3 Supercritical Water Oxidation Technology Indira Jayaweera SRI International, Menlo Park, California, U.S.A. I. INTRODUCTION Supercritical water oxidation (SCWO) is a waste treatment technology that uses supercritical water as a medium for oxidizing organic material. SCWO can also be described as an extension of the subcritical oxidation process, wet air oxidation (WAO) process, or widely known Zimpro process [1].* Both processes (SCWO and WAO) involve bringing together organic waste, water, and an oxidant (such as air or oxygen) to elevated temperatures and pressures to bring about chemical oxidations. Generic operational condi- tions for the two processes are as follows: WAO: 150–300jC, 10–200 bar SCWO: > 374–675jC, > 220 bar. Fig. 1 shows a graphical representation of these operational regions. WAO and SCWO processes are often referred to as hydrothermal oxida- tion technologies (HTOs). The major difference between the processes is that, in SCWO, organics are completely oxidized in a relatively short time (seconds to minutes), whereas in WAO, the reaction may require a longer time (minutes to hours). Furthermore, in WAO, some refractory organics are not completely oxidized because of the lower temperature of operation * The pioneering work of Fred Zimmerman in the 1950s led to the creation of the Zimpro process [1]. TM Copyright © 2003 by Marcel Dekker, Inc. All Rights Reserved. (<350jC), thus requiring a secondary treatment process. As information on WAO technology is readily available from other sources [2–5], the SCWO process is mainly discussed here. The SCWO process is ideal for the disposal of many aqueous hazardous materials (e.g., EPA priority pollutants, industrial wastewater and sludges, municipal sludges, agricultural chemicals, and laboratory wastes), but has also been demonstrated to effectively destroy military wastes (e.g., ordnance, rocket propellants, and chemical agents) [6–18]. The effluent from the SCWO process, consisting primarily of water and carbon dioxide, is relatively benign. Therefore, the SCWO process can easily be designed as a full-scale contain- ment process with no release of pollutants to the atmosphere. Compared with incineration and other high-temperature treatments, such as the plasma process, SCWO processes achieve high organic destruction efficiencies (>99.99%) at much lower temperatures (<700jC) and without NO x production. Sanjay Amin, a student of Michael Modell at Massachusetts Institute of Technology, first discovered in the mid-1970s the effect of supercritical water for decomposition of organic compounds without forming tar [19]. Figure 1 Phase diagram for pure water. Solid line: liquid–gas equilibrium. Jayaweera122 TM Copyright © 2003 by Marcel Dekker, Inc. All Rights Reserved. This information, together with the information available at the time from Connolly’s 1966 publication [20], which stated that organics can be solubi- lized in all proportions in high-temperature pressurized water, has led to the birth of the SCWO process. The breakthrough of the SCWO process seems to stem from the work of E. U. Frank, Karlsruhe, Germany, and Marshall and coworkers, Oak Ridge, TN , on the thermodynamics of binary mixtures of gases, organics, and inorganics in subcritical and supercritical water [21–23 and references therein]. Although the technology was invented in the late 1970s, much of the development work was conducted from 1980 to the early 1990s. During this period, researchers demonstrated the great utility of SCWO as a method for waste disposal without production of harmful products. However, during the same period, the major technical obstacles to commercialization of the process had also been discovered. The two major technical challenges were reactor corrosion and reactor plugging. Reactor corrosion is caused by the formation of acids during the processing, especially when waste streams containing acid-forming components (e.g., chlorinated organics) are treated. Reactor plugging occurs when inorganic salts present in the waste stream are precipitated during the processing. Thus, the major criteria for designing the process involve consideration of possible corrosion and reactor plugging, as most industrial waste streams contain inorganic solids or heteroatoms that form inorganic solids for a majority of the SCWO systems. In addition, the problems associated with salt plugging and corrosion vary with the SCWO operating conditions (or the type of SCWO system). In general, there are several diff erent versions of SCWO systems (low- and high-temperature SCWO, moderate and very high pressure SCWO, catalytic and -noncatalytic SCWO, etc.). Most of these different SCWO systems have been developed to overcome problems and to improve the performance of the process. However, only a few of those SCWO processes are commercially available and commonly practiced SCWO systems are discussed in this chapter.* II. BACKGROUND AND FUNDAMENTALS OF SCWO A. General Description The SCWO process involves bringing together an aqueous waste stream and oxygen in a heated pressurized reactor operating above the critical point of * Only the aboveground SCWO systems are discussed here. There is an underground SCWO system that uses hydrostatic pressure to avoid the need for high-pressure pumps. This system was designed by Oxydyne Corporation, Dallas, TX. Supercritical Water Oxidation Technology 123 TM Copyright © 2003 by Marcel Dekker, Inc. All Rights Reserved. water (374jC, 22.1 MPa or 218 atm). Under these co nditions, the solubility properties of water are reversed (i.e., increased organic solubility and decreased inorganic solubility*), and the viscosity of the media is decreased to a value similar to -gaslike values, thus enhancing the mass transfer properties. These unique properties of hot pressurized water allow oxygen and organics to be contacted in a single phase in which oxidation of organics proceed rapidly. At 400–650jC and 3750 psi, SCWO can be used to achieve complete oxidation of many organic compounds with destruction rate efficiencies of 99.99% or higher. A generic flow diagram for the SCWO process is given in Fig. 2. The aqueous waste is brought to system pressure using one or more high- pressure pumps. Compressed air or oxygen is added to the pressurized waste, and the waste–air mixture is initially heated to about 300jC using a preheater. The preheated mixture is directed to the main reactor operated at thedesiredtemperature(above374jC), where the complete oxidation occurs. The effluent from the reactor then travels through a heat exchanger, a pressure letdown valve, and a solid/liquid/gas separator. The preheater section of the system mimics a miniature WAO system because the reaction conditions in the preheater are similar to those of a WAO system except that WAO systems need longer reaction times. In the heat-recovery mode of operation, the SCWO uses the heat from the reaction to preheat the influent. As a rule of thumb, if the aqueous waste stream contains about 4 wt.% of organics, the SCWO can be processed under self- sufficient heat conditions. However, for dilute aqueous waste streams, the SCWO process may not be cost-effective because of the additional thermal energy required to maintain the reactor temperature in the 400–650jC range. Figure 2 A generic hydrothermal oxidation (WAO, SCWO) process flow diagram. * The details of the inorganic solubility are given in Sec. B.2., Phase Separations. Jayaweera124 TM Copyright © 2003 by Marcel Dekker, Inc. All Rights Reserved. B. In-Depth Treatment of SCWO 1. Fluid Characteristics The basic properties of water such as viscosity, dissociation constant, dielectric constant, compressibility, and the coefficient of thermal expansion play a major role in determining optimal reaction conditions for obtaining maximum benefits in both SCWO and WAO processes. The properties of water change dramatically with temperature, particularly near the critical point [24–26]. A well-known example, the variation of pK w with temperature at the saturation pressure, is shown in Fig. 3. The dissociation constant of water goes through a maximum around 250jC(pK w minimum), and then undergoes a sharp decline as the temperature approaches the critical point. The density and the dielectric constant of wat er also show sharp changes close to the critical point, as shown in Fig. 4. The rate-limiting properties of any organic reaction that includes the mixing of several components are the solubility of the contaminant in the liquid phase or its equilibrium solubility, and the mass transfer step (i.e., Figure 3 Variation of pK w with temperature at the saturation pressure. Supercritical Water Oxidation Technology 125 TM Copyright © 2003 by Marcel Dekker, Inc. All Rights Reserved. dissolution into the aqueous phase). Therefore, the transport properties of the reaction media are very important for efficient waste processing. The viscosity of water decreases with temperature, thus providing rapid diffu- sion. The conductance of heated water remains high in spite of the decrease in the dielectric constant because of the increased ion mobility brought about by the decreased viscosity. However, as the dielectric constant of water decreases with the increase in temperature, electrolytes that are completely dissociated at low temperature become much less dissociated at high temperature, particularly in the supercritical region. At the critical point (374jC, 218 atm pressure, dielectric constant, e=5), water acts as a mildly polar organic solvent, and thus supercritical water can readily solubilize nonpolar organic molecules. In fact, most hydrocarbons become soluble in water between 200j and 250jC [27], allowing opportunities to enhance reactivities of organics even under subcritical water conditions. The enhanced diffusivity and the decreased dielectric constant at elevated temperatures make water an excellent solvent for dissolving organic materi- Figure 4 Variation of density and dielectric constant with temperature at the saturation pressure. Jayaweera126 TM Copyright © 2003 by Marcel Dekker, Inc. All Rights Reserved. als that are tightly bound to solid material (important for treatment of solid waste). As an example, hot pressurized water has been shown to break and separate ‘‘highly stable’’ water–oil–solid emulsions, generated in petroleum wastewater treatment and other industrial operations [28]. Compared with ambient values, the specific heat capacity of water approaches infinity at the critical point and remains an order of magnitude higher in the critical region [26], making supercritical water an excellent thermal energy carrier. As an example, direct measurements of the heat capacity of dilute solutions of argon in water from room temperature to 300jC have shown that the heat capacities are fairly constant up to about 175–200jC, but begin to increase rapidly at around 225jC and appear to reach infinity at the critical temperature of water [29]. The static dielectric constant is a measure of hydrogen bonding and reflects the characteristics of the polar molecules in water. However, very little is known about the degree of hydrogen bonding under supercritical water oxidation conditions. The lack of data on the character of hydrogen bonding in water at high temperatures and pressures hinders the understanding of the structure and pr operties of supercritical water. The important question is: Up to what temperature can hydrogen bonding in water exist? It was initially believed that hydrogen bonds do not exist above 420 K. Later, Murchi and Eyring [30], using the approach of significant structures, showed that the hy- drogen bonds disappear above 523 K and that water above this temperature consists of free monomers. Later, Luck [31], studying the IR absorption in liquid water, extended the limit of hydrogen bonding at least up to the criti- cal temperature. Recently, a theoretical model developed by Gupta et al. [32] has shown that in supercritical water, significant amounts of hydrogen bond- ing are still present despite all the thermal energy and are highly pressure and temperature dependent. An interesting result has emerged from Sandia Na- tional Laboratories’ theoretical estimation of hydrogen bonding of super- critical media by calculating the equilibrium population of water polymers (dimers, trimers, etc.) [33]; however, this contradicts the Murch and Eyring findings above [30]. Their calculations have shown that at 450–650jC and 240–350 bar, the water polymer concentration can be as high as 40%. It is also cited in later work by Kalinichev and Bass [34] that hydrogen bonding is still present in the form of dimers and trimers in the supercritical state. More details and new theoretical discussions can be found in Refs. 35, 36, and the references therein. 2. Phase Separations It is important that the phase behavior of the influent at high temperature and pressure conditions be clearly understood for designing process compo- Supercritical Water Oxidation Technology 127 TM Copyright © 2003 by Marcel Dekker, Inc. All Rights Reserved. nents such as the main reactor. Under the operational conditions of SCWO, the conditions can be easily adjusted to attain a single phase when only organics are present. However, when inorganic salts are present (either as a reagent or as a by-product from the process) under SCWO conditions, it is challenging or even impossible to predict the phase behavior of the medium. The presence of electrolytes changes the saturation–vapor boundary line for water. Liquid–vapor equilibria in a soluble salt–water system above the critical temperature are complex. However, the situation below the critical temperature of pure water is simpler, at least for solutes that are so involatile at this temperature that their concentrations in the vapor phase are negligible. Liquid solutions of these solutes have vapor pressures that are lower at a given temperature than that of pure water. Equivalently, they have boiling points that are higher at a given total pressure than that of pure water. Fig. 5 shows the relationship between vapor pressure and temperature for Na 2 CO 3 –H 2 O and NaCl–H 2 O systems [36]. It is clear that these two systems have different phase behaviors under SCWO conditions. Because of the complex nature of the phase diagrams for salt-water systems and the Figure 5 The relationship between vapor pressure and temperature for the Na 2 CO 3 –H 2 O and NaCl–H 2 O systems. Jayaweera128 TM Copyright © 2003 by Marcel Dekker, Inc. All Rights Reserved. inconsistencies of the available literature data, only a brief discussion is given below with appropriate references. In recent years, studies of the phase behavior of salt-water systems have primarily been carried out by Russian investigators (headed by Prof. Vladimir Valyashko) at the Kurnakov Institute in Moscow, particularly for fundamental understanding of the phase behavior of such systems. Val- yashko [37,39,42,43], Ravich [38], Urosova and Valyashko [40], and Ravich et al. [41] have given a classification of the existence of two types of salts, depending on whether the critical behavior is observed in saturated solu- tions. Type 1 does not exhibit critical behavior in saturated solutions. The classic example of Type 1 is the NaCl–water system and has been studied by many authors [36,37,44–47]. The Type 2 systems exhibit critical behaviors in saturated solutions, and therefore have discontinuous solid–liquid–vapor equilibria. Table 1 shows the classification of binary mixtures of salt–water systems. In brief, the salts that are classified as Type 1 have increasing solubility with increasing temperature, whereas Type 2 salts show an opposite trend. For example, sodium carbonate, a Type 2 salt, has a 30 wt.% solubility under ambient conditions and its solubility near the critical point approaches zero [36] whereas sodium chloride, a Type 1 salt, has a 37 wt.% solubility under subcritical conditions at 300jC and about 120 ppm at 550jC [46]. In real systems, organic–inorganic multicomponent phase systems are possible, and the information gathered from binary or ternary systems cannot be extended to these real situations. Currently, Valyashko from Kurnakov Institute and Jayaweera from SRI International are jointly study- Table 1 Saltwater Binary Systems Type 1 salts Type 2 salts KF, RbF, CsF LiF, NaF LiCl, LiBr, LiI Li 2 CO 3 ,Na 2 CO 3 NaCl, NaBr, NaI Li 2 SO 4 ,Na 2 SO 4 ,K 2 SO 4 K 2 CO 3 , RbCO 3 Li 2 SiO 3 ,Na 2 SiO 3 Rb 2 SO 4 Na 3 PO 4 Na 2 SeO 4 CaF 2 , SrF 2 , BAF 2 K 2 SiO 3 SiO 2 ,Al 2 O 3 K 3 PO 4 CaCl 2 , CaBr 2 , CaI 2 BaCl 2 , BaBr 2 NaOH, KOH Source: Ref. 37. Supercritical Water Oxidation Technology 129 TM Copyright © 2003 by Marcel Dekker, Inc. All Rights Reserved. ing both the phase behavior and the morphology changes of salts precipitated from a salt-water system containing Na 2 CO 3 ,K 2 CO 3 , and NaCl [48]. Importance of Electrochemistry in SCWO Processing. In both WAO and SCWO processing, where the reactor surfaces experience extremes in pH and high inorganic salt concentrations under high temperature/high pressure conditions, enhanced electrochemical processes could cause corro- sion and rapid metallurgical degradations of the reactor vessels. Therefore, materials should be evaluated to determine if they could withstand the SCWO conditions. In general, researchers have been mainly focused on understanding the corrosion processes such as pitting corrosion (disruption of the protective oxide surface layer followed by the heavily localized dissolution of the underlying alloy, forming holes or pits), crevice corrosion (localized form of corrosion associated with stagnant solutions in crevices), and stress corrosion cracking (cracking induced by the combined influence of the tensile stress and corrosive medium) under SCWO conditions [49]; a detailed description of metallurgical aspects, material properties, thermody- namics of the corrosion process, corrosion kinetics, and corrosion phenom- ena under hydrothermal conditions can be found in Refs. 51 and 52. In predicting metal stability under aqueous environments, it is custom- ary to use electrochemical potential–pH diagrams (E h –pH or Pourbaix diagrams). Many workers have derived and published potential–pH diagrams for metal–water systems under varying temperature and pressure conditions [52–54]. Cr, Fe, and Ni systems are the most widely studied systems (the alloys currently used for WAO and SCWO studies contain Cr, Fe, and Ni, e.g., stainless steel 316 and Hast elloy C2-276). Under oxidative conditions, metal oxide films are formed on the reactor surfaces. Some metal oxides, such as Fe 3 O 4 ,Cr 2 O 3 , etc., will passivate the metal surface reducing the corrosion, and thus both immunity and passivation regions, where a process can be operated with minimal corrosion, are possible. In the case of chromium, the shift in the equilibrium line for the oxidative dissolution of Cr 2 O 3 Cr 2 O 3 þ 5H 2 O ! 2CrO À 4 þ 10H þ þ 8e À with increasing temperature is of practical importance for stainless steel, because it is the formation of chromic oxide (or at least a chromium- containing spinel, e.g., FeCr 2 O 4 ) that confers passivity to the alloy. Research- ers have tried to evaluate the effect of secondary metals on the primary metal in alloys by adding corresponding salts to the corrosion medium. For example, Fig. 6 shows the pos sible passivity, immunity, and corrosion areas for iron in the presence of CrO 4 2– under ambient conditions [52]. The potential–pH diagrams under ambient conditions cannot be used to predict the stability at higher temperatures. The passivation region for Jayaweera130 TM Copyright © 2003 by Marcel Dekker, Inc. All Rights Reserved. [...]... 20 03 by Marcel Dekker, Inc All Rights Reserved Chlorinated solvent Motor oil Gray water 30 102 638 32 50 95 10 1690 721 2240 11 0 8 1. 03 >99.998 0.6 . Li 2 SO 4 ,Na 2 SO 4 ,K 2 SO 4 K 2 CO 3 , RbCO 3 Li 2 SiO 3 ,Na 2 SiO 3 Rb 2 SO 4 Na 3 PO 4 Na 2 SeO 4 CaF 2 , SrF 2 , BAF 2 K 2 SiO 3 SiO 2 ,Al 2 O 3 K 3 PO 4 CaCl 2 , CaBr 2 , CaI 2 BaCl 2 , BaBr 2 NaOH, KOH Source: Ref. 37 . Supercritical. phenol and biphenyls), and the related products dibenzofuran and dibenzo-p-dioxin [82, 83, 87]. Li et al. [84] studied the intermediates from the oxidation of 2-chlorophenol and noted the produc- tion. diagrams for salt-water systems and the Figure 5 The relationship between vapor pressure and temperature for the Na 2 CO 3 –H 2 O and NaCl–H 2 O systems. Jayaweera128 TM Copyright © 20 03 by Marcel