Welcome to ULAB Plume

This website models plumes (air emissions from chimneys, a.k.a. stacks) from lead-acid battery (ULAB) smelters using the US EPA's regulatory atmospheric dispersion model AERMOD.

Lead exposure is estimated to cause 3.5 million deaths per year. Research suggests that lead pollution from ULAB smelters in low- and middle-income countries (LMICs) could be the dominant driver of global lead exposure (Crawfurd et al. 2026), but this is currently highly uncertain — significantly hindering the investment and political will needed to drive solutions. Key uncertainties include the pathways by which surrounding communities are exposed to lead pollution from smelters, the size of the exposed population, and the resulting cumulative health burden.

Our vision for ULAB Plume is to create an easy-to-use platform to model the distribution of lead air emissions around lead smelters in LMICs. Mapping lead air emissions has two key use cases that should be considered separately due to the achievable level of confidence:

• Assessing the compliance of lead smelters with air emission standards: a tool for evaluating regulatory compliance.
• Modelling the lead exposure pathways and the resulting health burden experienced by communities within 50 km of lead smelters: a risk characterisation tool.

See more on ULAB Plume use cases.

Air emissions regulatory screening tool

Many LMICs already have regulations for maximum lead concentrations from industrial emissions, see our Lead Air Legislation in LMICs summary. However, given the standards of lead smelters in LMICs and the lack of air monitoring infrastructure, we expect these regulatory limits are typically exceeded. The open-source code that ULAB Plume uses, AERMOD, was adopted as a regulatory tool in the US in 2005, used to assess compliance of industries with air emission standards. AERMOD spatially maps the air concentration and deposition of pollutants typically over a 50 km radius from industrial pollution sources. AERMOD has been proven as a regulatory enforcement tool for more than 20 years, and has even been used in high-profile court cases, successfully litigating lead smelters in the US for public health damages. Exide Technologies' lead-acid battery recycling smelter in Vernon, California closed in March 2015 to avoid federal criminal prosecution; the resulting cleanup of surrounding residential neighborhoods has cost over $270 million to date, the largest lead-contamination remediation in California's history (see AERMOD case studies).

AERMOD works — we don't need to prove that. Given the same input data, ULAB Plume already produces identical results to AERMOD (it uses the same code) — you can check for yourself, see method: validation. But like any model, the accuracy of the results depends on how well the input data reflects reality. The key hindrance currently preventing regulatory-grade results from ULAB Plume is the lack of reliable input data characterising emissions from typical lead smelters in LMICs, summarised in the methodology page.

The key data gaps include:

• Rate of lead emission from stacks: grams of lead per second.
• Rate of fugitive lead emissions: uncontrolled lead released from non-stack sources, e.g. battery breaking and handling, furnace charging, and leaks in exhaust gas routing.
• Particle size distribution (PSD) of lead emissions, which dictates how long the lead particles stay in the air and how far they travel. PSD is expected to change depending on furnace temperature, emission control technologies, stack characteristics and exhaust gas routing.
• Stack characteristics: height, diameter, temperature of gases as they exit the stack (which dictates how far the plume rises).
• Local weather / climate data: we're currently relying on NASA satellite data, and we're not yet sure how that will impact the accuracy of the results compared to using local weather stations.

We're working to address these data gaps and develop a feasible protocol for how to produce regulatory-grade results from ULAB Plume / AERMOD in LMICs (see timeline page). We hope to address these data gaps and verify our protocol with case study field tests, feasible by 2028. Until then, ULAB Plume's results are currently illustrative and not suitable for regulatory screening.

Modelling lead exposure pathways and characterising health risks

ULAB Plume also maps lead exposure pathways, extrapolating from AERMOD outputs to estimate the amount of lead that populations ingest via inhaling contaminated air, consuming contaminated food crops, and indirectly ingesting contaminated soil and dust. Modelling the health burden from smelters introduces significant uncertainties and is not necessary for regulating air emissions. However, we are optimistic that communicating the expected magnitude of the resulting health impacts from substandard lead smelters can help to inspire the political will and financial support needed to support interventions and regulatory enforcement.

Using biokinetic models, such as the IEUBK model, also from the US EPA, if we know the concentration of lead in different environmental media, we can predict the amount of lead a child or adult ingests and model the expected blood lead level (BLL) impact and cumulative health burden across populations. The US EPA label their biokinetic models as tools for the characterisation of health risks. By combining ULAB Plume and biokinetic models, we can map and characterise the health risks from pollution released from lead smelters. The output from AERMOD is a spatial map of lead air concentrations (µg/m³) and the rate of lead deposited on the ground (mg/m²/yr) within a 50 km radius around lead smelters. We extrapolate these outputs to:

• Air inhalation: air concentration (µg/m³) taken directly from AERMOD.
• Soil & dust ingestion: extrapolated from the rate of lead deposition (mg/m²/yr); the lead deposited on the ground is assumed not to be impacted by weathering and to accumulate in soil.
• Food crop ingestion: extrapolated from the rate of lead deposition (mg/m²/yr); a defined % of the lead deposited on the plant surface is assumed to be retained by the crop (crop transfer factor).

See our methodology for more information. Extrapolating the AERMOD outputs introduces uncertainties. The BLL impacts from air inhalation have the highest level of confidence (directly using the lead air concentrations from AERMOD). Food crop contamination has the highest uncertainty; however, we expect widespread low-level food crop contamination from long-range air emissions (>5 km) to represent a substantial health burden that has been overlooked in prior research. Investigating food crop contamination from lead smelters is exactly what inspired ULAB Plume — see my presentation at CGD's lead research conference (publications currently under review). We aim to develop ULAB Plume so that it can be used to highlight areas at risk of food crop contamination and evaluate the resulting health burden. However, the rate of transfer of lead air emissions to edible fractions of crops is currently highly uncertain, but we're planning to address that with our upcoming research.

Support ULAB Plume

For ULAB Plume to achieve its full potential as a tool for assessing compliance with air emissions regulations and characterising lead exposure health risks in the surrounding communities, we have a number of data gaps to address, and we are looking for academic collaborators and financial support. We are hoping to conduct a series of thorough environmental assessments, using a combination of isotopic analysis and air sampling techniques, to address these data gaps and validate ULAB Plume. The more data we can gather, the more reliable ULAB Plume's results will be.

We are particularly keen to gain academic collaborators with expertise in AERMOD and atmospheric pollution dispersion modelling, geochemists, and anyone with expertise in foliar lead uptake in crops. We are also looking for financial support to increase the scale of our field work and data collection.

Any review or feedback is also very welcome!

Please do reach out — happy pluming!

Chris Kinally
Senior Research Scientist
Pure Earth
chrisk@pureearth.org