Reusable versus single-use beer bottles: functional unit and reuse-loop boundary
A brewery delivers 1 million litres of beer per year to retail in 0.5 L glass bottles. It is comparing single-use bottles with a reusable bottle system. The single-use bottle weighs 300 g and is produced at 1.1 kg CO2e per bottle. The reusable bottle weighs 450 g and is produced at 2.1 kg CO2e per bottle; it is designed for 25 trips, but the brewery's return data show an average of 20 trips before breakage or loss, and 3% of reusable bottles are lost each year. Each reuse requires washing with 0.12 kWh electricity (0.40 kg CO2e/kWh), 1.5 L water and caustic soda. Empty bottles travel 200 km back to the brewery by truck; filled bottles travel 200 km to retail. The brewery's draft study credits the reusable system with avoided single-use bottle production for every trip and stops the assessment at the brewery gate. Retail and consumer transport are excluded, and no end-of-life is modelled.
1. Define the functional unit and reference flow, and identify the main boundary error in the draft study.
Read the worked answer
Functional unit: delivery of 1,000 L of beer to retail in 0.5 L bottles, including the bottle system required to deliver it. Reference flow: 2,000 bottle-uses per 1,000 L.
The main boundary error is that the draft stops at the brewery gate and credits avoided single-use production for every reuse, but does not include the full reuse loop: return transport, washing energy and chemicals, water, bottle losses, and end-of-life for both systems. It also excludes retail and consumer transport, which may differ because reusable bottles are heavier and may require different handling. A second error is the implicit assumption that each reuse avoids one single-use bottle without accounting for the 20-trip average and the annual loss. The correct approach is a cradle-to-grave or cradle-to-retail boundary, stated consistently for both systems, with the same functional unit.
2. Calculate the bottle-production burden per litre delivered for both systems, including the effect of the 20-trip average and the 3% annual loss. What data are still missing?
Read the worked answer
Per 1,000 L there are 2,000 bottles of 0.5 L.
Single-use production = 2,000 x 1.1 kg CO2e = 2,200 kg CO2e per 1,000 L, or 2.2 kg CO2e/L.
Reusable production per trip = 2.1/20 = 0.105 kg CO2e per bottle-use. The annual loss adds replacement production: 2.1 x 0.03 = 0.063 kg CO2e per bottle per year; if a bottle averages 20 trips per year, that is 0.00315 kg CO2e per trip. Total = 0.108 kg CO2e per bottle-use; per 1,000 L = 2,000 x 0.108 = 216 kg CO2e, or 0.216 kg CO2e/L.
Washing = 2,000 x 0.12 kWh x 0.40 kg CO2e/kWh = 96 kg CO2e per 1,000 L. Return transport: 2,000 x 0.45 kg = 900 kg = 0.9 t; at 0.08 kg CO2e/tkm over 200 km = 14.4 kg CO2e.
Missing data: wash water and caustic burden, caps and labels, retail and consumer transport, end-of-life, losses in the return loop, and the annual trip rate used to apply the 3% loss.
3. What end-of-life and use-phase issues must be included before claiming reuse is better, and how would you test the result?
Read the worked answer
Include end-of-life for single-use bottles (cullet recycling or landfill) and for retired reusable bottles; washing wastewater treatment; caustic production; and any extra consumer car trips to return bottles. The comparison must be cradle-to-grave with the same functional unit.
Test sensitivity on trips (15, 20, 25), loss rate (1-5%), electricity factor (grid versus renewable), wash temperature and detergent dose, and transport distance. Also test system expansion: reusable bottles may displace single-use bottles, but only if the return rate is actually achieved. If the reusable system wins only at 25 trips and the observed average is 20, the claim is fragile. Report the break-even number of trips and the break-even loss rate, and state that the result depends on consumer behaviour as well as brewery operations.