lable at ScienceDirect

Journal of Cleaner Production 64 (2014) 535e544

Contents lists avai

Journal of Cleaner Production

journal homepage: www.elsevier .com/locate/ jc lepro

Processing centres in artisanal gold mining

Marcello M. Veiga a,*, Gustavo Angeloci a, Michael Hitch a,Patricio Colon Velasquez-Lopez b

aNorman B. Keevil Institute of Mining, University of British Columbia, Vancouver, Canadab Technical University of Machala, Centre for Agricultural Research, Machala, Ecuador

a r t i c l e i n f o

Article history:Received 15 January 2013Received in revised form3 August 2013Accepted 4 August 2013Available online 19 August 2013

Keywords:Gold processingArtisanal miningMercuryCyanideProcessing centres

* Corresponding author.E-mail address: veiga@mining.ubc.ca (M.M. Veiga)

0959-6526/$ e see front matter � 2013 Published byhttp://dx.doi.org/10.1016/j.jclepro.2013.08.015

a b s t r a c t

The aim of this paper is to reflect on efforts made worldwide to construct processing centres whereartisanal miners can visit to have gold extracted from ore “free” or for a nominal fee. These centres useinefficient grinding and amalgamation processes to extract less than 30% of the gold to give back to theminers. As payment, miners typically leave the tailings (residues) at the centres that are processed bycyanidation to extract at least 90% of the residual gold. The cyanidation of contaminated tailings pro-duces mercuryecyanide complexes that are rarely recovered in the process of activated charcoal or zincprecipitation. As a result, tailings discharged into the local water streams carry mercury either as solublecyanide complexes or mercury droplets. There are a few processing centres, particularly in southernEcuador, which use responsible and cleaner gold extraction integrating mining and processing tech-niques, but the vast majority of the processing centres in developing countries are only creating morepollution. Processing centres from different countries were assessed: Nicaragua, Peru, Colombia,Indonesia and Ecuador and the technical procedures to process ores and extract gold are described. Themain environmental issues posed by poor practices are: (1) inappropriate cyanide management, (2)amalgamation of the whole ore, which increases the mercury losses with tailings, (3) usage of cyanide toextract residual gold from Hg-contaminated tailings, (4) dumping tailings with mercury, other heavymetals and cyanide into the local natural drainages, and (5) decomposition of amalgams without anycondenser or filter.

� 2013 Published by Elsevier Ltd.

1. Introduction

In the developing world, a combined 30 million individualsextract more than 30 different minerals using rudimentary tech-niques. Gold is the preferred mineral extracted by artisanal oper-ators due to its high unit value and its market price, which hasincreased almost sevenfold over the last decade. Between 10 and 15million people are directly involved in this activity, producing 300e400 tonnes of gold annually. This production, however, has causednumerous environmental problems, including high levels of riversiltation and mercury pollution (UNEP, 2008). For the purposes ofthis analysis, artisanal gold mining (AGM) is herein defined as“operations, independently of the size, that use rudimentarytechniques to prospect, mine and process gold ore” (Veiga, 1997).

Anthropogenic mercury emissions have been steadily increasingsince 1995. In 2010, it was estimated that annually, 1960 tonnes ofmercury were emitted into the air from all human activities (UNEP,

.

Elsevier Ltd.

2013). The AGM sector is responsible for 37% of these emissions.The increasing dependence onmercury by artisanal miners calls foraffirmative action to demonstrate non-mercury technologies(Hilson and Pardie, 2006) or at least methods capable of reducingmercury losses (Metcalf and Veiga, 2012). Unfortunately, there islittle will from either host governments or international aidagencies to invest in strategies which can address effectively theseharmful practices.

For four decades, AGM activity has intensified in virtually everydeveloping country. The environmental, social and health situation ofthe sector’s operators have deteriorated in the process. The maincause of this has been a lack of an effective and “democratic” distri-bution network capable of disseminating information to those whomost need it. Technical information about more efficient techniquestends to reside with those who hold the greatest financial and intel-lectual capital. The result has been the installation of well-capitalizedtoll treatment processing centres which have disproportionate in-fluence and coercive power over the poorer, simpler miners.

This paper describes the performance of processing centres inselected countries visited by the authors between 2003 and 2013under the auspices of the GEF/UNDP/UNIDO “Global Mercury

M.M. Veiga et al. / Journal of Cleaner Production 64 (2014) 535e544536

Project”, the U.S. Dept. of State project of “Reducing Mercury Useand Release in Andean Artisanal and Small-Scale Gold Mining” andthe UNIDO Colombia Mercury Project. All data generated from thecase studies are based on mass balances and chemical analyses oftailings and effluents as well as in-country interviews with minersand local authorities. Details of the analytical procedures are foundin the referred articles.

It is extremely difficult to obtain reliable, accurate informationabout the use of mercury and cyanide, gold production andmethods used to extract gold, largely because the AGM sector isusually informal and often considered illegal. The objective of thispaper is to outline and discuss some of the environmental problemscreated by the rampant growing of these processing centres, andalso to highlight an example of more desirable practice of somecentres in Ecuador which operate correctly, respecting miners andthe environment.

2. “Evolution” of the processing centres

Centralized processing centres for AGM first appeared in the1990s. The appeal of this model to the miners is that they do notneed to invest in expensive pieces of equipment to crush, grind,concentrate and amalgamate the gold; they simply take their oresto facilities where gold is extracted by specialized operators for afee. One early example of the establishment of processing centrescame in 1991, when the Government of Venezuela prohibited allamalgamation activities on board of dredges in the Caroni River(Veiga, 1997). The Government and private investors created threeprocessing centres on the shore of the river. These plants operatedto extract gold by amalgamation from gravity concentrates of al-luvial ore brought by the miners from their barges. The idea waspraised by many experts and during the existence of the centres,substantial reduction of mercury discharges into the river wasobserved. The centres were closed in 2002 by the VenezuelanGovernment when two new hydroelectric reservoirs were built. Allminers were forcibly removed from the river as they were blamedfor river siltation affecting the electricity generator turbines. TheCaroni River has less than 0.05% of particles smaller than 0.09 mmand these fine particles settle downwithin a distance of 140m fromthe source (barges). The allegations of turbine damage were

Fig. 1. Venezuelan miner amalgamating the whole ore usi

absolutely untrue and just a political manoeuvre of the local Gov-ernment to satisfy local environmentalists who had conflict withthe artisanal miners (Veiga, 1996).

The Shamva Centre in Zimbabwe, built in 1989 to process oresfrom 40 gold mines in the region, was another example of organizedprocessing centres. The centre was created out of an initiative by theNational Miners Association of Zimbabwe with the UK Non-Government Organization ITDG (Intermediate Technology Develop-ment Group, nowadays known as Practical Action). The ShamvaCentre worked to discourage miners from using mercury in theiroperations and moved tailings disposal away from rivers. The centrewas active for many years but the main challenges the miners facedwere the distant location from the mines and the lack of an efficientprocess to extract fine gold (Simpson, 2007). For further examinationof the Shamva case see Hilson (2006). The author also highlights thecomplications in a processing centre implemented by UNIDO andGhanian Government in Japa. Initially, the miners used the centre todecompose amalgams in retorts. However, like in the Shamva Centre,a high demand for services forced miners to wait for extended pe-riods and the miners preferred to return to the practice of burningtheir amalgams in open pans in the jungle.

The Shamva Centre model proliferated worldwide. The merit ofthe idea for the miners was clear: the processing centres provide,for “free” or for a symbolic fee, the whole service to generate, inshort time, a bar of gold in the miners’ hands. For the governmentsand funders of projects, the main merit was to avoid discharges ofmercury into the rivers. As a payment, miners leave their tailings(waste) in the centres’ facilities for further processing to extractresidual gold. Many other processing centres have been establishedin Zimbabwe and in many other developing countries to providesimilar services. However, these new centres rarely use clean orefficient techniques in processing the miners’ ore. The processingcentres use a crude grind to liberate fine gold particles beforeprocessing, negatively affecting gold recovery. They use rudimen-tary methods to concentrate the gold, or they simply amalgamatethe whole ore using copper-amalgamating plates (Fig. 1), or theygrind the orewith mercury in small ball mills. The amalgamation ofthe whole ore is the main reason for large mercury losses to theenvironment. With these inefficient processing techniques, thegold recovery is usually below 30% (Cordy et al., 2011). The centres

ng a hammer mill and a copper-amalgamating plate.

Fig. 3. Agitated cyanidation tanks in Portovelo, Ecuador.

M.M. Veiga et al. / Journal of Cleaner Production 64 (2014) 535e544 537

provide, in a short time frame, the “easily extractable” gold from theores that the miners spent usually three weeks to mine. When theminers go home with their gold, the centres’ owners use bettertechniques (e.g. cyanidation) to extract the residual gold from thetailings left behind. The process is unfair to the miners and highlypolluting as the material being re-processed contains mercury. Ifminers want to take the tailings with them, the processing centrescharge the actual value of their processing work. In Portovelo,Ecuador, this costs US$20 per 50-kg bag of tailings.

Tailings left in processing centres are leachedwith cyanide. Often,cyanidation is conducted in percolation vats as witnessed in Brazil,China, Colombia, Ecuador, Indonesia, Mozambique, Nicaragua, Peru,Tanzania and Zimbabwe (Fig. 2). Without proper aeration, the cya-nidation process can last up to 50 days, and gold recovery can be aslow as 50%. This is especially the case when the gold particles arelocked into other mineral’s grain and not exposed to the cyanidesolution (Sousa et al., 2010). In other operations, processing centresuse agitated cyanidation tanks (Fig. 3) to leach gold from tailings as isthe case in Ecuador, Indonesia, Philippines, Venezuela andZimbabwe. The process can take 24e60 h of cyanidation, dependingon the intensity of agitation and aeration of the tanks. Once the goldis dissolved in the cyanide solution, two techniques are commonlyused to recover it: (1) adsorption of gold onto activated charcoal, or(2) gold precipitation with zinc.

The following section introduces case studies of some processingcentres to illustrate the technical problems that are leading to morepollution. The description of the cases starts with the most primitiveprocessing centres such as in Nicaragua, Peru, Colombia, andIndonesia and concludeswith the description of themost progressivecentres in Ecuador. The Ecuadorean case is discussed in more detail.This is not the ideal case, as mercury and cyanide are still beingdumped into the rivers, but the government and investors’ attitudesseem to be the most appropriate to reduce pollution.

3. Processing centres in selected developing countries

3.1. Nicaragua

Nicaragua has been experiencing a gold rush for the last decade.The “official” gold production in 2011 was 7.46 tonnes (240,000 oz)

Fig. 2. Percolation vats in Zimbabwe; leaching tailin

according to the Ministry of Energy and Mines. In 2012, the goldproduction increased to 260,000 oz. (Molina et al., 2013). Thegovernment representatives believe that the 1.24 tonnes (40,000oz) of gold in 2011 was produced by 20,000 to 30,000 artisanalminers, locally known as “güiriseros”, and the remainder of goldwas produced by two organized mining companies, B2Gold andHemco. Not all gold produced by AGM is through amalgamation

gs left by the miners in the processing centre.

M.M. Veiga et al. / Journal of Cleaner Production 64 (2014) 535e544538

since there are many alluvial artisanal miners in the country whodo not use mercury. In Rozario, miners use 400 dredges to pump theore to sluice boxes with riffles concentrating only coarse gold.

Those (miners), extracting hard rock ore, the processing cen-tres charge US$70 per day (a 24-h) for the use of one “rastra” thatgrinds two tonnes of ore per day, and US$22 to US$35 for rastraswith capacity of half-tonne to one tonne per day, respectively.Rastras are grinding circuits, in which four rocks are dragged incircular movements by an electric motor over a bottom of rocks(Fig. 4). In 2012, 160 rastras were in operation in the countryreleasing likely four to five tonnes of mercury/a. Commonly, acopper plate with mercury is positioned at the discharge of therastra to trap any gold leaving the grinding media. To eliminatethe oxidation spots on the copper plate, miners drip a solution of2 g/L of sodium cyanide in neutral pH on the plate. It is interestingto notice that similar procedure was also observed in China andZimbabwe. The smell of HCN (gas) is noticeable around the ras-tras; the gas is harmful in low levels of long-term exposure and itis lethal in high levels of exposure. The owners of the processingcentres sell the Hg-contaminated tailings to other companies orprocessing centres, which use cyanidation to extract residual goldfrom tailings, usually by percolation in vat leaching. Analyses ofrastra tailings indicated grades of 7.17 g Au/t and 10.3 g Hg/t. Withfeed grades around 9 g Au/t. This indicates that the gold recoveryis less than 20% and as a result, part of the mercury in the tailingswill be later converted into mercuryecyanide complexes in thetailings reprocessing step.

Analyses of total mercury in the air using the LUMEX RA915þaround the gold processing areas using rastra (10 m away from thesource) are usually twice theWorld Heath Organization guideline forpublic exposure of 1000 ng Hg/m3 (WHO, 2007). This does notconsider themercury levels inairwhenamalgamsareburnedwithoutcondensers (retorts) or filters; a common practice in the country. Themercury levels in sites where stamp mills and copper-amalgamatingplates operate are usually around 30,000 ng/m3 (10-m distance).

3.2. Peru

In Peru, the number of artisanal gold miners is not well-known.Estimates range from 80,000 (Peru Support Group, 2012) to150,000 (estimated by some Peruvian authorities in the Ministry of

Fig. 4. “Rastras” amalgamating the whole ore

Energy and Mines) of people directly involved. Production esti-mates are about 30 tonnes/a of gold (Fowks, 2012). Brooks et al.(2006) estimated that, even with the efforts of internationalagencies to reduce mercury emissions, between 20 and 40 tonnesof mercury are still released to the environment by artisanal goldminers annually countrywide. This amount appears to be signifi-cantly underestimated, given the continued expansion of the arti-sanal gold mining sector in the country, in particular in Madre deDios (Ashe, 2012).

In 2010, we estimated that there were 132 mining sites and10,000 artisanal miners working in the Department of Piura, Northof Peru. Government authorities as well as mining associationshave not quantified accurate numbers, as miners are dispersed overan area of some 158,000 ha. The Processing Centres in the Piuraregion are extremely primitive and usually controlled by an asso-ciation of miners. Mercury is introduced in “chanchas” or small ballmills or “quimbaletes” (i.e. large pieces of stone rocked back andforth, by one or more people to grind the ore with mercury) (Fig. 5)to amalgamate the whole ore without previous gravity concentra-tion. Mercury losses can be anywhere between 30% and 40%,depending on the type of ore used.

In San Sebastian, there were approximately 1300 people (in2012) directly involved in mining and 46 cyanidation processingplants, each one with three percolation vats with capacity of15 tonnes. No agitation tanks have been observed in the region asthere is no access to grid electricity and most power is provided bysmall generators.

About 80% of the miners in Piura are currently selling the ore tobe processed in Nazca or Arequipa, at the South of Peru (1100 kmfrom the mines) by large processing companies. Similarly, theytransport the ore to Ecuador to be processed in one of the manyprocessing centres in Portovelo. This is done for several reasons,including the lack of knowledge about techniques to processcomplex sulphide ores, the lack of resources to acquire processingequipment, the lack of water and power in the region and imme-diate payment for the gold content in the ores. The companiesbuying ores in the region analyze the material, giving an equal splitto the miners, and paying 40%e50% of the value of gold in the ore.These companies also buy Hg-contaminated tailings to processwith cyanide, transferring the risk of mercury mobility and bio-accumulation to other regions.

in Nicaragua e a very primitive process.

Fig. 5. Quimbalete used by a Peruvian miner in a processing centre (Photo: AJ Gunson).

M.M. Veiga et al. / Journal of Cleaner Production 64 (2014) 535e544 539

3.3. Colombia

Officially, according to the Secretary of Mines of Antioquia,Colombia had 200,000 artisanal gold miners producing 30 tonnesAu/a in 2011. In the Department of Antioquia alone, there are 17mining towns housing between 15,000 and 30,000 artisanal goldminers. The mercury pollution in Colombia can be as high as150 tonnes/a giving to the country the shameful first position as theworld’s largest mercury polluter per capita exclusively from arti-sanal gold mining (Veiga, 2010).

The processing centres in Antioquia use small ball mills, like the“chanchas”, but locally known as “cocos” (Fig. 6) to amalgamatewhole ore without any previous concentration. About 60 kge70 kgof ore is added to each “coco” together with about 100 ge120 g ofmercury. Tailings with grades up to 5000 mg/kg of Hg indicate thata substantial part of the mercury added to the “cocos” is pulverizedand lost. Due to the historical presence of guerrilla in rural areas, allprocessing work is conducted in urban areas where the level of Hgin air reaches up to 40,000 ng/m3 on the streets and close to1 million ng/m3 inside some processing centres (Cordy et al., 2011).Miners also sell the amalgams with up to 50% Hg to gold shops andthe buyers evaporate the mercury without any filters or con-densers, ultimately contributing to the urban air pollution.Mercury-contaminated tailings are leached in percolation vats forup to 30 days. Some centres precipitate gold from cyanide solutionwith zinc and evaporate zinc (then contaminated with mercury) inopen furnaces to obtain gold. Again, all these vapours are releasedin the urban environment.

3.4. Indonesia

Indonesia has experienced an unprecedented gold rush in thelast ten years. There are in excess of 800 artisanal gold mining sitesin the country encompassing 250,000 miners (Balifokus, 2013 e

personal comm.), but this number seems underestimated. Esti-mates from the Indonesian Government suggest that between 40and 60 tonnes/a of gold are produced by AGM. The official gold

production of Indonesia in 2012 was about 95 tonnes (Chalze,2013), but it is unclear whether the Indonesian authoritiesconsider the AGM production in their statistics. In 2007, Indonesianartisanal miners were releasing 130e160 tonnes/a of mercury tothe environment (Telmer and Veiga, 2008). Recent estimates, fromIsmawati et al. (2013), state that the amount of mercury releasedinto the environment in Indonesia can range from 280 to560 tonnes/a. The main reason for large amount of mercury releaseis the common practice of whole ore amalgamation in small ballmills (Fig. 7). It appears that similar practices of mining and pro-cessing are found all over Indonesia. Groups of five to seven minerscommonly extract 100 to 1000 tonnes/day of partially weatheredsaprolitic ore from shafts up to 20 m deep. The ore is delivered to aprocessing centre where it is ground in small ball mills (load of 20e40 kg of ore), and dosed with 300 ge500 g of mercury and rivercobbles or steel balls or rods. The grinding step lasts four to fivehours, and the fine product is discharged into plastic bowls wherethe amalgam is separated by panning and ultimately burned with atorch in open air. Mercury is pulverized and lost with tailings. Thereason for using the whole amalgamation system in many AGMsites in Indonesia is that the gold is usually very fine and not highlyrecoverable by gravity concentration. Unfortunately, in this case,gold recovery using amalgamation often does not exceed 30%(Veiga et al., 2009). However, the material is very suitable for cya-nidation. The gold obtained by amalgamation is produced muchfaster than by cyanidation; this represents an advantage for theimpoverished miners (Krisnayanti et al., 2012).

Most miners in Indonesia sell or rent cyanidation tanks (i.e.US$400/tank of three-tonne capacity, as observed in Lombok) toprocess the Hg-contaminated tailings. The process consists ofaeration in tanks lasting from two to three days. Most of the tanksdo not have agitation, just poor aeration. Activated charcoal is thenadded to the leaching tanks and after three absorption cycles, thecharcoal is recovered by screening. The cyanidation tailings, alongwith mercury in solution, are deposited in rudimentary ponds orsimply discharged into the rivers. High levels of mercury have beenanalyzed in fish from a river in North Sulawesi receiving Hg-

Fig. 6. “Cocos” (small ball mills) in Colombia to amalgamate the whole ore e mercury is pulverized and lost with tailings to the environment.

M.M. Veiga et al. / Journal of Cleaner Production 64 (2014) 535e544540

cyanide tailings from a processing centre (Castilhos et al., 2006).This is clearly a result of bioaccumulation of mercuryecyanidecomplexes released with tailings into the river. Most operators donot have knowledge about the activated charcoal elution processand hence they burn it in drums, releasing contaminants (includingHg) to the atmosphere. The ash is then amalgamated or simplymelted with borax.

Fig. 7. Processing Centre in Lombok, Indonesia where the whole ore is amalgamated using sHg-contaminated tailings are at the back.

3.5. Ecuador

In 2001, it was estimated that there were around 100,000 arti-sanal miners actively working in Ecuador, with most of themmining gold (Sandoval, 2001). Currently, this number can bedoubled as the price of the precious metals increased substantiallyin the last 10 years. In 2010, an inventory of the Ministry of Energy

mall ball mills (capacity of 20 kg each) and 400 g of mercury; cyanidation tanks for the

M.M. Veiga et al. / Journal of Cleaner Production 64 (2014) 535e544 541

Mines (MEM), counted 1349 artisanal mines in the country wherethe large majority was extracting gold (MRNNR, 2013). In El OroProvince alone therewere 541 artisanal and small-scale goldmines.The MEM recognizes that this number can be up to tenfold larger asless than 20% of the miners are formalized. The official gold pro-duction of Ecuador, according to the MEM, is around 10 tonnes/abut it is recognized that this must be around 25 tonnes/a in whichcase 75%e80% is generated by AGM.

In Portovelo, El Oro Province, southern Ecuador, miners bringtheir ores to one of the many processing centres in the region. AGMinvolves directly about 10,000 people in the region (Velasquezet al., 2010) and, according to assessment of the Association ofOwners of the Processing Plants (APROPLASMIN), about ninetonnes/a of gold was produced in 2012. In 2010, there were about110 processing centres in Portovelo-Zaruma region, but the federalGovernment closed some that were illegal and nowadays there are87 processing centres. In the nearby town of Ponce-Enriquez,200 km from Portovelo, it is estimated that an additional 48 pro-cessing centres opened in 2012. Portovelo-area processing centrestreat 3000 tonnes per day of ore, mostly from local mines, but alsosome ore from Ponce-Enriquez and Suyo, a mining site located inPiura, in northern Peru. The ore brought from other mine sites ismuch richer in arsenic, lead, zinc and other heavy metals than thelocal ore. Mercury-contaminated tailings are also frequentlybrought to Portovelo from different locations as the processingcentres have more resources, such as agitated cyanidation tanks.About 10% of the processing centres process less than 10 tonnes ofore per day, 60% processes 10e50 tonnes/day, 25% processes 50e100 tonnes/day and only 5% processes more than 100 tonnes of ore/day, and they usually have their own captive mining operations.

A large majority of the processing centres processes material inChilean mills that typically consist of two or three heavy cementwheels with steel rims connected to a 20 hp electric motor (Fig. 8).The wheels rotate over a 25 cm wide, 2-inch thick steel plate tocrush and grind the material below 0.2 mm. The ground material isthen concentrated in sluice boxes with wool carpets and the

Fig. 8. Chilean mill with sluice boxes covered with wool ca

concentrate, around 200 kg, is then either panned to reduce themass (i.e. 15e20 kg) before being amalgamated manually in a pan.Alternatively, the concentrate is directly leached with cyanide. Inthe first case, the amalgamation in a “batea” pan (known as “pla-ton”) of small amount of panned concentrate represents low mer-cury losses with tailings, i.e. only 1.4% of the mercury introduced inthe process. The amount of gold extracted by the sluice boxesfluctuates between 40 and 50% (Velasquez et al., 2010), however,after panning and amalgamation of the sluice concentrates, minerstake home only 20e30% of the gold from their ore. The processingcentres blend tailings from the gravity concentration circuit withthose from amalgamation and further leach them with cyanide incarbon-in-pulp tanks. Some miners use a mercury-free optionconsisting of cyanidation of the gravity concentrates in rented14 m3 agitated cyanidation tanks. This costs the miners approxi-mately US$200 per batch. Concentrates are then leached for fivedays where the gold in solution is precipitatedwith zinc (i.e. Merill-Crowe recovery process), and then burned in open-gas furnaces.This procedure, in spite of polluting the urban air with zinc dust,has reduced almost 560 kg/a of Hg being released to the atmo-sphere when amalgams were burned without filters. Mercury isstill used in “chanchas” (small ball mills) to amalgamate the rela-tively coarse gold from small amounts of high-grade ores. In 2013,the miners changed their low-polluting manual amalgamationprocedure and are no longer using the “platon” for amalgamation.As observed in Portovelo, many processing centres offer the minersto amalgamate the concentrates from sluice boxes in the “chan-chas”. As discussed above, in this case, more mercury is pulverizedand lost with tailings during the grinding. This “new” trendmust beincreasing the mercury losses.

All Hg-contaminated tailings are leached with cyanide. Manyprocessing plants continue to dump their final cyanidation tailingsinto the local rivers. Over 880,000 tonnes of tailings and miningwaste are estimated to enter the Puyango River annually, contain-ing about 650 kg of Hg and 6000 tonnes of cyanide (Guimarãeset al., 2011).

rpets to concentrate gold e gold technique (Ecuador).

M.M. Veiga et al. / Journal of Cleaner Production 64 (2014) 535e544542

The introduction of flotation at some processing centres toconcentrate fine gold and copper sulphides has had positive eco-nomic and environmental effects in Portovelo. Firstly, copperminerals that were usually dumped into the rivers are now beingsold, and secondly, processors do not need to use mercury or cya-nide since the smelters in Peru and China pay for the contained goldin the concentrates. This technology improvementwas triggered byPeruvian mining engineers in Portovelo associated with the con-stant presence of UBC Dept. of Mining Engineering professors andstudents. Peruvian, Colombian, Canadian and American investorshave established 40 small processing centres that float copperminerals and gold from their own ores or from sluice box tailingsleft at the processing centres. These new centres have capacitiesranging from 30 to 100 tonnes of ore/day. In some of these centres,cyanide tailings are properly managed, impounded, and destroyedwith hydrogen peroxide. Despite the visible evolution of the cen-tres, many still dispose mercury and cyanidation tailings into thelocal rivers and other receiving environments.

Efforts to organize the processing centres by the Government ofEcuador have resulted in developing ways to explore a sustainablemanagement of processing plants that involve miners to collabo-ratively solve common problems such as tailing and waste man-agement. A working group encompassing engineers from thefederal and provincial governments and experts hired by theAPROPLASMIN, devised techniques to collect all processing tailingsin a 22-km pipeline to dispose them in a formally constructedtailing dam located a safe distance from the river. Design studieswere funded by the Government of Ecuador with a large decisionmaking process for its operation carried out with participation ofowners of processing plants. Beside the new tailing pond, a newindustrial park is being designed for the two main regions of arti-sanal and small scale gold mining of Portovelo-Zaruma and PonceEnriquez. By improving the mineral processing conditions at PonceEnriquez, there will be a marked reduction in the environmentalcontamination of town of Portovelo and the Puyango-Tumbes River(the latter is in Peru). All processing centres will eventually move tothe new mining industrial park, and they will be encouraged toimprove gold recovery techniques and to implement more sus-tainable waste management systems that include eliminating or atleast reducing mercury use. The Government is also establishing anew training centre for the local miners to improve their technical,economic, environmental and social skills.

4. Pollution from processing centres

Based on the field observations of many processing centres, themain pollution problems identified are as follows:

1. Lack of cyanide management (both use and disposal)2. Amalgamation of the whole ore resulting in increasing the

mercury losses with tailings3. Use of cyanide to extract residual gold from Hg-contaminated

tailings4. Tailings disposal with mercury, other heavymetals and cyanide

into the receiving environment5. Decomposition of mercury amalgams without any recovery

method or filter.

4.1. Lack of cyanide management

In spite of the efforts of organizations like the InternationalCyanide Management Institute (ICMI, 2012), sodium cyanide isopenly sold in all artisanal mining sites visited. Even brands fromcyanide manufacturers registered in the ICMI are easily acquired by

artisanal miners with little or no instruction or guidance. Locally,cyanide is often stored or used under unsafe conditions. There is noantidote for cyanide exposure in the plants, operators are nottrained and local doctors do not have any knowledge on how toproceed in cases of accidental cyanide ingestion or inhalation.Additionally, many cyanide containers are reused as water pales orgarbage cans. In many cases, there is no control of the pH of theleaching operations and operators do not know that cyanide inneutral or acidic pH generates HCN gas.

4.2. Amalgamation of the whole ore

The main source of mercury losses is associated with amalgam-ationof thewholeoreandnotwith theopen-airburningof amalgams.Amalgamation of the whole ore using copper-amalgamating platescauses losses ofupto2or3partsofmercuryperpartof goldproduced.When amalgamation occurs in ball mills, losses can be even higher inthe region of 15 parts of mercury per each part of gold produced. Onaverage, 46% of the mercury introduced in the ball mills are dis-charged as losses with tailings or leached with cyanide, and four percent –when amalgams are burned. The losswith tailings is due to theformation of small droplets of mercury (known as “flouring”) whenmercury is submitted to long grinding periods. According to Beard(1987), this loss due to lack of coalescence is caused by mercuryoxidation or impurities such as oil, grease, clay minerals, sulphates,and sulphides onmercury surface. Thefinedrops ofmercury lostwiththe tailings dumped into the rivers can move long distances associ-ated with the suspended particles (Adler-Miserendino, 2012).

4.3. Use of cyanide in Hg-Contaminated tailings

As mentioned above, Hg-contaminated tailings are oftenleached by the processing centres. In the cyanidation process,mercury, gold and other metals form soluble complexes with thecyanide, such as [Hg(CN)4]2- which is stable at pHs above 8.5 andHg(CN)2 (aq), stable at pH below 7.8 (Flynn and McGill, 1995).

In agitated cyanidation tanks in localities like Ecuador, approx-imately 90% of the gold from the contaminated tailings is dissolvedinto the cyanide solution, whereas only 40% of the mercury is dis-solved at the same leaching time. Kinetics of cyanideedissolution isslower for mercury than for gold, and mercury cyanide complexeskeep forming as the miners discharge the pulp into the receivingenvironment (Velasquez, 2010).

Gold cyanide complexes formed during the leaching step arethen adsorbed on activated carbon and removed from the leachingsystem without the need of filtration. Some species Hg(CN)2 aremore easily adsorbed on the activated carbon than [Hg(CN)4]2�

(Adams, 1991). Furthermore, the gold cyanidation process itselfoccurs at pH levels between 10 and 11. Under this condition, it isexpected that little mercury reports to the activated carbon. Thetailings fromwhich most of the gold was extracted with carbon arerich in mercuryecyanide complexes. In Portovelo, Ecuador, whereminers leach Hg-contaminated tailings (herein considered 100% ofthe mercury entering the cyanidation tanks), only 3.72% of themercury is leached and removed by the activated carbon, and 51%of the mercury remain with the pulp of contaminated tailings insolution. The rest of the mercury is not dissolved and remains asdroplets with the tailings to be discharged into the water streams.

Gold is also extracted from the cyanide solution by precipitationwith metallic zinc shavings (i.e. Merrill-Crowe process). The Au-enriched solution from the percolation vats or from agitatedtanks pass through PVC columns filled with zinc shavings, and thegold is precipitated. The cyanide solution returns to the tanks, andanother cycle starts. Precipitation with zinc is much more efficientto extract mercury from cyanide solution than activated carbon.

M.M. Veiga et al. / Journal of Cleaner Production 64 (2014) 535e544 543

Unfortunately, the portion of mercury precipitated along with thezinc, is released to the atmosphere when miners irresponsiblyevaporate the zinc shavings to obtain metallic gold (Velasquezet al., 2011).

4.4. Dumping mercury and cyanide into the drainages

The trend observed in many developing countries where pro-cessing centres are operating, is the discharge of cyanidation tail-ings into the rivers and other components of the receivingenvironment. Mercury-cyanide complexes increase the mercurymobility in water streams. Mercuric species are easily methylatedin the aquatic environment (Jensen and Jernelov, 1969). The prob-lem is that mechanisms to methylate mercuryecyanide complexesare not well understood. Whether the mercury cyanide species aremethylated in the sediments or directly bioaccumulated in aquaticorganisms remains unclear. In many sites where Hg-contaminatedtailings are leached with cyanide, fish contain high levels of mer-cury (Rodrigues-Filho et al., 2004; Castilhos et al., 2006; Sousa andVeiga, 2009; McDaniels et al., 2010). Guimarães et al. (2011) foundthat the mercury and cyanide discharge into the Amarillo River(upper Puyango River) in Portovelo, Ecuador, has inhibited meth-ylmercury production by killing the methylating bacteria. In thiscase, the high levels of mercury found in local fish (Velasquez, 2010)suggest the possibility of bioaccumulation of Hg-cyanide com-plexes. This is an interesting newarea to investigate since dischargeof mercuryecyanide complexes is becoming more common withthe spreading of processing centres worldwide.

The mobility of the mercury released by processing centres intothe Amarillo-Puyango-Tumbes River in Ecuador was studied usingisotopic analyses (Adler-Miserendino, 2012). The author suggestedthat the mercury used by the centres is responsible for elevatedmercury and other heavy metals distributed approximately 120 kmdownstream along the Puyango-Tumbes River delta. This situationis creating a diplomatic problem since the delta of the Puyango-Tumbes River lies in Peruvian territory.

4.5. Decomposition of amalgams

The most common process to decompose amalgams is byincreasing the temperature above 460 C when all mercury com-pounds are evaporated leaving behind solid gold (Angeloci, 2013).When the amalgams are burned in open pans and mercury evap-orated, one part of mercury is lost per one part of gold produced, asamalgams usually have 40%e50% Hg and 50%e60% Au and othermetals (e.g. Ag, Cu, etc.). Retorts have been promoted to condenseand reuse mercury and to avoid prolonged exposure to operatorsand public. Most processing centres visited do not have any effi-cient retorting process. Operators give the amalgam to the minersto burn in an adjacent space at the centre. Communal retorts wereintroduced in 1993 by a Swiss Development Agency in Ecuador(Hruschka, 2005). The retorts were fume-hoods with a strongexhaustion system. Theywere not perfect asmercury escapes in theprocess, but it was an important initiative to reduce air pollution.Under a more communal setting, appropriate techniques and safetyprocedures could be introduced. In some places like Indonesia andColombia, it is common for miners to walk away from processingcentres to sell amalgams to gold shops in urban areas. They preferto leave the amalgam-burning step with the buyers. In Segovia,Antioquia Department, Colombia, a gold shop introduced a com-plex system of condensers and activated charcoal filters to abatemercury vapours for roasting amalgams. The air inside of this shop,at the exit of the condensing system, had 100,000 ng of Hg/m3

(Veiga, 2010). These filters and condensers give the impression tooperators and public that they have safe conditions, but it is clear

that it is completely inadequate to burn amalgams in fume hoodseven with condensing systems. A sealed retort is needed, prefer-entially to be used away from the public. Many examples of indi-vidual retorts have been promoted (Veiga, 2006). Currently, theGovernment of Antioquia has established an area outside the urbanperimeter for the miners to burn their amalgams in individualsealed retorts before selling them to gold shops.

5. Conclusion

Processing centres are increasing rapidly around the world,following more or less the same model: offering rudimentary tech-niques for the miners to extract gold and then leaching their Hg-contaminated tailings with cyanide before discharging them intothe receiving environment. The labour division between “rich” pro-cessing centres’ owners and “poor” miners introduces a huge hurdlefor theevolutionof the cleaner technologies and forequitabledivisionof the gold production. This is, in fact, creatingmore pollution. Time isanother factor that creates resistance to cleaner techniques as arti-sanal miners are unwilling towait five ormore days of cyanidation tohave their Hg-free gold. They become trapped in the scheme of theprocessing centres and accept less than30%of the goldextracted fromtheir ores. There are many techniques of concentration and intensiveleaching that can be brought to their attention (Veiga et al., 2009), buttechnical skills and investment are needed.

Moreover, there are many examples of good progress, particu-larly in southern Ecuador. Unfortunately, not all governments ofdeveloping countries have the capacity to understand the impor-tance of the technical evolution of the AGM, and they do notparticipate as supporters or organizers of initiatives to changeminers’ behaviour (McDaniels et al., 2010). Governments try simplyto address the pollution problem associated with AGM by forcingformalization of the sector. Such formalization in developingcountries has been a fiasco. Out of the 80,000 to 150,000 PeruvianAGM in 2012, only 1000 have been fully formalized (TiempoMinero, 2012). In Colombia, of 200,000 AGM only 9196 areformalized (Puentes, 2013). One of the main problems is theformalization procedure. In Peru, for example, an informal minerhas to go through five steps of an intricate bureaucratic process toobtain the full mining title of a gold deposit that he/she discovered(Mineria Informal, 2013). Without education and organization, theformalization initiatives are only formalizing the pollution, asminers will not change their ways of working if they do not haveknowledge and capital. The owners of processing centres are takingadvantage of this situation.

The evolution witnessed in the southern Ecuador is occurringthanks to strong investment of the private sector. This is also due tointernational training co-operation efforts, active participation ofminers in the decision making process and more presence of thegovernment in the mining sites. Finally, the involvement of goodlocal technical people interested in the small mining business ismaking the greatest difference. The Ecuadorian Government islargely investing in the process of organization training andformalization of the miners as a way to improve the workingconditions and quality of life of the mining communities. As aresult, miners have increased their interest in creating associationsand are becoming more aware of the opportunities to make part-nerships to develop a small scale business in mineral resources.This provides jobs for a large contingent of local people and sustainsother social activities in the community.

Acknowledgements

The authors would like to acknowledge the financial support ofthe NSERC Discovery Grant #217089, the US Department of State,

M.M. Veiga et al. / Journal of Cleaner Production 64 (2014) 535e544544

UNIDO (United Nations Industrial Development Organization),Blacksmith Institute and Cassius Ventures Ltd for their financialsupport of this work. This work is dedicated to Dr. Dario Bermudez(in memoriam) from Universidad Experimental de Guyana for hispermanent support, friendship and dedication to improve theconditions of artisanal gold miners in Venezuela. We also expressour gratitude to Dr Gavin Hilson for his revisions of the manuscript.

References

Adams, M.D., 1991. The mechanisms of adsorption of Hg(CN)2 and HgCl2 ontoactivated carbon. Hydrometallurgy 26, 201e210.

Adler-Miserendino, R.A., 2012. Tracing Mercury Pollution in Aquatic Ecosystems:Implications for Public Health. PhD thesis. Johns Hopkins University, Baltimore,US, p. 199.

Angeloci, S.G., 2013. Myths and Realities in Artisanal Gold Mining MercuryContamination. MASc thesis. Department of Mining Engineering, University ofBritish Columbia, p. 180. Retrieved from https://circle.ubc.ca/bitstream/handle/2429/44361/ubc_2013_spring_angelocisantos_gustavo.pdf?sequence¼3 (June3, 2013).

Ashe, A., 2012. Gold Mining in the Peruvian Amazon: A View from the Ground.Retrieved from http://news.mongabay.com/2012/0315-ashe_goldmining_peru.html?utm_campaign¼Generalþnews&utm_medium¼Twitter&utm_source¼SNS.analytics (Jan. 7, 2013).

Beard, R.R., 1987. Treating Gold Ores by Amalgamation. Department of Mines andMineral Resources. Retrieved from http://mines.az.gov/Publications/circ027amalgam.html (March 25, 2013).

Brooks, W.E., Sandoval, E., Yepez, M.A., Howard, H., 2006. Peru Mercury Inventory2006. Open File report 2007e1252. U.S. Department of the Interior, U.S.Geological Survey.

Castilhos, Z.C., Rodrigues-Filho, S., Rodrigues, A.P., Villas-Bôas, R.C., Siegel, S.,Veiga, M.M., Beinhoff, C., 2006. Mercury contamination in fish from gold miningareas in indonesia and human health risk assessment. Science of the TotalEnvironment 368, 320e325.

Chalze, T., 2013. World gold production 2013. In: Mining Feeds. Retrieved fromhttp://www.miningfeeds.com/2013/04/22/world-gold-production-2013/ (May25, 2013).

Cordy, P., Veiga, M.M., Salih, I., Al-Saadi, S., Console, S., Garcia, O., Mesa, L.A.,Velásquez-López, P.C., Roeser, M., 2011. Mercury contamination from artisanalgold mining in Antioquia, Colombia: the World’s Highest Per Capita MercuryPollution. Science of the Total Environment 410, 154e160.

Flynn, C.M., McGill, S.L., 1995. Cyanide Chemistry e Precious Metals Processing andWaste Treatment. USA Bureau of Mines, NTIS Publication, p. 282. PB96e117841.

Fowks, J., 2012. El Intento de Perú de Atajar la Minería Informal Eleva el MalestarSocial. El País, Internacional. Mar 16, 2012. Retrieved from http://internacional.elpais.com/internacional/2012/03/16/actualidad/1331887231_484790.html (Jan.7, 2013).

Guimarães, J.R.D., Betancourt, O., Miranda, M.R., Barriga, R., Cueva, E., Betancourt, S.,2011. Long-range effect of cyanide on mercury methylation in a gold miningarea in southern Ecuador. Science of the Total Environment 409, 5026e5033.

Hilson, G., 2006. Abatement of mercury pollution in the small-scale gold miningindustry: restructuring the policy and research agendas. Science of the TotalEnvironment 362 (1e3), 1e14, 1.

Hilson, G., Pardie, S., 2006. Mercury: an agent of poverty in Ghana’s small scalegold-mining sector? Resources Policy 31, 106e116.

Hruschka, F., 2005. Ensuring Environmental Sustainability. Paper presented at TheMillennium Development Goals and Small-Scale Mining, Conference for Forg-ing Partnerships for Action, Washington, 2005. Retrieved from http://www.eisourcebook.org/cms/June%202013/MDGs%20&%20Small-scale%20Mining,%20focus%20on%20Environment.pdf (June 3, 2013).

ICMI e; International Cyanide Management Institute, 2012. International CyanideManagement Code for the Manufacture, Transport and Use of Cyanide in theProduction of Gold. Retrieved from http://www.cyanidecode.org/ (Nov. 19, 2012).

Ismawati, Y., Petrlik, J., DiGangi, J., 2013. Mercury Hotspots in Indonesia. Retrievedfrom http://www.ipen.org/hgmonitoring/pdfs/indonesia-report-en.pdf (May25, 2013).

Jensen, S., Jernelov, A., 1969. Biological Methylation of mercury in aquatic organ-isms. Nature 223, 753e754.

Krisnayanti, B.D., Anderson, C.W.N., Utomo, W.H., Feng, X., Handayanto, E.,Mudarisna, N., Ikramb, H., Khususiah, 2012. Assessment of environmentalmercury discharge at a four-Year-Old artisanal gold mining area on LombokIsland, Indonesia. Journal of Environmental Monitoring 14, 2598e2607.

McDaniels, J., Chouinard, R., Veiga, M.M., 2010. Appraising the Global mercuryproject: an Adaptive management Approach to Combating mercury pollution insmall-scale gold mining. International Journal of Environment and Pollution 41(3e4), 242e258.

Metcalf, S., Veiga, M.M., 2012. Using street Theatre to increase Awareness of andreduce mercury pollution in the artisanal gold mining sector: a case fromZimbabwe. Journal of Cleaner Production 37, 179e184.

Mineria Informal, 2013. Proceso de Formalización de Minería Muestra AvancesImportantes. Afirma Ministro del Ambiente. Retrieved from http://www.minam.gob.pe/mn-ilegal/index.php?option¼com_

content&view¼article&id¼165:proceso-de-formalizacion-de-mineria-muestra-avances-importantes-afirma-ministro-del-ambiente&catid¼1:notas-de-prensa&Itemid¼93 (June 13, 2013).

Molina, et al., 2013. El Oro, el Primer Producto de Exportación en 2013. Retrieved fromhttp://www.molinalaw.com.ni/_blog/Blog_de_Molina_y_Asociados_Central_Law_Nicaragua/post/el-oro-el-primer-producto-de-exportacion-en-2013/ (June13, 2013).

MRNNR e; Ministerio de Recursos Naturales Non-Renovables, 2013. Ecuador Tiene1349 Labores Mineras Artesanales. Retrieved from http://190.11.28.23/en/inicio/destacados/959-ecuador-tiene-1349-labores-mineras-artesanales.html (June 3,2013).

Peru Support Group, 2012. Artisanal and Small-scale Mining in Peru: A Blessing or aCurse?, p. 23. Retrieved from http://www.perusupportgroup.org.uk/files/fckUserFiles/file/Artisanal%20and%20Small-scale%20Gold%20Mining%20in%20Peru.pdf (Jan. 7, 2013).

Puentes, C.A.C., 2013. Formalización de la Pequeña Mineria Colombiana. Ministeriode Energia y Minas. Retrieved from http://www.minminas.gov.co/minminas/downloads/UserFiles/File/minas2/1-FormalizacionDeLaPequenaMineriaColombiana.pdf (June 10, 2013).

Rodrigues-Filho, S., Castilhos, Z.C., Santos, R., Yallouz, A.V., Nascimento, F., Egler, S.G.,Peregovich, B., Almeida Riberio, R.A., 2004. Environmental and Health Assess-ment in Two Small-scale Gold Mining Areas in Brazil (Sao Chico and Crepor-izinho). Technical report by CETEM to UNIDO, RT2004-004-02. Rio de Janeiro.Retrieved from http://www.globalmercuryproject.org (Dec. 12, 2012).

Simpson, J., 2007. The Shamva Mining Centre, Zimbabwe: Small-scale Mining CaseStudy. Retrieved from http://practicalaction.org/t4sl_casestudy_mining (Jan. 7,2013).

Sousa, R.N., Veiga, M.M., 2009. Using performance indicators to evaluate an envi-ronmental education program in an artisanal gold mining in the BrazilianAmazon. Ambio 38 (1), 40e46.

Sousa, R.N., Veiga, M.M., Klein, B., Telmer, K., Gunson, A.J., Bernaudat, L., 2010. Strategiesfor reducing the environmental impact of reprocessing mercury-contaminatedtailings in the artisanal and small-scale gold mining Sector: Insights from Tapa-jos river Basin, Brazil. Journal of Cleaner Production 18, 1757e1766.

Telmer, K., Veiga, M.M., 2008. World emissions of mercury from small scale arti-sanal gold mining and the knowledge Gaps about them. In: From Pirrone, N.,Mason, R. (Eds.), Mercury Fate and Transport in the Global Atmosphere: Mea-surements, Models and Policy Implications. UNEP-United Nations EnvironmentProgramme, pp. 96e129 (2008). Chapter 6.

Tiempo Minero, 2012. De 80 Mil Mineros Informales, Solo Mil Se Han Formalizado.Retrieved from http://www.tiempominero.com/index.php?option¼com_content&view¼article&id¼3623:de-80-mil-mineros-informales-solo-mil-se-han-formalizado&catid¼34:peru&Itemid¼59 (June 3, 2013).

UNEP e; United Nations Environment Programme, 2008. Mercury Use in Artisanaland Small Scale Gold Mining. Retrieved from http://www.chem.unep.ch/mercury/awareness_raising_package/E_01-16_BD.pdf (Nov. 1, 2012).

UNEP e; United Nations Environment Programme, 2013. Mercury Time to Act.Retrieved from http://www.unep.org/PDF/PressReleases/Mercury_TimeToAct.pdf (May 30, 2013).

Veiga, M.M., 1996. Advisory Assistance on Avoidance Mercury Pollution fromArtisanal Gold Mining Operations in Bolivar State, Venezuela. Report to UNIDOe United Nations Industrial Development Organization, p. 111. Retrieved fromhttp://www.globalmercuryproject.org/countries/venezuela/docs/venezuela%20unido%201996.pdf (July 03, 2013).

Veiga, M.M., 1997. Introducing New Technologies for Abatement of Global MercuryPollution in Latin America. UNIDO/UBC/CETEM pub., Rio de Janeiro, p. 72.Retrieved from http://www.globalmercuryproject.org/documents/non_country%20specific/HginLA-book.pdf (May 3, 2013).

Veiga, M.M., 2006. Retorts: Many Options and Many Barriers (Searching for Sus-tainable Solutions). Global Mercury Project, UNIDO, Vienna. Retrieved fromhttp://www.globalmercuryproject.org/documents/non_country%20specific/Retorts.pdf.

Veiga, M.M., 2010. Antioquia, Colombia: The World’s Most Polluted Place byMercury: Impressions from Two Field Trips, p. 26. Report to UNIDO. Retrievedfrom http://redjusticiaambientalcolombia.files.wordpress.com/2011/05/final_revised_feb_2010_veiga_antioquia_field_trip_report.pdf (May 21, 2013).

Veiga, M.M., Nunes, D., Klein, B., Shandro, J.A., Velasquez, P.C., Sousa, R.N., 2009.Replacing mercury use in artisanal gold mining: preliminary tests of mill-leaching. Journal of Cleaner Production 17, 1373e1381.

Velasquez, P.C., 2010. Mercury in Artisanal and Small Scale Gold Mining: IdentifyingStrategies to Reduce Environmental Contamination in Southern Ecuador. PhDDissertation at the Department of Mining Engineering of University of BritishColumbia, Vancouver, Canada, p. 184. Retrieved from https://circle.ubc.ca/handle/2429/27914 (Jan. 7, 2013).

Velasquez, P.C., Veiga, M.M., Hall, K., 2010. Mercury balance in amalgamation inartisanal and small-scale gold mining: finding ways to reduce environmentalpollution in Portovelo-Zaruma, Ecuador. Journal of Cleaner Production 18 (3),226e232.

Velasquez, P.C., Veiga, M.M., Klein, B., Shandro, J., Hall, K., 2011. Cyanidation ofmercury-rich tailings in artisanal and small-scale gold Mining: Identifyingstrategies to Manage environmental risks in southern Ecuador. Journal ofCleaner Production 19, 1125e1133.

WHO e; World Health Organization, 2007. Exposure to Mercury: a Major PublicHealth Concern, p. 4. Retrieved from http://www.who.int/phe/news/Mercury-flyer.pdf (June 3, 2013).