- Email: [email protected]
Past and future cadmium emissions from municipal solid-waste incinerators in Japan for the assessment of cadmium control policy
Journal of Hazardous Materials 262 (2013) 741–747
Contents lists available at ScienceDirect
Journal of Hazardous Materials journal homepage: www.elsevier.com/locate/jhazmat
Past and future cadmium emissions from municipal solid-waste incinerators in Japan for the assessment of cadmium control policy Kyoko Ono ∗ National Institute of Advanced Industrial Science and Technology (AIST), Tsukuba, Ibaraki, Japan
h i g h l i g h t s • • • • •
Cd emissions from municipal solid waste incinerators were estimated for 1970–2030. Emissions peaked in 1973 (11.1 t) at levels ten times that in 2010 (1.2 t). In the 1970s, the main source was pigments, but after 2000, it was Ni-Cd batteries. The effects of two Cd control policies were compared. Banning Cd use reduced emissions more than intensive collection of batteries.
a r t i c l e
i n f o
Article history: Received 9 May 2013 Received in revised form 10 September 2013 Accepted 14 September 2013 Available online 24 September 2013 Keywords: Cadmium Emissions Municipal solid waste incinerator Substance flow Time trend
a b s t r a c t Cadmium (Cd) is a harmful pollutant emitted from municipal solid-waste incinerators (MSWIs). Cd stack emissions from MSWIs have been estimated between 1970 and 2030 in Japan. The aims of this study are to quantify emitted Cd by category and to analyze Cd control policies to reduce emissions. Emissions were estimated using a dynamic substance flow analysis (SFA) that took into account representative waste treatment flows and historical changes in emission factors. This work revealed that the emissions peaked in 1973 (11.1 t) and were ten times those in 2010 (1.2 t). Emission from MSWIs was two-thirds of that from non-ferrous smelting in 2010. The main Cd emission source was pigment use in the 1970s, but after 2000 it had shifted to nickel-cadmium (Ni-Cd) batteries. Future emissions were estimated for 2030. Compared to the business-as-usual scenario, an intensive collection of used Ni-Cd batteries and a ban on any future use of Ni-Cd batteries will reduce emissions by 0.09 and 0.31 t, respectively, in 2030. This approach enables us to identify the major Cd emission source from MSWIs, and to prioritize the possible Cd control policies. © 2013 Elsevier B.V. All rights reserved.
1. Introduction Cadmium (Cd) is a nonessential toxic heavy metal with widely known environmental and health risks. Bergbäck et al. [1] pointed out that the risk posed by Cd could increase as a result of its anthropogenic use. Dispersive Cd emissions from solid-waste treatment of Cd-containing products could lead to an increase in its environmental concentration, and thus the reduction of Cd stack emissions is an important issue. Toward that end, some of the policies of the European community include end-of-pipe measures, phase-outs of certain applications, and collection of used products [2].
∗ Permanent address: Research Institute of Science for Safety and Sustainability (RISS), National Institute of Advanced Industrial Science and Technology (AIST), 161 Onogawa, Tsukuba, Ibaraki 305-8569, Japan. Tel.: +81 29 861 4854; fax: +81 29 861 8411. E-mail address: [email protected] 0304-3894/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jhazmat.2013.09.033
Incineration is one of the most important measures by which municipal solid waste is disposed of. Municipal solid-waste incinerators (MSWIs) are recognized as a major potential emission source of Cd owing to its low boiling point. Risk assessments of MSWIs have been reported extensively, and their potential to increase environmental Cd concentrations has been discussed [3–6] based on measurements. Although Cd deposition to soil around MSWI might be occurring, the apparent increase in the soil around MSWIs has not been observed by environmental monitoring [3,7–11] because the Cd concentrations are stable at low level due to improved air pollution control system of MSWIs. It is too small to be detected by the measurements used for the environmental samples collected in those studies. In Japan, there were more than 1200 MSWIs in operation in 2010 [12], and there have been no reports of increases in metal concentrations around MSWIs, in ambient air, or in soil since the 1990s. The Japanese government has been reporting the emission inventory of Cd (into air, water, and landfill) under the pollutant
742
K. Ono / Journal of Hazardous Materials 262 (2013) 741–747
release and transfer register (PRTR) legislation [13]. Although this includes Cd emissions from MSWIs, the reported emissions were all zero from the 1200 MSWIs, owing to both the low concentration of Cd in the exhaust gas from MSWIs and the detection limit of the analytical equipment. To consider Cd regulation (e.g., a ban on Cd in consumer products), it is essential to attribute the contribution of Cd emissions from MSWIs to the increase in the environmental concentration. However, as mentioned above, the contribution of MSWIs to the environmental concentration cannot be determined using existing monitoring data. Here, another quantitative research method, i.e., substance flow analysis (SFA) combined with environmental emission factors, is applied to identify emissions. SFA has been used extensively in the industrialized world to support environmental policy analysis [2]. If we can establish appropriate boundaries for the time scale of the analysis, we can compare results derived for a variety of substance flow scenarios. The substance flow of Cd has been studied by static and dynamic SFA [14–16], but these studies have focused on resource conservation. For example, Matsuno et al. [15] reported in-use stocks of Cd as potential resources to be recovered. However, from the viewpoint of assessing environmental risk caused by Cd, a study using dynamic SFA is limited. Here, we show that a dynamic SFA is applicable to emission estimation from the viewpoint of health risk analysis, instead of assessments based on measured data. To the best of the author’s knowledge, there are only two previous studies in which environmental emissions were calculated from in-use stocks of hazardous heavy metals [17]. Fuse and Tsunemi [17] showed an estimate for metal (lead, silver, etc.) emissions derived from the use of lead solders and lead-free solders, but their analysis did not include other important use categories. Our focus is on Cd. We have analyzed the all-use category related to the potential input to MSWIs during a longer period (1970–2030). We are interested in estimating past and future Cd emissions from MSWIs in order to ascertain the use category that has had the biggest contribution to ambient air emissions. We then want to compare the stack emissions that would result from various regulation scenarios. A complete ban on Cd use in portable batteries or semiconductors in solar panels is challenging because of the benefits that Cd use provides. Thus, in order to propose better (more cost-effective) countermeasures for reducing Cd risk, quantitative identification of the origins and routes of Cd transfer is necessary. The objectives of this study are as follows: to estimate potential emissions of Cd from MSWIs over time, and to compare the following three regulation scenarios for controlling Cd use: business as usual (BAU), intensive collection of used nickel-cadmium (Ni-Cd) batteries, and banning Cd use for batteries. We assess these Cd controlling scenarios, the countermeasures of which will be imposed after 2015. 2. Methods and data
Ni-Cd batteries, and others, for which the amounts were derived from the available annual statistics [18]. Other than Ni-Cd batteries, the production was assumed to be equal to the domestic supply. Statistics were available from 1950 to 2010. After 2011, quantities were assumed to remain constant until 2030. Details of these data are given in the Supporting Information, Fig. S1. The import and export of Ni-Cd batteries were not negligible for either battery cells or end-use products containing Ni-Cd batteries [19], so the ratios of Ni-Cd battery cells sold domestically in Japan to the total national production of Ni-Cd battery cells, considering the imported and exported amounts of the cells, were calculated. Thus, the quantity of Ni-Cd battery cells sold domestically was calculated using the following formula. Ni-Cd battery cells sold domestically in year t [Cd-t] = (Pt − Et + It ) × Rcell,t , ×Rprod,t . Here,
Rcell,t = (Pcell,t − Ecell,t + Icell,t )/Pcell,t
and
Rprod,t =
(Pi,t − Ei,t + Ii,t ) ∗ ri,t / (P ∗ ri,t ) . i i,t is the amount of Cd metal used for Ni-Cd batteries in year t [Cd-t]. Et and It are the amounts of Cd metal (ingot) exported and imported in year t, respectively. Rcell,t is the ratio of Ni-Cd battery cells sold domestically to the total production in year t. Pcell,t , Ecell,t , and Icell,t are the quantities of Ni-Cd battery cells produced domestically, exported, and imported, respectively, in year t. Rprod,t is the indirect trade ratio in year t to correct for the import and export of end-use products containing Ni-Cd batteries. Pi,t , Ei,t , and Ii,t are the quantities of domestic production, export, and import of end-use products containing Ni-Cd batteries (i = home appliance, communication equipment, general merchandize), respectively. ri,t is the rate of application (installation) of Ni-Cd batteries. The amount of Cd used in the production of Ni-Cd batteries was available from statistical data [18]. Additionally, it was assumed that all imported Cd ingot was used in the production of Ni-Cd batteries [20]. The annual export of Cd ingots is one-thousandth that of its import [21], and thus the exported amount was assumed to be negligible. The Cd content of a Ni-Cd battery was assumed to be constant [21]. The number of Ni-Cd battery cells was obtained from the Japanese Trade Statistics [21]. Pi,t , Ei,t , and Ii,t were based on data from the Japanese Trade Statistics [20]. The end-use products, for which indirect trade ratios (Rprod,t ) were calculated, were selected according to Matsuno et al. [15], and they are as follows: home appliances (video camera, rechargeable shaver), communication equipment (cordless phone), and general merchandize (electric power tools), for which annual import and export data were available from 1996 to 2008 [21]. For these products, ri,t was calculated using the data provided by BiPRO [22]. Since a declining trend was observed in these ratios, the data of 1996 and 2008 were used to calculate the ratios and were then interpolated assuming a linear decrease. Details of the data and calculation are shown in the Supporting Information, Table S1. i Pt
2.1. Estimating the amount of Cd in waste 2.1.1. Overview We estimated the annual amount of Cd contained in waste using a dynamic SFA [15]. First, we estimated the domestic supplies of Cd by use categories. The dynamic SFA applied in this study included manufacturing, market stock during use, and discarded material for an arbitrary year. The amount of Cd was expressed as a collective mass for all of Japan. 2.1.2. Estimating domestic supplies of Cd by use categories The following use categories were considered: cadmium plating, alloys, pigments, stabilizers, cathode-ray tubes, rectifiers, catalysts,
2.1.3. Estimating the amount of stock and discarded products containing Cd The dynamic SFA estimates the amount of stocks, flows, or discarded materials contained in end-of-life products over time, based on the annual production and the product lifetime [23]. Details of this method and the calculation procedure have been shown in the Supporting Information. In this study, the average lifetime was treated using the categorized parameters shown in Table 1. Lifetimes were retrieved from an existing database [24] or by expert judgment. We also conducted a sensitivity analysis for the assumed lifetime of a Ni-Cd battery.
K. Ono / Journal of Hazardous Materials 262 (2013) 741–747
743
Table 1 Categorized average lifetime (%). Use category
Categorized average lifetime
Plating Alloy Pigment Stabilizers Cathode-ray tube Rectifier Catalyst Ni-Cd battery Others
0 (year)
2.5 (year)
7.5 (year)
15 (year)
0 0 50 25 0 0 0 0 50
0 0 50 25 0 0 100 0 50
50 50 0 25 100 100 0 50 0
50 50 0 25 0 0 0 50 0
2.2. Waste treatment Total input of Cd to MSWI was calculated considering the amount of discarded products and representative waste treatment flows. 2.2.1. Municipal and industrial waste, incineration, and landfill We identified the amount of municipal waste by considering the methods of garbage treatment. Some products are collected for material recycling after use. The media by which Cd is released are incineration and landfill. To estimate the amount of Cd released into the environment through MSWIs, the waste treatment method must be specified and the amount of Cd treated by each method must be estimated. First, we determined the waste treatment method by use category. Actual waste treatment methods were determined by interviews with municipalities, and a common waste treatment flow for all use categories was designed; this is shown in Fig. 1. The products disposed of as industrial waste were treated specifically as industrial waste. Therefore, they were assumed not to have been mixed with municipal waste at waste disposal or treatment facilities. The ratios of industrial and municipal waste are specified in Table 2. We assumed that plating, alloys, cathode-ray tubes, and rectifiers were used for industrial use only, and therefore Cd input to MSWIs from these product use categories should be negligible. In Japan, municipal solid waste is separated into combustibles and non-combustibles. Usually, Ni-Cd batteries are classified as non-combustibles. However, input of Cd into MSWIs occurs when garbage is misclassified. The separation of garbage into combustibles and non-combustibles is not always complete and perfect; and the ratio of contamination of combustible garbage by non-combustible garbage was 12% [25].
b%
2.2.2. Recovery of Ni-Cd batteries JBRC [26] reported that the collection rate of used Ni-Cd batteries has remained almost constant at 30% since 1990. In this study, 30% was used as the collection ratio of used products, and the other 70% was assumed to enter the incineration process or the landfill process (Fig. 1). Because this assumption had a large uncertainty, a sensitivity analysis for the collection rate was conducted. 2.3. Estimation of Cd released into air During incineration in an MSWI, some of the volatilized Cd becomes particles during a cooling process. These are collected by an exhaust-gas treatment system (EGTS) and the excess is released into the air. In contrast, non-volatilized Cd is collected as bottom ash and treated as solid waste. Here, the emission factor (EF) was defined as follows: EF = (Cd volatilization rate) × (1 − particle collection efficiency). It was assumed that the Cd volatilization rate was 90%, which was an average of the reported distribution ratios of metals in bottom ash and fly ash [27–31]. The particle collection efficiency was set according to the characteristics of the type of EGTS. Three types of EGTS were considered: baghouse filtration (BF), electrostatic precipitation (EP), and centrifugal dust collection [multi-cyclone (MC)]. The collection efficiencies of these were assumed to be 99%, 98%, and 80%, respectively [31,32], but a sensitivity analysis was conducted because a large uncertainty was expected. In Japan, regulations controlling dioxin emissions from MSWIs were introduced in 1997; these regulations came into effect in 2000. As part of these regulations, baghouse filters have been installed in place of electrostatic precipitators since 1997. Fig. 2 shows the historical trend of the ratio of MSWIs with different
Collection for recycling
Yes
a%
Resource collection ?
100-b % Waste 100-a % Industrial
It %
No
Combustible Yes
Pre-treatment
c%
10 0 -It %
Separation of waste No
Incineration
5%
Municipal
100-c % Noncombustible
70 %
Pre-treatment
95 %
Landfil
30 % Fig. 1. Schematic diagram of municipal waste treatment flow in Japan. Time trend of ratio of incineration, It , is shown in the Supporting Information, Table S2.
744
K. Ono / Journal of Hazardous Materials 262 (2013) 741–747
Table 2 Parameters for Cd in waste treatment flow. Use category
Ratio of municipal use (%)/industrial use (%) (a)
Collection ratio of used products for municipal waste (%) (b)
Ratio of waste classified as “combustible” in municipal waste (%) (c)
Plating Alloy Pigment Stabilizers Cathode-ray tube Rectifier Catalyst Ni-Cd battery Others
0/100 0/100 100/0 100/0 0/100 0/100 0/100 70/30 100/0
– – 0 0 – – – 30 0
– – 50 50 – – – 12a 12a
a The ratio of waste classified as “combustible” was based on [24]. Ref. [24] reported that people are sometimes unsuccessful in separating garbage into combustibles and non-combustibles, and that the ratio of contamination of non-combustible garbage into combustible garbage was 12%.
EGTSs, and a rapid change from EP to BF between 1997 and 2001 can be seen. For each MSWI, the garbage incineration capacity and type of EGTS were available from various databases [33,34]. The total Cd emission into the air from all MSWIs was calculated using the following formula: Total emission =
Input ×
i
G i × EFi . i
Gi
Here, Input is the total input estimated above, Gi is the garbage incineration capacity for MSWIi , and EFi is the emission factor of Cd for MSWIi .
The amount of Cd contained in discarded products in Japan (Fig. 3a) increased to 2700 t in 2000. This is consistent with the results of Matsuno et al. [15], although the peak year in this study was slightly later. Since 2000, the amount of Cd discarded has decreased. In 2010, the amount of Cd contained in discarded products was 1700 t, and 95% of that material was derived from Ni-Cd batteries. The value of the estimated total input (total mass in Japan) was validated by comparing it with data from other studies. The Cd inputs into MSWIs in the 1990s and in 2000 were estimated in three previous studies [29,35,36] by mass balance calculation using the data of Cd concentration of fly ash and bottom ash, and the annual generation for fly ash and bottom ash. In the 1990s, Cd inputs into MSWIs were estimated as 60 t [35] and 200 t [36]. In 2000, Jung
2.4. Sensitivity analysis For parameters obtained from the literature that were based on limited data (Table 3), sensitivity analyses were conducted. In addition to the default values, upper and lower limits were assumed, and then differences in the Cd emissions were calculated for all of these values. 3. Results and discussion 3.1. Amounts of Cd disposed of via incineration The calculated amounts of Cd discarded for each use category are shown in Fig. 3(a and b). Fig. 3a shows the amount of Cd discarded over time, and Fig. 3b shows the total input of Cd to MSWIs. In the 1970s, the main sources were pigments and stabilizers, but after 2000, Ni-Cd batteries made the largest contribution.
Fig. 2. Historical trend of the fraction of exhaust gas-treatment system (EGTS) installed in MSWIs in Japan. BF: baghouse filtration, EP: electrostatic precipitation, MC: multi-cyclone.
Fig. 3. (a) Time course of the total amount of Cd discarded by use categories. (b) Time course of the Cd input into MSWIs by use category (total mass in Japan).
K. Ono / Journal of Hazardous Materials 262 (2013) 741–747
745
Table 3 Sensitivity analysis. Parameters
Default
Lower limit
Upper limit
Domestic supply of end-use products containing Ni-Cd batteries Average lifetime: 5 years/10 years for Ni-Cd batteries [%] Collection rate of used Ni-Cd battery [%] Ratio of municipal use [%] Particle collection efficiency of BF [%] Particle collection efficiency of EP [%] Particle collection efficiency of MC [%]
Estimated considering time trenda 50/50 30 70 99 98 80
Domestic supply = 50% 0/100 20 50 98.5 97 75
Domestic supply = 100% 0/100 40 100 99.5 99 85
BF: baghouse filtration, EP: electrostatic precipitation, MC: multi-cyclone. a The method which was shown in Section 2.1.2.
et al. [29] estimated Cd inputs into MSWIs as 112 t. The estimated value in this study was approximately 150 and 180 t in 1990 and 2000, respectively. These estimated values did not show complete agreement with the values from the previous report, but the order of magnitude was the same (150 t is between 60 and 200 t) in 1990, and it was one-and-a-half times higher in 2000. When assessment is implemented using this result, one should note the possibility of this range of error.
This method will be applicable to other metals, such as lead and arsenic, because the concept of calculation is simple and the required parameters for estimation are common. Furthermore, the emission would also be applicable for the calculation of local risk around an MSWI because the ambient air concentration could be calculated using an atmospheric dispersion model with the emissions as the model’s input. We can estimate an increment of ambient air concentration around MSWIs.
3.2. Emissions from MSWIs
3.2.2. Emission by category In the 1970s and 1980s, pigments and stabilizers, which were used prior to and during the 1970s (for use of Cd products in 1950–2010, see the Supporting Information, Fig. S1), were the major contributors, and emissions derived from Ni-Cd batteries were comparatively small. After 2000, the main source of Cd shifted to Ni-Cd batteries. The domestic supply of Ni-Cd batteries peaked in 1996 (Fig. S1) and discarded Cd peaked in 2000 (Fig. 3a). EGTSs were changed from EP to BF, which led to an increase in the particle collection efficiency (see Table 3). This is why emission derived from Ni-Cd batteries after 2000 had been restricted; BF became the major EGTS. Considering a summation of emission from 1970 to 2010, pigments was the largest contributor (88.2 t), with the second and third being stabilizers (44.7 t) and Ni-Cd batteries (43.4 t), respectively.
3.2.1. Total emissions As Tian et al. pointed out [37], stack emission may be the most problematic aspect of a solid waste incineration process due to dispersive emissions. Fig. 4 illustrates the stack emissions over time estimated in this study. Emissions peaked in 1973 (11.1 t) and decreased drastically after that. This is because the amount of Cd used in pigments and stabilizers decreased. Emissions were approximately 1.2 t in 2010. In the 1970s, there was no restriction on Cd use, but emission control for waste water was introduced. This was likely the biggest reason for avoiding Cd use at the time. The Water Pollution Prevention Act was introduced in Japan in 1970 (it went into effect in 1971), and it gave incentives for manufacturers to replace Cd with other materials voluntarily. Identification of Cd emission from MSWIs will help prioritize regulation policies, such as emission reduction policies for industrial use or consumer use. According to the Japanese PRTR, air and water emissions in 2010 were 1.4 and 2.1 t, respectively [13]. This air emission was derived from industry, except for emission from MSWIs, in which non-ferrous smelting made the largest contribution. Emission from MSWIs estimated here was two-thirds that from industrial activity such as the non-ferrous industry. This method enabled us to compare emissions from MSWIs with other sources and to compare current/future emissions with past emissions.
3.3. Sensitivity analysis Fig. 5 shows the results of the quantitative sensitivity analysis. The upper and lower limits assumed here (Table 3) were extreme but plausible values. The most sensitive parameters were the particle collection efficiency for EP in 1990 and that for BF in 2010. In 1990, the most common EGTS was EP, and in 2010 it was BF, as shown in Fig. 2. This change was reflected by the change in sensitivity ranking. The difference in emissions caused by using different particle collection efficiencies for BF in 2010 was ±40%. The fraction of Ni-Cd batteries treated as municipal solid waste was the second most sensitive parameter in 2010. The collection rate of Ni-Cd batteries was less sensitive than the particle collection efficiency, despite the large quantity of waste in 2010. Qualitatively, there were some uncertainties in the parameters used in this study. Regarding end-use products, there was no evidence of the percentage of end-use products that were fabricated with Ni-Cd batteries, or of detailed import and export data, and thus realistic values for these quantities were estimated. In addition, the emissions derived from pigments may have been underestimated because even if the contribution of outside use (such as from roof paint) and direct emissions into the soil were larger, this route was not considered here. 3.4. Future scenario
Fig. 4. Time course of Cd emissions from MSWIs by use category (total mass in Japan).
Scenario analyses were conducted, in which a ban on Cd use and a stricter collection of used Cd were imposed. In advance of these
746
K. Ono / Journal of Hazardous Materials 262 (2013) 741–747
Fig. 5. Results of sensitivity analyses. Ratios (percent) of difference in Cd air emissions for lower, upper, and default values are shown. (a) In 1990, (b) in 2010. NiCd B: Ni-Cd battery.
analyses, we assumed that there will be no change in the production amounts of alloys, pigments, Ni-Cd batteries, and “others” between 2010 and 2030. This assumption might be conservative, which may lead to an overestimation because the municipal use of Ni-Cd batteries will have been decreasing in the future. The amount of Ni-Cd batteries for municipal use had been decreasing until 2010 due to a decline in end-use products that were fabricated with these batteries. We assumed that an exchange of EGTS in a MSWI would not occur between 2010 and 2030. This is because most MSWIs and their EGTSs were renewed after 1997 (see Fig. 1), and the lifespan of MSWIs is 30 years on average; therefore, exchange or renewal of EGTS is not likely. The effect of stricter controls on the use on Ni-Cd batteries, i.e., mandatory 30% collection after use, was analyzed. Fig. 6 shows future Cd emissions under three scenarios: scenario 1, business as usual (BAU); scenario 2, promoting voluntary collection and recycling of used Ni-Cd batteries to reach levels of 50% as of 2015; and scenario 3, ban on using Cd in Ni-Cd batteries from 2015 in addition to Ni-Cd battery collection (30%). Emission in 2030 in each scenario is shown in the box by use category.
In 2030, Cd emissions under scenarios 1, 2, and 3 will be 0.46, 0.34, and 0.055 t/year, respectively. Compared to the BAU scenario, banning the use of batteries and intensive collection of batteries will reduce emissions by 0.40 and 0.11 t in 2030, respectively. A complete ban on Ni-Cd batteries would reduce emissions, but emissions from MSWIs would not be zero immediately after enforcement of the ban. As mentioned above, Cd use is Japan is already limited. Municipal use of Ni-Cd batteries has actually decreased because other batteries (i.e., Ni-MH or Li-ion batteries) are commonly available and have replaced Ni-Cd batteries. However, the emissions originating from pigments, which were mainly used in 1970s, will still contribute in 2030 (box in Fig. 6). Cd is an inevitable common byproduct of zinc refining. From the viewpoint of effective use of a limited resource, intensive collection of used Cd as a closed system might be a better approach than a complete ban of Cd use. Needless to say, a restriction of dispersive use of Cd will be most important. This method will provide quantitative evidence for such discussions. Dynamic SFA allowed for delayed disposal or emissions, which were derived from past use of Cd, to be estimated. This study quantitatively demonstrated
Fig. 6. Cd emissions under 3 scenarios in which Ni-Cd batteries are restricted. Scenario 1, business as usual (BAU); scenario 2, promoting voluntary collection and recycling of used Ni-Cd batteries to reach levels of 50% from 2015; scenario 3, ban on using Cd in Ni-Cd batteries from 2015 in addition to Ni-Cd battery collection (30%). Emission in 2030 by each scenario was shown in the box by use category.
K. Ono / Journal of Hazardous Materials 262 (2013) 741–747
the lag time and the amount by which the emissions would be reduced because of the countermeasures. Scenario analysis can show quantitative results regarding the differences in Cd emissions from MSWIs with and without controlling the use of the metal. Environmental monitoring around a MSWI cannot provide information on the use category of Cd emissions. 4. Summary In this study, Cd emissions from municipal solid waste incinerators were estimated for 1970–2030. The method enabled us to identify the major Cd emission source from MSWIs. Emissions peaked in 1973 at levels ten times that in 2010. In the 1970s, the main source was pigments, but after 2000 it was Ni-Cd batteries. We assessed Cd control countermeasures, such as intensive collection of used Ni-Cd batteries or a ban on the future use of Ni-Cd batteries. Banning Cd use reduced emissions more than the intensive collection of batteries. This method is applicable not only to Cd but to other metals as well because the concept of calculation is simple and the required parameters for the estimation are common. Acknowledgement The author is grateful to Mr. Hideki Wada of the Sustainable System Design Institute, who developed the SFA tool for this research. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.jhazmat. 2013.09.033. References [1] B. Bergbäck, S. Anderberg, U. Lohm, Accumulated environmental impact: the case of cadmium in Sweden, Science of the Total Environment 145 (1994) 13–28. [2] E. van der Voet, L. van Egmond, R. Kleijn, G. Huppes, Cadmium in the European Community: a policy-oriented analysis, Waste Management and Research 12 (1994) 507–526. [3] C.W. Hu, M.R. Chao, K.Y. Wu, G.P. Chang-Chien, W.J. Lee, L.W. Chang, W.S. Lee, Characterization of multiple airborne particulate metal in the surroundings of a municipal waste incinerator in Taiwan, Atmospheric Environment 37 (2003) 2845–2852. [4] D. Rimmer, C. Vizard, T. Pless-Mulloli, I. Singleton, V. Air, Z. Keatinge, Metal contamination of urban soils in the vicinity of a municipal waste incinerator: one source among many, Science of the Total Environment 356 (2006) 207–216. [5] M. Sakata, M. Kurata, N. Tanaka, Estimating contribution from municipal solid waste incineration to trace metal concentrations in Japanese urban atmosphere using lead as a marker element, Geochemical Journal 34 (2000) 23–32. [6] F. Zhang, S. Yamasaki, M. Nanzyo, K. Kimura, Evaluation of cadmium and other metal losses from various municipal wastes during incineration disposal, Environmental Pollution 115 (2001) 253–260. [7] M. Schuhmacher, M. Meneses, S. Granero, J. Llobet, J. Domingo, Trace element pollution of soils collected near a municipal solid waste incinerator: human health risk, Bulletin of Environment Contamination and Toxicology 59 (1997) 861–867. [8] J. Llobet, M. Schuhmacher, J. Domingo, Spatial distribution and temporal variation of metals in the vicinity of a municipal solid waste incinerator after a modernization of the flue gas cleaning systems of the facility, Science of the Total Environment 284 (2002) 205–214. [9] F. Cangialosi, G. Intini, L. Liberti, M. Notarnicola, P. Stellacci, Health risk assessment of air emissions from a municipal solid waste incineration plant – a case study, Waste Management 28 (2008) 885–895. [10] M. Meneses, J. Llobet, S. Granero, M. Schuhmacher, J. Domingo, Monitoring metals in the vicinity of a municipal waste incinerator: temporal variation in soils and vegetation, Science of the Total Environment 226 (1999) 157–164. [11] L. Morselli, M. Bartoli, B. Brusori, F. Passarini, Application of an integrated environmental monitoring system to an incineration plant, Science of the Total Environment 289 (2002) 177–188.
747
[12] Ministry of Environment (MOE), Annual Report on Waste Treatment in Japan, 2010 http://www.env.go.jp/recycle/waste tech/ippan/h22/data/ FY2010, disposal.pdf (in Japanese, accessed 14.03.13). [13] Ministry of Environment, Japan, PRTR Data Page, 2013 http://www2.env. go.jp/chemi/prtr/prtrinfo/e-index.html (accessed 14.03.13). [14] A. Lindqvist, F. von Malmbor, What can we learn from local substance flow analyses? The review of cadmium flows in Swedish municipalities, Journal of Cleaner Production 12 (2004) 909–918. [15] Y. Matsuno, T. Hur, V. Fthenakis, Dynamic modeling of cadmium substance flow with zinc and steel demand in Japan, Resources, Conservation and Recycling 61 (2012) 83–90. [16] K. Mason-Jones, H. von Blottnitz, Flows and fates of nickel-cadmium batteries in the City of Cape Town, Minerals Engineering 23 (2010) 211–218. [17] M. Fuse, K. Tsunemi, Assessment of the effects of the Japanese shift to lead-free solders and its impact on material substitution and environmental emissions by dynamic material, Science of the Total Environment 438 (2012) 49–58. [18] Japan Mining Industry Association, Bulletin of Japan Mining Industry Association, Tokyo. [19] M. Asari, S. Sakai, Battery recycling and flow analysis of NiCd and Li-ion batteries in Japan, in: The 7th International Conference on Waste Management and Technology, 2012. [20] Clean Japan Center, Research on the way of recycling-oriented economic system in the non-ferrous metal materials – research on the evaluation of the social system for waste reduction, Project supported by Japan Keirin Association Report FY 1998, 1999 (in Japanese). [21] Ministry of Finance (MOF), Trade Statistics of Japan, 2013 http://www.customs. go.jp/toukei/info/index e.htm (accessed 14.03.13). [22] BiPRO, Umweltbundesamt, Exemption for the use of cadmium in portable batteries and accumulators intended for the use in cordless power tools in the context of the Batteries Directive 2006/66/EC Final Report (revised), consortium ESWI (Expert team to Support Waste Implementation). 26 January 2010, 2006. [23] H. Hatayama, I. Daigo, Y. Matsuno, Y. Adachi, Outlook of the world steel cycle based on the stock and flow dynamics, Environmental Science and Technology 44 (2010) 6457–6463. [24] National Institute for Environmental Studies, Lifespan Database for Vehicles, Equipment, and Structures, 2013 http://www.nies.go.jp/lifespan/ isic search.php (accessed 14.03.13). [25] N. Tanigawa, T. Oikawa, K. Mimori, S. Awa, Annual Report of the Tokyo Metropolitan Research Institute for Environmental Protection, No. 28, 1998, pp. 65–68 (in Japanese). [26] Japan Portable Rechargeable Battery Recycling Center, Statistics of Recovery of Secondary Batteries, 2013 http://www.jbrc.net/hp/contents/recycle/ index.html (accessed 14.03.13). [27] H. Belevi, H. Moench, Factors determining the element behavior in municipal solid waste incinerator. 1. Field studies, Environmental Science and Technology 34 (2000) 2501–2506. [28] L.S. Morf, P.H. Brunner, S. Spaun, Effect of operating conditions and input variations on the partitioning of metals in a municipal solid waste incinerator, Waste Management and Research 18 (2000) 4–15. [29] C.H. Jung, T. Matsuto, N. Tanaka, T. Okada, Metal distribution in incineration residues of municipal solid waste (MSW) in Japan, Waste Management 24 (4) (2004) 381–391. [30] M. Lemann, R. Walder, A. Schwyn, Heavy metals in municipal solid waste incineration residues, Journal of Power Sources 57 (1995) 55–59. [31] K. Nakamura, S. Kinoshita, H. Takatsuki, The origin and behavior of lead, cadmium and antimony in MSW incinerator, Waste Management 16 (1996) 509–517. [32] A. Kida, S. Sakai, S. Sibakawa, A. Matsumoto, Heavy metals emission from municipal solid waste incineration system retrofitted for dioxins control, JEC 13 (2003) 51–67 (Abstract in English). [33] Ministry of Environment (MOE), Survey on Municipal Solid Waste Treatment, 2002 http://www.env.go.jp/recycle/waste tech/ippan/h14/data/seibi/ facility/01.xls (accessed 14.03.13). [34] Ministry of Environment (MOE), Report on Dioxin Concentration in Exhaust Gas from Municipal Solid Waste Incinerators http://www.env.go.jp/recycle/ dioxin/ippan/kekka11.html (accessed 14.03.13). [35] S. Yanagi, R. Sugawara, Y. Saito, The survey of inorganic salts in seeping water from controlled waste landfills and municipal solid waste incineration ashes, in: Annual Report of Miyagi Prefectural Institute of Public Health and Environment No. 24, 2006, pp. 79–83 (in Japanese). [36] Y. Suzuki, H. Takatsuki, S. Sakai, An effect of products recycling upon lead and cadmium flow in Japan, in: Proceedings of the 3rd Annual meeting for Society for Environmental Economics and Policy Studies, 1998, pp. 89–90 (in Japanese). [37] H. Tian, J. Gao, J. Hao, L. Lu, C. Zhu, P. Qiu, Atmospheric pollution problems and control proposals associated with solid waste management in China: a review, Journal of Hazardous Materials 252–253 (252) (2013) 142–150.
Contents lists available at ScienceDirect
Journal of Hazardous Materials journal homepage: www.elsevier.com/locate/jhazmat
Past and future cadmium emissions from municipal solid-waste incinerators in Japan for the assessment of cadmium control policy Kyoko Ono ∗ National Institute of Advanced Industrial Science and Technology (AIST), Tsukuba, Ibaraki, Japan
h i g h l i g h t s • • • • •
Cd emissions from municipal solid waste incinerators were estimated for 1970–2030. Emissions peaked in 1973 (11.1 t) at levels ten times that in 2010 (1.2 t). In the 1970s, the main source was pigments, but after 2000, it was Ni-Cd batteries. The effects of two Cd control policies were compared. Banning Cd use reduced emissions more than intensive collection of batteries.
a r t i c l e
i n f o
Article history: Received 9 May 2013 Received in revised form 10 September 2013 Accepted 14 September 2013 Available online 24 September 2013 Keywords: Cadmium Emissions Municipal solid waste incinerator Substance flow Time trend
a b s t r a c t Cadmium (Cd) is a harmful pollutant emitted from municipal solid-waste incinerators (MSWIs). Cd stack emissions from MSWIs have been estimated between 1970 and 2030 in Japan. The aims of this study are to quantify emitted Cd by category and to analyze Cd control policies to reduce emissions. Emissions were estimated using a dynamic substance flow analysis (SFA) that took into account representative waste treatment flows and historical changes in emission factors. This work revealed that the emissions peaked in 1973 (11.1 t) and were ten times those in 2010 (1.2 t). Emission from MSWIs was two-thirds of that from non-ferrous smelting in 2010. The main Cd emission source was pigment use in the 1970s, but after 2000 it had shifted to nickel-cadmium (Ni-Cd) batteries. Future emissions were estimated for 2030. Compared to the business-as-usual scenario, an intensive collection of used Ni-Cd batteries and a ban on any future use of Ni-Cd batteries will reduce emissions by 0.09 and 0.31 t, respectively, in 2030. This approach enables us to identify the major Cd emission source from MSWIs, and to prioritize the possible Cd control policies. © 2013 Elsevier B.V. All rights reserved.
1. Introduction Cadmium (Cd) is a nonessential toxic heavy metal with widely known environmental and health risks. Bergbäck et al. [1] pointed out that the risk posed by Cd could increase as a result of its anthropogenic use. Dispersive Cd emissions from solid-waste treatment of Cd-containing products could lead to an increase in its environmental concentration, and thus the reduction of Cd stack emissions is an important issue. Toward that end, some of the policies of the European community include end-of-pipe measures, phase-outs of certain applications, and collection of used products [2].
∗ Permanent address: Research Institute of Science for Safety and Sustainability (RISS), National Institute of Advanced Industrial Science and Technology (AIST), 161 Onogawa, Tsukuba, Ibaraki 305-8569, Japan. Tel.: +81 29 861 4854; fax: +81 29 861 8411. E-mail address: [email protected] 0304-3894/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jhazmat.2013.09.033
Incineration is one of the most important measures by which municipal solid waste is disposed of. Municipal solid-waste incinerators (MSWIs) are recognized as a major potential emission source of Cd owing to its low boiling point. Risk assessments of MSWIs have been reported extensively, and their potential to increase environmental Cd concentrations has been discussed [3–6] based on measurements. Although Cd deposition to soil around MSWI might be occurring, the apparent increase in the soil around MSWIs has not been observed by environmental monitoring [3,7–11] because the Cd concentrations are stable at low level due to improved air pollution control system of MSWIs. It is too small to be detected by the measurements used for the environmental samples collected in those studies. In Japan, there were more than 1200 MSWIs in operation in 2010 [12], and there have been no reports of increases in metal concentrations around MSWIs, in ambient air, or in soil since the 1990s. The Japanese government has been reporting the emission inventory of Cd (into air, water, and landfill) under the pollutant
742
K. Ono / Journal of Hazardous Materials 262 (2013) 741–747
release and transfer register (PRTR) legislation [13]. Although this includes Cd emissions from MSWIs, the reported emissions were all zero from the 1200 MSWIs, owing to both the low concentration of Cd in the exhaust gas from MSWIs and the detection limit of the analytical equipment. To consider Cd regulation (e.g., a ban on Cd in consumer products), it is essential to attribute the contribution of Cd emissions from MSWIs to the increase in the environmental concentration. However, as mentioned above, the contribution of MSWIs to the environmental concentration cannot be determined using existing monitoring data. Here, another quantitative research method, i.e., substance flow analysis (SFA) combined with environmental emission factors, is applied to identify emissions. SFA has been used extensively in the industrialized world to support environmental policy analysis [2]. If we can establish appropriate boundaries for the time scale of the analysis, we can compare results derived for a variety of substance flow scenarios. The substance flow of Cd has been studied by static and dynamic SFA [14–16], but these studies have focused on resource conservation. For example, Matsuno et al. [15] reported in-use stocks of Cd as potential resources to be recovered. However, from the viewpoint of assessing environmental risk caused by Cd, a study using dynamic SFA is limited. Here, we show that a dynamic SFA is applicable to emission estimation from the viewpoint of health risk analysis, instead of assessments based on measured data. To the best of the author’s knowledge, there are only two previous studies in which environmental emissions were calculated from in-use stocks of hazardous heavy metals [17]. Fuse and Tsunemi [17] showed an estimate for metal (lead, silver, etc.) emissions derived from the use of lead solders and lead-free solders, but their analysis did not include other important use categories. Our focus is on Cd. We have analyzed the all-use category related to the potential input to MSWIs during a longer period (1970–2030). We are interested in estimating past and future Cd emissions from MSWIs in order to ascertain the use category that has had the biggest contribution to ambient air emissions. We then want to compare the stack emissions that would result from various regulation scenarios. A complete ban on Cd use in portable batteries or semiconductors in solar panels is challenging because of the benefits that Cd use provides. Thus, in order to propose better (more cost-effective) countermeasures for reducing Cd risk, quantitative identification of the origins and routes of Cd transfer is necessary. The objectives of this study are as follows: to estimate potential emissions of Cd from MSWIs over time, and to compare the following three regulation scenarios for controlling Cd use: business as usual (BAU), intensive collection of used nickel-cadmium (Ni-Cd) batteries, and banning Cd use for batteries. We assess these Cd controlling scenarios, the countermeasures of which will be imposed after 2015. 2. Methods and data
Ni-Cd batteries, and others, for which the amounts were derived from the available annual statistics [18]. Other than Ni-Cd batteries, the production was assumed to be equal to the domestic supply. Statistics were available from 1950 to 2010. After 2011, quantities were assumed to remain constant until 2030. Details of these data are given in the Supporting Information, Fig. S1. The import and export of Ni-Cd batteries were not negligible for either battery cells or end-use products containing Ni-Cd batteries [19], so the ratios of Ni-Cd battery cells sold domestically in Japan to the total national production of Ni-Cd battery cells, considering the imported and exported amounts of the cells, were calculated. Thus, the quantity of Ni-Cd battery cells sold domestically was calculated using the following formula. Ni-Cd battery cells sold domestically in year t [Cd-t] = (Pt − Et + It ) × Rcell,t , ×Rprod,t . Here,
Rcell,t = (Pcell,t − Ecell,t + Icell,t )/Pcell,t
and
Rprod,t =
(Pi,t − Ei,t + Ii,t ) ∗ ri,t / (P ∗ ri,t ) . i i,t is the amount of Cd metal used for Ni-Cd batteries in year t [Cd-t]. Et and It are the amounts of Cd metal (ingot) exported and imported in year t, respectively. Rcell,t is the ratio of Ni-Cd battery cells sold domestically to the total production in year t. Pcell,t , Ecell,t , and Icell,t are the quantities of Ni-Cd battery cells produced domestically, exported, and imported, respectively, in year t. Rprod,t is the indirect trade ratio in year t to correct for the import and export of end-use products containing Ni-Cd batteries. Pi,t , Ei,t , and Ii,t are the quantities of domestic production, export, and import of end-use products containing Ni-Cd batteries (i = home appliance, communication equipment, general merchandize), respectively. ri,t is the rate of application (installation) of Ni-Cd batteries. The amount of Cd used in the production of Ni-Cd batteries was available from statistical data [18]. Additionally, it was assumed that all imported Cd ingot was used in the production of Ni-Cd batteries [20]. The annual export of Cd ingots is one-thousandth that of its import [21], and thus the exported amount was assumed to be negligible. The Cd content of a Ni-Cd battery was assumed to be constant [21]. The number of Ni-Cd battery cells was obtained from the Japanese Trade Statistics [21]. Pi,t , Ei,t , and Ii,t were based on data from the Japanese Trade Statistics [20]. The end-use products, for which indirect trade ratios (Rprod,t ) were calculated, were selected according to Matsuno et al. [15], and they are as follows: home appliances (video camera, rechargeable shaver), communication equipment (cordless phone), and general merchandize (electric power tools), for which annual import and export data were available from 1996 to 2008 [21]. For these products, ri,t was calculated using the data provided by BiPRO [22]. Since a declining trend was observed in these ratios, the data of 1996 and 2008 were used to calculate the ratios and were then interpolated assuming a linear decrease. Details of the data and calculation are shown in the Supporting Information, Table S1. i Pt
2.1. Estimating the amount of Cd in waste 2.1.1. Overview We estimated the annual amount of Cd contained in waste using a dynamic SFA [15]. First, we estimated the domestic supplies of Cd by use categories. The dynamic SFA applied in this study included manufacturing, market stock during use, and discarded material for an arbitrary year. The amount of Cd was expressed as a collective mass for all of Japan. 2.1.2. Estimating domestic supplies of Cd by use categories The following use categories were considered: cadmium plating, alloys, pigments, stabilizers, cathode-ray tubes, rectifiers, catalysts,
2.1.3. Estimating the amount of stock and discarded products containing Cd The dynamic SFA estimates the amount of stocks, flows, or discarded materials contained in end-of-life products over time, based on the annual production and the product lifetime [23]. Details of this method and the calculation procedure have been shown in the Supporting Information. In this study, the average lifetime was treated using the categorized parameters shown in Table 1. Lifetimes were retrieved from an existing database [24] or by expert judgment. We also conducted a sensitivity analysis for the assumed lifetime of a Ni-Cd battery.
K. Ono / Journal of Hazardous Materials 262 (2013) 741–747
743
Table 1 Categorized average lifetime (%). Use category
Categorized average lifetime
Plating Alloy Pigment Stabilizers Cathode-ray tube Rectifier Catalyst Ni-Cd battery Others
0 (year)
2.5 (year)
7.5 (year)
15 (year)
0 0 50 25 0 0 0 0 50
0 0 50 25 0 0 100 0 50
50 50 0 25 100 100 0 50 0
50 50 0 25 0 0 0 50 0
2.2. Waste treatment Total input of Cd to MSWI was calculated considering the amount of discarded products and representative waste treatment flows. 2.2.1. Municipal and industrial waste, incineration, and landfill We identified the amount of municipal waste by considering the methods of garbage treatment. Some products are collected for material recycling after use. The media by which Cd is released are incineration and landfill. To estimate the amount of Cd released into the environment through MSWIs, the waste treatment method must be specified and the amount of Cd treated by each method must be estimated. First, we determined the waste treatment method by use category. Actual waste treatment methods were determined by interviews with municipalities, and a common waste treatment flow for all use categories was designed; this is shown in Fig. 1. The products disposed of as industrial waste were treated specifically as industrial waste. Therefore, they were assumed not to have been mixed with municipal waste at waste disposal or treatment facilities. The ratios of industrial and municipal waste are specified in Table 2. We assumed that plating, alloys, cathode-ray tubes, and rectifiers were used for industrial use only, and therefore Cd input to MSWIs from these product use categories should be negligible. In Japan, municipal solid waste is separated into combustibles and non-combustibles. Usually, Ni-Cd batteries are classified as non-combustibles. However, input of Cd into MSWIs occurs when garbage is misclassified. The separation of garbage into combustibles and non-combustibles is not always complete and perfect; and the ratio of contamination of combustible garbage by non-combustible garbage was 12% [25].
b%
2.2.2. Recovery of Ni-Cd batteries JBRC [26] reported that the collection rate of used Ni-Cd batteries has remained almost constant at 30% since 1990. In this study, 30% was used as the collection ratio of used products, and the other 70% was assumed to enter the incineration process or the landfill process (Fig. 1). Because this assumption had a large uncertainty, a sensitivity analysis for the collection rate was conducted. 2.3. Estimation of Cd released into air During incineration in an MSWI, some of the volatilized Cd becomes particles during a cooling process. These are collected by an exhaust-gas treatment system (EGTS) and the excess is released into the air. In contrast, non-volatilized Cd is collected as bottom ash and treated as solid waste. Here, the emission factor (EF) was defined as follows: EF = (Cd volatilization rate) × (1 − particle collection efficiency). It was assumed that the Cd volatilization rate was 90%, which was an average of the reported distribution ratios of metals in bottom ash and fly ash [27–31]. The particle collection efficiency was set according to the characteristics of the type of EGTS. Three types of EGTS were considered: baghouse filtration (BF), electrostatic precipitation (EP), and centrifugal dust collection [multi-cyclone (MC)]. The collection efficiencies of these were assumed to be 99%, 98%, and 80%, respectively [31,32], but a sensitivity analysis was conducted because a large uncertainty was expected. In Japan, regulations controlling dioxin emissions from MSWIs were introduced in 1997; these regulations came into effect in 2000. As part of these regulations, baghouse filters have been installed in place of electrostatic precipitators since 1997. Fig. 2 shows the historical trend of the ratio of MSWIs with different
Collection for recycling
Yes
a%
Resource collection ?
100-b % Waste 100-a % Industrial
It %
No
Combustible Yes
Pre-treatment
c%
10 0 -It %
Separation of waste No
Incineration
5%
Municipal
100-c % Noncombustible
70 %
Pre-treatment
95 %
Landfil
30 % Fig. 1. Schematic diagram of municipal waste treatment flow in Japan. Time trend of ratio of incineration, It , is shown in the Supporting Information, Table S2.
744
K. Ono / Journal of Hazardous Materials 262 (2013) 741–747
Table 2 Parameters for Cd in waste treatment flow. Use category
Ratio of municipal use (%)/industrial use (%) (a)
Collection ratio of used products for municipal waste (%) (b)
Ratio of waste classified as “combustible” in municipal waste (%) (c)
Plating Alloy Pigment Stabilizers Cathode-ray tube Rectifier Catalyst Ni-Cd battery Others
0/100 0/100 100/0 100/0 0/100 0/100 0/100 70/30 100/0
– – 0 0 – – – 30 0
– – 50 50 – – – 12a 12a
a The ratio of waste classified as “combustible” was based on [24]. Ref. [24] reported that people are sometimes unsuccessful in separating garbage into combustibles and non-combustibles, and that the ratio of contamination of non-combustible garbage into combustible garbage was 12%.
EGTSs, and a rapid change from EP to BF between 1997 and 2001 can be seen. For each MSWI, the garbage incineration capacity and type of EGTS were available from various databases [33,34]. The total Cd emission into the air from all MSWIs was calculated using the following formula: Total emission =
Input ×
i
G i × EFi . i
Gi
Here, Input is the total input estimated above, Gi is the garbage incineration capacity for MSWIi , and EFi is the emission factor of Cd for MSWIi .
The amount of Cd contained in discarded products in Japan (Fig. 3a) increased to 2700 t in 2000. This is consistent with the results of Matsuno et al. [15], although the peak year in this study was slightly later. Since 2000, the amount of Cd discarded has decreased. In 2010, the amount of Cd contained in discarded products was 1700 t, and 95% of that material was derived from Ni-Cd batteries. The value of the estimated total input (total mass in Japan) was validated by comparing it with data from other studies. The Cd inputs into MSWIs in the 1990s and in 2000 were estimated in three previous studies [29,35,36] by mass balance calculation using the data of Cd concentration of fly ash and bottom ash, and the annual generation for fly ash and bottom ash. In the 1990s, Cd inputs into MSWIs were estimated as 60 t [35] and 200 t [36]. In 2000, Jung
2.4. Sensitivity analysis For parameters obtained from the literature that were based on limited data (Table 3), sensitivity analyses were conducted. In addition to the default values, upper and lower limits were assumed, and then differences in the Cd emissions were calculated for all of these values. 3. Results and discussion 3.1. Amounts of Cd disposed of via incineration The calculated amounts of Cd discarded for each use category are shown in Fig. 3(a and b). Fig. 3a shows the amount of Cd discarded over time, and Fig. 3b shows the total input of Cd to MSWIs. In the 1970s, the main sources were pigments and stabilizers, but after 2000, Ni-Cd batteries made the largest contribution.
Fig. 2. Historical trend of the fraction of exhaust gas-treatment system (EGTS) installed in MSWIs in Japan. BF: baghouse filtration, EP: electrostatic precipitation, MC: multi-cyclone.
Fig. 3. (a) Time course of the total amount of Cd discarded by use categories. (b) Time course of the Cd input into MSWIs by use category (total mass in Japan).
K. Ono / Journal of Hazardous Materials 262 (2013) 741–747
745
Table 3 Sensitivity analysis. Parameters
Default
Lower limit
Upper limit
Domestic supply of end-use products containing Ni-Cd batteries Average lifetime: 5 years/10 years for Ni-Cd batteries [%] Collection rate of used Ni-Cd battery [%] Ratio of municipal use [%] Particle collection efficiency of BF [%] Particle collection efficiency of EP [%] Particle collection efficiency of MC [%]
Estimated considering time trenda 50/50 30 70 99 98 80
Domestic supply = 50% 0/100 20 50 98.5 97 75
Domestic supply = 100% 0/100 40 100 99.5 99 85
BF: baghouse filtration, EP: electrostatic precipitation, MC: multi-cyclone. a The method which was shown in Section 2.1.2.
et al. [29] estimated Cd inputs into MSWIs as 112 t. The estimated value in this study was approximately 150 and 180 t in 1990 and 2000, respectively. These estimated values did not show complete agreement with the values from the previous report, but the order of magnitude was the same (150 t is between 60 and 200 t) in 1990, and it was one-and-a-half times higher in 2000. When assessment is implemented using this result, one should note the possibility of this range of error.
This method will be applicable to other metals, such as lead and arsenic, because the concept of calculation is simple and the required parameters for estimation are common. Furthermore, the emission would also be applicable for the calculation of local risk around an MSWI because the ambient air concentration could be calculated using an atmospheric dispersion model with the emissions as the model’s input. We can estimate an increment of ambient air concentration around MSWIs.
3.2. Emissions from MSWIs
3.2.2. Emission by category In the 1970s and 1980s, pigments and stabilizers, which were used prior to and during the 1970s (for use of Cd products in 1950–2010, see the Supporting Information, Fig. S1), were the major contributors, and emissions derived from Ni-Cd batteries were comparatively small. After 2000, the main source of Cd shifted to Ni-Cd batteries. The domestic supply of Ni-Cd batteries peaked in 1996 (Fig. S1) and discarded Cd peaked in 2000 (Fig. 3a). EGTSs were changed from EP to BF, which led to an increase in the particle collection efficiency (see Table 3). This is why emission derived from Ni-Cd batteries after 2000 had been restricted; BF became the major EGTS. Considering a summation of emission from 1970 to 2010, pigments was the largest contributor (88.2 t), with the second and third being stabilizers (44.7 t) and Ni-Cd batteries (43.4 t), respectively.
3.2.1. Total emissions As Tian et al. pointed out [37], stack emission may be the most problematic aspect of a solid waste incineration process due to dispersive emissions. Fig. 4 illustrates the stack emissions over time estimated in this study. Emissions peaked in 1973 (11.1 t) and decreased drastically after that. This is because the amount of Cd used in pigments and stabilizers decreased. Emissions were approximately 1.2 t in 2010. In the 1970s, there was no restriction on Cd use, but emission control for waste water was introduced. This was likely the biggest reason for avoiding Cd use at the time. The Water Pollution Prevention Act was introduced in Japan in 1970 (it went into effect in 1971), and it gave incentives for manufacturers to replace Cd with other materials voluntarily. Identification of Cd emission from MSWIs will help prioritize regulation policies, such as emission reduction policies for industrial use or consumer use. According to the Japanese PRTR, air and water emissions in 2010 were 1.4 and 2.1 t, respectively [13]. This air emission was derived from industry, except for emission from MSWIs, in which non-ferrous smelting made the largest contribution. Emission from MSWIs estimated here was two-thirds that from industrial activity such as the non-ferrous industry. This method enabled us to compare emissions from MSWIs with other sources and to compare current/future emissions with past emissions.
3.3. Sensitivity analysis Fig. 5 shows the results of the quantitative sensitivity analysis. The upper and lower limits assumed here (Table 3) were extreme but plausible values. The most sensitive parameters were the particle collection efficiency for EP in 1990 and that for BF in 2010. In 1990, the most common EGTS was EP, and in 2010 it was BF, as shown in Fig. 2. This change was reflected by the change in sensitivity ranking. The difference in emissions caused by using different particle collection efficiencies for BF in 2010 was ±40%. The fraction of Ni-Cd batteries treated as municipal solid waste was the second most sensitive parameter in 2010. The collection rate of Ni-Cd batteries was less sensitive than the particle collection efficiency, despite the large quantity of waste in 2010. Qualitatively, there were some uncertainties in the parameters used in this study. Regarding end-use products, there was no evidence of the percentage of end-use products that were fabricated with Ni-Cd batteries, or of detailed import and export data, and thus realistic values for these quantities were estimated. In addition, the emissions derived from pigments may have been underestimated because even if the contribution of outside use (such as from roof paint) and direct emissions into the soil were larger, this route was not considered here. 3.4. Future scenario
Fig. 4. Time course of Cd emissions from MSWIs by use category (total mass in Japan).
Scenario analyses were conducted, in which a ban on Cd use and a stricter collection of used Cd were imposed. In advance of these
746
K. Ono / Journal of Hazardous Materials 262 (2013) 741–747
Fig. 5. Results of sensitivity analyses. Ratios (percent) of difference in Cd air emissions for lower, upper, and default values are shown. (a) In 1990, (b) in 2010. NiCd B: Ni-Cd battery.
analyses, we assumed that there will be no change in the production amounts of alloys, pigments, Ni-Cd batteries, and “others” between 2010 and 2030. This assumption might be conservative, which may lead to an overestimation because the municipal use of Ni-Cd batteries will have been decreasing in the future. The amount of Ni-Cd batteries for municipal use had been decreasing until 2010 due to a decline in end-use products that were fabricated with these batteries. We assumed that an exchange of EGTS in a MSWI would not occur between 2010 and 2030. This is because most MSWIs and their EGTSs were renewed after 1997 (see Fig. 1), and the lifespan of MSWIs is 30 years on average; therefore, exchange or renewal of EGTS is not likely. The effect of stricter controls on the use on Ni-Cd batteries, i.e., mandatory 30% collection after use, was analyzed. Fig. 6 shows future Cd emissions under three scenarios: scenario 1, business as usual (BAU); scenario 2, promoting voluntary collection and recycling of used Ni-Cd batteries to reach levels of 50% as of 2015; and scenario 3, ban on using Cd in Ni-Cd batteries from 2015 in addition to Ni-Cd battery collection (30%). Emission in 2030 in each scenario is shown in the box by use category.
In 2030, Cd emissions under scenarios 1, 2, and 3 will be 0.46, 0.34, and 0.055 t/year, respectively. Compared to the BAU scenario, banning the use of batteries and intensive collection of batteries will reduce emissions by 0.40 and 0.11 t in 2030, respectively. A complete ban on Ni-Cd batteries would reduce emissions, but emissions from MSWIs would not be zero immediately after enforcement of the ban. As mentioned above, Cd use is Japan is already limited. Municipal use of Ni-Cd batteries has actually decreased because other batteries (i.e., Ni-MH or Li-ion batteries) are commonly available and have replaced Ni-Cd batteries. However, the emissions originating from pigments, which were mainly used in 1970s, will still contribute in 2030 (box in Fig. 6). Cd is an inevitable common byproduct of zinc refining. From the viewpoint of effective use of a limited resource, intensive collection of used Cd as a closed system might be a better approach than a complete ban of Cd use. Needless to say, a restriction of dispersive use of Cd will be most important. This method will provide quantitative evidence for such discussions. Dynamic SFA allowed for delayed disposal or emissions, which were derived from past use of Cd, to be estimated. This study quantitatively demonstrated
Fig. 6. Cd emissions under 3 scenarios in which Ni-Cd batteries are restricted. Scenario 1, business as usual (BAU); scenario 2, promoting voluntary collection and recycling of used Ni-Cd batteries to reach levels of 50% from 2015; scenario 3, ban on using Cd in Ni-Cd batteries from 2015 in addition to Ni-Cd battery collection (30%). Emission in 2030 by each scenario was shown in the box by use category.
K. Ono / Journal of Hazardous Materials 262 (2013) 741–747
the lag time and the amount by which the emissions would be reduced because of the countermeasures. Scenario analysis can show quantitative results regarding the differences in Cd emissions from MSWIs with and without controlling the use of the metal. Environmental monitoring around a MSWI cannot provide information on the use category of Cd emissions. 4. Summary In this study, Cd emissions from municipal solid waste incinerators were estimated for 1970–2030. The method enabled us to identify the major Cd emission source from MSWIs. Emissions peaked in 1973 at levels ten times that in 2010. In the 1970s, the main source was pigments, but after 2000 it was Ni-Cd batteries. We assessed Cd control countermeasures, such as intensive collection of used Ni-Cd batteries or a ban on the future use of Ni-Cd batteries. Banning Cd use reduced emissions more than the intensive collection of batteries. This method is applicable not only to Cd but to other metals as well because the concept of calculation is simple and the required parameters for the estimation are common. Acknowledgement The author is grateful to Mr. Hideki Wada of the Sustainable System Design Institute, who developed the SFA tool for this research. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.jhazmat. 2013.09.033. References [1] B. Bergbäck, S. Anderberg, U. Lohm, Accumulated environmental impact: the case of cadmium in Sweden, Science of the Total Environment 145 (1994) 13–28. [2] E. van der Voet, L. van Egmond, R. Kleijn, G. Huppes, Cadmium in the European Community: a policy-oriented analysis, Waste Management and Research 12 (1994) 507–526. [3] C.W. Hu, M.R. Chao, K.Y. Wu, G.P. Chang-Chien, W.J. Lee, L.W. Chang, W.S. Lee, Characterization of multiple airborne particulate metal in the surroundings of a municipal waste incinerator in Taiwan, Atmospheric Environment 37 (2003) 2845–2852. [4] D. Rimmer, C. Vizard, T. Pless-Mulloli, I. Singleton, V. Air, Z. Keatinge, Metal contamination of urban soils in the vicinity of a municipal waste incinerator: one source among many, Science of the Total Environment 356 (2006) 207–216. [5] M. Sakata, M. Kurata, N. Tanaka, Estimating contribution from municipal solid waste incineration to trace metal concentrations in Japanese urban atmosphere using lead as a marker element, Geochemical Journal 34 (2000) 23–32. [6] F. Zhang, S. Yamasaki, M. Nanzyo, K. Kimura, Evaluation of cadmium and other metal losses from various municipal wastes during incineration disposal, Environmental Pollution 115 (2001) 253–260. [7] M. Schuhmacher, M. Meneses, S. Granero, J. Llobet, J. Domingo, Trace element pollution of soils collected near a municipal solid waste incinerator: human health risk, Bulletin of Environment Contamination and Toxicology 59 (1997) 861–867. [8] J. Llobet, M. Schuhmacher, J. Domingo, Spatial distribution and temporal variation of metals in the vicinity of a municipal solid waste incinerator after a modernization of the flue gas cleaning systems of the facility, Science of the Total Environment 284 (2002) 205–214. [9] F. Cangialosi, G. Intini, L. Liberti, M. Notarnicola, P. Stellacci, Health risk assessment of air emissions from a municipal solid waste incineration plant – a case study, Waste Management 28 (2008) 885–895. [10] M. Meneses, J. Llobet, S. Granero, M. Schuhmacher, J. Domingo, Monitoring metals in the vicinity of a municipal waste incinerator: temporal variation in soils and vegetation, Science of the Total Environment 226 (1999) 157–164. [11] L. Morselli, M. Bartoli, B. Brusori, F. Passarini, Application of an integrated environmental monitoring system to an incineration plant, Science of the Total Environment 289 (2002) 177–188.
747
[12] Ministry of Environment (MOE), Annual Report on Waste Treatment in Japan, 2010 http://www.env.go.jp/recycle/waste tech/ippan/h22/data/ FY2010, disposal.pdf (in Japanese, accessed 14.03.13). [13] Ministry of Environment, Japan, PRTR Data Page, 2013 http://www2.env. go.jp/chemi/prtr/prtrinfo/e-index.html (accessed 14.03.13). [14] A. Lindqvist, F. von Malmbor, What can we learn from local substance flow analyses? The review of cadmium flows in Swedish municipalities, Journal of Cleaner Production 12 (2004) 909–918. [15] Y. Matsuno, T. Hur, V. Fthenakis, Dynamic modeling of cadmium substance flow with zinc and steel demand in Japan, Resources, Conservation and Recycling 61 (2012) 83–90. [16] K. Mason-Jones, H. von Blottnitz, Flows and fates of nickel-cadmium batteries in the City of Cape Town, Minerals Engineering 23 (2010) 211–218. [17] M. Fuse, K. Tsunemi, Assessment of the effects of the Japanese shift to lead-free solders and its impact on material substitution and environmental emissions by dynamic material, Science of the Total Environment 438 (2012) 49–58. [18] Japan Mining Industry Association, Bulletin of Japan Mining Industry Association, Tokyo. [19] M. Asari, S. Sakai, Battery recycling and flow analysis of NiCd and Li-ion batteries in Japan, in: The 7th International Conference on Waste Management and Technology, 2012. [20] Clean Japan Center, Research on the way of recycling-oriented economic system in the non-ferrous metal materials – research on the evaluation of the social system for waste reduction, Project supported by Japan Keirin Association Report FY 1998, 1999 (in Japanese). [21] Ministry of Finance (MOF), Trade Statistics of Japan, 2013 http://www.customs. go.jp/toukei/info/index e.htm (accessed 14.03.13). [22] BiPRO, Umweltbundesamt, Exemption for the use of cadmium in portable batteries and accumulators intended for the use in cordless power tools in the context of the Batteries Directive 2006/66/EC Final Report (revised), consortium ESWI (Expert team to Support Waste Implementation). 26 January 2010, 2006. [23] H. Hatayama, I. Daigo, Y. Matsuno, Y. Adachi, Outlook of the world steel cycle based on the stock and flow dynamics, Environmental Science and Technology 44 (2010) 6457–6463. [24] National Institute for Environmental Studies, Lifespan Database for Vehicles, Equipment, and Structures, 2013 http://www.nies.go.jp/lifespan/ isic search.php (accessed 14.03.13). [25] N. Tanigawa, T. Oikawa, K. Mimori, S. Awa, Annual Report of the Tokyo Metropolitan Research Institute for Environmental Protection, No. 28, 1998, pp. 65–68 (in Japanese). [26] Japan Portable Rechargeable Battery Recycling Center, Statistics of Recovery of Secondary Batteries, 2013 http://www.jbrc.net/hp/contents/recycle/ index.html (accessed 14.03.13). [27] H. Belevi, H. Moench, Factors determining the element behavior in municipal solid waste incinerator. 1. Field studies, Environmental Science and Technology 34 (2000) 2501–2506. [28] L.S. Morf, P.H. Brunner, S. Spaun, Effect of operating conditions and input variations on the partitioning of metals in a municipal solid waste incinerator, Waste Management and Research 18 (2000) 4–15. [29] C.H. Jung, T. Matsuto, N. Tanaka, T. Okada, Metal distribution in incineration residues of municipal solid waste (MSW) in Japan, Waste Management 24 (4) (2004) 381–391. [30] M. Lemann, R. Walder, A. Schwyn, Heavy metals in municipal solid waste incineration residues, Journal of Power Sources 57 (1995) 55–59. [31] K. Nakamura, S. Kinoshita, H. Takatsuki, The origin and behavior of lead, cadmium and antimony in MSW incinerator, Waste Management 16 (1996) 509–517. [32] A. Kida, S. Sakai, S. Sibakawa, A. Matsumoto, Heavy metals emission from municipal solid waste incineration system retrofitted for dioxins control, JEC 13 (2003) 51–67 (Abstract in English). [33] Ministry of Environment (MOE), Survey on Municipal Solid Waste Treatment, 2002 http://www.env.go.jp/recycle/waste tech/ippan/h14/data/seibi/ facility/01.xls (accessed 14.03.13). [34] Ministry of Environment (MOE), Report on Dioxin Concentration in Exhaust Gas from Municipal Solid Waste Incinerators http://www.env.go.jp/recycle/ dioxin/ippan/kekka11.html (accessed 14.03.13). [35] S. Yanagi, R. Sugawara, Y. Saito, The survey of inorganic salts in seeping water from controlled waste landfills and municipal solid waste incineration ashes, in: Annual Report of Miyagi Prefectural Institute of Public Health and Environment No. 24, 2006, pp. 79–83 (in Japanese). [36] Y. Suzuki, H. Takatsuki, S. Sakai, An effect of products recycling upon lead and cadmium flow in Japan, in: Proceedings of the 3rd Annual meeting for Society for Environmental Economics and Policy Studies, 1998, pp. 89–90 (in Japanese). [37] H. Tian, J. Gao, J. Hao, L. Lu, C. Zhu, P. Qiu, Atmospheric pollution problems and control proposals associated with solid waste management in China: a review, Journal of Hazardous Materials 252–253 (252) (2013) 142–150.