Battery Recycling Is a Far Bigger Source of Lead Exposure Than We Thought

Policymakers are increasingly interested in tackling global lead exposure. One critical obstacle is that we don’t have a clear sense of what the most common sources of exposure are. The single largest use of lead is in car batteries, and unsafe recycling of these batteries is harmful, but we don’t know exactly how harmful. Prior estimates suggested battery recycling is only a small part of the problem. However, a new CGD working paper shows that it could account for a third of all lead exposure in low- and lower-middle-income countries. This estimate comes with a lot of uncertainty, with a 95 percent confidence interval spanning from 20 percent to 90 percent of the burden, and an 80 percent confidence interval from 26 percent to 71 percent of the burden. But the bottom line is clear: lead exposure from battery recycling is not insignificant.

Lead-acid batteries still start most vehicles worldwide, and the lead inside them is valuable enough to encourage widespread recycling. In many low- and middle-income countries the smelting and breaking of batteries can heavily contaminate the surrounding environment. Growing evidence shows that this contamination has serious consequences for the health of communities living nearby.

We adapt a model originally developed by Ericson et al. (2017). Their model focuses on people with high levels of lead exposure in the 300 metres surrounding each recycling site. Soil lead in that area is usually orders of magnitude above safe levels, with children showing symptoms of severe lead poisoning. But several new quasi-experimental papers have consistently linked lower-level lead exposure within two to ten kilometres of a recycling site to negative impacts on children's health and education. That might seem like a large distance, but lead has been shown to have been deposited atmospherically across entire oceans and thousands of miles. So we extend the model to include these lower levels of pollution that spread much farther away from the sites over much larger populations.

The result? This mass population low-level exposure over vast areas ends up driving most of the harm, rather than the localised high-levels of exposure in specific hotspots.

Figure 1. Our updated model increases the outer range of pollution to 5,000 metres

Image

How our model works

We estimate exposure from unsafe recycling as a function of:

1. The number of polluted sites: We estimate the number of sites polluted by recycling in each country using two methods: one based on the number of vehicles (and therefore used car batteries), and another based on extrapolations from an exhaustive census count of polluted sites from Ghana. Since both methods have weaknesses, we take a midpoint between the two.
2. How exposed the affected population is: We use a global database of soil pollution around recycling sites to model distance decay of lead pollution around each site. We then convert from soil lead measures to blood lead levels using a biokinetic model (the All Ages Lead Model).
3. How many people are exposed: Using gridded population data, we count the average number of people living within five kilometres of each site in our database, and extrapolate to unmeasured countries based on population density.

Lead pollution decays with distance, but remains at low levels over a wide distance

Image

Note: This figure presents data fitted to 5,172 measurements from the Pure Earth toxic site identification program. The “0–5,000 m average” is the mean fitted soil lead level computed by averaging predicted values over a disk of radius five kilometres. The average is 217 mg/kg, above the 200 mg/kg US EPA soil lead concern threshold.

Our result: informal battery recycling causes about a third of lead poisoning

Using the latest estimates for global lead poisoning, our model suggests that unsafe battery recycling accounts for 33 percent of the lead exposure from all sources. This is up from the previous estimate of about 15 percent. Our estimate comes with substantial uncertainty across multiple different inputs. Monte Carlo simulations allowing for this uncertainty show a wide range of possible shares, though all larger than previous estimates.

The share of lead exposure from unsafe battery recycling

Image

Note: This figure shows the share of all lead exposure attributable to used lead-acid battery recycling. The figure shows the distribution of estimates across 50,000 simulations with varying independent draws for each of the four key parameters. Vertical dashed lines indicate the 95% confidence interval.

A 33 percent share of lead poisoning in low- and lower-middle-income countries is equivalent to eight million disability-adjusted life years, 800,000 annual deaths, or 180 million IQ points lost.

So what should policymakers do?

At present there is no proven blueprint for making battery recycling safe in low-income countries. Approaches that have worked in upper-middle-income countries like Brazil and China, such as “extended producer responsibility” rules with a mandatory buy-back for sellers, haven’t yet been tested in poorer settings with weaker enforcement capacity. Enforcement and closure of unsafe facilities may only be sustainable when there are safe alternatives. In Nigeria there is only one safe recycler. Ultimately we don’t really have all the answers, but our findings do demonstrate that the scale of the problem is bigger than previously thought, and big enough that it demands new ideas and new solutions.