Waste Management xxx (2015) xxx–xxx

Contents lists available at ScienceDirect

Waste Management journal homepage: www.elsevier.com/locate/wasman

Alkali activation of recovered fuel–biofuel fly ash from fluidised-bed combustion: Stabilisation/solidification of heavy metals Juho Yliniemi a,⇑, Janne Pesonen b, Minna Tiainen b, Mirja Illikainen a a b

Fibre and Particle Engineering Laboratory, P.O. Box 4300, University of Oulu, 90014, Finland Department of Chemistry, P.O. Box 3000, University of Oulu, 90014, Finland

a r t i c l e

i n f o

Article history: Received 10 February 2015 Accepted 18 May 2015 Available online xxxx Keywords: Recovered fuel Fluidized bed combustion Fly ash Sequential leaching procedure Alkali activation Immobilization

a b s t r a c t Recovered fuel–biofuel fly ash from a fluidized bed boiler was alkali-activated and granulated with a sodium-silicate solution in order to immobilise the heavy metals it contains. The effect of blast-furnace slag and metakaolin as co-binders were studied. Leaching standard EN 12457-3 was applied to evaluate the immobilisation potential. The results showed that Ba, Pb and Zn were effectively immobilised. However, there was increased leaching after alkali activation for As, Cu, Mo, Sb and V. The co-binders had minimal or even negative effect on the immobilisation. One exception was found for Cr, in which the slag decreased leaching, and one was found for Cu, in which the slag increased leaching. A sequential leaching procedure was utilized to gain a deeper understanding of the immobilisation mechanism. By using a sequential leaching procedure it is possible fractionate elements into watersoluble, acid-soluble, easily-reduced and oxidisable fractions, yielding a total ‘bioavailable’ amount that is potentially hazardous for the environment. It was found that the total bioavailable amount was lower following alkali activation for all heavy metals, although the water-soluble fraction was higher for some metals. Evidence from leaching tests suggests the immobilisation mechanism was chemical retention, or trapping inside the alkali activation reaction products, rather than physical retention, adsorption or precipitation as hydroxides. Ó 2015 Elsevier Ltd. All rights reserved.

1. Introduction The EU’s energy policy encourages the use of renewable biofuels and waste for energy production (Europe 2020, 2010). Recovered Fuel (REF) or Solid Recovered Fuel (SRF) is produced from municipal or industrial waste (SFS-EN 15359, 2011). By using REF for energy production, less waste ends up in landfills and the need for fossil fuels decreases. Suitable methods for producing energy from REFs are combustion, gasification and pyrolysis (VTT Report, 2000). In Finland, REF is mostly co-combusted with biofuels or peat in fluidized bed boilers (FBB) in combined heat and power plants (PlasticsEurope Report, 2008). The total amount of FBB ash generated annually in Finland is 600,000 t, including ash from biomass, peat and REF combustion (Emilsson, 2006). The temperature in an FBB is 800– 900 °C, which creates a stable combustion environment for different solid fuels. Because REF is produced from waste, it contains high levels of hazardous elements such as Pb, Zn and Sb. Hazardous elements ⇑ Corresponding author. Tel.: +358 443120686. E-mail address: juho.yliniemi@oulu.fi (J. Yliniemi).

are found in greater concentrations in fly ash (FA) generated from REF combustion. The high levels of heavy metals in REF FA prevent its use (FINLEXÒ, 2006). This forces energy producers to transport the FA to landfills, leading to greater cost and environmental risk. A workaround would be recovering heavy metals from FA (Ehsan et al., 2006; Isoyama and Wada, 2007; Moutsatsou et al., 2006; Mulligan et al., 2001; Schnoor, 1997). However, often, this is not technologically or economically feasible, so stabilization/solidification (S/S) is the next best option (Bobrowski et al., 1997; Glasser, 1997; Malviya and Chaudhary, 2006). Recently alkali activation, also referred to as geopolymerization, has received attention as the S/S method of choice (Bankowski et al., 2004a, 2004b; Fernandez-Jimenez et al., 2005; Guo and Shi, 2012, 2013a; Jaarsveld et al., 1998; Komnitsas et al., 2013; Lancellotti et al., 2010; Luna Galiano et al., 2011; Luna et al., 2009; Nikolic´ et al., 2014; Ogundiran et al., 2013; Perera et al., 2005; Phair et al., 2004; Provis, 2009; Tzanakos et al., 2014; van Jaarsveld et al., 1999, 1997; Xu et al., 2006; Zhang et al., 2008a, 2008b; Zheng et al., 2011, 2010). In alkali activation, solid aluminosilicate precursors such as metakaolin, coal FA or blast furnace slag are dissolved in an alkaline solution, usually sodium hydroxide or sodium silicate, to produce a three-dimensional, aluminosilicate

http://dx.doi.org/10.1016/j.wasman.2015.05.019 0956-053X/Ó 2015 Elsevier Ltd. All rights reserved.

Please cite this article in press as: Yliniemi, J., et al. Alkali activation of recovered fuel–biofuel fly ash from fluidised-bed combustion: Stabilisation/solidification of heavy metals. Waste Management (2015), http://dx.doi.org/10.1016/j.wasman.2015.05.019

2

J. Yliniemi et al. / Waste Management xxx (2015) xxx–xxx

network structure. Results of the alkali activation of FA and slag to stabilize/solidify heavy metals are encouraging. However, the precise S/S mechanism remains unknown. Possible mechanisms can be divided into physical and chemical immobilisation, although a clear-cut distinction is impractical because these mechanisms can operate jointly (Glasser, 1997). Physical immobilisation occurs on the micron scale. Zeolite-type aluminosilicate-structures could trap some atoms within their crystal structures. Other mechanisms by which physical immobilisation can occur include prevention of contact between the leaching medium and the heavy metals via a physical barrier. Chemical immobilisation occurs on the nanoscale. In chemical immobilisation, metals react with other elements and become part of the aluminosilicate structure, possibly by replacing silicon atoms (Schoenung, 2008). In addition, leaching could decrease because of adsorption forces or precipitation as hydroxides (Bernal et al., 2014). This paper examines the S/S of FA from the co-combustion of REF and biofuels using alkali activation. To date, alkali activation of REF FA has not been studied. Fly ash from fluidized bed–combustion of coal has been studied recently, and promising results have been obtained with other binding materials (Chindaprasirt et al., 2011; Li et al., 2012a; Liu et al., 2014; Slavik et al., 2008; Tyni et al., 2014; Xu et al., 2010). However, S/S of FBB FA by alkali activation has not been studied. The combustion temperature in FBBs is lower than that in coal-burning facilities (800–900 °C vs. 1200–1300 °C). A lower combustion temperature yields lower glassy-phase material and, thus, potentially less reactive FA in alkali activation (Chindaprasirt and Rattanasak, 2010; Li et al., 2012b). In addition, biofuel and peat FA contain significant amounts of alkali, earth-alkali, and iron compounds which complicate alkali-activation reaction chemistry. The method of choice in this study is to simultaneously granulate and alkali-activate REF FA with a sodium-silicate solution to produce spherical granules. If S/S is successful, these granules could be used in construction, civil engineering or as a substitute for sand and gravel in concrete. In addition to S/S capability, the physical properties and microstructure of the granules should be analysed to estimate their potential for use in civil engineering. These properties will be evaluated in further studies. The aim of this work is to define the efficiency of the S/S of heavy metals by leaching standard (SFS-EN 12457-3, 2002) and by a sequential extraction procedure (Bruder-Hubscher et al., 2002). In the sequential extraction procedure, various types of chemical reagents are applied to the samples in series, and each successive treatment is more drastic than previous one. This way, it is possible to estimate the mobility (i.e. bioavailability) of heavy metals in environmental samples. Loosely bound metals such as those that are water-soluble and fraction leachable with ammonium acetate or acetic acid are much more mobile in the environment than those that are leachable only with strong acids such as hydrogen-fluoric acid (HF) and aqua regia. Thus, the loosely bound fractions are much more likely to be released in the environment and are, thus, potentially bioavailable (Filgueiras et al., 2002). Additionally, considering S/S by alkali activation, the sequential extraction procedure can reveal more information about heavy metal species before and after alkali activation, thus unveiling the immobilisation mechanism.

(not PVC), cartons, paper and wood collected from industry and retail suppliers. The REF was of Class 1 quality according to the relevant standard (SFS-EN 15359, 2011). The fuel mixture used in the power plant comprised 50% REF and 50% biofuel (mainly wood bark). The FA sample was collected from electrostatic precipitators (ESP) into 10-L containers. The effects of blast-furnace slag (S) (Finnsementti, KJ400) and metakaolin (M) (Metastar 402, Imerys Minerals) as co-binders were studied by adding them in 20 wt% and 40 wt%, respectively, to the FA. The samples were named FA100 (100% – FA), FA60S40 (60% – FA, 40% – blast-furnace slag), FA80S20 (80% – FA, 20% – blast-furnace slag), FA60M40 (60% – FA, 40% – metakaolin) and FA80M20 (80% – FA, 20% – metakaolin). Sodium-silicate solution (ZeopolÒ 25, Huber) (Na-Sil) was used as an alkali activator. Na-Sil had a SiO2/Na2O molar ratio of 2.5, pH of 12.5 and water content of approximately 66 wt%. 2.2. Methods The chemical composition of the precursors was determined with X-ray fluorescence (XRF) (Omnian Pananalytics Axiosmax) from a melt-fused tablet. The particle-size distributions were measured with a Beckman Coulter LS 13320 and reported as volumetric-based sizes (d10, d50 and d90). Specific surface area measurement was based on the physical adsorption of gas molecules on a solid surface using a physisorption analyzer (ASAP 2020, Micrometrics) and the results were reported in a form of a BET isotherm. Dry matter content and loss-of-ignition (LOI) at 525 °C and 950 °C were determined using thermo-gravimetric analysis equipment (PrepAsh, Precisa). A field emission scanning electron microscope (FESEM, Zeiss Ultra Plus) was used to analyse the FA particles. The sample distance was 7.3 mm, and the acceleration voltage was 3.0 kV. The pseudo-total concentrations of elements in FA and alkali-activated granules were characterised by microwave-assisted wet digestion using a 3:1 mixture of HNO3 and HCl for 0.5 g of FA and determination was made using a inductively coupled, plasma-optical emission spectrometer (ICP-OES) (Thermo Electron IRIS Intrepid II XDL Duo, Thermo Scientific). Duplicate measurements were made for each sample. 2.3. Granulation

2. Experimental

A high shear mixer (Eirich R-01) was used to granulate the precursors. A high-shear mixer was chosen for this study because it can spread viscous fluids and produce dense and strong granules with a narrow size distribution within a short period (Reynolds et al., 2007). The granulator has a 5-L drum rotating clockwise at a speed of 45 rpm. Inside the drum is an impeller measuring in 10 cm in diameter and spinning counter-clockwise at a speed of 900 rpm. The tilt angle of the drum is 10°. There is a scraping blade inside the drum to remove any material stuck to the drum wall and to compact the balled material. The granulation procedure employed in this study is as follows: (1) dry precursors were weighed, mixed and added to the drum; (2) before switching on the drum and the impeller, approximately 15 g of Na-Sil was sprayed on the precursors to prevent dusting; (3) drops of Na-Sil were added until the target granule size (2–4 mm diameter) was achieved. Each batch was sealed in air-tight plastic bags and stored at room temperature for 28 days before the granules were analysed.

2.1. Materials

2.4. Leaching tests

The REF-biofuel FA used in this study was obtained from a heat and electricity power plant that uses a bubbling FBB to combust REF. The REF was made of packing material waste such as plastic

2.4.1. Standard EN 12457-3 The standard leaching method (SFS-EN 12457-3, 2002) was used in this study. This two-step extraction method uses water

Please cite this article in press as: Yliniemi, J., et al. Alkali activation of recovered fuel–biofuel fly ash from fluidised-bed combustion: Stabilisation/solidification of heavy metals. Waste Management (2015), http://dx.doi.org/10.1016/j.wasman.2015.05.019

J. Yliniemi et al. / Waste Management xxx (2015) xxx–xxx

as the extraction fluid. This standard method is used as a quality-control test in Finland for fly and bottom ashes used in construction (FINLEXÒ, 2006). In the first step, the sample is mixed in an end-over-end tumbler for 6 h with a liquid/solid (L/S) ratio of 2. In the second step, the sample from step 1 is mixed for 18 h with L/S = 8. Eluates from both steps were analysed with the ICP-OESmethod and a cumulative release of L/S = 10 was calculated. Deviating from the standard, a sample size of 17.5 g was used instead of 175 g, and the filter-paper retention size was 1 lm instead of 0.45 lm. For analysis, granule-size fractions between 2 and 4 mm were used. Duplicate samples of FA and each granule batch were analysed and the averages were calculated. 2.4.2. Sequential extraction procedure The sequential extraction procedure used in this study was based on the process outlined in (Rauret et al., 1999). An exception was made in the first step by adjusting the pH of the water to 4.0 with HNO3 according to Nurmesniemi et al., 2008; Száková et al., 2000. Water extraction (with pH adjusted to 4.0) as the first step was employed to mimic the leachability effect of acid rain in the environment. This modified sequential extraction procedure has been used to fractionate elements in FA (Nurmesniemi et al., 2008; Pöykiö et al., 2005), bottom ash (Nurmesniemi et al., 2008), lime waste (Pöykiö et al., 2006) and green-liquor dregs (Nurmesniemi et al., 2005). The concept underlying sequential extraction is the division of leachable elements into four fractions: (1) water-soluble fraction, which is leachable with H2O, (2) acid-soluble fraction, which is leachable with CH3COOH, (3) easily-reduced fraction, which is leachable with HONH3Cl, and (4) oxidisable fraction, which is leachable with a mixture of H2O2 and CH3COONH4. Reagents, mixing time and temperature are summarised in Table 1. The concentration of heavy metals in each leachate was analysed with ICP-OES. Duplicate samples from FA and each granule batch were analysed, and the average was calculated. The metals extracted in step F1, such as chlorides and sulphides (van Herck and Vandecasteele, 2001), are water-soluble, as are free ions and ions complexed with soluble organic matter (Filgueiras et al., 2002). This is the most readily available fraction and, thus, the greatest environmental concern (Filgueiras et al., 2002). The second step (F2, acid-soluble) indicates the amount of metals that are weakly adsorbed onto the particles via relatively weak electrostatic interactions, metals that can be released by ion-exchange processes, as well as acid-soluble metals, such as carbonates. Moreover, Fe and Mn oxides and hydroxides can be extracted in this step (van Herck and Vandecasteele, 2001). In the third step (F3, easily- reduced), hydrous oxides of manganese and iron are extracted together. The iron and manganese oxides act as cement, or are present as grains between particles or as coating on particles. The metals are bound strongly to these oxides but are thermodynamically unstable in anoxic conditions (Filgueiras et al., 2002). This means that metals in this fraction could be released under intensifying, reducing or oxidizing conditions in the environment. In the fourth step (F4, oxidisable), the metals bound to organic material or present as oxidisable minerals (e.g., sulphides) are

Table 1 Sequential leaching procedure. Step

Fraction

Reagents/1 g of sample

Mixing time and temperature

F1 F2 F3 F4

Water-soluble Acid-soluble Easily-reduced Oxidisable

40 ml 40 ml 40 ml 10 ml 10 ml 50 ml

16 h at 22 °C 16 h at 22 °C 16 h at 22 °C 1 h at 25 °C 1 h at 85 °C 16 h at 22 °C

H2O (pH = 4.0) 0.1 M CH3COOH 0.1 M HONH3Cl 30% H2O2 (evaporation) 30% H2O2 (evaporation) 1 M CH3COONH4

3

extracted. These metals can be leached under oxidizing conditions (Filgueiras et al., 2002). The residual metals are present in the minerals’ crystal lattice and inside the crystallized oxides (Filgueiras et al., 2002). This fraction is non-mobile and least harmful to the environment. 3. Results and discussion 3.1. Raw materials Table 2 shows the XRF-analysis and particle size-measurement results of the precursors. The combined amount of SiO2 and Al2O3 in the FA was approximately 45%. In addition, the FA contained a high amount of CaO. The dry matter content and LOI at 525 °C and at 950 °C of the FA was 100%, 0.0% and 2.1%, respectively. Blast-furnace slag (S) mainly contained CaO, SiO2, Al2O3 and MgO. The combined total amount of SiO2 and Al2O3 was 96% in metakaolin (M). FA had a significantly larger particle size than slag and metakaolin. The specific surface area (m2/g) of FA was 1.521. A FE-SEM- image (Fig. 1) shows that the FA contained asymmetrical particles with diameters ranging from a few micrometres to a few tens of micrometres, as verified by particle size measurement. A few spherical are particles present, but most particles have rough surfaces, which increases their surface area. 3.2. Granulation Table 3 lists the proportions of the samples’ precursors, their consumption of Na-Sil and their L/S ratios. The required amount of Na-Sil was between 113 g (FA60S40) and 160 g (FA60M40) for all samples. The small and platy-shaped of metakaolin (Provis et al., 2010) require more liquid than FA and slag particles, and this explains the higher consumption of Na-Sil by FA80M20 and FA60M40. The target granule size of all samples was acquired between 9 and 13 min, so the granulation process can be considered fast. Fig. 2 shows the prepared granules sealed in a plastic bag. 3.3. Leaching results 3.3.1. Standard EN 12457-3 Table 4 lists the pseudo-total concentration of elements, leachable elements in FA and legal solubility limits for heavy metals for the use of fly and bottom ash in civil engineering in Finland (FINLEXÒ, 2006). In this study, FA was found to have high contents of several heavy metals, such as Ba, Cu, Pb and Zn; this could potentially be harmful to the environment. The use of FA is controlled by a national legal limit (FINLEXÒ, 2006), and the EN 12457-3 leaching test is used for quality-control. In the tested FA, the leached amounts of Cr, Mo, Pb and Zn exceeded the legal limit. The leachable SO4 was almost 5 the legal leaching limit. The leachable chloride concentration was not determined, but according to Table 2, the total concentration of Cl was 1.2 wt.% in the precursors analysed by XRF. This indicates that the concentration of chloride was low, but it could exceed the legal leaching limit (800 mg/kg) nonetheless. Possibly, the leached amounts of As, Cd, Sb and Se exceeded their respective legal limits as well, but their concentrations were below the detection limit. Fig. 3 shows the theoretical leaching of heavy metals versus determined leaching. The dashed-outline column presents the leaching if the leachability of the metals had not been affected by alkali activation. This was calculated by identifying the leachable metals in the precursors by EN 12457-3, and multiplying the leachable content by the precursor weight percentages in the alkali-activated granules. The impurities in the sodium-silicate solution were determined by ICP-OES. The black column shows

Please cite this article in press as: Yliniemi, J., et al. Alkali activation of recovered fuel–biofuel fly ash from fluidised-bed combustion: Stabilisation/solidification of heavy metals. Waste Management (2015), http://dx.doi.org/10.1016/j.wasman.2015.05.019

4

J. Yliniemi et al. / Waste Management xxx (2015) xxx–xxx

Table 2 Chemical compositions and particle sizes of fly ash, slag and metakaolin.

FA S M

CaO

SiO2

Al2O3

Fe2O3

Na2O

K2O

MgO

P2 O5

TiO2

SO3

Cl

solidification of heavy metals.

Recovered fuel-biofuel fly ash from a fluidized bed boiler was alkali-activated and granulated with a sodium-silicate solution in order to immobilise ...
1MB Sizes 0 Downloads 15 Views