Estimating ‘Bottles per Hectare’ in an African City
How our plastic waste tracking platform helped calculate the volume of single-use plastic bottles discarded every day in Maputo
Did you know that more than 52 single-use plastic bottles can accumulate in one space, a hundred metres on each side, every month? That’s equivalent to 175 per square kilometre, every day. In a city the size of Maputo, that could mean over 61,000 plastic bottles discarded in ONE DAY.
Using ‘Bottles per Hectare’ to monitor plastic waste
Nothing prepares you for the first time you do a plastic waste audit with your bare hands. At a local beauty spot or a busy traffic stop, you put on some gloves and with a few other people and some reusable sacks, start collecting discarded bottles. You always knew there were a few bottles around — they gleam at you whenever you walk past this spot — but now you realise they are everywhere. Shiny new ones with a bit of fizzy drink still inside. Two week old ones with the label half torn off and some dirt around them. Ancient ones squashed flat, dried brittle by the sun and gone opaque from the scouring action of wind, sand and rain.
Within ten minutes your colleagues are bringing back the first full sacks and dumping their contents in the collection area for sorting and logging. Then they go out for more. Even in a relatively small area, there are just so many bottles. You begin to make some mental calculations to scale it up…if this little place has so many… how many does the whole city produce? Or the whole country?
This article describes Unwaste.io’s first steps toward finding a useful answer to that question. ‘Bottles per hectare’ (BPH) is a useful starting point based on data collected by Wastebase, our crowdsourced data collection portal.
Once we have refined a model for scaling up data about plastic waste, we’ll build it into Wastebase. That means anyone can find out the daily BPH of a Wastebase data area, just by panning and zooming the Wastebase map.
The ‘Bottles per Hectare’ concept
The number of plastic bottles discarded in a given location every day, expressed as a quantity per hectare (ha).
Of course we all remember that a hectare is a square 100m on each side, and that there are one hundred hectares in a single square kilometre, right?
In general, when comparing two different locations in terms of BPH, the location with a higher score is likely to have more visible litter, contribute more to plastic waste which ends up in the ocean (in coastal areas) and be less attractive to spend time in.
Methodology
This is a concrete example from our Wastebase Data Partner in Mozambique (Repensar Environmental Education Cooperative — Repensar MZ for short). It shows how we collected the data and then how we used it to calculate BPH.
The “Praça da Paz” (Peace Square) is a municipal public garden in central Maputo, just under 2 hectares in area (19,800 sqm). It sits between one of the city’s main arterial roads and a popular shopping mall. About five years ago the garden was refurbished and is generally maintained in good condition. It is popular with local residents and mall workers as a place to have lunch or a snooze. It is fenced around the entire perimeter, so protected from windblown bottles.
Repensar MZ performed a bottle audit in this garden on 9th April 2021, collecting and recording 254 PET bottles. So on that date, after the audit, the garden was completely empty of discarded bottles. You can see the data from this first audit here.
Repensar MZ performed a second audit in exactly the same location on 14th May (35 days later). Again, they swept the entire area to collect and scan all discarded bottles. You can see data from this audit (in this case, two separate teams) here and here.
In this second audit, 121 bottles were collected. This means that the garden accumulated an average of roughly 2.5 discarded bottles per day during that period. Dividing this by the surface area gives us:
1.75 bottles/ha/day …which is equivalent to… 175 bottles/km2/day
If we extrapolate this to the whole of Maputo City administrative area (347.69 km2) we can estimate 60,707 bottles being discarded per day.
If we expand to the greater Maputo conurbation (which includes the administratively separate population centres of Matola and Marracuene, a combined area of 1,387 km2) we could estimate 242,174 bottles per day.
Tracking BPH variation across different areas
The table below summarizes audits performed in April and May, across three locations in the greater Maputo Urban Area.
Room for improvement
We know this methodology has several rough edges. The BPH values in the table above must be taken with a pinch of salt. We welcome suggestions in the comments below.
Type of Area
Given the significant discrepancy between areas shown in the table above, a more accurate estimate would break the total urban area into types, each with its own BPH factor, for instance:
- Use type: Residential, Commercial, Leisure
- Urbanization level: High, Medium, Low, Rural
However even the above breakdown gives a possible twelve types. This creates a significant resource challenge in terms of operational monitoring (each monitoring requires between 3–5 site visits or audits).
Covered (inside) areas
Most people do not deposit rubbish in their own residence or place of work. In other words, areas covered with a roof. We cannot realistically extrapolate from a BPH for an open, public area to that part of a city which is covered. So we need to find a way of estimating the size of a non-covered area of a city.
‘Floated in’ bottles
In some areas, it is obvious when doing ‘bottle count’ audits (that is, the audits after the initial cleanup) that some of the new bottles detected are not recently discarded. We can assume this from their decayed physical condition. It might seem that these should be eliminated from the count, as they were not newly released to the environment during the time period in question.
However, we reckon it does make sense to include them, as the area being monitored is a representative sample of a wider area. If there are a lot of ‘old’ bottles in our sample area, we can reasonably assume that there will also be a lot in the wider area. Some of these have ‘floated in’ to our sample area. If there are few ‘old’ or ‘floated in’ bottles in our sample area, then we can reasonably assume that the wider area is cleaner of this older waste.
The ‘embarrassment effect’ of community collections
In one of the study locations, the audit team found that the bottle count decreased in the second visit, and continued to decrease in a third and fourth visit (not shown in the table above). This was attributed to an ‘embarrassment’ effect on the local authorities, who took note of Zero Waste activists prominently collecting waste from their area and put in some phone calls to the Municipal Solid Waste (MSW) Headquarters!
Next steps
We’ve planned several follow up actions for Q3 2021, to improve our data model.
Monitoring plastic by weight
In Wastebase, we can associate the exact weight of an empty bottle with the corresponding product barcode. Based on that, Wastebase can calculate GPH (grams per hectare). This is probably a more useful metric than the number of bottles, given that bottle size and weight vary (a small water bottle can weigh as little as 10 grams, whereas a 5 litre cooking oil bottle can be nearly 100 grams).
It is not that easy to get an accurate measurement of the weight of a bottle. Commercially available kitchen scales might have an accuracy of +/- 1 gram — which is a lot if you are weighing a 10 gram bottle. Better is to get a large number of identical bottles together, measure the total weight, and take the average. However one cannot easily put even five empty plastic bottles on a scale. More about this in a future article!
Informal Waste Collectors
In one of the locations, the team found three waste collection microenterprises active near the survey area. The team intends to compare another urban area with a similar commercial/residential usage mix, and similar level of affluence, but without such microenterprises, to get an idea of how much difference these collectors make to the overall level of plastic waste.
Conclusions
We’ve got a long way to go before we can confidently scale up data from local collections into regional or national representations of how much plastic waste exists in the environment. But we can already see how the Wastebase platform can amplify data from these local collections to give us more meaningful visualisation and help us understand both the discrepancies and opportunities for monitoring behaviour in different areas.
If we can better understand how, why and where people discard plastic waste, we can use data in a more targeted way to influence change. By working out a robust data model, we can also represent the true scale of the problem to everyone involved in making, distributing and consuming single-use plastic bottles.
References / Further Reading
World Bank 2017, Greater Maputo : urban poverty and inclusive growth (basis for the per-location income level classifications)
