Projected global mercury supply, demand, and excess to 2050 based on impacts of the Minamata Convention

There is a crucial need to understand the outlook for changes in mercury supply and demand in response to the 2017 entry into force of the Minamata Convention on Mercury. If demand exceeds supply there will be a shortage, but if supply exceeds demand there will be excess mercury, and ultimately the need to prepare for the final disposal of elemental mercury. This study estimates global mercury supply and demand from 2010 to 2050, by region, and projects excess mercury based on recent trends and consideration of impacts of the Convention. The analysis notes a downward trend in both supply and demand, but the supply of mercury from primary mining has not decreased as much as initially expected, while mercury recovery from by-products of nonferrous production has not yet been implemented extensively. Under a business-as-usual projection, there is excess supply until a shortage appears around 2030, but a surplus returns after 2035, and there is a projected excess supply of 59 tons annually in 2050. Relying on the level of mercury recovery in the nonferrous sector, global cumulative stock of excess mercury is projected at 14,439–20,943 tons in 2050.


Introduction
Following the entry into force of the Minamata Convention on Mercury in 2017, the use of mercury in industrial processes and products has been restricted based on a phase-out (or phase-down) schedule, so global demand for mercury is expected to decline.In Japan, domestic demand for mercury was already low, at 3.5 tons in 2016, due to the broad use of mercury substitution technologies.That same year, 65 tons of mercury were recovered from both mercury-added products and by-products containing mercury (unintentionally discharged from industrial processes), of which 70.8% was recovered from sludge generated by the nonferrous metal smelting industry [1].Most of Japan's recovered mercury is exported after being refined, but in response to an expected decline in international demand, elemental mercury, which was traditionally traded as a valuable commodity, will become waste and need to be disposed of as waste.
To be prepared for such a situation, measures for the permanent disposal of elemental mercury are being considered.In Japan, the Waste Management and Public Cleansing Act was revised in 2015 to set standards for the final disposal of mercury waste, but an actual final disposal system has not yet been established and will have to be developed in the near future [2].In Europe, as well, standards for the final disposal of elemental mercury were set by European Union Regulation (EU) 2017/852 in 2017, opening the way for final disposal [3].Given limited available capacity for the conversion of liquid mercury waste, however, the temporary storage of liquid mercury waste is still being allowed, until the capacity is in place to ensure the conversion and, if applicable, solidification of all mercury waste.In the United States, no method approved by the Environmental Protection Agency (EPA) has yet been adopted for treating elemental mercury for land disposal (as of October, 2022), but efforts have begun to consider new land disposal treatment technology to stabilize elemental mercury extracted from high-level mercury-containing waste [4].
Meanwhile, Article 3 (including Annex A) of the Minamata Convention allows the use of mercury for essential uses, such as products essential for civil protection and military uses.Also based on Articles 5 and 7, the use of mercury is allowed to remain for the time being in some manufacturing processes and artisanal and small-scale gold mining (ASGM).Therefore, in establishing systems for the final disposal of excess elemental mercury, there is an urgent need to understand the outlook for the generation of excess mercury, based on projections of international mercury supply and demand.
With regard to future supply and demand projections, the United Nations Environment Programme (UNEP) has estimated the amounts of excess mercury in Asia, South America, Eastern Europe, Central Asia, and Russia from 2010 to 2050 based on actual 2005 levels during the period 2009-2010 [5][6][7], and projected that all regions will have a surplus of mercury by 2020.Considering international trends in waste mercury management [8,9], the estimates were subsequently updated based on the impacts of mercury export bans in Europe and the United States and the situation after the adoption of the Convention, such as a restriction on the market supply of mercury stocks from the decommissioning of chlor-alkali facilities.The resulting projections suggested a cumulative global excess in 2050 of 24,000-25,000 tons.However, the projections revealed that trends in mercury recovery from the nonferrous smelting industry will have a significant impact, as the global projection of excess mercury ranged from about − 2 tons to + 35,000 tons based on a scenario of no versus active mercury recovery from the nonferrous smelting industry [10].Trends in China are also a key element for any projections due to the country's huge volumes of both demand and supply for mercury.An estimate of annual excess mercury was at 2725-2845 tons in 2016 [11], and according to one projection, cumulative excess mercury in China will reach around 10,000 tons in 2050, assuming no exports [12].
However, no estimation of the generation of excess mercury at a global-level incorporating recent changes in accordance with the effects of the Minamata Convention has been conducted so far.Therefore, this paper aims to estimate mercury supply and demand at a reginal basis and show future global prospects for the generation of excess mercury by examining the latest situation after the entry into force of the Minamata Convention and considering more plausible future scenarios for each region.

Mercury trade
In order to overview trends in international mercury flows, this study analyzes import and export trade statistics for Harmonized System (HS) code 280540 (mercury) in a UN trade database [13].Data for this study cover trade records from 2015 to 2021, the latest year available, and countries chosen were those that exported more than 20 tons per year for a minimum of two years during that period.
Although trade in mercury compounds and mercuryadded products also entails the transfer of mercury, this study focuses on flows of elementary mercury, since details of compounds and products cannot be identified by the UN trade data, and exporting countries of elementary mercury are expected to be significantly affected by any decline in demand for mercury, thereby leading to generation of future excess mercury there.The latest trends of mercury trade price were also examined.It is known that the UN trade data can provide the data source for trade analysis, but data discrepancies between exports and imports and rather low reliability are pointed out at the same time.Therefore, the trade data are not applied to estimation process in this research.

Global outlook
First, the status of primary mercury production was examined from the U.S. Geological Survey (USGS) Mineral Yearbook [14,15], and then, based on the supply of mercury to the world market as determined from the trade statistics mentioned above, an outlook of the mercury supply situation was developed, and sources of mercury that could become excess in the future were examined.
With regard to mercury waste disposal, interviews were conducted with officials of competent authorities on mercury waste in Japan, the EU, and the United States, as well as mercury recyclers.

Estimates and projections
Referring generally to the estimation methods of the study by  and Sodeno et al. [10], this study estimated the global mercury supply and demand from 2010 to 2050 at 5-year intervals by updating base data for 2015 and 2020 and future scenarios based on GMA reports, research papers especially focusing on China [32][33][34], government-and industry-published materials, and statistical data.
Since hazardous waste such as mercury is expected to be managed on a certain scale, estimates in this study were made on a regional basis in order to understand mercury supply and demand by region.Considering consistency with the GMA and similarities of the corresponding mercury management situation, the analysis was classified into seven regions: Europe, North America and Oceania, Asia (China is indicated separately because of its large impact.),Commonwealth of Independent States (CIS), Latin America, and the Middle East and Africa.The supply and demand in the Middle East and Africa were addressed for the first time in this study.
Table 2 shows detailed information on base data and scenarios used in this study.

Supply
The mercury supply sector covers primary mining and recycled mercury recovered from manufacturing processes (nonferrous metal production, natural gas refining, and vinyl chloride monomer (VCM)) and mercury-added products.Mercury eliminated by the decommissioning of chlor-alkali cells after the Convention entered into force is directly counted as excess mercury and not included in the supply data, since recovery, recycling, reclamation, and reuse from that source are prohibited under Article 3.5(b) of the Convention.
For future scenarios, primary mercury mining is assumed to be half of 2020 levels in 2025 and decreases toward complete phase-out in 2032, since mercury production has not decreased from the initial UNEP projection conducted in 2009.
For nonferrous metal production, this study includes zinc, lead, copper, and large-scale gold production.Japan's nonferrous industry has a unique mutually complementary system whereby one process's mineral output is another process's input.In Japan's nonferrous smelting processes, mercury concentration is increased by circulating minerals and mercury-containing by-products back into the process.
As a result, sludge containing high concentrations of mercury discharged from the exhaust gas treatment process is transferred to a mercury refiner to recover just the mercury, and the by-products are then cycled back into the smelting process.This recovered mercury is the source of mercury exported from Japan.Since only one company in Japan recovers mercury from the nonferrous smelting industry, government reports reflecting the results of interviews with recycling companies are considered more reliable for the amount of mercury recovered.For this reason, the reported values (36 tons in 2010, 55 tons in 2014, 46 tons in 2016) were used as the base data for Japan [35][36][37].
Meanwhile, UNEP estimates [5] applied Japan's mercury recovery rate of over 90% to the entire Asia region as a future scenario for mercury recovery from nonferrous metal smelting.As mentioned above, the mercury recovery system in the Japanese nonferrous industry is unique, however, and it may not be realistic for other countries to follow Japan's example and develop large facilities specialized for mercury recovery from sludge.Furthermore, there has been no report of mercury recovery by operators pursuant to Article 6 of the Mercury Export Ban in the EU [38], and in the view of EU and U.S. officials, sludge is to be disposed of in landfills as ordinary hazardous waste containing mercury, since recovery as elemental mercury would be subject to special regulations.Canada has reported that mercury-containing by-products from the smelting and refining process for base metal ores and gold are exported for further treatment in the United States [39].Therefore, for this study it is assumed that no active mercury recovery would take place in those countries unless sufficient demand exists for mercury.As a reference for the level of mercury recovery in China, the recovery rate calculated from the amount of recovered mercury in UNIDO's zinc smelting project in China was 30% [10,48].
Given this information, three mercury recovery profiles were created for this study, and the mercury recovery rates 50% reduction in 2020, 5% reduction per year thereafter China: phase-out in 2020 Chlor-alkali production 2010, 2015 data from WCC [31] a China: phase-out in early 2000s [34] Phase-out in 2025 VCM China: 2015, 2020 data are calculated from Lin et al. [34] India: 2015 data are calculated from activity data (VCM-P:GMA [16]) and Hg factor (low Hg catalyst [34]) 50% reduction from 2010 in 2020, then linearly decrease to zero in 2027

Batteries
After 2015, linearly decrease to zero in 2025 Measuring devices 2010 data for Asia, CIS, South America are from UNEP [5][6][7] After 2015, linearly decrease to zero in 2025 Lamps 2010 for other regions, 2015 are regional data from GMA [16,17] Gradual 20% reduction from 2015 by 2050 Electrical and electronic devices 2015 data for China and Asia: prorating GMA's 2015 estimates for the East and Southeast Asia with UNEP's estimates for 2015 [5] Gradual 55% reduction from 2015 by 2050 Others Europe: 2015 data [19] Gradual 25% reduction by 2020, another 50% by 2050.

Dental amalgam
After 2015, linearly decrease to zero in 2050 EU: phase-out in 2030 from by-products were revised as 30% for Profile 1 (except for Japan and Europe, at 1%), 5% for Profile 2 (except for China, at 10%), and 0% for Profile 3. It should be noted that by-products in the "recovery rate from by-products" in this study means the amount of mercury, excluding atmospheric emissions, from the total mercury input.Thus, by-products can include emissions to slag and water systems, and the "recovery rate from by-products" in this study appears lower than the general recovery rate from sludge of by-products.
The data for nonferrous metal production (zinc, lead, copper, gold) in 2020 were updated based on USGS reports [14,[40][41][42].As mentioned above, reported values were adopted for recovered amounts in 2010 and 2015 in Japan, so adjusting the volume recovered to the same level as reported values, the recovery rate in Japan was about 50-60%.The estimates of mercury recovery from nonferrous metal production in 2020 are indicated, by mineral, in Tables 3, 4, 5, and 6.The amount of recovered mercury (RHg) in metal i production industry is calculated by the following formula: where i represents the metal (Zn, Pb, Cu, Au), SP the smelter production (tons), UEF the Unabated emission factor (g Hg/ ton metal i produced), DF the Distribution factor to by-product and residue (= reduction efficiency in GMA) (%), and RR the recovery rate from by-product and residue (%).
With regard to future scenarios for the zinc production industry, Sverdrupa et al. (2019) predicted that the total supply of copper would reach a maximum in 2030-2045, zinc in 2030-2050, and lead in 2025-2030 [43].They assume that only poor-grade zinc ore was available after 2015 and that zinc mining would peak in 2025 and then decline steadily.Thus, for this study, the nonferrous smelting volume for zinc, lead, and copper was assumed to decrease by 1% annually after 2040.In addition, given the assumption that only poor-grade zinc ore would be available after 2015 [43] and the future outlook in the Japanese nonferrous smelting industry's decarbonization implementation plan assumes an annual 1% degradation in ore and concentrate grades [44], mercury emissions were assumed for this study to increase by 1% annually, due to gradual degradation in the grade of ore.Since the nonferrous metals industry is a large potential source of mercury, three different recovery rates from byproducts after 2030 were considered for this study, for major countries in the nonferrous industry: business-as-usual (BAU) scenario (RR = 0-30%), medium recovery scenario (RR = 5-30%), and high recovery scenario (RR = 10-50%).For natural gas refining, the International Energy Agency (IEA) future outlook [45] was reflected in the assumption of a downward trend in production, based on impacts of climate change responses and the 2022 invasion of Ukraine.Natural gas production was assumed to be reduced by 40% in Europe and 50% in Russia by 2030 and constant thereafter.
For mercury recovered from VCM, recovery rates were updated from Lin et al. [34] and GMA [16].For industrial processes, mercury recovery was assumed to maintain the status quo, since it is assumed that if mercury supply and demand are in equilibrium, there would be little incentive to recover mercury from by-products.
For mercury recovered from mercury-added products, considering a time lag between product shipment and disposal, demand data of mercury-added products from 5 years ago are used as the base data.For this study, three profiles were set according to the level of progress in mercury recovery with reference to the recovery rates set by GMA, and the amount of recovered mercury was calculated (except for the year 2010).Based on interviews with recyclers in the EU [19] and Japan [1], the actual recovery values for 2015 were higher than those estimated by GMA [16], suggesting that stock in the market may have been discarded.For the future scenario it is assumed that after 2025, more products will be recovered and the recovery rate will increase to 25% for Profile 1, 15% for Profile 2, and 5% for Profile 3.

Demand
The mercury demand sectors cover artisanal and small-scale gold mining (ASGM), chlor-alkali manufacturing, vinyl chloride monomer (VCM) production, and mercury-added products (batteries, measuring and control devices, lamps, electrical and electronic devices, other uses such as research and pharmaceuticals, and dental amalgam).
For this study, the future scenario for mercury-added products is basically the same as the UNEP scenario (Asia), except for the EU, where a feasibility study was conducted to accelerate the phase-out of dental amalgam [46], and therefore, the assumption of total elimination by 2030 is used.With respect to the future scenario for VCM, China is the largest user of mercury, but demand is estimated to peak at 903 ± 115 tons in 2014 and decline thereafter due to a shift to catalysts with lower mercury content [34], and hence, in accordance with the phase-out schedule of the Minamata Convention, demand is halved from 2010 levels by 2020 and phased out by 2030.

Prospects of excess mercury
For the outlook for global mercury supply and demand, a simple sum of mercury supply and demand for each region was calculated and evaluated.The intention of indicating simple sums of excess mercury by region is to contribute to the consideration of needs in developing a disposal infrastructure for respective regions in terms of their excess mercury capacities.In addition, a sensitivity analysis based on three scenarios was conducted for the supply of recovered mercury in the nonferrous metals sector, which involves significant uncertainties.
Based on the above results, an annual amount of excess mercury generated and accumulated mercury stock by region after 2020 until 2050 were estimated, using supplied mercury in excess of demand as the amount of excess mercury.Changes were assumed to occur linearly during every fiveyear period.Stored mercury, including mercury discharged from the decommissioning of mercury cells in the chloralkali sector and stored without being supplied to the market, was integrated into the inventory as excess mercury in 2020.
The results of projections were compared to previous studies.

Mercury trade
With regard to the price of elemental mercury, Fig. 1 shows the trade value price trends for mercury from 2015 to 2021.The average price is calculated in U.S. dollars per kilogram by dividing the total trade value by the total export volume for countries where both data series are available.The highest and lowest prices indicated in the table are calculated by dividing the trade value by the export volume for each country.Although the highest prices are often outliers and differ widely throughout the year, average mercury prices have declined since 2018, and in fact, average mercury prices in 2021 were a quarter of 2018 prices, possibly the impact of falling demand for mercury.Under the Minamata Convention, as shown in Table 1, 2018 is the phase-out year for acetaldehyde production (in which mercury or mercury compounds are used as a catalyst), and 2020 is the phaseout year for most mercury-added products, as well as the target year for reducing the use of mercury in vinyl chloride monomer production and sodium or potassium methylate or ethylate production.
Table 7 shows the trends of mercury exports in major exporting countries, excluding Singapore, Hong Kong, and United Arab Emirates, which are considered transit countries with large import volumes; India and Turkey, which are considered importing countries whose import volume mostly exceeds export volume; and EU countries having export bans of mercury outside the EU.Indonesia and Mexico, once major exporters, have recently stopped exporting mercury, while Japan, Switzerland, and Canada have been continuously exporting mercury, presumably recovered from byproducts and mercury-added products, though the amount has been declining, to 29 tons in Japan, 11 tons in Switzerland, and 1 ton in Canada as of 2021 [14].In response to the Convention, Switzerland and Canada introduced mercury export restrictions with the exception of research and analytical uses, in addition to allowing dental amalgam use until the end of 2027 in Switzerland [22,23].Japan limits exports to purposes permitted by the Convention, other than ASGM [24].Meanwhile, it seems that Nigeria has been rapidly increasing its exports since 2020.
In terms of mercury imports, India and Bolivia stand out as the largest players, importing more than 100 tons of mercury every year (except 2021 in the case of Bolivia), while other countries import only small amounts of mercury.
Nigeria reported exports of more than 100 tons of mercury in 2021 to each of Albania, Belgium, China, India, Malaysia, and the Netherlands.However, according to trade statistics on the importing country side, none of those six countries reported mercury imports from Nigeria, indicating a large discrepancy between data from the exporting and importing countries.Furthermore, according to the Minamata Convention Initial Assessment (MIA) report on Nigeria in 2017 [18], mercury discharged as general waste in Nigeria was estimated as 35 tons per year, mostly from consumer products with intentional use of mercury.Also, according to the national report in 2021 based on Article 21 of the Convention [25], there were no primary mercury mines and no mercury stocks exceeding 50 tons in Nigeria.Therefore, the reported export amounts from Nigeria in 2021 appear to be too high and should be considered unreliable.Aside from the 2021 data, this study found that total global exports of mercury declined gradually from 1284 tons in 2015 to 504 tons in 2020.Hence, the destination of excess mercury in exporting countries should be closely monitored in the future.It should be noted that this overview of the mercury trade does not include the transfer of mercury in the form of compounds or mercury-added products and illegal trade which cannot be seen in the official statistics.It should also be taken into account that, as the Nigerian case also shows, the UN trade data have been pointed out to be rather unreliable, including discrepancies between export and import data [49,50].

Outlook for global mercury supply
Under the Minamata Convention, new primary mercury mining has been prohibited since 2017, but mercury production from existing mines is allowed until 2032.Figure 2, created from USGS statistics [14,15], shows the trend of mercury mining in the world's top five mercury-producing countries, and the world total.Although the Minamata Convention encourages a reduction of reliance on primary mercury, primary mercury mining was not restrained after 2015, but in 2020 and 2021, there was finally a decline in global mercury production.
As observed above in 3.1, China accounts for most of the primary mercury production and is estimated to have produced 2000 tons in 2021.However, considering that China is also an importer of mercury, mining in China could be regarded as balanced so as to meet its own demand.Therefore, in terms of the supply of mercury to the world market, the study found that currently over 200 tons of primary mining mercury from Tajikistan, Mexico, and Peru, and 30-40 tons of recycled mercury recovered from by-products from Japan, Switzerland, and Canada are expected, in addition to temporary supplies from mercury stocks in other countries.
In China, recovered mercury from the VCM sector is the largest contributor to supply, with an estimated 654 tons going to recycling in 2011, all of which is considered to have been consumed domestically [34].The EU and the United States are reported to have large stocks of elemental mercury [38,47], but because of mercury export bans, they cannot be a source of supply for the global mercury market.

Outlook for global mercury demand
Artisanal and small-scale gold mining (ASGM) is expected account for the largest amount of mercury use.The Minamata Convention, based on Article 7, only requires each country to develop and implement a national action plan for taking steps to reduce the use of mercury in ASGM.Although significant uncertainties exist due to activities in the informal sector, GMA 2018 [16] estimates that 2059 tons (min.986, max.3132 tons) of mercury were used for ASGM purposes worldwide, with Indonesia accounting for 427 tons (2014), followed by Peru at 327 tons (2017) and Colombia at 175 tons (2014).China seems to have sharply reduced its use of mercury in ASGM applications due to regulations banning its use in ASGM [16,34].
For chlor-alkali usage, China already found replacements for mercury methods in the early 2000s [34].As of 2016, there were 34 mercury cell plants worldwide, 21 of which operated in Europe, which consumed 50 tons of mercury annually, compared to the global total of 94 tons [29].But under the Industrial Emissions Directive, this mercury-based production technology in Europe was decommissioned by the end of 2017 [33].Furthermore, mercury usage in chloralkali production will be phased out in 2025 in accordance with the Convention.As Article 3 of the Convention requires that measures be taken to prevent that excess mercury from being recovered and re-supplied to the market, it is necessary to carefully monitor the proper management of those large amounts of mercury that will no longer be needed as a result of the decommissioning of mercury cells.
For mercury-added products, overall mercury use is slowly decreasing, but mercury use increased in 2015 compared to 2010 for measuring instruments (estimated increase from 267 to 392 tons) and lamps (estimated increase from 112 to 173 tons) [16].According to the Minamata Convention, many mercury-added products were to have a phase-out in 2020, but demand for dental amalgam is still expected to continue due to the lack provisions for its elimination.

Regional estimates and projections
The results of future projections of mercury supply and demand trends are explained below.Detailed data by seven regions are shown in supplementary materials.
Europe In Europe, mercury demand and supply both declined steadily throughout the entire period.There is an estimated overall shortage of 35 tons in 2030, a level that can be adequately compensated by the recovery of mercury from nonferrous smelting if its mercury recovery rate increases to about 10%, which would result in a 19-ton supply surplus.According to COWI [19], the total amount of mercury recovered from waste in the EU in 2015 by the major recycling companies was estimated at 34-56 tons, and the projected demand for mercury in 2021 was estimated at 26-82 tons, while the potential for recycling was estimated at 12-39 tons.Compared to this estimate, the supply estimate in this study is similar, but with regard to demand, it is possible that the use of mercury in products and laboratory applications is being reduced at a faster pace than estimated in this study.It should be noted that Europe is characterized by a large mercury stock held by the chlor-alkali sector, but since mercury from cell decommissioning is not allowed to be supplied to the market, about 6000 tons of mercury stock reported at the end of 2014 are destined for temporary storage or final disposal.

North America and Oceania
In North America and Oceania, the supply of mercury begins to decline after 2030, while demand declines steadily throughout the entire period.The supply-demand balance is projected to be in a state of oversupply from 2020, with an annual excess of 84 tons per year in 2030.According to the DOE (2022) [4], in addition to the 1206 tons of mercury originally possessed by the U.S. government (National Nuclear Security Administration), 301 tons of mercury were reported to be in storage by commercial storage entities in February 2018, indicating that the supply was already in excess then.In North America and Oceania, the source of mercury is recovery from nonferrous production and mercury-added products after the discontinuation of mercury cells in the chlor-alkali sector.The DOE estimated that 130 tons of excess mercury are generated annually in the United States, but given that the estimates in this study are totals for the United States, Canada, and Oceania, it is possible that the demand for mercury in products is estimated too high in this study.

China
In China, mercury demand and supply both decline sharply throughout the entire period.The supply-demand balance shows a state of excess supply throughout the entire period, with a projected excess of 355 tons annually in 2030, but the degree of excess supply is expected to decrease to 93 tons in 2050 in accordance with mine closures.During the early part of the estimation period, primary mercury mining is extremely high, and the large amounts of mercury continue to be mined (as much as 2200 tons in 2020) even after the Minamata Convention comes into effect.However, the recent decline in primary mercury mining and the phase-down of VCM, which once accounted for much of the demand, indicate that the scale of both supply and demand is expected to shrink.China reported primary mercury min-ing in its first national report pursuant to Article 21 of the Convention, which is much lower than the USGS production data, at 109 tons in 2020 [28].Since not much mercury is exported from China [13] and it is assumed to be mostly consumed domestically [34], it is possible that mercury is stocked domestically, USGS mercury production data may be overestimated, or mercury may not be recovered from industrial processes to the extent estimated.The material flow in 2010 as estimated by Hui et al. [32] indicates a stock of 795 tons of mercury, which suggests that the over-supply estimated in this study seems reasonable.

Asia (excluding China)
In Asia (excluding China), both supply and demand are estimated to decline from 2015.The supply-demand balance of mercury is projected to remain in a state of excess demand until 2045, with a supply shortfall of 137 tons in 2030.The period of supply shortfall is shortened to 2040 for the medium-level recovery scenario and to 2035 for the high-level recovery scenario.Significant mercury demand is expected for ASGM in Asia (excluding China), but there is a large degree of uncertainty in the amount of demand.If ASGM applications are eliminated in 2030, supply will exceed demand under a high-level recovery scenario.
Latin America In Latin America, both supply and demand are estimated to decline after 2015.The mercury supplydemand balance is projected to remain in a state of excess demand after 2015, with a supply shortfall of 223 tons in 2030 and 74 tons in 2050.For the high-level recovery sce-nario, supply exceeds demand after 2045.Significant mercury demand for ASGM in Latin America is expected, leaving a large degree of uncertainty in the amount of demand.If ASGM applications are eliminated in 2030, supply will exceed demand under a high-level recovery scenario.

CIS
In the Commonwealth of Independent States (CIS) region, mercury demand and supply both decline throughout the entire period.The mercury supply-demand balance continues to exceed supply throughout almost the entire period, with a projected excess supply of 21 tons in 2030 and 7 tons in 2050.After 2030, supply and demand are generally in balance.

Middle East and Africa
In the Middle East and Africa, both supply and demand are estimated to decline after 2015.The supply-demand balance for mercury is projected to remain in a state of excess demand throughout the entire period, with a projected supply deficit of 207 tons in 2030 and 84 tons in 2050.Despite limited mercury supply sources in the region, mercury demand exists for ASGM, and even under a high-level recovery scenario, the supply of mercury is projected to be inadequate.This study suggests that a continued reliance on imports from outside the region may be unavoidable.

Global supply and demand trends for mercury
Estimates on global supply and demand trends for mercury by sectors are shown in Fig. 3.As mentioned in 3.1 and 3.2, during the mercury phase-out in progress under the Convention, primary mining in supply and ASGM and product uses in demand account for a large portion of total, while after the phase-out, mercury recovered from nonferrous sector and product uses become the main focus.As shown in Fig. 4 (business-as-usual scenario), throughout the entire period in this study, Europe, North America and Oceania, and the CIS show well-balanced supply and demand.China, on the other hand, sees mercury supply significantly exceed demand during the 2015-2025 period, which also has a significant impact on the global supply and demand.In the period around the entry into force of the Minamata Convention, the biggest shortfalls in mercury supply occurred in Asia (excluding China) and the Middle East and Africa in 2015, and in Latin America in 2000, showing a projected reduction in shortfalls thereafter.As indicated in 3.1, the total world exports were 1284 tons in 2015, and 504 tons in 2020.Taking into account those total world exports, it is assumed that the shortfall in supply in those regions was compensated by imports from other regions.
The global mercury supply-demand balance based on simple sums in each region is estimated to have been a state of 393 tons under-supplied in 2010 and 1086 tons over-supplied in 2015, and then projected to be over-supplied until 2028, switching back to under-supply afterward, with a supply shortage of 322 tons in 2035, and then an improvement in the supply-demand balance until 2050, when an excess supply of 59 tons is expected.For the period 2030-2045, it may be necessary to either utilize mercury stocks accumulated previously or implement more active mercury recovery from by-products compared to the BAU scenario.For the medium-level mercury recovery scenario, the mercury supply-demand balance turns to a surplus of mercury in 2040 and 2045, earlier than in the BAU scenario.
On the other hand, in the case of the high-level mercury recovery scenario shown in Fig. 5, global supply and demand continue to be in a state of excess supply from prior to 2015, and this study projects 568 tons in excess supply in 2030 and 907 tons in 2050.This suggests that if mercury recovery from by-products is actively promoted, surplus mercury stocks will grow steadily.
Regarding countries that do not have stabilization or final disposal facilities for mercury wastes, particularly in developing countries, the following directions for addressing this issue are recommended.That is, taking into account the small amount of metallic mercury disposed of, it is not realistic for each country to establish new treatment facilities for the final disposal of metallic mercury; therefore, it is suggested to establish some regional treatment facilities or export to existing treatment facilities in developed countries for recycling or disposal purposes in accordance with the Basel Convention and the Minamata Convention.Of course, for such a system to work, it is first necessary to understand the actual status of mercury waste disposal and establish an environmentally sound system for the collection of mercurycontaining or contaminated waste.

Prospects of excess mercury stocks
Future projections of cumulative excess mercury based on this study are shown in Fig. 6 (BAU scenario).With existing mercury stocks, including those derived from chloralkali mercury cells in Europe and the United States, the global cumulative excess mercury is estimated to have been 13,101 tons as of 2020.Subsequently, China's excess mercury increases, while Asia (excluding China), Latin America, and the Middle East and Africa regions continue to be in a supply deficit.Assuming that mercury continues to be supplied through the inter-regional mercury trade, the maximum cumulative amount of excess mercury is 17,390 tons in 2028, as a simple global total.Although there are export bans in Europe and the United States, surplus mercury in the rest of the world is sufficient to cover shortages in supply.After 2028, a gradual decrease in the amount of surplus mercury is observed, probably due to the impacts of a decrease in primary mining of mercury.Then, after reaching 14,326 tons in 2047, the cumulative excess mercury begins to increase again, resulting in a projected 14,439 tons of mercury in 2050.The EU and the United States were already stockpiling large amounts of mercury before the Convention entered into force, but in the future, mercury unintentionally generated as a by-product of industrial processes is expected to accumulate as excess mercury.
Depending on the mercury recovery scenario from nonferrous production, the amount of excess mercury is highly variable, generating 6.1 times more in the mediumlevel recovery scenario and 15.4 times more in the high-level recovery scenario than the amount of excess mercury in the BAU scenario.Relying on the level of recovery, the global cumulative amount of excess mercury in 2050 varies from 14,439 to 20,943 tons.Given that the GMA atmospheric inventory is expected to have an uncertainty of − 10% to + 30% [16], this study is expected to have the same level of uncertainty in base data.
As for previous studies, UNEP [51] projected 28,000-46,000 tons of excess mercury stock in 2050, and Sodeno et al. [10] projected the figure at 24,000-25,000 tons in 2050.The predictions of this study are as low as 31-60% of these predictions.The projected excess mercury are considered more confirming since this study reflects the situation today that the recovery of mercury from by-products is not being actively pursued.

Conclusion
Based on the global situation since the Minamata Convention entered into force in 2017, this study estimated global mercury demand and supply, by region, and projected the future accumulation of excess mercury.The global mercury trade after the Convention entered into force declined in the scale of both imports and exports.Among other sources, exported mercury includes mercury from mines in Mexico and Tajikistan and recycled mercury from Japan and Switzerland, while India and Bolivia remain the two largest importers as of 2020.UN statistics were used in this study to evaluate the mercury trade situation, but in some cases the reported values differ between the importing and exporting countries, and some values may involve misreporting, so any analysis requires a careful examination of the data.
After the Convention entered into force, mercury supply and demand were both on a downward trend.Mercury supply from primary mining did not decrease as much as initially expected, but a decrease was finally observed in China, the largest producer, starting in 2020.On the other hand, mercury recovery from by-products of nonferrous production was not implemented to the extent that had been anticipated.With regard to demand, chlor-alkali mercury cells were decommissioned in Europe ahead of the Convention's regulatory schedule, and demand reductions in the sector of mercury-added products also seemed to go more quickly than anticipated.The keys to future demand LaƟn America CIS Middle East + Africa Total trends will be the timing of the vinyl chloride monomer (VCM) phase-out and trends for the use of mercury in artisanal and small-scale gold mining (ASGM), both of which have particularly high uncertainty.
With regard to future projections for the mercury supply and demand balance, in the period covered by this study, supply and demand are generally in balance in Europe, North America and Oceania, and the CIS.On the other hand, Asia (excluding China), Latin America, and the Middle East and Africa are in short supply, while China is significantly oversupplied.This could be due to the large amount of mercury mining that continued even after the Convention entered into force, but since very few exports have been reported, it is unclear whether China's excess mercury is stocked domestically or excessively estimated in supply sectors.In terms of global totals, after the Convention enters into force, although the excess supply continues in the estimates and projections in this paper, a shortage begins around 2030 and the supply-demand balance improves after 2035, with a projected annual excess supply of 59 tons in 2050.In the high-level mercury recovery scenario in this study, the global supply-demand balance constantly has excess supply from 2015 onward and has 907 tons in excess supply projected for 2050.This suggests that if mercury recovery from by-products is actively promoted, surplus mercury stocks will continue to accumulate.
With regard to prospects for the generation of excess mercury, assuming that mercury continues to be supplied through inter-regional mercury trade, the maximum cumulative amount of excess mercury is projected at 17,390 tons in 2028 as a simple global total.After 2028, a gradual decrease in the amount of surplus mercury is observed, probably due to a decrease in primary mining of mercury.Then, after reaching 14,326 tons in 2047, the cumulative excess mercury stock begins to increase again, resulting in a projected 14,439 tons of mercury in 2050.Relying on the level of mercury recovery in the nonferrous sector, global cumulative stock of excess mercury is projected at 14,439-20,943 tons in 2050.
This study has several limitations in terms of uncertainties in base data and future scenarios associated with the mercury supply and demand estimates and prediction.Among uncertainties in base data associated with activity data, emission factors, and application of mercury control technologies, the crucial areas are considered to be ASGM operations, phase-out scenario in primary mining and VCM production, and mercury emission factors and future recovery rate in nonferrous sector; therefore, more reliable and higher quality data are required to be developed.
The EU and the United States were already stockpiling large amounts of mercury before the Convention entered into force, but in the future, mercury unintentionally generated as a by-product of industrial processes is expected accumulate as excess mercury.In the final disposal of elemental mercury, the EU and Japan have established disposal standards for permanent disposal, but challenges remain in providing final disposal sites, and the long-term stability of these treated mercury compounds will need to be verified further.In addition, mercury contained in by-products is often disposed in landfills as ordinary hazardous waste if it is not recovered from by-products, but the current status of disposal is not fully explained in national reports.Thresholds for mercury-contaminated waste are currently under consideration at the Conference of the Parties to the Minamata Convention, but contaminated waste containing high concentrations of mercury is at risk of leaching mercury, so the long-term stability of such waste will need to be examined.
A major finding of this study is that future estimates of cumulative stocks of excess mercury vary significantly depending on scenarios for mercury recovery from by-products.From this perspective, policies for mercury recovery from by-products should consider issues such as mercury recovery levels, the allowable extent of disposal as ordinary hazardous waste, and desired levels of intensive management (such as recovery and stabilization by sulfidation to control the risk of leaching), and policy makers must also be prepared for the anticipated generation of excess mercury.

Table 1
Time line for mercury reduction under the Minamata Convention VCM vinyl chloride monomer, SPME sodium or potassium methylate or ethylate a Not allowing the use of mercury 5 years after the COP has established that mercury-free process have become technically and economically feasible [19]file 1), 5% for Asia excl.China, CIS, Latin America (Profile 2), 1% for Africa, Middle East (Profile 3) Europe: 2015 data are average value from COWI[19]Japan: 2015 data are from national data

Table 3
Estimates of mercury recovery from nonferrous smelting in 2020 (zinc)

Table 4
Estimates of mercury recovery from nonferrous smelting in 2020 (lead)

Table 5
Estimates of mercury recovery from nonferrous smelting in 2020 (copper)

Table 6
Estimates of mercury recovery from large-scale gold production in 2020

Table 7
Exports of mercury by major exporters (2015-2021) *The exporting amount of Indonesia in 2017 was not avaiable, so it was calculated from trade value and average trade value per mercury in 2017 **The exporting data of Nigeria in 2021 seems peculiar and possible misreporting