Arsenic hazard on a global scale

I have been following the harrowing story of how arsenic gets into domestic water supplies for 20 years (see: Earth-logs Geohazards for 2002; 2003; 2004; 2005; 2006; 2008; 2009; 2011; 2013; 2017). In my opinion, it is the greatest natural hazard in terms of the numbers at risk of poisoning. In 2006 I wrote about the emergence in Bangladesh of arsenic poisoning on a huge scale during the mid 1990s for a now defunct Open University course. If people depend for drinking water on groundwater from tube wells driven into alluvium they would not know of the risk, unless the water is rigorously analysed for levels of As greater than 10 micrograms per litre (μg l-1), the WHO recommended maximum. The sad fact is that the affected population were advised to switch from surface water supplies, which carry a high risk of biological infection, to well water. That is because during downward percolation from the surface oxidation destroys bacteria and viruses as well as parasites. Opportunities provided by a massive UN-funded drilling programme and local well digging made the choice seemingly obvious. Most people came to prefer well water as gastro-intestinal infections and child mortality fell rapidly.

Arsenic adds no taste, which is why it was once the ‘poison of choice’. How it gets into groundwater is difficult to judge, unless wells are downflow of areas riddled with metal mines. Years of research uncovered an unsuspected mechanism. The most common colorant of mineral grains, and thus sedimentary rocks, is brownish iron hydroxide (goethite), and that is able to adsorb a range of dissolved elements, including arsenic. One would think, therefore, that groundwater should be made safe by such a natural ‘filtering’ process: indeed goethite can be used in decontamination. The problem is that iron hydroxide, which contains Fe-3, is only stable in water with a high capacity for oxidation. Under reducing conditions it breaks down to soluble Fe-2 and water, thereby releasing to solution any other element that it has adsorbed. In alluvium, beds containing organic matter are prone to this ‘reductive dissolution’ of goethite. If weathering upstream has released even seemingly insignificant amounts of arsenic during the build up of alluvium, there is a potential life-threatening problem as arsenic builds up in the goethite coating of sedimentary grains to become ‘locked in’, with the potential to be released in high concentrations if subsurface chemical conditions change. The colour of the alluvial sediments penetrated by wells is a clue. If they are reddish brown, groundwater is safe, if they are greyish and goethite-free then, ‘beware’. But it is rare to examine ‘cuttings’ from a drill site aimed at groundwater, unlike those aimed at ores or oil

Since the tragedy of Bangladesh, which resulted after 5 years or so in obvious signs of arsenicosis – dark wart-line keratoses on hands and feet or black blotches on facial and torso skin – several alluvial basins in large river systems have had their well water tested. But by no means all such basins have been screened in this way, and there are many less-obvious signs of arsenic poisoning. After long exposure to the lower range of dangerous arsenic levels a variety of cancers develop in known areas of arsenic risk. There are also high levels of endemic respiratory problems, cardiovascular disease, reduced intellectual development in children and even diabetes. Geochemical monitoring of all populated and farmed river systems is a huge task that is far beyond the resources of many countries through which they run. One approach to ‘screening’ for hazard or safety is to use geological, hydrological, soil, climate and topographic data. Those from known arsenic-prone basins and those where its levels are shown to be consistently below the 10 μg l-1 danger threshold help to develop a predictive model (Podgorski, J. & Berg, M. 2020. Global threat of arsenic in groundwater. Science, v. 368, p. 845-850; DOI: 10.1126/science.aba1510).

Modelled global probability of arsenic concentration in groundwater exceeding 10 μg l-1. Click to display a larger map in a separate browser tab. (credit: Podgorski & Berg; Fig 2A, with enhanced colour)

Rather than trying to model the full range of arsenic concentrations, Joel Podgorski and Michael Berg of the Swiss Federal Institute of Aquatic Science and Technology focussed on assessing probabilities that arsenic in well water exceeds the WHO recommended maximum safe level of 10 μg l-1. Their global map highlights areas of concern for environmental health. Thankfully, huge (blue) areas are suggested to present low risk, the pale, yellow, orange and red patches signifying areas of increasing concern. No populated continent is hazard-free. What is very clear is that Asia presents the greatest worries. Most of the Asian ‘hot zones’ are spatially close to large mountain ranges and plateaus. In the case of the Indus and Ganges-Brahmaputra plains the sources for excessive arsenic in groundwater implicated by previous geochemical investigations lie in the Himalaya. The factor common to all major hot spots seems to be rapid transport of huge amounts of sediment released by weathering from areas of high topographic relief, rather than local large-scale mining operations. There are hazardous areas related to historic and active mining, such as the Andes of Bolivia, Peru and Chile and the western USA, but they are tiny by comparison with the dominance of natural arsenic mobilisation.

Despite the WHO recommended maximum of 10 μg l-1 of arsenic, many countries base their policy on levels that are five times higher, largely because of the difficulty of analysing for the lower concentration without expensive analytical facilities. Field analyses are often done using simple semi-quantitative tests based on paper impregnated with reagents that show a colour range for different concentrations, which are unreliable for those lower than 100 μg l-1. Thankfully, despite the many risky areas, most of them have population densities less than 1 per km2.

If you are interested in the geological details of the arsenic problems of Bangladesh, the course text that I produced for the Open University (Drury, S. 2006. Water and well-being: arsenic in Bangladesh. The Open University: Milton Keynes, UK. ISBN 0-7492-1435-X), the course itself (S250 Science in Context) was withdrawn some years ago.  It may be possible to arrange a PDF for private use.

See also: Zheng, Y. 2020. Global solutions to a silent poison. Science, v. 368, p. 818-819; DOI: 10.1126/science.abb9746

Water-borne arsenic back in the news

In the 1980s grim news began to emerge from the Indian State of West Bengal and a decade later from neighbouring Bangladesh. Villagers from the low-lying delta plains of the Ganges and Brahmaputra river systems at the head of the Bay of Bengal began to present at clinics with disfiguring skin lesions or keratoses on hands and feet, loss of feeling in fingers and toes and dark skin patches on their torsos. The latter were colloquially known as ‘black rain’. The victims were often stigmatised, as their neighbours believed they were suffering from leprosy. These symptoms were followed a few years later by increased incidences of lung, liver, kidney and bladder cancers. The first medical practitioner to recognise these typical signs of chronic arsenic poisoning in 1983, Dr Depankar Chakraborti of Kolkata, was branded as a ‘panic monger’ by local authorities. His warnings, backed by evidence published by the World Health Organisation (WHO) in 1988 that there was a connection with high arsenic levels in West Bengal drinking water supplies from new tubewells, went largely unheeded for a decade. Tragically, as it turned out, thousands of tubewells had been sunk in the Bengali delta plains from the 1970s onwards, aimed at reducing the risk of disease from pathogens in the previously used surface water from ponds and streams. After a conference on the perceived problem, organized in Kolkata by Dr Chakraborti in 1995, the WHO declared the situation in Bangladesh to be a ‘Major Public Health Issue’, and the world’s press took up the story. Clearly, millions of Bengali villagers were at risk or were already suffering from chronic arsenic poisoning. By the late 1990s thousands of samples of tubewell waters from the delta plains had been analysed, many of which revealed arsenic levels far above the 10 μg l-1 safe threshold. In 2002, 400 Bangladeshi victims sued the British Geological Survey (BGS) for negligence. The BGS had analysed 150 water samples from the Bangladesh delta plains in 1992 and had not reported any risks, but arsenic was not among the elements being analysed. The civil action eventually failed.

Skin lesions or keratoses that are symptomatic of chronic arsenic poisoning

Almost two decades after the arsenic scandal on the eastern side of the subcontinent well-water analyses showing high arsenic values have been published from the Indus plains of Pakistan (Podorski, J.E. et al. 2017. Extensive arsenic contamination in high-pH unconfined aquifers in the Indus Valley. Science Advances, v. 3,; doi:10.1126/wsciadv.1700935). The Indus catchment having a similar Himalayan source and being at a similar latitude it has long been considered to be at potential risk from arsenic derived from its thick alluvial sediments. The Swiss-Pakistani-Chinese team have produced geochemical data from 1200 tubewells throughout the catchment within Pakistan. A swath from Lahore to Karachi, with the country’s greatest population density, is at high risk of water with arsenic concentrations above the WHO guideline safe concentration, suggesting some 50 to 60 million people being subject to its hazard.

Although the geological setting is similar to that in the Bengal plains, a different natural chemical process causes the high concentrations ultimately from the iron hydroxide veneer on sediment grains which selectively absorbs several trace elements, including arsenic, from river water. In Bangladesh arsenic is released from sediments as a result of highly reducing conditions due to organic matter buried in some layers of alluvium, by a process known as reductive dissolution – when insoluble ferric iron (Fe3+) hydroxide (goethite) is exposed to a ready supply of electrons the iron is reduced to the soluble ferrous (Fe2+) form and the mineral breaks down to release its absorbed trace elements. Most of the alluvium in the Indus plain contains little organic carbon, so another mechanism is implicated. The high arsenic levels correlate with high pH in the groundwater and therefore seem most likely to be released from goethite grain coatings by alkaline water. That, in turn, is often a product of high evaporation and salinisation from the massive irrigation using groundwater in semi-arid southern Pakistan. The alkaline water then returns to the underlying groundwater in the highly permeable Indus alluvium; i.e. it is a consequence of irrigated agriculture rather than of a natural geochemical process as in more humid Bengal.

Whereas a remedy in Bangladesh and West Bengal has been to sink new tubewells into oxidising alluvial strata (red coloured rather than the reducing grey sediments)  that yield water with safe arsenic levels, the risky areas in Pakistan may need expensive use of absorbent filters on a large scale to remove the hazard. Because irrigation using groundwater is on such a large scale on the Indus plain there is also a definite risk of ingesting arsenic from crops produced there, principally rice but also unwashed leaf vegetables

See also:

http://www.bbc.co.uk/news/science-environment-41002005

http://www.sciencemag.org/news/2017/08/arsenic-drinking-water-threatens-60-million-pakistan

http://www.dawn.com/news/1353482/50-million-at-risk-of-arsenic-poisoning-in-pakistan?preview

https://www.dawn.com/news/1354023

Estimating arsenic risks in China

Two weeks after Earth pages featured arsenic in groundwater below the Mekong Delta another important paper has emerged about modelling risk of arsenic contamination throughout the People’s Republic of China (Rodriguez-Lado, L. et al. 2013. Groundwater arsenic contamination throughout China. Science, v. 341, p. 866-868). Scientists based in the Swiss Federal Institute of Aquatic Science and technology and the China Medical University follow up the results of geochemical testing of groundwater from almost 450 thousand wells in 12% of China’s counties; part of a nationwide aim to test millions of wells. That is a programme likely to last for decades, and their work seeks to develop a predictive model that might better focus such an enormous effort and help in other large regions where well sampling is not so advanced.

As well as the well-known release of arsenic-containing ions through the dissolution of iron oxy-hydroxides in aquifers that exhibit reducing conditions, aridity that causes surface evaporation can create alkaline conditions in groundwater that also desorbs arsenic from similar minerals. The early results from China suggested 16 environmental  factors available in digital map form, mainly geological, topographic and hydrogeochemical, that possibly encourage contamination; a clear indication of the sheer complexity of the problem.  Using GIS techniques these possible proxies were narrowed down to 8 that show significant correlation with arsenic levels above the WHO suggested maximum tolerable concentration of 10 micrograms per litre (10 parts per billion by volume). Geology (Holocene sediments are most likely sources), the texture of soils and their salinity, the potential wetness of soils predicted from topography and the density of surface streams carrying arsenic correlate positively with high well-water contamination, whereas slope, distance from streams and gravity (a measure of depth of sedimentary basins) show a negative correlation. These parameters form the basis for the predictive model and more than 2500 new arsenic measurements were used to validate the results of the analysis.

Estimated probability of arsenic in Chinese groundwater above the WHO acceptable maximum concentration (Credit:Rodriguez-Lado, et al. 2013)
Estimated probability of arsenic in Chinese groundwater above the WHO acceptable maximum concentration (Credit:Rodriguez-Lado, et al. 2013)

The results graphically highlight possible high risk areas, mainly in the northern Chinese provinces that are partly confirmed by the validation. Using estimated variations in population density across the country the team discovered that as many as 19.6 million people may be affected by consumption of arsenic contaminated water. In fact if groundwater is used for irrigation, arsenic may also be ingested with locally grown food. It seems that the vast majority of Chinese people live outside the areas of risk, so that mitigating risk is likely to be more manageable that it is in Bangladesh and West Bengal.

As well as being an important input to environmental health management in the PRC the approach is appropriate for other large areas where direct water monitoring is less organised, such as Mongolia, Kazakhstan and Kyrgyzstan in central Asia, and in the arid regions of South America.

Yet another risk of arsenic exposure

The most widely feared risk of poisoning through natural causes, which grossly disfigures and kills through a range of cancers, is from chronic exposure to arsenic in drinking water. Tragically, the risk is highest from what has traditionally been considered safest source, groundwater. That was the gruesome lesson of a massive transfer in Bangladesh from drinking surface water containing organic pathogens to reliance on well waters. The greatest mass poisoning in history was eventually traced to shallow aquifers in the Ganges-Brahmaputra plains that were rich in organic matter. Their reducing chemistry broke down iron hydroxide coatings on sedimentary grains. Since these minerals are among the most accommodating adsorbers of ions from the environment, including a variety of arsenic-bearing ions, their dissolution releases potential poisons from otherwise safe storage. In Bangladesh and neighbouring West Bengal in India it was found that deeper aquifers have oxidising chemistry and so the iron minerals not only hold ionic pollutants fast by adsorption but help to extract them from groundwater. Deep wells together with various kinds of treatment of shallow groundwater, some using the very iron minerals whose breakdown caused the pollution, are helping to mitigate the perilous situation for people of South Asia.

Skin lesions from arsenic poisoning in Bangladesh
Skin lesions (keratoses) from arsenic poisoning in Bangladesh (Photo credit: waterdotorg)

Much the same kind of arsenic pollution has subsequently been revealed in groundwaters of lowland Vietnam and Cambodia. Yet the turn there to deep groundwater has revealed a new twist. That too is yielding increasingly high arsenic concentrations, but for a different reason (Erband, L.E. et al. 2013. Release of arsenic to deep groundwater in the Mekong Delta, Vietnam, linked to pumping-induced land subsidence. Proceedings of the National Academy of Science, doi/10.1073/pnas.1300503110). Scientists from Stanford University, California analysed waters from around 900 wells in the Lower Mekong Delta and found several tracts with arsenic contents well above levels deemed safe by the WHO. Some, as could be anticipated from South Asian studies, were from shallow wells along the present course of the Mekong. However, in the delta area to the southwest of Ho Chi Minh City (formerly Saigon) is a large cluster from wells 150 to 450 m deep, totally unlike the situation in other areas of thick Pliocene to Recent river sedimentation.

Comparing the distribution of affected wells with precise estimates of the subsidence rates of the land surface from orbital interferometric radar surveys shows a close correlation of arsenic contamination with rates of subsidence. This suggests that groundwater pumping from deep aquifers is causing compaction at depth, in much the same way as in the environs of Venice. But is this somehow drawing in arsenic polluted water from higher levels? It seems not. So the pollution seems most likely to be an effect of pumping itself. The authors suggest that most of the subsidence is due to compaction of clay-rich sediments rather than the sandy aquifers, well known by engineers to resist compression. They explain the increasing arsenic concentrations by the introduction into the aquifers of water expelled from the clays, either containing arsenic ions in solution or carrying organic compounds that create the reducing conditions to break down iron hydroxide grain coatings and release ions adsorbed on their surfaces.

This presents another grim prospect for South Asian people forced to make the choice between drinking polluted surface water and enteric disease and increasingly exploited deep groundwaters that seem to be safe as well as in very high volumes. Let’s hope that arsenic monitoring can be maintained in the Ganges-Brahmaputra plains in the long term.

South Asian arsenic update

Skin lesions from arsenic poisoning in Bangladesh
The first signs of chronic arsenic poisoning: skin keratoses. Image by waterdotorg via Flickr

That groundwater in West Bengal, India was polluted with arsenic to such levels that symptoms of poisoning had become endemic was reported by Depankar Chakraborti in 1983, leading to his being branded a ‘panic monger’ by the Indian authorities. The news broke internationally in 1993 as the now infamous tragedy in neighbouring Bangladesh emerged. Means of mitigating the effects – lesions or keratoses and skin discoloration, and later increases in incidence of several forms of cancer – and ideas of how the pollution had occurred had to await proper geochemical analyses of well waters and logging of the mainly alluvial sediments from which water was being withdrawn; another 8 years went by. Reports of arsenicosis began to emerge from other areas of alluvial sediments in SE Asia, revealing by far the worst mass poisoning in history and the likelihood that the lives of millions would be blighted by what Bangladeshis dubbed ‘the Black  Rain’ from the resemblance of the characteristic skin lesions to drops of black water.

Thanks principally to the work of water engineer Peter Ravenscroft with other geochemists, the source of arsenic in groundwater was narrowed down to the effect of reducing conditions in grey, carbonaceous sandstones and peats on the mineral goethite, an iron oxy-hydroxide that forms the main colorant in oxidised sediments and whose loose structure normally encourages the mopping-up by surface adsorption of a wide spectrum of dissolved ions, including those of arsenic. Goethite readily breaks down under reducing conditions, and when that happens all the adsorbed material is released into solution. The upper parts of the alluvial and deltaic sediments in the lower reaches of the Ganges and Brahmaputra rivers contain abundant organic remains picked up when vegetation burgeoned during the Holocene, which mixed with goethite-coated sand grains derived from erosion in the Himalayan stretches of the rivers. Purely natural sedimentary and hydrogeological processes created the dreadful plight of villagers. The terrible irony was that before the 1980s there were no signs of arsenicosis, yet mortality, especially of under-fives, was very high due to water-borne pathogens in surface water supplies. Indian and Bangladeshi authorities and UN agencies waged a campaign to sink shallow wells for drinking water rather than relying on river and pond supplies. At first rural people resisted the change since they regarded water from wells as the ‘Devil’s water’, but as infant mortality began to fall, the resistance turned to rapid construction nationwide of wells, both public and private. A few years later came the ‘Black Rain’.

In the attempts to mitigate the arsenicosis plague, filters containing adsorptive materials, including goethite, were installed on pumps. However, the geochemists showed that in the deeper wells there were consistently low concentrations of arsenic in sediments that were brown-coloured due to prevailing oxidising conditions and the presence of goethite. Although arsenic was present in the sediments it was safely locked in the goethite coatings of sand grains. Steadily major public supplies were transferred to deep, high-yield wells. Alluvial and deltaic deposits are generally highly permeable, so it was feared that as the deeper wells were pumped arsenic-rich water from the reduced shallow sediments would replace the safe groundwater. Thankfully, it seems that is not likely to be a problem (Radloff, K.A. and 12 others 2011. Arsenic migration to deep groundwater in Bangladesh influenced by adsorption and water demand. Nature Geoscience, v. 4, p. 793-798). The study injected As-bearing groundwater into a deep aquifer and monitored its arsenic concentration over time, once in place. Within a day, the concentration of dissolved arsenic fell by 70% and by 5 days had fallen below recommended maximum levels for drinking water; a dramatic demonstration of the clean-up power of even minute films of goethite in sediments, for that seems the only explanation for the fall. The US-Bangladeshi team verified this by testing samples of the deeper sediments from drill cuttings. They mixed highly contaminated groundwater with the cuttings, to find that arsenic sorption over  about a week was extremely high (~40mg kg-1).

Water well in Bangladesh. From http://www.flickr.com/photos/waterdotorg/3696304044

Rather than just publishing their reassuring findings, the team input them to hydrogeological models of the Bengal Basin, varying hypothetical pumping rates to assess the changes in deep-groundwater chemistry over time due to downward migration of the highly polluted near-surface waters. Sure enough, the As-rich waters would end up in the deep aquifer eventually to overwhelm the sorptive capacity of its goethite content; arsenic would once again enter well supplies. However, if deep extraction was limited to drinking water by limiting pumping for irrigation to intermediate depths, safe limits could be sustained theoretically for a thousand years or more, except in some areas especially prone downward intrusion of polluted shallow groundwater. (Use of highly contaminated shallow groundwater for irrigation would simply transfer the problem to crops.) Clearly, monitoring is obligatory, but one hopes this important study does resolve the horrifying plight faced by so many people in catchments fed by Himalayan waters.