Elemental Chlorine Production

Three chlor-alkali processes are used industrially: mercury cell, membrane cell and diaphragm cell. BAT for new chlor-alkali plants is generally considered to be membrane cell or non-asbestos diaphragm cell (EU IPPCB 2011). General descriptions of the main production methods can also be found in the BAT&BEP Guidelines.

EFAir: An EFAir of ND is proposed for stand-alone chlor-alkali facilities.

EFWater: An EFWater of 0.002 μg TEQ/ECU has been derived for Class 2c chlor-alkali facilities, and 1.7 μg TEQ/ECU for Class 2b facilities based on Dyke and Amendola (2007). Of the seven plants that were not associated with the EDC/VCM/PVC chain, emission factors were calculated based on the published data and production rates estimated to be 98% of plant capacity for the relevant years. For four plants of the seven, EFWater was 0.0016 μg TEQ/ECU. USEPA (2004) cited a median PCDD/PCDF concentration of 120 pg TEQ/L wastewater from U.S. chlor-alkali facilities. An EFWater of 17 μg TEQ/ECU is recommended for Class 2a, approximately 10-fold higher than Class 2b.

EFProduct: Data describing PCDD/PCDF concentrations in Cl2 or its co-products, H2 and NaOH, are not available, and an EFProduct of ND is recommended.

EFResidue: For metal electrodes, Dyke and Amendola (2007) reported transfers to secure landfill that range from 0.2 to 18 μg TEQ/ECU capacity (Median = 1.1 μg I-TEQ/ECU capacity) based on data gathered in 2000 and 2002. In the intervening time, the four largest generators have been closed. The average of the remaining three was 0.22 μg I-TEQ/ECU capacity. Based on the assumption that production rates were 98% of capacity, the following were derived: for Class 2c, an EFResidue of 0.3 μg TEQ/ECU; Class 2b, 1.7 μg TEQ/ECU, based on the median, and Class 2a, 27 μg TEQ/ECU, based on the highest in the dataset are recommended.

For graphite electrodes, PCDD/PCDF concentrations in sludge from chlor-alkali production using graphite electrodes have been reported as follows: up to 3,985 μg I-TEQ/kg in a sample from Germany (She and Hagenmeier 1994); from 13 to 28 μg N-TEQ/kg in three samples from Sweden (Rappe et al. 1991), and 21.65 μg I-TEQ/kg in one sample from China (Xu et al. 2000). For the Chinese dioxin release inventory, it is assumed that on average 50 kg of graphite sludge is generated per ton of alkali produced. With a default concentration of 20 μg TEQ/kg graphite sludge, an EFResidue of 1,000 μg TEQ/t of chlor-alkali is proposed, based on the most recent data. However, Sweden estimates 2.5 kg graphite consumed per ton of NaOH produced. With the latter sludge generation rate, an EFResidue of 40 μg TEQ/t can be derived.

Ethylene Dichloride (EDC), Vinyl Chloride (VCM), and Polyvinyl Chloride (PVC)

The major processes in the production of ethylene dichloride (EDC), vinyl chloride (VCM), and polyvinyl chloride (PVC) are as follows:

1. EDC production is accomplished by two processes - direct chlorination and oxychlorination, which may be used in combination as the “balanced process”.
   • Direct chlorination of ethylene is the catalyzed reaction of ethylene with elemental chlorine to produce EDC and HCl. The reaction is carried out at relatively low-temperatures (50-70°C) and pressures that are 4-5 times normal atmospheric pressure. Ferric chloride (FeCl3) is commonly used as a catalyst, although chlorides of other metals (aluminum, copper, antimony) may also be used.
   • Oxychlorination of ethylene entails the reaction of ethylene with HCl and either air or oxygen in the presence of a catalyst (usually copper chloride) in a fixed-bed reactor or a fluidized-bed reactor to produce EDC and water.
   • The balanced process for EDC production consists, in effect, of linking the direct chlorination and oxychlorination processes by recovering HCl produced during EDC thermolysis or “cracking” and using it as a reactant in the production of EDC by oxychlorination. When no EDC or HCl is imported or exported, then the EDC unit is called a ‘balanced unit’. The balanced process is used at many modern EDC/VCM and EDC/VCM/PVC facilities; however, import of HCl and oversized oxychlorination units have become common in recent years.
2. EDC purification usually begins with a water quench followed by caustic scrubbing. The water is returned to the process or is steam stripped prior to transfer to a water treatment system. EDC is further purified by distillation and sent to temporary storage before being used to produce VCM.
3. VCM production is carried out by thermolytic dehydrochlorination (“cracking”) of EDC to produce VCM and HCl along with various by-products. Cracking furnaces typically operate at temperatures of 450- 650°C and pressures about 20 times normal atmospheric pressure.
4. VCM purification occurs via a distillation process that separates VCM from high boiling residues containing PCDD/PCDF and other by-products. These residues and by-products are typically incinerated or otherwise destroyed, and, in some cases, HCl produced during destruction is recovered and recycled.
5. PVC production entails polymerization of VCM to produce PVC resin by one of four routes - suspension, dispersion/emulsion, solution or bulk (mass) – with suspension as the predominant route. In suspension and dispersion polymerization, water is the reaction medium. Solution polymerization is carried out in an organic solvent. In bulk polymerization, VCM itself is the reaction medium until the polymer precipitates and absorbs remaining monomer. The reaction finishes as polymerization in the solid polymer.

PCDD/PCDF formation in the oxychlorination process is well-acknowledged. However, PCDD/PCDF are also known to occur in thermal or combustion processes of the EDC/VCM/PVC production chain.

The following waste streams from the EDC, VCM and PVC production chain are among those that potentially contain PCDD/PCDF and/or release PCDD/PCDF to one or more vectors:

• Stack gases from vent and liquid/vent combustors, HCl furnaces, and, if allowed, flares (USEPA 2012a);
• Treated wastewater, which is discharged to surface waters, and wastewater treatment residue (sludge), which is commonly sent to landfills or to on-site or off-site thermal oxidizers or combustors;
• Liquid residues from the oxychlorination process, which are generated at the rate of 25-40 kg/t VCM (EU IPPCB 2003), and are commonly sent to on-site or off-site thermal oxidizers or combustors;
• Spent fixed-bed oxychlorination catalyst, which is generated at rates of 10 to 200 g/t VCM in plants using this technology (EU IPPCB 2003). This material may be landfilled, although some manufacturers subject the catalyst to a thermal treatment to destroy organics associated with the solid support;
• Fluidized-bed oxychlorination catalyst, which undergoes attrition so that small particles are carried over to product vapor and washed out in quench water that is sent to wastewater treatment where the catalyst is captured in wastewater treatment residues;
• Coke, which is generated during EDC pyrolysis and removed at a rate of 0.1-0.2 kg/t VCM (EC 2003);
• Spent lime, which is used in some plants for VCM neutralization (EC 2003);
• Maintenance waste, which includes a lengthy list of materials such as process filter media, used activated carbon, packing/saddles from scrubbers, incinerator solids, solids from various process streams, etc. (Vinyl Institute 2002).

Releases to air

These arise from:

• Combustion devices – vent and liquid/vent combustors and thermal oxidizers, halogen acid furnaces, flares, on-site hazardous waste incinerators and power boilers. Some of these devices are used to control process vent gases, gaseous emissions from processes residues such as heavy ends and, in the case of on-site incinerators, liquid and solid residues. Halogen acid furnaces and possibly other devices recover HCl as a salable or reusable product from these devices, while power boilers are used to generate heat and power.
• Certain process units, e.g., gases from EDC pyrolysis furnaces are vented directly to air and gases from other processes may also be vented directly to air for limited periods of time in order to accommodate downtime and trips on pollution control systems (EC 2003).

No emission factors for releases to air due to direct venting from EDC pyrolysis furnaces and other processes are derived due to lack of information. Releases from on-site hazardous waste incinerators not considered part of the process are addressed in Source Group 1 – Waste Incineration, and flares and power boilers are addressed in Source Group 3 – Power Generation and Heating.

PVC-only Vent Combustors

For U.S. facilities, PCDD/PCDF concentrations and emissions per ton of PVC capacity derived from measurements in stack gases of vent combustors at facilities that produce only PVC are shown in Table III.48.1 (Eastern Research Group 2011).

Table III.48.1 PCDD/PCDF concentrations in stack gases of PVC vent combustors and their release to air per ton of PVC at facilities in the U.S.

PCDD/PCDF concentrations in stack gases ng I-TEQ/m³		PCDD/PCDF Emissions (µg/t PVC)
Vinyl Institute (2002) at 7% O2	Eastern Research Group (2011) at 3% O2	Eastern Research Group Baseline (2011)A	Eastern Research Group after new emission limitB is in place (2011)C
Range; Average (Median)	Range; Average (Median)	Range; Average (Median)	Range; Average (Median)
0.0014-0.042; 0.013 (0.01)	0.0017–0.91; 0.08 (0.017)	0.000053-0.026; 0.005 (0.00061)	0.000053-0.020; 0.0023 (0.00050)

^A Emission factors were derived for 13 PVC vent combustors, based on flue gas concentrations measured in 2010 under current practice and relevant production rates. Total air emission from these combustors was 3.58 x 104 µg TEQ/y. Production of PVC in 2010 was 14.0 billion pounds (6.36 million tons) and capacity was 7.26 million tons (CMAI 2011b) yielding a capacity utilization factor of 87.6%.
^B 0.023 ng TEQ/Nm³ at 3% O2.
^C Combustors exceeding the limit were constrained to the new emission limit.

Vent and Liquid/Vent Combustors or Thermal Oxidizers at EDC, EDC/VCM and EDC/VCM/PVC Production Facilities

Emissions to air from these processes come from combustion. Generally, combustor data is reported or regulated as a concentration in air; e.g., ng TEQ/normal cubic meter (Nm³). Estimating the mass of PCDD/PCDF released to air thus requires knowledge of stack gas flow rates and hours of operation. Some EDC, EDC/VCM and EDC/VCM/PVC facilities operate liquid/vent combustors or thermal oxidizers and some operate vent-only combustors or thermal oxidizers. The latter operations may dispose of their liquid residues (heavy ends) in on-site hazardous waste incinerators or contract with others for such disposal.

Table III.48.2 PCDD/PCDF concentrations in stack gases of EDC/VCM vent and liquid/vent thermal oxidizers and EDC/VCM vent combustors in the U.S. (Vinyl Institute 2002, Carroll Jr. et al. 2001)

PCDD/PCDF Concentrations in Stack Gases of EDC/VCM Vent and Liquid/Vent Thermal Oxidizers and Combustors in the U.S. (ng I-TEQ/m³)
	Vinyl Institute (2002) at 7% O2		Carroll Jr. et al. (2001) at 7% O2
	Range	Average (Median)	Range	Average (Median)
EDC/VCM Liquid and Liquid/Vent Thermal Oxidizers	0.034-6.4	1.38 (0.3)	0.034-1.5	0.43 (0.096)
EDC/VCM Vent Combustors	0.01-10.3	2.47 (0.59)	0.01-0.59	0.15 (0.031)

Based on these measurements generated in 1996, the Vinyl Institute calculated estimated emissions from liquid and liquid vent combustors as 3.7 g I-TEQ/yr and from vent combustors as 6.9 g I-TEQ/yr. No data was given as to the rate of production of EDC or VCM each of these system types was serving; thus a separate emission factor for each cannot be calculated. Normalizing the total TEQ emissions for total VCM production at the time (6,173,000 tons), yields 1.7 µg TEQ/t VCM (Carroll Jr. et al. 2001).

Det Norske Veritas validated that all the EDC/VCM plants that are signatories to the ECVM charter met the charter commitment of 0.1 ng TEQ/Nm³ maximum in 2011 (Det Norske Veritas 2012). Plastics Europe has published ecoprofile information on a number of materials including VCM. Air emissions are listed as 4.16 x 10^-8 mg TEQ/kg VCM (0.042 µg TEQ/t VCM produced).

Halogen Acid Furnaces (HAF)

For HAF at U.S. EDC, EDC/VCM and EDC/VCM/PVC production facilities, data obtained between 1995 and 2001 show median emissions of 0.06 ng TEQ/dscm (0.02-0.53 ng TEQ/dscm, average of three runs) for HAF (USEPA 2005a,b). There are insufficient data to allow the derivation of emission factors for release to air. Consequently, releases are estimated based on stack gas flow rates and a concentration of 0.02 ng TEQ/dscm for Class 3 HAF, 0.06 ng TEQ/dscm for Class 2, and 0.53 ng TEQ/dscm for Class 1.

EFAir for Vent and Liquid/Vent Combustors or Thermal Oxidizers at EDC, EDC/VCM or EDC/VCM/PVC facilities

Either the suggested EFAir may be chosen, or an EFAir may be calculated by multiplying PCDD/PCDF concentration in the flue gas by flue gas flow rates, in m³/hour, and hours per year of operation.

• Class 1: an EFAir of 5 µg TEQ/t VCM based on Carroll et al. (2001) or a flue gas concentration of 5 ng TEQ/Nm³ is used, a factor of 10 higher than Class 2.
• Class 2: an EFAir of 0.5 µg TEQ/t VCM or a flue gas concentration of 0.5 ng TEQ/Nm³ is used, a factor of 10 higher than Class 3.
• Class 3: an EFAir of 0.05 µg/t VCM or a flue gas concentration of 0.1 ng TEQ/Nm³ is used based on EU BREF (EU IPPCB 2003).

EFAir for PVC-only facilities

Similarly, either the given EFAir can be used or air releases can be estimated by multiplying the value for PCDD/PCDF concentrations in the flue gas of these devices by their respective flue gas flow rates, and their respective hours per year of operation.

• Class 1: an EFAir of 1 µg TEQ/ton PVC or PCDD/PCDF concentration of 1 ng TEQ/Nm³ is based on the highest baseline value reported by Eastern Research Group (and a factor of 10 higher than Class 2).
• Class 2: an EFAir of 0.1 µg TEQ/ton PVC or emission concentration limit of 0.1 ng TEQ/Nm³, a factor of 5 larger than Class 3.
• Class 3: PCDD/PCDF emission of 0.021 µg TEQ/ton PVC constitutes the highest emission value in grams for a real plant still meeting the proposed US EPA emission concentration limit of 0.023 ng TEQ/Nm³ (Eastern Research Group 2011, CMAI 2011b).

For VCM production, air emission factors for HCB and PCBs of 72 µg/t and 8.9 µg/t, respectively, have been derived in Japan for one facility (Iwata et al. 2008).

Releases to water

Releases to water from the EDC/VCM/PVC production chain most commonly consist of treated effluents discharged from on-site, facility-wide wastewater treatment systems or systems that serve multiple facilities. Treatment of industrial wastewater commonly entails a sequence of processes: biological treatment, settling/clarification, equalization, neutralization, filtration, stripping (air and steam), chemical precipitation, and adsorption (USEPA 2004). The EU BREF (section 12.5.4.2) notes that BAT for water releases consists of extensive pre-treatment followed by final biological treatment that can achieve 0.1 ng TEQ/L in effluent (EU IPPCB 2003). OSPARCOM and the ECVM charter (EU BREF, section 12.4.7) cite 1 µg TEQ/ton annual oxychlorination capacity as BAT and industry commitment respectively; total effluent water flow variation plant-to-plant is an important factor in relating these values.

The Vinyl Institute (Vinyl Institute 2002) reported PCDD/PCDF concentrations in treated wastewater from the U.S. EDC/VCM/PVC industry, as listed below, together with the respective emission factors.

• For EDC/VCM and EDC/VCM/PVC facilities: concentrations in wastewater of 3.0 to 6.6 pg I-TEQ/L, with an average of 4.7 pg I-TEQ/L (ND = DL/2; congeners detected in all samples), and emission factors of 0.0032 to 0.032 µg I-TEQ/t EDC, with an average of 0.015 µg I-TEQ/t EDC; and
• For PVC only facilities, concentrations in wastewater of 0.0 to 2.0 pg I-TEQ/L, with an average of 0.43 pg I-TEQ/L (ND = 0; 4 of 6 samples with no detected congeners), and emission factors of 0.0 to 0.04 µg I-TEQ/t PVC , with a average of 0.0083 µg I-TEQ/t PVC.

Later, U.S. EPA requested more data from the industry, which was assessed as follows (USEPA 2004):

• For chlor-alkali, EDC, VCM, and other organic operations, concentrations in treated wastewater of 0.000104 to 110 pg I-TEQ/L, with a median of 55.2 pg I-TEQ/L; and
• For EDC, VCM, and other organic operations, concentrations of 3.12 to 174 pg I-TEQ/L, with a median of 34.3 pg I-TEQ/L; and
• PVC-only operations, one facility, 0.333 pg I-TEQ/L.

In 2007, The Vinyl Institute proposed to US EPA a voluntary program to measure PCDD/PCDF releases in wastewater discharges to surface waters from facilities manufacturing EDC by oxychlorination (Vinyl Chloride Producers 2007). In 2011, the results of the testing program were reported based on 2008/2009 data (Vinyl Chloride Producers 2011). The results demonstrated that each plant individually met the EU BREF limits. Summed across the 11 EDC/VCM plants tested, the total release to water was 0.028 g I-TEQ/day, in comparison to the value of 0.049 g I-TEQ/day that can be calculated using the EU BREF (0.1 ng TEQ/L) BAT and the wastewater release rate. This emission (0.028 g I-TEQ/day) is noted as a maximum emission, utilizing the highest flow rate for process water, a highly variable stream.

At 0.028 g TEQ/day, the total release to water s from the 11 facilities is 10.2 g TEQ/yr. These same 11 facilities reported their total release to water as 5.96 g TEQ in the 2009 Toxics Release Inventory (USEPA 2012a).

In 2007, European EDC/VCM facilities reported releases to water ranging from 0.0037 to 1.65 µg TEQ/ton oxychlorination capacity (OSPAR 2009). In 2011, Det Norske Veritas reported that all the EDC/VCM facilities that are signatories to the ECVM charter met the charter commitment of 1 µg TEQ/ton oxychlorination, with data unavailable for one plant (Det Norske Veritas 2012).

EFWater for EDC/VCM/PVC and EDC/VCM facilities:

• Class 1: an EFWater of 25 µg TEQ/t EDC for sites with oxychlorination reactors or 5 ng TEQ/L effluent waste water can be used, a factor of 10 higher than Class 2.
• Class 2: an EFWater of 2.5 µg TEQ/t EDC for sites with oxychlorination reactors is presented, or 0.5 ng TEQ/L TEQ/L effluent waste water can be used, a factor of five higher than Class 3.
• Class 3: an EFWater of 0.5 µg TEQ/ton EDC by for sites with oxychlorination reactors or 0.1 ng TEQ/L effluent waste water is presented, based on OSPAR and the ECVM Charter.

EFWater for Suspension and Dispersion/Emulsion PVC-only facilities:

• Class 1: an EFWater of 0.03 μg TEQ/t PVC or concentration of 0.01 ng/L is presented, a factor of 10 higher than Class 2.
• Class 2: an EFWater of 0.003 μg TEQ/t PVC is presented, based on the average EFWater derived by the Vinyl Institute, utilizing 1995 emissions and standalone PVC capacity, with non-detects evaluated at zero, or concentration of 0.001 ng/L (Vinyl Institute 2002).
• Class 3: an EFWater of 0.0003 μg TEQ/t PVC or concentration of 0.0001 ng/L is presented, a factor of 10 lower than Class 2.

Releases to Land

Releases to land can entail releases of residues to landfills, as well as releases to surface soils. Releases to surface soils from the EDC/VCM/PVC production chain are thought to be uncommon. However, the Vinyl Institute (2002) reported that, in the U.S. as of 1996, 6% of wastewater treatment sludge was disposed of by application to surface soils – “land farming”. USEPA (2006a) attributed the release to land of 1.45 g I-TEQ to EDC/VCM/PVC manufacturing in 2000, due to land application of wastes from one facility that ended this practice that same year. Carroll Jr. et al. (2001) reported releases to land by one U.S. EDC/VCM/PVC facility using “land farm” for disposal of wastewater treatment sludge and derived emission factors from two samples of material at of 0.054 and 0.11 µg TEQ/t EDC capacity. The average of these two emission factors is 0.08 µg TEQ/t EDC capacity.

EFLand for wastewater treatment sludge from EDC/VCM/PVC facilities: An EFLand of 0.08 µg TEQ/t EDC capacity (0.01 µg TEQ/t EDC production) based on the findings by (Carroll Jr. et al. 2001) is presented to be used only for those individual facilities that use land application for disposal of wastewater treatment sludge.

Releases to Product

PVC Resin

In 1994, the Swedish EPA reported PCDD/PCDF concentrations of 0.86 and 4.70 pg TEQ/g in two samples of suspension PVC from two Swedish PVC facilities (Swedish Environmental Protection Agency 1994). In 1996, Carroll et al. analyzed 26 resin samples and reported only two hepta-CDD and CDF congeners at concentrations above limits of quantification (detection limits were not reported and non-quantifiable concentrations were designated as “non-detects, ND”) in one sample of bottle resin (Carroll et al. 1996, 1998). In 1998, Wagenaar et al. (1998) analyzed eleven PVC resin samples, and found only hepta- and octa- PCDD/PCDF at concentrations that were above limits of detection but below limits of quantification. LOQ was not reported. Thus NA is the emission factor for PCDD/PCDF in resins for class 3 and ND for classes 1 and 2.

EFProduct for PVC resin:

• Class 1 and Class 2: ND
• Class 3: NA

VCM

For VCM, the only journal-published work remains that of European producers (Isaksen et al. 1996) who analyzed six samples of VCM. Congeners were detected at the ng/t level; however, due to that extremely dilute concentration (fg/g) the authors concluded that there are no process-generated PCDDs and PCDFs in VCM. There has been no more recent work to confirm or contradict this conclusion. Moreover, releases in VCM are relevant only for the very small fraction of VCM that is not used as monomer for PVC , so NA is presented as the EFProduct for VCM.

EFProduct for VCM:

• NA for all classes.

EDC

The following emission factors are relevant only for EDC sold outside the EDC/VCM and EDC/VCM/PVC product chains:

EFProduct for EDC produced via oxychlorination or mixed oxychlorination/direct chlorination:

• For Class 1: an EFProduct of 2.0 µg TEQ/t EDC is presented, a factor of 10 greater than Class 2.
• For Class 2: an EFProduct of 0.2 µg TEQ/t EDC is presented, based on the In-Process EDC sample reported by the Vinyl Institute.
• For Class 3: an EFProduct of 0.006 µg TEQ/t EDC is presented, based on the average of the Sales EDC samples reported by (Carroll Jr. et al. 2001).

EFProduct for EDC derived from Direct Chlorination, and only for product sold outside the VCM product chain

• For all facilities, ND.

Release to Residues

These are potentially the largest route of PCDD/PCDF release from EDC, VCM and PVC production, depending upon their treatment or destruction. Residues of greatest interest include process residues, wastewater treatment sludge, spent catalyst, and maintenance waste. Wastewater treatment sludge is reported to account for 26-33 percent of total PCDD/PCDF release from EDC/VCM/PVC production in the U.S.; spent catalyst, 19-25 percent; and maintenance waste, 12-20 percent (Vinyl Institute 2002).

Little is known about PCDD/PCDF levels in EDC/VCM/PVC process residues in part because they are commonly sent to liquid and liquid/vent thermal oxidizers or to on- or off-site incinerators. This may not always be the case. For example, heavy ends from Iranian EDC/VCM/PVC facilities have reportedly been sent to landfills (Ghaheri and Ghaheri 2007). Data compiled from European facilities show PCDD/PCDF concentrations ranging from 0.04 to 18 µg TEQ/kg in heavy ends from EDC/VCM production (Vinyl Institute 2002). In the U.S., concentrations of 6,365 µg TEQ/kg were measured in a sample of heavy ends from EDC distillation, 3.2 µg TEQ/kg in a waste sample described as heavy ends from VCM distillation, and 20 µg TEQ/kg in general process wastes (distillation residues, heavy ends, tars and reactor clean-out wastes) (Costner 1995, Stringer et al. 1995). Because of these high levels of contamination, the BAT&BEP Guidelines are clear: these streams must be destroyed to meet best practices.

Wastewater treatment solids (WWTS): A German EDC/VCM/PVC facility reported a PCDD/PCDF concentration of about 500 μg TEQ/t in WWTS (EU IPPCB 2003) which falls within the very broad range reported by the Vinyl Institute and presented in Table II.48.3, together with the respective emission factors.

The Vinyl Institute noted that PCDD/PCDF concentrations in WWTS from U.S. EDC/VCM facilities with fluidized bed oxychlorination reactors vary over a broader range and generally are higher than those from facilities with fixed bed oxychlorination reactors. In the Vinyl Institute study, ten WWTS samples were analyzed; two from PVC-only plants, four from fluidized bed oxychlorination plants and four from fixed bed oxychlorination plants (Carroll Jr. et al. 2001). The lowest concentration was found in the PVC-only WWTS. The highest concentration of a fixed bed sample was 250 μg I-TEQ/t sludge. Three of the four fluidized bed samples exceeded that value, and ranged to 12,000 μg I-TEQ/t sludge.

Accordingly, the Vinyl Institute developed upper bound emission factors based on the highest release rate per ton of sludge: for the fixed bed catalyst facilities, 0.75 µg I-TEQ/t EDC (1.5 µg I-TEQ/t EDC via oxychlorination) and for fluidized bed facilities, 4.0 µg I-TEQ/t EDC (7.9 µg I-TEQ/t EDC via oxychlorination). Most likely emission factors were also reported for fixed bed oxychlorination reactors (0.19 µg I-TEQ/t EDC) (0.37 µg I-TEQ/t EDC via oxychlorination) and fluidized bed reactors (1.96 µg I-TEQ/t EDC) (3.9 µg I-TEQ/t EDC via oxychlorination) (Table II.48.3). However, there was no clear difference between EDC/VCM and EDC/VCM/PVC facilities when the type of oxychlorination reactor - fixed bed or fluidized bed - was taken into account.

Table III.48.3– PCDD/PCDF concentrations in and emission factors for wastewater treatment sludge from U.S. EDC/VCM, EDC/VCM/PVC and PVC-only facilities in the U.S. (Vinyl Institute 2002)

Facility type	PCDD/PCDF content, μg I-TEQ/t sludge		Emission Factor, μg I-TEQ/t EDC by oxychlorination or PVC
	Range	Average (Median)	Most likely	Upper bound
PVC only	2.8-5.5	3.9 (3.9)A	0.023B	0.077B
EDC/VCM and EDC/VCM/PVC: Fixed bed oxychlorination reactors	93-12,000C	2200 (310)C	0.37	1.5
EDC/VCM and EDC/VCM/PVC Fluidized bed oxychlorination reactors			3.9	7.9

^A Two samples only
^B per ton PVC
^C Both fixed bed and fluidized bed oxychlorination reactors.

Spent catalyst: EDC/VCM/PVC and EDC/VCM production generates spent catalysts as residues from EDC production by both direct chlorination and oxychlorination processes. Of these two the oxychlorination catalyst is most significant for PCDD/PCDF (EU IPPCB 2003).

Oxychlorination: The oxychlorination process is carried out in either fluidized-bed or fixed-bed reactors using a metal catalyst, typically copper chloride:

• With fluidized-bed reactors, spent catalyst is removed continuously by the entrainment of fines that accumulate in sludge during wastewater treatment. The catalyst itself is seldom discarded.
• With fixed-bed reactors, spent catalyst is removed and collected as a discrete solid residue at the end of its useful life at a rate ranging from 10 to 200 g/t VCM (EU IPPCB 2003).

The Vinyl Institute (2002) reported PCDD/PCDF concentrations in oxychlorination catalyst ranging from 220 to 150,000 µg I-TEQ/t, I-TEQ (median 15,000, average 29,000 µg I-TEQ/t spent catalyst) and derived EFResidue ranging from 0.018 to 8.1 µg I-TEQ/t EDC.

Fixed bed catalyst sent to landfill was calculated to contain 4.7 g I-TEQ/yr and fluidized bed catalyst 0.21 g I-TEQ/yr. Dividing by EDC production from plants with fixed bed capacity (5,400,000 t/yr) gives 0.87 µg I-TEQ/t EDC from fixed bed facilities and 0.045 µg I-TEQ/t from plants with fluidized bed capacity (4,700,000 t/yr) (Vinyl Institute 2002).

Coke and spent lime: No PCDD/PCDF data were found for coke or spent lime.

Maintenance wastes: The Vinyl Institute (2002) surveyed all U.S. EDC/VCM manufacturing sites to identify the sources of process contaminated maintenance wastes and determined that the production of 11.115 million tons of EDC and 6.173 million tons of VCM was accompanied by the generation of 915.1 tons of maintenance wastes, of which, only 312 tons is sent to landfill (the rest is incinerated in normal course). This equates to a generation rate for maintenance waste of only 28 g waste/t EDC.

Waste Water Treatment Solids

EFResidue for EDC/VCM facilities for wastewater treatment solids:

• For Class 1: an EFResidue of 0.75 μg TEQ/t EDC produced via fixed bed oxychlorination or 4 μg TEQ/t EDC from sites with fluid bed oxychlorination is presented, representing the upper end of the distribution of plants in the Vinyl Institute study.
• For Class 2: an EFResidue of 0.2 μg TEQ/t EDC from sites with fixed bed oxychlorination or 2 μg TEQ/t EDC from sites with fluid bed oxychlorination is presented, representing the most likely value in the distribution of plants in the Vinyl Institute study.
• For Class 3: EFResidue of NA is accorded to facilities that incinerate wastewater treatment solids; for those that do not, 0.095 μg TEQ/t EDC from sites with fixed bed oxychlorination and 0.4 μg TEQ/t EDC from sites with fluid bed oxychlorination based on the lower values and the ratio between the two types in the Vinyl Institute (2002).

EFResidue for PVC-only facilities, wastewater treatment solids:

• For Class 1: an EFResidue of 0.095 µg TEQ/t PVC is based on the highest value derived by the Vinyl Institute (2002).
• For Class 2: an EFResidue of 0.06 µg TEQ/t PVC is based on the average of the values derived by the Vinyl Institute (2002).
• For Class 3: an EFResidue of NA is presented for facilities that incinerate waste water treatment solids or 0.005 µg TEQ/t PVC for those that do not is presented, based on the lowest value derived by the Vinyl Institute (2002).

EFResidue for EDC/VCM facilities with fixed-bed oxychlorination, as spent catalyst:

• Class 1: an EFResidue of 8 µg I-TEQ/t EDC based on the highest value derived by the Vinyl Institute.
• Class 2: an EFResidue of 0.85 µg I-TEQ/t EDC is the average of the values derived by the Vinyl Institute.
• Class 3: an EFResidue of NA is presented for facilities that thermally treat or incinerate spent catalyst; for those that do not, 0.02 µg I-TEQ/t EDC is presented, which is the lowest of the values derived by the Vinyl Institute.

Maintenance Waste

EFResidue for maintenance waste sent to landfill from EDC/VCM facilities:

• In lieu of an EFResidue, a surrogate of 15,000 µg/t waste, the median of the values reported for spent catalyst, as reported by the Vinyl Institute can be used in conjunction with actual waste generation rates or a surrogate of 28 g waste/t EDC capacity to estimate PCDD/PCDF release in maintenance waste.

Table III.48.4 Emission factors for EDC/VCM and EDC/VCM/PVC production

Class	µg TEQ/t VCM	µg TEQ/t EDC from Sites with Oxychlorination ReactorsA				µg TEQ/t Product
			WWTSB
	Liquid and Liquid/Vent Combustors	Treated Waste Water	Fixed Bed	Fluidized Bed	Spent Catalyst, Fixed Bed Oxychlorination	EDCC	VCM	PVC
1	5	25	0.75	4.0	8	2	NA	ND
2	0.5	2.5	0.2	2.0	0.85	0.2	NA	ND
3	0.05	0.5	0.095	0.4	0.02D	0.006	NA	NA

^A Assumes a balanced or nearly balanced direct chlorination-oxychlorination process. Sites operating direct chlorination only are ND.
^B Wastewater Treatment Solids.
^C Derived from oxychlorination or mixed direct chlorination and oxychlorination and sold for applications other than vinyl chloride. EDC derived from direct chlorination alone is NA.
^D Zero if combusted to remove organics.

Table III.48.5 Emission factors for PVC-only facilities

Class	µg TEQ/t PVC Product
	Air	Water	Wastewater Treatment Solids	Product
1	1	0.03	0.095	ND
2	0.1	0.003	0.06	ND
3	0.02	0.0003	0.005	NA

Chlorobenzenes

In evaluating production of chlorobenzenes by direct chlorination of benzene, Liu et al. (2004) determined PCDD/PCDF concentrations in collected six samples from the production process, as shown in Table III.48.6:

Table III.48.6 Concentrations of PCDD/PCDF in chlorobenzenes, intermediate and residue (Liu et al. 2004)

Sample	PCDD/PCDF (ng TEQ/kg)
Intermediate: mixture of 1,2- and 1,4-dichlorobenzene after distillation and separation from monochlorobenzene	620
Intermediate: mixture of di- and trichlorobenzenes	1850
Residue left from purification of 1,2,4-trichlorobenzene	3370
1,4-Dichlorobenzene: after distillation and crystallization (98.1%)	39
1,2-Dichlorobenzene: after distillation and crystallization	ND
Purified 1,2,4-trichlorobenzene	ND

Although no PCDD/PCDF were detected in 1,4-dichlorobenzene, total PCBs were present at a concentration of 1,797 ng/g. No information is available on releases to air or water.

Based on this study by Liu et al. (2004), the following emission factors are presented:

EFProduct: 39 µg TEQ/t for 1,4-dichlorobenzene.

Pentachlorophenol (PCP) and, Sodium Pentachlorophenate (PCP-Na)

Based on the analysis of PCP-Na, the Republic of China (2007) reported a PCDD/PCDF release of 25 g I-TEQ in 2,000 t of PCP-Na product, which indicates an EFProduct of 12,500 µg I-TEQ/t. In 2010, Tondeur et al. (2010) determined an average PCDD/PCDF content of 634 mg TEQ/kg in 20 samples of PCP from a U.S. production facility. In 1994, Bao et al. (1994) detected PCDD/PCDF levels ranging from 612-924 mg I-TEQ/g in thermolysis waste from PCP and PCP-Na production at a Chinese facility.

Data are inadequate to support the derivation of EFAir and EFWater.

EFProduct: For PCP, an EFProduct of 634,000 µg TEQ/t is presented, based on Tondeur et al. (2010).

EFProduct: For PCP-Na, an EFProduct of 12,500 µg TEQ/t is presented, based on data from the Republic of China (2007).

Further, Bao et al. (1994) calculate an EFResidue of 768,000,000 µg TEQ/t for PCP and PCP-Na production via alkaline hydrolysis of HCH, a process which is no longer in use.

2,4,5-Trichlorophenoxyacetic Acid (2,4,5-T) and 2,4,6-Trichlorophenol

The highest concentration of 2,3,7,8-TCDD reported in a 2,4,5-T product from Germany was 7,000 ng I-TEQ/kg. In one sample of 2,4,6-trichlorophenol, PCDD/PCDF were was found at 680,000 ng I-TEQ/kg (NATO/CCMS 1992).

Due to lack of data, emission factors have been derived only for releases in products:

EFProduct: For 2,4,5-trichlorophenoxyacetic acid, an EFProduct of 7,000 µg TEQ/t is presented (EC 1999b, NATO/CCMS 1992).

EFProduct: For 2,4,6-trichlorophenol, an EFProduct of 700 µg TEQ/t is presented (NATO/CCMS 1992, Fiedler 2003).

Chloronitrofen, Chlornitrofen or 2,4,6-Trichlorophenyl-4-nitrophenylether (CNP)

Masunaga (1999) reported a PCDD/PCDF concentration in samples of chloronitrofen of 7,100 ng I-TEQ/g active ingredient in a batch produced in 1978; 11,300 ng I-TEQ/g in a 1983 batch; 62 ng I-TEQ/g in a 1986 batch; 4.1 ng I-TEQ/g in a 1987 batch; and 4.9 ng I-TEQ/g in a 1989 batch. There is no further information on the synthesis and what might have led to decreased contamination in the more recent batches. Due to lack of information, only emission factors for release in products are presented.

EFProduct: For CNP produced using old technologies, an EFProduct of 9,200,000 µg TEQ/ton is presented, based on the concentrations reported for the two oldest samples analyzed by Masunaga (1999).

EFProduct: For CNP produced using new technologies, an EFProduct of 4,500 µg TEQ/ton is presented, based on the concentrations reported for the two most recently produced samples that were analyzed by Masunaga (1999).

Pentachloronitrobenzene (PCNB) (Quintozene)

PCB emission factors for PCNB production are shown in table III.48.7 below.

Table III.48.7 DL-PCB emission factors for source category 7d Pentachloronitrobenzene (PCNB, Quintozene)

7d	Pentachloronitrobenzene (PCNB, Quintozene)	Emission Factors (µg TEQ/t PCNB)
Classification		Air	Water	Land	Product	Residue
1	Low-end technologies	ND	ND	ND	2,400	ND
2	Mid-range technologies	ND	ND	ND	1,500	ND
3	High-end technologies	ND	ND	ND	680	ND

PCDD/PCDF concentrations ranging from 2.5 to 5.6 ng TEQ/g of PCNB, with a mean of 3.9 ng TEQ/g, were measured in three PCNB formulations in Australia (Holt et al. 2010). Use of PCNB for agricultural purposes in Australia was reported to be accompanied by the release to land of an estimated 27 g TEQ/year and ranking it as Australia’s sixth largest source (Holt et al. 2010). Concentrations of PCDD/PCDF and dioxin-like PCBs that were determined in five Chinese PCNB product samples are shown in Table III.48.8 (Huang et al. 2012). Hexachlorobenzene has also been identified as a contaminant in PCNB (USEPA 2010a).

Table III.48.8 Concentrations of PCDD/PCDF and DL-PCBs in two “raw pesticide” samples (R-1 and R-2) and samples of three PCNB formulations (F-1, F-2, and F-3)

		PCDD/PCDF, ng TEQ/g*	PCB, ng TEQ/g*
R-1	95% PCNB	0.26	0.68
R-2	95% PCNB	0.96	2.4
F-1	40% PCNB	0.18	2.0
F-2	30% PCNB, 15% Bromothalonil	0.73	0.82
F-3	20% PCNB, 20% Thiram	0.38	1.6

*Each value is the mean of lower bound values obtained from the analysis of duplicate samples.

In addition, the Ministry of Agriculture, Forestry and Fisheries (MAFF) of Japan reported PCDD/PCDF at 3.7 ng TEQ/g of PCNB and dioxin-like PCBs at 0.86 ng TEQ/g PCNB in Japanese PCNB formulations (MAFF 2002).

The following emission factors are derived:

EFProduct for PCDD/PCDF: 260 µg TEQ/t for PCNB produced by Class 3 facilities, based on the lowest value reported in the three studies, which was obtained by Huang et al. (2012); 2,600 µg TEQ/t for Class 2 facilities, based on the mean of the values reported in the three studies; and 5,600 µg TEQ/t for Class 1, based on the high value obtained by Holt et al. (2010). For agricultural uses of PCNB, each of these emission factors can also be used as EFLand.

EFProduct for DL-PCBs: 680 µg TEQ/t for PCNB produced by Class 3 facilities, based on the lowest value reported by Huang et al. (2012); 1,500 µg TEQ/t for Class 2 facilities, based on the mean of the high and low values reported by Huang et al. (2012); and 2,400 µg TEQ/t for Class 1, based on the high value reported by Huang et al. (2012). For agricultural uses of PCNB, each of these emission factors can also be used as EFLand.

2,4-Dichlorophenoxyacetic Acid and its Derivatives

PCDD/PCDF were detected in 2,4-D as long ago as 40 years (Woolson et al. 1972) and as recently as 2012 (A. Grochowalski, personal communication, 4 October 2012, Gullett et al. 2012). In the U.S., agricultural use of 2,4-D was associated with the release to land of 28.9 g TEQ/year in 1995. Lack of information prevented the preparation of a more recent estimate (USEPA 2006a).

As shown in Table III.48.9, PCDD/PCDF concentrations in 2,4-D and its derivatives have ranged from non-detect to 6,800 ng/kg of 2,3,7,8-TCDD, the most potent of the PCDD/PCDFs. As might be expected, PCDD/PCDF concentrations have generally diminished over the 40 years for which data are available. However, recent studies show that significant concentrations of PCDD/PCDF continue to be found in 2,4-D and its derivatives. For example, during 2008-2012, PCDD/PCDF were measured in samples of 2,4-D, 2,4-D esters and 2,4-dichlorophenol feedstock from a production facility in Eastern Europe. For 31 samples of 2,4-D produced primarily for export, a mean concentration of 102.7 pg TEQ/g was determined, while 21 samples of 2,4-D that had an unknown fate had a mean concentration of 5,688 pg TEQ/g. For both sample sets, most of the 2,4-D was produced on-site. However, beginning in 2010, crude 2,4-D was also imported from Asia for reprocessing. In addition, 51 samples of a variety of 2,4-D esters had a mean PCDD/PCDF concentration of 661.1 pg TEQ/g , and 17 samples of 2,4-dichlorophenol used as primary feedstock for production of 2,4-D and 2,4-D esters at this facility had a mean concentration of 116,365 pg TEQ/g (A. Grochowalski, personal communication, 4 October 2012).

Table III.48.9 PCDD/PCDF concentrations in 2,4-Dichlorophenoxyacetic acid (2,4-D) and its derivatives

	ng TEQ/kg	Point of production or purchase	Reference
2,4-D	<1,000 of 2,3,7,8-TCDD	Canada	Cochrane et al. (1981)
	6,800 of 2,3,7,8-TCDD	Germany	Hagenmaier (1986)
	4,800	Germany	Wilken et al. (1992)
	0 – 16	Japan	Masunaga et al. (2001)
	0.12-1.8	Australia	Holt et al. (2010)
	160-180
	~300* (est.)***	U.S.	Gullett et al. (2012)
	5.43–405	Eastern Europe	Grochowalski (pers. comm. 2012)
	1,080-18,500
Mixture of 2,4-D esters	661.1	Eastern Europe	Grochowalski pers. comm. (2012)
2,4-D dimethylamine salt	4,110	Germany	Wilken et al. (1992)
	160	Russia	Schecter et al. (1993)
2,4-D dimethylamine, 46.9%	8.7*	U.S.	Huwe et al. (2003)
2,4-D isooctyl ester, 61.7%	731*	U.S.	Huwe et al. (2003)
2,4-D isooctyl ester, 66.2%	2,627*	U.S.	Huwe et al. (2003)
2,4-D isooctyl ester, 67.2%	27.7*	U.S.	Huwe et al. (2003)
2,4-D isooctyl ester, 88.8%	1,379*	U.S.	Huwe et al. (2003)
Technical 2,4-D and 2,4-D Ester Herbicides	700**	U.S.	USEPA (2005a)
2,4-D Herbicides purchased in U.S.	1.9*, 2.4*, 82.3*	U.S.	Schecter (1998)
2,4-D Herbicides purchased in Palestine and Israel	96.4*, 828*	Palestine and Israel	Schecter (1998)
2,4-D Herbicide, Chimprom, Ufa, Russia	142*	Russia	Schecter (1998)

* ng TEQ/kg of ready-for-use product (active ingredient plus adjuvants).
** OCDD and OCDF were not assayed in 8 samples submitted by U.S. producers.
*** In as-purchased 2,4-D, Gullet et al. (2012) determined a ∑TCDF concentration of about 10 ng/g and noted that this value is consistent with Holt et al. (2010) and Masunaga et al. (2001). In two samples of as-purchased 2,4-D, Holt et al. (2010) reported mean concentrations of ∑TCDF of 0.135 and 2.6 ng/g that were associated with PCDD/PCDF lower-bound values of 0.0004355 and 0.0775 ng TEQ/g, respectively. Based on the ratios of ∑TCDF to total TEQ for the samples from Holt et al. (2010), total PCDD/PCDF of 0.30 ng TEQ/g can be estimated for this sample of as-purchased 2,4-D.

EFProduct for PCDD/PCDF: 0.12 µg TEQ/t for 2,4-D and its derivatives that are produced by Class 3 facilities, based on the lowest value reported by Holt et al. (2010); 170 µg TEQ/t for 2,4-D and its derivatives that are produced by Class 2 facilities, based on the mean of the higher-range values reported by Holt et al. (2010); 5,688 µg TEQ/t for 2,4-D and its derivatives that are produced by Class 1 facilities, based on the mean value reported by Grochowalski (2012).

Chlorinated Paraffins (CPs)

PCB and HCB emission factors for the production of CP are shown in tables III.48.10 and III 48.11 below:

Table III.48.10 PCB emission factors for source category 7d Chlorinated Paraffins

7d		Chlorinated Paraffins	Emission Factors (mg/t product)
Classification			Air	Water	Land	Product	Residue
1	Low-end production technologies		ND	ND	ND	210,000	ND
2	Mid-range production technologies		ND	ND	ND	165,000	ND
3	High-end production technologies		ND	ND	ND	40	ND

Table III.48.11 HCB emission factors for source category 7d Chlorinated Paraffins

7d		Chlorinated Paraffins	Emission Factors (mg/t product)
Classification			Air	Water	Land	Product	Residue
1	Low-end production technologies		ND	ND	ND	8,900	ND
2	Mid-range production technologies		ND	ND	ND	7,500	ND
3	High-end production technologies		ND	ND	ND	7	ND

Three samples of technical grade CPs from an East Asian country (with legislation in place limiting PCDD/PCDF in chemicals and products) were analyzed in duplicate, yielding lower-bound PCDD/PCDF concentrations that ranged from 132.9 to 545.4 pg TEQ/g and, mean concentrations for the three samples of 140.6, 228.6 and 490.8 pg TEQ/g (Takasuga et al. 2012).

In the same study, six samples of CPs produced from another East Asian country were found to have total PCB concentrations ranging from 140,000 to 210,000 ng/g, with a mean of 165,000 ng/g, as well as HCB concentrations ranging from 6,100 to 8,900 ng/g, with a mean of 7,733 ng/g. Considerable lower levels of PCB (40 ng/g) and HCB (7 ng/g) were detected in one sample from the East Asian country with legislation in place limiting PCDD/PCDF in chemicals and products.

In analyzing polyurethane foam and rubber materials used in a high volume air sampler pump, Takasuga et al. (2012) also detected high levels of PCBs and HCB in both of these materials and determined the main source of these contaminants as long-chain CPs that were produced in China and used in the rubber at levels of 2-6% as a flame retardant. Subsequent analysis of the technical CP used in the rubber found concentrations of total PCBs of 140-190 ppm and 6.8-8.7 ppm HCB. Polychlorinated naphthalenes (PCNs) and pentachlorobenzene (PCBz) were also detected in the rubber but were not quantified.

The following emission factors are derived:

EFProduct for PCDD/PCDF: 140 µg TEQ/t for CPs produced by Class 3 facilities, based on the lowest value reported by Takasuga et al. (2012); and 500 µg TEQ/t for Class 2, based on the mean of the values obtained with the most contaminated of the three samples.

EFProduct for PCBs: 40 mg/t for CPs produced by Class 3 facilities, based on the lowest value reported by Takasuga et al. (2012); 156,000 mg/t for Class 2 facilities, based on the mean of the values for five samples analyzed by Takasuga et al. (2012); and 210,000 mg/t for Class 1, based on the most contaminated of the samples analysed.

EFProduct for HCB: 7 mg/t for CPs produced by Class 3 facilities, based on the lowest value reported by Takasuga et al. (2012); 7,500 mg/t for Class 2 facilities, based on the mean of the values for the five samples analyzed by Takasuga et al. (2012); and 8,900 mg/t for Class 1, based on the most contaminated of the samples analyzed.

p-Chloranil (2,3,5,6-tetrachloro-2,5-cyclohexadiene-1,2-dione)

Ni et al. (2005) analyzed chloranil produced by two Chinese facilities and found PCDD/PCDF concentrations of 13 and 126 ng I-TEQ/kg. The disparity in their results was attributed to the use of different production methods, which were not described.

Liu et al. (2012) determined PCDD/PCDF concentrations in choranil samples from three other Chinese facilities, each of which used a production process involving the chlorination of hydroquinone. However, at each facility, the chloranil was purified to a different extent because of different intended uses. Chloranil produced for use as an intermediate in pharmaceutical products was most stringently purified and had a PCDD/PCDF concentration of 163 pg I-TEQ/g. PCDD/PCDF was found at 1,540,200 pg I-TEQ/g in the chloranil intended for use as an intermediate for dyes and pesticides. A PCDD/PCDF concentration of 26,368 pg I-TEQ/g was measured in moderate quality chloranil. Liu et al. (2012) proposed the average of these three values, 522,000 µg I-TEQ/t, as EFProduct for chloranil. Total PCB levels in the three chloranil samples ranged from 1,179.4 to12,413.7 pg/g (1.9-3.3 pg WHO-TEQ/g); pentachlorobenzene (PeCBz) ranged from 12.1 to 31.8 ng/g; and HCB from 3.8 to 391.5 ng/g.

No information is available to support the derivation of emission factors for releases to air, water, land and residue for chloranil. However, it is apparent that releases to residues will be greater with increasingly stringent purification.

Based on the findings of Ni et al. (2005) and Liu et al. (2012), the following emission factors are presented for p-chloranil production via chlorination of hydroquinone:

EFProduct: For minimal purification of chloranil, an EFProduct of 1,500,000 µg TEQ/t is presented.

EFProduct: For moderate purification of chloranil, an EFProduct of 26,000 µg TEQ/t is presented.

EFProduct: For high purification of chloranil, an EFProduct of 150 µg TEQ/t is presented.

An EFProduct of 400,000 µg TEG/t product is presented for p-chloranil production via direct chlorination of phenol (Fiedler 1998, 1996, BUA 1992).

Phthalocyanine pigments and dyes

HCB emission factors for the production of phthalocyanine pigments and dyes are included in Table III.48.12 below:

Table III.48.12 HCB emission factors for source category 7d – Phthalocyanine Pigments and Dyes Production

7d	Phthalocyanine-derived pigments and dyes	Emission Factors (g/t product)
Classification		Air	Water	Land	Product	Residue
1	Pigment Green 7 (CAS 1328-53-6)	ND	ND	ND	200	ND
2	Pigment Green 7 (BAT)	ND	ND	ND	10	ND
3	Pigment Green 36 (CAS 14302-13-7)	ND	ND	ND	10	ND
4	Pigment Green 36 (BAT)	ND	ND	ND	1	ND

As for PCDD/PCDF, these were detected in phthalocyanine copper and phthalocyanine green at concentrations of 73.28 and 1379.55 ng I-TEQ/kg, respectively, by Ni et al. (2005). No information was available to support the derivation of emission factors for releases to air, water, land and residue.Based on the values reported by Ni et al. (2005), the following emission factors for PCDD/PCDF release to products for the two phthalocyanine-based pigments and dyes are presented:

EFProduct for phthalocyanine copper: 0.07 µg TEQ/kg; and

EFProduct for phthalocyanine green: 1.4 µg TEQ/kg.

Tetrachlorophthalic acid (TCPA) and related pigments

TCPA is the primary feedstock for the production of a range of pigments. While no PCDD/PCDF data are available for TCPA, unintentional HCB concentrations as high as 3,000 ppm have been detected (Government of Japan 2006). However, BAT levels of less than 200 ppm and below 50 can be achieved by modification of production processes and recrystallization (Government of Japan 2006; Table III.48.13). With TCPA use, unintentional HCB is transferred to pigments and residues (Government of Japan 2007, 2006). TCPA-derived pigments include e.g. Pigment Yellow 110 (CAS 5590-18-1), Pigment Yellow 138 (CAS 30125-47-4), Solvent Red 135 and Solvent Red 162 (CAS 20749-68-2 and 71902-17-5).

Table III.48.13 HCB emission factors for source category 7d TCPA and related pigments

7d	TCPA and related pigments	Emission factors (g/t product)
Classification		Air	Water	Land	Product	Residue
1	Tetrachlorophthalic acid (CAS 632-58-6)	ND	ND	ND	2000	ND
2	Tetrachlorophthalic acid (BAT)	ND	ND	ND	200	500
3	Solvent Red 135 (CAS 20749-68-2)	ND	ND	ND	200	ND
4	Solvent Red 135 (BAT)	ND	ND	ND	10	ND
5	Pigments Yellow 110 (CAS 5590-18-1)& 138 (CAS 30125-47-4)	ND	ND	ND	200	ND
6	Pigment Green 7 (CAS 1328-53-6)	ND	ND	ND	200	ND
7	Pigment Green 7 (BAT)	ND	ND	ND	10	ND
8	Pigment Green 36 (CAS 14302-13-7)	ND	ND	ND	10	ND
9	Pigment Green 36 (BAT)	ND	ND	ND	1	ND

These emission factors are associated with a medium level of confidence, as they are based on a low data range; they are not based on expert judgment, but are derived from a limited geographical coverage.

Dioxazine dyes

Three dioxazine-based dyes that were produced using PCDD/PCDF-contaminated chloranil were analyzed by Williams et al. (1992) and found PCDD/PCDF concentrations as follows:

• Blue 106 (three samples): 19.5, 30.2 and 56.4 µg TEQ/kg;
• Blue 108: 0.1 µg TEQ/kg; and
• Violet 23 (Carbazole violet) (six samples): 1.4, 2.7, 2.7, 12.7, 16.0, and 18.9 µg TEQ/kg.

No data were available to support the derivation of emission factors for releases to air, water, land, and residue from the production of these dioxazine-based dyes.

Based on the values determined by Williams et al. (1992), emission factors for releases to product for these three dioxazine-based dyes are presented:

EFProduct for Blue 106: 35 µg TEQ/kg;

EFProduct for Blue 108: 0.1 µg TEQ/kg; and

EFProduct for Violet 23 (Carbazole violet): 12 µg TEQ/kg.

Triclosan [(5-chloro-2-(2,4-dichlorophenoxy)phenol]

PCDD/PCDF have been detected in triclosan, sometimes at relatively high concentrations. Menoutis and Parisi (2002) determined concentrations of 2,3,7,8-TCDD and 2,3,7,8-TCDF in triclosan samples from six producers in India and China and obtained results shown in Table III.48.14.

Table III.48.14 Concentrations of 2,3,7,8-TCDD and 2,3,7,8-TCDF in Triclosan produced in India and China

Sample	Origin	2,3,7,8-TCDD (pg/g)	2,3,7,8-TCDF (pg/g)	pg I-TEQ/g
1	India	17.2	0.70	17.27
2	China	95.4	7.13	96.1
3	India	111.8	3.43	111.2
4	India	41.5	8.51	42.4
5	India	1712.0	0.43	1712
6	India	18.9	207.3	39.6

More recently, Ni et al. (2005) measured a PCDD/PCDF concentration of 5.03 ng TEQ/kg from a Chinese producer, attributing this relatively low value to the use of raw materials that were not favorable to PCDD/PCDF formation. Zheng et al. (2008) included this value along with those obtained by other researchers in reporting PCDD/PCDF concentrations of 0.8 to 5.03 ng TEQ/kg in Triclosan produced in China.

No information was available to allow derivation of emission factors for releases to air, water, land and residue.

EFProduct using low-end production technologies: 1700 µg TEQ/t, based on the highest value from Menoutis and Parisi (2002).

EFProduct using mid-level production technologies: 60 µg TEQ/t of product, based on the five lowest values from Menoutis and Parisi (2002).

EFProduct using advanced production technologies: 3 µg TEQ/t of product, based on the values reported in Zheng et al. (2008).

Titanium tetrachloride (TiCl4) and Titanium Dioxide (TiO2)

Titanium dioxide is manufactured by two processes: the chloride process and the sulfate process. Only the chloride process produces PCDD/PCDF as incidental byproducts.

The chloride process begins with the conversion of titanium-bearing ore – rutile, which is 93% to 96% TiO2, and ilmenite, which may contain between 44% and 70% TiO2 – into TiCl4. This conversion is carried out in a fluidized bed chlorinator in the presence of Cl2 at a temperature of approximately 900°C, with the addition of petroleum coke as a reductant. The volatile TiCl4, along with other volatile metal chlorides, exits the chlorinator as overhead vapor. The non-volatile metal chlorides, unreacted coke and ore solids are removed from the gas stream and from the bottom of the chlorinator. TiCl4 is separated from the gaseous product stream and purified by condensation and chemical treatment. Vent gases from the chlorinator are scrubbed using water and caustic solutions then vented to the air. The purified TiCl4 is then oxidized to produce TiO2 and Cl2 that is driven off is recycled to the chlorinator. The pure TiO2 is slurried and sent to the finishing process which includes milling, addition of inorganic and organic surface treatments, and/or spray drying of the product TiO2. The product can be sold as a packaged dry solid or water-based slurry.

Typical wastes generated by the chloride process includes wastewaters from chlorinator coke and ore solids recovery, reaction scrubbers, chemical tank storage scrubbers, product finishing operations and wastewater treatment solids decantation. Waste sands from finishing (milling) of the TiO2 product, scouring of oxidation process units, and blasting of reactor internal surfaces prior to replacement of refractory are also generated.

PCDD/PCDF were detected at concentrations of 0.010 and 0.020 pg TEQ/L in treated wastewater from two TiCl4/TiO2 production facilities (USEPA 2006c). However, source reduction efforts have dramatically reduced generation of PCDD/PCDF as reflected in the trend of TRI reports by the USEPA. By 2010 values of 0.0012 to 0.1771 μg TEQ/t of product were representative.

PCDD/PCDF concentrations in residues from TiCl4/TiO2 production facilities were reported as follows: wastewater treatment solids, 402 ng TEQ/kg; chloride solids/waste acid, 812 ng TEQ/L; filter press solids, 2,615 ng TEQ/kg (USEPA 2001). By 2010 PCDD/PCDF in solid residues had been reduced to a range of 8 to 42 μg TEQ/t of product (USEPA 2010b).

In the U.S., a TiCl4/TiO2 production facility has been identified as a possible source of PCDD/PCDF in sediments and shellfish of St. Louis Bay, Mississippi (Elston et al. 2005a, 2005b).

Emission Factors

No data were available to allow derivation of an EFAir for PCDD/PCDF releases from TiCl4/TiO2 production.

EFWater: Values range from .0012 to .1771 μg TEQ/t of product based on 2010 TRI data reported by USEPA and production data from TZ Minerals International.

EFProduct: ND.

EFResdiue: Based on 2010 TRI data reported by USEPA and production data from TZ Minerals International current solid residues range from 8 to 42 μg TEQ/t of product.

Caprolactam (2-Azacycloheptanone)

For PCDD/PCDF, an EFAir of 0.00035 μg I-TEQ/t of caprolactam and, for HCB and PCB, EFAir of 3.2 and 8.1 μg I-TEQ/t, respectively, have been derived by Iwata et al. (2008).

PCDD/PCDF concentrations as high as 680 pg I-TEQ/L were measured in untreated process wastewater from a caprolactam production facility in Japan and a concentration of 1.6 pg I-TEQ/L was found in the facility’s treated combined wastewater (Kawamoto 2002). Lee et al. (2009) reported a PCDD/PCDF concentration of 0.045 pg I-TEQ/L in treated wastewater from a facility in Taiwan and derived an EFWater of 0.936 ng I-TEQ/t of caprolactam. The findings of both studies suggest that PCDD/PCDF may also occur in wastewater treatment residues. PCDD/PCDF have also been identified in air emissions from caprolactam facilities in China (Hong and Xu 2012). However, information is not available to derive emission factors for releases to land, product and residue.

Due to the 35-fold disparity in the two values obtained for PCDD/PCDF concentrations in treated wastewater from caprolactam, no EFWater is presented. Instead, the approximate mid-point of the two values, 0.50 pg TEQ/L can be used in conjunction with wastewater discharge rates to estimate PCDD/PCDF releases to water.

EFAir: An EFAir 0.00035 μg TEQ/t of caprolactam has been derived by Iwata et al. (2008).

EFWater: In lieu of an EFWater, a PCDD/PCDF concentration of 0.50 pg TEQ/L can be used to estimate PCDD/PCDF releases to water.

An HCB air emission factor of 3.2 µg TEQ/t of caprolactam and PCB air emission factor of 8.1 µg TEQ/t have been derived by Iwata et al. (2008).