SUPERCRITICAL WATER GASIFICATION OF SECONDARY PULP/PAPER-MILL SLUDGE AND SEWAGE SLUDGE
Linghong Zhang1, Chunbao (Charles) Xu2, *, and Pascale Champagne1
1Department of Civil Engineering, Queen’s University, Kingston, ON, CanadaK7L 3N6
2Department of Chemical Engineering, LakeheadUniversity, Thunder Bay, ON, CanadaP7B 5E1
Abstract: Supercritical water gasification of secondary pulp/paper-mill sludge with an initial water content of 98 wt% was conducted in an autoclave reactor at temperatures of 400-550oC over reaction times ranging between 20 and 120 min. Temperature had asignificant effect on improving total gas and hydrogen yields, particularlyin the range of 500 to 550oC. By contrast, increased reaction times only slightly enhanced gas formation. A notabledecline in the total gas and hydrogen yields were observedwhenthe water contentwas reduced through evaporation of water from the original sludge feedstock.For comparison, supercritical water gasification of three types of sewage sludge (primary sewage sludge, secondary sewage sludge, and digested sewage sludge) from the local municipal wastewater treatment plant was performed under similar experimental conditions. Secondary pulp/paper-mill sludge was found to exhibit the highest gas product yield, which could likely be attributed to its relatively high volatile matter and alkali metal contents.
Keywords: Supercritical water gasification; hydrogen;secondary pulp/paper-mill sludge; primary sewage sludge; secondary sewage sludge; digested sewage sludge.
1. INTRODUCTION
Sludge is the residue produced by treatment processes for wastewaters from domestic, industrial, or commercial sources. It is a liquid suspension that contains 0.25-12% solids, and is composed of biodegradable and recalcitrant organic compounds, as wells as pathogens andheavy metals.Sludge can also be considered as a waste biomass with a stored chemical energy content of 9-29 MJ per kgof total suspended solid that can be potentially recoveredby various biological and thermochemical processes. As with other types of biomass, sludge is a renewable energy source contributinglittle or no net greenhouse gas emissions to the environment because of its carbon-neutral lifecycle (Metcalf and Eddy, 2003;Fytili and Zabaniotou, 2008).
Since sludge has a very high water content (90% on wet mass basis), an energy-and-capital-intensive pre-drying step is typically required for most conventional thermochemical treatment processes, such as combustion, pyrolysis, and gasification (Furness et al., 2000). A specialized form of gasification, supercritical water gasification (SCWG), takes place in supercritical water (SCW) when pressure and temperature are increased to or above thecritical points for water (22.1 MPa and 374oC), and transforms biomass feedstock into a hydrogen-rich gas product. Ithas been considered as an advantageousmethod for directlytreating wet biomass feedstockseliminatingthe need for drying, as water participates in the steam reforming and water-gas shift reactions during the SCWG process, as shown byReactions 1 and 2. In addition, SCWG of biomass is characterized with a high H2 yield, reduced tar and coke formation, because of the very highsolubility of SCWfor both organic and inorganic compounds (Guo et al., 2007; Demirbas, 2004; Osada et al., 2006; Yanik et al., 2007).
Steam reforming reaction: CHxOy + (1-y) H2O → CO + (1-y+x/2) H2 (1)
Water-gas shift reaction: CO + H2O ↔ CO2 + H2 (2)
Although limited studies on SCWG of sewage sludge have been reported by other researchers (Xu and Antal, 1998; Xu et al., 1996; Zhang et al., 2007; Aye and Yamaguchi, 2006), there is no documented information on the SCWG behavior of sludge from industrial sources (e.g., from the pulp/paper industry). As such, theprimary objective of this study was to investigate the effects of reaction temperature, reaction time and sludge dry matter content (or water content) on the product yields inthe SCWG of secondary pulp/paper-mill sludge (SPP) with a batch reactorfor hydrogen production. For comparison, three types of sewage sludges: primary sewage sludge (PS), secondary sewage sludge (SS) and digested sewage sludge(DS) were also been examined under similar experimental conditions to those of SPP, to investigate the relationships between sludge compositions and the product yields.
2. MATERIALS AND METHODS
2.1 Materials
The sludge feedstocks used in this work were the SPP, supplied by the AbitibiBowater Thunder Bay Corporation, and PS, SS, and DSsupplied by the Thunder Bay Municipal Wastewater Treatment Plant. The DS is the sludge produced during the anaerobic digestion of the PS and SS. The compositions, in terms of water content, proximate and ultimate analyses, as well the mineral elemental concentrations of each type of sludge are given in Table 1 and Table 2. The solvents used in this work for product separation were distilled water and A.C.S. reagent-grade ethyl acetate and acetone purchased from the Canadawide Scientific and used as received.
2.2 Sludge supercritical water gasification
The SCWG tests were operated in a 75 mL Parr High-Pressure reactor, constructedof Hastelloy alloy, witha maximum working pressure of 41.37 MPa at 600C. For each experimental run, approximately 20 g of sludge was loaded into the reactor. The residual air in the reactor was completely removed with at least 3 cycles of vacuuming and nitrogen purging. The rector was pressurized to 2 MPa using high-purity N2to prevent the boiling of water during heating, then heated at about 10K/min to a specified temperature and maintained for a pre-determined period of time, which represented the reaction time in this research.During each experimental run, the pressure inside the reactor was recorded. Since the reaction pressure was dependent on the reaction temperature inside the sealed batch reactor, it could not be adjusted manually. However, our records demonstrated that all the reaction pressures throughout the study were above 22.1 MPa, indicating that water reached supercritical conditions. After the desired reaction time (20, 40, 60, and 120 min) elapsed, the reactor was cooled down to room temperature rapidly by using a
Table 1 Proximate and ultimate analyses of PS, SS, DS, and SPP
Type of sludge / water content(wt%)1 / Proximate analysis (wt%)(d.b.2) / Ultimate analysis (wt%)(d.b.2) / HHV4
(MJ/kg)
VM / FC / Ash / C / H / N / O3
PS / 97.2 / 67.8 / 10.7 / 21.5 / 40.3 / 5.2 / 3.3 / 29.7 / 15.75
SS / 95.5 / 60.1 / 12.2 / 27.7 / 37.9 / 4.8 / 5.9 / 23.7 / 15.43
DS / 97.2 / 48.9 / 9.3 / 41.8 / 31.6 / 3.7 / 3.9 / 19.0 / 12.57
SPP / 98.0 / 60.6 / 15 / 24.4 / 41.2 / 4.5 / 4.2 / 25.7 / 15.77
1.Water content = 100% - dry matter content
2.On a dry basis
3.By difference
4.Higher heating value (HHV) calculated by the Dulong Formula, i.e., HHV(MJ/kg) = 0.3383C + 1.422(H-O/8)
Table 2 Mineral elemental compositions of different sources of sludge1
Type of sludge / Major mineral elements in the sample (wt%)(d.b.2)Na / K / Mg / Ca / Mn / Fe / Zn / Al / Si / P / S
PS / 0.24 / 0.34 / 0.56 / 1.91 / 0.03 / 3.32 / 0.03 / 1.57 / 0.09 / 1.26 / 0.25
SS / 0.36 / 0.72 / 0.56 / 2.97 / 0.14 / 7.33 / 0.05 / 5.81 / 0.19 / 4.61 / 0.57
DS / 0.47 / 0.76 / 0.92 / 4.11 / 0.14 / 6.37 / 0.08 / 4.80 / 0.13 / 4.28 / 0.79
SPP / 7.04 / 0.53 / 0.81 / 1.42 / 0.03 / 0.48 / 0.06 / 1.65 / 0.08 / 0.87 / 2.33
- Determined by ICP-AES
- On a dry basis
piece of wet cloth towel to stop the reaction.Two to three duplicate runs were conducted for each experimental condition and the average yields were calculated and reported.The maximum errors between replicate runs were maintained within 5%for the gas and solid product yields, and 10% for the liquid yields. Better reproducibility for the liquid yields was challenging due to the difficulty in liquid productseparation and their low formation amounts in the SCWG process under investigation.
2.3 Separation and analyses of the reaction products
Once the reactor was cooled down to room temperature, the gas products were collected in a gas cylinder and analyzed with an Agilent 3000 Micro-Gas Chromatography (GC) equipped with dual columns (Molecular Sieve and PLOT-Q) and thermal conductivity detectors. The remaining solid/liquid products were recovered thoroughly from the reactor by washing with ethyl acetate solvent and filtered under the reduced pressure through a pre-weighted No. 5 filter paper. The solid retained on the filter paper was oven-dried at 105oCfor at least 12 hours and weighed to obtain the weight of the solid residue.The ethyl acetate soluble phase was separated in a separatory funnel,transferred into a pre-weighed evaporation flask and evaporated at a reduced pressure to completely remove the ethyl acetate. The remaining brown liquid was weighed and referred to as heavy oil (HO). The remaining water soluble product (WSP) in the aqueous phase was collected and diluted to 100 mL with distilled water in a 100 mL volumetric flask, from which 10 ml was sampled for further analysis. The yields of total gas, HO and solid residue were calculated as percentages (%) of the mass of each product in relation to the mass of dried matter in the sludge fed into the reactor prior to the reaction.The HO fraction was dissolved in acetone and analyzed with a Shimadzu gas chromatography/mass spectrometry (GC/MS)-QP2010Sequipped with a SHRXI-5MS capillary column (30m × 0.25mm × 0.25 m). The column temperature was initially 40oC, held for 2 min, and raised to 190oC at a heating rate of 12oC/min. The column was then heated at a heating rate of 8oC/minto 290oC, followed byanother isothermal hold for 30 min at this temperature. The obtained chromatographic peaks were identified using the WILEY8 library.
3. RESULTS AND DISCUSSION
3.1 Effect of temperature
The product yields obtained at various reaction temperatures(400-550ºC) are listed in Table 3. The gaseous products were mainly composed of H2, CO, CO2, CH4, and smallquantities of light C2 and C3 compounds, such as ethylene (C2H4), ethane (C2H6), acetylene (C2H2), propane (C3H8), and propylene (C3H6). As expected, highertemperature facilitatedthe formation of gas,which is in good agreement with Demirbas (2004) and Zhang et al. (2007). In addition, an increase in the yield of char anda decrease inthe yield of HO were also observed. Specifically, at low temperature400oC, the yield of gaseous products was 16.4%, whereas the yields of both HO and solid residue were as high as approximately 30%. As the temperature increased from 400oC to 500oC, the yield of gas products wasimprovedconsiderably to 30% (almost double that at 400oC) at the expense of lowerHO (17%) and solid residue (26%) yields.As the temperature wasincreased further, from 500 to 550oC, the gas yield was further increased at the expense of HO, suggesting cracking of the heavy oil product into gases. From a thermodynamic perspective, a high temperature favors gas formation, since the overall biomass SCWG process (as simplified byReaction 3) is endothermic (Guo et al., 2007).
CHxOy + (2-y) H2O → CO2 + (2-y+x/2) H2 (3)
As the temperature wasincreased from 500 to 550oC, however, a surprisingly large increase in the yield of solid residue, from 26% at 500oC to 39% at 550oC, was observed. Similar result where the formation of solid residue (or char) increased with temperature wasobserved in previous work by Xu and Lancaster (2008), when liquefying SPPin hot-compressed water in various temperature ranges. The increase in char formation at a higher temperaturescould be attributed to cellulose and lignin compounds in the sludge feed, which undergo hydrolysis and are decomposed to lighter fragments (intermediates) and could be further cracked into gaseous products or be condensed to chars (Osada et al., 2006; Yanik et al., 2007). For example, phenolic compounds and formaldehydes, both of which are hydrolysis products from lignin, maypolymerize intohigh molecular-weight condensed products or solidchars via condensation/cross-linking reactions. GC/MS analysesconducted in this study (spectrum not shown)
Table 3 Yields of gas, liquid and solid products and the formation of major gaseous components during SCWG of SPP at various experimental conditions
Temp.(oC) / Reaction time (min) / Dry matter content (wt%) / Yields of gas, liquid and solid products (%) / Yields of major gaseous components (mole/kg dried sludge)
Gas / HO / Solid
Residue / (WSPs+H2O)1 / H2 / CO2 / CO / CH4
400 / 60 / 2 / 16.4 / 28.8 / 29.1 / 25.7 / 1.50 / 3.21 / 0.19 / 0.22
450 / 60 / 2 / 18. 6 / 23.9 / 26.1 / 31.4 / 2.63 / 3.84 / 0.20 / 0.47
500 / 60 / 2 / 30.0 / 17.1 / 25.9 / 27.0 / 5.58 / 5.35 / 0.19 / 1.59
550 / 60 / 2 / 37.7 / 10.2 / 39.4 / 12.7 / 14.48 / 6.04 / 0.19 / 2.96
500 / 20 / 2 / 28.8 / 36.8 / 25.3 / 9.1 / 5.25 / 5.22 / 0.25 / 1.25
500 / 40 / 2 / 30.3 / 20.5 / 26.5 / 22.7 / 5.27 / 5.38 / 0.21 / 1.61
500 / 120 / 2 / 32.0 / 9.1 / 24.9 / 34.0 / 5.88 / 5.47 / 0.20 / 2.05
500 / 60 / 6 / 21.2 / 12.1 / 24.2 / 42.5 / 2.80 / 3.76 / 0.13 / 1.21
500 / 60 / 8.8 / 22.2 / 12.3 / 25.8 / 39.7 / 2.58 / 3.81 / 0.12 / 1.58
1 By difference.
confirmed that there weresignificant amounts of phenols and phenolic compounds, such as phenol, 2 (or 4)-methyl-; phenol, 4-methoxy-, phenol, 2,4-dimethyl-, phenol, 2(3, or 4)-ethyl- in the HO products. These phenol and phenolic compounds may not only be decomposed from lignin compounds, but also be converted from cellulose under subcritical condition via a series of complex ionic reactions (Kruse and Gawlik, 2003). Hence, it is likely that the increased gas and solids resulted from the cracking or the condensation of the phenolic compounds in the HO. Additionally, it has been widely accepted that as temperature increases, the dehydration of carbohydrates and some soluble organic acids, present in the WSP as a product of lignocellulosic feedstock hydrolysis, would occur at a high temperature to form HO which would further be condensed into char. For this end, the lumped yields of (WSPs and pyrolytic water), calculated simply by difference, are given in Table 3. As shown in Table 3, interestingly as temperature increased from 500C to 550C, the net increase in the yield of solid residue (~14%) was almost balanced by acorresponding decrease in WSPs+H2Oyield. This observation likely supports the possibility WSPs dehydration to form HO, followed by the condensation of HO into char at 500-550C.
The production of H2 and other combustible gases were of primary interest in this study. The yields of each main gaseous compound(e.g., H2, CO, CO2, and CH4) in mole per kg of dried sludge, are presented in Table 3. As was expected, the yield of all individual gas components (except CO)increased monotonically with temperature. In particular for H2 and CH4 formation,significant increases were observed between 500 and 550oC. Overall, increasing the temperature from 400oC to 550oC led to an almost 10-fold increase in the yields of H2 and CH4, the H2 yield attaining as high as 14.5 mol/kg dried sludge.
3.2 Effect of reaction time
Compared with temperature, reaction time showed a much smaller influence on solid product yields, as well as total gas yields including individual gas species such as H2, CO, CO2 and CH4 (Table 3). These results were generally consistent with previous work by Xu and Lancaster (2008) treating a similar SPP feedstock in hot-compressed water and observations that a higher gasification efficiency could be attained with longer reaction times (Williams andOnwudili, 2005;Zhang et al., 2007). The results (Table 3) indicated that reaction time significantly affected the distribution of the liquid products (HO and WSPs), where the WSPs+H2O yield increased at the expense of the HO yield as reaction time increased from 20 to 120 min at 500oC, suggesting the conversion of HO to WSPs with longer residence time probably via hydration reactions. Williams and Onwudili (2005) investigated the effect of reaction time on product distribution using glucose as the biomass model compound, and similar results were obtained where a significant decrease in HO yield corresponded to a higher WSP yield at a longer reaction time.
3.3 Effect of dry matter content (water content) in the sludge feedstock
In this study, the raw SPP with an original solid concentration of 2.0 wt% was thickened by vaporization of water to
condensed feedstockswith dry matter contents of 6.0 wt% and 8.8 wt%, respectively, to investigate the effect of dry matter content on product yields.It can be noted from Table 3 that the changeinfeedstock dry matter content did not result ina significant change in solid residue yield. However, the feedstock dry matter content did influence the distribution of gas and liquid products. Generally, the increase in feedstock dry matter content from 2 wt% to 8.8 wt% was accompanied by an increasedWSPs yield andcorresponding decreased HO yield, and total gas and individual H2 and CO2 yields (Table 3).The effect of feedstock dry matter content on the formation of CO and CH4 gases was found to be minimal. Similar results were reportedin other work by Kruse et al., (2003) andOnwudili and Williams (2007), wherea lowerwater density (or higherdry matter content) decreasedH2 production. As previously explained in theintroduction, water participates in the overall SCWG process as areactant. Thus, increasing the dry matter content would reduce the relative amount of water per unit of dried sludge in the reactor, which would shift Reactions 1 and 2in the direction leading to reduced H2 and CO2 yields as demonstrated by the results in Table 3.
3.4 Effect of different sources of sludge
Three types of municipal sewage sludges: PS, SS, and DS,were diluted with distilled water in this study to obtainthe similar solid matter content (~2%) as that of SPP to evaluate their potentials for gas production in SCW. SCWG of these different sludge feedstocks was performed at500oC for 60 min. The distribution of gas, HO, and solid residue products is provided in Figure1a and the yields of themajor gaseous components are illustrated in Figure1b. As can be seen from Figures 1a and 1b, product distribution and gas formation are strongly dependent on the type of sludge (or the properties of the feedstock). SCWG of SPP produced the highest gas, HO, H2, CH4 and CO2 yields. With respects to the production ofthese gas and liquid products, the four types of sludges tested in this work showed the following priority sequence: SPP > PS > SS > DS.The different behaviors of these feedstocks may be associated with their distinct compositional characteristics, such as the contents of volatile matter and ash as well as the ash compositions (alkali metal content). Specifically, the low yieldsof total gas and individual gas species of H2, CH4 and CO2 from DSwas probably due to its low content of volatile matter (Table 1), which has been partly consumed by the anaerobic bacteria. Moreover, alkali salts were proven to be an effective catalyst for improving biomass gasification and H2formation due to its distinct ability to promote the water-gas shift reaction (Reaction 2) (Zhang et al., 2007). The SPP feedstock contained a high content of alkali metals (Table 2), 7.04 wt% Na in relation to <0.5% Na for the other feedstocks (PS, SS, and DS). A high content of Na salts, added during the pulping process, may account for the higher performance of the SPP sludge in the course of SCWG for the production of gas and liquid products.
(a) (b)
Fig.1. Distribution of gas, HO, and solid residue products (a) and the yields of some major gaseous components (b)
in SCWG of different types of sludge feedstock at500oC for 60 min.
4. CONCLUSIONS
A comprehensive study on the SCWGof four types of sludge feedstocks, supplied from the pulp/paper-mill and municipal wastewater treatment plants, has been conducted. Various process parameters including reaction temperature, reaction time, dry matter content, and type/properties of the sludge were investigated to explore SCWG as a potential technology for the cost-effective disposal of high-water-content sludge and for energy recovery. The most significantconclusions of this work may be summarized as follows:
1)A higher temperature favored the gasification reactions, while decreasing the yield of HO products. The yield of H2was increased significantly from 1.5to 14.5 mol/kg-dried sludge by increasing the reaction temperature from 400 to 550oC.