Contents lists available at SciVerse ScienceDirect
Fuel 107 (2013) 113–121
Fuel
j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / f u e l
Study of catalytic post-treatment of the vapours from sewage sludge pyrolysis by means of c-Al2O3 M. Azuara a,⇑
, I. Fonts a,b, P. Barcelona a, M.B. Murillo a, G. Gea a
a Thermochemical Processes Research Group (GPT), Aragón Institute for Engineering Research (I3A), Universidad de Zaragoza, Mariano Esquillor s/n, E-50018 Zaragoza, Spain b Centro Universitario de la Defensa de Zaragoza, Ctra. Huesca s/n, 50090 Zaragoza, Spain
h i g h l i g h t s
" c-Al2O3 is useful for obtaining a better pyrolysis liquid from the point of view of its phase separation. " c-Al2O3 application makes it possible to lower the pyrolysis temperature. " A general improvement in the higher heating value of the organic phases has been achieved. " Energy content of the organic phase is comparable to a conventional fuel.
a r t i c l e
i n f o
a b s t r a c t
Article history: Received 9 February 2012 Received in revised form 31 January 2013 Accepted 7 February 2013 Available online 26 February 2013
The present work describes an experimental study on sewage sludge pyrolysis in a Fluidised Bed Reactor (FluBR) and the effect on the pyrolysis liquid product of passing the hot exhaust (pyrolysis) vapours through a Fixed Bed Reactor (FixBR) filled with gamma-alumina (c-Al2O3). Experiments have been done at temperatures of 450 and 550 (cid:2)C (FluBR) and 400 and 500 (cid:2)C (FixBR). In the FixBR experiments, the bed was filled with inert sand and two different bed quantities of c-Al2O3 (97.3 g, 194.6 g) giving a weight hour space velocity (WHSV) of between 0.6 and 1.8 h(cid:2)1. By means of this treatment, the pyrolysis liquid phase separation was improved, obtaining a liquid with two phases (organic and aqueous) instead of the three (two organic and an aqueous phase) usually obtained in sewage sludge pyrolysis without applying c-Al2O3. This organic phase is quite interesting from the point of view of a possible application as bio-oil. Although the product distribution results using c-Al2O3 show a slight liquid yield decrease for both tem- peratures (compared to sewage sludge pyrolysis without applying c-Al2O3), there is a clear increase in the organic phase yield at 450 (cid:2)C. Therefore, c-Al2O3 application makes it possible to lower the pyrolysis tem- perature in order to maximise the organic phase in the sewage sludge pyrolysis liquid. Furthermore, a general improvement in the higher heating value of the organic phases has been achieved.
(cid:3) 2013 Elsevier Ltd. All rights reserved.
1. Introduction
recognition as promising alternatives for the valorisation and reduction of this waste [3–5]. Nowadays, much attention in sewage sludge research focuses on pyrolysis liquid as a possible solution in the search for new renewable alternatives for fossil fuel [6–9].
Pyrolysis can be used as a sewage sludge management process in which three products are obtained: char, a gas mixture and a li- quid product with several phases. According to previous studies by the Thermochemical Processes Research Group (GPT), pyrolysis li- quid from sewage sludge obtained in a fluidised bed consists of two organic phases (a light one and a heavy viscous one) and an aqueous phase [10,11]. The organic phases have relatively high heating values and could be applied as sources of chemicals or as fuels [12].
Sewage sludge is the main waste produced during the purifica- tion of urban wastewater. Nowadays, due to the restrictive Euro- pean legislation and to the increase in the standard of living, the production of sewage sludge has augmented significantly during recent years. Specifically, in the EU more than 10 million metric tons (dry solids) of sewage sludge are produced annually [1,2]. Application on agricultural land and incineration are two of the most common processes to manage this waste; however, these practices are becoming increasingly difficult due to health, environ- mental and social concerns. Consequently, alternative thermo- chemical processes such as gasification or pyrolysis are gaining
⇑ Corresponding author. Tel.: +34 976762961.
The aqueous phase has a high ammonia content suggesting its application as a fertiliser [13,14]. The organic phases, and espe- cially the heavy phase, contain a significant amount of triglycerides and fatty acids [7,15,16]. Furthermore, in the light organic phase
Keywords: Pyrolysis Sewage sludge Gamma-alumina Biomass Catalytic treatment
E-mail address: mazuara@unizar.es (M. Azuara).
0016-2361/$ - see front matter (cid:3) 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.fuel.2013.02.017
2.1.2. c-Al2O3
The chemical composition of c-Al2O3 is shown in Table 2. The main physical properties are: 142 m2/g surface area (BET), 10.5 nm average pore size, 150–350 lm particle size and 0.4 cm3/ g mesopore volume (BJH).
For its activation, the c-Al2O3 was calcined at 600 (cid:2)C for 3 h and
there is a significant amount of aliphatic hydrocarbons. Triglycer- ides and fatty acids have an important presence in sewage sludge [17] and contribute substantially to the heating value of these phases, but they also increase their viscosity and their oxygen con- tent. These two properties limit their use as fuel due to the difficul- ties of injection and atomisation [18] and to reduced storage and thermal stability, respectively [19].
kept in a desiccator prior to the experiment.
2.1.3. Inert sand
The application of a catalyst in order to increase the pyrolysis liquid yield [20,21] and to improve its quality as fuel or chemicals is a common practice. Research in catalytic pyrolysis of wood is mostly focused on removing the oxygen content by using different catalysts such as Zeolites [20,22,23] or Al-MCM-41 (Mesoporous aluminosilicate molecular sieves) [24].
Inert sand, provided by Panreac, was mostly composed of SiO2 with a particle size between 0.25 and 0.30 mm. This material was used to compare the possible influence of the FixBR tempera- ture on gas and liquid yield without using the c-Al2O3.
2.2. Experimental system
The sewage sludge pyrolysis runs were carried out in a labora- tory scale (<1 kg/h) Fluidised Bed Reactor (FluBR) operating at atmospheric pressure with a continuous feed and ash removal sys- tem and heated by means of an electrical furnace. Further informa- tion about the FluBR lab-scale plant can be found elsewhere [16,30–32].
Research in catalytic pyrolysis of sewage sludge is quite limited, usually being based on previous works using lignocellulosic bio- mass. These studies also focus on the reduction of the oxygen con- tent of the pyrolysis liquids for improving the hydrogen to carbon ratio [25,26]. Owing to the fatty acids content in sewage sludge, other authors have pointed out the good performance of activated gamma-alumina catalysts (c-Al2O3) in the decomposition of these fatty acids and triglycerides. The aim is to produce aliphatic hydro- carbons such as alkanes and alkenes from lipids extracted from the sewage sludge [27–29].
Catalytic treatment with c-Al2O3 may cause the transformation of the triglycerides and fatty acids into aliphatic hydrocarbons with the consequent improvements in the fuel quality of the sewage sludge pyrolysis liquid (homogeneity, heating value or viscosity, among others).
Experiments with catalytic treatment were performed using a tubular Fixed Bed Reactor (FixBR) made of AISI 310 steel (length = 25 cm, inner diameter = 5.4 cm) (see Fig. 1). The FixBR was connected by means of a hot filter to the FluBR and placed next to the FluBR. Pyrolysis vapours flowed from the top of the FixBR to the bottom section. The available length of catalyst bed (L) in the FixBR, taking into account the distributor plate (placed 6 cm from the gas output), was between 0 and 15 cm. The FixBR was heated with electrical resistances and its temperature controlled using appropriately placed thermocouples and a PDI controller.
The main goal of this study is to know the effect of a c-Al2O3 catalytic treatment on sewage sludge pyrolysis distribution prod- ucts. Special attention is focused on the appearance of the liquid phase and on the higher heating value of the organic phase in lab-scale experiments. Furthermore, other characteristics of sew- age sludge pyrolysis products have been analysed including the higher heating value (HHV) of the char, the gas composition and its lower heating value (LHV).
2. Materials and methods
2.1. Materials
A liquid condensation system placed next to the FixBR ensured the collection of the pyrolysis liquids. This system was composed of two condensers cooled with ice and an electrostatic precipitator (ESP). The composition of the non-condensable gases produced during the pyrolysis was determined on-line using a micro gas chromatograph (Agilent 3000A) connected down flow below the pyrolysis liquid condensation system. A schematic diagram of the whole experimental set-up is shown in Fig. 2 where the spatial dis- tribution of the equipment can be seen.
2.1.1. Raw material
2.3. Experimental procedure
Anaerobically digested and thermally dried sewage sludge pro- vided by a Spanish wastewater treatment plant was used as raw material. Sewage sludge (SS) samples were ground and sieved be- tween 250 and 500 lm prior to each experiment. The proximate, elemental and heating value analyses of the sewage sludge are shown in Table 1.
114 M. Azuara et al. / Fuel 107 (2013) 113–121
First, the FluBR and FixBR temperatures and the solid feed rate (g/min) were adjusted. The pyrolysis experiments were carried out during 2 h using nitrogen as fluidisation agent. Char obtained in a previous experiment at the same pyrolysis temperature was used as the initial bed material inside the FluBR in order to minimise the effect of the non-steady state [33,34]. The mass balance was done for each pyrolysis experiment taking into account the yield obtained for each product: gas, liquid and char, as well as coke deposition on the c-Al2O3.
Table 1 Raw material analyses (as received). Proximate analysisa Analytical method wt.%
Total moisture Ash Volatile matter Fixed carbon ISO-589-1981 ISO-1171-1976 ISO-5623-1974 By difference 7.1 41.0 46.6 5.3 Table 2 c-Al2O3 chemical composition.a Compounds wt.%
a Analyses performed by the Instituto de Carboquímica (CSIC, Zaragoza, Spain).
a Data provided by Sasol Germany GmbH.
Elemental analysis (organic fraction)a C (wt.%) H (wt.%) N (wt.%) S (wt.%) O (wt.%) ASTM D 5373 ASTM D 5373 ASTM D 5373 ASTM D 5373 By difference wt.% (dry basis) 29.8 3.9 4.2 0.9 13.1 Heating value HHV (MJ/kg) ASTM D-3286-96 11.9 95.0 min 0.05 max 0.035 max 0.025 max 0.005 max 0.27 max Al2O3 C SiO2 Fe2O3 Na2O TiO2
At the end of each experiment, the char yield (g char = g char/ 100 g SS fed) was measured taking into account the char collected in the hot filter, char collector and cyclone. The higher heating va- lue (HHV) of the char product was determined by means of a C2000 IKA Bomb Calorimeter.
The pyrolysis liquid yield (g liquid = g liquid/100 g S.S fed) was determined by weight difference of the liquid collector vessels be- fore and after the experiment. The pyrolysis liquid was taken di- rectly from the collector vessels without using any solvent as wash liquid and stored in a flask in a fridge until the analyses be- gan (maximum 1 week). Firstly, the pyrolysis liquid samples were centrifuged in a Heraeus Megafuge 16 centrifuge at 4500 rpm dur- ing 30 min in order to enhance the phase separation. The volumet- ric proportions of each phase were then determined visually [30]. The densities were measured using a Densito 30 PX Mettler Toledo densimeter and each phase yield was calculated in wt.%. The or- ganic phase HHV was also determined by means of a C2000 IKA bomb calorimeter.
Coke deposition on the spent c-Al2O3 sample (g coke = g coke/ 100 g SS fed) was determined at the end of each experiment, by quantifying the weight difference of the spent c-Al2O3 sample after burning with air in a muffle furnace at 550 (cid:2)C. The possible depo- sition of coke on the wall of the stainless-steal reactor was not measured.
115 M. Azuara et al. / Fuel 107 (2013) 113–121
2.4. Operational conditions
During the experiments the flow of outgoing gas and the gas composition were measured with a flow meter and a microGC, respectively, in order to calculate the quantity and content of the gas produced (g gas = g gas/100 g SS fed). The lower heating value (LHV) of the outgoing pyrolysis gas was determined mathemati- cally using the gas composition.
During pyrolysis, the dry sewage sludge was subjected to tem- peratures of 450 or 550 (cid:2)C (FluBR-T). The selection of these tem- peratures was based on the results of previous research into maximising the liquid yield in sewage sludge pyrolysis [16]. For the FixBR filled with c-Al2O3, temperature values of 400 and
Fig. 1. Scheme of the Fixed Bed Reactor.
Fig. 2. Scheme of the FluBR+FixBR lab-scale pyrolysis plant.
yield, liquid yield, char yield and coke deposition obtained in each experiment.
500 (cid:2)C were used (FixBR-T). This temperature interval was set tak- ing into account previous works giving 450 (cid:2)C as an optimum tem- perature for the catalytic treatment of triglycerides and fatty acids with c-Al2O3 [27–29].
Table 3 shows the operational conditions of the experiments. A feed rate of 6 g/min of sewage sludge (solid feed rate per volumetric unit 0.339 kg/(s m3)) and a nitrogen flow of 4.4 dm3 (NTP)/min (in- let nitrogen rate 0.074 m/s) was kept in each experiment [16,32]. Since the physical characteristics of sewage sludge enabled good fluidisation behaviour itself, sand was not added to the bed [30]. Sets of experiments (A0 and B0) were carried out without applying the FixBR in order to compare the sewage sludge pyrolysis results using the FluBR alone with those obtained using FluBR+FixBR.
As expected, the FixBR-T and c-Al2O3 quantity did not have an influence on the char yield which only altered when the FluBR-T changed from 450 to 550 (cid:2)C. The char yield was considered a rep- resentative and reliable parameter in order to check if the pyrolysis process had performed correctly in the FluBR. The pyrolysis reac- tion appears to have been similar in all the experiments, since the analysis did not reveal significant differences in the char yields. The mean (±confidence interval 90%) value of the char yields ob- tained in the different experiments was 55.3 wt.% (±0.32) at 450 (cid:2)C (Exp. A) and 51.0 wt.% (±0.71) at 550 (cid:2)C (Exp. B). The char yield decreased at the higher temperature pyrolysis in the FluBR, as expected from the results of previous research [15,16].
The amount of c-Al2O3 was chosen in accordance with previous studies [28,29,35,36] investigating improvements in the yield and upgrading of the liquid, taking into account the weight hour space velocity (WHSV). The fixed bed had a similar tube diameter to the FluBR, the pyrolysis vapour flow remaining constant.
Furthermore, the higher heating value (HHV) of the char was greater in the runs done at 450 (cid:2)C. This value is directly related to the raw material conversion, being greater for less converted solids. The char HHV was around 6.6 (±0.26) MJ/kg for experiments A (450 (cid:2)C-FluBR-T) and 5.4 (±0.18) MJ/kg for experiments B (550 (cid:2)C-FluBR-T).
The WHSV was defined as the pyrolysis vapour (g/h)/catalyst (g) [h(cid:2)1] ratio and calculated experimentally for each run, being be- tween 0.6 and 1.9 h(cid:2)1. Depending of the dimensions of the fixed bed 97.3 g corresponds to 5 cm of bed, and 194.6 g corresponds to 10 cm. In the A1 and B1 experiments, an inert bed of sand was used to compare these experiments with experiments A0 and B0, carried out without FixBR, in order to know the possible influence of the vapours passing through the FixBR on the liquid distribution.
3. Results and discussion
In the pyrolysis experiments carried out without applying c- Al2O3, the liquid yield increased slightly from 31.5 wt.% (A1) to 33.6 wt.% (B1) and the gas yield increased by around 100 wt.% (from 5.2 to 10.4 wt.%) because of an increase of the pyrolysis tem- perature. Comparing this liquid yield with experiments A0 and B0, the results obtained were similar. This confirms that the use of a FixBR filled with sand at the studied temperatures (400 and 500 (cid:2)C) had no effect on the pyrolysis liquid and char distribution if it is compared with the experiments in which only the FluBR was used (A1 and B1).
Fig. 4 shows the product yields for the experiments carried out with c-Al2O3. When the amount of c-Al2O3 inside the FixBR in- creased, the liquid yield decreased at both FixBR-T (400 and 500 (cid:2)C).
The c-Al2O3 behaviour and the evolution of the gas composition over the 120 min duration of the experiments were studied as a previous step. Fig. 3 shows the evolution of the gas composition during one of the experiments carried out (Exp. B2). As can be seen, the gas composition remained constant during the whole experi- ment which may imply that the catalyst was not deactivated dur- ing the experiment. This occurred in all the experiments carried out, and therefore it could be said that the amount of catalyst cho- sen was sufficient to maintain the catalytic activity during the run time.
Before the start-up of the experiments, temperature profiles were tested in order to check the FixBR temperatures. The longitu- dinal temperature was approximately constant and the radial tem- perature had a maximum deviation of ±15 (cid:2)C with respect to the value selected. The possible limitations resulting from the external diffusion control in the process were checked by increasing the pyrolysis vapour rate through the FixBR at the highest temperature 500 (cid:2)C. In these tests, a similar pyrolysis product distribution and gas composition was obtained (results not shown). Consequently, any changes occurring in the process were assumed to be caused by the c-Al2O3 application.
3.1. Product yields
The influence of the following operational conditions (FluBR-T, FixBR-T and c-Al2O3 quantity, respectively) on the pyrolysis prod- uct distribution is described in this section. Table 4 shows the gas
The liquid yield decrease was higher in experiments B1–2 (from 33.6 to 29 wt.%) than in experiments A1–2 (from 31.5 to 29 wt.%). Furthermore, the drop in the liquid yield is more accused for both combinations of temperatures when c-Al2O3 is introduced in the fixed bed. Incidentally, the liquid yield remained approximately constant for the same c-Al2O3 quantity, which was 29 wt.% (Exp. A2–B2) and 26 wt.% (Exp. A3–B3). The constant liquid yield obtained at the two temperatures studied does not mean that temperature does not influence yields but that at these particular conditions the larger generation of liquid by pyrolysis at 550 (cid:2)C FluBR-T is bal- anced with a larger extent of cracking towards gases. This fact in- volves two possible aspects: c-Al2O3 works better at 500 (cid:2)C than at 400 (cid:2)C or the vapours are more cracked (species that undergo cracking are not the same) when the sewage sludge is pyrolised in the fluidised bed at 550 (cid:2)C. Furthermore, this liquid yield de- crease is more marked in experiments B1–2 (4.6 wt.% drop in liquid yield) than in the experiments B2–3 in which more c-Al2O3 was used (2.6 wt.% drop). This trend matched those obtained by other researchers who studied the catalytic pyrolysis of other kinds of biomass [37,38]. The results relating to the influence of the pyroly- sis temperature on the gas yield show greater gas production, increasing from 5.2 to 10.4 wt.% (Exp. A1–B1). This is consistent with previous research results [16].
116 M. Azuara et al. / Fuel 107 (2013) 113–121
As regards the c-Al2O3 effect on the gas yield, it can be observed in Fig. 4 that the increase in the gas yield with the amount of c- Al2O3 is greater in experiments A1–3 (5.2–10.2 wt.%) than in exper- iments B1–3 (10.4–11.4 wt.%).
Table 3 Operational conditions used in the pyrolysis experiments. Experiment A0 B0 A1 B1 A2 B2 A3 B3
a FixBR filled with sand.
In the run times used, the coke deposition was influenced in both sets of experiments by the amount of c-Al2O3, the deposition increasing with the bed length. In non-catalytic sewage sludge pyrolysis (in this case, without applying c-Al2O3) the optimal tem- perature to obtain a high liquid yield is around 550 (cid:2)C (33.5 wt.%)
FluBR-T ((cid:2)C) FixBR-T ((cid:2)C) c- Al2O3 (g) WHSV (h(cid:2)1) 450 – – – 550 – – – 450 400 0a – 550 500 0a – 450 400 97.3 1.60 550 500 97.3 1.84 450 400 194.6 0.66 550 500 194.6 0.96
117 M. Azuara et al. / Fuel 107 (2013) 113–121
Fig. 3. Gas composition during experiment type B2 (550 (cid:2)C T-FluBR, 500 (cid:2)C T-FixBR, 97 g c-Al2O3).
r
r
r
r
Balance
c-Al2O3 (g)
ggas (wt.%)
glíq (wt.%)
gchar (wt.%)
Coke (wt.%)
closure
⁄
⁄
⁄
⁄
⁄
⁄
Table 4 Product distribution. CI CI CI CI Exp. FluBR-T ((cid:2)C) FixBR-T ((cid:2)C)
0.4 ⁄ ±1.8 ⁄ 0.9 ⁄ ±4.0 ⁄
⁄
⁄
⁄
⁄
1.6 ⁄ ±2.8 ⁄ 1.0 ⁄ ±1.7 ⁄ 0.6 ⁄ ±1.0 ⁄ 0 0 0.3 ⁄ 0 0 ±0.5 ⁄ 5.2 6.6 10.2 ⁄ 92.3 91.0 91.6 ⁄ 0.2 ⁄ ±0.9 ⁄ 1.0 ⁄ ±4.4 ⁄
1.2 ⁄ ±2.0 ⁄ 1.0 ⁄ ±1.7 ⁄ 0.9 ⁄ ±1.5 ⁄ 0 0 0.2 ⁄ 0 0 ±0.3 ⁄ 450 450 450 450 550 550 550 550 – 400 400 400 – 500 500 500 – 0⁄⁄ 97.3 194.6 – 0⁄⁄ 97.3 194.6 10.4 10.8 11.4 31.1 31.5 29.0 26.8 33.4 33.6 29.0 26.4 55.3 55.6 55.4 54.6 51.7 51.0 50.7 53.0 0 0 1.7 2.7 0 0 1.5 3.0 95.0 92.0 93.8 A0 A1 A2 A3 B0 B1 B2 B3
Average values, n = 1 (Exp. A1, B1, A3, B3), n = 2 (Exp. A0, B0), n = 3 (Exp. A2, B2). r = Standard deviation. CI = 90% Confidence interval. (cid:3) Not determined. (cid:3)(cid:3) Filled with inert sand.
[39]. At this temperature a greater amount of gas and liquid (vola- tiles) pass through the FixBR filled with c-Al2O3. As a consequence, the coke content in B2 and B3 (550 (cid:2)C) should be higher than in A2 and A3 (450 (cid:2)C), but in fact the differences were not significant.
wt.%; gas compoundi g/100 g SS) in the experiments carried out at 450 and 550 (cid:2)C, respectively. In these figures, it can be observed that for all the experiments CO2 is the most abundant compound (4–6 wt.%) and H2 the less abundant one (0.01–0.13 wt.%).
3.2. Gas composition and LHV
Fig. 5 shows an increase in the CO (from 0.6 to 2.3 wt.%) and the H2 content (from 0.01 to 0.07 wt.%), possibly linked with secondary reactions of primary compounds as the amount of c-Al2O3 in- creased [40].
In Fig. 6, significant variations are only observed for H2S and H2, the yield to both compounds increasing with the amount of c- Al2O3. The yield of H2 increases from 0.07 wt.% to 0.13 wt.%. The
The pyrolysis gas obtained was composed mostly of CO, CO2, H2, H2S and some traces of light hydrocarbons (CH4, C2H2, C2H4 and C2H6). Furthermore, the rest of the gas (up to 100% v/v) was com- posed of N2. Figs. 5 and 6 show the gas compositions (expressed in
Fig. 4. Product yields vs. c-Al2O3 (g).
118 M. Azuara et al. / Fuel 107 (2013) 113–121
Fig. 5. Gas compounds distribution obtained in the experiments A. Error bars represent confidence interval (90%).
yield to H2S obtained in experiment B3 (around 0.9 wt.%) points out that most of the sulphur from the feedstock ends up in the gas at these operational conditions.
Comparing the gas composition obtained for each set of tem- peratures (Figs. 5 and 6), there are greater quantities of CO, H2S and H2 in experiments B than in experiments A (H2 in experiments A, 0.01–0.07 wt.% vs. H2 in experiments B 0.07–0.13 wt.%). The yields of the rest of non-condensable gases detected are not af- fected by the temperature.
optimum organic phase yield in sewage sludge pyrolysis liquid. One of the main characteristics of this pyrolysis liquid is the exis- tence of two or three phases depending on the temperature used in the pyrolysis process [30]. Sewage sludge pyrolysis in a FluBR at 450 (cid:2)C (Exp. A0) produced two phases, an aqueous (A.P.) and a vis- cous heavy organic phase (H.O.P.). This H.O.P. was apparently sim- ilar to the H.O.P. obtained at 550 (cid:2)C. However, in pyrolysis at 550 (cid:2)C (Exp. B0) three phases were determined: the aforemen- tioned H.O.P. in the middle of the liquid, the A.P.at the bottom and an organic phase at the top called the light organic phase (L.O.P.) [30]. These phases were also obtained when inert sand was placed in the FixBR (Exp. A1 and B1). As explained above, the liquids obtained were heterogeneous and separated into three phases in experiments B0 and B1 and into two phases in experi- ments A0 and A1. These phases had very similar colours (see Fig. 9-left side).
The LHV matches the gas yield trend observed with the amount of c-Al2O3 and with the temperature. At each set of temperatures the greatest lower heating values (Table 5) of the pyrolysis gas gen- erated in the experiments are obtained in the experiments A3 and B3, which were carried out with the maximum amount of c-Al2O3. The LHV of the experiments B are around twice higher than those obtained in experiments A, it is to say the LHV of the gas generated is higher when the temperature increases. The values obtained are quite similar to gas products obtained from the pyrolysis of other biomasses [41].
For ease of comparison and interpretation, schemes of the pyro- lysis experiments are depicted in Figs. 7 and 8 summarising the operational conditions, the results obtained for the phase distribu- tion (expressed in wt.%; phasei g/100 g SS) and their connection with the product yields.
3.3. Phase distribution and HHV of the pyrolysis liquid
Previous research [7,8,42] and studies carried out by the GPT [16,32] suggest a temperature of around 550 (cid:2)C to obtain an
As can be seen in Fig. 7, the results obtained in experiments A0 and A1 showed that there is no influence of FixBR-T on the H.O.P. and A.P. They appear in an approximate ratio of 1/3 wt.%, respec- tively, of the whole liquid.
Fig. 6. Gas compounds distribution obtained in the experiments B. Error bars represent confidence interval (90%).
In the experiments using c-Al2O3, the O.P. obtained is larger (from 8 to 11 wt.%) due to a clear A.P. decrease (from 24 to 18 wt.%). When increasing the amount of c-Al2O3, the O.P. yield is quite similar (11–10 wt.%) for both experimental sets (A2–A3), the same trend occurring with the A.P. (18–16 wt.%).
Table 5 Gas pyrolysis LHV. Exp. A1 A2 A3 B1 B3 B2 LHV [MJ/Nm3] 3.7 7.6 12.0 15.2
5.2 r (0.02) CI (±0.03) 12.2 r (0.02) CI (±0.03)
Fig. 8 shows that in experiments B0 and B1, the O.P. and the A.P. are in an approximate ratio of 1/2 wt.%, respectively, of the whole pyrolysis liquid. The liquid (around 33–34 wt.%) and phase yields show that there is no difference between experiments B0 and B1.
Average values, n = 1 (Exp. A1, B1, A3, B3), n = 3 (A2, B2). r = Standard deviation. CI = 90% Confidence interval.
119 M. Azuara et al. / Fuel 107 (2013) 113–121
Fig. 7. Product and phase yields in the experiments A. (450 (cid:2)C FluBR-T/400 (cid:2)C FixBR-T/0-97.3–194.6 c-Al2O3 g). Confidence interval (90%) is between brackets. Average values, n = 1 (Exp. A1, A3), n = 2 (Exp. A0), n = 3 (Exp. A2). aGlobal average of the char yield (n = 7).
The original O.P. yield (L.O.P. + H.O.P. in the experiment B0 and B1, O.P. in experiments B2 and B3) remains approximately constant (11–12 wt.%) when the amount of c-Al2O3 used is 97.4 g (Exp. B2).
In other words, the O.P. yield obtained in experiments B2 is similar to the sum of the H.O.P. and L.O.P. yields obtained in experiments B0 and B1.
Fig. 8. Product and phase yields in the experiments B. (550 (cid:2)C FluBR-T/500 (cid:2)C FixBR-T/0-97.3–194.6 c-Al2O3 g). Confidence interval (90%) is between brackets. Average values, n = 1 (Exp. B1, B3), n = 2 (Exp. B0), n = 3 (Exp. B2). aGlobal average of the char yield (n = 7).
120 M. Azuara et al. / Fuel 107 (2013) 113–121
Fig. 9. Pylolysis liquid samples, left picture belongs to a liquid obtained in pylolysis experiments B1 and right picture using c-Al2O3 Exp. B2.
However, there is a clear drop in the O.P. yield ((cid:2)5 wt.%) in experiment B3 comparing with experiment B2, the O.P yield decreasing from 11 to 6 wt.% possibly because of a higher gas yield (+0.6 wt.%), a higher coke yield (+1.5 wt.%) and an increase in the A.P. yield (2 wt.%) of the pyrolysis liquid.
Table 6 Organic phase HHV. Exp. A1 (H.O.P) A2 A3 B1 (H.O.P) B2 B3 HHV [MJ/kg] 28.7 38.0 33.6 40.3
38.4 r (0.5) CI (±0.8) 40.6 r (0.3) CI (±0.6)
Fig. 8 also shows a slight decrease in the A.P. yield in the exper- iments in which c-Al2O3 is used (from 22–23 to 18 wt.%). The c- Al2O3 application could influence the process by removing a cer- tain amount of water, generating a lower A.P. yield. The O.P. yield is in line with the liquid yield, both clearly decreasing in experi- ment B3 as the gas yield increases considerably.
liquid properties should also be studied in order to determine the optimal operational conditions.
Comparing both figures (Figs. 7 and 8), the organic and aqueous phases yields obtained for non-catalytic pyrolysis A0 and B0 are similar to those obtained in the other non-catalytic experiments A1 and B1, respectively. This fact indicates that the use of a FixBR filled with sand at 400 and 500 (cid:2)C respectively, do not affect the yields to these products. Therefore the differences found in the or- ganic and aqueous phases yields in the catalytic experiments would be caused only by the use of c-Al2O3.
Applying the c-Al2O3, the O.P. production increases in the experiments A (from 6–8 to 11 wt.%). However, at 550 (cid:2)C (experiments B) a drop in the O.P. yield is obtained in the experi- ment B3. In the experiments A, the liquid yield decrease is related with the A.P. loss apart from the gas and the coke yield augments. However, in experiments B the yield to the aqueous phase remains more or less constant while the whole O.P. yield decreases mark- edly. This fact could be due to excessive catalytic cracking at 500 (cid:2)C [43], in which heavy organic compounds could be converted into gas and coke decreasing the liquid organic compounds.
In the B2 and B3 experiments, only two phases of different colours were obtained when c-Al2O3 was used in the fixed bed. This reduc- tion in the number of phases and the change in appearance make their separation easier. Thus a significant advantage of using c- Al2O3 is the obvious improvement in the phase separation of the pyrolysis liquid of sewage sludge. Otherwise the properties and appearance of the whole pyrolysis liquid are different, especially the A.P. colour which is dark orange in non-catalytic pyrolysis and light yellow when c-Al2O3 is used in the process (Fig. 9-right side). The higher heating value of the O.P. has been studied and the results obtained are depicted in Table 6. The HHV is greater in the experiments at 550 (cid:2)C (Exp. B1) than at 450 (cid:2)C (Exp. A1). These results are consistent with those in the literature [27]. The O.P. higher heating value (HHV) is greater than that obtained in pyroly- sis liquids from other raw materials (wood, corn cob, rice husk) [44], some of these materials being around 40 MJ/kg (Exp. B2 and B3). Comparing pyrolysis with or without the use of c-Al2O3 at dif- ferent temperatures, the HHV increase is lower between experi- ments B1–B2 (20%) than A1–A2 (25%). Moreover, the influence of the pyrolysis temperature is mitigated in the experiments using c-Al2O3 because the HHV values are quite similar in both experi- mental sets (A2–B2 and A3–B3).
4. Conclusions
The data indicates that the maximum O.P. yields are obtained for a bed length of 5 cm, using any of the combinations of temper- atures studied. For both combinations of temperatures, major amount of c-Al2O3 (10 cm) does not mean an augment in the O.P. yield neither in the higher heating value of the organic phases. However, for 5 cm of c-Al2O3 the higher heating value of the O.P. of the experiment B2 is greater than the one of the experiment A2. The lower HHV of the O.P. of the experiment A2 ((cid:2)2.2 MJ/kg) might be compensated by the greater power needed to conduct the reaction at the highest combination of temperatures 550 (cid:2)C in the FluBR and 500 (cid:2)C in the FixBR (experiment B2). Apart from this, other
This work studies the influence on the product distribution of sewage sludge pyrolysis using a catalytic treatment of the vapours generated in a fixed bed filled with c-Al2O3.
Average values, n = 1 (Exp. A1, B1, A3, B3), n = 3 (A2, B2). r = Standard deviation. CI = 90% Confidence interval.
121 M. Azuara et al. / Fuel 107 (2013) 113–121
[14] Briens C, Piskorz J, Berruti F. Biomass valorization for fuel and chemicals production – a review. Int J Chem React Eng 2008;6:1–51.
[15] Pokorna E, Postelmans N, Jenicek P, Schreurs S, Carleer R, Yperman J. Study of bio-oils and solids from flash pyrolysis of sewage sludges. Fuel 2009;88: 1344–50.
[16] Fonts I, Juan A, Gea G, Murillo MB, Sánchez JL. Sewage sludge pyrolysis in influence of operational conditions on the product fluidized bed, 1: distribution. Ind Eng Chem Res 2008;47:5376–85.
The results have shown that c-Al2O3 is useful for obtaining a better pyrolysis liquid from the point of view of its phase separa- tion. Furthermore, in experiments in which c-Al2O3 is used, the yield to the organic phase of the pyrolysis liquid (which has the best fuel properties) increased under the operational conditions of 450 (cid:2)C FluBR-T/400 (cid:2)C FixBR-T/97.3–194.6 c-Al2O3 g compared to pyrolysis experiments of sewage sludge carried out at 450 (cid:2)C FluBR-T/400 (cid:2)C FixBR-T/Inert sand. Furthermore, a clear decrease in the organic phase yield in keeping with the gas yield increase in the pyrolysis process was obtained at 550 (cid:2)C FluBR-T/500 (cid:2)C Fix- BR/194.6 c-Al2O3 g,
[17] Thipkhunthod P, Meeyoo V, Rangsunvigit P, Rirksomboon T. Describing sewage sludge pyrolysis kinetics by a combination of biomass fractions decomposition. J Anal Appl Pyrol 2007;79:78–85. [18] Dyroff GV. Manual of significance of tests for petroleum products. ASTM Manual series MNL 1; 1989. [19] Mohan D, Pittman CU, Steele PH. Pyrolysis of wood/biomass for bio-oil: a critical review. Energy Fuels 2006;20:848–89. [20] Samolada MC, Papafotica A, Vasalos IA. Catalyst evaluation for catalytic biomass pyrolysis. Energy Fuels 2000;14:1161–7.
[21] Kersten SRA, van Swaaij WPM, Lefferts L, Seshan K. Options for catalysis in the thermochemical conversion of biomass into fuels. Catalysis for renewables; 2007.
As regards the product distribution, the results show a clear influence of the amount of c-Al2O3 on the liquid phase yields. One of the objectives of using c-Al2O3 is to lower the pyrolysis temperature in order to save energy consumption, while obtaining an organic phase with adequate characteristics for use as a bio-oil. It has been demonstrated that the organic phase increased using c- Al2O3 at 400 (cid:2)C FixBR-T and 450 (cid:2)C FluBR-T, reaching a similar or- ganic phase yield to that obtained in non-catalytic pyrolysis of sewage sludge at 550 (cid:2)C in FluBR-T and obtaining an improved li- quid in terms of its HHV.
[22] Ates(cid:2) F, Pütün AE, Pütün E. Pyrolysis of two different biomass samples in a fixed-bed reactor combined with two different catalysts. Fuel 2006;85: 1851–9. [23] Güllü D. Effect of catalyst on yield of liquid products from biomass via pyrolysis. Energy Source 2003;25:753–65.
[24] Antonakou E, Lappas A, Nilsen MH, Bouzga A, Stöcker M. Evaluation of various types of Al-MCM-41 materials as catalysts in biomass pyrolysis for the production of bio-fuels and chemicals. Fuel 2006;85:2202–12.
[25] Kim Y, Parker W. A technical and economic evaluation of the pyrolysis of sewage sludge for the production of bio-oil. Bioresour Technol 2008;99: 1409–16.
[26] Beckers W, Schuller D, Vaizert O. Thermolytical treatment of dried sewage sludge and other biogenic materials—including upgrading of pyrolysis vapours by a cracking catalyst and examination of heavy metals by X-ray fluorescence. J Anal Appl Pyrol 1999;50:17–30. [27] Maher KD, Bressler DC. Pyrolysis of triglyceride materials for the production of renewable fuels and chemicals. Bioresour Technol 2007;98:2351–68.
Taking into account the organic phase yield and its energy con- tent (comparable to a conventional fuel), sewage sludge pyrolysis combined with c-Al2O3 catalytic treatment may be considered an appropriate method of valorising this waste, representing a new way to research its possible application as a bio-oil. Furthermore, possible uses for the aqueous phase as a fertiliser or a source of chemicals could be the subject of future research. Organic phase properties such as viscosity, chemical composition, and stability, should be analysed for the purposes of comparison with other bio-oils from other biomass sources such as wood or animal wastes.
[28] Konar SK, Boocock DGB, Mao V, Liu J. Fuels and chemicals from sewage sludge: 3. Hydrocarbon liquids from the catalytic pyrolysis of sewage sludge lipids over activated alumina. Fuel 1994;73:642–6.
Acknowledgments
[29] Vonghia E, Boocock DGB, Konar SK, Leung A. Pathways for the deoxygenation of triglycerides to aliphatic hydrocarbons over activated alumina. Energy Fuels 1995;9:1090–6.
[30] Gil-Lalaguna N, Fonts I, Gea G, Murillo MB, Lázaro L. Reduction of water content in sewage sludge pyrolysis liquid by selective online condensation of the vapors. Energy Fuels 2010;24:6555–64.
The authors express their gratitude to the Spanish Ministry of Science and Innovation (MICINN) for providing frame support for this work (Research Project CTQ2010-20137/PPQ) and for the doc- toral grant awarded to M. Azuara (BES-2008-002004).
References
[31] Fonts I, Azuara M, Lázaro L, Gea G, Murillo MB. Gas chromatography study of sewage sludge pyrolysis liquids obtained at different operational conditions in a fluidized bed. Ind Eng Chem Res 2009;48:5907–15. [32] Fonts I, Azuara M, Gea G, Murillo MB. Study of the pyrolysis liquids obtained from different sewage sludge. J Anal Appl Pyrol 2009;85:184–91.
[33] Cousins A, Zhuo Y, George A, Paterson N, Dugwell DR, Kandiyoti R.
Development of a bench-scale high-pressure fluidized bed reactor and its
sequential modification for studying diverse aspects of pyrolysis and
gasification of coal and biomass. Energy Fuels 2008;22:2491–503. [1] Laturnus F, von Arnold K, Grøn C. Organic contaminants from sewage sludge
applied to agricultural soils False alarm regarding possible problems for food
safety? Environ Sci Pollut Res 2007;14:53–60. [2] Eurostat. Total sewage sludge production from urban wastewater. [23.08.11] [34] Aznar M, González AE, Manyà JJ, Sánchez JL, Murillo MB. Understanding the
effect of the transition period during the air gasification of dried sewage sludge
in a fluidized bed reactor. Int J Chem React Eng 2007;5.
[37] Zhang H, Xiao R, Huang H, Xiao G. Comparison of non-catalytic and catalytic fast pyrolysis of corncob in a fluidized bed reactor. Bioresour Technol 2009;100:1428–34. [5] Inguanzo M, Domínguez A, Menéndez JA, Blanco CG, Pis JJ. On the pyrolysis of sewage sludge: the influence of pyrolysis conditions on solid, liquid and gas fractions. J Anal Appl Pyrol 2002;63:209–22. [38] Ates(cid:2) F, Isikdag MA. Influence of temperature and alumina catalyst on pyrolysis [6] Piskorz J, Scott DS, Westerberg IB. Flash pyrolysis of sewage sludge. Ind Eng of corncob. Fuel 2009;88:1991–7. Chem Proc DD 1986;25:265–70. [39] Fonts I, Gea G, Azuara M, Ábrego J, Arauzo J. Sewage sludge pyrolysis for liquid [7] Kaminsky W, Kummer AB. Fluidized bed pyrolysis of digested sewage sludge. J production: a review. Renew Sust Energ Rev 2012;16:2781–805. Anal Appl Pyrol 1989;16:27–35. [8] Shen L, Zhang D-K. An experimental study of oil recovery from sewage sludge by low-temperature pyrolysis in a fluidised-bed. Fuel 2003;82:465–72. [40] Xu W-C, Tomita A. The effects of temperature and residence time on the secondary reactions of volatiles from coal pyrolysis. Fuel Process Technol 1989;21:25–37. [9] Karayildirim T, Yanik J, Yuksel M, Bockhorn H. Characterisation of products from pyrolysis of waste sludges. Fuel 2006;85:1498–508. [10] Fonts I, Kuoppala E, Oasmaa A. Physicochemical properties of product liquid [41] Chen G, Andries J, Luo Z, Spliethoff H. Biomass pyrolysis/gasification for product gas production: the overall investigation of parametric effects. Energ Convers Manage 2003;44:1875–84. from pyrolysis of sewage sludge. Energy Fuels 2009;23:4121–8.
[42] Park HJ, Heo HS, Park Y-K, Yim J-H, Jeon J-K, Park J, Ryu C, Kim SS. Clean bio-oil production from fast pyrolysis of sewage sludge: effects of reaction conditions and metal oxide catalysts. Bioresour Technol 2010;101:S83–5. [11] Samanya J, Hornung A, Apfelbacher A, Vale P. Characteristics of the upper phase of bio-oil obtained from co-pyrolysis of sewage sludge with wood, rapeseed and straw. J Anal Appl Pyrol 2011;94:120–5. [12] Czernik S, Bridgwater AV. Overview of applications of biomass fast pyrolysis [43] Wang P, Zhan S, Yu H, Xue X, Hong N. The effects of temperature and catalysts industrial wastes (herb residue). Bioresour Technol oil. Energy Fuels 2004;18:590–8. on the pyrolysis of 2010;101:3236–41. [44] Raveendran K, Ganesh A. Heating value of biomass and biomass pyrolysis products. Fuel 1996;75:1715–20. [13] Azuara M, Ábrego J, Fonts I, Gea G, Murillo MB. Ammonia content of bottom phase liquid from pyrolysis of sewage sludge in a bubbling fluidized bed. In: Proceedings of the 18th European biomass conference & exhibition; 2010.
