dioxin/cl/incineration: another study

[charlie.cray@green2.greenpeace.org]

----------
Original-TO:      dioxin-l@essential.org
----------
Source:  Thomas, V.M., and Spiro, T.G.  An estimation of dioxin 
emissions in the United States.  Toxicol. and Environ. Chemistry 
50:1-37 (1995)
 
   Total annual US air emissions of polychlorinated dibenzo-p-  
dioxins and polychlorinated dibenzo-furans (PCDD/F) from all    
known sources are estimated to be about 400 kilograms as of    
1989, almost entirely from combustion sources.  Municipal    
solid waste incineration is the largest known source of    
PCDD/Fs, as of 1989.  The next largest sources include 
   hospital waste incineration, forest and agricultural fires,    
and residential wood burning.  Anthropogenic emissions of    
PCDD/Fs are estimated to be an order of magnitude greater than    
emissions of PCDD/Fs from forest fires.  Dioxin emissions are    
shown to generally increase with the chlorine content of the    
combusted material, in the absence of effective pollution    
control systems.  Dioxin emissions from 1940 to 1970 are also    
estimated.  The estimates are verified through analysis of the    
concentrations of dioxin in soil, air and sediments.
 
   The data on environmental dioxin concentrations in air, soil   
 and sediments are mutually constraining, in that air 
   concentrations provide information on current emissions, soil  
  concentrations provide information, as a function of the    
dioxin life-time, on total past emissions, and sediment cores    
provide information on the quantity emitted in each year of    
the time reflected in the core.   Taken together, the estimates   
 of emissions and loadings are consistent, within the error    
limits of the data, and do not support the hypothesis of major    
unknown dioxin sources.  
 
   Dioxin emissions from combustion processes, responsible for    
the bulk of known dioxin emissions, can be highly variable,    
depending on the completeness of combustion, the combustion    
system and air pollution control equipment, and the material    
being burned.  While temperatures, oxygen levels, and 

   residence times in a well-operated combustion system are    
sufficient to destroy PCDD/Fs and their precursors, high    
emissions can result from poor combustion conditions such as    
pockets of gas with insufficient oxygen, or of low temperature    
pathways that allow some of the gases to escape complete    
combustion, and conditions that allow for the formation of    
dioxin in the post-combustion zone.
 
   Nevertheless, the data presented in section 4 will show that   
 dioxin emission factors from combustion of similar fuels, with   
 broadly similar combustion systems and pollution control         
  devices, typically vary by less than an order of magnitude.     
  This holds true, for example, for hospital waste 
 incinerators, sewage sludge incinerators, and industrial wood   
boilers.  Thus, for most processes dioxin emissions can be    
categorized by the combustion activity.  Exceptions are the    
incineration of municipal and hazardous waste, for which    
different types of combustion systems can have dioxin 
  emission factors that differ by more than an order of 
   magnitude.  
 
    ... we begin with Figure 1, in which the average dioxin    
emission factor for each source category is plotted against    
the chlorine content of the combusted  material.  The table    
shows that average dioxin emissions of combustion source    
categories tend to increase with the average chlorine content.  
   While a dependence on chlorine content might seem to be a    
foregone conclusion, there has in fact been doubt on this point   
 since limited data, on test-burns at municipal waste 
   incinerators, for example, have shown no correlation.  There   
 are, of course, many variables besides chlorine content,    
especially the operating conditions of the combustion system,    
and the application of pollution control technology.  Figure 1    
shows that favorable operating conditions (open symbols) can    
lower emissions by as much as two orders of magnitude.  But    
under poorly controlled conditions (closed symbols), there is    
a clear dependence on chlorine content.  Moreover, the 
   relationship is more than proportional since the slope of the  
  log-log plot is somewhere between one and two.  Since dioxin    
molecules contain more than one chlorine atom, a higher order    
dependence is not unexpected.
 
   The data in Figure 1 are too coarse grained to evaluate    
differences due to the chemical form of chlorine in the    
combusted material.  In hospital waste, a large fraction of    

the chlorine may be in the form of polyvinyl chloride, and in    
transformer fires the chlorine is in the form of 
   polychlorinated biphenyls (PCBs), while in municipal waste a   
 significant fraction of the chlorine may be in the form of    
sodium chloride, as well as PVC plastics.  The figure 
   indicates that, by and large, dioxin emissions are roughly    
independent of the original form of the chlorine in the    
combusted material.  According to the current understanding of    
dioxin formation, HCl is the chlorine source for dioxin    
formation in catalyzed reactions on fly ash, with a maximum at    
temperatures of about 400 C.  The HCl is formed during 
   combustion as part of the breakdown process of chlorine    
containing  molecules.  Thus the amount of HCl formed might    
not depend strongly on the type of chlorine compound, although    
the degree of conversion of chloride salts to HCl is 
   uncertain.
 
   The relationship between chlorine content and dioxin 
   emissions, demonstrated in Figure 1, provides a means to    
estimate how dioxin emissions are related to the presence of    
anthropogenic chlorinated compounds in materials that are    
burned.  With municipal and hospital waste combustion being the   
 largest dioxin sources that are affected by anthropogenic    
chlorinated organic materials (see Table 2), the relationship    
of total dioxin emissions to anthropogenic chlorinated 
   compounds will depend primarily on how the emission factors    
for municipal and hospital waste combustion would change in    
the absence of chlorinated compounds.  ...
 
   4.  DIOXIN EMISSION SOURCES
   4.1  Consumer Waste Incineration.
   Municipal waste incineration:  In 1989, about 20 million tons  
  of municipal solid waste (MSW), 15% of the total, were 
   incinerated in the US.
    ... The average emission factor, weighted for the amount of   
 waste combusted at each type of facility, is, to one 
   significant figure, 10 ug/kg.
 
   The chlorine content of municipal waste has been measured to   
 be 2500 ppm (dry basis) in Pittsfield, Massachusetts, 4500 ppm   
 in Baltimore, Maryland, and 8900 ppm in Brooklyn, New York    
[21,22].  For Figure 1, the average chlorine content is taken    
to be about 4000 ppm.
 
   Hospital incinerators:  ... Based on emissions from six    
hospital waste incinerators, as shown in Table 4, their dioxin    
emission factor is estimated to be 20 ug/kg, twice that for    
municipal waste incineration. (Emissions of 85 ug/kg or higher    
have been reported at other hospital waste incinerators.[23])     
Also shown in Table 4 is the chlorine emitted as HCl from    
each incinerator, which gives a lower bound for the chlorine    
concentration of the original waste.  
 
   4.2  Industrial Waste Processing and Incineration
   Hazardous waste incineration:  ... As of 1989, about 1.6    
million tons of hazardous waste was burned at commercial and    
private hazardous waste incinerators in the US, and about 2    
million tons at cement and aggregate kilns, boilers and other    
furnaces (Table 5).
 
    ... for most trial burns, chlorine content is typically 1 to  
  10 percent.  For dedicated hazardous waste incinerators, the    
dioxin emission factors shown in Table 6 range from 0.01 to    
6.8 ug/kg, more than two orders of magnitude.  The average is    
1 ug/kg, and the median is 0.2 ug/kg.  For soil incineration,    
the emission factors are lower, probably because the soils'    
concentrations of combustible materials were low.  For soil    
incineration, the average emission factor is 0.5 ug/kg, and    
the median is 0.009 ug/kg.
 
   For cement kilns ... the dioxin emission factors range from    
0.04 to 10 ug/kg, again a range of more than two orders of    
magnitude.  The average is 3 ug/kg, and the median is about    
0.7 ug/kg.  Table 7 also shows two measurements of emissions    
from cement kilns when hazardous wastes were not being 
   combusted, resulting in emissions of 0.01 ug/kg and 0.4 ug/kg, 
   respectively.
 
   For other types of boilers and industrial furnaces burning    
hazardous wastes, there is very little data available.  The    
data in Table 8 range over three orders of magnitude, from    
0.01 ug/kg to 40 ug/kg of hazardous waste.  The average is 8    
ug/kg (determined entirely by the highest measurement) and the    
median is 0.1 ug/kg.  
 
   Although it may be a substantial overestimate, total dioxin    
emissions from hazardous waste combustion will be estimated    
using the average emission factor from each subcategory.  This    
results in an overall emission estimate, to one significant    
figure, of 10 kg annually.
 
   4.4  Fossil Fuel Combustion
   Gasoline Combustion: ... While leaded gasoline may contain    
some trace quantities of chlorine (on the order of 10 ppm),    
leaded gasoline typically includes the additives 
   dichloroethane or dibromoethane with a chlorine to lead weight 
   ratio of about 1 to 3.  Leaded gasoline can contain up to 700  
  ppm chlorine (Table 10), with less chlorine in lower-lead    
gasolines.  Thus combustion of leaded gasoline can be expected    
to produce more dioxin than combustion of unleaded gasoline,    

and gasoline with higher lead concentrations can be expected    
to produce more dioxin than low-lead gasolines.
 
   Based on the data in Table 10, the emission factor for 
   unleaded gasoline combustion is estimated to be 0.003 ug/kg.   
  For cars burning leaded gasoline, the dioxin emission factors   
 appear to be higher for cars burning higher-lead (and higher-  
chlorine) gasoline.  Because leaded gasoline in the United    
States has a relatively low lead concentration (0.1 g/gal),    
and a low chlorine concentration (10-30 ppm), the dioxin    
emission factor for US leaded gasoline combustion is taken to    
be 0.03 ug/kg, corresponding to the emission factor from the    
one car tested using low-lead gasoline.  
 
   Coal Combustion:  The average chlorine concentration in US    
coal is about 200 ppm, ranging from 20 to 8000 ppm, and coal    
combustion can be expected to produce some dioxin.  
        
-end-
----------