourself and others to look how the various factors interact. I hope the passionate Barista Hustle crowd create their own versions of the model to create an intelligent debate.

    There are plenty of other complex issues around environmentally friendlier coffee so others should feel free to use Open Analyzer to create their own models and jointly explore those issues.

    Steven Abbott

    • BHLearn

      Thanks so much Professor Abbott for your generosity in sharing your software, and your profound wisdom with us 😀

Submit a Comment

30 Day Money Back Guarantee+
30 Day Money Back Guarantee

Signup for a personal BH Membership with a 30 day money back guarantee! Signup is risk-free and you can cancel your membership at any time!