Elsevier

Biomass and Bioenergy

Volume 120, January 2019, Pages 144-155
Biomass and Bioenergy

Research paper
Mass and energy balances of an autothermal pilot carbonization unit

https://doi.org/10.1016/j.biombioe.2018.11.009Get rights and content

Highlights

  • Charcoal yield in excess of 22 wt% were obtained in autothermal carbonization.

  • Yield of condensable organics was 4.91 wt% of the dry biomass.

  • Attained energy yield in autothermal carbonization exceeding 41%.

  • Lower stack emission than traditional carbonization process, high T heat as co-product.

  • Charcoal has FC = 80.8 wt%, VM 17 wt%, Ash 2.2 wt%, BET 135.4 m2 g−1, PAH <2.5 mg kg−1.

Abstract

The CarbOn pilot plant is a continuous biomass carbonization system, rated for a capacity of up to 50 kg h−1 and based on open top, downdraft technology, operating in oxidative pyrolysis in the temperature range of 500–650 °C and equivalence ratio (ER) between 0.1 and 0.2. In the reported validation tests, carried out on small size chestnut woodchips, charcoal mass yield in excess of 22.4 ± 0.7 wt% (dry base) has been achieved, with a fixed carbon content higher than 85 wt% (dry base). The fixed carbon yield (FCy) was 18.2 ± 2.2 wt% (dry base), the char carbon yield (CCy) 38.3 ± 1.6 wt% (dry base) and the net energy conversion efficiency to char (ε) equal to 41.2 ± 2.2% (wet base). Volume concentration of permanent gases in the pyrolysis vapors and condensable species were also measured before incineration and critically compared against literature data. The organic condensate from oxidative pyrolysis was obtained as 4.9 wt% of the dry biomass, and around 58 wt% of its constituents have been identified; in order of decreasing abundance, the organic fraction of condensate was composed of organic acids, aromatics, furans, anhydrous sugars, phenols, methanol, PAHs, acetaldehyde, ketones. Measured and calculated performance data shows that the pilot unit can produce high quality charcoal, meeting and exceeding the product specifications set by standard EN 1860-2 for BBQ lump charcoal as well as those set forth by international voluntary standards on biochar quality for soil application.

Introduction

Biomass carbonization is the process of converting solid, lignocellulosic biomass into charcoal; it is currently operated at several scales, encompassing manual, rudimental, batch methods and industrial continuous systems, with reported throughput up to several tens of thousands Mg (ton) per year [1]. Reported yields range from 8 to 12% for traditional kilns, 12–17% for brick kilns, 14–20% for standard industrial facilities and 25–33% for advanced industrial processes [2]. A major concern with charcoal production facilities is related to volatiles released in the atmosphere: while larger plant operates either a post-combustion of the volatiles or recovers the organic compounds that are being produced, smaller and more rudimentary systems directly vent the off-gases, generating plenty of harmful emissions, e.g. product of incomplete combustion, and greenhouse gases [3]. Along with improper post-harvest land and forest management [4], these emissions represents the main environmental impact of charcoal manufacturing in traditional systems.

Charcoal making is an established process, with a world production that exceeded in 2018 the 50·106 Mg according to FAO estimates [5]. Not only the produced char can be used to sequester carbon in the soil [6], or effectively impact on soil properties [7], but the process of charcoal making can also represent an attractive alternative to green power generation for small forestry companies, also in developed countries, as long as it can be profitable.

The bioenergy sector, and biomass gasification in particular, has been largely relying on public incentives and support schemes to sustain the tight economics of those systems. IRENA [8] recently reported installed costs between 2000 and 7000 USD(2016) kW−1 for fixed bed gasifiers deployed in Europe and North America, while prior estimated equipment cost for gasifier coupled to internal combustion engine have been reported in the range of 4321–5074 USD(2010) kW−1 [9] and 4150 USD(2010) kW−1 [10]. Considering that the expected cost reduction of gasification technology for power generation has not been fully achieved yet, and given the exceedingly high capital costs, the potential profitability of this kind of investment for a small size forestry company is questionable in industrialized countries, but to some extent also in rural and isolated contexts [11]. Focusing the attention on conversion products, rather than power, can make the manufacturing of charcoal from biomass attractive to small-medium sized forestry companies, while at the same time providing a sustainable heat generation solution that does not entirely relie on incentives and supporting policies, implementing a novel “combined heat and product” paradigm.

In the present EU policy context, where coherent, sustained and secure supporting schemes for power generation from biomass are missing or not in place yet, there is an opportunity for policy makers to promote the transition from energy to bio-products in the conversion of biomass, and open new business prospects, independent from subsidies and feed-in tariffs. With this aim, since 2013 RE-CORD has been developing an innovative biomass carbonization reactor for the small-to-medium sized forestry enterprises and is currently operating the first pilot plant, rated for a capacity of up to 50 kg h−1. The proposed pilot plant leverages the simplicity and effectiveness of autothermal operations and open top, fixed bed downdraft design, to bring to the small scale the performance of larger installations.

Smouldering is a non-flaming combustion mode, characterized by thermal degradation and charring of the virgin material, evolution of smoke and emission of visible glow [12]. The phenomenon is of interest in a broad range of fields, including wildfire prevention, peat fires, underground coal gasification, flight and confined-space safety, as illustrated by Rein in his introductory paper [13]. In the field of thermochemical biomass conversion, smoulderding is also referred to as oxidative or autothermal pyrolysis, that is heat for the process is internally provided by combusting part of the feedstock and pyrolysis gas inside the reactor.

In the context of pyrolysis processes for the production of liquid products, the autothermal regime has been proposed to facilitate the energy integration and reduce operating costs, improving process feasibility [14]. Beyond being credited of significant simplification in the process layout, autothermal operation has been proposed also for fast pyrolysis technologies as a mean to increase the process throughput, by increasing the capacity to process larger mass flow-rate of feedstock in the same reactor volume, or as mean to decrease capital expenditure (i.e smaller reactor) for the same plant capacity [15]. A significant number of studies at the reactor level have been carried out for the fast pyrolysis of biomass in autothermal conditions, including the work from Butt [[16], [17], [18]], Boukis et al. [19,20], Amutio et al. [14,21], Mesa-Pérez et al. [22] Kim et al. [23], 2015 Li et al. [24,25], which have been reviewed by Mazur [15].

Oxidative pyrolysis process has been modeled [14] and described [26,27] as the sum of pure thermal degradation and heterogeneous oxidation occurring at the same time, followed by char combustion reactions [28]. Several analytical and kinetic schemes have been proposed to model and explain the complex underlying chemistry and phenomena [29,30], and test of oxidative pyrolysis at the laboratory scale are available [31]. However, the body of prior experimental works reporting the technical application of oxidative/autothermal pyrolysis to maximize the biomass-to-charcoal conversion at the pilot scale in (potentially) continuous fixed bed reactors is, to the best of the Authors’ knowledge, very limited.

Researchers at CIRAD Montpellier [[32], [33], [34]] investigated the oxidative pyrolysis of biomass in a continuous downdraft reactor with a diameter of 0.2 m and height of 1.6 m, capable of working either in auto- or allothermal conditions (using a LPG burner), and even in batch mode. The maximum reported capacity of the unit was 6 kg h−1 with pelletized biomass, maintaining a bed height of 450 mm, the suitable size of the feed was between 3 and 13 mm, and charcoal was manually discharged from the bottom.

Milhé et al. [32] performed a comparative investigation on the CIRAD test setup under autothermal and allothermal conditions. Experiments were carried out with biomass particles in the range of 5–10 mm, attaining approx. 675 °C of maximum process temperature, 6 kg h−1 of biomass feed rate, residence time of around 60 min and ER (equivalence ratio, i.e. the ratio of actual on stoichiometric air flowrate) of 0.1. The Author observed that cracking reactions were enhanced during autothermal pyrolysis, as expressed by the slight increase in the production of phenolic and aromatic compounds at the expense of the guaiacolic compounds. According to the Author, in autothermal pyrolysis the fixed bed smouldered with a steep temperature gradient around an oxidation zone within the bed. The downward speed of the particles had to match the propagation rate of the combustion front in order to remain in steady state, so the air to biomass ratio cannot change much.

Daouk et al. [33] investigated the impact of air flux and bed bulk density on the behavior of the oxidation zone in terms of wood consumption, and yields of char, gas and tars for pine chips and wood pellets in autothermal operations. Experiments with woodchip were carried out with biomass particles in the range of 4–12 mm, attaining approx. 650 °C of maximum process temperature, 3.8 kg h−1 of biomass feed rate, residence time of around 90 min and ER range of 0.1–0.135. In stable continuous fixed bed conditions, oxidative pyrolysis of woodchips produced about 17% char, 10% tars, 25% water, and 55% permanent gases. The Authors attributed the low yields of tar during oxidative pyrolysis (8.5 wt% of the dry biomass in) to the thermal cracking and oxidation in the oxidation zone, where the temperature is higher (700 °C for an air flux of 0.022 kgm−2 s−1) than in the one during inert pyrolysis (575 °C), and tar cracking in the char bed zone, where heterogeneous and homogeneous reactions on the surface of the char take place. Air flux was shown to mainly affect char yield: a decrease of about 20% was measured when the air flux was doubled.

Pham et al. [34] investigated the behavior of the oxidation zone during batch oxidative pyrolysis of three distinct pelletized biomasses, namely pine, wheat straw and miscanthus. The Authors provided detailed measurement and assessment of the impact of feed properties on the propagation velocity, geometry, thickness and front temperature of the oxidation zone, and found that the front temperature is not dependent on the biomass species or, for the same species, by the moisture content, while the propagation rate and thickness of oxidation zone are strongly influenced by bed density, ash and moisture content.

The major differences between the CIRAD's reactor, reported in Refs. [[32], [33], [34]], and the reactor used for the present study are, notably, reactor layout, nominal capacity of the plant, open top design and the absence of preheating of the oxidant. For the first point, while in the CIRAD's reactor the pyrolysis gas flows through the whole biomass/charcoal bed and both solid and gas path are completely co-current, in our study the pyrolysis gas are withdrawn radially right after the oxidation zone, 0.18 m above the bottom of the cylindrical body of the reactor and 0.78 m above the end of the charcoal bed. The second difference is the capacity level as reported in the test; while Milhe et al. attained a biomass flowrate of up to 5.6 kg h−1 dry base (with wood pellet) and Daouk et al.reported 3.34 kg h−1 dry base (with wood chips), our plant was operated at a biomass flowrate of 36.5 kg h−1 dry base, that is more than 6 times the capacity of the former and mora then 10 times the latter. With respect to the third point, notably the biomass loading, the CIRAD system features an interlock system with two valves that isolate the biomass from the oxidant feeding, while our system is operated “open top” and slightly below the atmospheric pressure, providing an increased degree of safety toward the consequences of explosion (the system is intrinsically safe as it cannot go overpressure in case of backflame or fault). Finally, the CIRAD pilot plant features a preheating system that bring the flux of the oxidant at 140–150 °C before entering the reactor, while our plant is operated without preheating.

Garcia-Nunez et al. [1] has recently reviewed the proposed and commercial plants for large scale carbonization of biomass. However, when considering the set of systemic and financial constraints to which small scale forestry companies are subject in EU and comparable countries, i.e. strict regulations for stack emission, limited requirement of workforce for operation and on workforce specialization, and tight financial resources, the more established technologies (e.g. rotary kilns, auger reactors, Herreshoff furnaces), are either not affordable or exceedingly large for the scale of the local market. On the contrary, for a scenario of distributed small scale installations, fixed bed autothermal plants are simple enough to allow significant cost reduction against competing options.

The present work aims at extending the present knowledge on oxidative pyrolysis for charcoal production by testing and validating the operation of a pilot fixed bed carbonization unit working in autothermal regime, to a scale that, to the best of Authors knowledge, haven't been reported in literature yet, providing a comprehensive characterization of products and operating conditions. The data herein presented can potentially be used for the design and deployment of a larger scale unit and for process-development purpose.

Section snippets

Feedstock

Feedstock for the test was primarily from chestnut (Fratelli Travaglini, Italy); logs and branches were chipped in a conventional rotating-drum chipper. Woodchips were stored in a container, and sampled according to CEN/TS 14778–1:2005. Prior to the test, the material was left aground indoor to dry out, then packed in bags and weighed.

The feedstock had a median particle size (d50 - EN 15149) of 7.1 mm and a size-class P-31.5 (EN 14961). Moisture, fixed carbon, volatile matter and ash content of

Thermal profile

Bed ignition started at 9:45 with burner and ejector set to withdraw around 30 kg h−1, in order to allow warm up of the gas line; the test lasted 8 h from start-up to shut-down. Screw conveyor was activated once T2 attained 500 °C at around 10:30, and gas flow rate was then set at around 65 kg h−1.

Fig. 3 reports the measured temperature profile against time along reactor (top) and gas line (bottom). While the bottom of the reactor achieved a substantial stability already at around 12:00, as

Conclusion

A continuous biomass carbonization plant, operating in autothermal mode, have been designed, built and validated in operation; the plant is rated for a capacity of around 50 kg h−1, and was operated at 42 kg h−1 (wet base) at equivalence ration of 0.154 in the test herein reported.

At stable operating condition, the average charcoal properties were: fixed carbon content of 80.8 wt% (dry base), ash content of 2.2 wt% (dry base), volatile matters of 17 wt% (dry base), BET surface area of 135.4 m2 g

Acknowledgements

Authors wish to acknowledge the RE-CORD personnel who contributed with their help in carrying out the experimental campaign (Edoardo Miliotti, David Casini, Marco Buffi, Matteo Prussi) and laboratory analyses (Giulia Lotti, Lorenzo Bettucci, Silvia Pennazzi). MEng students Andrea Salimbeni and Francesco Matteini, who provided assistance on data analysis and elaboration during their final thesis, are also gratefully acknowledged.

References (46)

Cited by (0)

View full text