This data set comprises outcomes of tests and observations conducted to establish the viability of a low-cost method developed to determine the plastic resins used to produce plastic food and beverage packaging. The data were gathered with two aims. First, we aimed to establish whether the outcomes of simple tests and observations were consistent when made by different observers, and secondly we aimed to establish if the observations and tests resulted in the correct identification of resins. Samples comprised rigid packaging items (polyethylene terephthalate, high density polyethylene or polypropylene) and flexible items (polypropylene or low density polyethylene). Data include discrete outcomes of individual observations and tests, plastic resin identification outcomes based on observations and tests, and indication of which outcomes led to correct identification of the resin to hand.According to WHO/UNICEF, whilst 91.8% of urban households in Sub-Saharan Africa (SSA) had access to piped or protected groundwater sources in 2015, only 46.2% had safely managed water available when needed. Vendors provide a key role in supplying urban off-grid populations, with consumption of bottled or bagged water (sachets, water sold in 500ml plastic bags) growing in SSA. Whilst several studies show bottles and bags are usually free from faecal contamination, given that many off-grid urban populations lack solid waste disposal services, when people drink such water, there can be problems disposing of the plastic bags and bottles afterwards. This project aims to deliver evidence on the different ways that people sell water to off-grid populations and what this means for plastic waste management. We plan to do this in Ghana, where most urban household now drink bagged water, and by way of contrast, Kenya, where the government has banned plastic bags. In this way, we want to widen access to safe water and waste management services among urban off-grid populations, by supporting water-sellers and waste collectors to fill the gaps in municipal services. Both countries (and many others elsewhere) already have nationwide household surveys that collect data on the food and goods people consume and the services they have. However, as yet, these surveys have not been connected to the problem of waste management. We plan to visit marketplaces, buying foods and then recording packaging and organic waste. By combining this information with the household survey data, we can work out how much domestic waste like plastics gets collected and how much is discarded or burned, ultimately entering the atmosphere or oceans. In Ghana, we will also survey informal waste collectors in urban Greater Accra. We want to find out how much these small businesses support waste collection and recycling across this urban region (particularly plastic from bagged water), so we can help government identify gaps in waste collection coverage. We also believe highlighting the important role of small waste collectors could lead to greater business support for such collectors. We will also evaluate whether community education campaigns to encourage domestic waste recycling reduce the amount of waste and plastic observed in the local environment. Such campaigns are currently pursued by several local charities with support from the Plastic Waste Management Project. In Kenya, where water is usually sold in jerrycans rather than bagged, the jerrycan water often gets contaminated. We plan to find out whether this jerrycan water is safer under an arrangement known as delegated management. This involves a water utility passing on management of the piped network to a local business in slum areas, so as to reduce vandalism of pipes and bring water closer to slum-dwellers. We will compare water quality in areas with and without this arrangement to see if it makes the water sold safer. We also plan to bring water-sellers and consumers together to find and test ways of reducing contamination of water between a jerry-can being filled and water being drunk at home. Rather than imposing a solution, we want to work together with vendors and consumers on this issue, but there are for example containers designed to keep water cleaner that we could explore. Through these activities, we thus plan to develop evidence on different strategies for water-sellers to deliver safer water to people lacking piped connections, whilst managing plastic waste at the same time. In Ghana, this involves trying to increase recycling and waste collection for bagged water, which is relatively safe. In Kenya, this involves trying to reduce contamination of water sold in reusable jerrycans. Alongside our household survey evidence on how domestic waste is managed in slums, this should help governments plan waste and water services in poorer areas of Africa's expanding cities.

Data were collected via observations and tests carried out on a test set of plastic packaging items used for food and beverage products. Samples of plastic packaging comprised rigid (e.g. bottles, trays) and flexible (e.g. film, wrappers) items. Rigid packaging items comprised PET (polyethylene terephthalate), HDPE (high density polyethylene), or PP (polypropylene); flexible items comprised PP or LDPE (low density polyethylene). Samples were analysed by teams in Ghana (Greater Accra; 4 individuals) and Kenya (Kisumu; five individuals); all observers undertook preparatory training and instruction in conducting tests and observations prior to commencement of analyses. Observations and tests recorded attributes and qualities of samples appertaining to (1) format/texture, (2) roughness of an edge when cut, (3) appearance of an edge when cut, (4) density relative to water, (5) sound when rubbed, and (6) resistance to stretching. Sequences of tests and observations differed for items of rigid and flexile formats. Outcomes of tests and observations were used to identify the resin of each item by reference to their known qualities and attributes, structured within a framework in which sequences of tests and observations led in principle to discrete identification outcomes. Identification outcomes were compared with the resins as indicated by markings on packaging items or as determined through independent analysis.