Environmental Benefits of Refillable Beverage Containers

A transition from one-way to refillable beverage containers could have many environmental benefits. These potential benefits include reductions in:

  • greenhouse gas emissions,
  • carbon monoxide emissions,
  • solid waste generation,
  • energy consumption, and
  • water consumption.

The discovery of these environmental benefits has come through life-cycle analysis. Life-cycle analysis (LCA), which is also called life-cycle assessment or eco-balances, has become a useful methodology for evaluating the potential environmental impacts and natural resource demands of beverage containers.

Around 1970, Harry Teasley, Jr., of Coca-Cola directed one of the first LCA studies of beverage containers. Like the Coca-Cola study, most of the early LCA studies of beverage containers were investigations of a particular type of container that a company was using or was wanting to use. Such investigations can help a company comply with regulations, reduce waste, cut costs, or evaluate its use of scarce or expensive raw materials [DUDA]. In the 1990s, governments began conducting LCA studies in order to guide them in policymaking. One country, in fact, has directly incorporated LCA into its policy. Under the Dutch Packaging Covenant, bottlers cannot substitute refillable with one-way beverage containers unless they can demonstrate that the overall environmental impact of their proposed one-way containers is less than or equal to the impact of their refillable containers. The results of a recent LCA study, however, told Dutch soft-drink bottlers to continue refilling. Another application of LCA to policymaking is Denmark's packaging tax, whose rates are based on the environmental impacts of different packaging materials. For more details about these policies, see the Europe page.

Although ways to improve LCA methodologies are being investigated, LCA still has many methodological limitations. The usefulness of an LCA study is limited also because it may include some processes and exclude others and may evaluate only a limited set of environmental impacts and resource demands. Because the geographic scope, the containers under investigation, the beverages included, and other factors of LCA studies determine their findings, applying the findings of one LCA to other beverage markets is difficult. For these three reasons, the interpretation of LCAs on this web page refrains from concluding in absolute terms that one beverage packaging system is environmentally superior to another.

A life-cycle analysis study can serve as a comparison not only between refillable and one-way containers but also between refilling and recycling. Many LCA studies assume that one-way containers are recycled at a given rate but must also consider collection, sorting, and other recycling processes in order to effectively compare refilling to recycling. Comparisons between refilling and curbside recycling and between refilling and deposit-return systems for one-way containers would be useful.

Life-Cycle Analysis of Beverage Containers

Life-cycle analysis studies attempt to estimate the environmental impacts and natural resource demands of beverage containers per unit volume of packaged beverage. The natural resources usually include energy, water, minerals, timber, land, and fossil substances used either as fuel or as raw materials. Environmental impacts usually include solid waste, emissions to water, and emissions to air but may also include hazardous waste, organic waste, nuclear waste, noise, and dust. The LCA findings may be expressed in terms of individual pollutants and resource demands or in terms of impact categories such as the following: global warming, which includes greenhouse gas effects; acidification, which includes acid rain; and ground-level ozone formation, which contributes to smog. Finally, the sponsor of an LCA study may apply a ranking scheme to express the findings in terms of the relative acceptability of the environmental impacts and resource demands [BWIL].

Life-cycle analysis studies usually consider the entire life cycle of a beverage container, from the extraction of the natural resources that are used to manufacture it to the disposal of the container through recycling, landfilling, or incineration. This life cycle includes several processes that contribute to the environmental impacts and the resource demands of the container. Among these processes, an LCA of beverage containers may include the following: extraction of raw materials, production of container materials from both virgin and recycled materials, manufacture of the containers, filling, distribution, retailing, consumers' refrigeration of the product, return of the containers, washing, recycling, incineration, and landfilling. An LCA can also include the manufacture of secondary and transport packaging, the production of materials for this packaging, and the reuse, recycling, or other disposal of this packaging. Other processes may include the generation of energy by power plants, the production of chemicals for washing the bottles, the generation of energy and of solid waste by incineration, the generation of methane from landfills, and the dynamics of the market for recovered materials [CHAL, pp. 25-40].

Many of the processes that make up the life cycle of a beverage container are characterized by critical parameters which can significantly affect the findings of an LCA. These parameters include the trippage of refillable containers, the recycling rates of both one-way and refillable containers, the transportation distances involved in delivering beverages from the bottling plants to the points of sale, and the recycled content of new containers. The values of these parameters can be actual values or sets of arbitrarily-chosen values which are used to ascertain how the LCA results vary with these values. Reliable values are usually available for the critical parameters but may not be available for other input data. Studies conducted for individual companies can use the company's proprietary data [DUDA], but studies conducted for governments may not have access to such proprietary data and usually must use data from a variety of sources. The variety of sources of data for LCAs causes variation in the data and in turn limits the precision of the results of LCA studies. Normal statistical variation also affects the precision of LCA results. The variation in the data is one limitation of LCA methodology. Other limitations include the following [BWIL][CBA, p. 18].

  • The availability or the quality of data can limit the accuracy of an LCA study.
  • The methods used for estimating or evaluating the environmental impacts are limited by the assumptions on which they are based. Estimation or evaluation methods may not be available for all potential impacts.
  • The purpose of the study or other subjective influences may determine which processes are included, which data sources are used, and which impacts are included. Time and budget limitations may also determine the scope of the study.
  • The results of an LCA study become less useful over time as the manufacturing, filling, transportation, handling, disposal, and other processes become more efficient and as the weight, recycled content, and other intrinsic features of beverage containers change over time.

What Life-cycle Analyses Most Often Reveal

Because almost every LCA study of beverage packaging systems is unique, a useful way to present the results of several LCA studies may be to tally their findings for specific environmental impacts and resource demands. Such a presentation can suggest how consistently one type of beverage container compares to another with regard to a set of criteria but cannot conclude absolutely that one type is better than another. By tallying the results of eleven different LCA studies, the following presentation shows comparisons between beverage containers with regard to energy consumption, solid waste generation, water pollution, and air pollution. Five well-known air pollutants are considered. First, carbon monoxide (CO) mainly affects human health by impairing the respiratory and cardiovascular systems and in turn impairing many functions of the brain. Nitrogen oxides (NOx) also impair the respiratory system, and they contribute significantly to ground-level ozone formation and to acid rain. One of the sulfur oxides (SOx), sulfur dioxide (SO2), also contributes significantly to acid rain [EPA]. Finally, carbon dioxide (CO2) and methane (CH4) are greenhouse gases [EDF]. The following table lists the eleven LCA studies for our tallies.

Life-cycle Analysis Studies for Tallies
Author Sponsor Initials Year Sources
Lundholm and Sundstrom Tetra Pak, Inc. LS 1985 [HEC]
Franklin Associates NAPCOR* FA 1989 [HEC][SAPH]
Sundstrom Swedish Brewers Association GS 1990 [HEC][SAPH]
Deloitte & Touche Canada, Inc. Tetra Pak, Inc. DT 1991 [HEC][SAPH]
Proctor and Redfern, Ltd. Liquor Control Board of Ontario PR 1991 [HEC][SAPH]
First Consulting Group OMMRI** FCG 1992 [HEC][SAPH]
Schmitz, Oels, and Tiedemann German Fed. Env. Agency (UBA) UBA1 1995 [HEC]
Chalmers Industriteknik and Inst. for Prod. Dev. Danish EPA DEPA 1998 [CHAL][HEC]
Prognos, IFEU, GVM, Pack Force, and UBA German Fed. Env. Agency (UBA) UBA2 2000 [OKO]
Gesellschaft für Umfassende Analysen, GmbH Austrian Ministry of the Env. GUA 2000 [GUA]

* National Association for Plastic Container Recovery
** Ontario Multi-Material Recycling, Inc.

Before proceeding to the tallies, some background information about them is needed. As the table above indicates, the tallies rely on one source for information about several of the LCA studies considered here. Original copies of three other LCA studies were obtained. In regard to a particular environmental impact, the "number reporting the impact" refers to the number of studies for which results were available from the sources consulted. For each of the five types of air pollutants, the estimated emissions in grams or kilograms is the basis for comparison. For each LCA study for which water pollution estimates were available, the total weight of all of the water pollutants is the basis for comparison. However, no two LCAs included the same combination of water pollutants, and some appeared to have omitted some types of pollutants that could have tipped the balance in favor of the otherwise unfavorable container. Finally, the comparisons of refillable glass to refillable PET bottles involve glass bottles that are about 25-35 percent smaller in volume capacity than the PET bottles. These comparisons include 1-liter glass to 1.5-liter PET, 0.7-liter glass to 1-liter PET, and 330-ml glass to 500-ml PET.


Tally of Results of 8 LCAs--Refillable vs One-way Glass Bottles
Environmental Impact Air Pollution Water Pollution Solid Waste Energy
Number favoring one-way containers 1 0 0 0 0 0 0 2
Number favoring refillables 3 2 2 3 4 4 5 5
Number reporting the impact 4 2 2 3 4 4 5 7
Tally of Results of 7 LCAs--Refillable Glass Bottles vs Aluminum Cans
Environmental Impact Air Pollution Water Pollution Solid Waste Energy
Number favoring cans 0 0 0 0 1 2 1 3
Number favoring refillables 4 3 3 4 3 1 3 2
Number reporting the impact 4 3 3 4 4 3 4 5
Tally of Results of 5 LCAs--Refillable vs One-way PET Bottles
GS, DEPA 500 mL, DEPA 1.5 L, GUA mineral water, GUA soft drinks
Environmental Impact Air Pollution Water Pollution Solid Waste Energy
Number favoring one-way containers 1 0 0 0 0 0 0 0
Number favoring refillables 4 4 4 5 5 2 4 5
Number reporting the impact 5 4 4 5 5 2 4 5
Tally of Results of 3 LCAs--Refillable Glass vs Refillable PET Bottles
Environmental Impact Air Pollution Water Pollution Solid Waste Energy
Number favoring glass 1 0 0 2 0 0 0 0
Number favoring PET 2 2 2 1 3 2 2 1
Number reporting the impact 3 2 2 3 2 2 2 1

In terms of five types of air pollutants, the tallies indicate that the use of refillable containers puts less pollution into the air. The conclusion that refillables generate less solid waste per unit volume of packaged beverage should come as no surprise. The comparisons of refillable to one-way bottles, furthermore, reveal that refillables emit less water pollution and use less energy. For water pollution and energy use, however, the comparisons of refillable glass bottles to cans apparently favor cans. In terms of energy use and of the environmental impacts considered, finally, PET appears to be the better material for refillable bottles.

Because estimates of water consumption were available for only one of the LCA studies considered here, this resource demand is not included in the tallies. That particular LCA study, the Danish EPA study, found that 330-ml refillable glass bottles use less water than do 330-ml one-way glass bottles and do 330-ml aluminum cans. For both the 500-ml and the 1.5-L PET bottles, the DEPA study found that the refillable systems use less water. In addition, a review of some literature concluded that the amount of water required to wash refillable glass bottles is much less than the amount used to manufacture new one-way glass bottles for a given volume of beverage [SAPH, p. 33].

To better understand how refillable containers can generate less pollution and waste and can use less of the earth's precious natural resources, further reading about life-cycle analysis studies of beverage containers is highly recommended. The Danish EPA has an English-language version of its 1998 study that is written for a general audience [CHAL]. If you know German, then the UBA2 study would be worth reading [OKO]. In his 1994 book [SAPH], finally, David Saphire presents a thorough and worthy explanation of the environmental aspects of beverage containers that puts less emphasis on LCA studies themselves. However, understanding LCA studies themselves is necessary because they have been and will continue to be an important part of the debate over refillable beverage containers.

Environmental Cost-Benefit Analysis

Environmental Cost-Benefit Analysis (CBA) takes LCA a step further by assigning monetary values to the environmental impacts and natural resource demands of beverage packaging systems. While the assignment of these values has many methodological limitations [CBA, pp. 17-20][LEVY, pp. 79-81], its ethical limitations probably draw the most vociferous criticism. Many critics of CBA argue that the environment is something on which you cannot place a monetary value.

Two CBA studies are considered here. The tallies of LCA results use some of the findings from a CBA study that was completed for the Austrian Ministry of the Environment in 2000 [GUA]. In 2001, the consulting firms RDC-Environment and Pira International completed a CBA study for the European Commission (EC), who intended to use the findings to set new recovery targets for the EC Directive on Packaging and Packaging Waste. This study compared 330-ml refillable glass bottles with one-way glass bottles of the same size by investigating the container manufacturing, filling, distribution, and waste management processes under the following assumptions [CBA].

  • The return rate for the refillable bottles is 100 percent.
  • All bottle losses occur during washing and refilling.
  • The round-trip distance from the warehouse to the store is 100 Km.
  • Consumers recycle their commingled bottles and other containers only at drop-off centers. Industry bears all of the costs of recycling.
  • The portion of one-way bottles that are not recycled is split equally between landfilling and incineration.

The study concluded that refillable glass bottles cost less environmentally than one-way glass bottles do whenever the distance from the bottling plant to the warehouse is less than 3,500 Km with 20 trips for the refillable bottle and a 91 percent recycling rate for the one-way bottle; less than 4,200 Km with 20 trips and a 42 percent recycling rate; less than 2,300 Km with 5 trips and a 91 percent recycling rate; and less than 3,000 Km with 5 trips and a 42 percent recycling rate. The RDC-Pira study also attempted a similar comparison for PET bottles, but it apparently omitted the costs of washing bottles.


For more information about some of these sources, go to the annotated bibliography (B) or to the links.


  • [BWIL] Bothwell, George. "Life Cycle Assessment: How Precise?" Beverage World International Dec. 1993: 36.

  • [CBA] RDC-Environment and Pira International. Evaluation of Costs and Benefits for the Achievement of Reuse and the Recycling Targets for the Different Packaging Materials in the Frame of the Packaging and Packaging Waste Directive 94/62/EC. (Draft) Brussels: European Commission, 2001. (B)

  • [CHAL] Chalmers Industriteknik and Institute for Product Development. Life Cycle Assessment of Packaging Systems for Beer and Soft Drinks. Environmental Project No. 399. Copenhagen: Danish Environmental Protection Agency, 1998. (B)

  • [DUDA] Duda, Mark, and Jane S. Shaw. "Life Cycle Assessment." Society 35.1 (1997): 38.

  • [EDF] Environmental Defense. "Global and Regional Air." http://www.environmentaldefense.org/programs/GRAP Also see http://www.scorecard.org

  • [EPA] US Environmental Protection Agency. "Criteria Pollutants." http://www.epa.gov/oar/oaqps/greenbk/o3co.html

  • [GUA] Gesellschaft für Umfassende Analysen (GUA), GmbH. Volkswirtschaftlicher Vergleich von Einweg- und Mehrwegsystemen. Vienna: Austrian Ministry of Environment, 2000. (B)

  • [HEC] Lachance, Renuad, and Paul Lanoie. Refillable and Disposable Beer Containers: An Analysis of the Environmental Impacts. Montreal: École des Hautes Études Commerciales, 1999. (B)

  • [LEVY] Levy, Geoffrey M., Ed. Packaging, Policy, and the Environment. Gaithersburg, Maryland: Aspen Publishers, Inc., 2000. (B)

  • [OKO] Prognos GmbH, Institut für Energie und Umweltforschung Heidelberg, Gesellschaft für Verpackungsmarktforschung mbH, Pack Force, and German Federal Environment Agency (UBA). Ökobilanz für Getränkeverpackung II Berlin: UBA, 2000. (B)

  • [SAPH] Saphire, David. Case Reopened: Reassessing Refillable Bottles. New York, INFORM, Inc., 1994. (B)