Department of the Environment

About us | Contact us | Publications

Settlements Header ImageSettlements Header ImageSettlements Header Image

Publications archive - Waste and recycling

Disclaimer

Key departmental publications, e.g. annual reports, budget papers and program guidelines are available in our online archive.

Much of the material listed on these archived web pages has been superseded, or served a particular purpose at a particular time. It may contain references to activities or policies that have no current application. Many archived documents may link to web pages that have moved or no longer exist, or may refer to other documents that are no longer available.

CMPS&F - Environment Australia
Appropriate technologies for the treatment of scheduled wastes
Review Report Number 4 - November 1997


10. MOLTEN MEDIA PROCESSES


 

10.1 Overview

A variety of processes relying on destruction of wastes within a molten inorganic media have been developed. Three such processes are reviewed as part of this study:

Each of these processes relies on the destruction of wastes within a high temperature molten material phase (the molten material provides enhanced heat transfer to the waste) and a thermal mass that provides stability to the treatment processes. In addition, in some processes the molten material may act as a catalyst, increasing the rate of thermal destruction of wastes. The specific advantages and disadvantages of each process are outlined in the following sections. However, in general each of the molten media processes are expected to be relatively robust (in part due to the thermal mass of the system) and relatively indiscriminate in their destruction of wastes (ie. will destroy scheduled and non-scheduled components of the wastes). This can be important in the case of wastes such as mixed pesticides which may contain organic and inorganic pesticides. In this case the organic components would be destroyed thermally, whereas the inorganics may be found in the slag or other solid waste streams.

10.2 Molten Metal



10.2.1 Technology Description

The molten metal process has been proposed for use in Australia by Molten Metal Technology Inc. (MMT), and has been supported by Du Pont. The process uses a normal steel converter with molten iron or slag to heat a waste. Molten iron, which is the heat source for the process, has a high heat capacity, that is, it heats relatively slowly, retains heat, and cools slowly. The following description of the process has been adapted from information contained in the Report from the Independent Panel on Intractable Wastes (1992) and further information on the process obtained from MMT (1995).

The molten metal process is called "Catalytic Extraction Processing" (CEP), although it is not primarily a catalytic process or an extraction process, rather it may be described as a "fluid-phase, high temperature combustor". It uses a molten iron bath at a temperature of about 1650oC to dissociate wastes into their atomic constituents. The waste is injected with oxygen into a reactor containing the molten metal. The waste is bottom fed through tuyeres (or nozzles) and must therefore be either gas, liquid or finely divided solids. The process is shown schematically in Figure 10.1.

The bath is initially raised to the required temperature by electrical energy following conventional smelting practice, and the temperature can be maintained by the addition of organic waste and oxygen. Three products can be recovered depending on the waste feed: metals; gases such as chlorine, fluorine, hydrogen, and oxygen; and a ceramic slag phase that may be used as an aggregate or may require disposal by landfilling.

Figure 10.1
Schematic Diagram of the Molten Metal Process
(Independent Panel on Intractable Wastes, 1992)
Figure 10.1


In principle, the process can operate without discharge of gases to the atmosphere; carbon monoxide and hydrogen produced in the process can be used as a fuel or perhaps for synthesis. In practice, however, it is probable that the product gas will be combusted as an energy source for the process and excess gas flared. Because of this significant volumes of off-gases from the process can be expected to require treatment and conventional gas cleaning systems will be required involving considerable capital cost. According to MMT, a 5 tonne unit would produce 8 cubic metres of gas per second which is somewhat higher than expected (Independent Panel on Intractable Waste, 1992).

The nature of the gases released are claimed to be more predictable and "friendlier" than with a high temperature incinerator, for which the feed conditions are much more critical. With high chlorine content wastes containing alkali metals, consideration would have to be given to the production of a volatile sub-micron fume: such fume can be difficult to remove to the levels required to avoid a visible emission. For the waste to react completely, the rate of addition of the waste to the melt must be controlled.

Other emissions include metals, a ceramic slag and gases. The impact of acidic and basic compounds that are formed must be considered. Ferric chloride is formed when molten iron is used and the proponents suggest that nickel could be used to avoid the formation of ferric chloride, although it is not clear that this would confer an advantage (and may present difficulties with the formation of nickel carbonyl) (Independent Panel on Intractable Waste, 1992).

The process operates under reducing conditions and as such is not conducive to the formation of dioxins. In addition, MMT have designed a rapid quench system to avoid formation of PCDDs/PCDFs. Test work reported by MMT (1995) indicates that PCDD/PCDF concentrations were not detected to the German regulatory limit of 0.1 ng/m3 (TE).

MMT has used data from demonstrations and technology development programs to develop detailed thermodynamic models and engineering designs of the CEP system. The optimal design of the reactor (especially the height-to-diameter ratio, which affects the residence time and circulation rate in the bath) depends on the nature of the waste to be treated.

10.2.2 Performance

Information provided by MMT on the CEP system indicates >99.9999% destruction removal efficiency for chlorinated waste, no products of incomplete combustion detected, NOx < 100 ppm, SOx < 100 ppm, and TCLP results for the solid residue below regulatory levels.

10.2.3 Considerations in the Application of the Technology

The technology is applicable to a range of chlorinated wastes which are in reasonably homogeneous phases. The process is able to treat wastes containing iron and was developed to treat tyres and complete transformers. Wastes comprising predominantly inert material such as soil cannot be treated.

Although the molten metal process is a high temperature destruction process, the use of oxygen instead of air reduces the scale of the plant significantly, especially the gas cleaning plant and its associated stack. In addition, the process offers the potential to recover some constituents of the waste, such as metals. As such, it differs from high-temperature incineration sufficiently to make it more acceptable to the community.

The liquid-phase combustion process is in theory, capable of more intimate and complete contact of waste with the heat source. The process uses techniques that were developed for iron smelting, however management practices that recognise the priorities of waste destruction over smelting are required.

The 5 tonne reactor system is relocatable.

In February, 1992 ICI received the following quotes from MMT for processing HCB wastes (MMT made these quotes available to the Independent Panel on Intractable Waste, 1992):

10.2.4 Experience and Availability in Australia

The MMT system has been proposed for use in treating the HCB wastes held by ICI at Botany. Initial quotations were sought by ICI in 1992, however promotion of the MMT process in Australia in recent years has been limited. At this stage we are not aware of any specific plans to implement the MMT system in Australia in the foreseeable future. MMT are represented locally.

The MMT system has been supported by the US Department of Energy. Molten Metal has built a commercial-scale research and development unit in Fall River, Massachusetts. This facility houses four bench-scale units, four pilot-scale units and a commercial demonstration and prototype unit (MMT, 1995).

MMT has signed a letter of agreement to build a CEP facility to process and recycle biosolid wastewater treatment plant wastes from Hoechst Celanese's Gulf Coast Plant sites in Texas. The estimated plant cost is $15 - $25 million (MMT, 1995).

10.2.5 Summary

(a) Proponents

Molten Metal Technology Inc.

(b) Wastes Applicable

All waste types with the exception of soil, sludges and other solids with high mineral content. Electrical equipment can be destroyed.

(c) Contaminants Applicable

All scheduled compounds.

(d) Status

A commercial scale research and development unit has been built in the US.

(e) Timing for Commercialisation in Australia

This cannot be accurately determined at this point in time. However, full scale commercial operation in Australia is unlikely to occur for at least 3 to 5 years. Timing for implementation in Australia would be dependent on the a major waste treatment project proceeding (and the MMT system being selected for use), of which there are likely relatively few.

(f) Cost (example only)

Capital cost of $6 million and an operating cost of $370/tonne for a 10 tonne/hr unit (Independent Panel on Intractable Wastes, 1992).

Projected capital cost of US$25 - 35 million for 30000 tonne/year for Clean Harbours centralised facility (MMT) 1995.

(g) Safety/Environmental Risk

Off-gas volumes should be small compared to comparable incineration systems given indirect heating and oxygen are used although other information indicates significant gas volumes may be generated. The process produces a gas which is potentially explosive if mixed with air. Appropriate safety measures are required to avoid this. The process is able to achieve gas emissions with low concentrations of residual contaminants.

(h) Non-technical Impediments

Although this information was not provided it is likely that this process will be considered in a similar way to other intensive thermochemical processes. However, there is also the possibility that it may be viewed as a combustion process.

(i) Preferred Mode of Implementation

Relocatable system, or on-line to treat process wastes.

(j) Limitations

Significant volumes of off-gases from the process can be expected to require treatment using conventional gas cleaning systems which involve considerable capital cost. This process is applicable to wastes which are in reasonably homogeneous phases and as such, wastes comprising predominantly inert material such as soil cannot be treated.

10.3 Molten Slag



10.3.1 Technology Description

The molten slag process was initially developed for processing sludge and steel works dusts (Bradhurst, 1994). It is able to reprocess and utilise wastes such as sewage sludges, grit and screenings from sewage works, industrial sludges and other carbonaceous contaminated wastes, particularly those containing heavy metals, plastics and other hydrocarbons.

In this process the waste to be treated is blended with steelworks dust and fluxing agents, extracted, dried with heat from the furnace off-gases and fed into the foaming slag layer which forms at the top of the molten iron in an electric arc smelter at a temperature of around 1500oC. The waste sinks into the slag phase, metal oxides are reduced to metals and all organic materials return to their basic elements.

Inorganics move into the slag layer while iron and most heavy transition metals dissolve into the molten metal. Volatile metals like zinc, lead and cadmium fume off through the slag and are captured in the off-gas furnace systems.

Organic constituents are consumed yielding water vapour and carbon monoxide which converts to carbon dioxide. Compared with ordinary incineration systems, the molten slag process requires a much smaller furnace and produces much less emission gas. The molten iron bath maintains a stable temperature when large volumes of wastes are to be treated and there is no open flame to control. The energy transfer from the molten metal phase to the waste material is much more efficient than in high temperature incinerators, where energy transfer occurs between the gas phase (flame) and the waste material.

The process involves basic metal foundry industry practice. Excisting furnaces can be modified and used to implement the technology.

10.3.2 Performance

The destruction efficiency of the process needs to be confirmed, as the chlorinated organics may be volatilised and not be retained by the bath (eg. the lower temperature molten tin bath used in the Eco Logic process is designed to volatilise waste materials for subsequent treatment). The molten slag process may be limited in application to chlorinated organic wastes because of this. It is probably not cost effective to use the molten slag process to treat contaminated soils which have low organic content and low contaminant concentrations due to the large quantity of slag formed and the large energy requirement.

10.3.3 Considerations in the Application of the Technology

A trial unit has been constructed and operated in NSW. Test work has focused on wastes with a low chlorine content, although scheduled wastes have been trialed. Negotiations aimed at commercialisation of the molten slag process are currently under way (Bradhurst, 1995; Bradhurst, 1996). However, no update on progress was provided for the fourth review report. The object of these negotiations is to construct a commercial scale plant in the Illawarra region.

It is concluded that the molten slag process may have limited application for the treatment of scheduled wastes and further work is required before the process can be considered for commercial application in this field.

10.3.4 Summary

(a) Proponents (in Australia)

Illawarra Technology Corporation (Wollongong)

(b) Wastes Applicable

Organic liquids (high and low volatility) and other materials with a significant organic content. Slag formation and energy consumption is likely to make processing of significant quantities of soil uneconomic.

(c) Contaminants Applicable

Focus has been on wastes with low or no chlorine content (eg. sewage sludge). However, some scheduled wastes (eg PCBs) have been trialed. Generally applicable to most scheduled compounds although consideration should be given to the potential volatilisation with reduced destruction efficiency of materials added to the slag.

(d) Status

A trial unit has been constructed and operated in NSW. Opportunities to establish a commercial facility are being sought.

(e) Timing for Commercialisation in Australia

Uncertain.

(f) Cost (example only)

Treatment of scheduled wastes likely to be cost competitive in conjunction with the treatment of other wastes for which the process will be primarily operating, ie. sewage sludge and steel works dust.

(g) Safety/Environmental Risk

High thermal mass of the molten slag and iron ensure even heat transfer and destruction of wastes even in the case of power failure. Reformation of dioxins and other chlorinated organic materials cannot be discounted. Off-gas flow rates are much smaller than for a high temperature incinerator of the same capacity, resulting in easier control. Volatilisation of some chlorinated organic compounds without treatment is possible.

(h) Non-technical Impediments

There is some reluctance on the part of the proponents to treat scheduled wastes in the first instance, although this needs to be confirmed. Current activities are focused on the establishment of a commercial system for the disposal of sewage sludge and similar wastes.

(i) Preferred Mode of Implementation

Centralised system, established in conjunction with a steel works.

(j) Limitations

Possible limitations on destruction efficiency. Soils which have a low organic content and low contaminant concentrations are unlikely to be processed economically at present. Possible problems associated with corrosion when high chlorine content wastes are treated.

10.4 Molten Salt

10.4.1 Technology Description

The Molten Salt Oxidation (MSO) process utilises a sparged liquid bed of alkaline salt contained in a refractory lined metal reaction vessel (Schnittgrund, 1994 and Stewart, 1993). A simplified sketch of the MSO concept is shown in Figure 10.2. The process is based on the catalytic action of alkaline molten salts for the oxidation of organic materials. The molten salt bed acts as an efficient heat transfer and reaction media. Heat is liberated by the oxidation reaction sufficient to maintain the molten salt bed at a temperature of 900 to 1000oC.

Sparging of the bed typically involves air injected into the bottom of the unit from side penetrating nozzles. The same nozzles are also used to inject both make-up salt as well as the waste materials being processed. If the materials to be processed are solids, then generally the particle size must first be reduced to permit pneumatic conveying. Liquids are injected using a commercial oil gun inserted into an air nozzle to within several centimetres of the inside wall. Gases can be injected directly into the air stream before it enters the vessel (partial premixing with the sparge air) or they can also be injected using the oil gun concept (with or without premixing).

The reaction product gases pass through the salt bed and are discharged out of a top or a side port nozzle in the top of the vessel. The upper region of the vessel is designed to permit the efficient disengagement of the liquid salt from the gases. A melt withdrawal nozzle is provided at the top of the expanded salt bed to permit continuous injection of fresh sodium carbonate and tapping of the spent salt. The melt overflow can be collected in drums or it can be quenched to form an aqueous brine.

Control of the bed composition is effected by continuous injection of sodium carbonate and is based on the analysis of the feed and on periodic analysis of the melt overflow for ash and carbonate. The control has the objective of avoiding excess sodium chloride build up in the bed, as this will affect the freezing point of the salt mixture. The required sodium carbonate feed rate is 1.65 times the chlorine feed rate or approximately 4 times the ash feed rate, whichever is higher. The freezing point of the salt is a maximum (851oC) when the bed is pure sodium carbonate. This would be the case after preheating a cold vessel and loading the vessel with sodium carbonate. After many hours of operation with a chlorinated feed, the freezing point would be approximately 790oC, corresponding to 90 weight % sodium chloride. A normal range of temperatures for operation of the reaction vessel which permit proper waste destruction would be nominally 900 to 1000oC. The unit would normally not require an auxiliary fuel except for start up, stand by, or shutdown because of heat generated by oxidation of the waste. This is even true for low energy content wastes such as hexachlorobenzene (8300 KJ/kg, or 3570 BTU/lb).

Sodium carbonate is usually the preferred alkaline salt for making up the bed. As material is processed through the reaction vessel, any chlorine, sulfur, phosphorous, or ash products in the feed will be converted to inorganic salts and retained in the salt bed. In time, the composition of the bed changes as the sodium carbonate is converted to sodium chloride, and other salts and any ash is accumulated. A maximum operational limit for ash accumulation is approximately 20 per cent by weight, due to the increased bed viscosity caused by ash. A nominal minimum residual sodium carbonate content of 10 percent by weight is used to obtain a reasonable (economical) sodium carbonate utilisation. Hence, wastes with a high content of inert material, such as soil, cannot be treated in the reactor.

Figure 10.2
Schematic Diagram of the Molten Salt Process
(Independent Panel on Intractable Waste, 1992)
Figure 10.2


10.4.2 Performance

Test work for the MSO process indicates that particulate emissions from the process are relatively high, although it is claimed that this can be rectified using a better scrubber system, and should not be a limitation of the process. In general, metal chloride fume will usually be a very fine particulate and wet scrubbing systems would not normally be efficient enough to effect a high efficiency of capture. Fabric filtration (baghouse) systems can be expected to be required to achieve effective particulate removal.

Trials have been carried out at both the bench and pilot scale. Liquid 1,2,4 trichlorobenzene (58.6 weight % chlorine) was destroyed in molten sodium carbonate plus sodium chloride. The destruction efficiencies were 99.9999970% and 99.9999932% at bed temperatures of 900oC and 1000oC respectively.

A test program was conducted for the Canadian Electric Association to determine the MSO process conditions to meet the destruction efficiency of 99.9999%. The feed materials used were Aroclor 1254, Aroclor 1260 (liquid PCB products), and a solid mixture of PCB impregnated paper, plastic, and aluminium metal. Testing was carried out on bench scale equipment. Parametric studies were carried out to investigate the effects of temperature, bed depth, gas side residence time, oxygen enrichment, and two salt compositions (a normal sodium carbonate/sodium chloride mix, and a calcium oxide/calcium chloride mix).

The destruction efficiencies for the two transformer oils were in excess of 99.9999% except for one test, conducted at 800oC, which increased the destruction from 99.99973% to 99.999954%. However, decreasing the superficial gas velocity from 30 cm/s to 15 cm/s had virtually no effect (99.999954% to 99.999955%). The use of oxygen enrichment allowed a factor of three increase in throughput while increasing the destruction efficiency from 99.999955% (using air) to 99.999975%. Test results for the two different salt compositions were comparable (99.99957% for sodium carbonate, and 99.99983% for calcium oxide).

A test program to evaluate MSO destruction of chlorinated wastes was conducted for the US EPA, using chlordane (a liquid) and hexachlorobenzene (HCB, a solid) in both the bench scale and pilot scale equipment. HCB destruction removal efficiencies for the pilot facility were higher than 99.999% for five test runs including runs where the temperature was as low as 918oC and the HCB feedrate was as high as 122 kg/h. Gas analyses downstream of the baghouse were taken on three of the five runs and showed an improvement in the overall destruction and removal efficiency of HCB by about a factor of 100 (2 additional 9s). The bench scale tests showed similar results for similar run conditions. In addition, runs were made under deficient oxygen conditions (70% stoichiometric air), and at a low temperature (803oC). The deficient oxygen test achieved 99.999980% destruction. The low temperature test, however, gave surprisingly poor results; 99.9% destruction. At 902oC, the destruction removal efficiency was 99.99997%.

Chlordane destruction/removal efficiencies for the pilot facility were measured for five test runs. The lowest temperature was 896oC, and one air deficient trial was run. All of the destruction removal efficiencies were greater than 99.9999% with the exception of the first run. This run, made at 982oC and the lowest feed rate, gave a destruction removal efficiency of 99.99983%. The results of sampling in the stack downstream of the baghouse gave an actual release destruction removal efficiency of >99.9999988%, about the same two 9s improvement across the baghouse as seen in the previous runs.

For a small facility treating chlorinated organics, the total cost per tonne of wastes can vary from about $1200 to $2000 depending on the chlorine content. Costs are sensitive to the initial capital cost of the facility, how it is operated (effective throughput), the cost of soda ash and the chlorine content of the waste (amount of soda ash required per tonne of waste).

10.4.3 Considerations in the Application of the Technology

The molten salt concept lends itself to operation over a wide range of reaction stoichiometries (the ratio of the amount of oxygen made available to the amount of oxygen required for complete oxidation of the chemical waste) due to the stabilising effect of the large thermal mass of the salt. Both normal oxidising conditions (excess oxygen) and partially oxidising conditions (deficient oxygen) can be run without char or tar formation and with the reactions proceeding essentially to thermodynamically calculated equilibrium values.

This stability and predicability of operation allows the reactor vessel to be operated in the deficient oxygen mode with excellent destruction removal efficiencies of organic wastes without the need for a high temperature secondary incinerator, as is required with rotary kilns, and other flame based techniques. The residuals from the process are not useful, and must be disposed of properly in a secure landfill.

Generally, the cost of treatment with this technology is relatively high because of the high capital cost of the equipment, the labour requirements and the high energy cost. The cost per tonne is very dependent on the feed rate of the contaminant to the furnace. For a feed rate of 1000 kg/h, the cost is in the order of $1150/tonne. The above costs do not include effluent treatment costs, residuals and waste shipping costs handling and transport costs, analytical costs, and site restoration costs.

This treatment system is favoured for wastes contaminated with both chlorinated organics, and heavy metals such as cadmium, chromium, zinc etc. Other treatment processes such as BCD and hydrogenation can readily treat chlorinated organics but not heavy metals.

10.4.4 Summary

(a) Proponents (in Australia)

Rockwell International.

(b) Wastes Applicable

Organic liquids (high and low volatility) and other materials with a significant organic content. Soil and other materials with a high inorganic content that is not destroyed in the melt are not suitable for treatment in the molten salt system.

(c) Contaminants Applicable

All scheduled compounds.

(d) Status

Operational on a commercial scale in the USA.

(e) Timing for Commercialisation in Australia

Not stated. Establishment of the process in Australia could be achieved within a relatively short time frame, subject to an evaluation of commercial viability and regulatory approvals.

(f) Cost (example only)

Approximately $1150/tonne for a unit with a feed rate of 1000 kg/h. Costs may be higher (2 to 3 times) for treatment of smaller volumes in Australia.

(g) Safety/Environmental Risk

Operational experience in the USA indicates this process is relatively robust. The high thermal mass of the molten salt results in a relatively stable system. Care must be exercised when processing aqueous wastes due to the generation of steam and when processing highly chlorinated wastes due to the fuming of sodium chloride.

(h) Non-technical Impediments

Process probably viewed by the community in a more favourable light than incineration. This has not been confirmed.

(i) Preferred Mode of Implementation

Centralised system.

(j) Limitations

Not applicable to inert materials such as soil.

Chapter 9 - Eco Logic Chapter 11 - PCB Gone