How well are FOGO bins working: NSW food and garden bin audit analysis
Prepared for the NSW Environment Protection Authority. Analysis of 8,718 bins across 20 councils, audited between 2011 and 2017.
Summary published
Garden organics collection is close to solved. Food waste is not, and the two travel in the same bin.
Rawtec analysed 26 audited areas across 20 NSW councils to separate the two questions: how much of the available material is being captured, and what is getting in that should not be.
The finding that matters most for council budgets is about who participates rather than how well.
Of available garden organics diverted
Of available food waste diverted
Average contamination by weight
Of residents diverted no food waste at all
What the audits found
Garden organics is working, food waste is not
Households discarded 13.6 kg of food and garden organics a week and put 83% of it, 11.37 kg, in the FOGO bin. Split apart, 98% of available garden organics was diverted (10.14 kg per household per week) against 38% of available food waste (1.20 kg). Council results for food ranged from 5% to 78%.
The problem is participation, not diligence
Roughly 50% of residents were diverting no food waste at all, while only about 30% were contaminating their bin. The report’s conclusion is direct: food waste diversion is more likely to rise by increasing the number of participating households than by getting existing participants to be more thorough.
Bin size and collection frequency drive the result
Councils with a fortnightly general waste collection achieved higher food diversion than weekly ones, and councils with a 120 or 140 litre general waste bin did better than those with 240 litres. The best performing configuration, a small bin collected fortnightly alongside a weekly FOGO bin, reached 54%. The worst, a large bin collected weekly, managed 14%.
Services get better with age
Services running longer than a year averaged 45% food efficiency, against 25% for those under a year and 26% for trials.
Contamination is concentrated in a few bins
Average contamination was 2.6% by weight, or 0.30 kg per household per week, with council results from 0.04% to 17.83% and the middle half between 1.17% and 5.26%. In the two councils with bin-by-bin data, 68% and 87% of bins had no contamination at all. Among the bins that did, the average was 1.62 and 1.10 kg, against 0.20 and 0.35 kg when every bin is counted.
What the contaminants are
By frequency: plastic, containerised food, metals, other organics such as leather, rubber and oils, then miscellaneous items.
What the report suggested doing
Target the half of households putting no food in the bin. That is where the tonnage is, not in the households already doing it.
Because contamination sits in a minority of bins, a bin tagging programme aimed at those households is likely to work better than broad education.
A smaller general waste bin on a fortnightly cycle produced almost four times the food diversion of a large weekly bin.
A service under a year old is not a fair test. Efficiency nearly doubled once services passed the one year mark.
Seven things to fix in future audits
Rawtec found enough data quality problems in the source audits to make recommendations about the audits themselves.
Keep emphasising randomised sampling. Some audits in the dataset were not randomised, which limits what can be read from them.
Give clear guidance on stratified sampling for multi-unit dwellings. Several councils sampled them inadequately.
Emphasise representative sample sizes. The guideline minimum is 220 bins and several audits came in under it.
Put standard contamination definitions in the audit guidelines. Councils were using different ones, so the rates were not directly comparable.
Support auditor training on waste generation rate calculations. Rawtec found errors in them.
Record instances of gross contamination, above 20%, separately, so the effect of including or excluding them can be tested.
Require the key audit information in the report and retain the raw data. Missing metadata made some audits unusable.
Related reports
This analysis was updated in 2020 with a larger dataset.
The full report includes the configuration-by-configuration breakdown and the two bin-by-bin council analyses.
Working on something like this?
Most of our projects start with someone inside an organisation who already knows what needs to change and needs the evidence to back it in. If that is you, we would like to hear from you.