Air Qual Atmos Health DOI 10.1007/s11869-015-0325-8
Health risks and economic costs of exposure to PCDD/Fs from open burning: a case study in Nairobi, Kenya Yi-Hsuan Shih & Stephanie Jepng’etich Kasaon & Chao-Heng Tseng & Huang-Chin Wang & Ling-Ling Chen & Yu-Ming Chang
Received: 23 September 2014 / Accepted: 5 February 2015 # Springer Science+Business Media Dordrecht 2015
Abstract This study assesses the incremental health risk of exposure to dioxins and furans (PCDD/Fs) from indiscriminate burning of wastes in Nairobi and the potential economic benefits of reductions in dioxin-induced cancer mortality contributed by proper waste management. Fugacity models level III incorporated with the Human Health Risk Assessment Protocol (HHRAP) (USEPA 2005a) and CalTOX were utilized to simulate the PCDD/F levels in biotic environmental compartments. PCDD/F concentrations in samples of potatoes, eggs, beef, and long life milk were analyzed and compared with the modeled values. The PCDD/F concentration of 3.35 pg TEQ/ g in the milk sample was observed to rank the highest in food samples and exceeded the European Union criteria. Comparison results suggest that the level III + HHRAP is more conservative than CalTOX in health risk assessment. Regularities in the analyzed WHO-TEQ congener profiles for the food samples were discussed. The incremental dietary exposure to PCDD/Fs for the residents in Nairobi was estimated to be 0.08–2.15 pg TEQ/kg-day, falling within the WHO tolerable daily intake of 1–4 pg TEQ/kg-day. Potential excess cancers due to dietary exposure to PCDD/F associated with all illegal waste burning in Nairobi were estimated to be 636 cases over the 30-year time period or 21 cases/year, accounting for 0.05 % of cancer cases in the entire country of Kenya. With the waste recycling rate increased by 5 % and the opening of the new sanitary landfill that can reduce 50 % of waste disposed at the Dandora dumpsite, the economic benefits of avoided cancer deaths is expected to be US$ 0.16–1.93 Y.
million. These results indicate that additional actions on waste management, e.g., waste minimization and construction of sanitary landfill, should be implemented for the public health of Kenyans. Keywords PCDD/Fs . Fugacity model . Risk assessment . Economic costs . Waste management . Open burning
Introduction Chlorinated pollutants, such as pentachlorophenols, polychlorinated biphenyls (PCB), or DDT, were produced intentionally for industrial uses and were gradually phased out since the 1980s; while polychlorinated dibenzo-p-dioxins (PCDD) and polychlorinated dibenzofurans (PCDFs) were formed unintentionally and were deemed as the most toxic by-products attributed to the combustion of chlorinecontained municipal solid waste (Rappe et al. 1987; Fiedler 1996). Various kinds of wastes, e.g., wood, paper, plastics, tires, scrap metals, and motor oil, contain chlorinated compounds and are possible chlorine sources forming PCDD/Fs. PCDD/Fs, also called dioxins, are a group of persistent organic pollutants (POPs), of which molecular structure is not affected by transformation processes, e.g., biodegradation by microorganisms, hydrolysis, and photolysis (Fiedler 1996). Thus, the environmental and health impacts of dioxin could continue as long as life persists on Earth. Ingestion of dioxin-enriched foods through the food chain is known to be the dominant pathway of exposure (Kim and Moon 2013). Exposure to PCDD/Fs can cause immune system damage, severe reproductive and developmental problems, and interference with regulatory hormones. Carcinogenic and teratogenic effects can be induced at extremely low dose of dioxins (Ahlborg et al. 1992; Mizukami 2005; Aderemi and Falade 2012). Higher PCDD/F exposure is associated with an increased mortality risk in adults (Lin et al.
Air Qual Atmos Health
2012). Clinical manifestations as chloracne and hyperpigmentation with pruritus were referred to as the hallmark of high exposure to dioxin (Schecter et al. 2006). Health risk assessment works as a scientific basis for making regulatory decisions or policies. Several health risk assessments of PCDD/F emissions from municipal waste incinerators have been performed using air dispersion and multimedia transport models (Schuhmacher et al. 2001; Meneses et al. 2004; Cangialosi et al. 2008). Studies on occupational or human exposure examined PCDD/F levels in indoor environments (Hu et al. 2004; Kim and Moon 2013). However, risk assessment of exposure to PCDD/Fs emitted from open burning of municipal waste at a dumpsite is rarely presented. Relationships between outdoor and indoor levels of atmospheric PCDD/DFs and their emission from the anthropogenic sources have long been assessed as there are ongoing uncertainties over the contribution of different sources (Lohmann and Jones 1998; Lohmann et al. 1999). Compared with the use of air dispersion models in multimedia transport and health risk assessment (USEPA 2005a), few studies used fugacity model to calculate increased PCDD/F concentrations. At local scale simulation, atmospheric simulation results revealed the limited influence of industrial sources in local PCDD/F concentrations (Onofrio et al. 2011). The majority of the anthropogenic emissions tend to be transported at regional/global scales (Lohman and Seigneur 2001). It demonstrated that due to long atmospheric residence times, dioxins may undergo long-range transport and are deposited onto soils and sediments acting as ultimate sinks for them (Lohmann and Jones 1998). Moreover, due to toxicity, persistence and bioaccumulation, the levels of PCDD/Fs in the environmental media, especially biota in the food chain, are of particular concern. Therefore, regional/global multimedia-compartmental models have been developed and improving constantly. In an environmental system that is complex and dynamic, the empirical observations of PCDD/Fs in environmental media and biota, which receptors will be exposed to, are limited in assessing PCDD/F exposure of humans and ecosystems in most cases (Cowan-Ellsberry et al. 2009). Moreover, measurements cost time and money and only reflect the existing levels, of which variations with time and space cannot be validated. Furthermore, since from exposure to response is a long-term process, sample collection and analysis is inefficient in risk assessment. Thus, the evaluative POPs models like fugacity were introduced to help explain the complex behaviors of POPs and predict approximate concentrations of these chemicals in a generic environment (Mackay et al. 1992). Mackay et al. (1992) established the generic or evaluative fugacity models, e.g., levels I, II, and III, to assess and set priorities among POPs in hypothetical environment. They were further served as a simplified tool for health risk assessment in comparison with spatially resolved models (Wania and Mackay 1999; Mackay 2001). Whereas the earlier
fugacity models like levels I and II model assume the discharged chemical has achieved equilibrium between the different bulk compartments of the environment (air, water, soil, and sediment), level III describes non-equilibrium partitioning between media of different fugacities at a steady state. A report done by Webster and Mackay (2007) presented the application of RAIDAR, a fugacity-based model, to the simulation of PCDD/Fs generated from the waste incineration in northern Canada. However, validation for the fugacity model, which can be done by applying it to a similar situation in which monitoring data are available, was needed. Level III was chosen to be evaluated by comparing its modeled results with the measurements for PCDD/Fs in food samples. The fugacity-based model was further incorporated with the Human Health Risk Assessment Protocol (HHRAP) (USEPA 2005a) to simulate the PCDD/Fs levels in biota components. Moreover, CalTOX model developed by McKone (1993) is based on the level III fugacity concept and has been shown to produce similar estimates for the distribution of chemicals among the air, water, surface soil, and sediment (Maddalena et al. 1995). Nevertheless, further comparison of estimates in biota and health risk estimates from level III and CalTOX has not yet been carried out. Presently, in many Asian and African developing countries, a large percentage of waste cannot be properly classified, recycled, or successfully composted. Main disposal methods are indiscriminate incineration and landfilling. Refuse is dumped haphazardly into low-lying areas of open land whether onsite or offsite. Many landfills are downgraded and function as open dumps because municipalities do not have the financial capacity to maintain them. Particularly, landfill gas consists mainly of methane and carbon dioxide that are both major constituents of the world’s problem greenhouse gases (GHGs). The widespread dumping of wastes in water bodies aggravates the problems of generally low sanitation levels across the low-income settlements. Open dumpsites have long been recognized as sources of odor and breeding ground of disease-transmitting vermin and vectors, which cause malaria and dengue fever, and pose public health and environmental risks. Air, water, soil, and food contaminated by gases emissions and leachate discharges containing toxic chemicals like dioxins and heavy metals potentially cause serious health impacts including mutagenic and carcinogenic effects through hazardous waste disposal. Contagious disease and infectious pathogens like diarrheal diseases, hepatitis B, and human immunodeficiency virus (HIV) can be transmitted through the mismanagement of biomedical wastes (Henry et al. 2006). As the amount of solid waste increases coupled with an ineffective waste management system, there are high possibilities of negative short- and long-term impacts on human health and the environment in general. To insure public health and the environment, an integrated solid waste management system is crucial to developing countries, including Kenya, a slowly
Air Qual Atmos Health
industrializing country in Africa (UNEP 2010). There are social, environmental, and economical benefits in municipal solid waste recycling. Economic analysis can help raise awareness by publicity and enhancing the education regarding adverse effects of hazardous materials. Human health benefits of reducing dioxin emissions by enforced waste management can be expected. However, few studies attempted to quantify health costs of dioxin emissions (Kishimoto et al. 2001; Wright et al. 2001). The main goals of the present study are as follows: (a) simulate the PCDD/F levels in environmental media and biota associated with the PCDD/F emissions under different waste disposal scenarios using the fugacity models, (b) examine PCDD/F concentrations in selected food samples as a complement to the model estimates, (c) compare the level III estimates with the CalTOX for model verification, (d) quantify health risks through dietary exposure to PCDD/Fs of the residents living in Nairobi, and (e) estimate the health costs of cancer associated with exposure to PCDD/F. As few studies examined PCDD/F levels in food in Kenya, the analytical results can help estimate economic costs associated with cancer risks due to PCDD/F exposure for Kenyans. To the best of Fig. 1 Location of the Dandora dumpsite
our knowledge, this is the first paper to model the health and economic costs of PCDD/F emissions from open burning at dumpsites using the fugacity models. The local authorities in charge of environment and natural resources may use the estimated figures to come up with waste management policies to aid in the reduction of environmental pollution.
Materials and methods Study area Figure 1 illustrates the location of Dandora Municipal Dumping Site, the end disposal of Nairobi’s waste, what initially was to be refilling of an abandoned quarry 7.5 km east of the city center, occupies about 30 acres of land. There are legal loop holes when it comes to proper waste management at the dumpsite. As there are no clear procedural guidelines, medical and hazardous wastes can be found illegally dumped at the site. Open burning of municipal waste is widely practiced by the residents of Nairobi (Henry et al. 2006). Under the pilot study commissioned by the United Nations Environment
Air Qual Atmos Health
Program (UNEP), Kimani (2007) examined 328 children aged 2–18 years living around the dumpsite and revealed that 50 % of children examined who live and school near the dumpsite had respiratory ailments and high blood lead levels. The most obvious potential source of POPs releases at the site is the burning of chlorine-containing waste products such as commonly-found polyvinyl chloride (PVC) plastics. The high levels of unintentional POPs represent a concern for wider contamination since the Nairobi River passes below the dump and eventually drains into the Indian Ocean (ENVILEAD 2005). Due to the official designated dumpsite at Dandora reaching its full capacity, it has been noted to be responsible for gross environmental and public health hazards. Plans to decommission the Dandora dumpsite are underway under an initiative funded by Japan International Cooperation Agency. A new sanitary landfill at Ruai, 30 km east of central Nairobi (Fig. 1), is set to be commissioned. The proposed venue will reduce the waste disposed at Dandora dumpsite, and will be engineered for gas and electricity generation. These projects will be implemented in partnership with private companies as part of the council’s drive to open a new revenue stream under the devolved government. Scenario setting Four scenarios were proposed based on the projected amount of municipal solid waste in the year 2013. A forecast production of 3866 t/day of waste was left for disposal in Nairobi (UNEP 2010). In the Worst Case scenario, improperly and illegally disposed wastes at dumping sites other than Dandora in Nairobi were considered, and the proportion of total waste disposed and indiscriminately burnt was arbitrarily chosen to be 25 %. The business as usual (BAU) assumed 27 % of the waste generated was properly disposed at the Dandora dumpsite (UNEP 2010) and among them, 10 % of the waste was indiscriminately burnt. Table 1 summarizes the waste disposed and burnt at Dandora under each scenario in the year of 2013. There are two possible waste management plans the City Council of Nairobi could adopt given the observed trends (UNEP 2010). The Command and Control (C&C) scenario considered the increase of recycling level and the waste burning at the dumpsite was assumed to be reduced by 50 %. The Master Plan scenario considered the opening of the sanitary landfill at Ruai and the consequent decrease in the amount of waste disposed at the Dandora dumpsite by 50 %. Table 1 lists the PCDD/F emission rate under each scenario. For the open burning of domestic waste in uncontrolled conditions, an emission factor of 1000 μg toxic equivalent (TEQ) per ton of waste is applied (UNEP 2005). The TEQ levels of PCDD/F were calculated with toxic equivalency factors (TEFs) recommended by the World Health Organization (WHO) in 2005, as listed in Table 2. The TEQ percentage for
Table 1
Waste disposal scenarios in Nairobi
Scenario
Descriptions
Worst Case
All illegally disposed waste in Nairobi was considered; 25 % of the waste disposed in Nairobi was burnt. BAU 27 % of all waste generated was disposed at Dandora; 10 % of the waste disposed at Dandora was burnt. Command & 27 % of all waste Control generated was (C&C) disposed at Dandora; 5 % of the waste disposed was recycled and 5 % was burnt at Dandora. Master Plan 50 % of the waste disposed at Dandora was transited to Ruai; 5 % of the remaining waste was recycled and 5% was burnt at Dandora.
Table 2
Waste Waste PCDD/F disposed burnt emission rate (kg TEQ/h) 3866
951
4.95E−03
1044
104
5.44E−04
1044
52
2.72E−04
522
26
1.36E−04
TEQ percentage of 17 PCDD/Fs from open burning
No.
Chemical name
WHO 2005 TEFa
1 2 3 4 5
2,3,7,8-Tetrachloro-DD 1,2,3,7,8-Pentachloro-DD 1,2,3,4,7,8-Hexachloro-DD 1,2,3,6,7,8-Hexachloro-DD 1,2,3,7,8,9-Hexachloro-DD
1 1 0.1 0.1 0.1
6 7 8 9 10 11 12 13 14 15 16 17
1,2,3,4,6,7,8-Heptachloro-DD Octachloro-DD 2,3,7,8-Tetrachloro-DF 1,2,3,7,8-Pentachloro-DF 2,3,4,7,8-Pentachloro-DF 1,2,3,4,7,8-Hexachloro-DF 1,2,3,6,7,8-Hexachloro-DF 1,2,3,7,8,9-Hexachloro-DF 2,3,4,6,7,8-Hexachloro-DF 1,2,3,4,6,7,8-Heptachloro-DF 1,2,3,4,7,8,9-Heptachloro-DF Octachloro-DF
0.01 0.003 0.1 0.03 0.3 0.1 0.1 0.1 0.1 0.01 0.01 0.0003 Total
a
Values are based on van den Berg et al. (2006)
b
Values are derived from the USEPA (2005b)
TEQ percentageb 5.3 12.8 1.0 1.5 3.0
% % % % %
0.6 0.2 7.1 1.7 30.5 17.8 6.0 0.5 9.7 2.0 0.2 0.0 100 %
% % % % % % % % % % % %
Air Qual Atmos Health
each of the 17 PCDD/F compounds was calculated using the analytical results for the composition of PCDD/Fs emitted from the open backyard burning practiced in the USA (USEPA 2005b). Table 2 shows the TEQ profile for the PCDD/F emissions.
emissions to air were set to be 3.2E−04 mole/day and 3.0E −03 mole/day in CalTOX for BAU and Worst Case scenarios, respectively. Finally, evaluation of model performance was conducted by comparing level III values to CalTOX estimates. Health risk assessment
Fugacity model Level III version 2.80 (available at the Centre for Environmental Modelling and Chemistry (CEMC) website) was used to simulate the distribution of PCDD/Fs. The study area in Nairobi is 684 km2, while the modeled area in the standard equilibrium criterion (EQC) environment is 105 km2 (Mackay et al. 1996). Essentially, the model thus addresses a situation in which there is one dumpsite for every 400 km2. The modeled area was considered to contain 125 dumpsites evenly spaced over the total area. The PCDD/F emissions in the modeled area were scaled by a factor of 125 accordingly. The emission rates of 17 PCDD/F compounds were calculated from the TEQ emission factor using the estimated total PCDD/F emissions under the four scenarios listed in Table 1. The estimated actual emission rate was then input to the level III model as a modeling parameter, assuming a constant emission to air by 100 %. For model verification, this study simultaneously applied CalTOX to estimate PCDD/F concentrations in exposure media and health risks. Table 3 specified the site-specific landscape properties and human exposure factors for CalTOX simulation in Nairobi. Continuous inputs of PCDD/F Table 3 Nairobi
Site-specific parameter values used in CalTOX model for
Variables Landscape properties Contaminated area(m2) Annual average precipitation(m/day) Land surface runoff(m/day) Ambient environmental temperature(K) Yearly average wind speed(m/day) Exposure factorsa Body weight (kg) Fruit and vegetable intake (g/day) Grain intake(g/day) Milk intake(g/day) Meat intake(g/day) Egg intake(g/day) Fish intake(g/day) Exposure duration (year) Averaging time (year) a
Dietary intake were based on Steyn and Nel (2006)
Value
6.84E+08 2.81E−03 2.10E−03 2.93E+02 3.00E+05 60 200 220 30 30 20 0 30 59
Ingestion of PCDD/Fs through the diet was considered as the principal pathway of exposure (Kim and Moon 2013). Therefore, this study estimated PCDD/F levels in food commonly ingested by the population living in Nairobi. Inhalation or ingestion of air, water, and soil were not considered. PCDD/ F concentrations in the diet samples from Nairobi were examined. Four selected groups of food, namely, boiled eggs, potatoes, long-life milk, and canned beef, were randomly sampled from local markets. Due to strict regulations of shipping fresh meats and vegetables to Taiwan, canned and packaged food samples were shipped in. Analyses tests of PCDD/F contents were conducted by a standard laboratory. Samples were extracted, cleaned, and concentrated. Extracts were analyzed using high-resolution gas chromatography/highresolution mass spectrometry (HRGC/HRMS) (USEPA 1994). Table 4 summarizes the equations and parameters for calculating PCDD/F concentrations based on the HHRAP (USEPA 2005a). The simulated PCDD/F concentration in soil from level III was used as a key parameter. The simulated concentrations in the selected groups of food can be further used in exposure quantification. The average amounts of each kind of food ingested by residents in Nairobi were listed in Table 3. The daily food consumption is estimated on weight of each kind of food based on a 24-hour recall questionnaire. Accordingly, the measured concentrations expressed as pg TEQ/g fat has to be converted into pg TEQ/g fresh weight using percentage of fat content in each food sample. Body burden was calculated as average daily intake (ADI) of WHO-TEQ per unit body weight, assuming the average body weight of adults is 60 kg. In terms of adverse health effects, there are two quantifying measures: carcinogenic and non-carcinogenic risks. The non-carcinogenic risk was quantified as hazard quotient (HQ), which is specified as the sum of hazard index (HI) for each exposure to the pollutants. The tolerable daily intake (TDI) recommended by the WHO in 2000 is 1–4 pg TEQ/kg-day for chronic exposure to PCDD/Fs (van Leeuwen et al. 2000). The TDI is regarded as the value of reference dose and was then compared with the body burden to derive the HQ. If HQ is less than 1, non-carcinogenic effects are not of concern. On the other hand, the cancer risk was calculated as multiplying the ADI with the oral cancer slope factor (CSF). Carcinogenic risks below 10−6 are deemed acceptable to general population.
Air Qual Atmos Health Table 4
Equations and parameters for predicting PCDD/F concentrations in biota Equation
Parametersa
Root vegetable concentration (CRV)
CRV =Cs×Br
Beef concentration (Abeef)
Abeef =(CRV ×QPbeef +Cs×QSbeef)×Babeef ×MF
Cow’s milk (Amilk)
Amilk =(CRV ×QPcow +Cs×QScow)×Bamilk ×MF
Egg (Aegg)
Aegg =(CRV ×QPchicken +Cs×QSchicken)×Baegg
• Cs is the simulated bulk soil concentration; • Br is plant-soil bioconcentration factor of 0.046. • QPbeef is daily quantity of consumed plant of 12 kg/day; • QSbeef is daily quantity of ingested soil of 0.5 kg/day; • Babeef is biotransfer factor of 0.08 d/kg fat; • MF is metabolism factor of 1. • QPcow is daily quantity of consumed plant of 20 kg/day; • QScow is daily quantity of ingested soil of 0.4 kg/day; • Bamilk is biotransfer factor of 0.4 d/kg fat; • MF is metabolism factor of 1. • QPchicken is daily quantity of consumed plant of 0.2 kg/day; • QSchicken is daily quantity of ingested soil of 0.022 kg/day; • Baegg is biotransfer factor of 3 d/kg fat.
a
Values are based on USEPA (2005a)
Monetization of health costs In terms of health impacts arising from dioxin emissions, there are carcinogenic and non-carcinogenic effects to be quantified. However, there were insufficient data to provide solid evidence of dose-response relationship between dioxin exposure and non-cancer risks, e.g., male and female reproductive deficiency and mortality of ischemic heart disease (USEPA 2012). Hence, this study focused only on monetization of the health impacts of cancer, while omitting the quantification of non-carcinogenic hazard. The number of cancer cases caused by dioxin exposure was determined using N cancer ¼ ADI CS F POP
ð1Þ
where Ncancer is number of cancer cases caused by dioxin intake in food over an exposure duration, which was assumed to be 30 years; ADI is average daily intake dose (pg TEQ/kgday); CSF is cancer slope factor of 1.5×105 (mg/kg/day)−1 for PCDD/Fs (USEPA 1999); and POP is exposed population, of which number was approximately 3 million in 2009 in Nairobi. For monetization of cancer mortality, a value of a statistical life (VSL) was needed to estimate the economic costs of cancer. Studies regarding willingness to pay (WTP) to avoid risk of cancer related-death, i.e., the VSL, were compared to derive a best estimate for Kenya. A report of European Commission estimated the cost of a cancer death to be EUR 1.39–2.14 million (US$ 1.92–2.95 million) using benefits transfer method (Postle et al. 2003). A VSL of US$ 0.11–0.17 million adjusted for per capita GDP for Kenya was obtained. Another study based on hedonic wage method implied a VSL of US2000$0.55–0.68 million for an African country (Benkhalifa et al. 2012). This estimate was transferred to US$0.10–0.12 million for Kenya. Thus, a moderate VSL of US$ 117,055
was obtained using the benefits transfer method. The VSL estimate was applied to monetize the cancer-related economic costs under each scenarios of PCDD/F exposure affected by the implementation strategies for waste management system.
Results and discussion PCDD/F levels in exposure media Figure 2 compares the modeled and measured relative concentrations of 2,3,7,8-TCDD in food products. Measured concentrations for cow’s milk had the highest PCDD/F levels among the food, followed by beef, eggs, and vegetables. Notably, the measurement results indicate that the sample PCDD/ F concentration of 3.35 pg TEQ/g in long-life milk is higher than the European Union (EU) criteria of 2.50 pg TEQ/g for foodstuffs (EU 2011). The finding implies that daily consumption of cow’s milk may produce a harmful buildup of PCDD/Fs in the human body. A field investigation done by ENVILEAD (2005) revealed that the PCDD/F levels in egg samples collected in the vicinity of the Dandora dumpsite reached up to 22.92 pg TEQ/g of fat. This extremely high sample value is higher than our measured upper value of 0.22 pg TEQ/g fat by a factor of 2. Owing to limitations on shipment of food from Kenya to Taiwan, the number of samples analyzed was restricted to one package per kind of food product. Moreover, the food obtained from the supermarkets, especially beef and long-life (UHT) milk that can be stored for a long time, may be imported from places out of this study area and are probably less contaminated than crops and livestock grown or raised in the surroundings of the dumpsite. Given the effect of the limited number of samples, the measured value might represent an extremely low or high value rather than an average one.
Air Qual Atmos Health Fig. 2 Comparison of model estimates (△○) with sampling data ( ) for PCDD/F concentrations in food products. Measured upper and lower bound concentrations are calculated assuming that all values of the non-detected congeners are limit of detection (LOD) and equal to 0, respectively
Figure 2 shows that the model estimates associated with waste burning in Nairobi in the Worst Case and BAU scenarios were all below the EU criteria for foodstuffs. However, the fugacity models generate only mass balance among environment media without considering spatial or temporal deviation to the levels of POPs. The potential effects on food safety attributed to dioxin emissions from indiscriminate burning of wastes could be significant. The analytical results verified the model estimation and provided values that are of utmost importance. The results indicate that the government should become more vigilant in controlling PCDD/F emissions, which can be done by formulating and enforcing policies that restrict the indiscriminate burning of wastes commonly practiced in Nairobi. Figure 2 shows that the CalTOX estimates are the highest among all estimates in the four food, excluding milk. In CalTOX, the highest simulation values occurred in meat, followed by eggs, vegetable, and milk. In the estimates of level III with HHRAP, the ranking of PCDD/F TEQ levels is in the decreasing order of milk, beef, eggs, and vegetable; this trend is in agreement with the measured values, as well as with the findings of a study in Taiwan (Hsu et al. 2007). The comparison results imply that the level III estimates are closer to the measured concentrations than those of CalTOX. The results also indicate that PCDD/Fs tend to accumulate in biota enriched with lipid due to their lipophilic nature and would be concentrated in their milk several times after being consumed by the cow or other mammals. Figure 3 presents the measured WHO-TEQ congener profiles for PCDD/F concentrations in the food samples.
Although differences can be noted in the TEQ profiles from food to food, regularities can still be observed. TCDD, 1,2,3,7, 8-PeCDD, and 2,3,4,7,8-PeCDF contributed most to the TEQ levels. Moreover, 1,2,3,7,8-PeCDD was the largest TEQ contributor in food, with the exception of UHT milk. Furthermore, 2,3,7,8-TCDF contributed more than 2,3,4,7,8-PeCDF did for both UHT milk and potatoes. This results can be attributed to the fact that TCDD, 1,2,3,7,8-PeCDD, and 2,3,4,7, 8-PeCDF rank the top three in the WHO 2005 TEFs (Table 4 in Appendix). However, the TEQ profiles of the assumed landfill fire emission rates in Table 4 differ from the measured profiles in the food samples. The comparison result implies that other natural or man-made emission sources of dioxin with different TEQ profiles exit. It also implies that due to the complicity of the fate and transport of dioxins, current fugacity models are difficult to accurately predict dioxin TEQ profiles in biota. Human exposure to PCDD/Fs Figure 4 presents the upper and lower estimates of PCDD/F dietary exposure for adults in Nairobi based on the measurements. With only one sample for each item of food, these ADI estimates might only represent extremely low/high values. The modeled PCDD/F exposure under the BAU and the Worst Case scenarios were determined using CalTOX and level III incorporated with HHRAP. Among the simulated values, level III BAU estimate was closest to and approximately two times higher than the measurements. The dietary exposure to PCDD/Fs was estimated to be 0.08–2.15 pg TEQ/kg-day,
Air Qual Atmos Health Fig. 3 a–d Profiles of PCDD/Fs concentrations in food samples
falling within the WHO tolerable daily intake of 1–4 pg TEQ/ kg-day. This result implies that even when the illegally disposed and burnt wastes at all the dumping sites in Nairobi are considered, the PCDD/Fs exposure would not pose potential non-carcinogenic risks to the population. Compared with the total diet study of Hsu et al. (2007), the incremental daily intake of PCDD/Fs in the Worst Case scenario was close to the Taiwanese ADI of 1.48 pg TEQPCDD/Fs+dl-PCBs/kg-day. The dietary intake rate is expressed as g/day fresh weight (Table 2), whereas the PCDD/F concentrations estimated using level III with HHRAP are expressed as pg TEQ/g milk, pg TEQ/g of dry weight plant for produce, and pg TEQ/g of fat weight tissue for meat and eggs. Therefore, overlooking the fat contents in food might lead to overestimation of the ADI. Furthermore, the consumption of fats such as margarines and oils, which account for 30 % of energy intake in Kenyan women, was not considered in this study (Steyn and Nel 2006). Being fat-soluble and tending to bioaccumulate in fat tissues, PCDD/Fs in fat and oil are commonly examined and
Fig. 4 Estimated and modeled PCDD/F dietary exposure for Nairobi
non-negligible. Estimation of dietary exposure to dioxin will not be accurate enough without taking the consumption of fat into account. Further investigations of dioxin level in fat and taking fat consumption into ADI estimation are suggested. CalTOX estimates were the highest and were higher than that of level III by 50 %, which can be attributed to the fact that CalTOX modeled the highest estimates for PCDD/F concentrations in most of the food except milk (see Fig. 2). Therefore, level III with HHRAP was deemed more conservative on health risk assessment than CalTOX. Cancer risks and economic costs of cancer deaths Table 5 summarizes the estimated cancer risks and economic costs under each scenario posed by the PCDD/F emissions from waste open burning in Nairobi using level III incorporated with HHRAP. All the cancer risks due to the open burning at dumpsites significantly exceeded the acceptable probability of 1×10−6 for the general population. By Eq. (1), we predict a maximum of 636 excess cancers over the exposure duration of 30 years or 21 cases/year due to dioxin exposure. Given that the annual number of new cancer cases was 39,000 with more than 27,000 deaths per year in Kenya, we calculate that a maximum of 0.05 % cancer cases in Kenya might be directly linked to dioxin emissions in the open burning of waste in Nairobi. Given the lack of health-care facilities, 80 % of cancer cases in Kenya are diagnosed at their late stages. Using this percentage as the mortality rate of cancer cases in Nairobi, a maximum of 17 deaths/year excess cancer deaths was estimated as the potential outcome related to dioxin exposure.
Air Qual Atmos Health Table 5 PCDD/F impact scenarios on cancer in Nairobi
a
The lower and upper estimates were margin compared with the BAU and Worst Case scenarios, respectively
Worst Case
BAU
Command and Control
Master Plan
Cancer risks Lifetime cancer cases Annual cancer cases Annual cancer deaths Cancer mortality costs (US$)
2.1E−04 636 21 17 1,986,555
2.3E−05 70 2 2 218,053
1.2E−05 35 1 1 109,026
5.8E−06 17 1 0 54,580
Economic benefits (US$)a GDP percentage (%)
– 0.017 %
– 0.002 %
109,026–1,877,529 0.001 %
163,473–1,931,975 0.000 %
Notably, the USEPA (2010) estimated that the oral CSF for dioxin based on human studies ranges from 1.0×105 to 1.3× 106 (mg/kg-day)−1, ranging over an order of magnitude. Moreover, the vulnerability to cancer posed by dioxin differs among the populations. In view of these factors, the confidence interval of the number of the estimates of cancer deaths due to dioxin emissions would also be extensive. Table 5 shows that the economic cost of increased cancer mortality related to the dioxin exposure was approximately US$2 million in the Worst Case scenario. Given that the GDP value of Nairobi was US$12 billion in 2012, the predicted costs will take 0.017 % of the GDP in Nairobi. Moreover, the economic benefits of the decrease in number of cancer mortality could be remarkable. Compared with the BAU, the economic benefits under the C&C and the Master Plan were estimated to be US$0.11–1.88 million and US$0.16–1.93 million, respectively (Table 5). Notably, other environmental and health risks resulting from illegal dumps and poor waste management could be more substantial than those of open burning alone. Non-cancer hazard posed by dioxins, such as reproductive and developmental abnormalities and endocrine disruption, might be more harmful to public health and even pose more socially extensive impact than cancer hazard. Health end-points such as HIV/AIDS and diarrheal diseases, the top and third causes of death in Kenya, can be transmitted to slum scavengers who live on waste recycling by frequent contact to contaminated waste. Thus, the concern of overestimation on economic benefits of the proposed waste disposal scenarios is insignificant. Other health impacts of deteriorated public health due to mismanagement of municipal waste should be taken into account for a more comprehensive analysis. Although the opening of a sanitary landfill in Ruai would aid in the reduction of waste burnt at Dandora and significantly reduce the PCDD/F emission levels, cancer risk for the population who are least exposed under the Master Plan scenario is still comparably higher than 1 × 10−6 by 5.8-fold. Moreover, due to longer transport distances, the direct disposal of waste in Ruai will generally be too costly for most of Nairobi’s residents. Therefore, more aggressive interventions should be carried out to reduce the amount of waste generated.
Regular examination of the PCDD/F concentrations in food products, especially in lipid-rich items, is deemed necessary for the prevention of contaminated food consumption in Nairobi. Further investigation on the amount of food ingested for different genders, ages, and socioeconomic groups is suggested to estimate dietary exposure more accurately. Notably, the emission characteristics of the uncontrolled combustion sources vary by waste contents and combustion conditions and have much more temporal diversity than the stationary or point sources. Therefore, determining the emission factors of miscellaneous sources, namely the open burning of household wastes, which are truly area sources, is difficult. Site-specific data on PCDD/F emission rates for the open burning at dumpsites are required to verify the accuracy of the model’s assumptions and lower the uncertainty of the results.
Conclusion The measured PCDD/F concentration in milk samples from Nairobi was 3.35 pg TEQ/g fat, which exceeds the 2011 EU criteria of 2.50 pg WHO-TEQ/g fat. However, due to the limited number of samples, the value might represent only an upper-bound estimate rather than the average value. The modeled concentrations and their trend in the food, except milk, using level III incorporated with the HHRAP were closer to the measured values than the CalTOX. The comparison results suggest that the level III fugacity model is probable and can be a simplified tool in health risk assessment. The CalTOX estimates for dietary exposure to PCDD/Fs were higher than those from the level III, which in BAU was close to the estimates based on measured values for food. This finding shows that the indiscriminate burning of wastes at dumpsites leads to a significant increase of PCDD/F concentrations in the exposure media. The results indicate that consideration should be given to more aggressive actions. Taking account of open burning of waste as commonly practiced in Nairobi, a maximum of 0.05 % of cancer cases in Kenya, or 17 deaths/year excess cancer deaths, was estimated as potential outcomes related to the dioxin emissions.
Air Qual Atmos Health
However, as the oral slope factor for dioxin based on human studies ranges over an order of magnitude and the vulnerability to cancer posed by dioxin differs among the populations, uncertainty regarding these estimates exists. The economic benefits of reducing the amount of waste burnt at dumpsites and the development of sanitary landfill in Nairobi range from US$ 0.16–1.93 million, depending on what scenario they are compared with. The results pointing to considerable health benefits of avoided cancer deaths can be achieved through improved waste management in the future. Moreover, the above figures may represent underestimates of the full economic benefits because the approach adopted producing the above figures does not address the benefits of reductions in the number of cases for non-carcinogenic and infectious diseases. Finally, in light of the fact that PCDD/Fs can be transported over long ranges, additional regulation or regional agreement should be made regarding the illegal burning of wastes at the dumpsite to decrease PCDD/F levels in the region. Greater public involvement through intolerance to waste mismanagement should be implemented. Acknowledgments The authors acknowledge the financial support of the National Science Council of the Republic of China, Taiwan under contract no. NSC 100-2628-E-027-007-MY3. We would also like to express appreciation to Faiza Ramadhan, Marion Amulyoto, and Joseph Kasaon for sample collection and exportation.
References Aderemi AO, Falade TC (2012) Environmental and health concerns associated with the open dumping of municipal solid waste: a Lagos, Nigeria experience. Am J Environ Eng 2(6):160–165 Ahlborg UG, Brouwer A, Fingerhut MA et al (1992) Impact of polychlorinated dibenzo-p-dioxins, dibenzofurans, and biphenyls on human and environmental health, with special emphasis on application of the toxic equivalency factor concept. Eur J Pharmacol 228(4):179–199 Benkhalifa A, Ayadi M, Lanoie P (2012) Estimated hedonic wage function and value of life in an African country. Available at http://www. hec.ca/iea/cahiers/2012/iea1201_lanoiep.pdf Cangialosi F, Intini G, Liberti L et al (2008) Health risk assessment of air emissions from a municipal solid waste incineration plant—a case study. Waste Manag 28(5):885–895 Cowan-Ellsberry CE, McLachlan MS, Arnot JA et al (2009) Modeling exposure to persistent chemicals in hazard and risk assessment. Integr Environ Assess Manag 5(4):662–679 EU (2011) European Commission Regulation No. EU 1259/2011 amending Regulation (EC) No 1881/2006 as regards maximum levels for dioxins, dioxin-like PCBs and non dioxin-like PCBs in foodstuffs. Accessed at http://www.efsa.europa.eu/en/topics/topic/ dioxins.htm Fiedler H (1996) Sources of PCDD/PCDF and impact on the environment. Chemosphere 32(1):55–64 Henry RK, Yongsheng Z, Jun D (2006) Municipal solid waste management challenges in developing countries—Kenyan case study. Waste Manag 26(1):92–100
Hsu et al. (2007) A total diet study to estimate PCDD/Fs and dioxin-like PCBs intake from food in Taiwan. Chemosphere 67(9):S65–S70 Hu S, ChangChien G, Chan C (2004) PCDD/Fs levels in indoor environments and blood of workers of three municipal waste incinerators in Taiwan. Chemosphere 55(4):611–620 Kim S-J, Moon H-B (2013) Occurrence and human exposure of PCDD/ Fs and dioxin-like PCBs in house dust from Busan, Korea: comparison with seafood consumption. Toxicol Environ Health Sci 5(3): 155–162 Kimani NG (2007) Environmental pollution and impacts on public health: implications of the Dandora Municipal Dumping Site in Nairobi, Kenya. United Nations Environment Programme (UNEP). Accessed at http://www.kutokanet.com/Storage/UNEP_ Dandora_Environmental_Pollution_and_Impact_To_Public_ Health_2007.pdf Kishimoto A et al (2001) Cost effectiveness of reducing dioxin emissions from municipal solid waste incinerators in Japan. Environ Sci Technol 35(14):2861–2866 ENVILEAD (Environmental Liaison, Education and Action for Development) (2005) The International POPs Elimination Project (IPEP): a study on waste incineration activities in Nairobi that release dioxin and furan into the environment. Accessed at http:// www.ipen.org/ipepweb1/library/ipep_pdf_reports/4ken% 20kenya%20waste%20burning%20and%20incineration.pdf Lin YS, Caffrey JL, Hsu PC et al (2012) Environmental exposure to dioxin-like compounds and the mortality risk in the U.S. population. Int J Hyg Envir Heal 215(6):541–546 Lohman K, Seigneur C (2001) Atmospheric fate and transport of dioxins: local impacts. Chemosphere 45:161–171 Lohmann R, Jones KC (1998) Dioxins and furans in air and deposition: a review of levels, behaviour and processes. Sci Total Environ 219: 53–81 Lohmann R, Green NJL, Jones KC (1999) Atmospheric transport of PCDD/Fs across the UK and Ireland: evidence of emission and degradation. Environ Sci Technol 33:2872–2878 Mackay D (2001) Multimedia environmental models: the fugacity approach, 2nd edn. Lewis Publishers, Boca Raton Mackay D, Paterson S, Shiu WY et al (1992) Generic models for evaluating the regional fate of chemicals. Chemosphere 24(6):695–717 Mackay D et al (1996) Evaluating the environmental fate of a variety of types of chemicals using the EQC model. Environ Toxicol Chem 15: 1627–1637 Maddalena RL, McKone TE, Layton DW, Hsieh DP (1995) Comparison of multi-media transport and transformation models: regional fugacity model vs. CalTOX. Chemosphere 30(5):869–889 McKone TE (1993) CalTOX: a multimedia total-exposure model for hazardous-wastes sites. California Environmental Protection Agency, CA, US Meneses M, Schuhmacher M, Domingo JL (2004) Health risk assessment of emissions of dioxins and furans from a municipal waste incinerator: comparison with other emission sources. Environ Int 30(4): 481–489 Mizukami Y (2005) Frontier density pattern of dioxins. J Mol Struc THEOCHEM 713(1–3):15–19 Onofrio M, Spataro R, Botta S (2011) The role of a steel plant in northwest Italy to the local air concentrations of PCDD/Fs. Chemosphere 82:708–717 Postle M et al. (2003) Assessment of the impact of the new chemicals policy on occupational health: final report. European Commission—Environment Directorate-General. 96pp. Norfolk, UK. Available at http://www.greenpeace.org/luxembourg/ PageFiles/344421/assessment-impact.pdf Rappe C et al (1987) Sources and relative importance of PCDD and PCDF emissions. Waste Manag Res 5(3):225–237 Schecter A, Birnbaum L, Ryan JJ, Constable JD (2006) Dioxins: an overview. Environ Res 101:419–428
Air Qual Atmos Health Schuhmacher M et al (2001) The use of Monte-Carlo simulation techniques for risk assessment: study of a municipal waste incinerator. Chemosphere 43(4–7):787–799 Steyn NP, Nel JH (2006) Dietary intake of adult women in South Africa, Kenya and Nigeria with a focus on the use of spreads. Available from: http://www.mrc.co.za/chronic/kenyareport.pdf UNEP (2005) Standardized toolkit for identification and quantification of dioxin and furan releases, 2nd edition. UNEP Chemicals, Geneva, Switzerland. Accessed at http://www.chem.unep.ch/POPs/pcdd_ activities/toolkit/default.htm UNEP (2010) Solid Waste Management in Nairobi: a situation analysis, technical document accompanying the integrated solid waste management plan. Prepared for the City Council of Nirobi by the University of Cape Town. Prepared by Kasozi A, von Blottnitz H. Accessed at http://www.unep.or.jp/ietc/GPWM/data/T3/IS_6_1_ Nairobi_SWM_SituationAnalysis.pdf USEPA (1994) Method 1613: tetra-through octa-chlorinated dioxins and furans by isotope dilution HRGC/HRMS. Office of Water, Engineering and Analysis Division (4303). Accessed at http:// water.epa.gov/scitech/methods/cwa/organics/dioxins/upload/2007_ 07_10_methods_method_dioxins_1613.pdf USEPA (1999) Human health and ecological risk assessment support to the development of technical standards for emissions from combustion units burning hazardous wastes: background document. Accessed at: http://www.epa.gov/osw/hazard/tsd/td/combust/pdfs/ rabdmain.pdf USEPA (2005a) Human Health Risk Assessment Protocol (HHRAP) for hazardous waste combustion facilities, final. Report EPA520-R-05006, Office of Solid Wastes, Washington, DC, USA
USEPA (2005b) The inventory of sources and environmental releases of dioxin-like compounds in the United States (External Review Draft 2005). EPA/600/p-03/002A, EPA, National Center for Environmental Assessment. Accessed at: http://www.epa.gov/ncea/ pdfs/dioxin/2k-update/ USEPA (2010) The U.S. EPA’s Draft Oral Slope Factor (OSF) for 2,3,7,8Tetrachlorodibenzo-p-dioxin (TCDD). Prepared by G. Rice, Science Advisory Board Dioxin Review Panel Meeting; Washington, DC, USA USEPA (2012) EPA’s reanalysis of key issues related to dioxin toxicity and response to NAS comments, Volume 1. U.S. Environmental Protection Agency, Washington, DC, EPA/600/R-10/038F. Available at http://www.epa.gov/iris/supdocs/1024index.html van den Berg M et al (2006) The 2005 World Health Organization reevaluation of human and mammalian toxic equivalency factors for dioxins and dioxin-like compounds. Tox Sci 93:223–241 van Leeuwen FXR et al (2000) Dioxins: WHO’s tolerable daily intake (TDI) revisited. Chemosphere 40(9–11):1095–1101 Wania F, Mackay D (1999) The evolution of mass balance models of persistent organic pollutant fate in the environment. Environ Pollut 100(1–3):223–240 Webster E and Mackay D (2007) Modelling the environmental fate of dioxins and furans released to the atmosphere during incineration. Available at http://www.trentu.ca/academic/aminss/envmodel/ Wright JC, Millichamp P, Buckland SJ (2001) The cost-effectiveness of reductions in dioxin emissions to air from selected sources: economic analysis for section 32 of the resource management act. Ministry for the Environment: Wellington, New Zealand. Available at http:// www.mfe.govt.nz/publications/hazardous/dioxin-reduction-costeffectiveness-aug01.html